I

H I S T O R Y OF C I V I L I Z A T I O N S OF

I The development of sedentary and nomadic civilizations: 700 B .c. m A. D. z 5

! A
t

i
I

I ,
I
I

I'
1 lX v
 History of civiliz,ations of Central Asia

Volume I
T h e dawn of civilization:
earliest tirnes t o 700 H.C.

Volume I /
T h e development of sedentary and nomadic civilizations:
700 KC. t o A.U. 250

Volume I I I
T h e crossroads of civilizations:
A.D. 250 t o 750

Volume IV
T h e age of achievement:
A.D. 750 t o the end of the fifteenth century:
Part I:
The historical, social and economic setting
Part 11:
The achievements

Volume V
Development in contrast:
from the sixteenth t o the eighteenth century

Volume V I
Towards contemporary civilization:
from the beginning of the nineteenth century
to the present time
 H i s t o r y of
civilizations
of C e n t r a l
Asia
Volume I1
The development of sedentary and
nomadic civilizations:

Editor: Jn'nos Hamatts

Co-editors: B. N. Puri and G. F. Etemadi

U N E S C O P u b l i s h i n g
The authors are responsible for the choicc and the presentation
of the facts contained in this book and for the opinions
expressed therein, which are not necessarily those of U N E S C O
and d o not commit the Organization.

The designations employed and the presentation of material

throughout this publication d o not imply the expression
of any opinion whatsoever on the part of U N E S C O concerning
thc legal status of any country, territory, city or
area or of its authorities, or concerning the delimitation
of its frontiers or boundaries.

First published in 1994 by the United Nations Educational,

Scientific and Cultural Organization
7 place de Fontenoy, 75352 Paris 07 S P
Composed by U N E S C O Publishing, Paris
Printed by DarantiPre, Quetign)r

Second impression 1996

ISBN 92-3-10284-4

0 Unesco 1994
Printed in France
 Federico Mayor
Director-General of U N E S C O

NE of the purposes of UNESCO, as proclaimed in its Constitution, is

'to develop and to increase the means of communication between . . .
peoples and to employ these means for the purposes of mutual
understanding and a truer and more perfect knowledge of each other's lives'.
The History of the Scientific a n d Cultural Development of Mankind, published
in 1968, was a major early response on the part of U N E S C O to the task of ena-
bling the peoples of the world to have a keener sense of their collective destiny
by highlighting their individual contributions to the history of humanity. This
universal history - itself now undergoing a fundamental revision - has been fol-
lowed by a number of regional projects, including the General History of Africa
and the planned volumes on Latin America, the Caribbean and on aspects of
Islamic culture. The History of Civilizations of Central Asia is an integral part
of this wider enterprise.
It is appropriate that the second of UNESCO's regional histories should
be concerned with Central Asia. For, like Africa, Central Asia is a region whose
cultural heritage has tended to be excluded from the main focus of historical
attention. Yet from time immemorial the area has served as the generator of
population movements within the Eurasian land-mass. The history of the
ancient and medieval worlds, in particular, was shaped to an important extent
by the succession of peoples that arose out of the steppe, desert, oases and
mountain ranges of this vast area extending from the Caspian Sea to the high
plateaux of Mongolia. From the Cimmerians mentioned in Homer's Odyssey,
the Scythians described by Herodotus, the Hsiung-nu whose incursions led the
emperors of China to build the Great Wall, the sixth-century Turks who
extended their empire to the boundaries of Byzantiun~,the Khitans who gave
their name to ancient Cathay, through to the Mongols who erupted into world
history in the thirteenth century under Genghis Khan, the nomadic horsemen
of Central Asia helped to define the limits and test the mettle of the great civil-
izations of Europe and Asia.
 Preface

N o r is it sufficient to identify the peoples of Central Asia simply with

nomadic cultures. This is to ignore the complex symbiosis within Central Asia
itself between nomadism and settlement, between ~astoralistsand agricultura-
lists. It is to overlook above all the burgeoning of the great cities of Central Asia
such as Samarkand, Bukhara and Khiva, which established themselves in the late
Middle Ages as outstanding centres of intellectual inquiry and artistic creation.
The seminal writings of the philosopher-scientist Avicenna (a native of Buk-
hara) and the timeless masterpieces of Timurid architecture epitomize the
flowering of medieval culture in the steppes and deserts of Central Asia.
The civilizations of Central Asia did not, of course, develop in a vacuum.
The impact of Islam was pervasive and fundamental. The great civilizations on
the periphery of the Eurasian continent likewise exerted an important influence
on these lands. For some 1,500 years this arid inland sea - far removed from the
earth's true oceans - was crucial as the route along which merchandise (notably
silk) and ideas flowed between China, India, Iran and Europe. The influence of
Iran - although the core of its civilization lies in South-West Asia - was particu-
larly strong, to the extent that it is sometimes difficult to establish a clear boun-
dary between the civilization of the Iranian motherland and that of the outlying
lands of Central Asia.
T o the rich variety of peoples of Central Asia was thus added a multi-
plicity of external influences. For century after century, the region experienced
the influx of foreign art and ideas, colliding and merging with the indigenous
patterns of Central Asia. Migrations and the recurrent shock of military inva-
sion, mingling and displacing peoples and cultures, combined to maintain the
vast region in flux.
The systole and diastole of population movements down the ages add to
the difficulty of delimiting a region whose topology alone does not prescribe
clear boundaries. Thus, when, at the nineteenth session of its General Confe-
rence, U N E S C O decided to embark on a History of Civilizations of Central
Asia the first problem to be resolved was to define the scope of the region
concerned. Subsequently, at a U N E S C O meeting held in 1978, it was agreed
that the study o n Central Asia should deal with the civilizations of Afghanistan,
north-eastern Iran, Pakistan, northern India, western China, Mongolia and the
former Soviet Central Asian republics. The appellation 'Central Asia', as
employed in this History, refers to this area, which corresponds to a clearly dis-
cernible cultural and historical reality.
UNESCO's call to specialists, and particularly to scholars native to the
region, to participate in the undertaking met with a wide and generous res-
ponse. The project was deemed by academics to be an excellent opportunity to
draw back the curtain that had veiled Central Asia for so long. However, none
were in any doubt as to the huge dimensions of the task.
An ad hoc International Scientific Committee was formed in 1980 to plan
and prepare the work, which it was agreed should cover, in six volumes, the
 Preface

history of Central Asia from earliest times to the present day. The Committee's
initial task was t o decide where prc-eminence should be given in the very wide
canvas before it. In due course, a proper balance was struck and teams of cditors
and authors were selected.
The preparation of the History of Civilizations of Central Asia is now
well advanced. The best resources of research and archaeology have bccn used
to make the work as thorough as possible, and countless annals consulted in
major centres throughout the region. It is my sincere wish that this, the second
volume, and thosc that follow will bring instruction and pleasure to readers all
over the world.
It remains for me to thank the President, Rapporteur and members of the
International Scientific Committee, and the editors, authors and teams of spe-
cialists who have collaborated t o shed new light on Central Asia with this
detailed account of its vital and stirring past. I am sure it will prove a notable
contribution to the study and mutual appreciation of the cultures that are the
common heritage of mankind.
Description of the project
Mohammad S. Asimov, President, International Scientific Committee . 11
Members of the International Scientific Committee . 15
List of contributors . . 17
Introduction J. Harmatta . . 19
1 Ancient Iranian nomads in western Central Asia
A. Abetekov and H . Yusupov . . 23
2 Media and Achaemenid Iran M. A . Dandamayev . 35
3 Alexander and his successors in Central Asia
A. H . Dani and P. Bernard . . 67
4 The Greek kingdoms of Central Asia P. Bernard . 99
5 Parthia G. A. Koshelenko and V. N. Pilipko . . 131
6 Nomads in eastern Central Asia N. Ishjamts . . 151
7 The Yiieh-chih and their migrations
K. Enoki, G. A. Koshelenko and 2. Haidary . 171
8 The Sakas and Indo-Parthians B. N. Puri . . 191
9 The culture of the Xinjiang region Ma Yong and Wang Binghua . 209
10 The Western Regions under the Hsiung-nu and the Han
Ma Yong and Sun Yutang . . 227
11 The Kushans B. N. P14ri . . 247
12 Economy and social system in Central Asia in the Kushan age
A . R. Mukhamedjanov . . 265
13 Cities and urban life in the Kushan kingdom
B. A. Litvinsky . . 291
14 Religions in the Kushan Empire
J. Harmatta, with the contributions of B. N. Puri, L. Lelekov,
S. Humayun and D. C. Sircar .
15 Kushan art
G. A. Pugachenkova, S. R. Dar, R. C. Sharma and M. A. Joyenda,
in collaboration with H . Siddiqi
16 Languages and scripts in Graeco-Bactria and the Saka kingdoms
J. Harmatta
17 Languages and literature in the Kushan Empire J. Harmatta .
18 States in north-western Central Asia N. N. Negmatov .
19 The nomads of northern Central Asia after the invasion
of Alexander
Y. A. Zadneprovskiy .
20 The rise of Sasanian Iran
B. A. Litvinsky, with the contributions of M. Hussain Shah and
R. Shabani Samghabadi
Conclusion J. Harmatta .
Maps .
Bibliography and references .

Index . . 555
 D E S C R I P T I OONF THE PROJECT

Mohammad S. Asimov

T HE General Conference of UNESCO, at its nineteenth session (Nairobi,

October, November 1976), adopted the resolution which authorized the
Director-General to undertake, among other activities aimed at promot-
ing appreciation and respect for cultural identity, a new project on the preparation
of a History of Civilizations of Central Asia. This project was a natural conse-
quence of a pilot project o n the study of Central Asia which was approved during
the fourteenth session of the U N E S C O General Conference in November 1966.
The purpose of this pilot project, as it was formulated in the U N E S C O
programme, was to make better known the civilizations of the peoples living in
the regions of Central Asia through studies of their archaeology, history, lan-
guages and literature. At its initial stage, the participating Member States
included Afghanistan, India, Iran, Pakistan and the former Soviet Union. Later,
Mongolia and China joined the U N E S C O Central Asian project, thus enlarging
the area to cover the cultures of Mongolia and the western regions of China.
In this work, Central Asia should be understood as a cultural entity deve-
loped in the course of the long history of civilizations of peoples of the region
and the above delimitation should not be taken as rigid boundaries either now
o r in the future.
In the absence of any existing survey of such large scope which could
have served as a model, U N E S C O has had to proceed by stages in this difficult
task of presenting an integrated narrative of complex historical events from ear-
liest times to the present day.
The first stage was designed to obtain better knowledge of the civiliza-
tions of Central Asia by encouraging archaeological and historical research and
the study of literature and the history of science. A new project was therefore
launched to promote studies in five major domains: the archaeology and the his-
tory of the Kushan Empire, the history of the arts of Central Asia, the contribu-
tion of the peoples of Central Asia to the development of science, the history of
ideas and philosophy, and the literatures of Central Asia.
 Mohammad S. Asimov

A n International Association for the Study of Cultures of Central Asia

(IASCCA), a non-governmental scholarly organization, was founded on the
initiative of the Tajik scholar B. Gafurov in 1973, assembling scholars of the
area for the co-ordination of interdisciplinary studies of their own cultures and
the promotion of regional and international co-operation.
Created under the auspices of U N E S C O , the new Association became,
from the very beginning of its activity, the principal consultative body of
U N E S C O in the implementation of its programme o n the study of Central
Asian cultures and the preparation of a History of Civilizations of Central Asia.
The second stage concentrated on the modern aspects of Central Asian
civilizations and the eastward extension of the geographical boundaries of
research in the new programme. A series of international scholarly conferences
and symposia were organized in the countries of the area to promote studies on
Central Asian cultures.
T w o meetings of experts, held in 1978 and 1979 at U N E S C O Headquar-
ters, concluded that the project launched in 1967 for the study of cultures of
Central Asia had led t o considerable progress in research and contributed to
strengthening existing institutions in the countries of the region. The experts
consequently advised the Secretariat on the methodology and the preparation of
the History. O n the basis of its recommendations it was decided that this publi-
cation should consist of six volumes covering chronologically the whole history
of Central Asian civilizations ranging from their very inception up t o the pres-
ent. Furthermore, the experts recommended that the experience acquired by
U N E S C O during the preparation of the History of Scienttfic and Cultural De-
velopment of Mankind and of the General History of Africa should also be
taken into account by those responsible for the drafting of the History. As to its
presentation, they supported the opinion expressed by the U N E S C O Secreta-
riat that the publication, while being a scholarly work, should be accessible to a
general readership.
Since history constitutes an uninterrupted sequence of events, it was
decided not to give undue emphasis to any specific date. Events preceding or
subsequent to those indicated here are dealt with in each volume whenever their
inclusion is justified by the requirements of scholarship.
The third and final stage consisted of setting u p in August 1980 an Inter-
national Scientific Committee of nineteen members, who sit in a personal capac-
ity, to take reponsibility for the preparation of the History. The Committee
thus created includes two scholars froin each of the seven Central Asian coun-
tries - Afghanistan, China, India, Islamic Republic of Iran, Pakistan, Mongolia
and the former USSR - and five experts from other countries - Hungary, Japan,
Turkey, the United Kingdom and the United States of America.
The Committee's first session was held at U N E S C O Headquarters in
December 1980. Real work on the preparation of the publication of the History
of Civilizations of Central Asia started, in fact, in 1981. It was decided that
 Description of the project

scholars selected by virtue of their qualifications and achicvemcnts relating to

Central Asian history and culture should ensure the objective presentation, and
also the high scientific and intellectual standard, of this History.
Members of the International Scientific Committee decided chat the new
project should correspond to the noble aims and principles of U N E S C O and
thereby should contribute to the promotion of mutual understanding and peace
between nations. The Committee followed the recommendation of the experts
delineating for the purpose of this work the geographical area of Central Asia to
reflect the common historical and cultural experience.
The first session of the International Committee decided most of the prin-
cipal matters concerning the implementation of this complex project, beginning
with the drafting of plans and defining the objectives and methods of work of
the Committee itself.
The Bureau of the International Scientific Committee consists of a presi-
dent, four vice-presidents and a rapporteur. The Bureau's task is t o supervise
the execution of the project between the sessions of the International Scientific
Committee. The reading committee, consisting of four members, was created in
1986 to revise and finalize the manuscripts after editing Volumes I and 11.
Another reading committee was constituted in 1989 for Volumes I11 and IV.
The authors are scholars from the present twelve countries of Central
Asia and experts from other regions. Thus, this work is the result of the regional
and of the international collaboration of scholars within the framework of the
programme of the United Nations Educational, Scientific and Cultural Organ-
ization (UNESCO).
It is our sincere hope that the ~ublicationof the second volume of the
History of Civilizations of Central Asia will be a further step towards the pro-
motion of the cultural identity of the peoples of Central Asia, strengthening
their common cultural heritage and, consequently, will foster a better under-
standing among the peoples of the world.
 M E M B E R SO F T H E I N T E R N A T I O N A L
S C I E N T I F I CC O M M I T T E E
(in alphabetical order)

D r F. R. ALLCHIN(United Kingdom) Professor LIU CUNKUAN(China)

Professor M. S. ASIMOV(Tajikistan) D r L. I. MIROSHNIKOV
(Russian
President Federation)
Editor of Volume IV (Parts I and 11) Professor S. NATSAGDOKJ
(Mongolia)
D r N. A. BALOCH(Pakistan) Professor B. N. PURI(India)
Co-editor of Volume I1
Professor M. BASTANIPAHIZI(Islamic
Republic of Iran) Professor M. H . 2. SAFI(Afghanistan)

Professor S. BIKA(Mongolia) Professor A. SAYILI(Turkey)

D r R. SHABANISAMGHABADI
Professor A. H . DANI (Pakistan)
(Islamic Republic of Iran)
Editor of Volume I
Co-editor of Volume I11
Professor K. ENOKI(Japan) Professor D. SINOR(United States
Professor G. F. ETEMADI of America)
(Afghanistan) Professor B. K. THAPAK(India)
Co-editor of Volume I1
Professor ZHANGGUANG-DA
Professor J. HARMATTA
(Hungary) (China)
Editor of Volume I1 Co-editor of Volume I11

New members:
Professor H.-P. FRANCFORT (France)
Professor I. TOGAN(Turkey)
A. ABETEKOV Toyo Bunko
Institute of History of the Kyrgyz The Oriental Library
Academy of Sciences Honkomagome, 2-Chome, 28-2 1
Bishkek, Kyrgyzstan. Tokyo, Japan.

P. BERNARD 2. HAIDARY
~ c o l ePratique des Hautes Ltudes Director-General
IVc Section International Centre for Kushan Studies
Sciences Historiques et Philologiques Ministry of Information and Culture
Sorbonne Kabul, Afghanistan.
45 rue des ~ c o l e s
75005 Paris, France. J. HARMATTA
Vice-president
M. A. DANDAMAYEV Section of Linguistics and Literary
Oriental Institute Sciences
Dvortsovaya naberejnaya, 18 Hungarian Academy of Sciences
St Petenburg, Russian Federation. 1052 Budapest V, Hungary.

A. H. DANI S. HUMAYUN
Director Faculty of Literature
Centre for the Study of Civilizations Kabul University
of Central Asia Kabul, Afghanistan.
Quaid-i-Azam University
M. HUSSAIN
SHAH
Islamabad, Pakistan.
Head of the Department of History
S. R. DAR Faculty of Social Sciences
Director Kabul University
Lahore Museum Kabul, Afghanistan.
Shara-e-Quaid-e-Azam
Lahore, Pakistan. N. ISHJAMTS
Institute of Oriental Studies
K. ENOKI Academy of Sciences
Director Ulan Bator, Mongolia.
M. A. JOYENIIA Mahanagar
Institute of Archaeology Lucknow 226 006, India.
International Centre for Kushan Studies
R. SHABANI
SAMGHABADI
Kabul University National University of Iran
Kabul, Afghanistan. Shahid Beheshti University
G. A. KOSHELENKO Tehran, Islamic Republic of Iran.
Institute of Archaeology
Russian Academy of Sciences Director
Moscow, Russian Federation. Indian Museum
27 Jawaharlal N e h r u Road
L. LELEKOV
Calcutta 700 016, India.
T h e All-Russian Institute of Restoration
Krestyanskaya pl. 10 H. SIDDIQI
Moscow, Russian Federation. Institute of Archaeology
International Centre for Kushan Studies
B. A. LITVINSKY
Kabul University
Chief
Kabul, Afghanistan.
Section of Oriental Studies
Russian Academy of Sciences D. C. SIRCAR
Moscow, Russian Federation. 645 '0' Block
N e w Alipore
MA YONG Calcutta 700 053, India.
Institute of History
Secretary of the Chinese Society for SUN YUTANG
Central Asia Institute of History
5 Jianguomen Nei Da Jie Chinese Academy of Social Sciences
Beijing, China. 102 Seventh Building
Yung A n N a n Li
A. R. MUKHAMEDJANOV
Jian Kuo Men Wai
Director of the Institute of Archaeology
Beijing, China.
Uzbek Academy of Sciences
Samarkand, Uzbekistan.
Xinjiang Institute of Archaeology and
N. N. NEGMATOV
Cultural Relics
Institute of History
16 South Beijing Road
Tajik Academy of Sciences
Urumqi, Xinjiang 83001 1, China.
Dushanbe, Tajikistan.

V. N. PILIPKO
H. Yusurov
Institute of History of the T u r k ~ n e n
Institute of History of thc Turkmen
Academy of Sciences
Academy of Sciences
Ashgabat, Turkmenistan.
Ashgabat, Turkmenistan.

G. A. PUGACHENKOVA Y. A. ZAUNEPROVSKIY
Institute of the History
Uzbek Acadeniy of Sciences
of Matcrial Culturc
Tashkent, Uzbekistan.
Russian Acade1ny of Sciences
B. N. Pun1 Dvortsovaya nabcrejnaya, 18
B-58, Sector A St Petersburg, Russian F e d e r ~ t i o n .
v OLUME I1 of the History of Civilizations of Central Asia deals with the
historical period from c. 700 B.C. to c. A.D. 250. This was characterized
by the development of the nomadic and sedentary civilizations, created
by the Iranian tribes who lived partly in the steppe and wooded-steppe zone,
partly in the oasis zone. The ancestors of the Iranians, the Indo-Iranian
tribes, had separated from the Proto-Balts and the Proto-Slavs in the fifth
millennium B.C. and began to infiltrate the territory of Central Asia during
the fourth millennium B.c., but remained in close contact with the Finno-
Ugrian tribes, who borrowed a considerable number of important terms from
them. Indeed it is from such loan words and linguistic contacts that their
migrations can be reconstructed. The first mass migration of the Indo-Ira-
nians during the second millennium B.C. was connected with the rise of ani-
mal husbandry, in particular horse-breeding, and with the invention of the
two- o r four-wheeled vehicle. In their movements towards the Caucasus, the
steppes of southern Siberia and Central Asia, they already used the war cha-
riot. As a result they were able to reach Mesopotamia, the Iranian plateau,
India, the Altai mountains, and even China and Korea. The second wave of
Indo-Iranian migrations was marked by the emergence of the equestrian
nomads that gave rise to the ethnic image of Central Asia and led to the for-
mation of the first syncretistic civilizations.
In the eighth and seventh centuries B.C. there were two different zones of
civilizations in Central Asia - the northern zone of the Iranian nomads and the
southern one of the Iranian husbandmen, who settled on the soil of the oases.
This contrast between nomadic tribes and sedentary population has left deep
traces in the historical tradition of the Iranians, and was strengthened by the rise
of the Achaemenid Empire, which established the Syr Darya as its northern
frontier against the nomadic Sakas. South of the river, inside the Achaemenid
Empire, the transition from nomadism to agriculture accelerated, and different
forms of co-existence between nomads and villages of husbandmen developed.
Beyond the empire the influence of Achaemenid culture in Central Asia spread
towards the east as far as the Sakas of the Altai region.
The invasion of Alexander the Great transformed the scene. The nomads
were driven back, sedentary civilization was strengthened and, due to the nume-
rous Greek settlements, quickly became red om in ant. With the rise of the
Graeco-Bactrian kingdom, the civilizations of Central Asia underwent a funda-
mental change. Greek culture became their constant and essential constituent. It
transformed the character of life as well as the mode and conditions of produc-
tion. Under its influence a new syncretistic civilization came into being, and
Greek-influenced art and architecture developed. The use of the Greek script
and language spread throughout Bactria; urbanization made rapid progress;
houses, public buildings, Greek-type sanctuaries and the pantheon of Greek
divinities appeared everywhere in its cities. Greek influence was not confined to
Bactria: its general importance is well illustrated by the fact that the Mauryan
emperor Aioka had a Greek translation of his edicts ~ r e ~ a r e d .
The Graeco-Bactrian kings made the first attempt t o create a powerful
state in Central Asia, which could control the trade routes leading t o China, the
Indian subcontinent, the Seleucid Empire and the steppes of eastern Europe.
They conquered ancient north-western India, extended their power u p t o the
Phryni and the Seres and led military expeditions against the nomads in the
north and west. It was in this period that the first objects of Chinese origin
reached Central Asia and that the name of the Ts'in dynasty (221-207 B.c.)
emerged in the form of Cin among the Iranians. Coming from Central Asia to
Iran and India, it furnished the base for the European name of China. The
Graeco-Bactrian kingdom played a catalysing role between the cultural goods
of distant peoples, and created a syncretistic culture which became the basis of
the civilizations of Central Asia u p t o the Arab conquest.
About 130 B.C. the Graeco-Bactrian kingdom fell t o the invasions of the
Iranian nomads, the Sacaraucae, Asiani and Tochari. During the second century
B.C. the expansion of the nomadic Hsiung-nu Empire led t o the migrations of
these Iranian nomads, w h o overthrew the Graeco-Bactrian state and interrupted
the commercial relations established by its kings. In the beginning of the first
century B.c., however, the Chinese were able t o open the Silk Route across the
Tarim basin, leading t o the intensive caravan trade that developed between
China and Central Asia and between Central Asia and its other neighbours.
In the first century B.C. the Sakas established a series of kingdoms in east-
ern Iran and north-western India; but in the first century Am., these territories
came under the control of the Indo-Parthians, who ruled an empire that
stretched from Sistan to the Indus and beyond. Meanwhile t o the north of the
Hindu Kush mountains the unification of the five Tocharian tribes under the
rule of the Kushans had already begun; and about A.D. 50 thcir king, Kujula
Kadphises, who had made hiinself master of Bactria, ousted the Indo-Parthians
and annexed their Indian provinces. During the second century A.L).,his succes-
 Introduction

sors, Vima Kadphises and the Great Kushans, ruled the first great empire in
Central Asia - an empire that also controlled north-west India as far as Kau-
iambi and the Indus valley down to the ports of the Arabian Sea. As a result the
Kushans were able to establish strong commercial relations with the eastern
provinces of the Roman Empire by the maritime routes between north-west
India, the Red Sea, and the Persian Gulf; and with the pcoplcs of the Caucasus
and steppes of eastern Europe, by the land routes along the Oxus river and
beyond the Caspian Sea.
Consequently, Central Asia played an important intermediary role of
world dimensions in the transmission of artefacts, culture and ideas. Chinese
silk was sold in Rome. In return Roman gold was exported to north-west India
and Central Asia, and provided the precious metal for the superb coinage of the
Great Kushans. Parallel with the traffic in commodities cultural exchanges
increased. The influence of Graeco-Roman art is t o be seen in Central Asia
where it was a major factor in the formation of Gandhiran art. The spread of
the great religions began with Buddhism, which travelled from India across
Central Asia as far as China. The Kushan pantheon, consisting of pre-Zoroas-
trian, Zoroastrian, Graeco-Roman, Indian Hindu and Buddhist divinities, is a
good illustration of the syncretistic civilization of Central Asia at the time. The
Kushans, who were themselves of nomadic origin, re-established relations be-
tween Central Asia and the northern nomads. They in turn played a more and
more considerable role in the formation of the civilizations that subsequently
came into being o n this territory. Indeed the Kushan synthesis determined the
future of the cultures of Central Asia up to and beyond the adoption of Islam.
Traffic along the Silk Route brought quite exceptional prosperity and the
Great Kushans tried t o exclude their neighbours from this rich transit trade. T o
establish direct contacts with more distant lands, the states bordering on the
Kushan Empire tried to conquer the starting-points o r important sections of the
Silk Route, and were sometimes successful. About A.D. 234, Sasanian Iran con-
quered the western provinces of the Kushan Empire up to Sogdiana, Gandhara
and the Indus delta. Somewhat later, around A.D. 270, the Chionites (descen-
dants of the western Hsiung-nu) made themselves masters of K'ang-chii (a
nomadic state lying to the north of the Syr Darya) and conquered Sogdiana.
Although the Kushan Empire declined, it bequeathed its syncretistic civiliza-
tion, imbued with Greek culture, t o the Chionites, the Guptas, the Hephthalites
and eventually the Tiirk dynasties which replaced the Hephthalites during the
seventh century A.D.

Middle Asia is the territory belonging to the Commonwealth of Independent

States (former Soviet Central Asian republics).
 A. Abetckov and H. Yusupov

HE territory of Central Asia, which consists of vast expanses of steppe-

land, desert and semi-desert with fine seasonal pastures, was destined by
nature for the development of nomadic cattle-breeding. Between the
seventh and third centuries B.C. it was inhabited by a large number of tribes,
called Scythians by the Greeks, and Sakas by the Persians.
The history of the Central Asian nomads is inseparable from that of the
nomadic and semi-nomadic peoples of the Eurasian steppe zone. Their political
and economic life was closely linked, and their material culture had much in
common. It should also be noted that, despite their distinctive qualities, the
nomadic tribes were closely connected with the agricultural population of Cen-
tral Asia. In fact, the history and movements of these nomadic tribes and the
settled population cannot be considered in isolation; each had its impact o n the
other, and this interdependence must be properly understood.

Literary sources on the ancient Iranian nomads

of Central Asia
The term 'Tiira" is the name by which the Central Asian nomadic tribes were
known in one of the earliest parts of the Avesta. The T i r a s are portrayed as
enemies of the sedentary Iranians and described, in Yajt XVII (prayer to the
goddess ASi), 55-6, as possessing fleet-footed horses.? As earlv as 641 o r 640
B.C. the nomads were known in Assyrian sources as the S a k a ~ . ~

::.See Map 1.
1. Litvinsky, 1972, p. 156.
2. The mention of the Tiiras' fleet-footed horses is possibly a hint at their link with the
territory of Turknienistan, which iron1 time immemorial was famous for its splendid
swift horses, the ancestors of the Turkmen breeds (Isroriya T14rkmt~tskoy
SSR, 1957, p. 104).
3. Litvinskp, 1972, pp. 156-7.
 A. Ahcteko,~and H. Yusupov

Many Greek writers referred to all the nomads of Eurasia, including those
of Central Asia, as Scythians; and the Persians designated all the nomadic tribes
of the Eurasian steppes, i~lcludingthe S c ~ t h i a n sas
, the Sakas. These broad clas-
sifications were based on the similarity of the culture and way of life of all the
nomads who spoke Iranian languages. The question of the actual distribution of
the different nomadic tribes or tribal groups is debatable, largely because of the
dearth of written sources. Moreover, it is well t o remember that nomadic life
characteristically entailed frequent migrations, with the result that different
tribes successively occupied one and the same territory. When it is considered
that these tribes were culturally very close to one another it is easy t o under-
stand why classical writers sometimes associated different tribes with the same
historical events. For example, in their description of Cyrus' war against the
Central Asian nomads, Cyrus fought against the Massagetae according t o Hero-
dotus; against the Sakas according t o Strabo; against the Abiae according to
Quintus Curtius; against the Derbices according t o Ctesias; and against the
Dahae according t o Berossus.
It is now generally agreed that the ancient nomads of Central Asia were
descendants of the Bronze Age cattle-breeding tribes who had inhabited the
same t e r r i t ~ r y which
,~ does not exclude, however, the probability of consider-
able ethnic intermingling and movement within and beyond the borders of the
region. These trends must have become particularly marked at the start of the
first millennium B.c., when a number of tribes changed from cattle-breeding to
a purely nomadic way of life.
This view is confirmed by anthropological studies. Between the seventh
and fifth centuries B.c., the Sakas of the Aral Sea region seem t o have a mixed
population, consisting of a Europoid, mainly Andronovo stratum with a signifi-
cant admixture of Mongoloid forms of Central Asian origin. Anthropological
materials of the Saka period from eastern Kazakhstan are heterogeneous, show-
ing genetic similarities with the population of the T'ien Shan and the Altai
mountains with a Mongoloid admixture already apparent. The Sakas of the east-
ern Pamirs occupied a place apart, among the other Saka tribes o r those akin to
them.5
The question of the distribution of the Saka tribes is extremely complex.
Current literature presents the most varied and contradictory points of view,
due principally to the paucity of written sources. The location of the different
tribes can only be determined from the extant archaeological data, and any pic-
ture of the distribution of the Central Asian tribes belonging t o the Saka-Massa-
getae community remains tentative and incomplete. The Naqsh-i Rustam
inscription of Darius I lists three Saka tribal confederations: (a) the Saki Hau-

4. Sec V o l u ~ n cI, Chapters 14 and 15.

5. Litvinsky, 1972, p. 184.
 Ancient Iranian nomads In western Ccnlral Asla

mavarga, in Ferghana, wherc they began to change over to a settled form of

life; (b) the S a k i Tigraxaudi in the region beyond the Syr Darya and in Semi-
rechye; and (c) the Saka tayaiy paradraya, o r Europcan Sakas (Scythians). In
his list of Darius' satrapies, Herodotus also mentions thc Caspians and Sacac
as belonging to the fifteenth province. They are usually located along the
southern and eastern shores of the Caspian Sea up to the mouth of thc now-
dried-up Uzboi. It is possible that the Dahae and a number of the other
groupings of the Hellenistic period may have derived from the earlier Massa-
getian Confederation.
N o kurgans or burial mounds of thc Scythian period (seventh-fifth centu-
ries B.c.) on the Uzboi are known so far, though the association of the Massage-
tae with the area seems t o be well f o ~ n d e dThe
. ~ kurgans that have been investi-
gated (dating from the fourth-second centuries B.c.) show that they are
connected with Massagetian tribal groups of a later period, perhaps the Dahae.
Their material culture is unquestionably of local origin and contains elements
common to the Prokhorovo culture of the lower Volga and Ural regions.

Society and economy of the Iranian nomads

of Central Asia
The eighth to sixth centuries B.C. witnessed the development of a class society
both among the nomadic tribes and in the settled oases. The development of a
specialized nomadic cattle-breeding economy obviously led t o major economic
and social changes, though the written sources throw little light on the social
and economic relationships that existed among the ancient nomads. The trans-
ition t o a nomadic way of life in the eighth and seventh centuries B.C. occurred
at much the same time over the whole of the Central Asian and southern Rus-
sian steppes, which were then populated by nomadic tribes showing many simi-
larities in material culture, customs and mores. A nomadic cattle-breeding econ-
omy, in which cattle were put t o pasture on a succession of different grazing
grounds, led to a sharp increase in the number of cattle. This, in turn, led t o
more tribes shifting t o a nomadic way of life. The basis of their prosperity lay in
their large herds; horses now began to ~ l a va greater role; tribes living on the
banks of the rivers engaged in fishing; and skilled craftsmen produced arms,
ornaments, sumptuous carpets and other objects.' The only indication of the
political system of administration operating among the nomads is the mention
made of the Saka-Massagetian kings, who were identified with the leaders of the
nomadic confederations. The fact that these confederations constituted an

6 . P'yankov, 1964, 1972; Yusuyov, 1976.

7. Istorijla Turkmenskoj~SSR, 1957, p. 69.
 A. Abetekov nnd H. Yusupov

organized military force provides justification for speaking of an aristocracy in

contrast to the rank-and-file of free nomads.
There may also have been slaves among the nomads, but they clearly did
not take any significant part in productive life. The range of grave goods found
in barrows also provides evidence for the existence of social differentiation
among the Central Asian nomads. In some regions such differences are particu-
larly marked in barrows from the Late Bronze Age, which contained basic types
of arms, horse-harness fittings and artefacts typical of the time. Still more strik-
ing evidence of social differentiation is to be seen in the very rich mausoleum
complexes of leaders - in the Tagisken and Uygarak cemeteries o n the lower
reaches of the Syr Darya,"n the Chilik kurgan, the barrows in central Kazakh-
stan and Kyrgyzstan and the huge royal sepulchre of Arzhan.'Judging from the
fact that, during this period, there was a shift from large burial mounds to
smaller clusters of kurgans, we may presume that changes of some sort occurred
in the family and tribal structure of nomad society.
Most scholars believe that the social organization of the steppe tribes was
based on a clan and tribe structure. Based mainly o n data in classical authors,
this conclusion is confirmed by archaeological material - barrows in the steppe
belt are usually sited in separate groups, representing the burial grounds of
members of different clans. By this stage, however, the original equality of the
steppe societies had already been lost, and there were considerable social and
property distinctions.
The burial complexes of ordinary herdsmen are found under small earth-
en o r stone mounds, whereas grandiose structures, sometimes as high as 20 m,
were erected over members of the tribal aristocracy. Complex burial structures
have been found in tombs of nobles, with different layouts in different parts of
the steppe zone. Among the European Scythians they normally took the form
of deep complex catacombs with several chambers. In Semirechye and the Altai
regions, the excavated graves of aristocrats contain monumental timber tombs.
Thus, in the Besshatir burial ground in the Ili river valley, log-built burial vaults
were found, consisting of three parts: a corridor, an antechamber and the burial
chamber proper. The walls rose to a height of 4 m and were built of horizontal
trunks of Schrenk spruce, strengthened by timber uprights dug into the
ground.1°
A large number of people must have been employed to erect these com-
plex structures. Obviously the persons buried in them were the chiefs of large
tribes or even tribal confederations. The Arzhan tumulus (Fig. I), already men-
tioned, is of interest in this connection. Its stone-built mound, 120 nl in diam-

8. Vishnevskaya, t 973; Vishncvskaya and Itina, 1971; Itina, 198 I .

9. Gryaznov, 1975.
10. Akishev and Kushaev, 1963, p. 35.
 Ancicnt Iranian nomads rn wostrrn C e n ~ r a Aua
l

FIG. 1. Arzhan: plan of a royal tomb

eter, covered a timber burial structure consisting of seventy radiall). arranged

chambers. In the central chamber the leader and his close conlpaniolis were
buried, while the other chambers held the remains of representatives of subject
tribes and possibly the offerings of friend117 tribal groups. Each o i the burials
 A. A b c t ~ k o vand H. Yusupov

was accompanied by many different articles, but a considerable proportion of

these were unfortunately plundered in antiquity. It has been calculated that
about 160 saddle horses were buried in the kurgan and another 300 horses eaten
at the funeral feast."
Social differentiations between steppe tribes of the Scythian period are
also mentioned by classical authors. For instance, Lucian states that the Pontic
Scythians were divided into a royal clan, pilophoroi, thought to be members of
the military aristocracy o r priests and the 'eight-legged ones' - ordinary herds-
men with a pair of oxen and a cart. Some believe that this division corresponds
to the division of society into three class o r caste groups that characterized all
the ancient Indo-Iranians." A similar division can also be traced in the distribu-
tion of graves in archaeological sites, for example, in the Saka burial ground at
Uygarak on the lower Syr Darya.13 Classical sources bear witness to the exis-
tence of slaves among the Scythians and the use of slave labour in the economy.
In the Scythian world, however, slavery did not develop to any considerable
extent, remaining mostly domestic and patriarchal, and 'slavery never became
widespread among the nomads' as a basis of production.14 Written sources also
suggest that women held a comparatively high position among the nomadic
Saka-Massagetae. For example, Herodotus describes Queen Tomyris as their
leader in the war against Cyrus. It is not, however, clear whether a parallel can
be drawn in this respect between Saka-Massagetian society and its western
neighbour, the matrilineal Sarmatian society.

Culture of the Iranian nomads of Central Asia

Arcl~aeologicalremains of the first millennium B.C. in the Eurasian steppes have

been studied since the nineteenth century. Initially it was thought that this evi-
dence supported the idea, based on an acquaintance with classical tradition, that
the whole population of the steppe belt belonged to the same ethnic stock. They
seemed to demonstrate cultural uniformity throughout the area. Everywhere
burials were found in barrows (kurgans) containing similar weapons, horse
trappings and works of art. The choice of motifs and their style - known as the
'animal style' - pointed to a uniform cultural pattern. All this helped to give rise
to the concept of a single Scythian culture, present throughout the Eurasian
steppes, which had spread from a single centre and belonged to one tribe or

1 1. Gryaznov, 1980.
12. Grantov~ki~ 1960,
, pp. 14-15.
13. Vishnevskaya, 1973, pp. 67-8.
14. Markov, 1976, p. 303.
 Ancient Iranian notnads in western Central Asia

people. This interpretation seemed to be in conformity with the statement of

Herodotus (IV.ll) that the Pontic Scythians came from Asia. O n this evidence,
the original homeland, common to all peoples who displayed this culture,
should be sought somewhere in the Asian part of the steppe-lands.
Some modern scholars share this point of view and are paying special
attention to dating the archaeological remains in various parts of the steppe
zone. They hope to find the area where Scythian culture had its origins in those
places where the forms of weapons, horse furniture and objects worked in the
animal style appeared earliest. It has thus been suggested that one of the earliest
complexes of this type is the famous Chilik barrow in eastern Kazakhstan.15 In
recent years the Arzhan barrow in Tuva has also attracted close attention. Its
dating is a moot point but supporters of an earlier date (ninth-eighth centuries
B.c.) believe that it is precisely this site that points to Central Asia as the zone in
which the Scythian culture that spread across the Eurasian steppes first took
shape.Ib
However, as new material accumulates and is carefully scrutinized there
seems to be increasing evidence in support of a different concept, that is, that in
the Scythian epoch there existed in the steppe zone not one but rather a whole
series of distinctive cultures belonging to different peoples. Even the features
that stamp these cultures as similar show appreciable local variations, while
their other characteristics are equally specific. The common features are due not
only to their having come from a single source o r to ethnic affinities but also to
close contacts between the steppe tribes. These factors account for similar econ-
omic structures, bringing in their train an outward unification of life-style, and
leading t o the formation of a Scytho-Siberian cultural entity. Within this entity,
every culture pattern is 'completely distinct and original by virtue of its own
articular historical past and the articular conditions ruling in the country in
which it is found'." This of course in no way rules out the existence also of eth-
nic ties between some of these peoples.
A whole range of such Scythian-type cultures in the Eurasian steppes has
already been studied. In addition to Scythian remains in the Black Sea area,
these studies have covered Sarmatian complexes in the country round the lower
reaches of the Don and Volga'Qnd various groups of Saka sites in Central Asia:
on the lower Syr Darya," in the Pamirs20and in Semire~hye,'~ and a whole series
of cultures whose assignment t o any particular people mentioned in the sources

15. Chernikov, 1965.

16. Terenozhkin, 1976, pp. 2 10-1 1.
17. Gryaznov, 1978, p. 18.
18. Smirnov, 1964.
19. Tolstov and Itina, 1966; Vishne\:ska!a, 1973.
20. Litvinsky, 1972.
21. Akishev and Kushaev. 1963.
 A. ALetekov and H. Yusupov

is problematical, that is, the Tasinol culture in central Kazakhstan,'? the Pazirik
culture in the Altai,?' the Tagar culture in southern Siberia.
Among the steppe peoples of the S c ~ t h i a ngroup the predominant econ-
omic activity was nomadic herding, but in some areas the economy was of a
more complex nature.
Herodotus (IV.17-18) describes the Pontic Scythian tribes partly as
sedentary agriculturalists but he says (1.216) that the Massagetae of Central Asia
'sow no grain but live by keeping herds and fishing. . . . They also drink milk.'
Investigation of the large fortified settlement of Chirik-Rabat, o n the north-
western confines of the Kyzyl Kum Desert, a settlement connected with the
Massagetae, certainly contradicts the statement that they led a purely nomadic
life and shows that in their economy the ancient traditions of fishing were com-
bined with tillage of the land and semi-nomadic stock-raising.
Close ties between the nomadic and agricultural societies of Central Asia
can be traced not only in the political and ethnic but also in the cultural and
productive spheres. As is usually the case with nomads in general, craft produc-
tion among the nomads of Central Asia was not so well developed as in the set-
tled agricultural provinces, from which they obtained the wares they needed. A
social division of labour thus grew u p between the nomads and the settled agri-
culturalists and craftsmen.
Classical writers were much impressed by the excellent quality of the
arms of the Central Asian nomads. Quintus Curtius (IV.9.3) noted that they
had coats of mail made 'of iron plates'. According to Arrian (III.13.4), the Cen-
tral Asian warriors went into battle 'carefully covered' with a metal coat of mail.
They also used metal helmets and shields of various shapes and sizes. According
to Herodotus (I.215), the Massagetae's horses were protected by breast-plates.
It has been suggested that it was in Central Asia that equine armour first
appeared.?" In the absence of any archaeological evidence for the advanced pro-
duction of weapons by the nomads themselves, it may be supposed that some of
their arms, especially defensive armour which required much workmanship,
were imported from the provinces inhabited by a sedentary population.
O n the other hand, evidence of ceramic production by the nomads them-
selves is provided by the so-called 'barbaric ceramics', distinguished by an
extremely coarse texture. Typical of the tribes in the Uzboi region, in particular,
were the large trough-shaped vessels, used as ossuaries; in the oases, they were
unknown. Another point suggesting that they were locally produced is their
usually large size (over 1 m in length), which would have made it difficult to
transport them over long distances on account of their fragility.
Judging by Hcrodotus' account of the religion of the Massagetae, they

22. Margulan, 1966, pp. 303 et scq.

23. Rudcnko, 1953.
24. Gafurov, 1972, p. 92.
 Ancient lranian nomads in western Central Asra

practised the cult of the supreme sun god - Mithra - associatcd with various
forms of fire- and horse-worship. Some scholars arc of the opinion that as
Zoroastrianism spread, some of thc Sakas adopted its tcachings.15 Thc qucstion
of the Sakas' religion may be approached, it would scem, in thc samc way as
that of their culture. There could not have been only onc religion in such a vast
region. The chances are that there were local interpretations of similar bcliefs
and rites, these being reflected in varying burial ccrcmonials in different prov-
inces.
Some information about the religious view of the nomads of north-west
Turkmenistan is provided by the I ~ h i a n l i , ~
a "monumental stone building dating
from the fifth t o second centuries B.C. This was, in all likelihood, a cult centre
for the nomadic tribes of the plateau beyond the Uzboi.?' The plan of the build-
ing is rectangular (35 x 40 m) with rounded corners to the south-east and south-
west. The upper part of the building is topped by a thick and intricately con-
structed stone structure on which a hot fire had burned for a long time. The
eastern and western part along the slope of the central elevation is traversed by
parallel arched rows of vertically standing slabs, the gaps between which are
filled with ashes of the 'sacred' fire that were brought here. South of the central
high ground were outbuildings and passageways lined with large stone slabs
standing o n their edges. The surviving walls are as much as 2 m high. O n the
northern side, there is a semi-enclosed right-angled area with two altars. Other
buildings contain large hearths or altars.2RTraces of the prolonged action of fire
are visible everywhere, and slag, the bones of animals (predominantly horse
skulls and hooves), and a considerable number of bronze arrow-heads have
been found. T w o distinct types of pottery have been discovered in the complex:
the local Daha-Massagetian earthenware and the roundware brought in from
the south-western and southern regions. The large quantity of horses' bones
inevitably brings t o mind Herodotus' observation that the only god the Massa-
getae worshipped was the sun, t o which they sacrificed horses (1.216). The
horse, reflecting the ideology that was taking shape among the nomadic ~ e o ~ l e s ,
was widely represented in the distinctive Scythe-Saka-Massagetian art known as
the 'animal style'. Tacitus (VI.37) also noted the ideological significance of the
horse image in speaking of the sacrificial slaughter of horses as a Parthian cus-
tom; so did Philostratus, who observed that the Parthian king Vardanes sacri-

25. Litvinsky, 1972.

26. Allnost in the centre of the burial ground containing barrows of different periods
located on a hill to the south of the Dordul heights (a plateau be!.ond the Uzboi river)
dominating the sands of the Kara Kum lowlands, the ruins of a stone building were
discovered, which excavations showed to have n o connection with the burial ground.
27. Yusupov, 1976, p. 42.
28. The fact that Ichianli had a number of peripheral 'altars' in addition t o a central altar
suggests that the former, unlike the latter, were directly associated with burial grounds.
 FIG.2. Decor'ltion cut out from Icather of n saddle
covering: first kurgnn a t Pn7.irik
(mountainous Altai). (After Rudenko, 1953.)

ficed a white horse of the best Nisa breed. T h e horse was equally popular in
both Scythian 'lnd Saka art.2"
T h e art of the Achaelnenids also owes much t o the nomadic art from
which it borrowed s o many features. A t the same time, Achaemenid works in
turn had a strong influence o n the culture of the nomads. Evidence of the cultu-
ral and trade relations between Achaemenid Iran and Central Asia and the
regions t o t l ~ enorth-east of it is provided by various objects (everyday and cere-
n~onial)discovered in the excavations of kurgans, such as the Arzhan burial
mound (eighth-seventh centuries B.c.) and the Tuva and the Pazirik barrows
(sixth-fourth centuries B.c.) in the Altai. Here, d u e t o the permanent layer of
ice, articles of leather (see Fig. 2), wool and thick felt have survived in an excel-
lent state of preservation. Of considerable interest arc pieces of woollen cloth
and a short pile carpet with woven designs which suggest that they were of Ira-
. .
nian orlgln, though a Middle Asian provenance is not altogether excluded.'' It
seems reasonable to infer that their basic type of dwelling was the portable vurt.
Burial sites at different localities s h o w differences in form of construction, the
objects they contain and the manner of burial. I n the vicinity of the Sarikamish
delta of the Ainu Darya, in the lower redches of the Syl- Darya and in Semi-
rechye, alongside shallow-ditch graves were the huge barrows of the aristoc-
racy, with complex wooden constructions o r sophisticated structures in
unbaked brick typical of the architecture of t11c seventh t o third centuries 15.c..
In north-wcst Turkmenistan and in the eastern P'lmirs, mon~imental stone
vaults built at ground level were quite colnmoli and widespread, ser\.ing '1s
family o r trib'll tombs. Such tombs have a rich v,~rict\.-of articles - n7ea-
pons, o r n a n e n t s (including some in the animal st~sle),and horse tr'lppings cry

1972, pp. 90-7.

29. .Ckifi I Sn~-tr~rz!i,
30. Gryaznov, 1975, p. 10; Istorijla 7M1-k~nc.r/skojl
.?.?A, 1'157. pp. (,C1-7 1
 Ancient Iranian nomads in western Crntral Asia

similar t o those found in the Sarmatian monuments in the Ural region. Besides
local earthenware, they contain bronze weapons and ornaments, imported
beads (carnelian and lazurite) and pottery vessels brought in from the oases,
providing further evidence of the links between the Saka-Massagetae and the
Central and Western Asian worlds. T h e lion and panther motifs in the art of
south Tagisken and Uygarak locate the Sakas of the Aral Sea region in the area
t o which the Scytho-Siberian animal style had spread." Overall the nomadic
tribes made a very significant contribution t o the development of the Central
Asian peoples. Military and political vitality, vigorous economic development,
fostered b y the commerce essential t o the nomadic way of life, and strikingly
original art were their characteristic features.

31. Vishnevskaya, 1973; Vishnevskava and Itina, 1971; Itina, 1981.

33
 M E D I AA N D A C H A E M E N I DIRAN::-
M. A. Dandamayev

The immigration of the Median and Persian tribes

When and how the Medes and Persians reached the Iranian plateau is still an
open question, though it has been discussed in scholarly literature for decades.
Until recently, some scholars held that the original homeland of the Iranians lay
in Middle Asia, from which some of the tribes were thought to have reached the
Iranian plateau between the ninth and eighth centuries B.C. But many now
consider that they came there via the Caucasus from the steppes of southern
Russia.' V. I. Abaev, for one, considers that the Iranian tribes were in southern
Russia in the early second millennium B.C. and that, subsequently, some of
them left for Iran via the Caucasus and for Middle Asia via the north Caspian
coast, while the Scythians, who were also of Iranian stock, remained in southern
RussiaSz
The Medes and Persians certainly appeared in Iran as early as the begin-
ning of the first millennium B.C. There were indeed places where the older, non-
Iranian tribes - the Kassites, Kutians and others - remained politically predomi-
nant during the ninth and eighth centuries B.C. But from the second half of the
seventh century the Iranians formed the majority in many parts of western Iran,
including the region that was to become the Median kingdom and the lands to
the west. When the Iranians appeared there, they already had advanced cultural,
social and economic traditions; they engaged in both pastoralism and agricul-
ture, were thoroughly acquainted with metals, reared horses and used the
chariot. Like the Later Achaernenid Empire, the Kingdom of the Medes arose in
a region where Iranian speakers re dominated and was rooted in the previous
development of the Iranian tribes.

:' See Map I .

1. Grantovskiy, 1970, pp. 7 et seq.
2. Abacv, 1969, pp. 121-4.
 M . A. Dandamayev

The early history of the Iranians is only scantily reflected in written

sources. Assyrian texts show that the Medes had settled in north-western Iran at
the beginning of the first millennium B.C. In the ninth century B.c., this region
had scarcely begun to change from a tribal to a class society, and was divided in-
to scores of petty princedoms, ruling alike over the Medes and the indigenous
peoples of Kutian o r Kassite descent.
The first reference in Assyrian sources to the Persians also relates to the
ninth century B.C. An inscription of King Shalmaneser 111, written around 843,
mentions the province of Parsua; in 834 the Assyrians levied taxes from twenty-
seven 'kings' of that province. Until recently it was widely assumed that Parsua
was near Lake Urmia, but Levine has recently demonstrated that it was most
probably in the central Zagros mountains.'
At that time the Persians were not yet united but were led by many sep-
arate chieftains. Assyrian texts of the late eighth century B.C. speak of the land
Pariumai to the east of the modern Sulaymaniyah, that is, north-west of Elam.
The Persians are thought to have parted from the Median tribes around 800
B.c., and gradually to have moved south-eastwards. In 714 they are mentioned
as subjects of the Assyrian monarch Sargon 11. With the passage of time they
came to occupy the ancient land of Elam in south-west Iran, which was named
Pdrsa after the new arrivals.
This region is roughly equivalent to the modern Iranian province of Fars,
an Arabization of Middle Persian Pdrs going back to the O l d Persian name
Parsa used to designate the land and people of the ancient Persians as well as
their capital, Persepolis. The name 'Persis' is derived from 'Persia', the Greek
transcription of O l d Persian Pdrsa. The forerunner of the country's modern
name, Iran, was first mentioned by the Greek author Eratosthenes in the third
century B.C. as Ariane, deriving from O l d Iranian drydnam xiaeram meaning
'land of the Aryans', since the Persians and Medes held themselves to be Aryan
tribes. Both they and other Iranian tribes such as the Bactrians, Chorasmians,
Sogdians and Sakas acknowledged their common origins and the kinship of the
languages they spoke.
Archaeologists such as Ghirshman suggest that the route the Iranian
tribes took is further indicated by changes in material culture, more especially in
particular forms of burial o r the decoration of horse harness and pottery.
Ghirshman's view is that rather than conquering Elam, the Persians acquired
lands there by entering the service of local rulers as cavalry, the latter being un-
known in Western Asia before they reached the Iranian plateau.'
Before the early 640s B.C. the Persians were dependent on the kings of
Elam, briefly becoining tributaries of the Assyrians. Apparently they were even

3. Levine, 1974, pp. 106 et seq

4. Ghirshman, 1977, p. 51.
 Mrdiu and Achaemrnid Iran

then organized in a tribal alliance headed by chieftains of the Achacmcnid

lineage. The founder of the dynasty is traditionally held to be Achacmenes.
From 675 to 650 B.c., the Persian alliance was led by CiSpiS (Teijpes in Greek
transcription), whom later tradition held to be Achaemenes' son. The kingship
then passed to his son, Cyrus I, who was, as is clear from an Assyrian inscrip-
tion, the lord of Pariumai and about 646 B.C. sent his son as a hostage to Nini-
veh, the capital of Assyria.
Some of the Persians adopted a sedentary life-style, while others remained
nomadic pastoralists. Gradually the tribes came to occupy the greater part of
the Iranian plateau. The Medes and Persians were then merely part of the grca-
ter Iranian world that stretched from the northern coast of the Black Sea to
what is now Afghanistan. The ethnically related Cimmerians and Scythians
lived to the north of the Black Sea. Herodotus (VII.64) states that the Persians
called all the Scythian tribes 'Sakas', while the Greeks called the nomadic tribes
of southern Russia and Middle Asia 'Scythians'. In modern scholarship the
name 'Sakas' is reserved for the ancient tribes of northern and eastern Central
Asia and Eastern Turkestan to distinguish them from the related Massagetae of
the Aral region and the Scythians of the Pontic steppes. These tribes spoke Ira-
nian languages, and their chief occupation was nomadic pastoralism.

Media
The need to resist the marauding forays of the Assyrians hastened the unifica-
tion of the petty Median princedoms. In 672 B.c., the Medes, supported by
Cimmerians and Scythians who had thrust into Western Asia from the Pontic
steppes at the end of the eighth century and beginning of the seventh century
B.c., rebelled against Assyria. The Assyrian king Esarhaddon persuaded the
Scythians to abandon the rebels, but the Medes fought on and won their inde-
pendence, setting up their own state. By the middle of the seventh century B.C.
Media was a major kingdom ranking with Elam, Urartu, Mannai and, of course,
Assyria.
In 653 B.C. the Medes mounted an attack on Assyria, but the Scythians,
who were allies of Assyria, fell on the Medes. Pressed o n two fronts, the Medes
were defeated, and from 653 to 624 B.C. the Scythians ruled Media. In 624 B.C.
King Cyaxares defeated the Scythians and finally united all Median tribes into a
single state whose capital was Ecbatana. Cyaxares soon established a powerful
regular army, reorganizing it by type of weapon into spearmen, bowmen and
cavalry, rather than as the previous tribal levies.
The Medes could then turn against their time-honoured enemy, Assyria,
which had already been at war with Babylonia for over ten years. In 614 B.C.
they seized Aiiuv, the ancient capital of Assyria; and in 612 B.c., helped by the
Babylonians, they stormed its chief city, Niniveh. The Assyrian Empire lay in
 M. A. Dandamayev

ruins and the Medes took eastern Asia Minor and northern Mesopotamia, the
heartland of Assyria.
C ~ a x a r e s called
, the 'founder of dominion over Asia' by the Greek trage-
dian Aeschylus, set about expanding the frontiers of his state at the expense of
his southern and eastern neighbours. O n e of the first blows fell on Persia
around 624 B.C. Judging by later indirect evidence, Cyaxares also succeeded in
taking Parthia, H ~ r c a n i at o the east of the Caspian Sea, and Armenia.
About 590 B.C. he annexed Mannai, a major state t o the west of Media. At
the same time the Medes subjugated Urartu. When in 590 B.C. the Median army
reached the River Halys, Alyattes, ruler of the flourishing state of Lydia in Asia
Minor, was alarmed by Cyaxares' conquests and opposed him. The war be-
tween the two kingdoms lasted five years, with neither side gaining a decisive
victory. O n 29 May 584 B.C. an eclipse of the sun during a battle o n the Halys
was interpreted by both sides as an ill omen. They therefore stopped the war,
and made a peace treaty establishing the River Halys as the boundary between
Lydia and Media. In the same year Cyaxares died, bequeathing a powerful state
to his son Astyages. During the following century, Media was the centre of Ira-
nian material and intellectual culture, which the Persians subsequently took up
and developed. Median art in particular was one of the chief components in
subsequent Achaemenid art.

The rise of the Persian Empire

Between 600 and 559 B.C. Persia was ruled by Cambyses I, a vassal of the
Median kings. In 558 his son, Cyrus 11, became king of the sedentary Persian
tribes, the foremost of whom were the Pasargadai. The Persian confederation
also included the Maraphioi and the Maspioi. The heartland of the Persian king-
dom lay around the city of Pasargadae, built chiefly in the early part of Cyrus'
reign (Fig. 1). The hill and plains tribes - the Kyrtioi, the Mardoi (some of
whom also lived in Media), the Sagartioi and some nomadic tribes - and also the
settled Karmanioi, Panthialoi and Derusiai, were later subjected to Cyrus, appa-
rently after the war with Media.
Persia's social organization at this time can be described only in outline.
The fundamental social unit was the nmdna o r large patriarchal family. The
nmdnapati, the head of the family, was a kind of paterfamilias with unlimited
temporal and spiritual power over all his kin. The totality of families formed the
clan (vis). The clan commune, like the later rural commune, consisted of a num-
ber of families and was governed by its elder (vispati). It remained a powerful
force for many centuries. The clans were united in a tribe (zantu) led by a chief
(zantupati), and several tribes made up a province (dahyu) governed by a king.
The chief occupation was agriculture and animal husbandry, particularly the
breeding of horses.
 Media and Achaemenid Iran

When Cyrus became King of Persiz~,four major powers remained in the

Near East - Media, Lydia, Babylonia and Egypt, Xn 553 a x , Cyrus revolted
against Astyages, King of Media, to whom the Persians had formerly been sub-
ject. The war lasted for three pears, ending in 550 with a complete victory far
the Persians, Ecbatana, the former Median capital, became om of (I-ym~' royal
residences. After subduing Media, Cyrus formally retained the Median king-
dom and adopted the Median king's official titles: Great King, King of Kings,
King of the Lands'. With Media conquered, Persia, previously a little-known
outlying province, entered the main stage of world history and was for the next
two centuries to play a palitically dominanr role.
The Persians took the whole of Elam and in 549-348: B.C. extended their
dominion to the lands that had been part of the Median Empire - Parthia, Hyr-
cmia and probably Armenia. Meanwhile Croesus of Lydia had observed Gyms'
rapid successes with alarm and began to prepare for war. At the initiative of the
Egyptian Phafaah, Amasis, Croesus coa~ludedan, alliance with him; but the allies
failed to grasp the need for swift and decisive acrion, while Persia grew daily in
power. In late October 548 B,C* there was a bloody battle between the Pessians
and Lydians on the River Halys, but the outcome was u m r e and neither
hazarded a further fight. Croesus retired to his capital -of Sardis, and the next bartie
was fought outside its wails. Pressed by superior forces, the Lydian3 had to take
refuge in the city, After a siege of fourwen days, Sardis fell to the Persians in May
547, and the Lydian kingdom tame to an end. It was then the mrn of the Greek
city-states in Asia Minor, which were s o w forced to acknowledge Cyrus' rule.
The pre-Achaemenid states in Central Asia
The chronology of Cyrus' next campaigns is not fully known. H e instructed his
commanders to complete the subjugation of Asia Minor while he himself went
to Ecbatana to prepare for the conquest of Babylonia, Egypt, Bactria and the
Sakas. As is well known, the Persians conquered Egypt only after Cyrus' death.
H e took Babylonia in 539 B.C. Bactria and the Sakas were certainly subdued
during his lifetime, as the Bisutun inscription, made around 518 B.c., lists as
Persian possessions Margiana, Bactria, other Central Asian countries, Gandhira
(Old Persian GandZra, O l d Indian Gandhdra) and Sattagydia in the east. It is
thus clear that Persian rule had been extended t o the Indus and Jaxartes (now
Syr Darya) under Cyrus. During the reign of his son, Cambyses 11, there were,
apparently, no wars in those parts. From Pliny's Historia Naturalis (VI.92) we
know that Cyrus sacked the city of Capisa (KZpiia) (north of modern Kabul);
Arrian writes in his Anabasis (VI.24.3) of Cyrus' attack o n 'the land of the
Indians', where the Persians lost a large part of their forces; and both Arrian in
his Anabasis (111.27.4) and Diodorus (XVII.81.1) speak of the Ariaspoi, a tribe
on the southern border of Drangiana, which provided Cyrus' army with food
during its campaign and was rewarded with exemption from paying taxes to the
Persian king.
When did the Persians win these lands? Some scholars believe that Cyrus
only conquered Central Asia after the subjugation of Babylonia, but that
appears improbable. According to Herodotus (1.177-8), Cyrus successively
subdued all the people of Asia while his commander, Harpagus, was ravaging
the cities of Asia Minor, and only then attacked Babylonia. It may thus be
assumed that he won his Central Asian provinces after his victory over Lydia
but before the war with Babylonia. Highly interesting in this regard is a report
by the Babylonian historian Berossus (third century B.c.), who probably drew
on Babylonian sources for his description of Cyrus' campaigns. H e writes:
'Cyrus attacked Babylonia after he had reduced all the rest of A ~ i a ' ,that
~ is to
say, only after capturing his most distant provinces in the north and east be-
tween 545 and 539 B.c., Drangiana, Margiana, Chorasmia, Sogdiana, Bactria,
Aria, Gedrosia, the Saka tribes, Sattagydia, Arachosia and Gandhira, did Cyrus
turn his attention to Babylonia.
There are unfortunately no reliable written sources for the history of
Middle Asia prior to the Achaemenids. Even Astyages of Media may have had
to contend to some extent with the Middle Asian tribes. Ctesias says that the
Sakas were under Median rule, but the sources neither support nor refute him.
There is no trustworthy information about earlier clashes between the Middle
Asian tribes and peoples living to the west of them. Diodorus (11.4) and
 Mrdia and Achaernenid Iran

Justin (1.1) speak of a siegc of Bactra by the legendary Assyrian monarch Ninus
and its capture by the equally legendary Semiramis; but the available sources
suggest that Assyrian forces never penetrated farther cast than Mcdia proper.
It has often been suggested that various organized states existcd in prc-
Achaemenid Middle Asia. In the last century M. Duncker wrote that an ancient
state of Bactria had arisen as early as the ninth century B.C. Although his view
was rejected by others, J. PriSek later argued that there was no reason to dispute
the existence of an ancient Bactrian kingdom, since the Avesta spoke of the Bac-
trian monarch Viitaspa, the legendary patron of Zoroaster. In his view, Bactria
must have been an independent state before the Persian conquest, since it was a
major administrative province under the Achaemenids. H e further suggested
that Margiana also had its own kings prior to the Achaemenids."urther proof
that there was an ancient Bactrian kingdom is sometimes seen in Ctesias' report
of the Bactrians' resistance to Cyrus, and in Herodotus' suggestion that
together with Babylonia, Egypt and the Sakas, Bactria was the major obstacle to
Persian world conquest.
From the foregoing, it will be seen that the question of the existence of an
ancient Bactrian kingdom remains open. Fresh evidence about the level of de-
velopment in Middle Asia has come from excavations by archaeologists from
the former Soviet Union. V. M. Masson considers that as early as the first third
of the first millennium B.c., an urban civilization had grown up in Middle Asia
on the oases of major irrigation systems, and that towns with citadels had been
built on man-made platforms.' A case in point is ancient Hyrcania, in the south-
west of modern Turkmenistan, where settlements up t o 5 ha in area with cita-
dels have been found. Masson believes that an early class society had begun to
spring u p in settled oases, and that this corroborates the information in the
Avesta that major political units already existed in Central Asia. O n the other
hand M. M. D'yakonov argues that there were no large organized states in Mid-
dle Asia before the Persian conquest and both the farming oases and the barbar-
ian periphery with its nomadic Saka population came with the decay of military
democracy. In the middle of the first millennium B.c., irrigated agriculture in
the major river valleys had given birth t o states in Chorasmia, Sogdiana, Mar-
giana and Bactria; but their borders coincided with those of the irrigation sys-
tems. More specifically, D'yakonov postulates that Margiana had no tradition
of monarchy, since in the Bisutun inscription the leader of the rebellion there at
the start of Darius 1's reign is called a chief, while the rebel leaders in Persia,
Media, Elam, Babylonia and other countries are self-proclaimed kings. D'vako-
nov points out that the characteristic occupation of the society described in the
Avesta was pastoralism and that agriculture ~ l a y e donly a subordinate role."

6. PriSek, 1906-10, pp. 51-4.

7. Masson, 1959, pp. 58 et seq., 122 et seq.
8. M. M. D'yakonov, 1961, p. 75.
 Avestan society represents one of the most difficult problems. It has long
been established that its material culture was archaic. The Avesta makes no
mention of iron, although bronze was in use; and a sophisticated urban life,
enduring states o r crafts ~ r a c t i s e dseparately from farming are unknown. The
GdOds, the earliest part of the Avesta, which reflect the material culture and
social relations of eastern Iran and Middle Asia in the pre-Achaemenid period,
depict a society of sedentary herdsmen and farmers, still preserving a system of
clan and tribal relationships. As the clan communes disintegrated, so early class
units began to spring up, and the emergence of classes is reflected in the GdOds,
which contain a protest against the rule of the tribal ilite.
The country in which Zoroaster preached is called a i r y d n m va@ in the
Avesta. Many scholars have located that country in Chorasmia, supposing this
to have been the homeland of Zoroastrianism, from which it subsequently
spread to Sogdiana, Margiana, Bactria and other countries. D'yakonov suggests
that the Avesta was written in the Helmand/Tedzhen/Hari-rud alley,^ and the
same view is held by G . Gnoli, who considers that the home of the Avesta was
Sistan and the adjacent regions.1° At any rate, the society in which Zoroaster
taught arose in eastern Iran, where the settled tribes and the Iranian nomads
met.
J. Marquart argued that a i r y d n m vae/6 was a major pre-Achaemenid
state, centred o n Chorasmia, that was destroyed by Cyrus - a hypothesis later
supported by W. B. Henning, I. Gershevitch and S. P. Tolstov. Henning held
that the state's original centres were Merv and Herat (ancient Aria, that is, the
Haraiva of O l d Persian inscriptions). The chief basis for conjectures about a
'Greater Chorasmia' is the report in Herodotus (111.17) that a dam o n the River
Akes (thought to be the modern TedzhenIHari-rud, the valley of which bor-
dered on Parthia and Drangiana) belonged to Chorasmia and that, under the
Achaemenids, Chorasmia, Parthia, Aria and Sogdiana made u p a single satrapy.
These scholars feel that the latter's boundaries were originally those of a state
conquered by the Persians."
Archaeologists, however, consider that in Chorasmia proper substantial
progress in the development of irrigated agriculture may be observed only in
the sixth century B.c., while in the eighth and seventh centuries B.C. the country
had neither a numerous population nor an advanced irrigation system. The rise
of major settlements such as Kalali-gir was definitely an Achaemenid phenome-
non. O n these grounds, Gnoli feels that a 'Greater Chorasmia' is unlikely to
have existed.12
Judging by the evidence available to date, the first Middle Asian cities

9. I. M. D'yakonov, 1971, p. 142.

10. Gnoli, 1975, pp. 386 et seq.
11. Hcnning, 1951, pp. 42 ct scq.; Gershevitch, 1959, p. 14.
12. Gnoli, 1975, pp. 230-3.
 Media and Acharmenid Iran

began to rise only in the middle of the first millennium r3.c:. - the capitals of
Sogdiana, Bactria and Margiana, which were some dozcns of hcctares in arca
and possessed citadels. An advanced farming culturc based on artificial irriga-
tion had appeared in these regions as carly as the seventh century i j . ~ : . ,but in all
probability no large organized states existed there at that timc.

The Achaemenid Empire world power

In the autumn of 539 B.C. the Persians captured Babylonia. All the lands to thc
west as far as the Egyptian border - Syria, Palestine and Phoenicia - voluntarily
submitted t o Cyrus. This accomplished, Cyrus resolved to secure his north-
eastern frontiers from invasion by the Massagetae. These forays had caused con-
siderable damage t o the settled parts of the O l d Persian Empire. T o put an end
to the threat of Scythian invasions, Cyrus set up a number of fortified border
settlements that classical writers usually call cities. O n e such, founded in Sog-
diana in the basin of the Jaxartes where Ura-tyube now stands, existed until the
Macedonian invasion and was called Cyropolis by classical authors.
In 530 B.C. Cyrus mounted a campaign against the Massagetae, the
nomadic tribes living on the plains north of Hyrcania and east of the Caspian.
During a battle beyond the Amu Darya, Cyrus was defeated and killed, prob-
ably at the beginning of August. His defeat left a profound imprint on classical
literature. According t o ancient Greek authors, Cyrus lost 200,000 soldiers - a
figure that is, of course, grossly exaggerated. There are several different accounts
of the death of Cyrus. According t o Herodotus (1.205-14) Cyrus took a camp
of the Massagetae by subterfuge; but their main force, commanded by their
queen, Tomyris, then inflicted a major defeat on the Persians and Cyrus'
severed head was flung in a sack full of blood. Herodotus writes that this was
the fiercest battle ever fought by 'barbarians', o r non-Greeks.
Berossus and Ctesias give a somewhat different picture of the encounter.
According t o Berossus, Cyrus died fighting the Dahae, a Scythian tribe of Mid-
dle Asia, while according to Ctesias his last battle was against the Derbices, sup-
ported by Indians using battle elephants. In the fighting an Indian speared
Cyrus in the liver, the wound proving fatal three days later. O n hearing the
news, the Scythiail king Amorges dashed t o the Persians' aid with 20,000 tribal
horsemen and, after a fierce battle, the Derbices were defeated.
The differences in the accounts of which tribes defeated Cyrus are due t o
the fact that the Derbices were part of a ~ o w e r f u ltribal confederation of the
Massagetae living in the steppes between the Caspian and Aral seas. In Ctesias'
time they were the most famous anlong the Massagetae. But long before the
time of Berossus (third century B.c.), the Dahae had replaced the Massagetae on
the stage of history, and that is why he named them as Cyrus' adversaries.
Cyrus' eldest son Cambyses I1 came to the Achaemenid throne in
 M . A. Dandamayrv

530 B.C. and soon began preparations for an attack on Egypt. The Egyptian
army was quickly routed, its fleet surrendered without a fight and in May 525
B.C. Egypt became a Persian satrapy. Calnbyses died in March 522 B.c.; and
after a seven-month interval during which Gaumata the Magus ruled, the Per-
sian throne was seized by Darius I. At the start of his reign the peoples of Baby-
lonia, Persia, Media, Elam, Margiana, Parthia, Sattagydia, the Middle Asian Saka
tribes and Egypt all rose against Darius. The revolts were bloodily put down in
the course of a year o r so.
In 519 B.C. after he had restored the empire of Cyrus t o its former bar-
ders, Darius led a campaign against the Scythian tribe known as the Saki
Tigraxauda, that is, 'the Sakas who wear pointed caps', described in the fifth
column of the Bisutun inscription. In some passages, however, the inscription is
damaged, and scholars have restored the missing characters in different ways.
According to J. Harmatta, Darius reached the Aral Sea at the mouth of the
Araxia, which can be equated with the Araxes of Herodotus, that is, the Oxus
of the Hellenistic period (modern Amu Darya)." It has frequently been argued
that the fifth column of the Bisutun inscription refers t o Darius' famous cam-
paign against the Black Sea Scythians - but that view is untenable if only
because all Achaemenid inscriptions list the Sakd Tigraxaudd, against whom the
campaign was mounted, together with the Saka Haumavargd and other Middle
Asian tribes and satrapies. Thus the Saka Tigraxauda and Sakd Haumavargi
alike must both have dwelt in Middle Asia. The Black Sea Scythians figure in
Achaemenid inscriptions as the 'Overseas Sakas' o r Sakd tayaiy paradraya, in
the same context as Thrace (Skudra).
In the earliest inscriptions, when the Persians had only one Scythian tribe
to contend with, they called them simply the Sakas. In other words, they
invested the collective name 'Sakas' with a definite ethnic connotation. Later,
when they had subdued other Scythians, they began to distinguish between
three tribes: the Saka Haumawarga, the Saka Tigraxauda and the Sakd tayaiy
paradraya, the Overseas Sakas of the Black Sea and of Middle Asia. The Saki
Haumavarga of Middle Asia appear to have been reduced first, under Cyrus.
Skunxa, the chief of the Saka Tigraxauda, against whom Darius I campaigned in
519 B.c., is shown o n the Bisutun relief (see Fig. 2) as a captive wearing a sharp-
pointed cap some 30 cm high. Darius replaced hiin by another chief of the same
tribe. The Sakd Tigraxaudd (who wear pointed caps) were known to Greek
authors as the Orthokorybantioi, a direct translation of the Old Persian name.
They differed from other Scythians in Central Asia (and from the Chorasmians
and Bactrians) in their pointed headgear. In other respects they all dressed simi-
larly in a short tunic with a broad belt and narrow trousers.
The eastern Iranians figured prominently in the Achaemenid wars. Bactria

13. Harmatta, 1979, p. 27

 Media and Achaemcnid Iran

FIG.2. Rack relief of Darius I at Bisutun.

(Photo: Deutsches Archaologisches Institut, Abteilung Tehcran/Trinpclrnann.)

alone provided the Persian army with 30,000 horsemen, while the Saka tribes
supplied large numbers of mounted bowmen, who served in Persian garrisons
in Egypt, Babylonia and other lands. Together with the Persians, Medes and
Bactrians, the Sakas formed the core of the Achaemenid army, and distin-
guished themselves for their bravery in the major battles of the Graeco-Persian
wars. Terracotta statuettes of Sakas, Bactrians, Chorasmians and Sogdians wear-
ing hoods and long narrow trousers have been found during excavations in
many cities of the Old Persian Empire, from Egypt to Central Asia. The Persian
army's chief weapon was the Scythian composite bow, which had far better bal-
listic properties than those of other peoples. That is why the Medes and the Per-
sians adopted the mounted archery tactics of the Scythians.
Having conquered the SakZ TigraxaudZ, the Persians took Thrace, Mace-
donia and ancient north-western India between 519 and 512 B.C. By the end of
the sixth century B.C. their empire stretched from the Indus in the east to the
Aegean in the west, and from Armenia in the north to the first cataract on the
Nile. Thus the greatest power of the ancient world came into being, uniting
dozens of countries and peoples under the Persian kings. The social and econ-
omic institutions and cultural traditions established in the Achaernenid period
played a great part in world history and endured for centuries, seming the states
of Alexander the Great, the Seleucids, Parthians and Sasanians.
The Achaemenid Empire, however, soon began to weaken. During the
wars with Greece in the first half of the fifth century B.C. the Persians suffered a
number of major reverses in mainland Greece and at sea. In the fifth century,
Egypt, Babylonia, Media, the Asia Minor provinces and others often revolted
against Persian rule. In the early fourth century B.C. the Persians lost Egypt,
which was recovered only in 342 B.c., shortly before the empire colIapsed,
Finally the Indian satrapy was also lost, while Chorasmia, Sogdiana and the
Sakas became allies rather than subjects of the Persian kings. In addition, from
the late fifth century B.C. the satraps of Asia Minor engaged in constant feuds
from which the Achaemenids generally remained aloof. Some satraps frequently
rebelled against the kings and, relying o n the help of Greek mercenaries,
attempted to become independent monarchs. Lastly, the court nobility came to
wield great influence and intrigued against the kings it disliked.
Military setbacks in the Greek wars forced a radical change of diplomacy.
They began to set states against each other, using bribery to that end. During
the Peloponnesian war, Persia, still interested in weakening Greece, helped first
Sparta and then Athens.
While the klite of Persia's aristocracy was engaged in palace intrigues and
coups, a dangerous adversary was looming o n the political horizon. In the
spring of 334 B.c., Alexander's Macedonian army set out against Persia.
Although Persia had the largest army, it was considerably weaker than
that of the Macedonians, and was n o match for Alexander's heavy infantry.
Although Persian commanders had long known that Greek and Macedonian
soldiers had better weapons and tactical skills than their Persian counterparts,
they had done nothing to improve their army and had ignored all the achieve-
ments of Greek military art. Their units of Greek mercenaries were now the
strongest part of the Achaemenid army. After several Persian defeats, the deci-
sive battle was fought on 1 October 331 B.C. at Gaugamela in Syria. The
Persians were completely defeated and could n o longer offer any systematic
resistance to the Macedonian army. A year later, the Achaemenid Empire came
to an end.

The Achaemenid economy

The Achaemenid Empire was marked by widely differing social and economic
structures. It included Asia Minor, Elam, Babylonia, Syria, Phoenicia and
Egypt, countries that, long before, had their own institutions of state. But
together with these economically advanced countries, the Persians had subdued
the Massagetae and other tribal peoples.
T o administer such heterogeneous territories Darius I embarked on his
renowned administrative and financial reforms around 518 B.C. H e created a
stable system of state government for the conquered countries and systematized
tax collection. This led, inter alia, to the establishment of a new administrative
system that underwent little change until the end of the empire. But even after
 Mrdia and Achaemrnid Iran

Darius' reforms, each satrapy remained essentially autonomous in social and

econon~icmatters, endowed with its own social institutions and internal struc-
ture, and preserving its old local laws and traditions.
For administration and taxation Darius divided his empire into twenty
regions known as satrapies, each governed by a satrap. This title had existed
under Cyrus and Cambyses, but at that time both civil and military functions
were combined in the hands of the same person, the satrap. Darius introduced a
sharp distinction between the functions of the military commander and those of
the satrap, w h o became purely a civil governor, responsible within his province
for administration, justice, the economy, taxation and the supervision of offi-
cials. The army, conversely, was subject t o military commanders who were
independent of the satraps and subordinate directly to the king. After the death
of Darius I, however, the sharp demarcation between military and civilian func-
tions was not strictly observed.
The larger satrapies might also include countries that enjoyed internal
autonomy. This was particularly true of the distant provinces in whose internal
affairs the Persian administration rarely interfered, governing them through
local princes and confining itself t o the receipt of taxes.
T o implement these new reforms a large central civil service was estab-
lished with an imperial chancellery. The central state administration was in Susa,
the administrative capital of the empire. The imperial court spent the autumn
and winter in Babylon, the summer in Ecbatana and the spring in Susa, while
during the major festivals it met in Persepolis, Pasargadae o r Susa. The satraps
and military commanders were closely linked with the central civil service and
were under the constant surveillance of the king and his functionaries. Everyone
in the centre and the provinces was watched by police officials known as 'the
king's ears and eyes' who were independent of the satraps and other local auth-
orities, and reported directly t o the king on any seditious words o r deeds.

Old Persian inscriptions

The Persian tribes that inhabited the south-west of the Iranian plateau during
the first millennium B.C. spoke different dialects of Old Persian belonging to the
Iranian branch of the Indo-Iranian o r Aryan languages. In addition t o Old Per-
sian, the O l d Iranian language group included Median (of which only isolated
glosses have come down t o us), Avestan (which has left a substantial body of lit-
erature), Parthian, Sogdian and Scythian. The cuneiform inscriptions of the
Achaemenid kings were written in O l d Persian. Old Persian cuneiform repre-
sented a huge step forward in the development of writing. Unlike Assyrian-
Babylonian script, which used more than 600 signs, it consisted of only thirty-
six syllabic signs and eight logograms (i.e. signs denoting individual words such
as 'king', 'god', etc.) and hence was logo-syllabic. The idea for it probably came
from Aramaic writing, which consisted of twenty-two silnply formed signs,
It adopted many of the featurcs of Aramaic script but took the shape of its
signs from Assyrian-Babylonian cuneiform, probably via the Elamites or
Urartians.
It is still difficult to say with any certainty when O l d Persian cuneiform
arose. Most scholars date its invention to the reign of Cyrus 11, but others such
as W. Hinz date it to the reign of Darius I. Struve and D'yakonov have sug-
gested that it was invented in Media in the pre-Achaemenid period under the
influence of Urartian writing, and hence that the Persians received it in an
already finished form. According to these scholars the style of the Achaemenid
inscriptions took shape under the influence of the Urartian annalists via the
Median inscriptions.
The hypothetical Median origin of O l d Persian cuneiform cannot be
proved, however, so long as n o Median inscriptions in cuneiform script have
been found. That writing existed in the Median state is hard to doubt, but noth-
ing is actually known about it. Furthermore, the possibility cannot be ruled out
that another foreign-language script existed in Media, as occurred in many Near
Eastern countries in ancient times.
About 200 Old Persian inscriptions are known so far. Many of them are
accompanied by Elamite and Akkadian translations and some also by an Egyp-
tian translation. This was to some extent a mark of respect for the historical tra-
dition represented by languages that had been used for writing for several thou-
sand years prior t o the rise of the Achaemenid Empire. The inscriptions were
displayed o n major trade routes, royal tombs, palace walls and pillars, o r carved
on metal tableware, weapons, stone vases and seals. Some have been discovered
in the foundations of palaces, where they had been placed as foundation plates.
The majority of O l d Persian inscriptions have been found in Persia, Elam and
Media. The most famous inscription - the vast Bisutun relief (Fig. 2) - records
the stormy closing years of the reign of Cambyses I1 and the early years of the
reign of Darius I (c. 522-519 B.c.). It is written in Old Persian, Elamite and
Akkadian, and the content is virtually the same in all three versions. It is located
30 km east of Kermanshah on the ancient caravan route between Babylon and
Ecbatana, the Median capital. It is carved on a sheer rock face about 105 m from
the ground, and its size makes it strikingly visible from the road that passes
beneath. It is 7.8 m high by 22 m wide overall. It contains over 1,000 lines, each
on average 2 m long.
The Bisutun inscription was translated into many other languages and
sent out to all the satrapies of the Achaemenid Empire, as the inscription itself
records. At the beginning of the present century, poorly preserved papyrus doc-
uments with an Aramaic translation of the Bisutun inscription were found dur-
ing archaeological excavations on the island of Elephantine in southern Egypt -
a text intended for dissemination in the western part of the empire. In 1899 a
fragment of a stone block with part of an Akkadian version of the Bisutun
 M ~ d i aand Achaemenid Iran

inscription was discovcrcd at Babylon in the ruins of a royal palaw. 'I'hc

inscription consists of an introduction sctting out the genealogy of IIarius I, a
historical part proper rccounting events, and a conclusion. The cxact datcs and
placcs of battles are indicated, and also, in the Akkadian and Aramaic versions,
the number of Darius' enemies killcd or taken prisoner. It may bc concluded
from this that the accounts of the main battles were compiled immediately after
the battles had taken place, indicating unquestionably the authcnticity of much
of the information provided.
Above the inscription there is a rclief 3 m high by 5.48 m wide, depicting
the victory of Darius I over the peoples of the Acliaemcnid Empirc and their
leaders who had risen in rebellion in 522-521 u.c:. With his lcft hand the
supreme god of the Persians, Ahura Mazda, holds out a ring to Darius, symbo-
lizing his investiture with royal power, and blesses him with his raiscd right
hand. Darius is depicted life-size (1.72 m) (Fig. 3). His right hand is raised to
Ahura Mazda in a gesture of prayer; in his left hand he holds a bow; and with
his left foot he crushes Gaumata, who briefly seized the Achaemenid throne. T o
the left, behind Darius, two of his courtiers can be seen - Gobryas his spear-
bearer and Aspathines his bow-bearer. They arc smaller than Darius (1.47 m)
but taller than the rebel leaders, who hardly come up to Darius' chest ( I . 17 m).
Directly behind Gaumata are shown the eight usurping impostors and the
leader of the Saki Tigvaxaudi. Their hands are tied behind their backs, and they
are chained together by a single long chain.
The other major inscriptions of Darius I are t o be found at Naqsh-i Rus-
tam, a few kilometres north of Persepolis. At the entrance to the Achaemenid
royal sepulchres hewn out of the rock are two trilingual inscriptions. O n e con-
tains the royal genealogy and a list of the countries under Persian rule; the other
sets out the legal and ethical principles framing Darius' rule. There is also a
relief depicting Darius. Gold and silver foundation plates with inscriptions of
Darius I have been excavated in Persepolis. Many examples of such inscriptions
have been found on palace buildings in Susa, some written on marble, others on
clay tablets and bricks. A statue of Darius nearly 3 m high but with the head
missing, bearing an inscription in Old Persian, Elamite, Akkadian and a particu-
larly detailed Egyptian hieroglyphic text, has also been found in Susa (Fig. 4).
Stone inscriptions of Xerxes have been found in Persepolis and Pasargadae.
Among these tlie Daiva inscription relating his efforts to ban the worship of
false divinities (the daivas) is tilost important.
Among the Achaemenid inscriptions found in Egypt mention should be
made of the three stele of Darius I bearing inscriptions about the construction
of a Suez canal, written in Old Persian, Ela~nitc,Akkadian and Egyptian.
The decline of O l d Persian cuneiform can already be seen under Xerxes'
successors. Although some inscriptions from tlie Late Achacinenid ~ e r i o dhave
survived, oilly a fcw are of real historical value. More than tcn gold and silver
vessels bearing Old Persia11 inscl-iptions o r the usual trilingual inscriptions are
FIG.3. Darius I on the Bisutun relief. (From Heinz Luschey,
Archiologische Mitteilungen aus Iran, N.F. 1, 1968.)
known. Beside uninscribed specimens (Fig. 5) on a series of Achaemenid royal
seals (Fig. 6) cuneiform inscriptions have also survived.
The official written language of the empire was Aramaic, used for corn-
munication between chancelleries throughout the state. Official documents
written in Aramaic were sent out from Susa to all corners of the empire. On
receiving them the local scribes, who knew two o r more languages, translated
them into the native language of the governors. In addition to Aramaic, which
was common to the entire state, the different countries used local languages for
drafting official documents.

FIG.5. Cylinder seal of Artaxerxes I1 (?). (Photo: The Hermitage, St Petersburg.)

State administration

T o help run the satrapies, there was a regular postal service. O n the major high-
ways there were state-protected relay stations and inns at intervals of a day's
march; and on important passes there were strongly garrisoned watch-towers.
Thus the road from Sardis t o Susa, some 2,470 km in length, had 111 relay sta-
tions. By changing mounts and couriers, up to 300 km could be covered in a
day, and the entire journey from Sardis t o Susa could be done in seven days.
Elamite texts from Persepolis, written in the late sixth century B.c., pro-
vide a wealth of information about the delivery of state mail to the various
satrapies. Extant documents include official letters, reports by senior officials to
each other or the king and the king's instructions. Reports addressed to the
monarch were usually sent t o Susa and were probably destined for the imperial
chancellery. From Susa, couriers bearing royal orders were sent out to virtually
 FIG.6. Achaemenid cylinder seal with the name of Darius I

every satrapy. The regular delivery of state instructions required a considerable

body of professional couriers who were maintained entirely at state expense. At
the relay stations there were royal stores from which food was provided for the
couriers and others travelling on official business. In those far-off days the pos-
tal service was used only for official mail; private letters were sent either by
some chance expedient o r by private messenger.

The economy
Under Cyrus and Cambyses there was as yet no properly established taxation
system based on the economic potential of the countries making up the empire.
About 518 B.c., Darius introduced a new system. All satrapies were obliged to
pay money taxes in silver, the amount of which was strictlv fixed for each
satrapy and determined on the basis of the area of cultivated land and its fertil-
ity as calculated through the mean annual yield. Herodotus provides a detailed
list of the taxes paid by tlie satrapies. Thus Sattagydia, Gandhira and Arachosia,
which formed a single province for taxation purposes, paid 170 talents of silver
(1 talent = 30 kg), Bactria 300 talents, the Sakas 250 talents, while Parthia. Cho-
rasmia, Sogdiana and Haraiva paid 300 talents.
Darius I introduced a standard inonetarv unit throughout tlie enipirc -
the gold daric weighing 8.42 g (Fig. 7), which forrned the basis of the Achaeme-
nid monetary system. The minting of gold coins was a prerogative of the Per-
sian king. The usual rnediuin of commerce nras the sil\rer shekel, 5.6 g in n-eight,
 M.A. Dnndamayev

FIG.7. Achaemenid gold daric. (Photo: The Hermitage, St Petersburg.)

with some 95 per cent pure silver. It was minted chiefly in the Asia Minor satra-
pies in the king's name. Silver and smaller copper coins of various values were
also struck by the autonomous cities, the dependent princes and the satraps.
Minted Persian coins were little used outside Asia Minor; the usual medium of
trade was unminted silver ingots, with Persian coinage playing only a secondary
role. This explains why the hoard of silver coins found in Kabul in 1933, which
proves that minted coinage was used in Afghanistan (it was buried in roughly
380 B.c.), contains only eight minted Persian shekels. At the same time it con-
tains worn Greek coins from virtually every part of Greece and every period,
from archaic square, stamped ingots to staters and tetradrachms. Coins were
first introduced into Central Asia during the Achaemenid period. Darics and
other Persian coins have been found there, but there is no reason t o believe that
they were common. Precious metal, which belonged to the state, was minted at
the king's discretion and most remained unminted. Thus the proceeds of taxa-
tion were stored for decades in the imperial treasuries and removed from circu-
lation.
The relative political calm throughout Western Asia under the Achaeme-
nids, together with the availability of good sea and land routes, promoted the
development of international trade o n an unprecedented scale. Another impor-
tant factor in the flourishing of commerce was the expedition by Scylax of
Caryanda in Asia Minor, whom Darius I (c. 518 B.c.) ordered to explore the
possibility of opening sea links between India (i.e. modern Pakistan) and other
countries of the empire. Scylax's vessels sailed down the Indus to the ocean,
along the southern shores of Iran and, rounding Arabia, reached the Red Sea
coast in 30 months.
In Achaemenid times there were many major caravan routes. Particular
importance was attached to the road which, crossing the Zagros mountains,
linked Babylon with Ecbatana and ran o n to Bactria and the borders of India.
Iran was linked with the Indus valley by a road through Makran. A further
aspect in the developement of commercial links was the differing natural and
climatic conditions of the countries making u p the Achaemenid Empire. From
India gold, ivory and incense were imported; from Sogdiana and Bactria lazurite
 Media a n d A c h a e m e n u i Iran

and carnelian were taken to Westcrn Asia; and from Chorasmia, turquoise.
Judging by the Achaemenid art products found in Sarmatian tumuli from the
end of the fifth century B.C. near Orsk in thc Urals - including a trilingual
inscription of the Persian ruler Artaxemes I - the nomads of thc southern Urals
maintained commercial contacts with the Central Asian satrapics; Ccntral Asia
has even yielded artefacts made by Greek craftsmen from Naukratis in thc Nile
delta. Further evidence of Iran's commercial links with Ccntral Asia and thc
lands t o the north-east has come from excavations of fifth-century-n.c.. tumuli
in the Altai, where artefacts preserved in the permafrost include a trimrncd pile
carpet, apparently of Median or Persian origin.

Iranian culture in the Achaemenid period

Persian conquests and the fact that the empire united dozens of peoples helped
its subjects to broaden their intellectual and geographical horizons. The Achae-
menid period was one of intensive ethnic mingling and syncretism in cultures
and beliefs. The prime reason was that contacts between different parts of the
empire had become more regular than in the previous period. More specifically,
the sources report frequent visits by state functionaries from Arachosia,
Haraiva, Gandhira, Bactria and other eastern Iranian or Central Asian countries
to Susa and Persepolis.
Iran, which had since time immemorial acted as an intermediary in East-
West cultural exchange, maintained its historical role under the Achaemenids.
A t the same time, the Iranians created their own original and sophisticated civil-
ization. O n e of its achievements was the adaptation of the cuneiform script for
writing Old Persian (see above). The chief official written language was Ara-
maic; under the Achaemenids, standard formulae were devised t o render Ara-
maic terms and clerical expressions into the different Iranian languages; and
from the official written Aramaic of the Achaemenids, the later written forms of
Parthian, Middle Persian, Sogdian and Chorasmian were derived. It was in this
period that the peoples of Central Asia first became acquainted with Aramaic
script. This, too, was the period when a number of O l d Iranian words - chiefly
socio-economic, military and administrative terms - were borrowed by Indian
languages.
Among the outstanding achievements of Old Iranian civilization was
Achaemenid art, which is known above all from the monuments of Pasargadae,
Persepolis and Susa, the Bisutun rock reliefs, the Persian royal tombs at Naqsh-i
Rustam, and from large quantities of metal and stone carvings. The subjects
may be military triumphs o r hunting exploits by Persian kings and warriors,
combat between heroic monarchs and various monsters symbolizing evil, or
palace and religious rituals. It was the characteristics of this art that took shape
at the turn of the sixth and fifth centuries B.C.
 M. A. Dandamayev

F I G . 8. S a k i Tigraxaudi at Persepolis.
(Photo: Deutsches Archaologisches Institut, Abteilung TeheranIWalser.)

Persepolis impresses through the size of its platform, the height of the
columns, the reliefs in the apadana o r grand hall (Figs. 8-15). The canons laid
down under Darius I were in no way violated in the later palaces at Persepolis,
the architectural decoration of the rock sepulchres of the Persian kings or the
carved metalwork of the fifth to fourth centuries, though new motifs and
images were added. Persepolis was the home of the imperial Achaemenid style
which was to symbolize the might and grandeur of the kingship and that sub-
sequently spread far afield, creating a form of cultural unity from the Indus to
the coasts of Asia Minor. Metalwork and particularly rhytons, made by craft-
smen from Media, Asia Minor o r eastern Iran, are canonical in form, decoration
and even dimensions, regardless of their geographical origins (Figs. 16-21).
Analysis of Achaemcnid art reveals the influence of the Egyptian hypo-
style hall o r echoes of Ionia in the design of the columns, while Urartian building
techniques are plain to see in the huge, man-made platforms. But the art itself is
far from the sum of its borrowed components, as the borrowed forms rapidly
lose their original qualities. In other words, while the details of a given image or
 Frt-;, 9. Bacrrians at: Persepalis.
(Photo: Dewtsches Arxthaologiscl~eslnstitut, Abteilung TcI~eran/T~-inpelmann.)

structure may be known from previous eras or other countries, the image itself
is completely new and specifically Achaemenid. All the material aspects of the
art remain essentially original, and it is individualistic, the result of specific his-
torical circumstances, a particular ideology and social life which imparted new
functions and significance to the forms borrowed.I4

Old Iranian religion

Zoroastrianism, the religion founded by Zoroaster, arose in eastern Iran in the
seventh century B.C. It may confidently be stated that Zoroaster lived before the
Persians conquered Central Asia. Achaemenid rule is known to have had a pro-
found impact o n all the peoples of the Old Persian Empire. Achaemenid admi-
nistrative and cultural terms were adopted in their languages. But the Avestn,

14. Nylander, 1970, pp. 144 et seq.

 M. A. Dandnrnayev

FIG, 10. Bactrian camel at Persepolis.

(Photo: Deutsches Archiiologisches Institut, Abteilung Teheran/Walser.)

the Zoroastrian holy book, bears no trace of Achaemenid terminology, nor is

there any mention of the Achaemenid money, taxation system o r kings. The
Avesta is a composite work. Its earliest parts, the GaOiis (gdOZ = song), differ in
form and content from the rest of the baok, They were written in verse in an
archaic dialect, being the sermons of Zoroaster himself. H e urged his listeners to
protect their livestock from the marauding forays of the nomad tribes, opposed
the predatory killing of stock and sanguinary mass sacrifices, and instructed
every believer to rear and defend useful animals.
The greater part of the book is what is known as the Younger Avesta. Its
core appears to have been written in the last quarter of the fifth century LC.,
and much of it belongs to the still later Arsacid period. Because of its long
development Zoroastrianism underwent a complex evolution. According to the
GaBds, Zoroaster received from the god Ahura Mazda a mission to renew reli-
gion and break with ancient beliefs. H c introduced a radical religious reform,
accentuating belief in the final victory of Ahura Mazda, rejecting some of the
dacvas or tribal gods and setting the others beIow Ahura Mazda. He taught that
Ahura Mazda (Ormazd in Middle Persian) is the sole, omnipotent and ubiqui-
 Media a n d Achaemcnid Iran

FIG. 11. Detail of a relief at Perscpolis.

(Photo: Dmtsches ArchZ~logischesInstitut, Abteilung Teheran.]

tous god of good and the incarnation of light, life and truth. H e existed before
the world and is its creator. From the outset, however, together with Ahura
Mazda there existed the evil spirit, Angra Mainyu ( A ~ r Mainyuf)
o or Ahriman,
who incarnates darkness and death, and with his datva helpmates, works evil.
Ahura Mazda struggles constantly with Angra Mainyu, relying in that
combat o n his assistants who incarnate good thought, truth and immortality,
the triad of the Zoroastrian ethic. Man was created by Ahura Mazda but is free
to choose good or evil, and is consequently open t o the influence of evil spirits.
By his thoughts, words and deeds man must resist Angra Mainyu and his adhe-
rents, the spirits of evil.
The Zoroastrian priests created a complex eschatology, according t o
which the world would last 12,000 years. The first 3,000 had been the 'golden
age' which knew n o cold, heat, sickness, death or ageing. The earth had been
full of sheep, goats and cattle. That was the period of Ahura Mazda's reign.
Then the 'golden age' came to an end, and Angra Mainyu had created hunger,
sickness and death. But a saviour o r sao;yant of Zoroaster's kin would come t o
the world, and at the end good would triumph over evil and the ideal kingdom
would arise, in which Ahura Mazda would hold undivided sway over heaven
and earth, the sun would shine for ever and all evil would vanish.
Some time after its birth Zoroastrianism began t o spread to Media, Persia
and other countries of the Iranian world. But in Persia it began t o take hold
only towards the end of the sixth century B.c., and the Achaeinenid kings, while
appreciating the advantages of Zoroaster's teachings as a new established
 FIG.12. Arachosians at Persepolis.
(Photo: Deutsches Archaologisches Institut, Abteilung TeheradTrinpelmann.)

FIG.13. Gandhiirans at Persepolis.

(Photo: Deutsches ArchSologisches Institut, Abteilung Tcheran/Trinpelrnann.)
 M . A. Dandamaycv

religion, nevertheless did not reject the cults of the ancient tribal gods. Zoroas-
trianism had not at that tiine become a dogmatic faith with rigid standards, and,
naturally, various modifications of the new religion appeared. With this in
mind, Achaemenid religion of the tiine of Darius I may be said to have been a
form of early Zoroastrianism.
The Achaeinenids none the less worshipped Egyptian, Babylonian, Greek
and other alien gods. In the temples of those gods sacrifices were made in the
name of the Persian kings, who wished t o attract the benevolence of the local
deities. This was due not only to political considerations, but above all to the
fact that the ancient religions were not dogmatic o r intolerant towards the
beliefs of other ~ e o p l e s .

FIG.15. A Persian guardsman at Persepolis. (Photo: T h e I-Icrmitagc, St Petersburg.)

 Media and Achacmenid Iran

FIG. 16. Silver rl~yton,fifth t o fourth century B.C.

(Photo: The Hermitage, St Petersburg.)

FIG.17. Horseman in gold. (Photo: The Hermitage, St Petersburg.)

FIG.18. Gold-handled bowl. (Photo: T h e Hermitage, St Pctersburg.)

FIG. 19. Handle of a silver vessel, Achaelnenid period.

(Photo: The Hermitage, St Petersburg.)
FIG.20. G o l d rquarc ncckl,lcc button with Ahura M a ~ d a .
(Photo: T h c H c r ~ n i t a g c St
, I'rtcr\I~urg.)

FIG.21. Seal-ring wit11 winged horned lion.

(Photo: Thc Hermitage, St Petersburg.)
 A. H. Dani and P. Bernard

Part One

(A. H. Dani)

Alexander advanced into Central Asia in the follow-up operations against the
Achaemenid monarch, Darius I11 Codomannus, whom he had defeated in three
successive battles at Granicus (334 B.c.), Issus (333 B.c.) and Gaugamela (331
B.c.). This had fired his imagination to pursue the retreating monarch and to put
an end t o Achaemenid power by crushing the remaining source of its strength
in Central Asia. The eastern forces of the Achaemenids are described in Arrian's
Anabasis (111.8.3-7):
The Indians w h o were coterminous with the Bactrians, as also the Bactrians them-
selves and the Sogdians, had come to the aid of Darius, all being under the com-
mand of Bessus, the satrap of the land of Bactria. They were followed by the Sakas,
a Scythian tribe belonging to the Scythians w h o dwell in Asia. These were not sub-
ject to Bessus but were in alliance with Darius. . . . Barsaentes, the satrap of Ara-
chosia, led the Arachosians and the men who were called Mountaineer Indians. . . .
There were a few elephants, about fifteen in number, belonging to the Indians who
live this side of the Indus. With these forces Darius had encamped at Gaugamela
near the River Bumelus, about 600 stades from the city of Arbela.

The elephants probably belonged to Porus, the ruler of Jhelum region, and
among the 'Mountaineer Indians' was possibly the local chief Sisicottus who is
known t o have helped Bessus.'
The whole of Central Asia opposed Alexander and resisted his march at
every stage. The Achaeinenids had built a strong empire and much of Central
Asia shared their cultural heritage for some two centuries. It was in defence of
this heritage that they rallied against the invaders with courage and strength.
Alexander's father Philip had advanced from his native Macedonia, to establish
his supremacy over Greece, and had then brought the Greeks of Asia Minor

:' See Map 2.

1. Dani, 1980, pp. 11 7-30.
 A. H. Dani and P. Rernnrd

under his control. It fell to the good fortune of his son Alexander to win his
first great battle at Granicus, which enabled him t o possess the Mediterranean
coastal region of Asia. But it was only after his subsequent success at Issus that
he could properly measure his growing strength against that of the Achaemenid
monarch.

In reply to a letter from Darius, Alexander had declared his political manifesto:

Your ancestors invaded Macedonia and the rest of Greece, and without provoca-
tion inflicted wrongs upon us. I was appointed leader of the Greeks, and crossed
over into Asia to avenge these wrongs; for you were the first aggressors.

This motivation of revenge has been construed by some historians2 as being a

'crusade' against the Persians, but Ghirshman3 has modified this view. Alexan-
der's failure t o capture Darius stirred him t o pursue the Achaemenid king and
crush his power completely. Military victory was not enough. The change in ~ o l -
icy adopted by Alexander t o ~ a c i f yPersia has been explained by Frye4 and Tarn5
as the 'fusion of Greeks and Persians, o r better, it should be said, Hellenes and
Iranians'. This is far from the an-Hellenic ideal' which modern historians had
attributed to Alexander. It was this double approach, political and military, that
dragged the war into the heart of Asia. If Alexander was t o become King of
Greece and Persia, he must destroy, root and branch, all the sources of Achaeme-
nid power, and establish a series of strong garrisons right up t o the Oxus and Jax-
artes. It was in the pursuit of this aim that he won over t o his side Sisicottus and
welcomed Ophis, ruler of Taxila, who held out the prospect of conquering the
Indus region. Alexander was thus lured t o the farthest reaches of Achaemenid
territory. While he succeeded in destroying Persian power, his death in 323 B.C.
put an end t o the dream of a Hellenic empire. The struggle in Central Asia roused
the dormant spirit of its people. While they were willing t o benefit from cultural
contacts, they soon threw off the political yoke and absorbed the Greek popu-
lation that Alexander had left behind in the garrison cities of his new empire.

Alexander's campaigns can be divided into several phases. His first object was
to pursue the retreating Achaemenid monarch. The second phase was to meet

2. Olmstead, 1948, pp. 495 et seq

3. Ghirshman, 1954, p. 208.
4. Frye, 1963, p. 127.
5. Tarn, 1951, p. 137.
 Alexander and his successors in Central Asla

the challenge of the Arians, Arachosians, Bactrians and Sogdians who formcd
a united opposition and continued t o fight even when Bessus was captured
and killed. The third phase was the new game of diplomacy and war against
the local chiefs of the Indus region, and finally his retreat through the descrt
of Baluchistan to Susa and Babylon involving the reappraisal of his long, ardu-
ous campaigns. The Greek historians (Arrian VI.1 et seq.) havc created a myth
about his return march from the River Beas and havc not givcn a proper
assessment of the situation when Alexander's Grcek appointees were cithcr
killed o r illegally usurped power, while Alexander was busy with his Indus
campaigns o r on his retreat. It was no surprise that when Alexandcr unexpcct-
edly died, his faithful commanders fished for power in troubled waters. The
Indus region fell to the rise of the Mauryans and later when Seleucus Nicator
tried to recover this lost territory, he had further to cede to Candragupta
Maurya the provinces of Aria, Arachosia, Gedrosia and the Paropamisadae.
Before sixty years had passed, Parthia, Hyrcania and Bactria became indcpen-
dent and rejected the Seleucid hegemony. A fresh power struggle began in
Central Asia; a new era of adjustment and cultural assimilation was inaug-
urated.

THE M U R D E R O F DARIUS

In the first phase, Darius tried t o remuster his forces at Ecbatana (modern
Hamadan) and was joined there by Bessus of Bactria, Barsaentes of Arachosia,
Satibarzanes of Aria, Nabazarnes, Artabazus and many others including his
Greek mercenaries. But Alexander's advance was too swift for Darius to reor-
ganize the support of his eastern provinces. H e chose to retreat and paused to
fight at the Caspian gates. Here history took a new turn. Alexander usurped the
title of the Great King and called himself 'Lord of Asia', disregarding the fact
that the Achaemenid monarch was still alive - a signal of great historic change.
Although the Greek historians give no details except for the final disowning
and, later, the murder of Darius by his eastern commanders (the crime being
attributed to Barsaentes and Satibarzanes), the split between the Greek mercen-
aries of the Achaemenid monarch and the eastern commanders can be seen
clearly - a division that may have been brought about by Alexander's own
diplomatic moves. The result was the return of the Greek mercenaries and a
switch in their loyalty. The eastern commanders, on the other hand, defended
their national home, which lay entirely in Central Asia. Bessus emerged as the
great leader and retreated to his home province of Bactria to meet Alexander's
challenge. Satibarzanes and Barsaentes returned to their provinces of Aria and
Arachosia, respectively, t o make defence preparations. Meanwhile Alexander
consolidated his gains in the Caspian region. Hyrcania lay at his feet and the
Greek deserters submitted to him with the Inen and material they commanded.
 A. H. Dani and P. Bernard

H e made Zadracarta (modern Astrabad) the headquarters of Hyrcania, the

strong centre and spearhead for his conquest of the East.

T H E F A L L OF: A R I A A N D A R A C H O S I A

The defence of the East seems to have been well planned. Alexander6 began to
pursue Bessus who, with the support of his people, had assumed the upright
tiara and made himself known as 'Great King'. Satibarzanes feigned submission
to Alexander and accepted his general Anaxippus and a Macedonian military
garrison in Aria. But when the latter arrived, Satibarzanes, with the full support
of his people, killed the Greek general along with his whole force. The war of
liberation had now begun. Both Aria and Arachosia were u p in arms in alliance
with Bactria and probably with Sogdiana. Alexander had t o forgo his ~ l a nof
advance on Bactria and turned back to deal with the new situation. We only
read about his victorious march t o Artacoana, through Aria and further south
to Drangiana. But he could not capture the local chieftains. Satibarzanes is said
to have gone to Bessus, while Barsaentes left Drangiana and escaped to the
'Indians' in eastern Arachosia, bringing in the new support of the 'Mountaineer
Indians'. Here Alexander further changed his policy. H e appointed Arsames, a
Persian, as satrap of Aria; but while he advanced into Arachosia, this newly
appointed satrap proved t o be in league with Satibarzanes, and led Aria in a
further revolt. Alexander was able t o defeat and kill them, putting a new satrap,
Stasanor, in charge of Aria. Alexander's advance into Arachosia cannot be
explained unless his immediate objective was t o capture Barsaentes, who was
ultimately caught and put to death. N o details are available about the name of
the local chief o r the resistance that Alexander met; but we d o know that he was
forced t o march round the central massif of Afghanistan, through the Helmand
and Arghandab valleys and over the high range of the Hindu Kush simply to
reach Bactria - a circuitous journey which was hardly necessary unless dictated
by some political o r strategic reasons that remain unknown t o us. W e d o how-
ever know the measures he adopted to control these provinces. H e founded or
refounded cities, peopled with a Macedonian and Greek population and strong
garrisons, all named after himself - Alexandria in Aria (Herat), Alexandria
Prophthasia in Drangiana (Phrada), Alexandropolis (Kandahar), Alexandria in
Arachosia (Ghazni) and Alexandria ad Caucasum (now identified with ancient
KapiSa, modern Begram, near Charikar). This chain of posts, prrisoned by
Alexander's own troops, were meant to safeguard the route and cover the rear
of his advancing army.

6. Tarn, 1951, pp. 61 et seq.

 Alexander and his successors in Central Arla

'I'IiE <:AI'TUHI: 0 1 ' UESSUS

Bessus enjoyed quite a strong position in Bactria, as hc had bccn ablc to gain thc
support of Oxyartes and Spitamenes (the two great chiefs of Sogdiana) and of
Satibarzanes who had fomented the revolts in Aria. Alexandcr decided to makc
a strategic move and take Bessus by surprise. But Bessus was not prcparcd to
give open battle, and withdrew to the other side of thc Oxus. Alcxandcr
appointed the veteran Artabazus as satrap of Bactria and marched to the Oxus
with the intention of crossing the river, but he found that Bessus had dcstroycd
all the boats, followed a scorched-earth policy and joined up with Spitamcncs.
As Alexander advanced, Spitamenes retrcated towards Bukhara, but Bessus
stood his ground and was eventually captured.

R E S I S T A N C E I N B A C T R I A A N D I T S SUI'I'RESS1C)N

In Sogdiana the local chiefs decided to follow a wait-and-strike tactic, and to

employ desert manoeuvres. They gained the support of nomads beyond the
Jaxartes, and Spitamenes 'found allies in the nomads of the Kyrgyz stcppe, west
of the Polytimetus river - part of the great Saka confederacy known as the Mas-
~ a g e t a e ' .Initially
~ Alexander occupied Maracanda (Samarkand), the royal sum-
mer residence of Sogdiana. Then, worried about the Saka hordes beyond the
Jaxartes, he advanced northward past the fortress of Cyropolis occupying seven
fortresses on the way t o the Jaxartes, the boundary of Achaemenid territory. It
was in these engagements that he was wounded in the tibia and lost part of the
bone. This fitted in well with the tactics of the local chiefs, which only became
known to Alexander when the whole country behind him had risen in revolt.
His garrisons in Cyropolis and the seven fortresses were all massacred, and he
had t o turn back to reconquer this territory. Meanwhile Spitamenes, who had
retired into the desert, besieged the citadel of Maracanda, and the Saka nomads,
who were allies of Spitamenes, swarmed round the Jaxartes. This was the first
time that Alexander was faced with a new encircling tactic by the Sakas. H e es-
tablished a strong garrison on the Jaxartes by founding Alexandria Eschate
(Alexandria the Farthest), now identified with Khojand. H e then crossed the
river and broke through the encircling Sakas with the help of his archers and
cavalry. Alexander had sent his commander Pharnuches, a Lycian, to relieve
Maracanda. Spitamenes withdrew down-river and lured the Macedonian force
after him t o the edge of the desert. This gave Spitamenes his opportunity to
make a frontal attack and driving the whole Macedonian force back to the river
he annihilated it and again besieged Maracanda. Alexander had to rush perso-
nally to relieve the town and although he was able to destroy the hawest of the
countryside, Spitamenes' desert tactics roved more than a match for him.

7. Ibid., p. 69.
 A. H. Dani and P. Bernard

Alexander was obliged to retreat to Bactria while Spitamenes celebrated his "ic-
tory at Bukhara, the royal winter residence of Sogdiana. Alexander consolidated
his position in Bactria and received large reinforcements from his home country
and from the satraps he had appointed in the western provinces. H e was joined
by Pharasmanes, the ruler of Chorasmia south of the Aral Sea, who was rob-
ably won over because of his opposition to Spitamenes, and by Sisicottus, an
older friend of Bessus. It was here also that Ophis, the ruler of Taxila, jealous of
his powerful neighbour, Porus, came to offer his alliance, opening up before
Alexander the rosy picture of the conquest of the Indus valley. Alexander
'assumed the state of Great King, surrounded himself with eastern forms and
pomp, exacted self-abasement in his presence from oriental subjects, and
adopted the maxim that the king's person was divine. H e was the successor of
Dariu~.'~

THE C O N Q U E S T O F TRANSOXANIA

The territory north of the Oxus had yet to be conquered. Spitamenes was a
strong force in Sogdiana, and four other chiefs - Oxyartes, Chorienes, Catanes
and Austanes - were in arms in the Paraetacene (modern Hissar) hills. Alexan-
der himself advanced, dividing the army into five columns, which swept across
the plains and reunited at Maracanda. While Alexander was building fortified
garrisons at various points, Spitamenes, in league with the Sakas, overwhelmed a
Bactrian border post and appeared before Bactra itself. In the winter of 328 B.c.,
Alexander put Coenus in charge of western Sogdiana with two battalions of the
phalanx, two squadrons of the Companions (his personal bodyguard), and the
newly raised Bactrian and Sogdian horse. Spitamenes, helped by the Massagetae,
attacked him but by now Coenus had mastered his tactics and was able to over-
power and defeat Spitamenes. We d o not know what diplomatic moves fol-
lowed, but we read of the estrangement of the Sogdians from Spitamenes and
their surrender to Alexander. Later the Massagetae lost heart, cut off Spita-
menes' head and sent it to Alexander. Thus was the end of the great defender of
Sogdiana. Alexander arranged for Spitamenes' daughter Apama to be married to
Seleucus Nikator, and she became the mother of Antiochus I. Alexander was
not yet master of the whole of Sogdiana. While he held the plains, the great
chiefs were strong in the hills. Late in 328 Alexander advanced to Oxyartes'
stronghold, 'the Sogdian rock' near Derbend, which was very strongly
defended. Oxyartes was not present. While we read of Alcxander's assault on
the rock and its surrender, we have no details of Oxyartes' reconciliation except
that his captured daughter Roxane was marricd to Alexander. The way in which
Oxyartes was made to accompany Alexander to the siege of other strongholds
suggests that some historical facts have not been recorded and a deliberate tradi-

8. Tarn, 1951, p. 77.

 Alexander and his successors in C,'cntral Asia

tion was established that Alexander had fallen in love with Roxanc. I.ater whcn
we find Oxyartes sccuring thc surrender of Chorienes, who had a strong fort on
the Vakhsh river south of Faizabad, his political role should be clearly under-
stood. As we note later that the same Oxyartes was made satrap of the Paropa-
misadae, the political trend becomes clear. Alexandcr did not himself advance to
subdue the two remaining chiefs of the hills but entrusted the task to Cratcrus
who was successful in his mission.

It was in Bactria that Alexander planned to conquer the Indus provinces of the
Achaemenid E m ~ i r e Three
.~ local chiefs had their own rcasons for supporting
him. O n e of these, Sisicottus, came from Swat, and was later rewarded by an
appointment in this locality. Sangaeus from Gandhira had a grudge against his
brother Astis, and t o improve his own chances of royalty, sided with Alexander.
The ruler of Taxila wanted t o satisfy his own grudge against Porus. In this way
Alexander's new push towards the Indus was preceded by considerable diplo-
matic activity, of which very little is known. Whether his original intention was
to explore the southern sea is difficult to say, but his advancc in that direction
and the information he received from local chiefs must have increased his
curiosity. However, to say that mere curiosity brought him t o this part of the
world would be wide of the mark. Certainly such curiosity cost him dear, and
the assistance of the local chieftains was of no great consequence from a military
point of view.

In the early summer of 327 B.C. Alexander started from Bactra and found to his
surprise his city of Alexandria ad Caucasum in some disorder. Arrian
(IV.22.6-8) gives details of his route from Alexandria to the Indus, the strategy
he followed and the help he received from local chieftains. Taxiles and the oth-
ers came t o meet him, bringing gifts reckoned of value among the Indians. They
presented him with the twenty-five elephants they had with them. Alexander
divided his army, sending Hephaestion and Perdiccas into the land of Peucelao-
tis, towards the River Indus with the brigades of Gorgias, Clitus and Meleager,
half of the Companion cavalry, and the entire cavalry of Greek mercenaries. H e
gave them instructions t o capture by force places en route, or to compel them t o
capitulate; and when they reached the Indus, they were t o make all necessary
preparations for the passage of the army. Taxiles and the other chiefs marched
with them. When they reached the Indus they carried out all Alexander's orders
but Astis, the ruler of Peucelaotis, revolted, bringing ruin on himself and on the

9. Dani, 1980, pp. 11 7-30.

 A. H. Dani and P. BernLzrd

city to which he had fled for refuge, when Hephaestion captured it after a siege
of thirty days. As far as the route is concerned, P. H. L. Eggermont,I0 relying on
the Geogrdphy of Strabo, makes Alexander cross the Cophen (Kabul) river but
this is not borne out by other historians. If Alexander crossed the Kabul river,
the only route open for him was through the Khyber pass - a route that he defi-
nitely avoided because it lacked water and was less inhabited. It is therefore rea-
sonable to believe that Alexander marched north of the Kabul river across the
Ningrahar valley into Bajaur. It is also clear that Hephaestion and Perdiccas,
who were sent ahead, did not follow the Khyber route. As their target was Peu-
celaotis (modern Charsadda, north of the Kabul river), they must have come
down through Mohmand territory. Both routes lay north of the Kabul river and
clearly indicate the stategy followed by Alexander. Peucelaotis was occupied
with the help of Sangaeus and Taxiles, but not without a great fight against
Astis. Alexander himself went north. We are not informed w h o was his guide,
but as we hear later that Sisicottus, a chief of this great region, was appointed to
administer the area, it is reasonable to believe that Alexander must have fol-
lowed his advice. Arrian (IV.23) calls this 'the land of the Aspasians, Guraeans
and Assacenians'. In modern geographical terminology, it embraces Nawagai,
Bajaur, Dir and Swat.

THE BATTLE W I T H THE A S P A S I A N S

The way was now blocked by the Aspasians, who followed a scorched-earth
policy and gave Alexander a tough fight, finally retreating into their mountain
fastness. Arrian (IV.24.6-25,4) describes it thus:

Then crossing the mountains Alexander descended to a city called Arigaeum [iden-
tified with Nawagai], and found that this had been set on fire by the inhabitants,
w h o had afterwards fled. There Cratcrus with his army reached him, aftcr accom-
plishing all the king's orders; and becausc this city seemed to be built in a conve-
nient place, he directed that general to fortify it well, and settle in it as many of the
neighbouring people as werc willing to live there, together with any of the soldiers
w h o were unfit for service. H e then advanccd to thc place whcre he heard that
most of the barbarians of the district had fled for refuge. . . . When the enerny who
were occupying the commanding heights saw the Macedonians approaching they
dcscended into the plain, being emboldened by their superiority in number and
despising thc Macedonians, because they were seen to be few. A sharp contest
ensued; but Alexander won thc victory with ease. . . . Ptolcmy indeed says that all
thc Inen were capturcd, to a number exceeding 40,000 and th'it over 2,300,000 oxen
were also taken, of which Alexander picked out the finest, becausc thcy seemed to
him to excel both in beauty and size, wishing to send them t o Macedonin to till the
soil.

10. Eggerniont, 1970, pp. 63 et seq.

 Alexander and his successors in Central A s k

That such a great booty in cattle was collected shows the great prosperity of the
region and the reason why the local tribe put up such a stiff resistance.

FIGHT WITH THE ASSACENIANS

Alexander then crossed the River Guraeus (the Panchkora, in Dir District).
Beyond the Karmani pass lies the Talash valley. The Assacenians, identified
with the ASvakas of Sanskrit literature, tried t o defend themselves. According t o
Arrian (IV.25.7-26,l):

When the barbarians perceived Alexander approaching, they durst not take their
stand for a battle in close array, but dispersed one by one to their various cities
with the determination of preserving these by fighting from the ramparts. The
most important of them was Massaga.

The ponderous ruins of Massaga occupy a conspicuous height near Ziarat about
16 km north of Chakdara fort. Here on a bare hill the walls of later-period ram-
parts have stood through the centuries t o speak of the brave defence that
the people put up against Alexander, as described by Quintus Curtius
(VIII.10.23-9):

An army of 38,000 infantry defended the city which was strongly fortified both by
nature and art. For on the east, an impetuous mountain stream with steep banks on
both sides barred approach to the city, while to south and west nature, as if design-
ing to form a rampart, had piled up gigantic rocks, at the base of which lay sloughs
and yawning chasms hollowed in the course of ages to vast depths, while a ditch of
mighty labour drawn from their extremity continued the line of defence. The city
was besides surrounded with a wall thirty-five stadia in circumference which had a
basis of stonework supporting a superstructure of unburnt, sun-dried bricks. The
brickwork was bound into a solid fabric by means of stones so interposed that the
more brittle material rested upon the harder, while moist clay had been used for
mortar. Lest, however, the structure should sink, strong beams had been laid on
top, supporting wooden floors which covered the walls and afforded a passage
along them.
Alexander while reconnoitring the fortifications, and unable to fix on a plan
of attack, since nothing less than a vast mole, necessary for bringing up his engines
to the walls, would suffice to fill up the chasms, was wounded from the ramparts
by an arrow which chanced to hit him in the calf of the leg. When the barb was
extracted, he called for his horse, and without having his wound so much as ban-
daged, continued with unabated energy to prosecute the work on hand. But when
the injured limb was hanging without support and the g a d u a l cooling, as the
blood dried, aggravated the ~ a i n he
, is reported to have said that though he was
called, as all know, the son of Jupiter, he felt notwithstanding all the defects of the
weak body. H e did not, however, return to the camp till he had viewed everything
and ordered what he wanted to be done.
 A. H. Dani and P. Bernard

It was at Massaga that we learn of Alexander's ploy. After the besieged had
agreed to surrender, Diodorus (XVII.84.1) informs us:

When the capitulation on those terms had been ratified by oaths, the Queen [of
Massaga], to show her admiration of Alexander's magnanimity, sent out to him
most valuable presents, with an intimation that she would fulfil all the stipulations,
Then the mercenaries at once, in accordance with the terms of the agreement, evac-
uated the city, and after retiring t o a distance of eighty stadia, pitched their camp
unmolested, without thought of what was to happen. But Alexander, who was
actuated by an implacable enmity against the mercenaries, and had kept his troops
under arms ready for action, pursued the barbarians, and falling suddenly upon
them made a great slaughter of their ranks. T h e barbarians at first loudly protested
that they were attacked in violation of sworn obligation, and invoked the gods
whom he had desecrated by taking false oaths in their name. Alexander, with a
loud voice, retorted that his covenant merely bound him to let them depart from
the city, and was by n o means a league of perpetual amity between them and the
Macedonians. The mercenaries, undismayed by the greatness of their danger, drew
their ranks together in a ring, within which they placed the women and children to
p a r d them on all sides against their assailants. As they were now desperate, and
by their audacity and feats of valour made the conflict in which they closed hot
work for the enemy, while the Macedonians held it a point of honour not to be
outdone in courage by a horde of barbarians, great was the astonishment and alarm
which the peril of this crisis created. For as the combatants were locked together
fighting hand to hand, death and wounds were dealt out in every variety and form.
Thus the Macedonians, when once their long spikes had shattered the shield of the
barbarians, pierced their vital organs with the steel points of these weapons, and on
the other hand the mercenaries never hurled their javelins without deadly effect
against the near target presented by the dense ranks of the enemy. When many
were thus wounded and not a few killed, the women, taking the arms of the fallen,
fought side by side with the men for the imminence of the danger and the great
interests at stake forced them to d o violence to their nature, and to take an active
part in the defence. Accordingly some w h o had supplied themselves with arms did
their best to cover their husbands with their shields, while others w h o were with-
out arms did much to impede the cnemy by flinging themselves upon them and
catching hold of their shields. T h e defenders, however, after fighting desperately,
along with their wives, were at last overpowered by superior numbers, and met a
glorious death which they would have disdained to exchange for ;I life with dis-
honour.

Massaga was only one of the fortress cities of the Assacenians. T w o more places
of great importance were then the target of attack. Arrian continues (IV.27.5-
28,l):

Thence hc dispatched Coenus to Bazira, entertaining an opinion that the inhab-

itants would surrender when they heard of the capture of Massaga. H e also dis-
patched Attalus, Alcetas, and Demetrius, the cavalry officcr, t o another city,
 Alexander and his successors in Central Asla

named Ora, with instructions to blockade it until he himself arrivcd. l'hc men of
this city made a sortie against thc forccs of Alcetas; but thc Macedonians casily
routcd them, and drove them into the city within the wall. But affairs at Bazira
were not favourable to Coenus, for the inhabitants showed no sign of capitulating,
trusting to the strength of the place, bccause not only was it situatcd on a lofty
eminence, but it was also thoroughly fortificd all round. When Alexander lcarnt
this, he started off to Bazira, but ascertaining that some of thc ncighbouring bar-
barians were about to get into the city of O r a by stealth, being dispatched thither
by Abisares for that very purpose, he first marched to Ora. H e ordercd Coenus to
fortify a certain strong position to serve as a basis of operations against thc city of
Bazira, and then t o come to him with the rest of his army, after leaving in that
place a sufficient garrison to restrain the men in the city from enjoying the free usc
of their land. But when the men of Bazira saw Coenus departing with the larger
part of his army, they despised the Macedonians, as not being ablc to contend with
them, and sallied forth into the plain. A sharply contested battle ensued, in which
500 of the barbarians fell and over seventy were taken prisoner. But the rest, flee-
ing for refuge into the city, were now more securely shut off from the country by
the men in the fort. The siege of O r a proved an easy matter to Alexander, for he
n o sooner attacked the walls than at the first assault he got possession of the city,
and captured the elephants which had been left there. When the men in Bazira
heard the news, despairing of their own affairs, they abandoned the city about the
middle of the night, and fled to the rock which is in their land, and is called
Aornos.

Both these forts lie on the left bank of the River Swat and to attack them Alex-
ander's forces had t o cross the river. Stein" identified O r a with Udegram and
Bazira with Barikot. From Udegram retreat upstream along the Swat river was
possible, but if Udegram had already fallen t o Alexander, the only escape from
Bazira t o Buner was across the Karakar pass. Abisares' forces could reach either
place through this pass as Alexander was coming from across the river. As a
stand was not possible, the forces must have planned to retreat t o Buner where
they could again get help from Abisares, who must have been planning t o stop
Alexander on that side of the River Indus. Although the Greek historians relate
that the Assacenians were defeated, they were not conquered, but took up new
positions at Aornos.
The very fact that Alexander did not advance directly to Aornos suggests
that by then he was well aware of the geography of the terrain where the local
chiefs had taken refuge and was well guided by other local chiefs. Who were
these local chiefs and what was their affiliation? We know the names of two of
them - Cophaeus (probably a ruler of the lower Kabul valley) and Assagates
(probably Aivagupta) - but they are not heard of again, except in connection
with Alexander's march towards Aornos.
Arrian informs us that Alexander fortified O r a and Massaga to keep the

11. Stein, 1929.

 A. H. Dani nmd P. Bernard

land in subjection and also the city of Bazira. H e appointed Nicanor as 'viceroy
of the land on this side of the River Indus', fortifying yet another city, Orobatis,
identified with V a r u ~ a(modern Shahbazgarhi, in Mardan District)
where the ASokan rock edicts are t o be found. As it lies in a strategic position
that could be used as a base for operations against enemy forces in Buner, its
fortification can be well understood. It was probably here that Alexander
planned his future campaign to dislodge his enemies from Aornos and to pre-
vent Abisares from interfering in the region.

THE CAPTURE OF A O R N O S

Stein12 takes 'Aornos' to be linguistically identical with (Mount) Una on the

Indus in the Indus Kohistan, near Thakot o n the modern Karakorum highway;
but 'Buner' could be a corrupted form of 'Aornos'. In that case, any high peak
in Buner (and there are several) could be identified with this last siege. Alexan-
der was bent on taking possession of Aornos, towards which he now moved.
The first city he reached is called Embolima, which is identified with ancient
Ambulima. Stein takes it for A m b because it is situated o n the route t o Mount
Una following closely the River Indus. O n the other hand Eggermont13 identi-
fies it with Ambela, an important pass that can be reached from Shahbazgarhi.
Both are strategically located. As a good strategist Alexander left the expe-
rienced Craterus with part of the army and himself advanced towards Aornos.
If Aornos was situated o n the right bank of the Indus, as the narratives of Dio-
dorus and Curtius state, one alternative would be t o accept its identification
with Pirsar on Mount Una, which can be reached from A m b along the river, or
from Ambela through the Kaghlum and Chakesar passes. But it is difficult to
understand how he could get t o Pirsar by the latter route when Erices (Assace-
nus' brother) was holding the district of the Assacenians, which was Buner
itself. Classical historians would have us believe that Alexander tackled Erices
after capturing Aornos. Eggermont would have Erices take refuge o n Mount
Elam, if he could be identified with the leader of the forces fleeing from Barikot,
because Elam lies t o the north-west by the side of the Karakar pass. In either
case the capture of Aornos was part of Alexander's Buner campaign, and it was
after this victory that he fortified the place, 'committing the superintendence of
the garrison t o Sisicottus' (Arrian IV.30.4). After such an arduous campaign
Alexander was finally able t o subjugate the Assacenians and the whole area west
of the Indus. When Nicanor was appointed viceroy we read of two local chiefs,
Sangaeus in Peucelaotis and Sisicottus (Sasigupta) in the farthest corner of
Bune r.

12. Stein, 1929.

13. Eggermont, 1970, pp. 63 et seq.
 Alcxandcr and his successors in (,'cntral Asla

After his arduous Buner campaign Alexander returned to cross the Indus at the
point where a bridge had been built. Arrian (V.3.5) describes his activities:

When Alexander arrived at the River Indus, he found a bridge made over it by
Hephaestion, and two thirty-oared galleys, bcsidcs many smaller craft. H e more-
over found that 200 talents of silver, 3,000 oxen, above 10,000 sheep for sacrificial
victims, and thirty elephants had arrivcd as gifts from the Indian l'axilcs; 700
Indian horse~nenalso arrived from Taxiles as a reinforcement, and that princc scnt
word that he would surrender to him the city of Taxila, the largest town between
the Rivers Indus and Hydaspes.

Arrian (V.8.2 et seq.) continues the story:

Then starting from the Indus, he arrived at Taxila, a large and prosperous city, in
fact the largest of those situated between the Rivers Indus and Hydaspes. H e was
received in a friendly manner by Taxiles, the governor of the city, and by the
Indians of that place; and he added to that territory as much of the adjacent coun-
try as they asked for. Thither also came to him envoys from Abisares, King of the
Mountaineer Indians, the embassy including the brother of Abisares as well as the
other most notable men. Other envoys came from Doxares, the chief of the prov-
ince, bringing gifts with them. Here again at Taxila Alexander offered the sacrifices
which were customary for him t o offer, and celebrated a gymnastic and equestrian
contest. Having appointed Philippus, son of Machetas, viceroy of the Indians of
that district, he left a garrison in Taxila, as well as the soldiers who were invali-
dated by sickness, and then marched towards the River H ~ d a s p e s .

It is very significant that Arrian calls Abisares 'King of the Mountaineer

Indians'. If this information is correct, should we not suppose that the forces of
the Mountaineer Indians who fought at Gaugamela were sent by him - an infe-
rence supported by the fact that Abisares had opposed Alexander all along,
while his enemy, Omphis, had been on Alexander's side? Abisares, who had
sent his troops t o support the Assacenians in O r a and Buner against Alexander,
now sent only his embassy of goodwill t o Taxila and never went in person to
attend on Alexander, who nevertheless tacitly accepted the pretence of submis-
sion. Quintus Curtius (VIII. 12.12-1 6) gives more information:

When Alexander asked him [Omphis] whether he had more husbandmen or

soldiers, he replied that as he was at war with two kings he requircd more soldiers
than field labourers. These kings were Abisares and Porus, but Porus was superior
in power and influence. Both of them held sway beyond the River Hydaspes, and
had resolved to try the fortune of war whatever invader had come. O m ~ h i s with
,
Alexander's permission, and according to the usage of the realm, assumed the
ensigns of royalty along with the name which his father had borne. His people
called him Taxiles, for such was the name which accompanied the sovereignty on
 A. H. D a n ; n n d P. Bernard

w]loIllsoevcr it devolved. Whcn, therefore, hc had entertained Alexandcr for three

da),s with lavish hospitality, he showed him on the fourth day what quality of corn
had supplied to Hephaestion's troops, and then presented him and a]] his
friends with golden crowns, and eighty talents besidcs of coined silver. Alexander
was so exceedingly gratified with his ~ r o f u s generosity
e that he not only sent back
to Olnphis the presents he had given, but added a thousand talents from the spoils
which he carried, along with many banqueting vessels of gold and silver, a vast
quantity of Persian drapery, and thirty chargers from his o w n stalls, caparisoned as
when ridden by himself.

Quintus Curtius (VIII.13.3) informs us:

Alexander had now resolved to cross the Hydaspes, when Barsaentes, who had
instigated the Arachosians to revolt, was brought t o him in chains, along with
thirty captured elephants. . . . Samaxus was also brought in chains, the king of a
small Indian state, w h o had espoused the cause of Barsaentes. Alexander, having
put the traitor and his accomplice under custody, and consigned the elephants to
the care of Taxiles, advanced t o the River Hydaspes.

This information brings a new perspective of the battle at Gaugamela vls-a-vis

the princes from the Indus region. It is not clear where Barsaentes was caught,
but if the identification of Samaxus with Sambus in Sind (as suggested by Egger-
mont) is accepted we can understand the purpose of the entire campaign of
Alexander in the Indus region; and the capture of Barsaentes at this stage does
focus on the part that was played by the princes from the Indus region in the
great battle at Gaugamela.
T o these events Plutarch adds:

Alexander, therefore, after having received many presents from Taxiles, and given
him more in return, at last drank t o his health, and accompanied the toast with the
present of a thousand talents of coined money.

A t this time in Taxila there was a certain Kautilya, the author of the well-known
book on Indian policy, the A ~ t h a i i s t ~ aw,h o was t o become famous as the
teacher of Candragupta, the founder of the Mauryan Empire. Plutarch records:

Sandrocottos [Candragupta] himself, w h o was then but a youth, saw Alexander

and afterwards used to declare that Alexander could easily have taken possession
of the whole country, since the king [i.e. onc of the Nanda kings of the Gangetic
valley] was hated and despised by his subjects for the wickedness of his disposition
and the meanness of his origin.

Justin (XV.4.15) adds:

This man was of humble origin, but was sti~nulatedto aspire to regal power
by supernatural encouragement; for having offcndcd Alexander by his boldness
 Alexander and his suc-ces>orsin (.'cntrul Asia

of speech and orders bcirig givcn to kill him, he saved himself by swiftness of
foot.

Unfortunately we d o not know whcn and whcre Candragupta mct Alcxandcr,

but as he was in Taxila, that is the most likcly placc of thcir meeting

ALEXANLIEII AN11 I'OKUS

From Taxila t o the Hydaspes Alexander had the choice of two main roads.
Either would be practicable provided the passes were in the hands of allics. T h e
principal chain of the Salt range commences in the lofty hills of Chcl formcd by
the convergence of three spurs, two of which extended as far as the Himalayan
out-liers. T h e first is traversed by the Grand Trunk Road at Bakrala and, 32 km
lower down, by the Dhudial-to-Jalalpur road at the gap through which the
Bunhar Nullah flows. T h e spur o n which the fort of Kohtas stands is tcrminatcd
at one end by the Bunhar and at the other by the N u h a n Nullah, which flows
through the Pubbi range near the apex of the triangle. T h e lowcr road, which
emerges near Jalalpur, is narrow and was perhaps under the control of
Sopeithes, while the northern route - the Grand Trunk Road - was under Abi-
sares' control. T h e route followed b y Alexander depended upon the relation-
ship of these t w o chiefs t o Porus. In spite of his feigned submission t o Alexan-
der, Abisares was ready t o support Porus with w h o m he was in league, and
therefore the northern route appears t o have been less ~ r c f e r r e d .O n the other
hand, Sopeithes, the ruler of the Salt range, was too weak t o stand against Alex-
ander and more likely t o yield. O n this ground the southern route was prefer-
able, as Stein has argued o n other, geographical, grounds. O n the other side of
the H y d a s ~ e sUhelum) lay the Kingdom of Porus (Fig. 1). T h e narnc appears t o

FI<;.1. Silver coin nrith tlie figure of Porus (BMC 191.61).

(Photo: 0 British h l u s c u n ~ . )
 A. H. Dani and P. Bernard

have been derived from the ancient Puru tribe, which at this time must have
spread from Jhelum eastward beyond Chenab, probably u p t o the River Ravi,
because the younger Porus, nephew of the former, ruled here. H e was antago-
nistic t o his uncle and is reported t o have offered Alexander help against him,
hoping to be installed as ruler over the whole area. The territory of the elder
Porus between the Jhelurn and Chenab was well defended. Strabo tells us that it
was an extensive and fertile district containing nearly 300 cities. Diodorus men-
tions that Porus had an army of more than 50,000 foot soldiers, about 300
horses, over 1,000 chariots and 130 elephants. The story of the battle against
Porus is related in detail by the Greek historians.
The special strategic moves made by Alexander f o r the battle, the severity
with which the battle was fought, the daredevil courage shown by Porus, and
his final treatment by Alexander indicate the importance of the war. It appears
that all along Porus was the main target of attack on this side of the Indus, just
as the Assacenians had been on the other. As Sisicottus was Alexander's suppor-
ter on the west, so Omphis was on the east, while Abisares played a political
game on both fronts. As we shall see later, he not only saved his life but pre-
served his territory and never submitted personally t o Alexander.
After his victory over Porus, Alexander (Fig. 2) founded the city of
Nicaea, as well as another city, Bucephala, in memory of his horse Bucephalus,
who died there. Alexander not only restored his territory t o Porus but also
added the neighbouring region beyond the Chenab that belonged to the 'Glau-
sians' and as far as the Ravi, where the younger Porus had risen in revolt as his
uncle had been honoured by Alexander. A t a campsite in this area Alexander
effected a reconciliation between Taxiles and Porus. And there arrived a second
ambassador from Abisares, who, seeing the failure of his political manoeuvres,
again feigned submission by sending his brother with a gift of forty elephants. It
is strange that even then he did not come in person. We are told:

A t this time Phrataphernes, viceroy of Parthia and Hyrcania, came t o Alexander at

the head of the Thracians w h o had been left with him. Messengers also came from
Sisicottus, the viceroy of the Assaccnians, t o inform him that these people had
slain their governor [Nicanor] and revoltcd from Alexander. Against thcse hc dis-
patched Philippus [the viceroy of Taxila region] and Tyriespis, with an army, to
arrange and set in order the affairs of their land.

This was the first revolt by local people in the territory of the Assacenians.
It was o n the banks of the Ravi that Alexander met the Cathaeans (with
their stronghold at Sangala). There were many other independent tribes
between the Ravi and the Beas. In these campaigns Porus acco~npanicdAlexan-
der and helped him with the elephants. After a great siege Sangala was captured
and razed to the ground. Alexander advanced up t o the Beas which was prob-
ably the limit of the Achaemenid Empire. Beyond lay the Gangetic kingdom of
the Great Nandas.
 A. H. D a n i a n d P. Bernard

We are informed that here the Greek soldiers revolted and would not proceed
further to conquer the Gangetic region. Although Alexander unwillingly
acceded to their demands, it is strange that they did not insist on returning by
the route they had come. Was this story of revolt concocted by the Greek his-
torians to put all the blame on the soldiers and prove that Alexander wished to
be a world conqueror but stopped at the limit of the Achaemenid Empire? The
army certainly stood by him so long as he was only subjugating the lands that
had belonged to the Achaemenids. If the army only wanted an easy return
home, it is difficult to understand why Alexander tackled new hazards by going
south.
It is said that Alexander made Abisares viceroy over his own country and
the region belonging to Arsaces. Thus to the east of the Indus, Porus and Abi-
sares, who were enemies of Taxiles, were made stronger. O n the other hand
Philippus, who was appointed 'viceroy of the district' at Taxila, became 'viceroy
of the country beyond the Indus extending to Bactria' after the murder of Nica-
nor. Later his authority was extended u p to the territory of the Malli, that is, to
the confluence of the Indus and the Chenab. Beyond this point as far as the sea,
and extending over the Makran coast, Pithon was appointed viceroy; but that
area remained to be conquered. Meanwhile Alexander went back to the River
Hydaspes.

A L E X A N D E R FIGHTS HIS W A Y TO T H E SOUTH

Arrian (11.1.1-2,2) continues the story:

Alexander now resolved to sail d o w n the Hydaspcs to the Great Sea, after he had
prepared o n the bank of that rivcr many thirty-oared galleys and othcrs with one
and a half banks of oars, as well as a number of vessels for conveying horses, and
all the other things requisitc for the easy convcyance of an army on a river. . . .
With himself he placed on board all the shield-bearing guards, the archers, the
Agrianians and the bodyguard of the cavalry. Craterus led a part of the infantry
along the right bank of the Hydaspes, while along the other bank Hephaestion
advanced at the head of the most numerous and efficient part of the army, includ-
ing the elephants, which now numbered about 200. These generals were ordered
t o march as quickly as possible to the place whcrc the palace of Sopeithes was
situated.

They reached this o n the second day.

Alexander then proceeded down the Jhelurn to its confluence with the
Chenab. In the southern Panjab he had to fight against four tribes - the Sibi,
Agalassi (Diodorus XVII.98), Sudracae and Malli. The Sibi occupied the Shor-
kot region in Jhang District, and the Agalassi (or the Agrasura) must have been
 Alexander and his successors in Central Asia

close neighbours. The other two tribes, the Malli (or Milavas) and the Sudracac
(Sudvaka o r Ksudraka), made a joint defence against Alexander. It was in the
fortress city of the Malli, which was heavily defended, that Alexander was badly
wounded. The city seems to have been Milavasthina (probably modern Mul-
tan) and the Sudracae must have lived in the Bahawalpur region northward
along the Ravi.
The onward journey t o the sea was interrupted by two more geographical
features - the great Indus gorge at Sakkhar and the head of the delta below thc
hillock of Thatta. In the first area Alexander met two important tribes, that of
Musicanus, probably with their headquarters at Al-Ror, near Sakkhar, and the
second of Sambus, with their city called Sindimana. In the name 'Musicanus' it
is not difficult t o see the ancient tribe of Miisikas, or Mausikas, and in the name
'Sambus' the later Sindhi tribe, Sammas. It is said that the Brahmans instigated
their fight against Alexander, who defeated them and destroyed their cities.
The next important place was Patala, where 'the water of the Indus is
divided into two large rivers, both of which retain the name of Indus as far as
the sea. Here Alexander constructed a harbour and a dockyard.' Quintus Cur-
tius (IX.8.28) further writes:

From there they came to the next nation, that of the Patalii. Their king was Moeris,
who had abandoned his city and taken refuge in the mountains. Alexander took
the town and pillaged the fields. From there great booty was driven off, in the
form of flocks and herds, and a great store of grain was found. Then taking guides
acquainted with the river he sailed down to an island which arose in the middle of
the channel.

There has been a vain attempt to identify the city of Patala. If 'Patala' is not
taken as a proper name but only refers t o a city, it can be corrected t o 'Pattana',
that is, city o r port city par excellence, a term applied in a later period t o Thatta,
which is ideally situated in the way the Greek historians describe. King Moeris
has been taken by Eggermont to be Mauryas - but without any reasonable
foundation, and he is better regarded as the head of the local tribe Med or Mehr,
which is well known in the Sind coastal area.

ALEXANDER MARCHES BACK ACROSS BALUCHISTAN

When the exploration of the Indus was complete, Alexander ~ r e ~ a r efor

d the
return journey. H e had already, according to Arrian (VI. 17.3),

sent Craterus into Carmania with brigades of Attalus, Meleager and Antigenes,
some of the archers and as many of the Companions and other Macedonians as,
bcing now unfit f o r military service, he was dispatching to Macedonia by the route
through the lands of Arachosians and Zarangians.
 A. H. Dani and P. Bernard

Eggermont rightly points out the line of march along the ancient caravan trail
from Al-Ror through the Bolan pass to Kandahar and from there to Sistan,
According to Strabo (613.3 et seq.), ~ l e x a n d e r

hilnself set out with one division through Gedrosia. H e kept away from the sea, no
more than 500 stadia at most, in order that he might at the same time equip the
seaboard for the reception of his fleet; and hc often closely approached the sea,
although its shores were hard to traverse and rugged.

The fleet he gave over to Nearchus and Onesicritus, the latter his master pilot,
giving them orders to take an appropriate position, and t o follow, and sail
alongside, his line of march.
Alexander's retreat from the Indus delta has been reconstructed by Egger-
mont who has evaluated the two possible routes - the northern one, suggested
by Stein, and the southern one given by Holditch. H e has opted for the latter to
identify the port town of Alexandria in the Oreitae country near the mouth of
the Hingol river - the most important river in Baluchistan, separating the east-
ern part held by the Oreitae tribe from the western part - Gedrosia proper.
Alexander entered via the River Arabis (the H a b river flowing between the Kir-
thar and Pab ranges that run in a north-south direction). The focal point here is
the central Kalat area, the southern part of which is drained by the Porali river.
The Oreitae tribe appears to have occupied this entire zone. They put up a stout
resistance and rose in revolt after the departure of Alexander, but were brought
to book by Leonnatus. Even later, Diodorus (XVII.105.8) informs us, when
Alexander 'was on the march, some of the Oreitae, having attacked the troops
commanded by Leonnatus and slain a good many men, escaped unscathed into
their own country'.
Arrian (VI.27.1-2) tells the story of Alexander's last appointments:

When he arrived at the capital of the Gedrosians he gave his army a rest. Apol-
lophanes he deposed from his satrapy because he found out that he had utterly dis-
regarded his instructions. H e appointed Thoas t o be satrap over the people of this
district, but as he was taken ill and later died, Sibyrtius occupied the vacant post.
The same man had also recently becn appointed by Alexander satrap of Carmania,
but n o w the government of the Arachosians and Gedrosians was committed to
him, and Tlepolemus, the son of Pythophancs, got Carmania. T h e king was
already advancing into Carmania when tidings reached him that Philippus, satrap
of the Indian Country, had been treacherously ~nurdercdby the mercenaries; but
that his Macedonian bodyguards had put to dcath his murderers whom they had
caught in the very act, and others whom they had afterwards seized. O n learning
what had occurred he sent a letter t o India addrcssed to Eudemus and Taxiles
directing them to assurne the administration of thc province previously governed
by Philippus until he could send a satrap to govern it.
 Alexandrr and his succc~ssorsin C'cntral Am

That opportunity nevcr came. The finale of Alexander's march is givcn in thc
words of Justin (XV.l .lo-15):

Seleucus Nicator waged many wars in thc cast after thc partition o f Alcxandcr's
empirc among his generals. H e first took Babylon and then with his forcer aug-
mented by victory subjugatcd the Bactrians. H e thcn passcd ovcr into India, which
after Alexander's death, as if the yoke of servitude had been shakcn off from its
neck, had put its prefects t o death. Sandrocottus was the leader w h o achicvcd their
freedom, but after his victory he forfeited by his tyranny all titlc t o thc nanlc of the
liberator, for hc oppressed with servitude the very people whom hc had cmanci-
pated from foreign thraldom.

Alexander returned t o Susa only t o find that the satraps appointed by him had
enrolled mercenaries and acted as independent rulers while some of the Persian
satraps had ill-used and murdered their subjects. 'One trouble, a revolt of Greek
mercenaries in Bactria, was not really overcome; Amyntas was replaced by
another Philippus, but the discontent simmered till Alexander died.'I4 Alexander
was struck down by fever and died in Babylon on 13 June 323 B.C.
H o w far Alexander succeeded in uniting his empire is difficult t o say
because the men that he posted as satraps in the different provinces could not
remain in power after his death. H e certainly succeeded in bridging the rift be-
tween the Greek and Persian worlds, and, by bringing the two into one imperial
system, he fulfilled the aim that once inspired the Achaemenids; the voyage of
his admiral Nearchus must have added information t o that already gained by
Skylax in the time of Darius I, and the new silver currency issued by Alexander
must have accelerated trade and commerce. The new cities he founded in Asia,
and the Greek population he settled in them, ~ l a n t e dthe seeds of Hellenistic
culture and inaugurated a new spirit of cultural exchange.
However, the empire that he founded did not survive him. Its unity was
destroyed, and for forty years after his death his own companions and comrades
indulged in mutual strife. The one who emerged successful in Asia was Seleucus
Nicator. The major claim t o independence from the Seleucids came from the
Indus region where Candragupta Maurya, with the support of the Parvataka
(probably the Paurava tribe), overthrew the Greeks and gained the provinces of
Aria, Arachosia, Gedrosia and the Paro~amisadae,allnost all the eastern areas
where troubles had been brewing in Alexander's lifetime. Within half a century
the Seleucids lost Parthia and Hyrcania, provinces situated t o the south-east and
east of the Caspian Sea, and probably at about the same time Bactria threw off

14. Tarn, 1951, p. 110.

 A. H. Dani and P. Bernard

their suzerainty. The Seleucids continued t o control Iran until the Roman
menace roused the national consciousness of the Iranians. The Parthians gave
the final signal, recovering Iran from the Greeks and stabilizing their western
frontier on the Euphrates.

Part Two

(Paul Bernard)

Rare as they are for this period, historical sources and coins help t o pierce the
darkness surrounding the fate of the Greek colonies in Central Asia during the
twenty years between the death of Alexander (323 B.c.) and the conquest of
Central Asia by Seleucus I (c. 305-304 B.c.). The survival of this Greek presence
in regions far away from the Mediterranean and apparently isolated is primarily
explained by the fact that it had put down roots. Even if we reject the theory of
a genuine Greek colonization prior t o Alexander of political exiles settled in
these provinces by the Achaemenid kings,I5 it is nevertheless evident that the
number of colonists left behind by Alexander was far from negligible. Classical
texts mention 13,500 soldiers in the Oxus valley (Arrian IV.22) and 4,600 in
Arachosia alone (Curtius VII.3-4), the centre af troop disposition south of the
Hindu Kush. T o this number must be added the pensioned soldiers, who were
settled in the newly founded towns, as the conquest progressed.

The announcement of Alexander's death caused 23,000 Greek mercenaries set-

tled, n o doubt against their will, in the upper satrapies of the Iranian plateau and
in Central Asia to rebel and attempt t o return t o their distant homeland (Diodo-
rus XVIII.7).'" But the dead king's lieutenants, w h o had sided with the regent
Perdiccas, immediately instructed the general responsible for suppressing the
rebellion t o put the rebels to death, convinced that they could count on other
troops remaining at their posts. Far from feeling imprisoned in a hostile envi-
ronment behind their ramparts, the colonists w h o had not abandoned their
posts were sufficiently sure of themselves t o join in the power struggles then
taking place in Western Asia. The consideration shown t o their satraps on the
occasion of three successive reorganizations of the empire (at Babylon in 323
B.c., at Triparadisus in 321 and at Persepolis in 316), and which amounted, for

15. Narain, 1957, pp. 1-6.

16. On this revolt, see Koshelenko, 1979, pp. 181-221.
 Alexander and his successors in Central Asia

the eastern provinces, t o little more than a few staff changcs indicated thcir con-
siderable political importance in the balance of power. In 31 7 ~ . c : . ,the satraps of
Central Asia, Bactria, Aria-Drangiana, Arachosia, the Paropamisadae and Gand-
hira joined forces t o check Pithon, their powerful colleague in Media, whose
ambitions threatened their own position. Their 6,500-strong army (further proof
that the rebellion of 323 B.C. had not drained the country of colonists) was victo-
rious. It was with this army that they sided with Eumenes in his struggle against
Antigonus, and, in spite of the defeat suffered by the coalition in Iran (316 B.c.),
Antigonus was careful not t o undermine their power. These Clitc soldiers were
also political creatures who knew how to attend to the effective running of their
provinces, where they had no doubt rallied the local nobles t o their cause to
ensure local support. In this context, a highly significant comment is made by
the Greek historian, Hieronymus of Cardia, who was personally involved. I-Ie
notes that the reason behind Antigonus' decision to confirm the satraps of Car-
mania and Bactria in their offices was 'because they would not allow themselves
to be dismissed by a mere letter, given the many partisans at their service among
the local populations, whose allegiance they had won through their fine admi-
nistration' (Diodorus XIX.48). The silver coins struck in the Greek manner
which appeared at this time in the Oxus valley (either imitating Athenian coins
or with an eagle design o r bearing the name of Sophytes)," minted for local use,
indicate that these satraps were also concerned with the economic development
of their provinces. The ground had been well prepared for the fresh wave of
Greek colonization, which was t o be initiated by the Seleucid kings.
After a gap of some ten years, the satrapies of Central Asia reappear in the
history of the Hellenistic kingdoms, through their inclusion in the empire of
Seleucus I. Seleucus, a Macedonian noble, had ~ u r s u e dan uneventful military
career among the companions of Alexander, gradually rising through the officer
ranks. At the division of Triparadisus (321 B.c.) he was allotted the important
satrapy of Babylonia. After many turns of fortune, in which he demonstrated
his tenacity, political acumen and administrative talents, Seleucus eventually
consolidated his power over the entire region of Mesopotamia and northern
Syria. Before engaging in the final struggle with his rival Antigonus, who had
withdrawn t o Anatolia, Seleucus had to be certain of his eastern borders. In 307
B.C. he therefore decided t o ensure acknowledgement of his authority in the
satrapies of the Iranian plateau and Central Asia, and his expedition does not
appear to have encountered any serious resistance from the Greeks living north
of the Hindu Kush.

MAURYAS A N D C E N T R A L ASIA

South of the Hindu Kush, however, Seleucus came up against a new, non-
Greek, power - the Indian Empire of the Mauryas. Its founder, Candragupta,
17. Mitchinet-, 1976, pp. 21-4.
 A. H. Dani and P. Bernard

had recently extended his power beyond the Indus, up to the eastern edge of
Iranian plateau. In spite of the vagueness of the historical texts, the treaty con-
cluded between Candragupta and Seleucus seems to acknowledge the fait
accompli of Indian control of a large part of the territories west of the Indus,
comprising Gedrosia, the Paropamisadae (the region of Kabul and Begrarn) and
Arachosia (the Kandahar region). The Greek colonies in these regions, particu-
larly Alexandria ad Caucasum (Begram) and Alexandria in Arachosia and Alex-
andropolis (Kandahar), thus became subject to a foreign power, but this depen-
dence did not prevent them from flourishing while remaining true to their
ancestral traditions. T o safeguard the interests of the Greeks and Macedonians
who had come under foreign rule, Seleucus concluded a convention with the
Indian rulers which guaranteed full rights to children born of intermarriages
with local Iranian women. Having secured his eastern frontier by incorporating
into his kingdom the provinces north of the Hindu Kush, and by stabilizing
through an alliance his relations with the Mauryas, Seleucus returned west with
the 500 elephants he had obtained from Candragupta. There he eliminated his
rival Antigonus at the battle of Ipsus (301 B.c.) and extended his empire over a
large part of Anatolia.

POLITICAL UPHEAVALS

Slightly before 290 B.C. the provinces of Central Asia were rocked by upheavals
that destroyed several cities, particularly in Margiana and Aria.18 These were
caused not by local revolts against the Greek colonists, but by a wave of
nomadic invaders. This may be supposed from the military expedition led by an
important servant of the Seleucid state, Demodamas of Miletus, who advanced
to the Jaxartes (modern Syr Darya). O n the banks of this river, beyond which
lay the territory of the nomads, Demodamas erected altars honouring Apollo of
Didyma, protector of the Seleucid dynasty, as a symbolic affirmation of Greek
presence. Seleucus made his son Antiochus viceroy, and put him in charge of
the upper satrapies. The royal coins bearing the combined names of the two
sovereigns, which were struck by the mint at Bactra,I9 bear witness to a special
relationship between Antiochus and the Bactrian satrapy. Antiochus probably
lived in Bactria for a time, using it as a base for supervising the reconstruction of
the devastated provinces. Seleucid concern for these provinces continued until
the death of Seleucus I (281 B.c.). During his own reign (281-261 B.c.), Antio-
chus I was no doubt too occupied with the difficulties he faced in Anatolia and
his rivalry with Ptolemaic Egypt to give them the same attention. There was
ample reason for the interest shown by the Seleucids in these Central Asian

18. Wolski, 1947, pp. 13-70.

19. Newell, 1978, pp. 23 1-6.
 Alexander and his successors in Central Asia

satrapics. Their strategic importance lay in their rolc as the bulwark of thc
empire against the continuing threat from Asiatic nomads. Their wealth carnc
from thcir oases, especially those in the valleys of the Oxus and Polytimctus
(modern Amu Darya and Zcrafshan), which enjoyed an agricultural surplus
from the expansion of irrigated land, and prosperity from the metals and prc-
cious gems found in the mountains of the region. This interest was strengthened
by family ties. Seleucus had married the Bactrian princess, Aparna, whose
father, Spitamenes, had organized the resistance against Alexander; Seleucus
named several of the towns that he founded after her; and Antiochus was
Apama's son.

U R B A N I Z A T I O N A N D CITY L I F E

The founding or refounding of cities bearing Antiochus' name testifies to his

determination t o consolidate and develop the urban fabric of these satrapies. In
the oasis of Merv (ancient Margiana) the Achaemenid settlement at Erk-kala,
transformed by Alexander into Alexandria in Margiana, was incorporated by a
new city (the Gyaur-kala site) whose massive unbaked-brick ramparts encloscd
a vast square 1,500 m across, fortified at each corner by a bastion. Inside the
ramparts, whose irregular contours reflect those of the site, two main streets
linking the four gates crossed at the centre of the city.?"ccording to Strabo, the
king was so impressed by the fertility of the oasis and anxious to protect it from
nomadic incursions like the one that had recently devastated it, that he ordered
the city t o be surrounded by a rampart 1,500 stadia (250 km) long. Lengthy sec-
tions of this defence work, consisting of a light rammed-earth wall punctuated
with towers, have been discovered on the northern boundaries of the oasis." At
Maracanda (Samarkand), the main site of the Zerafshan oasis, a rampart with a
corridor inside, following the irregular contours of the Achaemenid city and
dated by pottery of the first half of the third century B.C. can be attributed t o
the period of Antiochus I, in spite of the differences in architectural technique.??
The presence of a Greek colony whose origins go back t o the period of Seleucid
rule is confirmed by the discovery there of a Greek name (Nikias) engraved on
a vase.2JThe city of Antioch in Scythia, mentioned by a Byzantine author, may
be the former Alexandria Eschate (modern Khojand-Leninabad) on the Jaxartes,
refounded in the name of Antiochus I by Demodamas during his expedition
against the nomads. Further south, in the Hari-rud valley, Antiochus restored
the ramparts of Artacoana, headquarters of the satrapy of Aria (possibl~.Alex-
andria in Aria). The resent-day town of Herat represents - in the quadrilateral

20. Fila~iovich,1974, pp. 1 et seq.

21. Viazigin, 1949, pp. 260-75; Merezhin, 1978, pp. 11-15
22. Shishkina, 1969, pp. 62-75, 1975, pp. 60-7s.
23. Shishkina, 1975, p. 69, Fig. 9.1.
 A. H. Dani a n d P. Bernard

form of its medieval ramparts, and in its four gates and two main streets inter-
secting at right angles near the town centre - the topography and grid pattern
typical of Hellenistic cities." T w o other towns, also in the satrapy of Aria, owe
their existence to Antiochus I: Soteira, from his surname Soter (meaning
'saviour') and Achaea, founded by Alexander and refounded by the Se]eucid
king. At Bactra (the capital of Bactria) the most ancient rampart of the acropo-
lis, a solid mass equipped with ~ r o j e c t i n gtowers, perched o n a high base of
adobe, probably dates from the Greek ~ e r i o dthough
, it is not possible to deter-
mine whether it was built under Seleucid rule.25In contrast, the remains of a
lesser rampart surrounding the oasis, also constructed of adobe but flanked with
rectangular towers, recalls the rampart built by Antiochus I t o defend the Mew
oasis, and may well have been erected on his ~rders.~"he Greek city about
which we are best informed, as a result of extensive French excavations, is the
town located on the site of A y Khanum, o n the eastern borders of Afghan Bac-
tria, at the confluence of the Oxus and its southern tributary, the Kokcha. It is
not known whether the city was founded by Alexander o r Seleucus, but we do
know that its rise began under Seleucid rule. During this period the natural
defences of the vast site (1,800 x 1,500 m), formed by the two rivers and a natu-
ral hill which acted as its acropolis, were completed by the construction of
massive ramparts built of unbaked brick, reinforced with full rectangular tow-
ers, and by a citadel erected in the south-east corner of the acropolis. The basic
layout of the town (Fig. 3) was designed so that the main street, at the foot of
the acropolis, left room for the broad expanse of the lower town with its vast
palace (Fig. 4). The residential area, with patrician mansions, was laid out in the
triangle formed by the junction of the two rivers, while the most important
sanctuary of the city was located on the side of the main street.27
During the first half of the third century B.c., under the reigns of Seleucus
I (31 1-281), Antiochus I (281-261) and Antiochus I1 (261-246), the Greek
provinces of Central Asia were part of an empire centred around the ancient
Greek lands of Anatolia and Hellenized Western Asia. It was a crucial period
for these colonies as their Hellenism was then nourished by frequent contact
with Mediterranean influences which were able to penetrate freely, propagated
by officials, soldiers, merchants, artists and intellectuals, such as the Aristotelian
philosopher Clearchus who, on his way from Greece to India t o investigate Ira-
nian and Indian religions, left behind at Ay Khanum a copy of the aphorisms
embodying the most venerable Greek wisdom engraved in the sanctuary of
Apollo at Delphi.z8

24. Lezine, 1963/64, pp. 1 2 7 4 5 .

25. Dagens et a]., 1964, pp. 61-104.
26. Pugachenkova, 1976, pp. 1 3 7 4 3 .
27. Bernard et al., 1973, Annual reports in CRAI, 1965-80; for a overview, see
Bernard, 1981, pp. 108-20.
28. Robert, 1973, pp. 21 1-37.
Alexander and his successors in Central A s k

FIG.3. Plan of Ay Khanum.

I;I(,. 4. Pl'ln of t h e palace of A y Klinnum (reconstruction).
 Alexander and his successors in Ccntrnl Asla

New colonists, many of whom probably came from the Scleucid posses-
sions of Asia Minor, strengthened thc Greek presence in thc Ccntral Asian
satrapies. Under Sclcucid administration, the towns of thcsc satrapics learnt
how to conciliate the respect due t o the monarchical powcr and the practice of
municipal institutions such as they wcrc found in any Greek city, within the
limits of autonomy allowed by the royal authorities. The activity of thc mints of
Bactra and A y Khanumz9indicates the economic prosperity of the wholc rcgion.
Almost 40 per cent of all bronze coins discovcred at Ay Khanum wcrc struck
by the first three Seleucid kings. The West, in return, exhibited a curiosity about
this new world. During their joint reigns, Seleucus and Antiochus ordered
Patroclus, one of their generals, to explore the Caspian, and Demodamas wrotc
a treatise devoted t o the geographical observations he madc during his timc in
Central Asia.

THE END OF SELEUCID POWER I N CIENIKAL ASIA

Seleucid power in Central Asia fell victim to the very success of the colonics
that it had so strongly fostered. Having increased the Greek elements of their
population, enjoying the resources of a booming economy and benefiting from
the support of local nobles and the mass of peasants under their rule, these col-
onies must have grown increasingly impatient with the monarchy, whose con-
cerns were predominantly directed towards Western affairs, and eventually fclt
strong enough t o take their destiny into their own hands. The break, which
occurred gradually without provoking a reaction from the central authorities,
was instigated by Diodotus, the satrap of Bactria-Sogdiana. Diodotus struck
coins still bearing the name of his sovereign, Antiochus 11, but he substituted his
own emblem (Zeus wielding a thunderbolt) and portrait in the place of his mas-
ter's. The complete break came with the king's death in 246 B.C. (or according
to some sources, slightly later, in 238 under Seleucus IT), when Diodotus took
the final step of striking coins in his own name with the title 'king'.'Trom that
time onwards, the Greek territories north of the Hindu Kush formed an
independent kingdom, t o which modern historians have lent the name Graeco-
Bactria.
From the late third century B.c., the Greek colonists south of the Hindu
Kush in Arachosia, the Paropamisadae and Gandhira had been subjects of the
Mauryan Empire and were t o remain so for over a century, until around 200
B.c., when the conquests of the Graeco-Bactrians brought them back into a
Greek state. In 205 B.c., Antiochus I11 even renewed with the Indian sovercign,
Sophagasenus, the treaty concluded in 303 B.C. by his ancestor, Seleucus 1,
confirming Indian sovereignty over these territories. Far from being a source of

29. Mitchiner, 1975, pp. 28-32; Bernard, 1985, pp. 6 et seq.

30. Newell, 1978, pp. 245-9.
 A. H. Dani and P. Rcmavd

hostility or conflict between the Mauryan and Seleucid empires, the presence of
the Greek colonies on the western borders of India fostered neighbourly
tions between the two.ll The Seleucid kings regularly sent ambassadors to the
court of P ~ y a l i ~ u t r-a first Megasthenes and then Daimachus. The name o f ,
representative of Ptolemaic Egypt, Dionysius, has also come down. Emissaries
sent by ASoka to spread Buddhist doctrine in the West visited the states ruled
by Antiochus I1 and other western kingdoms.
Even though it formed a minority among the indigenous p o p ~ l a t i ~ ~ ,
whose language and culture were Iranian, the Greek element, concentrated in
the towns and administrative centres, probably continued under Mauryan rule
t o play the leading role it had enjoyed during the last quarter of the fourth ten-
tury B.C. in the early days of colonization. The vigour with which the traditions
of Hellenism were maintained in these regions is a cultural phenomenon with
its roots deeply embedded in politics. When Emperor ASoka ordered his edicts
to be engraved in a Greek translation at Kandahar, he gave clear evidence of the
importance of the Greek colonists whom he addressed in their own language.
The discoveries made over the last twenty years o n the site of the old town at
Kandahar provide striking examples of the firmly rooted Greek culture in this
Indo-Iranian setting. In contrast t o the new cities that were built on virgin
ground, the Greek settlement at Kandahar is interwoven with the remains of the
Achaemenid town.j2
The texts discovered there are just as eloquent as those found at Ay Kha-
num. A votive inscription offered by the son of Aristonax3' provides evidence
that people still knew how t o write Greek verse there during the early third
century B.C. T w o other Greek inscriptions (one accompanied by a version in
Aramaic, the language of the Achaemenid administration) paraphrase some of
the fourteen rock edicts written in the Indian language and engraved on rock at
different Indian sites, in which the Mauryan emperor ASoka (268-237 B.c.)
directed his subjects t o observe the law of the Dhavrna and to practise the vir-
tues it sought t o inspire - non-violence, compassion, tolerance and the service
of others. The intimate knowledge of the current language of Greek philosophy
shown by the Greeks - for good translations can only be made into one's
mother tongue - in their search for the closest equivalents t o Indian concepts is
a clear indication that the Hellenism of this Greek colony was nourished by the
loftiest and liveliest Western thought. It was through these Greek colonies
under Mauryan rule that the Indian and Mediterranean worlds entered into
contact, and that a mutual curiosity arose between them. The story of Emperor
Bindusira's request t o his colleague Antiochus I for a philosopher, some wine
and some figs is well known, as is the Greek's mocking reply (Atheneus

3 1. Schwarz, 1970, pp. 267-316.

32. Reports on excavations in Afghan Studies, 1978, 1979, 1982.
33. Fraser, 1979, pp. 9-21.
 Alexander and his successors in Central Asia

~ I V . 6 5 2 - 3 ) .A Greek romance of the Hellenistic period also tells the story of a

'Philhellenic' Mauryan emperor w h o rescues a shipwrecked Greek on the coast
of Bengal and has him escorted through his territories as far as the Persian bor-
der (Diodorus 11.55-60). Megasthenes, Seleucus 1's ambassador t o the court of
Piyaliputra, collected material f o r his book, which became the indispensable
work of reference o n India f o r the entire ancient world. Modern scholarship has
reaffirmed the reliability of many of his observations on the geography, eth-
nography and society of the s ~ b c o n t i n e n t . 'Thus
~ the century-long annexation
of the territories south of the Hindu Kush by the Mauryan Empire created no
obstacles t o the implantation of Hellenism in these regions.

34. Bongard-Levin, 1971, pp. 109-22; Derrett, 1979.

97
 P. Bernard

Political history
Written sources for the history of Greek rule in Central Asia are scarce and
fragmentary. The works of classical antiquity that dealt with the subject have
been lost; all that remain are some fragments: Polybius' account of the expedi-
tion of Antiochus I11 which survives in mutilated form, X.27-31, 49 and XI.39,
some indirect references (the History of Parthia by Apollodorus of Artemita, on
which Strabo drew for his Geography) and a synopsis (Justin's synopsis of the
Historiae Philippicae, Book XLI, by Pompeius Trogus). Bactria does not appear
in the Chinese chronicles, the Shih-chi, the Han-shu and the Hou Han-shu,
until after the collapse of Greek rule in the Oxus valley. The Indian texts that
refer to the Yavana (i.e. Greeks or westerners) are not truly historical in nature,
and cannot easily be interpreted. The archaeological record also has gaps. Exca-
vations were not undertaken in this field until relatively recently and, except for
Begram and Taxila, date from after the Second World War. Very few have
deliberately focused on this historical period (Ay Khanum, Charsadda and
Shaikhan Dheri).' Although the data they have provided are neither as abundant
nor varied as we would wish, these excavations have nevertheless enabled us to
form an idea of the civilization created by the Greeks of Central Asia. O u r
knowledge of the political history of the states they founded continues to be
based almost entirely on the study of the coins they issued. From these, numis-
matists and historians have been able to reconstruct - not without uncertainties
- the sequence of various reigns and their approximate duration, as well as the
location and boundaries of different kingdoms.

See Map 3.
1. For such reconstructions of the history of the Greek kingdoms of Central Asia, the basic
works that cover the whole period are Tarn (1951), Narain (1957) and now Bopearachchi
(1991). The contributions of Rapson (1922) and, above all, Macdonald (1922) should not
be overlooked either.
 P. Brrnnrd

As we have already noted in Chapter 3 it was towards the middle of

third century B.c., with the death of Antiochus I1 (246 B.c.), that the final break
came between the Seleucid Empire and its Central Asian possessions. The satrap
. ..
of Bactria, Diodotus, took the ~nitiativcin transforming these territories into an
independent kingdom. Bactria was its vital centre, around which gravitated sag-
diana in the north, Margiana in the north-west and Aria in the west. In the east,
the Greek presence reached its farthest limit o n the Syr Darya, where the site of
Alexandria Eschate (subsequently refounded as Antioch of Scythia) has been
identified beneath the medieval layers of the Khojand citadel. It is doubtful
whether the Greek armies ever entered the Tarim basin; the 'Seres' and 'PhryniJ
referred to in a passage of Apollodorus of Artemita (Strabo X I . l l . l ) must be
regarded as 'neighbours' of the Greeks in the broad sense of the word, because
they lived outside their sphere of influence. For half a century, under the first
three Graeco-Bactrian kings, Diodotus I, his son Diodotus 11, and Euthy-
demus I, the Graeco-Bactrian kingdom was confined to the north of the Hindu
Kush. During this period, however, the kingdom consolidated its position and
gained power as a result of the wealth of its land, particularly in Bactria.
According to Apollodorus of Artemita (Strabo X I . l l .I), the fertility of Bactria's
soil created for the Graeco-Bactrians the power that led them to undertake the
conquest of India. Recent studies o n the development of irrigation during the
period of Greek rule confirm that agricultural production was the basic factor
. ~ country's prosperity under Euthydemus I is
in its economic e x p a n s i ~ n The
directly reflected in the abundance of his coins. His issues of silver and bronze
coins were by far the commonest found at A y Khanum, after those of the Seleu-
cid period.' It was no doubt this economic affluence, with a wise administration,
that ensured the allegiance of the Greek colonists and local nobles. Without
this, Euthydemus would not have been able to hold out for two years against
the Seleucid king, Antiochus 111, when he launched his campaign to reconquer
the lost eastern provinces, defeated Euthydemus' cavalry on the banks of the
Arius (Hari-rud) and besieged him at Bactra. Euthydemus' stubborn resistance,
and the threat posed by nomads seeking to turn the conflict t o their own advan-
tage, forced Antiochus I11 to abandon the siege and t o acknowledge Euthyde-
mus' independence. After this abortive attempt to win back the Greek prov-
inces, Central Asia disappeared for ever from the political horizon of the
Seleucid kings. Antiochus 111 took the southern route back to the west, stop-
ping on the way to renew the treaty of friendship concluded a century earlier by
his ancestor Seleucus I with Candragupta, which recognized Mauryan sover-
eignty over the lands lying between the Indus and Helmand rivers.
It was during the period just mentioned that there took place the first of a
long series of power struggles between ambitious rivals which were to punc-

2. Gentelle, 1978; Gardin and Lyonnet, 1978/79.

3. Petitot-Biehler, 1975; Holt, 1981; Bernard, 1985.
 The Grcck kingdoms of Ccntral Ash

tuate the history of the Greeks of Central Asia and which, by dividing their for-
ces, contributed t o their downfall. It was by assassinating the legitimate sove-
reign, Diodotus 11, son of the founder of the Graeco-Bactrian kingdom, that
~ u t h y d e m u shad ascended the throne.
Under Euthydemus 1's son, Demetrius I, a movement of expansion began
towards the territories south of the Hindu Kush and ancient north-west India.
The founding of a Demetrias in Arachosia indicates that this highly Hellenized
province (see Chapter 3), and probably Drangiana as well, had by that time
become part of the Graeco-Bactrian Empire. The kings who followed pushed
the conquest towards India; but the history of its various stages has given rise t o
many different theories. The most famous of these kings was Eucratides (c. 171-
145 B.c.), whose brilliant career as a military leader led him to be compared to
Mithradates the Great, his contemporary. After a fierce power struggle, he
wrested power from Demetrius and went campaigning in India. The exceptional
personality of Eucratides is suggested by the originality of his coinage, which is
full of iconographic innovations, by the creation of his own specific era revealed
by an inscription at Ay Khanum4 and by his grandiose schemes t o embellish the
palace there. His outstanding career met with a tragic end. H e was assassinated
by his own son who desecrated his father's body.

The Indian campaigns

It was during the reign of Menander (150-135 B.c.), one of the few Indo-Greek
sovereigns t o be remembered in the classical tradition, that Greek rule spread to
its farthest limit and included the greater part of the Panjab as far as the banks
of the Ravi. If the Indian texts are to be believed, the Greek armies penetrated
deep into the Ganges valley as far as Madhyadeia and Magadha. The grammar-
ian Patafijali, when illustrating a particular rule, used two phrases that referred
to the towns of Saketa and Madhyamika being besieged by the Yavana (i.e. the
Greeks). Kalidiisa's drama Mdlavikdgnimitra preserved the memory of a victory
won during the reign of Pusyamitra (184-148 B.c.), founder of the Sunga
dynasty, by the Indian armies over the Greeks on the banks of the Sindhu
(probably a tributary of the Chambal before the latter flows into the Jamuna).
Lastly there is the Yuga Purdna of the Gargisamhitd, which relates that the
Greeks raided and destroyed P i t a l i ~ u t r a ,the capital of Magadha. Claims have
recently been made that traces of this expedition, led by Menander, are t o be
found in the destruction levels at various sites in the Ganges valley from Hasti-
napura to Piraliputra itself.5 It is unlikely, however, that the Greeks made any

4. Bernard, 1980, pp. 22-7.

5. Sharma, 1980. T h e cl.rronology of the destruction levels o n different sites that the author
associated with Menander's campaigns are insufficiently substantiated t o authorize such
conclusions.
 P. Bernard

permanent settlements in the Ganges valley. The various hoards of Greek coins
that have been found there are t rob ably no more than an indication of the fact
that Greek money was highly ~ r i z e din the regions that traded with the Greek-
ruled Panjab. It is difficult to know whether the Greeks exercised any direct
control over Sind towards the south and the coastline between the Indus delta
and the Gulf of Cambay (see Strabo XI.l.l). In all likelihood, it was not until
the discovery of the monsoon at the end of the second century B.C. and the
institution of fully fledged international maritime trade between Egypt and
India6 that these regions began to be of interest t o the Greeks, who until then
had probably been content with nominal rule over them. According to fie
Periplfis of the Erythraean Sea7 drachms of Menander and Apollodotus were
still in circulation at the close of the first century A.D. in the Broach-Barygaza
region. The presence of these coins there may be explained by the local needs of
trade with the West which was developing at that time.

The last phase

Excessive territorial expansion, which spread the Greek population too thinly
for adequate control, forced the Indo-Greek sovereigns to delegate too much
authority to viceroys, who were tempted to play their own games. This inevit-
ably led to the disintegration of the Greek power in India into a number of
independent principalities. This explains why during the two and a half centu-
ries between Diodotus I and the last Indo-Greek king Strato I1 (A.D. 10) the
names of more than thirty kings have been recorded. Some of them bore the
same name (much to the consternation of numismatists). Apart from those
already cited, the most notable kings were: Apollodotus I1 (as opposed to Apol-
lodotus I), whose coins were still in circulation in the first century A.D. in the
coastal region of Broach-Barygaza; Antialcidas, whose name appears on an
Indian monument - the votive column at Besnagar - set u p by one of his ambas-
sadors to the court of a Sunga king; and Strato I, whose reign lasted for several
decades at the end of the second century B.C. Most of these kings reigned exclu-
sively south of the Hindu Kush, for the territories north of the mountains had
slipped out of Greek hands by the third quarter of the second century B.C.
The northern border of the empirc had been weakened by the expansion
into India, the multiplication of centres of power, and the struggles between
rival factions. It was the first to receive the shock waves of nomadic peoples
migrating from the north-western regions of China, one following on the heels

6. T h e basic study is by Dihle, 1978.

7. For arguments against later datcs f o r T h e Periplus of t h e ErJ,rhracnn Sca, see Dilile, 1965.
 The. G r c ~ kkingdoms of Crntral Asia

of another. The abandonment of Ay Khanum around 145 n.c:., a date that

apparently coincides with the death of Eucratides, was most likely causcd by
the arrival of one of the tribes, called the Yiieh-chih, in castern Bactria. Ilclio-
cles, Eucratides' successor, was apparently the last Greek king to reign in Rac-
tria (c. 145-130 B.C.)."~ then Bactria had also lost the two provinces on its
western flank, which had been invaded by the Parthians. When the Chinese
ambassador, Chang Ch'ien, visited the Oxus valley in 129-128 [LC., he found
the Yueh-chih settled on the northern bank of the river, and in control of thc
southern bank, though they had not yet occupied it. Chang Ch'icn's description
of southern Bactria as 'a region bereft of central power, with nulnerous local
chieftains and little armies of poor military value' seems t o apply to a country in
which the political structures created by Greek colonization had already dis-
appeared. The Yueh-chih occupation of the southern bank was completed
around 100 B.c., as related by the Chinese chronicles, the Hun-shu and the Hou
Hun-shu. West of Bactra, the former Greek territories were seized by another
nomadic tribe, possibly the Sacaraucae. The Greek principalities south of the
Hindu Kush enjoyed 100 years' respite before they too gave way t o the new
influx of nomadic tribes. Having been expelled from the high valley of Ili, the
Sakas crossed Chinese Turkestan and the Pamirs and descended through the
Gilgit and Swat valleys into Gandhiira where, around 85 B.c., under the leader-
ship of a chieftain named Maues, they occupied Taxila. The remaining Greek
possessions in Gandhiira were divided into two parts, which eventually disap-
peared. In the principality of KipiSa-Begram, the last Greek ruler was Her-
maeus, w h o succumbed around 70 B.C. to the attacks of other nomadic tribes
who had taken the western route around Bactria and conquered, successively,
Sistan and Arachosia. Further t o the east, between the Chenab and Sutlej rivers,
the Greek power survived, with Strato 11, until A.D. 5-10.

Graeco-Bactrian sites

There is n o shortage of sites where finds of coins, pottery or other artefacts

indicate the presence of settlements dating from the Hellenistic period: at
Gyaur-kala (Alexandria in Margiana);y at Afrasiab (Maracanda in Sogdiana);'"t
Khojand (Antioch in Scythia);I1in Bactria at Termez;12 at Kobadian;I3 at Takht-i

8. O n the question of Greek enclaves, thought to have existed in Bactria later than the
reign of Heliocles, see page 126 below.
9. Filanovich, 1974.
10. Fila~iovich,1973; Shishkina, 1969, 1975, 1985.
11. Negn~atovand Belyaeva, 1977, p. 569; Belyaeva, 1979, 1980.
12. Kozlovskiy and Nekrasova, 1976.
13. D'yakonov, 1953.
 P. Bernard

Sangin;" at Tepe-i Dinistan;15 at Emshi-tepe;lb at Tepe Nimlik;" at Dilberjin;li

and at Bactra itself."
Again to the south of the Hindu Kush range: at Begram,?O at Kandaha+l
and at Taxila ( S i r k a ~ ) .There
~? are also sites where a town plan is still visible at
ground level, which shows the characteristics of this period: at Herat," Taxila
(Sirkap),zi Charsadda and Shaikhan DheriZ5There is even a place where an
ancient name has survived through the centuries as a testimony of a Greek pres-
ence down to the Timurid period: a crossing o n the Oxus was called 'PardagwiJ
from its Greek name pandocheion ( h ~ s t e l r y ) . ?However,
~ the thickness and
number of layers dating from later ~ e r i o d spresent an obstacle to the extensive
excavation of deep layers, so that in most cases the vestiges of the Greek period
in Central Asia consist of just a few sections of walls. A y Khanurn in northern
Afghanistan represents a fortunate e x c e p t i ~ n . The
~ ' remains of a Greek town are
still visible there at g o u n d level, as the site was never reoccupied after it had
been abandoned by the Greeks.?#It has therefore become the site of extensive
excavations, which for the first time provided an overall view of an urban layout
of this period.

The Greek settlements

The presence of colonists of Greek extraction is clearly established at Ay
Khanum by some fifteen proper names.29 Some like Hermaeus, Hippias,
Callisthenes, Cosmas, Niceratus, Philoxenus, Philiscus, Sosipatrus, Strato,

14. Litvinsky and Pichikyan, 1979, pp. 89-109, 1981a, 19816; Pichikyan, 1980.
15. Denisov, 1975.
16. Kruglikova and Sarianidi, 1971, pp. 163-7.
17. Schlumberger, 1947, pp. 241-3.
18. Kruglikova, 1974, 1976, 1979; Kruglikova and Pugachenkova, 1977.
19. Dagens et al., 1964; Gardin, 1957.
20. Ghirshman, 1946.
21. AS, 1978, pp. 9-66, 1979, pp. 1-8.
22. Marshall, 1951; Ghosh, 1947/48.
23. Lezine, 1963/64.
24. Marshall, 1951; Ghosh, 1947/48.
25. Wheeler, 1962; Dani, 1965/66.
26. Minorsky, 1967. E. V. Rtveladze (1977, pp. 182-8) has identified the site with
Surob-kurgan-Kampyr-tepe, some 30 km west of Termez.
27. Bernard et al., 1973; annual reports in CRAI, 1965-72, 1974-76, 1978, 1980; ~ e r n a r det
al., 1976, 1980. C o ~ n p l c t ebibliography u p to 1981 in BCH, No. 1, 1982, p. 23.
28. Immediately after the Greek I-ule the site of Ay Khanum was briefly reoccupied by local
populations but this reoccupation did not result in substantial modification of the Greek
buildings, except for the partial destruction of thc palace.
29. Rapin, 1983, pp. 3 15-72; Grenet, 1983, pp. 373-81.
 The Greek kingdoms of Central Asia

Theophrastes, Timodemus, Zeno and Isidora are common to the entire Greek
world. Among these families, many must have come from Greek Asia Minor
and particularly from the Meander valley, like King Euthydemus, who had
migrated from the town of Magnesia ad Meandrum. The Meander valley con-
nection is further suggested by a statuette discovered at Takht-i Sangin which
represents the River Oxus as Marsyas playing the double flute, the iconographic
form in which Greek art depicted the Meander's main source. Northern Greece
was the other main source for Greek colonists as indicated by a group of names
characteristic of that region (Kineas, Molossus, Triballus) and Macedonia in
particular (Lysanias). Most of the persons whose names have come down to us
were palace officials. But as in the rest of Hellenized Asia, many of the colonists
would have been landowners who lived off the tracts of land (kleroi) allotted to
them when they first settled there. At Kandahar, the name has been preserved of
a certain Aristonax,-'O who belonged to one of the Greek families of Alexandria
in Arachosia to whom the imperial edicts of the Indian king ASoka were
addressed.
It is known from the classical authors that Greek colonists were not the
only inhabitants of the cities of that time, either new or ancient, but that in
some towns at least they lived alongside the indigenous population. This was
particularly true of Alexandria ad Caucasum (Arrian IV.22.5; Diodorus
XVII.83.2) and Alexandria Eschate (Arrian IV.4.1; Quintus Curtius VII.6.27).
This information was confirmed by the excavations at Ay Khanum, where the
names from inscriptions and graffiti also reveal the presence of Bactrians bear-
ing Iranian names (Oumanes, Xatranos, Arixares) who must have lived within
the city itself. Some of them were even officials at the palace treasury (Oxyba-
zos, Oxeboakes, Aryandes), a further indication that the Greek colonists had
managed to achieve a certain symbiosis with the local population. There is,
however, no doubt that the government of the cities lay in the hands of the
Greek communities. It is significant in this respect that at Ay Khanum, officials
of local extraction do not appear to have occupied the highest posts in the hie-
rarchy. The great bulk of Greek colonists arrived during the Seleucid rule, as
must also have been the case for the Near East generally,3' and the deeper Helle-
nization of the Oxus valley occurred with this second wave of settlers.

Greek and the local language: epigraphic documents

The colonists had preserved intact the vehicle for their cultural identity, namely
their ancestral language and script. The evidence ~ i e l d e dby the AJ- Khanum

30. Fraser, 1979, p. 10.

31. An example under Antiochus I is the founding of Antioch in Persia by colonists from
Magnesia ad Meandrum, Orientis Graeci Insmiptiones Select/re, (1903-05), 233, 1. 15.21.
excavations ranges from simple names inscribed 011 vases to elaborate inscrip-
tions cut in stone. There are now four examples of the latter at Ay Khanum,not
to mention two papyri and some thirty brief financial records inscribed on
vases. T w o further inscriptions are known from other sites in Bactria (Takht-i
Sangin and Jigs-tepe) and three others in Arachosia.
After extensive excavations, carried out in different areas of the site, it ;,
indeed surprising that, except for the economic graffiti, so few inscriptions have
been found at Ay Khanum compared with the wealth of texts one would expect
to find on the site of a Hellenistic city in the Mediterranean area. While this
may possibly be explained by the element of chance encountered in all excava-
tions o r by the pillage of the ruins, the real reason can more likely be traced
back to the nature of the town itself. Ay Khanum was essentially a royal city
whose administration, centred around a palace, was probably not very condu-
cive to the type of honorific epigraphy that flourished in the relatively autono-
mous cities of the Mediterranean kingdoms and normally furnished the bulk of
inscriptions. The inscriptions that have come down to us are nevertheless suffi-
cient to show that the Greek language used by the colonists of Central Asia
does not contain the slightest hint of barbarization. This is seen as well in the
simple stereotyped administrative formulae,j2 dedication^,)^ funeral epitaphs,j~
some versified according t o the rules of Greek traditional prosody, or even phi-
losophical texts such as the As'okan rock edicts discovered at Kandahar. Here
the translator demonstrates an intimate knowledge of the Greek philosophical
language into which, thanks t o a keen sense of proper equivalent, he was able to
transpose into Greek the concepts of Indian Buddhism. The lettering itself fol-
lows the same evolution as that of characters in current use in the Mediterra-
nean area. This applies to both the cursive capital used for economic inscrip-
tions on vases in the ~ a l a c etreasury at A y Khanum and t o the various types of
lettering cut in stone. In some cases these remain close to the cursive capital
lettering but in others are more ornate. The language of the colonists was by no
means a withered bough but rather a flourishing branch constantly irrigated by
the sap of close contacts with the mother tongue. These contacts were fostered
not only by the ~ o l i t i c a lties that existed while the provinces of Central Asia
still belonged to the Seleucid Empire, but also by the constant circulation of
men of all professions, and by the penetration of ideas and literary texts.

32. Vase inscriptions recording sums of money rcceivcd and the storagc of various goods at
the treasury of Ay Khanum (Rapin, 1983, 1992; Grenet, 1983).
33. T h e votive inscription to Hermes and Heracles at the gymnasiu~nof Ay Khanum
(Robert, 1973, pp. 207-11); the votivc inscription of Clcarclius on the same site (see
note 36 below); also the dedication t o the god O x u s at Takl~t-iSangin (Litvinsky and
Pichikyan, 1981a, pp. 202-4); and the dedication at Kandahar (Frascr, 1979, pp. 9-21).
34. Examples of funerary epitaphs arc t w o unpublished inscriptions at Ay Khanurn, one in
verse, and a funerary epigram at Jiga-tepe. ( O n the latter, scc Pugachcnkova, 197%
pp. 74-5; Kruglikova, 1977, p. 245, Fig. 16.)
 The G'rec~kkingdoms of Central Ash

These included a philosophical treatise of the third century n.c:. and the frag-
ment of a poem, one written on a papyrus, the other on a parchment, which had
disintegrated but had left the letters in ink printed in the soil in the debris of the
palace at Ay Khanum.I5 Even when around 250 B.C. the Parthian Empire grew
up between the Greek cities of Central Asia and those of the Mediterranean, the
circulation of people and ideas was probably not completely interrupted. The
parthian sovereigns were well disposed towards the Greek communities in their
own territory, as is demonstrated by the title 'Philhellenic' which figures on the
coins struck by Mithradates I (171-138 B.c.) and his successors. Moreover, the
military expedition of Antiochus I11 in the late third century u.c. provided an
opportunity for Graeco-Bactrian Hellenism t o reinforce direct contact with the
source of its national traditions. Antiochus I11 was obliged t o stay in Bactria for
two years with his army and, very probably, with all the artists and men of let-
ters who used to make up the retinue of a Hellenistic king.
Apart from the two inscriptions at Kandahar, in which ASoka addressed
his Greek subjects in their own language, the most revealing text concerning the
Hellenism of the Greek colonists is undoubtedly the inscription discovered at
Ay Khanum in the heart of the city inside a funerary m o n ~ m e n t . The' ~ philos-
opher Clearchus of Soli, a well-known figure in the history of the Aristotelian
school, o n a visit from Greece around 275 B.c., had set up in this monument a
copy of the famous Delphic maxims to serve as a code of good conduct. The
maxims were a collection of some 150 aphorisms kept on display at the sanctu-
ary of Apollo at Delphi embodying the ideals of Greek life. The stone stele on
which they had been inscribed at Ay Khanum has disappeared, but the base on
which it stood has survived. It preserves not only Clearchus' own dedication,
but the final maxims of the series, inscribed on the base for lack of room on the
stele itself. These read as follows: 'In childhood, learn good manners; in youth,
control thy passions; in middle age, practise justice; in old age, be of good coun-
sel; in death, have no regrets.' The fact that the municipal authorities allowed
this moral and civil code - the quintessence, as it were, of Greek wisdom - to be
displayed publicly in the centre of town provides a striking example of the
determination of the Greek colonists to place their city under the protection of
the traditional patron deity of the Greek colonization and t o remain faithful to
their national heritage.
The site at Ay Khanum has also yielded two non-Greek epigraphic docu-
ments. The first is an ostracon written in Aramaic script noting various pay-
ments. The rudimentary form of the text, devoid of inflections and syntactic
markings, makes it difficult t o decide whether the language is Aramaic, that is, a
continuation of the official language of the Achaemenid government, or some

35. CRAI, 1978, pp. 456-8; Rapin, 1992, pp. 115-23.

36. Robert, 1973, pp. 21 1-37.
 P. Bernard

local Iranian dialect. Even more enigmatic is the inscription engraved on a silver
ingot discovered in an archaeological context dating from a brief post-Greek
reoccupation of the site.I7 Both its language and its script, which suggests ~~~i~
letters, are unknown. This tantalizing text might possibly represent the language
of the nomadic invaders.

Towns and urbanization

Under Greek rule Central Asia experienced such unprecedented urban growth
that its fame as 'the land of a thousand cities' spread t o the West (Strabo ~v.1.3;
Justin XLI.1 and 4). The cities were necessary instruments in the process of col-
onization, fulfilling many different roles. They served as: military bases; ad-
ministrative centres (multiplied by dismembering the Achaemenid satrapies);
economic centres of the various regional units; communication centres and trad-
ing posts along the international and local trading routes; and, not least, cultural
centres diffusing Greek traditions. Some of the new towns were built entirely
from scratch o n virgin soil such as A y Khanum3%nd Alexandria ad Caucasum,)9
often in the vicinity of a previous local settlement which could be abandoned
once the new town was settled. The Indo-Greek city of Taxila (Sirkap) replaced
the Bhir Mound, which dated from the Achaemenid and Mauryan period,'O and
Puskalavati-Peucelaotis was moved under Menander from Bala Hissar (Char-
sadda) to Shaikhan Dheri." Other new towns were built around a pre-existing
fortified site (Antioch in Margiana around Erk-kala,42and Bactra around a pre-
vious citadeld3).Others were built o n the site of the former town itself with
Greek ramparts superimposed on those dating from earlier periods such as
Alexandria in Arachosia (Kandahar)" and Maracanda ( A f r a ~ i a b )In .~~
the latter
case of existing cities reoccupied by Greek colonists, the archaeological evidence
is too scanty to give us an accurate idea of the impact of the Greek settlements,
but at least we know from the excavation at Ay Khanum how an entirely new
Greek city, built on virgin soil, appeared.
As in the case of Peucelaotis (Shaikhan Dheri), the founders of Ay

37. Bernard et al., 1980, yp. 27-9; Rapin, 1992, pp. 139-42.
38. For the town plan and architecturc of Ay Khanum, scc thc studics citcd in note 27
above.
39. Ghirshman, 1946.
40. Marshall, 1951; Ghosh, 1947148.
41. Wheeler, 1962; Dani, 1965/66.
42. Filanovich, 1974.
43. Dagens et al., 1964; Gardin, 1957.
44. AS, 1978, pp. 9-66, 1979, pp. 1-8.
45. Filanovich, 1973, pp. 85-94; Shishkina, 1969, 1975, 1985.
 The Greek kingdoms of Central Asla

~ h a n u m ~ ~ advantage
ook of thc natural dcfences provided by thc conflucncc of
two rivers, the Oxus and one of its southern tributaries, the Kokcha. A natural
hill, some 60 m high, closed off the third side of a vast triangular arca rncasuring
1.8 x 1.5 km. The site, with its acropolis formed by thc flat top of the hill and its
lower town laid out between the hill and the two rivers, was perfectly suitcd to
the needs of the Greek town-planners. The natural dcfences were strcngthcned
by a solid rampart of unbaked brick, which ran around the entirc perimeter of
the town, skirting the banks of the rivers and following the outer edge of the
acropolis. Special care was taken to ensure that the northern tip of the lower
town was particularly well fortified, for it was there that thc town lacked nat-
ural defences. At this vulnerable point, the wall was built 7 m thick with rec-
tangular towers (19 x 11 m).+' This type of massive rampart built solid through-
out, where defensive action took place exclusively on the top of its towers and
curtains, and whose effectiveness resided mainly in the height and mass of its
masonry capable of withstanding the assaults of siege machinerv, was characte-
ristic of the Greek period in Central Asia.48The same kind of rampart is found
at Gyaur-kala, Begram and also at Sirkap, where the Indo-Scythian wall prob-
ably imitates a Greek rampart not yet discovered. O n the contrary the ramparts
of Maracanda with their hollow curtains represent the continuation of a local
tradition of fortification. Within the city walls a citadel, generally built up
against the rampart, provided the town with a last refuge in case of an assault.
The citadel at Ay Khanum was built in this manner in the south-east corner of
the acropolis.
At Ay Khanum most of the buildings were concentrated in the lower
town, which was less windswept than the acropolis and could be supplied with
water by a branch of one of the canals on the plain (see Chapter 3, Fig. 3). The
plan of this lower town does not conform to the traditional Hellenistic grid pat-
tern such as may be seen, in a simplified form, at Taxila (Sirkap) and at Peuce-
laotis (Shaikhan Dheri), where the layout of the town is divided by ~ a r a l l e l
streets at right angles to a main thoroughfare. The particular features of the
urban layout at Ay Khanum result from the character of the city itself, which
was the seat of a royal palace. T o make room for the ~ a l a c eand avoid too close
a proxin~itywith other buildings, the main street extending across the lower
town was diverted along the foot of the acropolis on a raised strip of ground
that separated it from the lower town. The ~ a l a c e(see Chapter 3, Fig. 4) was
thus able to spread out across the entire width of the lower town in its southern
half, so that it covered an area of 350 x 250 m. The only section where ~ a r a l l e l
streets are to be observed is in the area of residential mansions at the south-west

46. For the town plan and architecture of Ay Khanum, in addition t o the studies cited in
note 27 above, see also Bernard, 1976, pp. 245-75, 1981, pp. 107-20.
47. Leriche, 1986.
48. Francfort, 1979, pp. 23-30.
corner of the town where the rivers meet. The overall plan of the city was there-
fore dictated by the special role it was meant to play.

Architecture: the palace at A y Khanum

As the Greek architects had n o prototype of their own to imitate for the designof
the palace at Ay Khanum, they drew their inspiration largely from the Nee-
Babylonian and Achaemenid models they had seen when they first entered the
region with the Greek armies. Like the Achaemenid palace of Susa, the palace at
Ay Khanum consists of a massive conglomeration of courtyards (see Fig. 1) and
buildings for official, residential and economic use. The main courtyard, through
which the palace was entered from the north, struck visitors by its imposing site
(137 x 108 m) and by the rows of stone columns, crowned with Corinthian cap-
itals, which formed the fagades of its four porticos. Behind the southern portico,
a vestibule with eighteen Corinthian columns three rows deep, reminiscent of the
spirit of some Achaemenid architectural compositions, provided a monumental
entrance to the palace itself. At the western end of the palace was a second court-
yard less imposing and of a more private character with its four porticos lined
with sixty Doric columns. Among the buildings for official use in the south-east
corner, one in particular is remarkable for its size and plan. It is a huge square
50 m each side, divided by two corridors at right angles into two pairs of similar
units. In each pair the eastern unit features an audience hall decorated with pil-
asters topped by painted capitals, while the western unit is composed of office
rooms. The south-west corner of the palace is occupied by three sets of private
apartments recognizable as such by the presence of forecourts, kitchens and
bathrooms. West of the great northern courtyard lies the treasury, composed of a
series of store-rooms grouped around a central courtyard. The purpose of the
building is clear from its layout and the artefacts found there. These include stor-
age vessels, debris of precious stones (agate, onyx, carnelian, rubies, garnets, lapis
lazuli, turquoise, beryl and pearls), both worked and unworked, and inscribed
vases which once contained the cash reserve of the palace. Judging by the Indian
coins and remains of precious objects from India that have been found there, it is
possible to suggest that the palace, in its final and most monumental stage,
described above, may have belonged to King Eucratides, who is known to have
made conquests in India; and Ay Khanum may well have been Eucratidia, the
city that was named after him.

The originality of Graeco-Bactrian architecture

The palace at Ay Khanum typifies the character and originality of Graeco-
Bactrian architecture; its walls were built of unbaked brick, sometimes on a
baked-brick base. The roofs were flat and made of earth, as in all Oriental archi-
 FIG;. 1. Courtyal-d in the north of t h e palace at A \ Khanurn.
Remains of tlic south colonnade.

tecture, but o n the main buildings one o r two rows of Greek-st!-le tiles were
added t o the roof as a border. T h e use of stone was reserved for doorwavs and
architectural supports. T h e base and drums of the colulnn were cut o n a kind of
lathe which ensured rapid and standardized results. While the layout of the
buildings was largely inspired by Iranian and Central Asian architecture, the
dCcor remained faithful t o Greek taste, making use of the three classical orders
of column (Doric, Ionic and, particularly, Corinthian), as well as of decorative
terracotta antefixes, generally with the Greek palmette to line the edge of the
roofs. T h e buildings exhibit a sense of the grandiose sometimes overdone, an
effective usc of repetition, a taste for s~rmmetr!~ that verges on the mechanical,
tirelessly playing with parallel and ortllogonal axes. They show a practical imag-
ination capable of designing the simplest and most functional solutions but
lacking a sense of beauty and delicacy. All of this goes t o make the architectural
style of the palace typically imperial - powerful, proud and cold.
'The buildings of AY Khanurn, the gymnasium (Fig. 2) and the
thcatre, answered the needs of a population leading a t\.picall\, GI-cck litc. The
g y m n ~ s i u m \vllich
, w ~ dedicated
s to hot11 the intellectual ~ n ph\.sical
d aspects of
Greek cducation, ,lnd thus constituted the most effecti\.e instrument for the
 FIG.2, General view of the gymnasium of dy Kharmm.

diffusion of Hellenism, was composed of courtyards and buildings that

stretched along 350 m of the bank of the Oxus. Its northern building, probably
reserved for teaching, covered a square of 100 m by 100 m. While the basic con-
cept of its plan adheres t o a typical Greek gymnasium (a courtyard surrounded
by various buildings and porticos), it is remarkable for several distinctive fea-
tures: its considerable size, the dogged symmetry of its architectural composi-
tion, each side of the courtyard being occupied by a colonnaded porch flanked
by two long rooms, and the apparent lack of differentiation of the rooms.
The theatre, built against the inner slope of the acropolis, spread the fan of
its unbaked-brick tiers over slightly more than a semi-circle, with a radius of 42 m
and a height of 17 m. Its seating capacity of about 5,000 is somewhat greater
than that of the only other Hellenistic theatre so far excavated in Hellenized
Asia at Babylon, and slightly smaller than the famous theatre at Epidaurus in
Greece. The presence of royal boxes set half-way up the tiers, a feature un-
known in Greek theatres, indicates a society in which differences in social status
were more clearly marked and where the democratic ideal so cherished, even by
Greek cities under royal rule, was already seriously weakened. There can be no
doubt, however, that the repertoire of Greek plays was performed there. Indi-
rect but indisputable evidence of this is provided by one of the carved spouts of
the Oxus fountain, which represents the traditional comic mask of the
 7'he Greek kingdoms of'Ccntra1 Ash

slave It is, therefore, quite probable that the colo~listsof Ay Khanurn,

like their Mediterranean cousins, were familiar with the Greek new comedy,
and in articular with the plays of the most famous Greek comic writer,
~ e n a n d e r .A passage in Plutarch (De Alexandri fortuna aut virtutc, 328 I).),
referring to the fact that in the parts of Asia conquered by Alexander, the chil-
dren of Persia, Susa and Gedrosia were learning to recite the tragedies of Sopho-
cles and Euripides, confirms, in spite of rhetorical exaggeration, the spread to
Asia of that typically Greek literary genre and social phenomenon: the tlieatrc.
Another kind of construction at Ay Khanum that is equally characteristic
of the Greek urban landscape is a stone fountain, decorated with carved spouts
and fed by underground streams situated at the foot of thc ramparts along the
bank of the 0 ~ ~ sAy . ~Khanum
' possessed an arsenal, set like the theatre on the
edge of the main street, at the foot of the acropolis. It was a vast edifice measur-
ing 140 x 110 m, with store-rooms grouped around a central courtyard. Its pres-
ence and size emphasize the role of Ay Khanum as a military base on the east-
ern marches of Bactria.
At Ay Khanum, as in other Greek towns, the dead were buried outside
the city walls and the families had their mausoleums made up of several vaulted
chambers arranged on either side of a central passage. There massive rectangular
structures of unbaked brick, half submerged in the earth, represent a kind of
funerary architecture unknown in the Greek world and that appeared here for
the first time in Central Asia. In accordance with Greek custom, honorary
burials were allowed within the city walls. This privilege was granted t o bene-
factors of the community so that their memory would be for ever present
among the living. T w o such mausoleums in the shape of small Greek temples
were discovered near the entrance to the ~ a l a c e The . more monumental of the
two contained an underground stone vault and was ~ r o b a b surrounded l~ by a
row of columns. The more modest, which was also the more venerable because
it contained the mortal remains of Kineas, one of the city's founding fathers,
had a simple falade with two wooden columns.

Domestic architecture
It was, paradoxically, in domestic and religious architecture - the two types
most directly involved in the personal life of the citizens and which ought,
therefore, to have been the most conservative - that we encounter the most
far-reaching innovations. The traditional Greek house had a central courtyard
around which the living room and service quarters were arranged. This was
replaced by a house with a courtyard in front of the body of the building and

49. CRAI, 1976, pp. 310-13, Fig. 18.

50. Leriche and Thnraval, 1979, pp. 171-205.
 P. Bernard

with the building itself firmly centred around the main living-room. A peripheral
corridor set off the living-room from the other rooms that formed a horseshoe
around it. This living room opened out into the front courtyard through a two-
columned porch. This layout subordinated all other rooms to the main living-
room, which became the focal point of the architectural composition, while the
courtyard acted as a kind of private annexe. In this plan we may have evidenceof
a hardening of the hierarchical relationship between the master of the house and
his subordinates. Although the houses were radically different from those of the
Mediterranean, they retained a typically Greek feature - the bathroom - and an
even more important place was reserved for it than in their Mediterranean
counterparts. Constructed with particular care, the bathroom consisted of two
or, more often than not, three complementary rooms, which led in a row from
one to the other. The floors were of flaptones o r mosaics and the walls plastered
with red stucco. There were usually a dressing room, a bathing room, where one
could take a shower, and a water supply with cauldrons, from which one drew
the hot and cold water. This type of domestic architecture was particular to the
patrician mansions in the southern quarter of the town and its northern suburbs,
and to Bactria in general. It probably incorporates elements borrowed from local
domestic architecture, and is unknown south of the Hindu Kush, where the few
Indo-Greek houses uncovered at Taxila (Sirkap) reflect the principles of the
traditional Greek plan with a central courtyard that is found in the Mediterra-
nean and throughout Western Asia.

Religions and religious monuments

In the absence of texts, coins in general and the religious monuments discovered
at Ay Khanum are practically the only sources known about the religion prac-
tised by the Greeks of Central Asia. With very few exceptions, the official state
pantheon was entirely Greek, as illustrated by the images o n the coins which
depict its various gods in association with the reigning m ~ n a r c h . Among
~' the
relatively small range of deities most frequently represented, we find Zeus,
Poseidon, Apollo, Heracles, the Dioscuri, Artemis and Athena (the latter por-
trayed mostly in her typically Macedonian form as Athena Alkidemos), as well
as Nike and Tyche, personifications of victory and good fortune. Examples of
cross-influences with local divinities are rare, f o r example: the crown of radiat-
ing spikes that surrounds the head of Artemis, perhaps suggesting the halo of
light worn by the Iranian goddess Anahita; the Persian cap worn by Zeus-
Mithra, also surrounded by rays of light, o n the coins of Amyntas and Her-
maeus; and lastly, the wheel, the Indian symbol of universal kingship, found on
one lone copper coin of Menander. The significance of the use of the bull and

51. For coil1 types, see I,ahiri, 1965; Mitchincr, 1975.

 The Greek kingdoms of Central Asia

the elephant on the coins remains ambiguous. Both these animals arc as
common to Greek symbolism (they are featured on Seleucid coins) as they arc
to Indian (where they are the animals sacred to Siva and Indra). Agathocles, one
of the first kings t o penetrate into the Panjab, is unique in that, on the coinage
minted in his Indian dominions, an important place was givcn to local Hindu
deities: the brothers Krsna (Krishna), holding a wheel, and Balarima holding a
plough, as well as an Indian goddess holding a flower.52
O n the basis of the coinage, one would have expected t o find Greek-style
temples in Bactria. It therefore came as a great surprise that the architecture of
the temples discovered at Ay Khanum owed nothing to Greek tradition. O n e of
the most important, if not the principal, sanctuary of the city, both in terms of
size (60 x 60 m) and location - on the main street, not far from the palace - con-
tained a massive temple 20 m by 20 m raised up on a high, three-stepped base
with its outer walls decorated with indented niches. Inside the temple, a large
vestibule led into a smaller chapel flanked by two sacristies. Opposite the
entrance stood the cult image. Outside the city walls, not far from the main gate,
stood another temple with a closely related plan, also standing on a similar high
podium, with its outside walls similarly decorated with indented niches, but
containing three chapels opening into an open-air vestibule. There was also
another sanctuary at the south-west corner of the acropolis, built around a monu-
mental stepped platform in the open which was clearly used as an altar. This
last place of worship recalls directly Iranian religious sites, where, according to
the descriptions of classical authors, the Iranians worshipped the forces of
nature in high open places, without erecting any statues to personify them. Even
if we were ready to admit that this sanctuary was specially built for the local
population, particularly for the troops stationed on the acropolis, and that the
temple outside the city walls was also erected for a local cult, this supposition
would not apply t o the main temple with the indented niches which was
obviously used by the Greek colonists themselves. The only significant fragment
of its cult image that has survived - a foot clad in a Greek sandal decorated with
winged thunderbolts, carved in an impeccably Greek style - seems to indicate
that the divinity in question, whose identity remains unknown, was ~ o r t r a ~ in
ed
Greek form ( ~ e r h a a~ Zeus).
s However, the burial of votive vases at the foot of
the edifice indicates a ritual unparalleled in Hellenistic religion. The ~ a i n t e d
images of the Dioscuri at the entrance t o the shrine of the temple at Dilberjin in
the ancient oasis of Bactra5>confirm that divinities of Greek origin were wor-

52. Bernard and Audouin, 1974, pp. 7 4 1 . Pantaleon, whose coinage was similar to that o f
Agathocles but less rich, also issued bronze coins depicting a goddess holding a flower.
53. For the oldest level, possibly dating back to the end of the Graeco-Bactrian period, see
Kruglikova, 1977, pp. 407-10. For the painting of the Dioscuri, see Kruglikova, 1976,
pp. 87-93; Buriy, 1976, yp. 110-13. The layout of the sanctuar). of the god Oxus at
Takht-i Sangin is also non-Greek. See Litvinsk? and Pichikyan, 1981a, pp. 197-200 (the
Kushan level); Pichikyan, 1991.
 P. Bernard

shipped in temples built in a purely Oriental style. It is not impossible,

however, that these Greek gods may have been identified with local divinities.
Since such a combination of Greek deities with Oriental temples has been
observed at only two sites, neither of which have been fully excavated, it should
not be set up as a general rule. The undeniably Greek inspiration of the monu-
mental temple of Jandial at Taxila,'.' both in terms of its plan and its decor of
Iollic columns, in spite of its ~ e c u l i a features
r and its probable post-Greek date,
indicates that religious buildings in the Greek style did indeed exist in the Indo-
Greek area. This possibility should not be ruled out for Bactria either, which
was the true cradle of Central Asian Hellenism, and where the ~ u b l i cmauso-
leums of Ay Khanum faithfully preserved the memory of this traditional reli-
gious architecture.

Local cults and Buddhist influence

We have little information regarding the indigenous cults during this period, at
least as far as the lands north of the Hindu Kush are concerned. The Oxus river,
master of fertilizing waters, certainly occupied an important place in local reli-
gious thought. This is suggested by the use of its name in the composition of
personal names, and by the discovery of a statuette at Takht-i Sangin which
represents the Oxus in the form of the satyr Marsyas playing the double flute.jj
This very unusual manner of depicting a river-god suggests a direct link with
the way Greeks used to represent the source of the Meander, the great river in
Anatolia. This in turn leads t o speculation that the valley of the Meander may
have provided Bactria with contingents of colonists during the period of Seleu-
cid rule. At Ay Khanum two naked women figurines carved in bone with exag-
geratedly feminine features, and standing in hieratic frontality, undoubtedly
represent a local fertility goddesss"ather than the purely Greek Aphrodite. At
Takht-i Sangin indirect evidence for the cult of fire, which plays such an import-
ant part in Iranian religions, is found in a personal name, Atrosokes, which
means 'fire-brand'.s7 The monumental altar o n the acropolis at Ay Khanum is
evidence of a non-Greek cult, but its precise nature remains obscure. The docu-
ments concerning Greek territories south of the Hindu Kush offer information
on the relationship between the Greek colonists and Indian religions. Their
impact on the new masters of the territories was felt even in the most exalted
circles of society. Heliodorus of Taxila, ambassador of the Indo-Greek king
Antialcidas, who served at the court of the king of the V i d i S ~region, was a

54. Marshall, 1951, pp. 222-9.

55. Litvinsky and Pichikyan, 1981a, pp. 202-4.
56. CRAI, 1974, pp. 302-3; Francfort, 1984, pp. 15-17.
57. Litvinsky and Pichikyan, 1981a, p. 202; Pichikyan, 1991, pp. 170-2.
 The Greek kingdoms of Central Asia

follower of the cult of Vishnu, as can be seen from the confession of faith that
he inscribed on a votive column dedicated to Visudeva at the site o f Besnagar."
Above all it was Buddhism that penetrated Greek society, and the reasons for
its success can be looked for in the fact that its concepts contained many points
in common with that of the philosophy of Epicurus" and that it had
from the active protection of the Mauryan emperors. The conversion of Menan-
der, the most famous of the Indo-Greek sovereigns, to the 'Good Law' as
described in the Indian work entitled Milindapal?hah'may or may not be a his-
torical fact, but it does bear witness, at least, to Menander's personal sympathy
for that doctrine and more generally to the strength of the message that had
been passed on by the Emperor Aioka to the Greeks in his north-western prov-
inces more than 100 years earlier. Towards the end of the Greek occupation of
Gandhira o r shortly thereafter, a local governor, the Meridarch Theodorus,
dedicated a Buddhist reliquary in Swat." The extraordinary flowering of what is
called Graeco-Buddhist art in the first centuries A.D. would probably never
have occurred had there not been a large number of Buddhists already in the
Greek community of the southern provinces and among its artists.

The figurative arts

In contrast to the many examples of creative originality in architecture, the
figurative arts were, generally, much more dependent on Western models of a
latter-day classicism. The early date (towards the middle of the second century
B.c.) at which the Greek Empire went into decline left little time for the
figurative arts t o be revitalized by the upsurge of inspiration that characterized
the mainstream of Hellenistic art in the second and first centuries B.C. In many
respects, it is even possible to speak of Central Asian Greek art as traditional-
istic, as in the case of the mosaics discovered at Ay Khanum. Instead of being
composed of small stone cubes, which could be laid down in tight patterns to
create skilful effects of depth and colour, the mosaics were made by the old
technique of setting ~ e b b l e sin a bed of c e ~ n e n t .The
' ~ ~ loose spacing of the peb-
bles and the limitation of colours, white for the background and brown-red
with a few isolated touches of black for the designs, reduced the decorative
motifs to simple outlines, while the repertoire itself remained conventional. The
same conservative spirit may be seen in the stone statuary, which was mainly
used in small-size works, such as a woman leaning on a short column,h3a male

58. R a ~ c h a u d h u r i ,1920, pp. 58-6 1.

59. Schlumberger, 1972, pp. 188-99.
60. Foucher, 1951, pp. 289-95.
61. CII, 1929, 2, 1, 1.
62. Bernard et al., 1976, pp. 16-24.
63. CRAI, 1972, p p 628-9, Fig. 15.
nude wearing a crown of lenves,(n4which is a fine anatomical study in the best
Greek tradition (Fig. 3), the bust of an old man set on top of a pillar,"^ the funer-
ary relief of a naked youth, with his cloak thrown over his back6"nd a gargoyle
representing a comic mask.(17The general standard of craftsmanship is high. There
is, however, one important innovation that we owe t o the Graeco-Bactrian
artists. In the execution of large-scale statues and reliefs for decorating the walls
of certain buildings, these artists systematically used and perfected the technique
of modelling raw clay or stucco o n a framework of wood and thin lead wires used
only to a limited extent in the Mediterranean area. With this technique the sculp-
tors were able to develop a more personal style, particularly in the art of portrai-
ture (heads found at Ay Khanum and Takht-i Sangin (Figs. 5 and 6)).6sOther
superb examples of this development of skilled portraiture are seen in the work
the Graeco-Bactrian and Indo-Greek coin engravers and in the bust of a veiled
woman from a terracotta mould (Fig. 4) found at A y Khanum.
The abundance, diversity and quality of the products of the so-called
minor arts are characteristic features of Greek civilization in Central Asia. The
sites of Ay Khanum and Takht-i Sangin have yielded a mass of objects of every
kind - utilitarian, decorative and votive - made from a wide range of materials.
The decorative bowls made of dark schist, carved with simple designs and
encrusted with coloured stones, which were discovered at A y K h a n ~ r n , "are ~
typical of these local crafts. Owing to the proximity of Bactria to India, the
working of ivory in that province was particularly well developed and produced
a wide range of fittings from furniture to arm-fittings. The Greek tradition was
forcefully expressed in this field as can be seen in parts of thrones and beds,7c
sword hilts7' and figurative carvings for sword handles and scabbard endings
such as a head of Heracles wearing his lion scalp head-dress,72an image of the
same Heracles trampling on an adversary7' and a fantastic feminine water deity,
half centaur, half triton, holding an oar. T w o bronze statuettes, one of Heracles
crowning himself,7Jthe other of the god Oxus portrayed as Marsyas playing the
double flute,75exemplify the reworking of Western themes in a provincial style
not without a certain rustic flavour (particularly in the case of the Silenus). It

64. CRAI, 1969, pp. 3 4 1 4 , Figs. 17-18.

65. Bernard, 1967, pp. 90-1, Plates XIX-XX; Veuve, 1987, pp. 91-3.
66. CRAI, 1972, pp. 623-5, Fig. 13. For a foot from the cult ilnagc in the temple with
indented niches, see also page 115 above and CRAI, 1969, pp. 338-41, Figs. 15-16.
67. C R A I, 1976, pp. 3 10-13, Fig. 18; Leriche and Thoraval, 1979, pp. 196 et seq., Figs. 16-18.
68. CRAI, 1969, p. 344, Figs. 19-20; Litvinsky and Pichikyan, 1981a, pp. 204-6, Figs. 9-10.
69. Francfort, 1976, pp. 91-8, 1984, pp. 21-6.
70. Bernard, 1970, pp. 327-43.
71. Pichikyan, 1980, pp. 202-12.
72. Litvinsky and Pichikyan, 19816, pp. 212-15, Fig. 17.
73. Ibid., p. 207, Fig. 12.
74. CRAI, 1974, p. 302, Fig. 13.
75. Litvinsky and Pichikyan, 19816, pp. 202-4, Figs. 6-7.
was indeed in this field of the minor arts that local artists were most liltcly to
play a role, opening the way t o Oriental conceptions. ]:or example, wc have at
A y Khanutn terracotta figurines of a local goddess drcsscd with heavy robes
and weighed d o w n with jewellery, as well as bone statuettes in which a naked
and plump goddess is represented in a hicratic nudity.'' T h e most inlportant
~ c , discovered at
work in this Gracco-Oriental style is a gilded silver p l ; ~ c l ~ also
A y Khanurn, which depicts the Greek goddess of nature, (:vbclc, riding over a
--
rocky terrain in her chariot drawn by lions and driven bj. a winged Victor):

76. C K A I , 1974, pp. 302-5, I:ig. IS; 1:ranctol-t, 1'184, p. 1.5; G u i l l , ~ u m c. ~ n d~ o u ~ c u l l 1987,
e,
p p 00-3.
77. C K A I , 1970, pp. 33'1-17, I:iK. 31; I:I-,inctost, 1'181, pp. '13-104.
1 . 1 ~ 5.
~ . 1Icail ot ,
I S ~ , I c u c ~I\lnS.
d -I CI.11 and J I ~ L L I \ ~ C I .
~c.lcwco\1 (?), T ~ h l ~ tS.lng~n.
(Photo: 6' \ ' I . I J I ~ I'Tcl
I chc111n.)
 P. Bernard

T~~ priests dressed in the traditional robes of the servants of Cybele take part
in the scene, one walking behind the chariot and holding a parasol to provide
shade, the other burning incense on the top of a high stepped altar of an Orien-
tal type. The sun, the moon and a star shine in the sky. The allegorical represen-
tation of Victory, the personification of the sun as the bust of Helios, and the
robes worn by Cybele and the Victory are all inspired by the Greek aesthetic
tradition, but the conventions of Oriental art are deeply felt in the absence of
in the flatness of composition, and in the rejection of three-parter
views in favour of absolute frontal it^ o r strict ~ r o f i l e This
. work nevertheless
remains an exception, and it has t o be admitted that the figurative arts lack the
originality and homogeneity of style that we find in the architecture of the
period, and the Graeco-Buddhist art of G a n d h ~ r aof later times.

Everyday implements
For everyday needs, the colonists relied heavily o n implements created by
Greek technology. Apart from the common grinders to be found in any civil-
ization, heavy grain millstones have been discovered in the colonists' houses.
These millstones, whose area of distribution coincides with that of the Greek
colonization, are of a sophisticated design. The upper grinding stone, cut out to
act as a hopper, was moved back and forth by means of a horizontal lever
allowing the grain to flow automatically from the hopper t o the lower stone
where it was crushed. The wine presses and ink-wells7xwere copied from West-
ern models, and Greek-style sundials with a hemispherical section were used to
tell the time.7yAn equatorial sundial was found in the gymnasium at Ay Kha-
num whose form, though quite distinct from any previous known model, was
nevertheless obviously inspired by the Greek theory of solar clocks." A large
part of the ceramic vessels imitated specifically local shapes (for example, the
tulip bowl in the Panjab and Gandhara area and the cylindro-conical drinking-
cups in the Oxus valley), but new types inspired by Greek models were con-
stantly introduced, such as high-footed craters, fish platters, hemispherical
bowls, carinated bowls, bowls with moulded designs (the so-called Megarian
bowls) or applied designs (the so-called Pergamene vessels), amphorae, pitchers,
etc. Even the g e y i s h black slip of certain series is a deliberate substitute for the
black-glaze monochrome pottery of the Mediterranean areae8'

78. CRAI, 1978, pp. 462-3, Fig. 21, Guillaulnc and Kougcullc, 1987, pp. 47-8.
79. Veuve, 1982, pp. 23-36, 1987, pp. 86-8.
80. Veuve, 1982, pp. 36-51, 1987, pp. 88-91.
81. Gardin, 1973, pp. 121-88; Bernard et al., 1976, pp. 45-51.
l>l(i. 0. 1 lc<\Ciill <I'l\. ,1,,c1 ~11<1~>.15tcl-
\\.it11 13Ll'~tl~i<lll-\t\-lc
l l c < ~ ' ~ - C i~1-Jl,llt-i
l ~ ~ ~ ~ .S < l l l ~ l l l .

(Plioro: s \'I~~ii~iii~-*I~L~~C~~CII~I~.~
 P. Bernard

Trade and trade routes

The distribution of silver coins is a good indication both of their use as inter-
national currency beyond the borders of the countries in which they were
issued, and of the geographical range of that country's commercial activities,
The area in which Graeco-Bactrian tetradrachms are found (mainly of Euthyde-
lnus I and 11, Eucratides I and 11, and Heliocles) reached as far as S y r i a - M ~ ~ ~ .
potamia with finds at Baarin," Susa" and the Kabala hoard in Caucasia.8' Indo-
Greek coins circulated as far as the heart of the Ganges valley at Panchkora.8j ln
contrast, the Western silver coins that reached Bactria were mainly Seleucid (up
to Antiochus 111) o r posthumous issues struck in the name of Alexander from
mints in Asia Minor, Syria and P h o e n i ~ i aA
. ~large
~ hoard of Indian coins, each
stamped with several punch marks, and Indo-Greek drachms of Agathocles was
discovered at A y Khanum" but it probably tells us little about the trade between
Bactria and the Greek provinces of north-west India, because the hoard was
found in the palace treasury, and seems to represent taxes and duties levied in
those regions. The same is n o doubt true of a mother-of- earl plaque whose
decoration made of incrusted coloured glass is typically Indian in style," and of
fragments of agate and onyx used for furniture decoration, which were disco-
vered in the same place, and whose Indian origin is equally indisputable. The
existence of close trade links with the Indian subcontinent is clearly seen in the
widespread use of ivory in local workshops for the production of a range of arte-
facts. Western imports were extremely rare, for local craftsmen were successful
in responding to every kind of demand, and produced items that were tho-
roughly Greek in style. Among the finds from the Mediterranean area which
were made at A y Khanum we might mention scraps of literary papyri,89the
stamped handle of an amphora, unique among hundreds and thousands of
sherd~,~O some fragments of black glazed pottery and plaster casts taken from
metallic vessels of probable Western origin. For obvious reasons, it is even more
difficult to detect Graeco-Bactrian and Indo-Greek exports. It is quite likely that
part of the ivory destined for the workshops of Hellenized Asia came from India
and was transmitted through Bactria. The possibility that the famous carved

82. Seyrig, 1973. (Hoard 28, N o . 3.) A coin of Eucratides has cven been found in a hoard
discovcred in Italy, and another has been found on the northern coast of the Black Sea.
83. Mrarkholm, 1965, pp. 127-56; Houghton and LC Rider, 1966, pp. 11 1-17; Strauss, 1971,
pp. 109-40; LC Rider, 1965, p. 198.
84. Dadashcva, 1976, pp. 106-9.
85. Sharma, 1980, pp. 80-3.
86. Curiel and Fussman, 1965; Petitot-Biehler, 1975, pp. 40-51; Holt, 1981, pp. 12-13.
87. Bernard and Audouin, 1973, pp. 238-89, 1974, pp. 7 4 1 .
88. Bernard et al., 1980, pp. 29-36; Rapin, 1987, pp. 185-232.
89. Ibid., p. 24, n. 35; sce also above, p. 107.
90. Bernard, 1965, pp. 636-9.
 Thc Grcek kingdoms of Central Asia

r h y t o n ~of Nisa in Parthia were produced by Gracco-Bactrian craftsmen cannot

be ruled out." But the first real trade links between Central Asia and China
were established much later than the mission of Chang Ch'ien and lie beyond
the scope of this chapter.
In the third and second centuries KC., the trade we have just mcntioncd
was carried out exclusively by overland routes, for the sea route linking the
north-western coast of India t o the Persian Gulf and Egypt had not yet opened.
The great east-west caravan route was controlled by the Greek kingdoms. Start-
ing at Pitaliputra, it made its way up the Ganges valley t o cross the Panjab,
through Taxila and Puskalivati. Once it had reached Alexandria ad Caucasum
(Begram), it crossed the Hindu Kush to descend into Bactria, and from there, it
veered westward towards Hecatompylos, Ecbatana, Seleucia on the Tigris, or
Antioch or even Asia Minor. At Alexandria ad Caucasum a secondary branch of
this old route forked off south of the Hindu Kush to cross Arachosia and join
the main route again in Aria. The Parthian Stations, a handbook for travellers
by Isidore of Charax written around the beginning of the Christian era,
describes this itinerary from the crossing of the Euphrates to Arachosia. Cara-
vans travelling along these roads halted not only at the great urban and trading
centres, like those mentioned above, but at simple staging posts, which were the
forerunners of the Islamic caravanserais. O n e such staging post located on the
right bank of the Oxus near Termez has been identified thanks to the Iranian
form of its name 'Pardagwi' which has preserved the Greek word pandocbeion,
'hostelry'.y2The river- and sea-trade route that was said t o allow the shipping of
goods along the Oxus t o the Black Sea, via the Caspian and the Caucasian isth-
mus, was never anything more than a grandiose theoretical scheme submitted to
Seleucus I by one of his generals, Patroclus, and based on the mistaken idea that
the Oxus flowed into the Caspian Sea.y3All the merchandise leaving the Oxus
valley and bound for 'Western Asia, even when it was destined to go t o the
Black Sea, ancient Albania or Iberia (Georgia and Azerbaijan), took the land
route through E ~ b a t a n a . ~ ~

Coinage
Numismatics plays a crucial role in our knowledge of the Greek kingdoms of
Central Asia. Indeed, it is through the study of coins that it has been possible to
reconstruct the broad outlines of the history of these kingdoms, and the abun-
dance of their coinage bears witness to the political and economic power they

91. Masson and Pugachenkova, 1959.

92. Minorsky, 1967, pp. 45-53.
93. Tarn, 1951, pp. 486-90.
94. Manandian, 1965, pp. 47-52.
 P. Bernard

once held.'s The Greek coinage of Central Asia, like that of the Seleucids fro,
it originated, was based on the silver standard. Gold was only struck in
exceptional c i r c u m s t a n ~ e s . ~ T hcoinage
e of this period was exclusively royal,
that is to say, it was issued by sovereigns in their own name, even down to [he
issues of bronze coins intended for minor purchases. It may be divided into two
main series that were geographically distinct. N o r t h of the Hindu Kush lay the
area of what is known as Graeco-Bactrian coinage, which represents the direct
continuation of the Seleucid series that it succeeded in this region and whose
Attic standard it preserves (with a theoretical drachm weight of 4.4 g and tetra-
drachm weight of 17.5 g) as well as the exclusive use of Greek for the king's
name. South of the Hindu Kush lay the area of Indo-Greek coinage, which had a
weight standard considerably lighter than that of the Attic coinage (with a theor-
etical drachm weight of 2.4 g and tetradrachm weight of 9.8 g) and much closer
to that of the Indian punch-marked coins that were common in those territo-
ries. The Indo-Greek coins bear bilingual inscriptions, one Greek and the other
a translation of it into Prakrit, the Indian language of ancient north-western
India, written in Kharosrhi script. The most commonly used denomination in
this coinage was the drachm, which was sometimes minted in the square shape
of ancient Indian punch-marked coins, a form that was also used for bronze
coins. The Indo-Greek coinage, which was minted in the territories of north-
western India after they had been conquered by the Graeco-Bactrian kings, did
not begin until around 180 B.c., long after the first Graeco-Bactrian issues, but
continued to circulate until the time of the last Indo-Greek king, Strato I1 (c. 55
B.c.), long after the Graeco-Bactrian issues had come t o a halt.
The recent discoveries, notably in the famous Qunduz hoard of Graeco-
Bactrian tetradrachms and decadrachms issued by the late Indo-Greek kings of
whom only bilingual coins were previously known (Amyntas, Antialcidas,
Archebius, Diomedes, Hermaeus, Lysias, Menander, Philoxenus, Theophilus), do
not necessarily prove that these kings retained enclaves within a Bactria overrun
by nomads.97They only imply that these late Graeco-Bactrian coins were struck
by these sovereigns for commercial use with the northern territories, already lost
to them, where the former Graeco-Bactrian issues continued to circulate, or to
pay tribute to the nomadic c o n q ~ e r o r s . ' Except
~ in the case of Pantaleon and

95. For catalogues of the coins, sec note 51 above. For the historical interpretation of the
coinages, see the works cited in note 1.
96. The magnificent 20-stater piece (167 g) of Eucratides I, now in the Cabinet des
Midailles, Paris, is the most remarkable of thcsc rare gold coins: scc Scltman, 1955
Plate LV, N o . 5; Bopedrachchi, 1991, p. 69, Plates 16 and 25.
97. Curiel and Fussman, 1965, pp. 61-4; Bopcarachchi, 1991, pp. 75-6.
98. The late date of these kings rendcrs untenable thc theory that they would have kept
territories in Afghan Bactria, especially since the region o f Badalthshan or eastern
Bactria, in which these last outposts of thc Greck presence north of the Hindu Kush arc
supposed t o have been located, had already fallcn into nomadic 11.1nds by 145 I~.c.,as the
excavations at A y Khanurn havc shown.
 The Greek kingdoms of Ccntral Asia

~gathocles,who had admitted into thcir coinage ccrtain Indian divinities, thc
gods represented on the Indo-Greek issues remain Greek. The wheel depicted
on a unique copper coin of Menander is virtually thc solc conccssion to Indian
symbolism. Despite the progress of Indianization of the colonists, official
life remained Greek until the very end. The superb royal portraits adorn-
ing these coins, whether bareheaded, helmeted or wcaring the kausia (a Macc-
donian felt cap), are some of the most striking manifestations of Central Asian
Hellenistic art t o survive.

The Greek heritage in Central Asia

The influence of Greek tradition on the empire of the Kushans, successors to
the Greeks in Central Asia, should neither be over- nor underestimated. This
influence was profound, as one would expect, even though in certain fields it
met with an anti-Hellenistic reaction. These former nomads, having subjected
an empire of sedentary peoples, adopted some of the ways of urban life and
civilization in an environment that had been shaped by two or three centuries of
Greek domination. Greek ceased t o be the official language, and was replaced
by local Iranian and Indian languages; but to write Bactrian, which until then
had been only a spoken language, the Kushans made use of the Greek alphabet
with minor modifications. Gold replaced silver as the monetary standard, but
the highly developed monetary system of Kushan Central Asia, based on bime-
tallism (gold and bronze) and using coins depicting the king on the obverse and
various deities on the reverse, was a Greek invention and in no way a product of
the Iranian or Indian East.
In religion there can be n o doubt that local cults quickly gained the upper
hand over the pantheon of the colonists. This change must have been made all
the easier by the fact that the Greeks themselves had probably never, except in
their own official state religion, put up any barriers between their own gods and
those of their subjects, and so had paved the way t o their progressive assimila-
tion. The Kushan coinage is indeed dominated by an Indo-Iranian pantheon
within which only a few gods are occasionally designated by their Greek name,
but most of the deities are portrayed in a style that, if not overtly Greek, is at
least Hellenized. The goddesses are shown draped in a chiton and a cloak, occa-
sionally holding a cornucopia. O n e of them (Risto) even retained the martial
costume and attributes of an Athena. The gods are generally dressed in a knee-
length tunic with a short cloak thrown over the shoulder, or in a long robe with
a full-length cloak. The same trend can also be detected in both the sculptures
and paintings of the ~ e r i o d An
. image of Athena modelled in clay dating from
the first century R.C. was found at K h a l c l ~ a ~ ainn Bactria. Several centuries later
Dalverzin in north Bactria also yielded a figure of a goddess retaining the fea-
tures of a helmeted Athena; and from Tepe Shotor on the Hadda lain comes a
 P. Bernard

statue of a seated Vajrapini, one of the Buddha's attendant deities, which faith-
fully reproduces a prototype of Heracles as he is represented on the coinageof
the Graeco-Bactrian king Euthydemus.
~t is in the architecture and figurative arts that the Greek heritage is to be
most clearly seen. Certain Graeco-Bactrian building styles were incorporated
into Kushan architecture. At Dalverzin in northern Bactria this was the case
both for the great mansions with their front courtyards and central living-
rooms, and for a mausoleum containing many vaulted chambers set on either
side of a central corridor. However, it is primarily in architectural decoration
that the persistence of the Graeco-Bactrian and Indo-Greek tradition reveals
itself. It is apparent in the columns set o n Attic-Asiatic bases (sometimes
arranged to form a true peristyle), in the pilasters used t o decorate walls, in the
ubiquity of the Corinthian order with its acanthus leaves, in the rows of ante-
fixes along the roof edges and in the decorative use of relief and sculpture
modelled in clay o r stucco. Even the technique of cutting the bases of columns
with a turning machine survived into the Kushan period.
In the figurative arts the Greek conquest brought about an equally pro-
found change. It introduced a certain naturalism in the representation of the
human body and its drapery, an attempt to produce volume and perspective, the
use of three-quarter views and asymmetrical attitudes. Even when Central
Asian art had largely outgrown these outside influences, their underlying pres-
ence could still be felt. For a long time, it has been thought that the art style
called Graeco-Buddhist, which developed in Gandhira and the Oxus valley in
the first centuries of our era, derived its markedly Western features from
influences passing from the Roman Empire along the trade routes. The dis-
covery over the past twenty years of a vigorous and authentically Greek civil-
ization that had flourished under the rule of the Graeco-Bactrian and Indo-
Greek kingdoms seems to indicate that the real roots of the classical influences
evident in Graeco-Buddhist art are to be sought in this easterilmost branch of
Hellenism, while Rome's contribution was merely secondary. Since at Tepe
Shotor an artist of the fourth century A.D. can portray a Vajrapini so similar to
the Heracles o n the Graeco-Bactrian coins of Euthydemus, and since Indo-
Scythian coins provide an intermediate link," there is no need to look for proto-
types in some distant country beyond the sea when local traditions ~rovide
them. The fabulous royal tombs discovered at Tillya-tepe in southern Bactria
have brought to light another descendant of the Greek art in Central Asia, an
art of the steppes enriched with Hellenistic iconography and traditions in the
form of gold jewellery, richly set with semi-precious stones,
turquoise.
The fact that we now have roof of the existence of Greek theatres in
Central Asia, where Greek plays were actually ~ e r f o r m e d ,sllould lead to 2

99. Tarzi, 1976, pp. 394-7, Figs. 10-1 1 .

 The Greek kingdoms of Central Asia

re-examination of thc origins of Indian theatre, whose first crcations were

roughly contemporary with the end of the Greek kingdom, and of possible
Greek influence o n it. In the same manner, the lively interest that Greek colo-
nists of Central Asia took in astronomy - as shown by the discovery at Ay
Khanum of sundials, one of a highly sophisticated type - raises othcr questions
of a similar nature. W e are indeed entitled to ask ourselves if the Babylonian
astronomical conceptions that are manifest in Indian astronomy may have
reached India earlier than has been supposed, that is t o say, in the Hellenistic
via the Greek kingdoms of Central Asia.
 PARTHIA::.
G. A. Koshelenko and V. N. Pilipko

HE Kingdom of Parthia emerged as a result of the socio-economic crisis

affecting the Seleucid state in the mid-third century B.c.' In the course
of that crisis the governors of the extreme eastern satrapies - Diodotus
in Bactria and Andragoras in Panhia - seceded from the Seleucid kingdom.
While in Bactria an independent Graeco-Bactrian kingdom came into being, the
situation in Parthia was much troubled by incursions of nomads belonging t o
the confederation of the Parni who had occupied land along the edges of the
agricultural oases from the Caspian Sea to the River Tedzhen. The Parni, with
Arsaces at their head, invaded Parthia. In the ensuing struggle Andragoras was
killed and control of the country passed to the nomadic aristocracy of the Parni
headed by Arsaces.? The Parni soon seized Hyrcania (a region on the southern
and south-eastern Caspian seaboard), and this boosted the economic and mil-
itary potential of the infant state.

Parthia and the Seleucids

The Parthian kingdom still had t o survive a stubborn struggle with the Seleucids
before its independence was firmly established. Between 230 and 227 B.C. the
Seleucid king Seleucus I1 undertook a campaign to recover his eastern satrapies.
The Seleucid troops were initially successful, but the Parthians, backed by the
nomadic tribes (Apasiacae), finally carried the day when the revolts which had
broken out in regions t o the west of the Seleucid Empire prevented Seleucus
from continuing the war. The next stage in the struggle with the Seleucids took
place during the reign of Antiochus 111. In 209 B.C. he began his famous Eastern

:' See Map 3.

1. D'yakonov, 1961, pp. 180 et seq.
2. Koshclenko, 1976, pp. 31-7.
 G. A. Kosh~lenkoand V. N.Pilipko

which resulted in the defeat of Parthia and forced Artabanus (Ardavin

in Middle Persian) to recognize Seleucid overlordship. But when the Romans
defeated the Seleucids in the battle of Magnesia (192 B.c.) the Parthians were able
to take advantage of the situation and reconquered the provinces south of the
Caspian Sea. Parthia again became independent and was able to resume its expan-
sion both eastward and westward. This was made easier by the fact that the
Graeco-Bactrian and Seleucid kingdoms lacked political stability and were open
to bitter internal strife. The Parthian ruler Mithradates I (171-139 B.c.) made
clever use of these favourable circumstances. In the east he attacked Bactria and
took from it a number of dependent region^.^ Media was then captured - a con-
quest that opened the way for the Parthians t o the west and south, towards Meso-
potamia, Susiana and Elymais. The political chaos in these regions enabled the
Parthians to invade central Mesopotamia in 141 B.C. and seize that major centre
of the Hellenistic east, Seleucia o n the Tigris. After this Susa also came under
their suzerainty..' These conquests, however, presented the Parthians with a very
complex problem. The new territories that came under their control contained
important concentrations of Greek and Hellenized inhabitants, who were now
deprived of the privileged position they had enjoyed in the Seleucid Empire. For
the next two centuries the Greek cities within the Parthian state were the main
opposition forces, hostile to central rule. Counting o n the support of this Greek
population, the Seleucid ruler Demetrius I1 attempted to recover Mesopotamia in
140 B.C. but was unsuccessful. There was a further Seleucid attempt to subdue
Parthia in 131/130 B.C. The Seleucid ruler Antiochus VII Sidetes, relying on the
support of the Greek cities, inflicted severe defeats on the Parthians and pene-
trated into the innermost regions of Parthia. But eventually the Seleucid forces
were routed and Antiochus VII himself was killed in battle in 129 B.C.This was a
turning point in the history of Hellenistic Central Asia. The Seleucid state ceased
to exist as a world power and its kings became the petty rulers of rival warring
states in northern Syria.j
'The Parthians recovered all the lands they had earlier lost and the way
westward into Syria now lay open to them. At that time, however, the situation
was once more aggravated on the eastern borders of Parthia. The major move-
ment of nomadic tribes in Central Asia which had brought about the downfall
of the Graeco-Bactrian kingdom (between 140 and 130 LC.) was bound to
affect Parthia as well. In 130 B.C. the Saka tribes invaded the eastern regions of
Parthia and individual detachments penetrated as far as Mesopotamia. The Par-
thian king Phraates I1 lost his life in the struggle against the nomads (129 B.c.)~
as did his successor and uncle, Artabanus 1 (123 B.C.).~Parthia also faced

3. Koshelenko, 1972.
4. Le Rider, 1965.
5. Fischer, 1970.
6. Debevoise, 1938, pp. 35-8.
 Parthia

substantial problcms in the west where Hyspaosines, King of Characcnc (a

sl.nall region on the northern shore of the Persian Gulf), had seized most of
Mesopotamia. Thus after a period of resounding success against the Selcucids,
Parthia found itself on the verge of collapse.

Mithradates I1 and his successors

Mithradates I1 (123-87 B.c.) managed to stabilize this difficult situation. He was
able to subjugate Characene and re-establish calm in the Greek cities. Parthia
followed a more moderate policy towards the Greeks than hitherto, and they,
now lacking any support from the west, became much more reconciled to Par-
thian rule.
The problem of the nomads on the eastern front was solved by a mixture
of military means and diplomacy. They were displaced from Parthian territory
proper and settled around Lake Hamun on the lands of Arachosia and Dran-
giana - the region later called Segistan (modern Sistan). The emerging petty
states under nomadic leaders came very much within the Parthian sphere of
influence, and some of them became vassal dependencies. Parthian influence in
the east was considerably extended and came to include the greater part of mod-
ern western Afghanistan.'
From the beginning of the first century B.C. the Parthian state had
achieved unprecedented strength and had become the foremost power in West-
ern Asia. But the latter years of the reign of Mithradates I1 were marked by
further complications: the struggle for power between various representatives of
the Arsacid house; the interference in Parthian affairs of the Armenian king-
dom; and the relentless eastward expansion of Rome. The first century B.C. saw
repeated military conflict between Rome and Parthia, as well as clashes between
old rivals who were meddling in each other's internal disputes.

The pattern of decline

During this period there were two major political forces in Parthia, and the
struggle between them had a profound influence on the course of Parthian his-
tory as well as on relations between Rome and Parthia.x O n e group consisted of
the leaders of the Greek and local towns of Mesopotamia and the Parthian aris-
tocracy who had settled there. The major force in the coalition was provided by
the Greek towns. The eastern progress of the Romans gave the Greeks new

7. Sarianidi and Koshelenko, 1982.

8. Koshelenko, 1963, pp. 56-8.
 G. A. Koshclenko and V. N . Pilipko

hope. They were attracted by Rome's call to defend 'Hellenism' from 'barbn.
rism' and the prospect of social peace secured by the Roman legions. In a region
with age-old traditions of slavery, the ruling classes of both the Greek and east-
ern towlls (which were close in character to Greek city-states) saw orthodo\
slave-holding as the main source of their enrichment. Constant wars, which dis.
rupted economic activity, halted trade and led to disturbances and uncertainty
. I

were equally unwelcome.

The second group consisted of the nobility of the purely Iranian regions,
was closely linked with the peripheral nomadic tribes. This section of the
Parthian ruling class wished t o pursue a broad expansionist policy, and with the
ordinary fighting men dependent o n them, formed the nucleus of ParthiaYs
armed forces. They looked to enrichment from major conquests. Their popular
slogans were 'the return of the Achaemenid heritage' and 'the conquest of all
Asia'.

The Romans and the Parthians

The series of wars between the Romans and the Parthians took place against the
background of active conflict between these two groups of the ruling class of
Parthia, with the Romans actively supporting those claimants to the Arsacid
throne put forward by the 'Western' group. The major wars between the
Romans and the Parthians in the first century B.C. were indecisive. In 53 B.C.
the Romans faced a humiliating defeat at Carrhae in Mesopotamia; the Parthian
invasion of the West in 40 B.C. ended in their disastrous defeat at the battle of
Gindara (38 B.c.); and a further Roman invasion under Mark Antony in 36 B.C.
failed likewise. At the end of the Roman civil war in 29 B.c., when Octavian
became the first Roman emperor - Augustus - Rome's Eastern policy under-
went important changes. The aggression that had marked the period of the
Republic gave way to a quest for peace. O n e of the main reasons for this was
the realization that Rome lacked the military resources for any large-scale con-
quests in the East. Parthia thus played an outstanding role in the history of the
East by halting Roman aggression. The new Roman policy of peace also co-
incided with the interests of Parthia, which was not strong enough to undertake
major conquests in the West. However, Rome's changed policy in the East did
not signify an end to attempts to interfere in the domestic affairs of Parthia. It
was a period of intense conflict between the two groups of Parthia's ruling class-
The Romans actively took sides in the dispute, supporting the candidate who
was most favourable to their own interests in Parthia. Their success in placing
Vonones on the Arsacid throne was a considerable achievement for Ro~nan
diplomacy; but the pro-Roman policy of Vonones brought about a reaction
which consolidated all the 'patriotic' forces, under Artabanus 11, founder of the
Later Arsacid dynasty. His main support came from thc eastern regions oi
~ a r t h i aand he had close links with neighbouring nomadic tribes. His suppor-
ters opposed the transformation of Parthia into a Roman vassal and vicwcd thc
~onflictwith Rome as a struggle to restore the power of Cyrus, the Achaernc-
nid. In domestic policy Artabanus I1 sought to limit the self-rule of thc (;reek
towns. In cultural matters there was a reoricntation towards ancient Iranian tra-
ditions, and the Hellenistic cultural heritage was rejected.

Successors of Artabanus
Although the reign of Artabanus I1 was an important landmark in the history of
Parthia, being marked by significant political and cultural changes in the Arsa-
cid state, it did not mean an end t o internal conflicts. For many ycars after the
death of Artabanus, wars were waged between two of his successors, Vardanes
and Gotarzes. Some stability was achieved under Vologases I, who conducted
an active foreign policy and sought to restore Parthian control over Armenia.
The long and variable struggle between Rome and Parthia over this ended with
an agreement in A.D. 63 that the brother of the Parthian king Vologases should
be proclaimed King of Armenia and crowned in Rome by the Roman emperor
Nero. This agreement was extremely important since it led to a long period of
peace on the frontier between Rome and Parthia interrupted by only minor dis-
putes.

Later Roman-Parthian relations

Peace was next broken in A.D. 114 when the Roman emperor Trajan began his
carefully prepared campaign against Parthia.y The Romans initially had consid-
erable success, capturing Ctesiphon, the capital of Parthia. The Roman army
marched t o the Persian Gulf, and the Roman fleet sailed down the Tigris. The
success of the Romans owed much to the bitter conflicts within Parthian society
between rival claimants to the Arsacid throne, and to the revolts that had bro-
ken out in Elymais and Persia. But at the height of the Roman success the situ-
ation radically changed. In all the Parthian territories conquered by the
Romans, insurrections broke out, triggered off by the introduction of the
Roman system of ~rovincialadministration, which strictly controlled towns,
taxes and requisitions, and by the discontent of the petty rulers who had recog-
nized Rome's authority and had subsequently been stripped of the remnants of
their independence. The rival representatives of the Arsacid house united
against the invader and in A.D. 117 the Romans were compelled to abandon all

9. Lepper, 1979.
 G. A. Koshclenko and V . N . Pilipko

their conquests in Parthia. Although the Roman frontier was peaceful again,pa,-
this was still not secure and faced severe complications o n its northern and east-
ern borders. It appears that Hyrcania finally achieved independence; &, sep
aratist trends of other regions of the state became more marked; and Parthia!s
northern provinces suffered incursions from the Alani. The emergence and
growth of the powerful Kushan Empire created a permanent danger in the East,
Exhausted by internecine wars and constant difficulties with Rome, Parthia
sought to reduce tension in the East to a minimum. The stumbling block in rela-
tions between Rome and Parthia, however, remained Armenia, where in the time
of Vologases I11 there was a bitter clash in A.D. 161-63. The northern flank of the
Roman defence collapsed and Parthian troops invaded Syria. Rome, alarmed that
there might be a general uprising against its rule in the East, mustered its strength
to stabilize the situation and then to launch a counter-offensive. The peace treaty
concluded at the end of the war was harsh for the Parthians, since the whole of
Mesopotamia as far as the River Khabur was ceded t o Rome. Even harsher for
Parthia were the consequences of the war which broke out in A.D. 195. The Par-
t h i a n ~had supported Pescennius Niger in the Roman civil war. The Romans
found that Vologases IV (A.D. 191-207), who seemed t o have invaded eastern
Iran, had at the same time to oppose the large-scale revolts that had broken out in
Media and Persia. The Roman military expedition dealt a heavy blow to Parthia:
the richest parts of the country were devastated and some 100,000 inhabitants
were taken to Syria and sold into slavery. The last war between Rome and Parthia
began in A.D. 216. The conflict between Vologases V and Artabanus V, the two
pretenders t o the Parthian throne, made the conditions ripe for Roman interven-
tion. The Romans, under their emperor Caracalla, invaded Parthian territory and
laid waste much of Mesopotamia and part of Media. In the summer of A.D. 217,
Artabanus V, who had mustered sizeable forces, started t o wage a resolute cam-
paign against the Romans. Caracalla fell at the hands of conspirators and Macri-
nus became emperor. After a decisive battle at Nisibis the Romans had to sue for
peace. However, this was the Parthians' last success. The ruler of Persia,
Ardashir, united with a number of other local rulers to raise a revolt against the
Arsacids. In 223 he defeated and killed Vologases V. A few years later Artabanus
V was defeated and killed at the battle of Hormizdagan, and the entire territory of
the Arsacids soon passed into the hands of the new dynasty of the Sasanians.'"

Socio-economic systems
The Parthian period saw considerable economic development in the countries
that made up the Arsacid state. Archaeological investigations on the Susa plain.

10. Lukonin, 1961.

in the Diyala vallcy, the Kopct Dag foothills and elsewhere in I'arthia" indicate
that thcrc was a sharp increase in the number and sizc of settlcn~cnts(I:igs. 1-4)
and in the extent of irrigated land, cornparcd with the preceding l'he
basis of Parthia's economy was agriculture; and irrigated cultivation was
broadly developed in most parts of Parthia. Cereals wcrc grown and thcre were
large areas of vineyards and gardens. Technical crops such as cotton and sesarnc
were well known. At the confines of the cultivated land many nomadic tribes
were engaged in cattlc-raising. It seems likely that good relations wcrc cstab-
lished between the inhabitants of the oascs and the nomadic cattle-raiscrs,
though a number of tribes, chiefly in mountain areas, still led a communal cxis-
tence within a natural economy.

Flc;. 1. Slte of Nisa. (Photo: 8 Vladimir Tcrebcn~n.)

Trade and commerce

Crafts were well developed in Parthia. Some items gained international recogni-
tion, including lined fabric from Borsippa (in Babylonia), carpets from the indi-
genous areas of Parthia, and the iron of Mal-giana in the form of
weapons and armour). In the Parthian period trade and monetary relations were
also developed. Parthia took part in extensive international trade and had trad-

11. Wenke, 1975176.

 G. A. Koshelenko and V. N. Pilipko

FIG.2. Nisa. Throne-mom. [Courtesy af 1. Iskender-Mochiri.)

ing links with the Roman Empire, mainly via Palmyra, which served as a kind
of 'buffer' and acted as a commercial intermediary. Palmyrene caravans went
both to the various towns of Parthia and t o the coastal towns of Syria.l2The
great quantities of Parthian coins (particularly from the first century A.D.)
found in Transcaucasia suggest that Parthia maintained fairly close ties with
those regions. Trade with India was conducted both by sea, via Spasinu-Charax
on the Persian Gulf, and overland via southern Iran o r from Merv by the south-
east portion of the 'royal way' leading to India via Sistan and Kandahar. In the
Indian trade there was constant competition between the Parthians and the
Romans, who carried goods mainly by sea via the Red Sea ports. A special role
in international trade was played by the 'Silk Route', which linked the countries
of the Far East and the Mediterranean through Central Asia. As Parthia con-
trolled much of the route and was able to enjoy a monopoly of its trade, Par-
thian merchants made huge profits by reselling Chinese wares, mainly silk, on
the markets of the Roman Empire. The Parthian administration therefore tried
to prevent direct contacts between Chinese and Roman merchants. This inter-
mediary trade enriched the Parthian merchant class, and the state obtained a
substantial revenue from trading dues. Archaeological finds from Seleucia on
the Tigris and written records point to the existence of taxes on trade in slaves,

12. Koshelenko, 197117, pp. 761-5.

 and other items." Local trade also developed and local markets were estab-
lished, as attested by the vast quantity of small bronze coins minted to mect the
of those markets. The process was most in evidcncc in the econo-
mically more developed regions of the country (Mesopotamia, Susiana,
Margiana, etc.).14

FIG.3. Aerial view of the site of Merv, (Photo: 63 Vladimir Tcrebenin.)

Parthian coinage
The Parthian monetary systemI5 was based on the silver drachm (weighing
about 4 g). Coins were first struck in the reign of Arsaces I, the founder of the
state. The main denominations were the tetradrachm (struck in quantity bv
Western mints) and the drachm (typical of the Eastern mints). Small bronze
coins were also issued as change. The obverse side of the drachm usually carried
the bust of the king facing left, while the reverse bore Greek lettering around
the edge and portrayed a seated figure in nomadic attire with a bow held at
arm's length (Figs. 5-10, as identified by D. G . Sellwood in An Introduction to
the Coinage of Pavthia, London, Spink & Son, 1971). The tetradrachms and, i l l

13. McDowell, 1935b.

14. McDowell, 1935a; Le Rider, 1965.
15. Sellwood, 1980.
 G. A. Koshelrnko and V . N . Pilipko

particular, the bronze coins were more varied in type. From the reign of vole-
gases I onwards the drachm also bore Parthian letters. Starting in the seventies
of the first century B.C. the reverse side of coins regularly carried monogums
which, in the view of a number of researchers, are abbreviations for the names
of mints. Coins struck by the Merv mint, for instance, were marked with rht
Greek letter 'pi'.'"

FIG.4. Merv. Plan of the city: A - Alexandria in Margiana enclosure (Seleucid age);
B - new enclosure (Parthian age). (Courtesy of V. M. Masson.)

16. Pilipko, 1980.

FIG.5. Drachm of Arsaces I1 (21 1-191 ($.( .).
(Courtesy of Christian Schaack.)

FIG.6. Drachm of Artabanus I (127-123 H.c.).

(Courtesy of M. I. Mochiri.)
 G. A. Koshelenko and V. N. Pilipko

FIG. 7. Drachm of Phraates IV (38-2 B.c.).

(Courtesy of M. I. Mochiri.)

FIG.8. Drachm of Phraataces and the Queen Musa (2 B.C. to A.D. 4).
(Courtesy of Christian Schaack.)
 Parthia

FIG,9. Drachm of Osroes I (AD. 109-129).

(Courtesy of M. I. Mochiri.)

FIG. 10. Drachm of Vologases V (A.II. 191-208).

(Courtesy of M. I. Mochiri.)
 G. A. Koshelenko and V. N.Pilipko

The socio-economic fabric

The question of the character of socio-economic relations in Parthia is
extremely complex and cannot yet be regarded as resolved. Parthia did not
represent a uniform whole so far as the level and character of S O C ~ O -
relations were concerned. T w o basic regions are to be distinguished, one being
Babylonia, Mesopotamia and Elymais, the other the Iranian uplands, the eastern
Iranian steppe and the southern part of present-day Turkmenistan. The differ-
ences between the two regions lay in the ethnic composition of the population
(with various Iranian-speaking peoples in the east and Semitic-speaking peoples
in the west) and in their very distinct socio-economic systems. In the western
regions of Parthia, civilizations based o n class divisions had been in existence
for a long time. In the eastern regions, the switch to a class society only took
place at the beginning of the first millennium B.C. In that and particularly in the
Achaemenid period, slavery became a widespread practice in the eastern Iranian
regions. It seems, however, that the principal influence o n the social structure
the indigenous Parthian regions was exerted by the nomadic Parni." Society in
the eastern regions of Parthia apparently consisted of three basic social groups
which can be defined as classes. The upper class was made u p of the dzdt (free
people), descendants of the nomadic aristocracy of Parni who had become the
ruling class in Parthia and filled the top posts in the state administration. In the
Parthian army it was they who represented its main striking power as the hea-
vily armed and mailed cavalry. They received the lion's share of the spoils of
war and thus had a special interest in an aggressive foreign ~ o l i c y The . second
social class consisted of the descendants of rank-and-file Parni. They belonged
to the conquerors and so to the ruling class, but they were dependent on the
Parni aristocracy. They formed the bulk of the Parthian army, the mass of
lightly armed archer horsemen. The position of the mass of the sedentary agri-
cultural population was determined by the fact of the nomadic conquest. Its
members were formed into communities, where they enjoyed full
legal rights. They owned specific property, which they could buy and sell. But
they were not considered entirely free, and there was a clear distinction between
them and the upper class. They depended on the ruling class, and that depen-
dence was collective. Ancient authors emphasized that in n o case could they
obtain full ~ e r s o n a freedom.
l Exploitation of that social class was apparent and
primarily took the form of payment of taxes. It can be presumed that the pre-
vailing trend in social development had increased dependence of the peasant
communes o n the state. Another move worth recording was the attempt of the
ruling class to abuse its administrative powers for ~ e r s o n a ends.
l This iilvolved
exploitation of the ~ r o d u c e r s At
. the bottom of the social scalc were the slaves,

17. Koshelenko, 1980, pp. 17-99.

 I'a r t h ia

but our sources arc silent about their position in economic production. In the
western part of the Parthian kingdom the level of dcvelopmcnt of socio-
economic relations was somewhat different, and the influence of 'classical'
forlns of slaver? (handcd down from the Seleucid period) was more in evidcncc.
The major slave-holding centres wcre the Greek city-states. Manumission
records found in Susa reflect the considerable extent of ancient slavery,lx while
material from Babylonia confirms transactions involving thc buying and selling
of slaves. O t h e r forms of dependence continued from previous periods, includ-
ing temporary debt-discharge s l a ~ e r yIt. ~may be supposed that in the Scleucid
period it was the rule t o make the surrounding rural territory dependcnt on the
Greek city-states. This social institution was also inherited by thc Parthians. O n
the whole, the social structure of society in the western regions of Parthia can
be reckoned t o have been extremely complex. In contrast t o the East, the circle
of fully free people was much broader, and included the citizens of the Greek
towns, members of the municipal religious communities, the Parthian settlers
and possibly a portion of the peasantry. At the same time, slavery was more
strongly developed than in the East, and there was a much wider variety of
other (non-slavery) forms of dependence.

The king and his council

The Parthian state system displayed a number of original features that were due
to peculiarities of its socio-economic and political development. Most impor-
tant was the fact that the system emerged from three sources: the heritage of
Achaemenid Persia; the principles of Hellenistic statehood undcr the Seleucids;
and the traditional institutions of the nomadic Parni. The king stood at the head
of the state, but royal power was regarded as the collective property of the
Arsacid family, and only members of the Arsacid family could occupy the
throne - a principle that led t o rival claimants and frequently involved dynastic
conflicts. Written sources record the existence of two royal councils, which
seem t o have limited the king's power. One, consisting of kinsmen of the king,
wielded considerable influence, particularly in matters of succession t o the
throne. A survival from the time of the Parni, it became an instrument by which
the nobility could influence state affairs. T h e second council, of priests, o r magi,
seems t o have enjoyed less influence. Alongside the Arsacid fanlily, six other
noble families played an important role in the Parthian state - an arrangement
possibly due t o the influence of Achaemenid tradition.

18. Koshelcnko a n d Novikov, 1979, pp. 41-54.

19. Welles et al., 1959.
 G. A. Kosl~c~knko
and V . N. Pilipko

Administration
The Parthian administrative system was fairly complex. Part of the kingdo,
was divided into satrapies ruled by satraps appointed by the king. The rest con-
sisted of vassal kingdoms.?O The Parthian satrapy was much less extensive than
the Achaemenid one. In some cases power over a number of satrapies (usUally
along the frontiers) was concentrated in the hands of the same person. The low-
est administrative unit was the stathmos (in Greek) o r diz (in Parthian),
represented a group of a few villages. The stathmos also had a small military
post. This administrative unit was headed by a dizpat. There was an extensive
and developed bureaucracy, as attested by ostraca from Nisa and by the Par-
thian and ostraca from Dura-Europos. The towns occupied a rpe-
cial place in the system of state rule.2' The Greek city-states in Parthia were a
survival from the Seleucid period. Under the Parthians they formally retained
their autonomy though their position changed very much for the worse. Their
aristocracy, which had previously consisted solely of Greeks and Macedoniq
lost its exclusiveness and was broadly penetrated by members of the local popu-
lation. The urban system became increasingly oligarchical, popular assemblies
declined in significance, and power was concentrated in the hands of a council
made u p of representatives of a few of the richest families. Less is known of the
character and evolution of other types of town. The old Babylonian towns (e.g.
Uruk Warka) enjoyed a position in Seleucid times which had made them some-
what comparable to the city-states. Their population was divided into citizens
enjoying full rights and the rest, the under-privileged, without such rights. The
fully ~rivilegedaristocracy formed a religious and municipal commune enjoying
a measure of self-rule. These towns also owned a land district. It may be pro-
posed that these religious and municipal communes underwent gradual changes
during the Parthian ~ e r i o dsimilar to those affecting the Greek city-states.
While sources also speak of 'Parthian towns', in contrast to Greek ones, there is
no specific information about their internal life. It can only be conjectured that
they did not enjoy autonomy and were under the full control of the local Par-
thian administration.
Apart from the territories forming part of the royal domain and governed
through satraps, much of Parthia consisted of vassal kingdoms. Their role and
importance constantly increased as more and more of the states (Persia, Ely-
inais, Margiana, Sistan, etc.) slipped from the direct control of the central
government. The Arsacid family turned out to be a palliative. separatism
remained; only its colour changed. What basically fuelled separatism was the
very narrow social support for the dynasty. The Arsacids were unable to create
a sufficiently broad unity of the ruling classes or to alter the structure estab-

20. D'yakonov, 1961.

21. Koshelenko, 1979; Sarkisyan, 1952.
 Parthia

at the time of the emergence of the Parthian state, in the specific circunl-
stances of the nomadic conquest.

The fiscal system

The fiscal system of the Parthian state is still inadequately known. What wc d o
know is that there was a general state cadastrc for the lands of the royal domain.
The state fixed and strictly controlled tax revenue. Records found in excavations
at Nisa provide evidence of different types of tax collection, depending o n the
category of the land. T w o categories are known - patbiz and uzbari. Patbiz
was collection in kind f o r the use of the king. It is less clear what thc other catc-
gory was. There are also indications of the existence of special levies for the sup-
port of religious activities, somewhat similar t o tithes.2'

Parthian culture
The study of Parthian culture still requires further research. Until compara-
tively recently the prevailing view among scholars was that Parthian culture
was eclectic, lacking in originality and devoid of a creative basis. Parthian
culture developed through the interaction of a number of factors - the Achae-
menid heritage, the conceptions of the Hellenistic period, the contribution of
the nomadic Parni, and the particular cultural traditions of the peoples w h o
made u p the Parthian state. T h e basic trend in the development of Parthian
culture was the synthesis of Greek and local sources. This synthesis, which had
started in the preceding period, assumed substantial proportions in Parthian
times when the Greek population lost its privileged position. It took many
forms and made itself felt in various spheres, notably architecture, sculpture
and painting (Figs. 11-13). Parthian sculpture and painting are best known as a
result of the excavations at Dura-Europos. They are marked by a particular
style, which suggests that this Parthian-Mesopotamian art was one of the
sources of early Christian art. Parthian architecture was marked by monumen-
talism and a break with the traditions of the Greek architectural orders, such
styles being used f o r purely decorative purposes. T h e most widespread form of
artistic craft was the fashioning of terracotta statuettes. They abound in Meso-
potamia and Margiana, and ~ r o v i d emost important material for the study o f
popular beliefs. Although the Arsacid dynasty was, o n the whole, marked by
tolerance, and various religions coexisted in Parthia, there can be no doubt that

22. Koshelenko, 196617, 1977; Pugachenkova, 1967; Rostovtzeff, 1938; Ghirshman, 1962,
1976.
FIG. 11. Parthian pnilcc frorn thc tcmple of Sbami (scoand cc~~tur*y
u.c.).
(Courtesy of A hmnd Tellrani- y e Moghaddam, Irar7-e Unsta~iMuscum.)
 I:Ic;. 12. Nis.1. lih\.tc)li. I \ ~ o r \(I'lioto:
. lC \'l,~dimi~-
Tcl.ehcnin.)

Zoroastrlan~siii gradu'ill\ g r e w in impol-tancc. A c c o r d i n g t o t r , ~ d i t i o n tlie

Avesta w a s first codified u n d e r Vologascs I. A t that time tlie s\.lnhols of
Zoroastrianism (in t h e f o r m of fire a l t , ~ r s )first a p p e , ~ r c don coilis. possibly i~idi-
cating its a d o p t i o n as tlie official religion. In t h e i ~ i d i g e n n u sP a l . t l i i ~ ~lands
i tlic
Zoroastrian calendai- \vCis ,ilso used beside tlie Seleucid c r ~ h. l o r e o \ ~ e rZOI-0'1s-
trian principles f o u n d their w,I\. i n t o t h e official idcolog!. of tlic d\-n.ist!-. \Y'liilc
. . .
royal p o w c r ~nltl,ill\. based its autlioi-it\ ~311tlie right of c o n q u e s t , it suhsc-
q u e n t l ~ rs o u g h t its justification in I-eligious s , ~ n c t i o n .T h e P,11-tlii'in period prob-
abl17 saw t h e iiitl-oduction of t h e p ~ - ~ i c t i o
cfc kindling tlie c o r o n a t i o n fires, n.hiCli
W'IS later a d o p t e d 171- tlie S,~sanians." I n tlie eastern sector of t h c Pal-tliian king-

dom B u d d h i s m \\.as sp,-cading tast'j . ~ n dtlicl-c \\-ere P . ~ ~ - t l l i .s~cnl ~ o l a r sof Budcl-

 FIG.13. Nisa. Detail of a scenc o n a s h y t o n . Ivory. ( P h o t o : O Vladimir Tcrcbenin.)

hism w h o went t o China and participated in literary and missionary activities.

Unfortunately, very little is as yet known about the development of literature.
It may, however, be supposed that this was a time when many epic cycles took
shape, apparently including the Rustam cyclc.
 N.Ishjamts

Early Mongolia
In the first millennium B.C. Mongolia and the adjoining regions of Central Asia,
unlike China and other countries with a settled way of life, constituted a dis-
tinctively original nomadic world inhabited by aboriginal tribes and clans, who
kept sheep, goats and cattle. O n e of their principal occupations was the breed-
ing of horses, in particular the Przhevalski horse - which had been domesticated
earlier - a small, stocky animal with unusual endurance, widely used by the
Huns, Tiirks and Mongols. The two-humped Bactrian camel was of great im-
portance in the climatic conditions of the Gobi Desert. South of the Gobi Des-
ert, a small number of donkeys and mules were bred. It is interesting to note
that the wild ancestors of these horses, camels and asses were still found at that
time in the south-western part of the Mongolian Gobi east of the Altai, in
Dzungaria and Kazakhstan. In the period under review - 700-300 B.C. - the ter-
ritory of Mongolia and other parts of Inner Asia knew a fully developed
nomadic way of life, often referred to as Central Asian nomadism.
The people who then lived in the territory of what is now Mongolia,
Inner Mongolia, Dzungaria and southern Siberia have left behind them an enor-
mous number of 'stone-slab' graves, 'reindeer stones' and other material
vestiges of their existence.
The stone-slab graves are so described because at ground level they are
bordered by a rectangular wall of stone slabs sunk edgewise into the g o u n d . At
one corner of a stone-slab grave there is often a stone column, sometimes dec-
orated with the images of animals. The skeletons found in these graves lie on
their backs, usually with their heads turned towards the east, and are accompa-
nied by the bones of domestic animals, clay vessels and other articles. Some of
the vessels are made of reddish clay with handles; others are of brownish-grey

" See Map 4.

clay and are covered with hatchings like the clay vessels of the following
Hsiung-nu period.
The peoples buried in these graves had fully mastered casting techniques,
Their graves contain beautiful bronze objects and iron articles (or vestiges of
them). The Scythian-type bronze pots, axes, daggers, arrowheads, bronze and
iron horse's bits from the stone-slab graves of Mongolia bear a striking resem-
blance to similar articles found in the graves of the region beyond the Baika]
and in Ordos. Mongolia also boasts many specimens of reindeer stones - stone
columns decorated with images of galloping reindeer, sun discs and weapons,
which in technique and design have much in common with the 'animal style9of
the ancient monuments of representational art found in other parts of the steppe
belt of Europe and Asia.
In addition, large numbers of cowries from the Indian Ocean, white
cylindrical beads made of prophyllite, fragments of Chinese three-legged ves-
sels, and ornaments of nephrite (rings, discs and half-discs) and mother-of-pearl
have been found in the stone-slab graves of Mongolia.' For the most part, these
objects reached Mongolia through trade with China, Central Asia, Khotan and
Afghanistan. The country's cultural links extended through Sogdiana to India
and across Kazakhstan as far as the Black Sea and eastern Europe.
The various tribes of the zone, w h o undoubtedly spoke different lan-
guages and were ethnically and culturally different, possessed many articles that
were similar in shape. This is especially true of the weapons, horse harness and
ornaments. Initially the red om in ant type of weapon was the bronze-socketed
arrowhead with a flat tip (striking area), oval o r rhomboid in shape. This was
subsequently replaced by the socketed trihedral o r ~ y r a m i d a l arrowhead.
Horse's bits showed striking similarities. At first, bits with stirrup-shaped end
rings were exclusively used, but were later supplanted by bits with rounded
rings. With the introduction of bridles and metal bits it became ~ o s s i b l eto ride
on horseback over long distances, and this led to much closer relations between
tribes and significantly strengthened economic and cultural contacts between
far-flung provinces of the steppe.
In terms of the general level of development, the culture of stone-slab
graves and reindeer stones of Mongolia and other parts of the Eurasian steppe
belt of the seventh to third centuries B.C. coincided with the Late Bronze and
Early Iron Ages. Already by 400-300 B.C. iron articles were widespread in
Mongolia and throughout Inner Asia and heralded the beginning of the next
stage in development.
According to the ancient Chinese bone inscriptions, the famous ~hih-chi
(Historical Records) of the scholar Szti-ma Ch'ien, and other sources, the terri-
tory of present-day Mongolia, Inner Mongolia and Dzungaria was in times long
past inhabited by the Hsien-yiin, the Hsiung-nu and other nomadic cattle-

1 . Okladniltov, 1955, p. 261.

 Nomads in castern C'cnrral Asia

breeding tribes. In the first millennium B.c., these territories were succcrsivcly
inhabited by the H u , the Tung H u , the Hsi H u and the Hsiung- nu.^ Amongst
the above-mentioned peoples, the H u and the Hsiung-nu occupied the territory
of Mongolia. T h e T u n g H u (which means 'the Eastern H u ' in Chincsc) lived in
eastern Mongolia and western Manchuria, whereas the Hsi H u ('the Western
H ~ ' )lived in the area t o the south-west and west of Mongolia.

The Hsiung-nu Empire

~rchaeologicalevidence from the seventh t o the third century B.C. provides a
picture of nomadic societies with a patriarchal-clan organization, using slaves
obtained through capture o r purchase. With the further spread of horsc-brced-
ing and the development of bronze culture, the tribal-clan Plite grew in strength,
while the rank-and-file members of the tribal community were more con-
stricted. T h e development of property and social differentiations in society led
to the disintegration of the clan structure, and with the onset of the Iron Age,
quite large nomadic tribal unions came into being.
In the period f r o m the seventh t o the third century B.c., morc powerful
tribal unions arose in Inner Asia - the Hsiung-nu in Ordos and central Mongo-
lia, the Tung H u in eastern Mongolia and western Manchuria and the Yiich-
chih in Gansu and the lands between Dzungaria and Ordos. T o the south, Tan-
gut-Tibetan tribes led a nomadic way of life in the vicinity of Koko N o r (Qing-
hai). The Central Asian nomadic world was increasingly becoming a military
and political power.
Relations between China and the nomads were at times anything but
peaceful. China of the Ch'in dynasty (246-207 B.c.) built the famous Great
Wall,3 a military fortification running along the length of its frontier. In 214 B.C.
the Ch'in court sent an army of 100,000 men against the Hsiung-nu, captured
Ordos from them and then the foothills of the Yin-shan. Towards the close of
the third century B.c., as a result of their rising prosperity from cattle-breeding,
the development of their iron industry and military skill, the twenty-four Hsi-
ung-nu tribes increased considerably in strength; and from their tribal union the
powerful Hsiung-nu Empire emerged.
T h e dramatic events that attended the emergence of the nomadic
Hsiung-nu state find literary, albeit somewhat legendary, expressions in the
sources. A t the close of the third century B.c., a certain tribal chief, T'ou-man
by name, with the title of shan-yii - which meant 'the greatest' o r 'the best'-
headed the Hsiung-nu tribal union. According t o legend, he had two sons from
different wives. T o secure the throne f o r his favourite younger son, he handed

2. Sima Q i a n , 1931.
3. Bai Shouyi, 1980, p. 1 16.
over his elder son, Mao-tun, as a hostage to the Yiieh-chih. T'ou-rnan [Ilen
attacked the Yiieh-chih, hoping that they would kill their hostage, but Mao-tun
managed to steal a horse and return home. His father put 10,000 families under
his control. Mao-tun forthwith set about training his cavalry in the arts of war
and ordered all his horsemen t o shoot their arrows only in the wake of his whis-
tling arrow. Failure to comply with the order was punishable by death. &'hen
he saw that his warriors were adequately trained, Mao-tun, while hunting, shot
his father with an arrow and killed him o n the spot.
After 209 B.c., when Mao-tun proclaimed himself shan-yii, the H ~ i u n g - ~ ~
state rapidly became a powerful nomadic empire. Lung-chhg, 'The Dragon
Site', the nomadic tribal encampment and headquarters of the H ~ i u n ~ - ~ U
shan-yii, was located in the south-east spurs of the Khangay mountains, in a
region where Karakorum and other political centres of the Turkic and Mongol
peoples were later to come into being. The leader of the Hsiung-nu became the
keeper of the nephritic seal which was inscribed with the words: 'The state seal
of the Hsiung-nu shan-yii'. The rise of a Hsiung-nu state system, with a capital
for the shan-yii, a seal, flag, border guards and the other attributes of sover-
eignty, marked the beginnings of a distinctive nomadic power.'
The Tung H u , who heard that Mao-tun had killed his father, decided to
take advantage of the resulting confusion and demanded that Mao-tun should
surrender to them his treasured argamak (a fleet-footed horse) and his beloved
wife. Mao-tun agreed to both demands. They then demanded an uninhabited
strip of the desert, unsuitable for cattle-breeding, but Mao-tun answered: 'Land
is the foundation of a state. H o w can it be surrendered?' H e then launched a
military campaign against the Tung H u , who were taken completely unawares
and routed. O n his return, he attacked the Yiieh-chih, driving them westward,
subjugated the Wu-sun of Semirechye and recaptured Ordos.
In 198 B.c., a treaty of 'peace and alliance' was concluded between the
Hsiung-nu and China. The H a n emperor officially recognized that the Hsi-
ung-nu Empire enjoyed power comparable to that of his own empire, and that
its sovereignty extended over all the northern lands beyond the Chinese bor-
ders. The ruler of the Hsiung-nu, in turn, recognized the sovereignty of the
Chinese emperor over all territory behind the Great WalLS The treaty further
provided that the Han court should give the emperor's daughter in marriage to
the ~ h a n - ~and
i i should send him every year a lavish quantity of gifts - silks,
fabrics, handicrafts, rice, gold and money, which was regarded by the
Hsiung-nu as a form of tribute. The Hsiung-nu also received tribute from the
Wu-huan and subjugated other peoples; they sent their royal daughters to the
Wu-sun and held hostages.
Controlling a key section of the Silk Route, the caravan trade link

4. Konovalov, 1976, p. 3.
5. Taskin, 1968, pp. 25, 42, 47, 48.
 Nomads in eastern Central Asia

between China and the West, the Hsiung-nu reaped great profits from its inter-
national trade. They zealously fought to maintain control of these routes and
vied with China f o r the hegemony of Central Asia.
From the time of Mao-tun, there was regular trade between the kIsi-
ung-nu and the Chinese, the Hsiung-nu exchanging cattle, wool and furs for
Chinese Between 129 and 90 B.c., however, the H a n emperor Wu-ti
changed the policy of his predecessors and launched a number of major military
campaigns against the Hsiung-nu, but he was unsuccessful. Whcn the Han court
proposed that the Hsiung-nu should become a vassal state, they detained the
Chinese ambassador and refused t o discuss the matter, decapitating their own
master of ceremonies w h o had allowed the ambassador to enter his yurt.
Between 70 and 60 B.C. there was internecine war between various fac-
tions of the Hsiung-nu leadership seeking the throne. Whcn H ~ - h a n - ~ e h
became shan-yii, one of his brothers proclaimed himself shan-yii and attacked
him. Hu-han-yeh was obliged t o acknowledge his subordination t o China in 53
B.c., but managed t o preserve Hsiung-nu statehood with all its svmbols of
sovereignty. When he finally crushed his rival, Hu-han-yeh was able t o act
more and more independently and transferred his headquarters t o the Ulan
Bator region, where it remained after his death. But in A.D. 48, as a result of
worsening internal dissension, the Hsiung-nu split into two factions. The elders
of the eight southern tribes proclaimed the aristocrat Pi as shan-yii, migrated t o
China, fell under the sway of the H a n court and moved to Ordos, north and
west Shaanxi.
The Northern Hsiung-nu stubbornly defended their independence, but
eventually, in A.D. 93, they were attacked by an alliance of the Chinese and
Southern Hsiung-nu froin the south, by the Ting-ling, red-bearded, blue-eyed
giants from the north, and by the Hsien-pi from the east. The last Northern
shan-yii, descended from Mao-tun, was killed, and Mongolia was taken over by
the Hsien-pi. Some of the Hsiung-nu, notably those of the western branch, did
not surrender.
The Hu-yen, an ancient H u n tribe, assumed leadership and marched
westward. In contrast t o the Southern and Northern Huns, they may be
described as Western Huns, whose descendants would later reach Afghanistan,
India and the Roman Empire.

The Hsien-pi state

The Hsien-pi, w h o took over control of Mongolia after the fall of the
Hsiung-nu state, had emerged as a powerful tribal union as early as the first
century B.C. T h e main clan of tile Hsien-pi had set u p their nomadic cainp in
south-east Mongolia and lived along the middle course of the Liao-ho river.
A large number of H ~ i e n - n~oiw settled in central Mongolia and over 100,000
 N.Ishjamts

Hsiung-nu families, who had settled there earlier, adopted their tribal name.
T'an-shih-huai, leader of the Hsien-pi tribal union, in A.D. 155 established [he
Hsien-pi state, which rapidly became one of the most powerful empires of its
day, as powerful as the previous Hsiung-nu Empire. The Han court considered
that the Hsien-pi's horses were swifter and their weapons sharper than those of
the Hsiung-nu, and the Hsien-pi, too, managed to acquire good-quality iron
from the border regions of China. Their political centre, the headquarters of
T'an-shih-huai, was in the south-east near the Darkhan mountains but was later
moved to the former shan-yii's headquarters in the Khangay mountains."
Between A.D. 155 and 166, T'an-shih-huai conducted a series of major mil-
itary campaigns that led t o the extension of Hsien-pi power over the Great Steppe as
far as southern Siberia and from Ussuri to the Caspian Sea. Until the third decade of
the third century A.D. the Hsien-pi state was the leading power in Central Asia.
Under their rule Mongolia saw a complex ethnocultural development. From
the mixing of the Huns and Hsien-pi a new culture emerged with its own linguistic
particularities, which was later to serve as the point of departure for the formation
of the early Mongolian ethnic group with its distinctive language and culture.'
Subsequently the Hsien-pi state split into several parts. Until the close of the
third century, it only effectively controlled central and south-east Mongolia. The
Mu-yung, T'o-pa and T'u-yii-hun, kindred tribes of the Hsien-pi, seceded to set up
separate states in the south. At about the same time another Mongolian-speaking
people appeared on the scene in central Mongolia - the Ju-jan, who were destined to
play a key role in the history of the ~ e r i o dfollowing that of the Hsien-pi.g

The economy, social structure and state organization

of the Hsiung-nu
The major achievement of the ancestors of the Hsiung-nu was the opening up
of the steppes and the Mongolian Gobi. Until then, the Great Steppe and the
desert, like the sea, had divided the inhabited wooded steppe into two distinct
and separate belts. The inhabitants of each belt - agriculturists, sedentary cattle-
breeders and forest hunters - had n o way of crossing the Gobi, and the steppe
grasslands went t o waste unused. The Hsiung-nu bred a large number of horses
and draught oxen and introduced as a mobile home the covered wagon on high
wheels. They were the first t o engage in nomadic cattle-breeding (Fig. I ) and in
organized hunts - infinitely more productive than individual hunting - and by
the third century B.C. were practising falconry.'

6. Pcrlcc, 1961, p. 21.

7. Bira, 1977, yp. 379-80.
8. Ishjamts, 1974, pp. 24-6.
9. Gumilcv, 1960, p. 96.
 Apart from cattle-breeding and hunting, they engaged in agriculture, as
can be seen from the grain hullers found in Mongolia and in the regions of
Great Wall. War prisoners and deserters from China and other settled countries
were widely used as agricultural labourers.
Their covered wagons on high wheels, in which they lived, were comfor-
table, providing good protection from wind and frost, and more security,
because in the event of danger, the wagon-dwellers could flee with all their Pos-
sessions t o other encampments.
Although the Hsiung-nu Empire was made up of a large number of dif-
ferent peoples, the Hsiung-nu themselves were divided into twenty-four major
tribes, each consisting of kinship groups, clans and patriarchal families. The
Hu-yen, Hsu-pu and Lan were regarded as the oldest and most prominent, but
from the time of Mao-tun, the Hsiung-nu shan-yii came from another noble
family, the Luan-ti, that became the most distinguished of them all.
The tribal nobility formed the aristocratic Clite, while the rank-and-file
members of the tribe were relatively poor. There were quite a number of slaves
engaged in agriculture, handicrafts and cattle-breeding,1° but they were more
like domestic servants. It will be seen that Hsiung-nu society was in a state of
transition from a tribal t o a class system, and the Hsiung-nu Empire represented
a particular form of class-based state organization."
The ruler of the empire was the shan-yii. H e called himself 'ch'eng-li ku-
t'u' (Son of Heaven).I2 His power was considerable and hereditary but by no
means absolute. In their administration it is even possible t o identify several
classes of officials or, to be more precise, nobles divided into eastern and west-
ern groups, terms which also signify 'senior' o r 'junior'. The first class consisted
of the Chu-ch'i princes ('chu-ch'i' meaning 'wisdom'). The Eastern Chu-ch'i
prince was supposed t o be the heir apparent, but at times his right to succeed to
the throne was disregarded. The second class consisted of the Lu-li princes; the
third class, the Great Leaders; the fourth class, the Great Tu-YU;the fifth class,
the Great Tang H u . In addition, the Eastern and Western Chu-ch'i princes and
the Lu-li princes were called 'four horns' and the 'great leaders', Tu-yu and
Tang H u were called 'six horns'. These high-ranking figures were always mem-
bers of the shan-yii's clan.
Alongside this aristocracy of blood there grew up an aristocracy of talent
- the service nobility (not related t o the shan-~ii'sfamily). They were known by
the name of Ku-tu-hou, and were aides of the highest-ranking nobility, per-
forming all the administrative tasks. Apart from this top-level aristocracy, there
was the clan nobility - princes affiliated exclusively with the clans, sui gcneris
clan chiefs o r elected elders.

10. De Guigncs, 1756-58, Vol. I, Part 2, p. 15.

1 I . Harmatta, 1952, p. 287.
12. Ban Gu, Vol. 13, Chaptcr 973: 7a.
 Nomads in eastern Central Asia

Hsiung-nu society possessed its own customary legal systen~and Chinesc

authors have noted that 'their laws were simple and easily executed'. Major
crimes, such as the drawing of a sword, were punishable by death and theft was
punished by confiscation not only of the thief's property but also that of his
family. Minor crimes were punished by cuts o n the face. Trials lasted no more
than ten days, and at n o one time were there ever more than a few dozen people
under arrest. Apart f r o m the customary law a system of public law began t o
emerge under Mao-tun. Violation of military discipline and evasion of military
service both carried the death penalty. These extraordinary laws contributed
greatly to strengthening the cohesion of the Hsiung-nu, turning them into the
most powerful state in Central Asia.

Hsiung-nu burials and the finds from Noin-Ula

The main sources f o r the study of the Hsiung-nu are their graves and settle-
ments, the latter t o a lesser degree in view of their nomadic way of life. Many of
them are to be found in Mongolia, southern Siberia and Ordos. There are four
major Hsiung-nu burial sites: two in central Mongolia and two in the south
beyond Lake Baikal. T h e largest, the Khunui-go1, is located in a remote area of
the Khangay mountains, in the basin of the River Khunui. It was here, in 1956,
that T. Dorzhsuren and other Mongolian archaeologists counted over 300
burials." In the mountains of Noin-Ula, 122 km north of Ulan Bator, 212 burial
grounds have been recorded. A t the end of the last century, the Russian archae-
ologist Y. D. Tal'ko-Grintsevich located 214 graves at Sudzhinsk beyond the
Baikal, 10 k m east of Kyakhta. H e also excavated the Derestui burial ground in
the former Selenga District beyond Lake Baikal, with some 260 burials. It seems
likely that these four large burial grounds belonged t o the four noble clans of
the Hsiung-nu, and that Mao-tun and other early shun-yii were interred in the
Khunui-go] burial ground, which contains some of the most magnificent Hsi-
ung-nu tombs, while H ~ - h a n - ~ eand h his descendants were buried in Noin-
Ula.
Of particular interest are the contents of the burial chamber of a tomb in
the Noin-Ula mountains accide~ltallydiscovered in 1912 by a Mongol Gold
engineer and scientifically excavated in 1924. Like other tombs nearby, Noin-
Ula Tomb N o . 6 was square, with sides measuring 24.5 m and an embankment
1.62-1.95 m high. O n the south side it was also protected by a long bank. T h e
sides of the square and the bank were faced with stone and aligned to the points
of the compass. In the inner chamber stood coffins, pointing south, o n a floor of
planks which showed faint traces of lacquer and paint. Among the objects
found was a woollen canopv covering the ceiling of the outer chamber and a

13. Dorzhsuren, 1958, p. 6 .

heavy felt carpet, with scenes of animals locked in combat, lying under the
fin. A woollen cloth with embroidered plant motifs and figures of different
mals was affixed to the ceiling of the outer chamber, covering practically it,
whole surface. The fortunate discovery of a Chinese lacquer cup with two
inscriptions made it possible to date T o m b N O . 6 quite closely to the beginning
of the first century A.D. It is, in fact, the tomb of Wu-chu-lii, the ~ h a n - ~ofi ithe
Hsiung-nu Empire.
T o judge from the finds in the Noin-Ula tombs, permanent dwellings of
the Hsiung-nu were equipped with plank beds, and their mobile dwellings were
furnished with low tables on short legs. The height of these tables indicates that
people sat round them o n the floor, which was covered with heavy felt. ~t is
interesting t o note that such small, low tables were extensively used by Central
Asian cattle-breeding peoples.
The Noin-Ula tombs contained a large variety of Hsiung-nu vessels of
wood, metal and clay. The most remarkable metal vessel was a bronze oil-lamp
mounted on three legs with a conical stem for a wick in the centre. There were
fragments of a big bronze kettle for cooking meat, and a smaller kettle, notable
for its handles in the shape of animal heads. Among other finds were minute
pyrite crystals, pierced with holes and used as dress ornaments, beads of mala-
chite and glass of different colours and amber beads of various shapes and sizes.
Of particular interest was an amber bead in the shape of a lion's head. Apart
from the beads contained in Hsiung-nu graves of ordinary type, mainly of
women, bronze mirrors and various dress ornaments that belonged to the
various shan-yii have come t o light.
Their main form of transport was the saddle-horse. Horse bridles, bits
and saddles resembled those of the Altai in Scythian times. Saddles generally
consisted of two leather cushions padded with wool, but a few had a wooden
frame with pommels at the front and back, and stirrups. The stirruped saddle
represented a major step forward.
In spite of the fact that iron and smelting techniques were known to the
Hsiung-nu and that they could ~ r o d u c evarious kinds of iron objects, they did
not strike fire from a flint, but obtained it by rubbing two sticks together or,
more exactly, by boring one into the other. The Noin-Ula graves, thanks to the
fine state of preservation of the wood found there, ~ r o v i d ea full range of arti-
cles used for fire-making.
The most remarkable piece of bone jewellery work was a carved cylinder
representing a winged and horned mythical wolf. Especially notable were the
embroidered felt carpets of local manufacture found in the Noin-Ula tombs.
The seams of the middle section of the carpets were embroidered in a distinctive
spiral pattern, and their borders were covered with a design in which scenes of
fighting animals alternated with tree patterns.
The Hsiung-nu were in direct and close contact not only with China but
with neighbouring peoples t o the cast and west, who were culturally very much
 Nomads In eastern (it,ntrul A m

FIG.2 . Hsiung-nu felt carpct from Noin-Uln (Mongoli,l).

like them. A remarkable bronze crown in the shape of a wolf's head from Noin-
Ula may be compared t o the figures of wolf heads in the art of the Altaic Scl-th-
ians. Other Hsiung-nu articles in Mongolia and Ordos display striking similari-
ties with southern Siberian works of art. Animals are portraved with protruding
haunches - a style very characteristic of the art found here.
Of particular interest are the scenes of combat betwcen a !yak and a
horned 'lion' and a griffin attacking a deer, on the felt carpets of Noin-Ula
(Figs. 2-4). T h e yak is highly stylized with an abnormally large head hanging
low and a protruding tongue. Scenes of beasts of prey attacking artiodactyla
have been characteristic of the art of the peoples of Western Asia since earl!.
times - a motif that entered Asia Minor from Mesopotamia and spread through
the Sakas to southern Siberia and then t o the Hsiun,--nu.
Motifs borrowed f r o m the plant kingdom are exceedingly rare in the art
of Eurasian nomadic tribes, such as the Hsiung-nu. All the more interesting,
therefore, are the conventional representations of trees on the Noin-Ula carpets
in the spaces between the animal combat scenes. Here nre have a replica of the
'sacred tree', a typical feature of Assyrian art. Hsiung-nu felt carpets were dcc-
orated with borders of squares, crosses, 'battle-axes' and other figures. The
most widespread motif on the seams of the felt carpets consisted of rhomboids
or spirals arranged in t w o varieties. It is interesting to note that this Hsiung-nu
lllotif later found widespread application in the art of the Mongolian and Turkic
peoples and can still be seen o n the felt carpets and the protective covers
of the Buryats, Kyrgyz and Kazakhs.
While there is 110 real trace of the influence of Chinese art on ohiects
 FIG.3. Hsiung-nu felt carpet from Noin-Ula (Mongolia).

FIG.4. Hsiung-nu felt carpet from Noin-UIa (Mongolia).

found in the graves of common people, traces are t o be found in those of the
nobility. T h e mythological animal embroidered o n the silk fabrics found in
Noin-Ula is essentially Chinese in character. Its body resembles an eagle with
upraised tail, of which the tuft is abnormally large. T h e animal's paws are like
the tiger paws in the art of the Altaic and southcrn Siberian tribes of Scythian
date. T h e front of the chest is represented by a succession of superimposed
 Nomads in eastern Central Asia

scales,and the wings are in the Persian style of the Achaemenids. 'Thus we have
here an example of the influences of Hsiung-nu, Altaic, Scythian and Persian art
fabrics of Chinese character.

Hsiung-nu customs, religion and culture

The influence of the Hsiung-nu and also of ' ~ i d d l eand Western Asia on China
was especially great in the military field. As early as 307 B.c., Wu Ling Wang, a
prince of the C h o u dynasty, introduced the use of the dress of the nomadic H u
into China and began t o instruct his subjects in the art of shooting with bow-
and-arrow. The Emperor Ch'in Shih-huang-ti introduced large cavalry detach-
ments into the Chinese army and thereby ensured the success of his operations
against the Hsiung-nu at MCn Ch'ien Yang. His cavalrymen were heavily armed
and armoured, like those of the Assyrian army which had introduced cavalry as
an arm of their military organization, and already used chain mail, plate armour
and p-otective armour for horses.I4
When Mao-tun reorganized the Hsiung-nu army, he replaced its heavily
armed horsemen by light cavalry, armed with long composite bows, creating a
military force with much greater manoeuvrability. H e reorganized his army,
subjected it t o strict discipline, and introduced major improvements in military
strategy and tactics. These developments in military science and weaponry were
of great importance both for China and other countries. The composite bow
and the stirruped saddle were widely adopted throughout the Eurasian steppe,
Parthia and Syria; and the descendants of the Hsiung-nu with their traditional
bows later became one of the most dangerous adversaries of the Roman legion-
aries in Pannonia.I5
The emergence of its powerful empire had a great influence 011the Hsi-
ung-nu's material and spiritual life. Although it could not radically change their
nomadic ways, it led t o the establishment of an entirely new central head-
quarters for their ~ h a n - ~ini i the Khangay mountains where, in addition t o his
residence, they erected a sanctuary and other buildings. Settlements appeared in
the steppe - fortified places f o r agricultural and craft communities like Gua-dov
(367 x 360 m), Baruun dereegiin kherem (345 x 335 m) in Mongolia and the set-
tlement at Ivolginsk (348 x 200 m) beyond Lake Baikal. In the Talas \.alley
under the rule of Chih-chi11 a fortress showing distinct Roman influence was
built, and more than 100 foot-soldiers were garrisoned there. It has been sug-
gested that they may have been Roman legionaries from the defeated army of
Crassus who had surrendered t o the Parthians after the battle of Carrhae in
53 B.C. and been sent t o serve on their eastern frontier at Merv, from where they
became mercenaries of the Hsiung-nu. Nevertheless, the shan-yii of the ~ ~ i -
ung-nu did not alter their way of life. They continued t o receive ambassadors in
their yurts which were now more presentable than in earlier days. The over-
whellning majority of the Hsiung-nu, especially the rank-and-file, also conti-
nued to live in yurts. The image of a yurt of that period, a covered wagon ol,
high wheels drawn by oxen, found on one of the south Siberian stone sculp-
tures, and the miniature images of a harness yoke found among the south Siber-
ian and north Chinese bronze articles, need not cause any surprise.'b In these
yurts of the Hsiung-nu, as later in those of the Mongols and Tiirks, the left side
of the entrance was for men and the right side for women.
Hunting and archery played a major role in their daily life and existence;
and a reference in the Shih-chi shows that they actively encouraged their chi]-
dren to learn hunting from an early age: 'As soon as a boy is able to ride a ram,
he shoots birds and small game with a bow, and when he gets t o be a little older,
he shoots foxes and hares.'" The Hsiung-nu women were not only remarkable
horse-riders but had bows and arrows, and assisted their husbands in defending
children and old people from enemy attack. O n the wall of the Talas fortress
they fought valiantly beside their husbands and the Romans against the Chinese
troops, and they were the last to leave their posts. The participation of women
in defence and the training of children to handle the bow-and-arrow were evi-
dently dictated by the necessities of life. Only in this way could the nomadic
Hsiung-nu, a numerically small group, defend their independence and way of
life.
The Hsiung-nu practised exogamy, but their shan-yii could only take
wives from a limited number of noble clans. They practised ~ o l ~ g a m and
y levi-
rate marriage was customary, that is, on the death of a father or elder brother, a
wife was transferred to a surviving son o r younger brother, ~ r o v i d e dthat she
was not that man's mother. The shan-yii's court also complied with this prac-
tice.
The Shih-chi says that on the death of a Hsiung-nu emperor his close rela-
tives and concubines were buried with him, but archaeological excavations do
not bear this out. If such a custom had ever been practised by the Hsiung-nu,
the actual burial of ~ e o p l ehad long been replaced by symbolic actions. After
the death of her husband, a Hsiung-nu woman would lace a lock of her hair in
her husband's grave as a sign of mourning, symbolizing her journey to the next
world t o accompany him.
. ..
The Hsiung-nu lnltlally believed in animism, totemism and in life beyond
the grave. From the time of Mao-tun, Shamanism became the state religion. The
chief shaman was chosen from the sorcerers and served the s h a n - ~ i i ,his clan
and relatives. The head-dress of a shaman was found in one of tllc graves of

16. Bira e t a]., 1984, p. 48.

17. Bichurin, 1950, Vol. I, p. 40.
 Nomads in eastern Central Asia

Noin-ula and was very reminiscent of the Mongol darkhansP head-dress, the
,,ly difference being in the frontal representation of the totem spirits. On
Hsiung-nu head-dress there is a bird, which may represent the face of
a n t h r o p ~ m ~ r pspirit
h i ~ of an ancestor.
The Hsiung-nu worshipped the sun, the moon and other heavenly bodies
and made sacrificial offerings t o the heavens, the earth, spirits and their ances-
tors. The shan-yii described himself as 'born of heaven and earth, brought forth
by the sun and moon'. T h e Shih-chi says: 'At daybreak the s h ~ n - sets ~ i out
from camp t o worship the rising sun, at nightfall to worship the moon.vx The
~ s i u n g - n unobles gathered at the shun-yii's headquarters in the fifth lunar
month and made sacrificial offerings t o their ancestors, the heavens, the earth
and the spirits. Three times a year they congregated at the shrine of the moon
where, on the day of the 'dog' of the first, fifth and ninth months, they offered
sacrifices to the heavenly spirit.
It is not known whether the Hsiung-nu had images of their ancestors o r
spirits. Of great interest, in that regard, was the discovery, in the Noin-Ula
tomb, of a translucent stone 5 m m thick, with a schematic incised representation
of a human figure. Three holes bored into the figure indicate that it was attached
to something. I t may have been some kind of anthropomorphic amulet. The
Hsiung-nu prince w h o ruled over the region of Koko N o r and Gansu wor-
shipped a huge 'golden idol'.
In reckoning time the Hsiung-nu used a duodecimal animal cycle in
which the days of the 'dog' and the 'snake' were regarded as auspicious for wor-
ship. This duodecimal cycle, which reached the Hsiung-nu from India o r Baby-
lon via Sogdiana, remained in force for as long as the Hsiung-nu existed. Their
basic system of calculation, however, was decimal, and they used this in their
military organization. They customarily launched a military expedition at the
time of the full moon, but its outcome hinged o n the prophecies of the shamans,
sorcerers and soothsayers. Like the Mongols, they used a ram's shoulder-bone
to predict the future, placing the bone in a fire, and reading the future from the
lines which appeared o n it.
The Chinese sources say that the Hsiung-nu did not have an ideographic
form of writing as the Chinese did, but in the second century B.C. a renegade
Chinese dignitary by the name of Yue 'taught the shan-YUhow t o write official
letters to the Chinese court o n a wooden tablet 31 cm long, and t o use a seal and
large-sized folder'. But the same sources indicate that when the Hsiung-nu
noted down something o r transmitted a message, they made cuts on a piece of
wood (k'o-mu) and they also mention a ' H u script'. The fact is that over twenty
carved characters were discovered among the objects at Noin-Ula and other
Hun burial sites in Mongolia and the region beyond Lake Baikal (Figs. 5 and 6).
Most of these characters are either identical o r very similar t o letters of the

IS. Ibid., p. 50.

 N.Ishjamts

FIG.5. Runic characters of Hsiung-nu-Hsien-pi script

(Mongolia and Inner Mongolia).

FIG.6. Runic characters of Hsiung-nu-Hsien-pi script

(Mongolia and Inner Mongolia).
 Nomads in eastern Central Asia

orkhon-Yenisey script of the Turks of the Early Middle Ages that occurs now
and again in the Eurasian steppes. From this some specialists hold that the
Hsiung-nu had a script similar to ancient Eurasian runiform, and that this
itself later served as the basis for ancient Turkic writing.
~ ~ t hlegends
s , and other forms of oral literature occupied an important
place in Hsiung-nu spiritual life. Tradition has it that in front of t h e head-
quarters of the shan-yii there was an artificial pool, the dwelling place of a dra-
gon who had fallen from heaven and become an object of worship. It was not
only the Chinese w h o worshipped the dragon. While the sources portray the
~ s i u n g - n uas a redoubtable nation of fierce warriors, they were actually fun-
loving people. They would gather before the headquarters of the ~ h a n - ~and ii
the temple of their ancestors to organize amusements like horse-jumping and
camel races and other festivities. They would sing slow songs, a custom that
later became widespread among the Mongolians. The sounds of flutes and
drums and the strains of a few types of string instruments were heard through-
out the steppes. As early as the beginning of the second century B.C. Hsiung-nu
music and dances were favoured by the Han emperor. The k'ung-hou and the
fife, which had come at an earlier time to Inner Asia from Sogdiana, were
adopted by the Chinese from the Hsiung-nu.
Many Chinese silks and embroidered fabrics as well as cloth of Western
origin were found in the Noin-Ula tombs. There was one woollen wall carpet
of Western manufacture and two of Yueh-chih or Wu-sun origin, a series of
embroideries provisionally described as 'Graeco-Bactrian', two portraits dis-
playing similar workmanship and, finally, a tapestry from Parthia or Asia
Minor. There were also some remarkable fragments of wall hangings, decorated
with patterns representing horsemen, a child warrior flinging a spear or a trident
at an eagle, lion griffins and garlands of flowers that have been described in
detail by K. V. Trever.Iy The two horsemen depicted on the largest fragment are
of Europoid type, with moustaches and tufts of hair over their foreheads remi-
niscent of Gandhiran sculpture. Their dress and head-dress are typically Ira-
nian. The horses are notable for their coats of different colours, their cropped
manes, their elegant long-necked heads and their long slender legs. They were
probably the famous thoroughbred Parthian war-horses or the remarkable
'thousand-li' horses of the Hsiung-nu. Along the lower edge of the fragment,
between two broad bands, is an embroidered garland of ~ a l r nleaves inter-
spersed with acacia flowers, their tendrils entwined - a border that is Greek or
Graeco-Indian in design.20 Of the two portraits found on the fragments, one has
preserved the face of a man whose features are not Mongolian, but bear a much
closer resemblance t o the Turanian type found in Central Asia and Persia.
Trever compared this portrait to the heads found on the bas-reliefs of Gand-

19. Trever, 1940, pp. 141-3, Plates 3 9 4 4 .

20. Rostovtzeff, 1929, p. 87.
 N.Ishjamts

haran art and concluded that it represented one of the peoples of Central Asia
whose culture was in contact with the Hellenized culture of Central Asia, such
as that of the Graeco-Bactrians. It is clear that these works are representative of
a great artistic culture and consummate craftsmanship.
This nomadic civilization of the Hsiung-nu exerted an influence even on
the more advanced cultures of China.*' We might mention in this context [he
Hsiung-nu treatment of the tiger, one of the prototypes of the Chinese tao-ti,
of winged wolves, goats and horses and, finally, of the bizarre 'aquiline griffon',
a fabulous animal with the body of a winged lion and the head of a phoenix.

Hsien-pi culture
Hsien-pi culture was also suggestive of Hsiung-nu culture in many ways; but so
far it has not been adequately studied. Between the first and third centuries A.D.
it attained a similar level t o the culture of the Hsiung-nu. According to histori-
cal sources, the Hsien-pi also recorded events by incising wooden tablets. In
their practice of Shamanism, they initially worshipped a wooden idol. However,
later on, the casting of idols from iron and other metals (Fig. 7) became wide-
spread among the Hsien-pi, as among many other nomadic peoples of Central
Asia. In later centuries, other branches of the Hsien-pi tribe, the Mu-yung,
T'o-pa and T'u-yii-hun in Inner Mongolia and northern China, created a more
advanced form of culture, inheriting many elements from the Hsiung-nu and
also borrowing a great deal from the neighbouring countries of Central Asia,
China, East Turkestan and southern Siberia. In turn, the Hsien-pi also exerted
an influence on them.
The original and distinctive culture of the Hsiung-nu and of the Hsien-pi
together constituted the first important stage in the formation of the nomadic
civilization of Central Asia, playing an important role between East and West
and linking China and Central Asia, while, at the same time, remaining dis-
tinctive - a very significant contribution to world culture.

21. Serodzhav, 1977, pp. 4, 22, 106, 109-1 I .

168
 Nomads in eastern Central Asia

FIG.7. Hsien-pi terracotta from Inner Mongolia.

 THE Y U E H - C H I H
A N D T H E I R MIGRATIONS:'

K. Enoki, G. A. Koshelenko
and 2. Haidary

The Yiieh-chih in Gansu

The Yueh-chih are the people who, from the latter half of the second century
B.C. to the beginning of the first century A.D., occupied and ruled what is now
Middle Asia and Afghanistan. They destroyed a country called Ta-hsia -
usually identified with the Bactrian kingdom under the Greeks. Their original
home is said t o have been in the western part of Gansu Province in China, from
which they migrated via the northern part of the T'ien Shan mountain range.
They were forced t o migrate because of the invasion of the Hsiung-nu, w h o
became predominant in Central Asia. The Yiieh-chih settled to the north of the
Amu Darya and ruled the country of Ta-hsia, which lies south of the same
river. There they established five hsi-hou (yabghu) o r governor-generalships. At
the beginning of the first century A.D., the Kushans, one of the five yabghu,
became powerful and took the place of the Yueh-chih. This is an outline of the
history of the Yueh-chih. Nothing more is known about them except that the
Chinese emperor Wu-ti (140-87 B.c.) of the Former H a n sent Chang Ch'ien t o
the Yueh-chih t o negotiate a military alliance against the Hsiung-nu, but this
was not concluded.
Many theories have been published about who the Yueh-chih were. The
oldest theories were based solely o n the resemblance of the name of the Yiieh-
chih to that of the Goths o r Massagetae. But the Goths could have had nothing
to do with a people in Central Asia in the second century B.C. N o r can the iden-
tification of the Yiieh-chih with the Massagetae, who lived in the plain east of
the Caspian and Aral Seas in the fifth century B.c., be maintained.
The identification of the Yiieh-chih with Casia' is also based o n the simi-

:' Sec Map 4.

1. Egarni, 1948, pp. 84 et seq., 1951, pp. 123 et seq.
 K. Enoki, G. A. Koshelrnko a n d 2. Haidary

larity of the names Yueh-chih and Casia, but there is some additional positive
evidence to support it. 'Casia' is the name given by the Greeks in the first ten-
tury A.D. to the Kunlun mountain range in the south of the Tarim basin and t~
the region stretching north of it, which is famous even today for the production
of jade. According t o the Book of Kuan-tz6, jade was produced either in the
country of the Yu-chih, who are considered t o be identical with the Yijeh-chih,
o r in the mountains on their frontier. The Book of Kuan-tz6 is some time before
the third century B.c., when the Yiieh-chih dominated the greater part of Man-
golia. So it is quite possible that 'Yu-chih', 'Yueh-chih' and 'Casia' represent the
same name; and that the Yueh-chih were known to the Chinese to be associated
with jade.' Presumably jade was known by the name of casia because it was pro-
duced in the country of the Yueh-chih, or the Yueh-chih were known by the
name of Casia because of their jade. In a place near modern Khotan in the
ancient region of Casia, jade is still called gutscha; and 'gutscha' is very similar
t o the old ~ r o n u n c i a t i o nof Yueh-chih, which may have been 'zguja" or some-
thing like that. If the jade was called casia because of the Yueh-chih, the country
of Casia might have been the place where the Yueh-chih originated. But the
Yueh-chih were a great horde of pastoral people, and had 100,000 or 200,000
cavalrymen, according t o the Shih-chi (Book 123), when they reached the Amu
Darya. This makes it unlikely that they could have originated in a place such as
Casia where the oases could only support a population of a few thousand at the
time of the H a n dynasty. It must also be remembered that n o other nomadic
people has ever risen t o power in any part of the Tarim basin where Casia was
situated. If the Yueh-chih were called by the name of Casia, because of the casia
o r jade they produced, they must also have had another name of their own.
What is certain, however, is that the region of Casia and other countries in the
Tarim basin were under the control of the Yueh-chih; although it is most likely
that Casia was the native lace of the Yueh-chih.
Recently Yang Hsien-i4 has identified Chu-chih in the Tso-chuan under
the fourteenth year of Duke Hsiang (559 B.c.) with the Yueh-chih. There is
indeed a resemblance between the two names, and there is the statement of
Wu-li, the ancestor of Chu-chih, who was deported t o Kua-chou, which the

2. Concerning the Yu-chih as described in the Book of Kuan-tzci, see Matsuda, 1939, and
Kuwabara, 1940, pp. 8-9, 71. Sec also Wang, 1927, and Pelliot, 1929, p. 150. Thc latest
publication on the Book of Kuan-tzci conccrning the parts in which references are made
to the Yu-chih is Ma, 1979, Vol. 1, p. 255; Vol. 2, pp. 41 I, 429, 460, 462, 560, 569-70.
According t o Profcssor Ma, these parts of the Kuan-tzci were c o ~ n ~ i l eind its present form
at various times from the beginning of thc Former H a n t o thc rcign of Wang Mang, that is
t o say, from 200 K C . to A . I I . 12. Professor Ma also trics to establish thc identity of the
Yu-chih with the Yueh-chih in his other book (Ma, 1982, pp. 476-7).
3. Reconstruction made by Haloun, 1937, p. 316. See also a new rcco~~struction made by
Pulleyblank, 1966, p. 17.
4. Yang, 1983a, yp. 232-3. Another book of thc same content was publisl~cdby thc same
author (Yang, 198317, pp. 232-3).
 The Yueh-chih and their migrations

~ " ~ h - c h ioccupied
h in a later period. But Chii-chih is called lung-tz8, o r a bar-
barian of the West, in the Tso-chum, and the tribe t o which he bclonrcd
ch9iang-jung,o r the Ch'iang barbarians of the West. In other words, C h ~ - ~ h i h
iflot a tribal name, but a personal one. Moreover Kua-chou, t o which
ancestor of Chii-chih was exiled, was not the place where he lived permanently.
For these reasons, even if Chii-chih (a personal name) can represent the same
sound as ~ i i e h - c h i h(a tribal name), Chii-chih cannot be regarded as identical
with the Yiieh-chih.
According t o Strabo, the Bactrian kingdom was destroyed by the invasion
of four peoples: the Asioi, the Pasianoi, the Tocharoi and the Sakarauloi, all of
whom came from beyond the Syr Darya. According to Pompeius Trogus, thc
Asiani were lords of the Tocharians and conquered the Sacaraucae. Szii-ma
Ch'ien, w h o undoubtedly derived his information from the report of Chang
Ch'ien, states that the Yueh-chih conquered and ruled Ta-hsia. Attempts have
been made in the past t o reconcile the three statements by identifving the Yiieh-
chih with one of four peoples, in the belief that 'Ta-hsia' meant the Gracco-
Bactrian kingdom and that Szb-ma Ch'ien tells us of the conquest of that king-
dom. Some scholars think that the Asioi, the Asiani and the Pasianoi are one
and the same people, w h o are t o be identified with the Yiieh-chih, 'Pasianoi'
being a corruption of 'Gasianoi'. Others suggest that the Tocharoi should be
identified with the Yueh-chih because the Yueh-chih occupied the western part
of Gansu Province, which is called Thogara in the Geography of Ptolemy;
because Tocharistan was where the Bactrian kingdom was situated; and because
the Asiani, as described by Pompeius Trogus, were the ruling family of the
Yiieh-chih.
It is, however, uncertain whether the country of Ta-hsia in Szii-ma Ch'ien
means the Bactrian kingdom under the Greeks. According t o Szii-ma Ch'ien,
the country of Ta-hsia had n o big o r ~ o w e r f u lking but only small chiefs in a
number of cities. If the country of Ta-hsia was the Bactrian kingdom, it must
have been under a king w h o controlled the whole of Bactria. Szii-ma Ch'ien
also states that the people of Ta-hsia were skilled traders, but their soldiers were
weak in warfare and disliked battle, and, for these reasons, Ta-hsia was con-
quered by the Yiieh-chih. These statements are not applicable t o the Greek
kingdom of Bactria, whose soldiers seem t o have been skilful in fighting. As
'Ta-hsia' is an exact transcription of 'Tochara' (which was the central part of the
Bactrian kingdom), if the Yiieh-chih were the Tocharians, the conquest of Ta-
hsia by the Yiieh-chih means the conquest of the country of Tochara by the
Tocharians, which seems rather strange. The evidence of Szb-ma Ch'ien shows
that Ta-hsia cannot be the Bactrian kingdom, but was the country of Tochara
divided into several small units at the time of the Yiieh-chih invasion.
In other words the Graeco-Bactrian kingdom had already been destroyed or
divided when the Yiieh-chih arrived. Therefore, there is no need t o accept the
identification of the Tocharas with the Yiieh-chih.
 K. Enoki, G. A. Koshrletzko and Z. Haidary

A third group of scholars maintain that the Yiieh-chih were the Kushans,
As already mentioned, the Kushans were originally one of the five p b g h u
established by the Yueh-chih. This view therefore holds that
the other four governorships were also of the same stock. It seems convincing
when the word 'Kushan' is explained as a genitive plural form of the root
'Kuia', which means 'Yiieh-chih'. But the statement in the Han-shu (Annals of
the Fornler Han) about the establishment of the five tribal chiefs suggests that
they were natives of Ta-hsia who were conquered by the Y i i e h - ~ h i hMoreover,
.~
if the name 'Kuiin' is identical with 'Yiieh-chih', we wonder why the author of
the Han-shu uses different Chinese characters with different sound values to
represent the two names.

The Yiieh-chih and the Scythians

It seems most plausible to identify the Yueh-chih with the Scythians - a solu-
tion that fits the situation of Central Asia in the third and second centuries B.C.
better than any other theory. The name of Yueh-chih in Archaic Chinese, 'zngi-
wit-t'ia', the barbarian prototype of which might have been "Zguja, can be
considered as a transcription of the name of the Scythians. This suggestion was
first made in 1935 by Haloun, who held that the Chinese knew the Tocharian
people under the name of the Yueh-chih o r Scythians. It may also be proposed
that the Yueh-chih were not only called Scythians, but were Scythians them-
selves. According t o Szti-ma Ch'ien, up t o the beginning of the third century
B.c., it was the Yueh-chih and the Tung H u , the two dominant powers in the
Mongolian plain, who pressed the Hsiung-nu from the west and east respec-
tively. At that time, the Hsiung-nu were a small community of tribes. Only
when they were unified under the able leader Mao-tun could they push the
Yueh-chih further westwards and the Tung H u eastwards. Mao-tun himself was
taken as a hostage by the Yueh-chih in his early days.
Mao-tun was inspired by the unification of China by the Emperor Ch'in
Shih-huang-ti in 221 B.c.; the first blow of the Hsiung-nu against the Yiieh-chih
was given about this time. It resulted in the withdrawal of the Yueh-chih to the
western part of Gansu Province. The Hsiung-nu were originally a pastoral peo-
ple in the steppes north of the Yin-shan mountain range. Szti-ma Ch'ien states
that the Yueh-chih were the only people who ~ r e s s e dthe Hsiung-nu from the
west. This may mean that the Yueh-chih were seeking t o control the greater
part of the Mongolian plain.

5. Kuwabara, 1940, pp. 42-6.

 The Yueh-chih and their migrations

The Yiieh-chih Empire

204-200 B.C. Mae-tun conquered Mongolia and subjugated
peoples. In 176 B.C. he defeated the Yueh-chih in the western part of Gansu
province. In his letter t o the Han, Mae-tun said that the H ~ i u n ~ had - ~ udes-
troyed the Yueh-chih; and LOU-Ian, Wu-sun, Hu-chieh and twenty-six other
in the neighbourhood were subjugated to the Hsiung-nu. It is an
exaggeration t o say that they destroyed the Yueh-chih, but it is clear that the
yiieh-chih were driven from the west of Gansu and probably moved from the
north of the T'ien Shan mountains. What is important is that all these countries
were subjugated as a result of the defeat of the Yueh-chih, that is, they had becn
under the control of the Yueh-chih u p t o that time.
Lou-Ian, later called Shan-shan, is a country near Lop N o r in the eastern
part of the Tarim basin. T h e Wu-sun were a pastoral people in the region of the
River Ili and Lake Issik-kol, north of the T'ien Shan mountains. The twenty-six
other countries seem t o have been small states in the Tarim basin, probably
including Casia mentioned above. As regards Hu-chieh, no definite identifica-
tion has been made. T h e late T. Fujita was of the opinion that it is identical with
the Uighur of a later p e r i ~ d According
.~ t o the Hun-shu, in the mid-first cen-
tury B.c., five kings competed with each other for the leadership of the
Hsiung-nu Empire, one of them named Hu-chieh-wang, o r King of Hu-chieh,
which lies t o the west of the Hsiung-nu Empire. A few years later Chih-chih
shun-yii became independent from the Hsiung-nu Empire, occupied the north-
west part of the Mongolian plain and successfully defeated the army of the W u -
sun. Subsequently he marched northwards, subjugated the U-chieh, and, turn-
ing to the north, conquered the Ting-ling. The U-chieh are said to be identical
to the Hu-chieh. Taking these statements into consideration, the Hu-chieh may
be laced in what is n o w Dzungaria o r the upper waters of the River Selenga.
Fujita's view, therefore, seems t o be quite close t o the truth.
The dominion of the Yueh-chih also extended as far south as the upper
waters of the Yellow River. Szti-ma Ch'ien states that a small part of the Yueh-
chih, who could not follow the migration of the main horde towards the west,
settled themselves in Nan-shan, the southern mountains, which separate the
upper waters of the Yellow River people from the Mongolian plain.
The date of migration of the Yueh-chih t o Ta-hsia is not clearly known.
The Hsiung-nu gave a third blow t o the Yiieh-chih during the reign of Lao-shang
~ h a n - ~(c.
i i 174-161 B.c.). But it is generally believed that the conquest of Ta-hsia
was made some time between 139 and 128 B.c., that is, between the departure of
Chang Ch'ien and his arrival at the court of the Yueh-chih. So it is not certain
when the small part of the Yiieh-chih settled in Nan-shan, though it is quite
likely that it was in 176 B.C. when the Yiieh-chih were forced to evacuate the

6. Tizai Koshishi no kc~zkj~rj,

Saiiki-hsl4, 1943, pp. 61, 64.
 K. Etzoki, G. A. Koshelenko and Z. Haidary

western part of Gansu Province. In any case, it is generally understood that this
s~nallpart of the Yueh-chih asked the Ch'iang (or Tibetans on the upper waters
of the Yellow River) for protection and whether they could stay with them.
may, however, be suggested that these Ch'iang people had been under the rule of
the Yiieh-chih and that the sillall group of the Yueh-chih who were later called
the Little Yiieh-chih, t o distinguish them from the Great Yueh-chih (or Ta-
Yiieh-chih) in Ta-hsia, were earlier rulers of these Ch'iang people. They never
asked for protection, but actually lived with the Ch'iang who were their subjects,
When China was unified by Ch'in Shih-huang-ti, the upper waters of the
Yellow River (modern Gansu Province) did not form part of the Chinese
Empire, but were under the domination of the Yueh-chih. According to the late
S. Wada, the Yueh-chih realm included the north-western Mongolian plain and
the upper waters of the Yellow River.'
Szfi-ma Ch'ien locates the original place of the Yueh-chi11 between
Tun-huang and Ch'i-lien, which is a mountain in the Nan-shan range. He
believes that the Yueh-chih migrated from this restricted locality into the coun-
try of Ta-hsia o r Bactria (Shih-chi, Book 123). Further, as pointed out earlier,
the Yueh-chih need not be looked upon as a small community of people located
in a small area. They ruled the greater part of the Mongolian lai in, possibly
Dzungaria, the north of the T'ien Shan where the Wu-sun lived, countries in the
Tarim basin and the upper waters of the Yellow River. Their principal territory
may have been between Tun-huang and K'ang-chu, which were two of the most
important places on the so-called Silk Route, but we should not consider that
this region was the only territory held by them.
The position of the Yueh-chih as a red om in ant power in Central Asia,
before being pressed by the Hsiung-nu, is also evident from the statement of
Szii-ma Ch'ien. H e records that the authority of the Yueh-chih ruler was
accepted by the countries in Central Asia. They provided free food to envoys
moving through their territories, but messengers from China had to pay for their
food as well as their horses. This was one of the reasons why the Emperor Wu-ti
decided to send an expedition to the country of Ta-yuan (present-day Ferghana)"
in order to demonstrate China's military superiority over the Central Asian
countries. This could also mean that the countries between the Wu-sun and Par-
thia had been under the sway of the Yueh-chih until they were displaced by the
Hsiung-nu.
It may be suggested that the Yiieh-chih in the third century B.C. were
similar t o the T'u-chueh (Turks) of the sixth and seventh centuries A.I>. in terri-
tory and power; and that the migration of the Yueh-chih was not that of a
group of people from one place t o another, but a withdrawal from the eastern
and northern frontiers of the Yueh-chih Empire.

7. Wada, 1939, pp. 236-7, 1942, y . 278; see also his personal rcmar-lc to tlic .~uthor.
8. Pulleyblank, 1966, pp. 25 et seq.
 The Yiieh-chih and their migrations

The Yiieh-chih and Pazifik

We know that southern Siberia, Mongolia and Middle Asia were in the spIlere
of the so-called Scythian civilization, and it is in southern Siberia that most
archaeological sites o f the Scythian period have been found, chiefly in the
northern part of the Altai range. T h e tombs at Pazirik, excavated by Profcssor
5.1. ~ u d e n k o are
, very well known, and are dated between the fifth and third
centuries B.c.' T h e third century B.C. coincides with the date when the y u c h -
were at the apex of their power, before they were challenged b y the
Hsiung-nu. The Altai region in fact was part of the Yiieh-chih Empire, and the
sites at Pazirik should be related t o the Yueh-chih.
It is stated in the Shih-chi (Book 123), that, when Chang Ch'ien was cap-
tured by the Hsiung-nu o n his way to the Ta-yueh-chih in about 129 B.c:., he
was told by Chun-ch'en shun-yii, the King of the Hsiung-nu at the time, that
the Yueh-chih were t o the north of his domain and it was not possible for the
Han to contact them without crossing his territory, which required his pcrmis-
sion and approval. At that time, the authority of the s h ~ n - ~extended
ii to the
north of Tai and Yu-chung or, roughly speaking, the northern part of the prov-
ince of Shan-hsi. Chang Ch'ien and his party were, however, captured near
Lung-hsi (in Gansu Province), and it is not clear whether they were brought to
the court of the shun-yii o r t o the western part of the territory of the
Hsiung-nu, where the shun-yii happened to be staying. In any case the region of
Ili, where the Yiieh-chih are usually believed to have settled at that time, lies
west of the place in which the shun-yii is supposed t o have been. But if the
shan-yii meant the direction of the Altai mountain range, this may be more
likely than the Ili region. This could be supported by another reference from the
Han-shu (Book 52) in which is recorded the controversy between Han An-kuo
who opposed the expedition against the Hsiung-nu and Wang H u i who insisted
on the necessity of it. T h e Emperor Wu-ti accepted Wang Hui's opinion and
dispatched more than 300,060 troops to Ma-i in the present province of Shan-
hsi. The expedition ended in failure, and Wang Hui committed suicide in 133
B.C. According t o Wang Hui, if the expedition had succeeded, the H a n would
have been able t o subjugate the Yueh-chih in the north. Here, the Yiieh-chih are
placed to the north of the H a n a few years later than the departure of Chang
Ch'ien to the Yiieh-chill. These examples show that the region of the Altai range
is better suited than the Ili valley as the dwelling- lace of the Yueh-chih.
The ethnic identity of the Pazirik people is still t o be established. It is not
yet known if they were of the same stock as the S c ~ t h i a n sin south Russia. But,
in this connection, it may be pointed out that the so-called 'Detached Scythians'
are described by Herodotus as having lived in the vicinity of the Altai region.

9. Potapov, 1953, p. 16; Kisclc\,, 1951, pp. 361-92, especially 391, cf, A,IIA, 1949, Val. 9,
P. 216; Rudenko, 1953, pp. 346 et seq.; cf. Rudenko, 1951, p. 90; Rice, 1957, p. 200.
 A'. Enoki, G. A. Koshclcnko and Z . Haidar~j

Again according to Herodotus, these Detached Scythians were rebels against

Royal Scythians in south Russia, but they were of the same stock. on,
wonders if they were the Pazirik ~ e o p l eand the ancestors of the Yijeh-
chih.

The Yiieh-chih and the Detached Scythians

According to Herodotus, there was a great migration of peoples in the seventh
century B.c., which resulted in the occupation of south Russia by the Scythians
who had originally lived further east and were pushed westwards towards the
Issedones. Could not the Detached Scythians be considered as those Scythians
who did not move west with their main horde? It may be proposed that the
Detached Scythians built up a large empire which included a greater part of the
Mongolian plain, regions t o the north of the T'ien Shan range, the Tarim basin
and the upper waters of the Yellow River. They were known t o the Chinese as
the Yueh-chih (i.e. Scythians). A portion of them migrated into - or removed
their centre to - Middle Asia, while the rest were destroyed by the Hsiung-nu
or stayed in Nan-shan with the Ch'iang.
If the Yueh-chih were Scythians, then what was the relationship between
them and the four peoples w h o are said t o have come down from beyond the
Jaxartes (the present Syr Darya) t o destroy the Bactrian kingdom? If the expla-
nation given above is correct, the country of Ta-hsia, which was conquered by
the Yueh-chih, cannot have been the Bactrian kingdom, which had already been
destroyed before the arrival of the Yueh-chih. Therefore, the invasion of these
four peoples must have taken place some time before the coming of the Yiieh-
chih. In the third and second quarters of the second century B.C. the Greeks in
Bactria were fighting with the Greeks in India as well as with the Iranians of
Parthia. According t o W. W. Tarn, Bactria was up t o about 141 B.C. under the
control of Heliocles, who is believed t o be the last king of the Bactrian king-
dom.I0 So the invasion may have taken place in that year or some time later and
it must have been before the coming of the Yiieh-chih who occupied the Sog-
diana-Bactria region between 136 and 129 (or 128) B.C. Strabo tells us that the
Bactrian kingdom was destroyed by the Tocharians and three other ~eoples,
and, according to Sz6-ma Ch'ien, the country which the Yueh-chih conquered
was Ta-hsia. As 'Ta-hsia7 is believed to be a transcription of 'Tochara7, and
if these two statements are accepted, it cannot have been the ~ u e h - c h i hwho
conquered the Bactrian kingdom.

10. Tarn, 1938, py. 272-3; Narain, 1962, p. 14 1 .

178
 The Yiieh-chih and their migrativnr

The Yiieh-chih conquest of Bactria

In the history of Central Asia, the third and second centuries B.(;. constituted a
P eriod in which the nomadic peoples inhabiting its northern and north-eastcrn
borderlands made a great impact o n the course of political events. The mid-
century B.C. saw the emergence of the Graeco-Bactrian and Parthian king-
doms. It was the period when Diodotus in Bactria and Andragoras in Parthia
freed themselves f r o m Seleucid rule. In Bactria this gave rise to the Graeco-
Bactrian kingdom, but events took a different course in Parthia, where the
tribe of the Parni, under the leadership of Arsaces, overthrew Andrag-
oras and laid the foundations of the Parthian kingdom (see Chapter 5 ) . In his
eastern campaign, the Seleucid ruler Antiochus I11 had besieged Bactra, the
~raeco-Bactrian capital. T h e Graeco-Bactrian king, Euthydemus, offering
peace on honourable terms, put forward a number of arguments that included
the following (Polybius XI.34.5):

If Antiochus does not accede to my request, the situation of both parties will
become insecure. Huge hordes of nomads are massed on the border, posing a
threat to both of us, and should the barbarians cross the border they will undoubt-
edly conquer the land.

Euthydemus suitably impressed the Seleucid ruler, peace was concluded, and
the border between Graeco-Bactria and the nomadic tribes remained intact. The
incident is most significant because it clearly demonstrates the strong pressure
exerted by the nomads o n the political borders of the sedentary states in the late
third century B.C.
Euthydemus' warning was t o prove no hollow utterance, for events soon
occurred which led t o the downfall of the Graeco-Bactrian kingdom and altered
the entire political situation in Central Asia. These were triggered off by the
conflict of two nomadic tribes, the Hsiung-nu and the Yueh-chih. It is clear
from Chinese sources that the Hsiung-nu ruler Mao-tun (206-1 74 B.c.) attacked
the Yueh-chih tribe, whose leader was killed in battle. The Yiieh-chih were
forced to retreat westwards. The Hsiung-nu ~ u r s u e dthem and the son of the
Yiieh-chih leader was also killed in an encounter with them." The movement of
the Yueh-chih set off a whole series of displacements of nomadic ~ e o p l e sin
Central Asia. O n e such 'secondary' displacement caused by the movement of
the Yueh-chih is mentioned in Chinese sources. They moved into the Ili basin,
from where they drove off the Sai (Archaic Chinese Sak) people; but the Yiieh-
chih were in turn attacked by the Wu-sun and compelled to resume their west-
ward migration.'? T h e diversity of ~ e o p l e sinvolved in these migrations is also

11. Bichurin, 1950, Vol. 11, pp. 147, 151.

12. Narain, 1962. 1.3~.133-5.
 k: Enoki, G. A. Kosh~>lerrkoand 2.Haiddry

confirmed by ancient writers. The movement of the Yueh-chih filially to a

halt in Bactria. Although the political events and migrations that led up to
nomadic conquest of Bactria remained unknown t o Greek and Roman writers,
the fact itself did not escape their attention. It is very significant that the classical
writers in question mention the participation of several peoples in the conquest
of Bactria. Describing the situation in the nomadic zone of Central Asia, Strabo
enumerates the tribes that 'took Bactria from the Greeks': the Asioi, the Pasianoi,
the Tocharoi and the Sacarauloi (Strabo XI.8.2). A little later he mentions the
Sakai in connection with the conquest of Bactria (XI.8.4). In connection with this
event, Pompeius Trogus (Trog. Comp., Prolog. XLI) speaks of the Scythian
tribes of the Sacaraucae and the Asiani and subsequently (XLII) of the Asiani -
'kings of the Tochari' - and the Sacaraucae. Scholars have taken great pains to
correlate Chinese and Greek ethnic names and determine the role of the various
peoples in these movements. What remains apparently in dispute is identification
of the Chinese Sai with the Greek Sakai and the Indian Sakas. More problematic,
though highly plausible, is the identification of the Yueh-chih with the Tochari.
During the eighties of the second century B.C. the migration of the Yiieh-
chih eventually displaced the Saka tribes of the eastern part of Central Asia.
They migrated south and crossed the Pamirs, having touched o n eastern Bactria.
This stage in the migration of the Sakas was apparently completed in the early
first century B.C.
The Yueh-chih proper, having drawn into their orbit a number of other
nomadic peoples, passed through Ferghana (Ta-yuan) and reached the borders
of Bactria. The completion of the first stage of their migration is recorded in the
Chinese sources.'<hang Ch'ien, w h o had been sent by the Chinese to per-
suade the Yueh-chih t o join China in attacking the Hsiung-nu, stayed among
them for some time during the ~ e r i o d130-125 B.C. The situation he recorded"
was that the Great Yueh-chih lived a nomadic life, ranging over the area imme-
diately north of the Oxus river, while the country south of the Oxus (southern
Bactria) was subordinated to the Great Yueh-chih but retained its political
autonomy, divided into a large number of city-states each with its own ruler. It
can thus be supposed that at the time northern Bactria and Sogdiana were fully
under the control of the Yueh-chih while in southern Bactria towns persisted
under the supreme suzerainty of the nomads.

The nomads and Parthia

Almost simultaneously the nomads began migrating westwards from Bactria
and came into conflict with the Parthians. This clash sprang from a particular

13. Davidovich, 1976, pp. 56 et seq.

14. Bichurin, 1950, Vol. 11, p. 151; Pulleyblank, 1968; Ziirchcr, 1968.
 The Yiieh-chih and their migrations

The parthian king Phraates hired some nomads (Justin calls them Scyth-
to do battle with the Seleucid king, Antiochus VII Sidetes. They
lare l hen the Parthians had already routed the Seleucid forces and King
phraates refused t o pay them the agreed sum. In retaliation, the nomads devas-
tated the eastern half of Parthia. Phraates marched against them, with an army
[hat included Greek soldiers w h o had been taken prisoner by the Parthians. I n
the heat of the battle against the nomads in 128 B.c., the Greeks went over t o
the enemy, the Parthian army was annihilated and Phraates was killed (Justin
~ ~ 1 1 . 1 . 5The
) . nomad detachments then left Parthian territory. The new king,
~ r t a b a n u s(Phraates' uncle), tried t o mount an offensive against them. It came
to nothing, and in a battle with the Tochari (here Justin gives a precise ethnic
name instead of his previous vague 'Scythians'), Artabanus was mortally
wounded (Justin XLII.2.2), dying in 124 o r 123 B . c . ' ~ His son Mithradates 11
succeeded t o the Parthian throne and put a final stop t o the nomad incursions.
He waged many valiant wars against his neighbours and joined many peoples t o
the Parthian kingdom. H e also several times vanquished the Scythians and
avenged the wrongs done t o his forebears (Justin, XLII.2.4-5).
By the turn of the first century B.C. the Saka tribes had settled in Gand-
hira and the Great Yiieh-chih in northern Bactria. South Bactria, though still
under the authority of minor Greek rulers, was also under the control of the
Saka tribes; and o n the eastern borders of Parthia (Margiana, Aria and Dran-
giana) military operations were under way against them. Subsequent political
developments involved a Parthian offensive t o the east. This is the historical
context that must be borne in mind for a proper understanding of Strabo's
report that the Parthians also 'seized part of Bactria, having driven away the
Scythians' (Strabo XI.9.2). T h e course of those wars is attested by three unique
series of Parthian coins with the Greek inscription 'KATA ETPATEIA' ( O n
campaign) and the name of three regions: Aria, Margiana and Traxiana. The
dating of these coins has been disputed for many years. The most ~ l a u s i b l e
view, however, seems t o be that of David Sellwood, w h o attributes them to
the seventies of the first century B.C. and t o the mint of an unknown Parthian
ruler.Ih It is reasonable t o suppose that as a result of these campaigns, the Par-
thian~gained control of Aria (Herat oasis), Margiana (Merv oasis) and pos-
sibly the area of Mashhad. Sellwood thinks that this was the region called
Traxiana. The further conquests of the Parthians in the east are attested in Isi-
dore of Charax's Mansiones Pavthicae. After 'Areia' and the 'Anau6n chora'
(part of Aria), the author names the Parthian possessions Zarangiane, 'Sakastane
Sakon Skyth6n' (Sakastan of the Saka Scythians) and Arachosia (Kandahar

15. Debevoise, 1938, p. 38.

16. Sellwood, 1980, pp. 97 ct seq.
 K. Enoki, G. A . Koshelcnko and 2. Haidary

region)." In modern works it is thought that as a result of these military

ations the Parthians halted the nomads, brought them under control and settled
them on the land. After these events Drangiana was called Segistan, modern Sis-
tan. The Sakas seem t o have retained some form of their political organization
under Parthian sovereignty while beyond the confines of Parthia there were
some minor nomad possessions dependent o n the Parthians; there is, for
instance, numismatic evidence for a ruler Sapadbises.Ix Such was the situation of
the advance of the nomads in the western zone at the turn of the Christian era,
Further east, in the central regions of Bactria, the situation is described in
the Chinese chronicle, the Han-shu. It was already markedly different from that
recorded by Chang Ch'ien.19 The whole of southern Bactria had already been
occupied by the Yueh-chih, so that now the Great Yiieh-chih bordered on the
south with the land of Chi- in. The supreme leader of the Great Yueh-chih had
his residence in a town north of the Oxus river. The entire territory dependent
on the Great Yueh-chih was divided into five hsi-hou. The state of the Great
Yueh-chih represented a confederacy of five tribes. These were the five former
territories for nomadism which had been transformed into five minor vassal
states dependent on the central authority ~ersonifiedby the King of the Great
Yueh-chih. At the same time, it must be remembered that the state that had
been formed was still very unstable and loose-knit.

Archaeological remains of the nomads

in northern Bactria
In northern Bactria the burial places investigated by Mandel'shtam at Tulkhar,
Aruktau and Kokkum20 are thought to be those of nomads who conquered Bac-
tria. Material from the graves indicates that tribes of various origins took part in
the conquest. Very significantly, the material culture suggests northern links, in
particular with the Sarmatian tribes. This is borne out by evidence from
nomadic graves in the Zerafshan valley in Sogdian territory."
A very important point, deduced mainly from nomadic graves in Bactria,
is that the graves are situated on the edge of the valleys and d o not encroach on
the irrigated and tilled areas. The very plausible suggestion has been made that
the nomadic conquest left no havoc in its wake because the nomads did not des-
troy the irrigation systems and did not take u p irrigated land for grazing. We

17. Sarianidi and Koshclcnko, 1982, pp. 307 et seq.; Masson atid Romodin, 1964, pp. 135 et
seq.
18. Sarianidi and Koslielcnko, 1982, pp. 310 ct scq.
19. Davidovich, 1976, pp. 60 et seq.; Ziirclier, 1968, pp. 367 et seq.
20. Mandel'shtam, 1966, 1975.
2 1. Obel'chenko. 1974.
 The Yiiclh-chih and their migrations

,-an safely surmise that having moved into the conquered territories, they took
over land not used by farmers, and preferred to exploit the subjugated popula-
tion, leaving the existing economic structure intact.
A close link is t o be noted between the nomads and the settled oases. *rhc
vast majority of ceramic vessels found in the nomad graves were made i n
and came t o the nomads from the oases. We d o not of course know
how this happened; perhaps the nomads received the manufactured items
(including pottery) as tribute exacted from the farmers or acquired then,
barter. T h e fact remains that very close ties existed between the nomads
and the farmers.
The sensational discoveries of the Russian-Afghan expedition (led by V. I.
sarianidi) at the necropolis of the ancient town of Tillya-tepe in northern
Afghanistanz2have shed light o n the upper social stratum in the period when
state-like formations, headed by the descendants of leaders of the nomadic
tribes, emerged in the conquered territories. The coins from Tillya-tepe enable
it to be fairly accurately dated between the second half of the first century B.C.
and the first half of the first century A.D." Six excavated graves yielded some
20,000objects made of precious metal (Fig. 1). The huge riches that accompa-
nied the dead contrast with the more than modest sepulchral structures. It is
particularly significant that there is n o mark over the graves indicating a burial
site. Sarianidi assumes that these were secret burials.
Archaeological evidence from the Pamirs and its comparison with other
material confirm the evidence of written sources that in migrating to India the
Sakas did pass through this mountain region. Another important point is the
sharp decline in the population in the Pamirs in the subsequent period, doubt-
less due to the fact that the bulk of the population moved ~ o u t h . 'Excavations
~
at Ay Khanum have shown that the fall of the Graeco-Bactrian city was indeed
the result of the nomadic conquest; the population that settled o n its ruins was
very different culturally f r o m its ~redecessors.There is unfortunately very little
archaeological material from the more southerly parts of the region t o throw
light on nomadic migration.
Indian historical tradition is of no help, since it merely brushes over these
events.25There is also insufficient material t o throw light o n the history of the
cities and inhabited rural localities in the region during this 'dark age'. N o
doubt a number of cities such as Ay Khanum perished in the course of the
nomadic conquest. But it is equally clear that the nomads did not make a special
effort to wreck the irrigation svstems and the towns and villages. So far as we

22. Sarianidi, 1984.

23. Sarianidi and Koshelenko, 1982.
24. Litvinsky, 1972.
25. Sharma, 1980; Serditikh, 1983, p p 84-8.
 K. Enoki, G. A. Koshelerzko 'ind Z. H'iiddry

c..ln iudge, the nomads sought t o exploit rather than destroy the existing ecoll-
omic structure.
Material from northern Bactria indicates that the nomadic conquest was
soon followed by fresh material prosperity and, in particular, urban develop-
ment. O n arcl~aeologicalevidence f r o m the Surkhan Darya valley, V. M.Mas-
son collcluded that the rapid rise in the number of settlements there began in
first century B.c.?' Yiieh-chi11 times saw the rise of what was later to becolne
the lllajor centre of Dalverzin-tepe,I7 as well as many smaller centres such as
Zar-tepe. Furthermore, significant reconstruction operations at the anciellt
town site of KhalchayanZx can be ascribed t o the Yiieh-chih period; the
renowned Sistan sanctuary of Kuh-i Khwaja also seem t o have been rebuilt.!q
T h e most revealing material regarding the history of the Central Asian city in
those 'dark ages' has been obtained, however, in the course of archaeological
exploration at Taxila.'O

E'ic;. I . Gold necklncc. 'I'illY,l-tcpc. (Coul-tcs). of V. S'lrinnidi.)

26. Masson, 1974; Masson, 1976.

27. Dnl'verzintepa, 1978.
28. Pugachenkova, 1966.
29. Gullin;, 1964.
30. Marshall, 1951; Il'in, 1958.
 The Yiieh-chih and their m~grations

The 'dark ages'

The scarcity of sources makes it hard to trace political and social dcvclopmcnt
in the ~ o u n t r i e sof Central Asia during the so-called dark ages. It is clear froln
what has been said earlier that the states that were formed in the of
nomadic conquest were insufficiently stable, their frontiers were constantly
changing, and some states were emerging as others declined. Thcrc was no
structure within the states. O n e reason for this was thc character of
statehood that emerged as a result of the nomadic conquests. The bcst-known
instance was the state of the Yueh-chih, which was divided into five hsi-hou
babghu)It can be assumed that this division into five possessions represented a
division of the territories f o r the nomadism of the five tribes that took part in
the conquest of Bactria. It may be supposed that the situation was similar in
other regions conquered by the nomads.
Another factor that weakened these new emergent states was the frequent
co-existence of two political structures - an old one retained from earlier times
and a new superimposed nomad structure. O n e example of this was the situ-
ation in Bactria during the first stage of the Yueh-chih conquest, when the terri-
tories north of the O x u s were under the direct rule of the Great Yueh-chih,
while southern Bactria was divided between a number of minor rulers of indivi-
dual cities.
An important role in the political structure of the new states was played
by satraps, an institution inherited from the Achaemenids. However, at this
time many satraps, while nominally plenipotentiaries of the central government,
were in fact almost o r fully independent rulers w h o founded dynasties of their
own, and the title 'satrap' o r 'great satrap' was often the mark of an independent
ruler. The nomadic conquest clearly had an effect o n the developnlent of social
relations in the countries of the region.
A number of Indian works describing the calamities which befell the
country in the 'Kali Age' include among these disasters the disruptions of the
established social order, the flight and liberation of slaves, the increased wealth
of the Sudras and the general weakening of caste (varva) stability. It is perhaps
significant that the texts concerning these changes are usually set in the context
of foreign conquest - Greek, Saka, Parthian and Kushan."
Despite political instability, economic relations in the Saka-Parthian
period developed very successfully. Busy sea-routes were opened u p between
Egypt and India, the first direct contacts between Egypt and India dating back
to the late second century B.C. (apparently between 120 and 115). In about 100
B.c., Hippalus discovered the lnechanism of the monsoon winds, and thereby
made it possible f o r ships t o sail regularly t o and from the shores of India (Prri-
PIUS 57). Trade between Egvpt and India was most actively developed in the late
31. Yadava, 1978/79.
 K. Enoki, G. A. Koshelcnko and 2. Haidary

first century r3.c. When in the twenties of the first century B.C. Strabo jour-
neyed in Egypt, hc 'learnt that some 120 ships complete the voyage from the
Straits of Hormuz t o India' (Strabo 11.5.12). H e also observed in that connec-
tion that 'great fleets now set off as far as India' (Strabo XVII.1.13) and, further,
that merchants sailing from Egypt via the Nile and the Arabian
Gulf to India. . . have been going as far as the Ganges' (Strabo XV.1.4). In the
first century A.D. knowledge of the coasts of India and the ability to use the
monsoon winds had attained such a standard that many new Indian peoples and
cities were discovered, trading links were established with them, and ships from
the ports of southern Arabia and the H o r n of Africa went to particular ports of
India with particular goods (Pliny, Nut. Hist. VI.101-6)." Large vessels carried a
detachment of armed guards (Pliny, Nat. Hist. V1.lO1; Philostratus, Life of
Apollonius of Tyana 111.31.1 c.). This was clearly why the rulers of Indian states
permitted merchant ships t o moor only in particular places (Periplus 52) and
introduced a number of other limitations (Philostratus, Life of Apollonius of
Tyana 111.35.1 c.). The Indian trade ~ i e l d e dhuge ~ r o f i t sfor the merchants from
Egypt since the goods were resold a hundred times dearer (Pliny, Nat. Hist.
VI.101). The overall trade balance was negative for the Romans, who ruled
Egypt from 30 B.C. Pliny wrote: 'India annually swallows up from our state no
less than 55 million sesterces' (Pliny, Nut. Hist. VI.lO1). In the second half of
the first century A.D. Indians were fairly frequent visitors to Egypt. On one
occasion in the theatre of Alexandria in Egypt some time in the period A.D.
71-75, Dio Chrysostomus said (Ad Alexandr. 40):

I see before me here not only Hellenes, Italics and even people from Hither Syria,
Libya, Cilicia and distant Ethiopians and Arabs, but also Bactrians, Scythians, Per-
sians and a number of Indians, all of whom customarily come together and form
the audience here with you.

In the territory of India the Roman trading station at Arikamedu3j was well
known; and in Egyptian territory, in the 'White Haven', Indian ceramics and
fragments of pottery with Tamil inscriptions in Briihmi script have been
found."
Overland caravan trade was actively developed during this period. The
'royal way' passed through Parthian territory, starting from Zeug~naon the
Euphrates. It is described by Isidore of Charax. Cutting across Mesopotamia
and passing through the territory of Iran it reached Merv. From there it turned
south and led on through Aria, Drangiana and Segistan to Arachosia (in the
Kandahar region), then through Ghazni t o the Indian subcontinent. Another

32. Bcrzina, 1982, pp. 31 ct seq.

33. Wheeler et al., 1946.
34. Whitkornb and Johnson, 1980.
 The Yiieh-chih and their migrations

important branch of the highway ran from Merv to Amul on the Oxus and then
samarkand, where it merged with the Silk Route from the oases of Tarim
basin, The Chinese sought t o establish permanent trade links with Parthia as
early as the second century B.C. T h e official Chinese annals contain an account
by Chang Ch'ien of his visit t o Parthia,35from where he brought vines and
alfalfaseeds t o China. From the late second century B.C. the Silk Route func-
tioned more o r less regularly, passing through Central Asia and playing an
important part in its economic ties. The Parthians protected trade along the
route, deriving considerable profit from the payment of taxes, and did every-
thing they could t o prevent direct links between China and Rome since their
intermediary role was extremely pr0fitab1e.j~The Parthians also carried on
maritime trading with India via the Persian Gulf ports, notably Spasinu-Charax.
There were north-south caravan routes from Bactria to India; and many routes
via the Hindu Kush linked the northern and southern parts of Central Asia.
Central Asia was thus provided with a network of maritime and overland
routes, both international and local, which ensured the development of inter-
national and local trade. Control over the trade routes procured financial
resources for the state treasury, and could be used as a political weapon. Thus,
the Parthian government tried t o ensure that caravans from Palmyra (a major
centre of the caravan trade) went not t o Seleucia o n the Tigris, a city hardly not-
able for loyalty t o the Arsacids, but t o the cities more closely connected with
the central authority, such as Ctesiphon, Vologaesocerta and Spasinu-Charax.3'
The expansion of international and domestic trade was obtained by means
of a developed system of monetary circulation. Coins of many dynasties circu-
lated in Central Asia and, with growing international trade and political
upheavals, were often t o be found far from the place of minting. Much of the
money in circulation was accounted for by coins issued by the early Graeco-
Bactrian kings. Minted from high-standard silver in accordance with the Attic
system of weights and issued in large quantities, they remained in circulation for
a number of centuries. Another major part of the total amount of money in cir-
culation was constituted by the so-called Indo-Greek coins issued by the
Graeco-Bactrian kings w h o had conquered part of India. Beginning with Deme-
trius, a number of these kings had possessions both north and south of the
Hindu Kush, and some ruled solely in the territory of the Indian subcontinent.
These coins usually followed the O l d Indian system of weights. O n e side of the
coin carried a Greek legend and usually a typically Greek portrait; the other
side bore a reverse type with a Kharo!ihi legend. Also in circulation were coins
of the last Graeco-Bactrian and Indo-Greek kings, issued when the bulk of the
region was alreadv under nomad control.

35. D'yakonov, 1961, p. 204.

36. Colledge, 1967, p. 80.
37. Koshelenko, 1971.
 o 2. Haiddry
K. Enoki, G. A. K o s h ~ ~ l r n kand

Very soon after reaching the settled oases of the region, the rulers of the
nascent nomad states began issuing their own coinage. This was often dictated
not by economic necessity but by the desire to proclaim to the world a new
independent state. The first local coins in the region north of the Oxus were
various types of imitations of Seleucid and Graeco-Bactrian ~oins.~K We know of
coins modelled on the Seleucid drachms portraying Alexander (place of rninring
unknown); imitations of Antiochus I drachms (in the Zerafshan valley); imita-
tions of tetradrachms of the Graeco-Bactrian king Euthydemus (western part of
the Zerafshan valley); imitations of the tetradrachms of the Graeco-Bactrian
king Eucratides (in Chorasmia or the middle valley of the Jaxartes); imitations
of the obol of Eucratides (in the Kafirnigan basin); silver and subsequently
bronze imitations of the tetradrachms and drachms of Heliocles. These stages in
the development of coinage in that part of the region correspond to the second
and first centuries B.C.
Subsequently coins of another type began t o be minted on behalf of the
new authorities. O n e of the most striking examples of these new coins are those
of Heraus, who was evidently one of the predecessors of the Great Kushan
kings. In Sogdiana, coins began to be minted in the first century A.D. depicting,
on the reverse, a standing archer and, on the obverse, the profile of a king with
the name 'Aitam'.
The period of the 'dark ages' was one of the most fruitful in the develop-
ment of Central Asian art, when the previously rather isolated artistic schools
came into close mutual contact. As the Greeks from Bactria campaigned and
settled in India, they became familiar with Indian art but they also brought with
them forms of their known and customary art. In Bactria there was evidently
little interaction between Greek and local art before the nomadic conquest.
Greek art was that of the conquerors and was alien to the bulk of the ~ o ~ u l a -
tion, who maintained their own traditions. Neither, on the other hand, did local
art have any significant effect on the art of the Greeks. The nomadic invasion
abruptly changed the situation. The social barriers dividing the world of the
Greeks from that of the Bactrians were swept away, there being then no longer
anything t o prevent contact between the two artistic traditions. The noinads
also brought with them their art, which spread in their wake in Bactria, Sog-
diana and the Indian subcontinent. As the power of thc Indo-Parthian rulers
extended, so the Parthian art forms that had already taken shape at this time
began t o affect the artistic life of many parts of Central Asia. In a word, the
'dark ages' constituted a time of intense interaction between many trends in art.
This is most clearly seen in the works of decorative and applied art, found
in the excavations of the Tillya-tepe necropolis." Among mass items brought to

38. Zeirnal, 1978.

39. Sarianidi, 1984.
 The Yiich-chih and their mjgrations

light, a number of distinct trends are clearly evident. The first thing that strikes
the eye is the large quantity of works of clearly nomadic origin (scenes of ani-
mals seizing their prey are vigorous and expressive). There is undoubtedly a
close similarity between these works and those from the burial mounds of
[he Altai mountains. T h e genetic link between this category of work and the
'Siberian animal style' is beyond doubt. Another trend is represented by typi-
cally Greek objects connected with Hellenistic art traditions. Chinese items o r
items displaying Chinese influence occur in small quantities. Some works of art
clearly reflect ancient Bactrian traditions in which Parthian influence makes
itself felt. Finally, there are specimens representing a synthesis of various tradi-
tions.
In the case of Bactria, the best-known relic of monumental art of the
pied of the 'dark ages' is t o be found at Khalchayan.'O Pugachenkova regards
the building in Khalchayan as a palace. A more plausible view, however, is that
the palace in Khalchayan represents a temple of deified ancestors. The opinion
is sometimes expressed that the Khalchayan palace must date from Kushan
times, but this can hardly be so. T h e palace clearly dates from the 'dark ages'
though the exact point has not yet been determined.+' What arouses the greatest
interest in Khalchayan is the relief compositions, including depictions of repre-
sentatives of the local dynasty and a carved frieze of Dionysian character. The
carving is in clay and has been thought t o represent the dynasty of Heraus, but
this is hard t o accept when the possessions of Heraus very probably lay south
of the Oxus.

40. Pugachenltova, 1966, 1971.

41. Koshelenko, 1974.
 T H E S A K A SA N D I N D O - P A R T H I A ~ ~ : : -
B. N.Puri

H E term 'Saka' o r 'Scythian' is given generally to the nomads who occu-

pied the northern regions of Asia and Europe. The earliest references t o
them are in the Histovide of Herodotus (VII.64) and in the cuneiform
inscriptions of Darius I, in a rather vague manner (see Chapter 2) denoting thrce
different and widely separate tribes. T h e history of the Sakas' is closely linked
with tribal movements from the neighbourhood of China (see Chapter 7). They
were forced by the Yueh-chihz t o move south-west and occupied Bactria. Sub-
sequently the Yueh-chih put an end t o Saka rule in Bactria, causing their ruler
to flee to Chi-pin (Kapiia). T h e Indo-Greeks in Kabul impeded further Saka
progress and compelled them t o move westwards in the direction of Herat and
thence to Sistan. This country was finally named Sakastan after them.? Parthia
now acted as a barrier against any tribal movement from Upper Asia, and the
stream of invasion was thus diverted into another channel, forcing the Sakas t o
move into the country of the lower Indus (Indo-Scythia).

'" See Map 3.

1. For the history of the Sakas and their relations with contemporary tribes, see: Minns,
1913, 1908-25 (1961), 1925, p. 187; McGovern, 1939; Tarn, 1951; Rapson, 1922; Smith,
1907; Thomas, 1906; Lohuizen-de Leeuw, 1949; Puri, 1965; Majumdar, 1951; Konow,
1929; Nilakantha Sastri, 1957.
2. On the Yiieh-chih, a comprehensive bibliography of the Chinese sources is given by
Vallke Poussin, 1930, pp. 328 et seq. See also Konow, 1929, pp. liii et seq.; Puri, 1965, p. 9,
n. 16. The principal Chinese texts supplying relevant information about these tribal
movements are, according to chronological sequence, the Shih-chi, 123, of SzG-ma Ch'ien
(c. 90 B.c.), containing the report of the Chinese ambassador Chang Ch'ien's visit (c. 126
B.c.) to the West; thc Hdn-shu of Pan Ku (c. A . n . 92), containing the annals of the Former
Han dynasty covering the from 206 n . c . to A.D. 24; and lastly, the Hou Hun-sh~4
of Fan Yeh, recording the annals of the Later Han dynasty. spreading over the period
A . D . 25 to 220. See Pullevblank, 1968; Narain, 1957, pp. 130 et seq.
3. Konow, 1929, pp. xvi et seq.; Narain, 1957, pp. 1 3 2 4 ; Puri, 1963.
The Sakas in India
The Saka invaders of the Iildian subcontinent did not come through the Kabul
v a l l e ~ .Neither
~ is thcre evidence that Begram was ever ruled by the
dynasty.5 It has often been suggested that the Sakas could not have entered India
from the north across the high mountain ranges of the Himalayas, Karakorum
and the Pamirs, and that they must have reached India by the Bolan pass from
Drangiana (modern Sistan) and Arachosia (Kandahar) over the Brahui moun-
tains into the country of the lower Indus (Sind).6 However, the major discov-
eries by A. H. Dani, following the construction of the Karakorum highway
across the mountains from Pakistan t o Chinese Turkestan, have brought to light
important new evidence that requires a fundamental reappraisal of earlier theo-
ries. Dani has established the major routes used by the traders, soldiers and pil-
grims who crossed the high Karakorum mountains in the early historical period
and left large collections of Saka petroglyphs at the principal river crossings at
Shatial, Chilas, Gilgit and Hunza. The petroglyphs at Chilas include drawings
of Saka soldiers and horsemen as well as representations of stupas and the ibex.'
The sacred rock of Hunza again has numerous mounted horsemen and ibex of
the same period with a series of Kharosthi inscriptions which include the names
of Saka and Pahlava rulers."~ now seems clear that when Maues suddenly cap-
tured Taxila from the Indo-Greek king Apollodotus 11, he must have used the
north Karakorum route,' even if he were subsequently joined by other Sakas
who entered the Indus valley from Sistan. Maues belongs to the early group of
Saka rulers in the Indus valley w h o preceded the main dynasty of Azes I, whose
era dates from 57 B.C. A date for Maues and the Saka invasion in the ~eriod
85-70 B.C. coincides with the indications from Parthian history that any move-
ment of the Sakas from Sistan t o join the northern invaders should be some time
after the death of Mithradates I1 in 88 B . C . ~ ~
In Sanskrit literature the Sakas belong to the north-west. The Mahabh-
drata (11.32.17) locates them with the Yavanas (Greeks) and Pahlavas (Indo-
Parthians) in the far north-west beyond Sagala (modern Sialkot). The Kalakacd-
ryakathdnaka, a Jaina work of unknown date," ~ r o v i d e sfurther information
about the Saka dynasty in the Indus valley. The Jaina teacher Kilaka, whose sis-
ter had been abducted by Gardabhilla, King of Ujjain (ancient Ujjayini), sought
the help of the Saka King of Kings, who crossed the Indus by boat and

4. Rapson, 1922, pp. 563 et seq.; Konow, 1929, p. xxxi.

5. MacDowall, 1985, pp. 555-66.
6 . Konow, 1929, p. xxxi.
7. Llani, 1983, pp. 91-128.
8. Dani, 1985, pp. 5-124.
9. Bivar, 1984, pp. 5-15.
10. Debevoise, 1938; Sykes, 1922; Camel-on, 1937; Ghirshninn, 1961.
11. Jacobi, 1880, yp. 247 ct scq.
 The Sakas and Indo-Parthians

proceeded to Kathiawar. In the autumn the Sakas attacked, took Gardabhilll

risoner and established a Saka as over-king ( ~ a y a h i r i ~ aAfter
) . some time, Vik-
P
ramadityaythe King of Malwa, ousted this Saka dynasty and established his own
which came t o be known as the Vikrama era of 58 H.C. Vikramiditva9s
dynasty was in turn uprooted by another Saka king, who founded an era o f his
own when 136 years of the Vikrama era had elapsed - the Saka era of *.I,. 78.
The classical authors speak of the Sakas in north-western India as Indo-
Scythians. Ptolemy (VII.1.55) states that all the country along the course of the
Indus was called by the general name of Indo-Scythia. It included Patalene,
Abiria and Surashtra. Dionysius Periegeta (V.1088) speaks of the southern
Scythians as settled on the Indus and his commentator, Eustathius, says that
these were the Indo-Scythians.
The Indus valley, particularly the Panjab, Swat and the foothills of Kash-
is the area where the series of Kharosthi inscriptions of the SakasI2 and
of Maues and the kings of the Azes dynasty have been found.]' Large
numbers of copper coins of Maues and the Azes dynasty were found in the
excavations at Taxila."

MAUES

The earliest Saka ruler in the Indus region was Maues (Moga), who belongs t o
the first quarter of the first century B.C. The Taxila copper plate of Patika,
which records the action of Patika, son of Liaka Kusuluka the satrap of
Chukhsa, who established a Buddhist relic and a samgh~rima,ljis dated in the
reign of the Great King, the Great Moga, in Year 78 of a Graeco-Bactrian era,
probably of Eucratides (see Chapter 17). A defaced inscription from Maira in
the Salt range, about 160 km south of Taxila and similar in palaeography, seems
to be dated Year 58 and to contain the word 'Moasa'. Other inscriptions in the
same era are known from Fatehang, 16 km south of Taxila, dated Year 68 and
Shahdaur in the Agror valley perhaps dated Year 60. This leads Konow to sug-
gest that the Indo-Scythian Empire in the Indus was established around 88 B.c.,
the period of the death of Mithradates I1 of Parthia. The coins of Maues are
sometimes overstruck on, and in turn overstruck by, coins of the late Indo-
Greek Apollodotus 11, who seems to have been the immediate ~redecessorof
Maues at T a ~ i l a . 1Maues
~ imitates Indo-Greek coin types such as those of Apol-
lodotus 11, his immediate predecessor, and Demetrius, who was roba ably the
first Graeco-Bactrian king t o rule the 1ocality.l' O n his coins that copy the

12. Konow, 1929, pp. 11 et seq.

13. Jenkins, 1955.
14. Marshall, 1951.
15. Konow, 1929, pp. 23-9.
16. Jenkins, 1955, pp. 1 et seq.
17. MacDowall, 1989, pp. 29-30.
Elephant Head and Caduceus types of Demetrius, Maues is styled simply a
'Basileos Mauou' in Greek alone, 'of King Maues'; subsequent issues follow [he
Indo-Grcek practice, adding the equivalent titles 'maharajasa Moasa' in
. . on the reverse; and two of his silver issues adopt a new and more ele-
Kharosthi
vatcd forrn of titulature, the Greek 'Basileos Basileon Megalou Mauou', 'of [he
Great King of Kings Maues' with its corresponding Prakrit version in
Kharogihi, 'rijatirijasa mahitasa Moasa'. The higher title of 'Great King of
Kings' is used regularly by his successors Azes I, Azilises and Azes 11. From
this numismatic data, it has been deduced that Maues was the earliest among [he
known Saka rulers of the Panjab, and that his coin types were continued by
Azes 11 and Azilises, w h o also borrowed some other Indo-Greek types not
copied by Maues.'"

SUCCESSORS OF MAUES

The numismatic data'? suggest an outline of the history of Maues' successors. ~t

appears that Vonones, probably a younger contemporary of Maues, was ruling
in the eastern border country of Iran and Arachosia in association with his
brother Spalahora and his nephew Spalagadamas Vonones' successors included
Spalyrises (the Greek form of 'Spalahora') with his son Spalagadama. Then
began the reign of Spaliriia, first as the 'brother of the king', then as the sole
ruler, and lastly jointly with someone called Azes. During this period the Saka-
Pahlava kingdom may have extended from Arachosia t o the Paropamisadae
(Kabul). The relationship between Maues and Azes I cannot be satisfactorily
explained. It has been proposed by some scholarsZ0that the two belonged to dif-
ferent racial stock. According t o Konow,*' Maues was a Saka and Azes a Pah-
lava who succeeded the Saka king Moga (Maues) in Taxila. Tarn,zz however,
thinks that Azes I was Spaliriia's son and both were of Saka race. Rapson2'
describes Maues, Azes I and Azilises as the first three Saka kings belonging to
the same class and having close numismatic affinity.

T H E D Y N A S T Y OF: A Z E S

The major Indo-Scythian dynasty ruled an empire based on the Panjab and
Indus valley from c. 50 B.C. t o A.D. 30. Numismatists have distinguished the
existence of two kings called A ~ e sThe
. ~ ~silver coins of Azes I have the obverse

18. Nilakantha Sastri, 1957, p. 198.

19. Whitehead, 1914, pp. 91 ct seq.
20. Nilakantha Sastri, 1957, p. 207.
21. Konow, 1933, p. 24.
22. Tarn, 1951, p. 346, n. 3.
23. Rapson, 1922.
24. Jenkins, 1955, pp. 1 et seq.
 The Sakas and Indo-Parthians

type of the king o n horseback holding a couched spear while Azes II has the
horseman holding an upright whip. Copper coins of Azes I are overstruck by
Azilises, showing that Azes 1 preceded Azilises. Belonging to the Azes dynasty
is a group of Kharogthi inscriptions dated in the Old Saka era. ~ ~ h ~ i ~
~~~~w~~ relates the era t o the Yiieh-chih conquest of Bactria in 129 I ~ . C .but
,
most scholars refer them t o the Vikrama era of 58 13.c.The Taxila silver vase,
referring to a 'Great Kushan King'z6like the Shahdaur inscription," has a num-
ber date with 'Ayasa' that MarshallZ8interpreted as 'in the era of Azes9. The
Kharop~hiinscriptions published by Fussman," which have dates 'Ayara ali-
dara7(of the deceased Azes) and 'Ayasa purvakalisa' (of Azes of time past), now
confirm Marshall's argument of an era founded by Azes I. It is, however, puz-
zling that Azes I should use an era established t o mark the expulsion of the
Sakas from Malwa; therefore it more probably refers t o the accession date of
Azes I about the same time, when the Saka Empire was consolidated and
reformed o n the Indus after Vikramaditya stemmed their further advance.
The coinage of Azes I was struck at three principal mints: at Puskalivati
in Gandh~ra,at Taxila and in the middle Indus province, but not in Arachosia
where finds rarely contain his coinage.'O H e retained the silver denominations
and square coppers that the Indo-Greeks had used in the provinces, but used as
his obverse type the Saka king o n horseback, rather than a traditional royal por-
trait.
Azilises succeeded Azes I as King of Kings in the Indus provinces. Tarn''
argues that Azes I associated Azilises with himself as co-ruler because of some
joint coins struck in both names, but the supposed joint coinage is rare and may
simply be the continued use of old dies after Azilises' accession. Azilises struck
coins in the three mints used b y Azes I. His silver tetradrachms have been
found in hoards with those of the Indo-Greek king Hippostratus in north
Pakistan and Kashmirj2 and with coins of Azes I, Azes 11, Hermaeus and the
Parthian king Orodes I1 (57-37 B.c.) in the hoard from Mohmand.)' In Stupa IV
at the Dharmarijika at Taxila, a Roman denarius of Augustus struck between
2 B.C. and A.D. 14 was found with a drachm of Azilises in a stupa deposit. T h e
evidence of overstrikes and stratified finds at Taxila shows that Azilises suc-
ceeded Azes I and was followed by Azes 11. H e continued t o use the Azes I
obverse type showing the king with a couched lance, but subsequently adopted

25. Lohuizen-de Leeuw, 1949, pp. 1-50.

26. Konow, 1929, pp. 70-7.
27. Ibid., pp. 16-17.
28. Marshall, 1936, pp. 973 et seq.
29. Fussman, 1980, pp. 143.
30. Jenkins, 1955, pp. 1 et seq.
31. Tarn, 1951, p. 348.
32. Whitehead, 1923, p. 338.
33. Jenkins, 1955, pp. 23-5.
the new obverse type of the king holding a whip which was continued by his
successor, Azes 11.
Azes 11 succeeded as King of Kings in the lndus kingdom, and probably
added Jalalabad and Gardez to his empire. Hoards of copper and silver coins of
Azes 11 have been reported from Jalalabad and G a r d e ~ . ~H' e used the Zeus
Nikephorus reverse type for his silver issues from Taxila and the Pallas types
for his mint in Gandhira. At the end of his long rule there was a major debase-
ment of the Saka silver coinage datable before A.D. 42.j5 Drachms of Azes 11 of
the Zeus Nikephorus type, first in base silver and then in billon (a debased alloy
of silver and copper in which silver constitutes only about 20 per cent), were
struck in large numbers and are very common in finds from north Pakistan.
the same time the coinage is copied by a series of local satraps and strategoi - the
satraps Jihonika (Zeionises) and Rajiivula and the strategoi Indravarma and
Aipavarma. Subsequently the Kushan king Kujula K a d ~ h i s e scopied some of
the coin types of Azes 11; and both their Indo-Parthian and Kushan successors
continued to use the era of Azes for some time, no doubt because they wished
to claim continuity with the Azes dynasty.

The Pahlavas in India

The dynasty that reunited the fragmented empire of Azes I1 was the Pahlava or
Indo-Parthian kingdom of Gondophares. The various forms in which the name
of this ruler appears on the coins are merely attempts to render in Greek the
outcome of the O l d Persian title 'Vindafavnah' (the winner of glory).'' Most of
Gondophares' coins in the Indus valley struck in billon retained the obverse
type of the tetradrachms of Azes I1 of a horseman holding a whip, adding a cha-
racteristic tamgha - the Gondopharan symbol. The Takht-i Bahi inscription of
Gondophares is dated in Year 103 and the twenty-sixth year of his reign, fixing
his accession in o r about A.D. 20 and giving him a long rule up to A.D. 46 or
1ater.j' This context is consistent with the legend associated with St Thomas in
the Apocryphal Acts of T h ~ r n a s , ' which
~ mention King G u n d a ~ h a rand his
brother Gad in connection with Thomas' enforced stav at his court. The death
of the king's brother, his restoration to life and finally their conversion were
instrumental in making Gondophares known to the Western world.

34. MacDowall, 1973, p. 212.

35. Ibid., p. 228.
36. Rapson, 1922, p. 577.
37. F o r his inscription from Taxila dated Year 103, see Konow, 1929, pp. 57 et seq., and also
Introduction, p. xlviii; his coins are recorded by Whitehead, 1914, Vol. I, Plate XV. For
refercnce t o Gondopharcs in the legend of St Thomas, see Smith, 1924, pp. 245 et seq. A
comprehensive bibliography on the subject is provided by Va1lC.e Poussin, 1930, p. 277.
38. James, 1924, pp. 373-5.
 The Sakas and Indo-Parthians

philostratus (Llfe of Apolfonius of Tyana I I , p. 20 ct seq.) writing around

A.D, 217 describes Apollonius' visit to Taxila in A.D. 43-14, which should bc
within the period when Gondophares was King of Kings in the Indus prov-
inces. The King of Taxila had a Parthian name, Phraotes, and was independent
,f vardanes, the King of Kings of Parthia. H e was powerful enough to exercise
suzerainpowers over the satrap of the Indus (Gandhira), who was also not sub-
ject to the Parthian King of Kings. The king spoke Greek well and discussed
The king comments that he shares his wealth with his enemies:

The barbarians were perpetually making raids into my territories, but I kcep them
quiet and control them with money so that my country is patrolled by them and
instead of invading m y dominions they themselves keep off the barbarians o n the
other side of the country.

The reconstruction of the political history of the Indo-Parthian dynasty of

Gondophares depends almost entirely on numismatic evidence and the way in
which the coin series of the dynasty are arranged. In some cases inscriptions on
the coins give a statement of relationship to a predecessor, presumably to stake
a claim to legitimacy. Whether such titles indicate a blood relationship or
simply honorific titulature does not greatly matter, because they should indicate
the sequence of rulers. R a p ~ o n suggested
'~ that Gondophares immediately fol-
lowed Azes I1 because both rulers were associated with the same strategos
Aipavarma, son of the strategos Indravarma. But the coins of ASpavarma are
not contemporary issues of Azes 11. They seem to be copying the types of Azes
I1 including the Greek legend of Azes I1 after the great debasement, when local
rulers and officials asserted their independence; and the supposed joint issues of
Aipavarma and Gondophares are in fact coins that have the titles of ASpavarma
alone on both obverse and reverse with an additional symbol, which has been
called the G o n d o ~ h a r a nsymbol because it is also found on billon tetradrachms
of Gondophares and his successors in the Indus provinces. Major new discov-
eries of Indo-Parthian coins and fresh studies based on them have made it
necessary to revise substantially the reconstruction of Indo-Parthian history
that Rapson suggested some seventy years ago.
MarshalPo found in excavations at Taxila a hoard of base silver drachms of
the later Indo-Parthian kings, Sasan, Sapedana and Satavastra. A further speci-
men was found at Saudpur in Sind." This series with the reverse type of Nike
(Victory) holding a wreath can be attributed to the province of Sind. The exca-
vations at Taxila" yielded substantial numbers of billon Indo-Parthian tetra-

39. Rapson, 1922, p. 590.

40. Marshall, 1936, pp. 27-46.
41. Ibid., p. 95.
42. Marshall, 1951, Vol. I, pp. 751-842.
drachlms with the mounted-king obverse - the currency of the upper Indus
provinces, Taxila and Gandhira - of Gondophares, Abdagases (the nephew of
Gondophares) and Sasan, who is sometimes given the double name of ' ~ ~ d ~ ~ h
ara Sasa' in his K h a r o ~ f hlegends.
i Simonetta" analysed the Indo-Parthian
drachms of Parthian type - the currency of Aria and Sistan - and published a
sequence of kings, Orthagnes, Gondophares Autokrator, Abdagases, Pakores
and Sanabares, followed by cruder drachms struck in copper in the names of a
later Sanabares and a later Gondophares. T o this series Alram4' has added new
types of a Gondophares, surnamed Sae, and of Ubuzanes the son of Orthagnes.
Ma~Dowall.'~ has shown that the Nike-type copper tetradrachms of the Indo-
Parthians were the currency of Arachosia and Sistan, and that the sequence of
Gondophares Soter, Abdagases, Orthagnes, Pakores, a second GondophareS
styled Megas, Sarpedanes and Satavastra was followed by Ardamitra who intro-
duced the new reverse type of an early Sasanian fire altar. This series was clearly
the coinage of the later Indo-Parthians in Arachosia and Sistan, the types found
in the British excavations at Kandahar from which Kushan coppers are con-
spicuously absent.4hCribb17 has published new types of billon drachms copied
from debased issues of Strato I1 and Rijuvula with a crude thundering Pallas
reverse for kings G o n d o ~ h a r e s ,Abdagases, Sases, Ubuzanes and Sarpedanes.
Sases, Ubuzanes and Sar~edanesalso have the second name 'Guduvhara' sug-
gesting that this title 'winner of glory', used as a personal name by the founder
of the dynasty, became a family name and title for some of the later Indo-Par-
thians.
The presence of overstrikes ~ r o v i d e suseful confirmatory evidence. Cop-
per imitation tetradrachins of Hermaeus, struck in his name long after his death,
are overstruck by Arachosian Nike tetradrachms of G o n d o ~ h a r e s which
, are in
turn overstruck by Kujula Kadphises and by Jihonika..'"
The nameless king Soter Megas copies the local-type billon tetradrachms
of Sasan in Gandhira,4')and copper didrachms of the nameless king resu sum ably
the later stages of the long reign of Kujula K a d ~ h i s e s )are overstruck by
Pakore~.~O
The full picture of Indo-Parthian political history must await the detailed
numismatic study of each of these Indo-Parthian coin sequences and of their
interrelationships with the rulers who receded and followed them. In the
meantime we can see that there werc separate Indo-Parthian currencies in Aria1

43. Simonetta, 1957, pp. 44-64.

44. Alram, 1983, pp. 69-74.
45. MacDowall, 1965, pp. 137-48.
46. Helms, 1982, p. 9.
47. Cribb, 1985, pp. 282-300.
48. Mitchiner, 1976, p. 734.
49. MacDowall, 1968, pp. 29 et seq.
50. Simonctta, 1957, pp. 53-9.
 Thc Sakas and Indo-Parthians

Sirtan, ~ r a c h o s i a ,the Indus valley provinces, Sind and the east Panjab, each
distinctive obverse and reverse types. We see the emergence of Gondo-
phares, 'the winner of glory' and founder of the dynasty in the Indus prc~vinccs,
Kapisene and east Panjab. His immediate successors Abdagases and Saran lcjst
Kapisene to Kujula Kadphises but retained the Indus provinces, Arnchosia
east panjab, and Sasan controlled Sind. In the latter half of the first century *.I,.
later ~ndo-Parthianrulers continued t o hold Arachosia, Drangiana and Aria up
to the Sasanian conquest. Even after they had lost Gandhira and Taxila, later
do-Parthian rulers are t o be found in east Panjab and Sind. The Periplur
(Chapter 38), rob ably dated towards the end of the first century A.D., says that
the provinces of the lower Indus, still called Scythia, were ruled by Parthians
who were continually expelling each other. There is no doubt that this feuding
and civil warfare makes the sequence of rulers here so complex.
At its height the empire of Gondophares covered substantially more terri-
tory than the Indo-Scythian dynasty of the House of Azes had done, extending
from Aria and Sistan in the west t o Mathura in the east and including Kabul and
Begram in the north and Kandahar and the mouth of the Indus in the south.

The administrative system of the Sakas and Pahlavas5'

The coins of the Sakas f r o m the time of Maues show that their rulers did not use
the simple titles of 'Basileus' and 'mahirija' (king) but assumed the grander
titulature of 'Basileus Basileon' and 'mahirija rijatiriija' (King of Kings). The
use of these higher titles seems t o have been copied from the usage of contem-
porary rulers of Parthia, where the title 'Basileus Basileon' had been introduced
by Mithradates I1 about 123-88 B.c.,~? and was regularly used on drachms and
tetradrachms from the time of Mithradates 111, 57-54 B.c., and Orodes 11, 57-38
B.c.~' The title had some substance in the Indus provinces, because the Sakas,
like the Pahlavas, seem t o have left local hereditary monarchs in office, w h o
acknowledged their suzerainty and acted as local rulers, such as Visnuvarma,
King of Apraca (Bajaur), the father of Indravarma, w h o dedicated a reliquary in
Year 63 Ayasa (A.D. 6),54and Ajitasena, King of Odi, who also dedicated a reli-
q~ary.~~
At several stages the coins seem t o show a system of joint rule, where a
senior king has a junior colleague w h o was intended t o succeed his senior

51. The study of the administrati\~esystem of the Sakas and Parthians is covered in a wider
perspective in Puri, 1968a, Vol. I, Chapter IV, pp. 85 et seq.; see also ~ i l a k a n t h aSastri,
1957, pp. 218 et scq.; Rapson, 1922, pp. 574-5.
52. Sellwood, 1980, pp. 70 ct seq.
53. Ibid., pp. 110 et seq.
54. Fussman, 1980, pp. 2 et seq.
55. Fussman, 1986, pp. 1 et seq.
 B. N.Puri

partner after death. The fully developed system is seen in the succession ,f
kZatrapas and mahikratrapas among the Saka rulers in Mathura and western
India (see below). The concept of dvairdjya (double kingship) was known to
Indian political thinkers.i6 It seems to have been based on political requirements
to contain disruptive forces near the throne. It is usually thought to be indicated
by the joint coins of Vonones with Spalahora, Vonones with Spalagadama, Spa-
liriia with Spalagadama, Spalyrises with Azes, and Azes with Azilises. In the
Pahlava dynasty of the House of Gondophares it was thought that there were
joint coinages of Gondophares and Gadana, Orthagnes and GondophareS
Gadana, Gondophares and S a ~ a n ; ~but ' it now seems that 'Gondophares' and
'Gudaphara' were 'winner of glory' titles, which became a sort of family name
for many subsequent members of the family. There can be no doubt that the
supposed joint coinage of Hermaeus with Kujula Kadphises and of Azes I1 with
ASpavarma are essentially coinages of Kujula Kadphises copying the types of
Hermaeus and ASpavarma copying the types of Azes I1 after a major debase-
ment of the denomination.
Major districts of the empire were governed by k~atrapas(satraps) and
mahaksatrapas, a system dating back to the time of the Achaemenids. 'Satrap' is
the ~ e i l e n i s t i cvariant of the O l d Persian 'xiagapavan' (protector of the realm).
These provincial governors were men of standing and position, appointed by
the 'Great King' to maintain law and order in their territory. They owed alle-
giance to the supreme monarch, but were otherwise for ~racticalpurposes inde-
en dent. The weakness of central control over the satraps encouraged centrifu-
gal tendencies and the strife between the satraps appointed by Alexander was
largely responsible for the breakup of his empire under his successors. The
Indo-Greeks retained the system of government through subordinate and asso-
ciated kings. The system of government through satraps continued under the
Saka-Pahlava rulers.5RIn the Taxila copper t late inscription of Year 78, Liaka
Kusulaka was satrap of Chukhsa, the great lain of C h a ~ h Kusulaka .~~ Patika
was mahdk~atrapain the Mathura lion capital inscription." Early in the first
century A.D., in Year 191 of the Graeco-Bactrian era (c. A.D. 30), ~ihonika,son
of Manigula, was satrap of Chukhsa6' and issued a series of base-silver tetra-
drachms and copper coins in his own name,62when he seized independence at
the end of the reign of Azes 11. From Jalalabad in Afghanistan an inscription of
Year 83 (A.D. 25) refers to a satrap, Tiravharna, who must have been an early
Indo-Parthian satrap under G o n d o ~ h a r e s .

56. Puri, 1968a, p. 89.

57. Mitchiner, 1976, pp. 740 et seq., 755 cr seq.
58. VallCe Poussin, 1930, pp. 268 ct seq.
59. Konow, 1929, pp. 23-9.
60. Ibid., pp. 3 0 4 9 .
61. Ibid., pp. 81-3.
62. Whitehead, 1914, pp. 157 ct seq.; MacDowall, 1973, pp. 215-30.
 The Sakas and Indo-Parthians

From the Panjab the satrap Kharahostes is known from the copper coins
he issued early in the first century A.D.~' The coin inscriptions describe him as
the son of Arta, and he is identified by most scholars with the Yuvaraja (heir
apparent)Kharaosta of the Mathura lion capital inscription,64whose daughter
Ayasia Kamuia was the Chief Queen of Rajula (i.e. Rijuvula). A seal inscription
of Sivasena shows him as satrap of Abhisira, which has been identified with the
hill country between the Jhelum and Chenab.65
The titles of 'ksatrapa' and 'mahikratrapa' were regularly used by the
Saka rulers of Mathura in the first centuries B.c./A.D. Coins are known for the
earlier group of Saka rulers, the ksatrapas Sivaghoga, Sivadatta, Hagimasa and
Hagina.b6In the later group, coins are known for Rijuvula as ksatrapa and
mahik!atrapa, for his son Sodisa as ksatrapa and mahik!atrapa and for another
kptrapa [Tora]nadan6' The Mathura lion capital inscriptionbRrecords the relig-
ious gift of the kratrapa Sodisa, son of the mahiksatrapa Rijuvula, and is in
honour of the m a h i k v t r a p a Kusuluka Patika and the ksatrapa Mevaki Miyika
in honour of the whole of Sakastan. Sodisa subsequently became mahZkratrapa,
as we can see from four inscriptions that record religious benefactions, one of
them dated to Year 42 (A.D. 72).69At this stage the mahiksatrapa was assisted
by a satrap who eventually succeeded him.
The rulers of the western group of Sakas in Malwa and Kathiawar also
used the titles 'ksatrapa' and 'mahiksatrapa'. The inscriptions of Nahapina
from Nasik, Karle and Junnar dated Years 41 to 46 (probably A.D. 120-25)
show him first as ksatrapa and then as mahiksatrapa, while his coins refer to
him as rajno and Basileus (king).'O In the Girnar inscription Rudradiman claims
that he himself acquired the title 'mahiksatrapa' by virtue of his conquests."
After the dynasty of the Western Satraps became firmly established, the long
series of their coins dated in the Saka era of A.D. 78 shows a mahiksatrapa and
k~atraparuling together, the ksatrapa occupying the role of heir apparent and in
due course succeeding to the office of mahiksatrapa, as we saw earlier at Math-
ura. By the second and third centuries A.D. these mahiksatrapas seem to have
been independent rulers.72

63. Whitehead, 1914, p. 149.

64. Konow, 1929, p. 46.
65. Ibid., p. 103.
66. Allan, 1936, pp. cxi et seq., 185 et seq.
67. Ibid., pp. cxii et seq., 185 et seq.
68. Konow, 1929, pp. 30-49.
69. Rosenfield, 1967, p. 264.
70. Rapson, 1908, pp. Ivi et seq., pp. 65 et seq.
71. Kielhorn, 1905106, pp. 36 et seq.
72. For the history of the Saka kZatrapas of western India, see Rapson, 1908, ~ntroduction;
also Rapson, 1922, p. 577. See also: Nilakantha Sastri, 1957, chapter IX; ~ o s e n f i e l d ,
1967, pp. 130 et seq. A summary o f the records dealing with the history of the ~ a k a and
s
 The strategos (general o r military governor) is another official in the Sakn
and the office seems to have been hereditary in some cases, A[
the end of the reign of Azes I1 coins are struck by Indravarma Apraca Raja and
his son the strategos Aipavarma, probably as an independent ruler.71 Rapson"
thought that this Greek title 'strategos', which is equivalent t o the Indian 'sena-
pati' (lord of the army), was inherited by the Sakas and Pahlavas from the Indo-
Greeks. But Tarn75points out that Indian satrapies governed by generals (strate-
goi) simply reflected the continuation of Seleucid practice that one would expect
in a successor state.
~t a lower level, administration was conducted by a mevidarchos (district
officer). Kharosthi. . inscriptions record a meridarchos Theodorus on a Swat relic
vase and an unknown meridarchos from a copper plate inscription from %apur
near T a ~ i l a both
, ~ ~ apparently of the Greek ~ e r i o d .

Phil
The Sakas in India, especially the Indo-Scythians under Maues and the House
of Azes in the Indus valley, progressively occupied provinces that had been
ruled by the Indo-Greeks since the time of Menander. They inherited and con-
tinued to use the Greek political institutions and culture that they found. They
retained the Greek provinces and the Greek system of administration, even
retaining the Greek names for officers such as 'strategos' and 'meridarchos'. In
town planning they retained the Greek chessboard form of town plan at Sirkap
(Taxila). Their buildings, such as the Saka temple at Jandial, were Greek in plan
and decoration. Greek had ousted Mauryan art at Taxila, but it had become
increasingly moribund. Removed from cultural contact with the West, it
became increasingly Indianized under the Sakas. The process is seen in the stupa
of the double-headed eagle in Sirkap, where some niches still have the pedi-
mented front of a Greek temple but others have ogee Indian arches and the
form of Indian toranas. Toilet trays under the Sakas retain the stones the Greeks
had employed and copy Greek models but progressively introduce Indian
motifs like the lotus into their background designs.
Parthia proper7' was a Seleucid successor state (see Chapter 5) and the

their relations with contemporary powers is given by Kapson, 1908, pp. Ivi c t seq.,
Nos. 31-7, covering the inscriptions of the Ksaharita dynasty, and Nos. 3 8 4 2 , covering
with those of the Western Satraps of Cas~ana'sfamily.
73. Mitchiner, 1976, pp. 601 et seq.
74. Rapson, 1922, p. 577.
75. Tarn, 1951, p. 241.
76. Konow, 1929, pp. 2-5.
77. The history of Parthia is rccordcd by: Llebevoise, 1938; Tarn, 1930; Ghil-sliman, 1961;
Lozinski, 1959; Huart, 1927.
 The Sakas and Indo-I'arthians

pahlavas who succeeded the Sakas in the Indus provinces brought with them
imporrant Hellenistic elements that were valued in Parthia proper. ~t home,
parthians also retained established Greek cultural traditions and institutions
inherited from the Seleucids. Although the sizeable concentrations of Greeks
and Hellenized peoples had lost the privileged position they had once enjoyed,
Greek remained the lingua francs f o r commerce in Western Asia. Greek speech
and culture were appreciated b y the educated Parthians. Orodes spoke
Greek and Greek tragedies were played at his court, as Plutarch relates. Among
their titles such as 'Dikaios' (the just) and 'Euergetes' (the benefactor) Parthian
kings used the title 'Philhellenos' (lover of Greek culture) on their coins, occa-
sionally from the time of Mithradates I and regularly from the middle of the
first century B.C. When the Indo-Parthians came t o control the empire of Azes
11 in the first decades of the first century A.D., the Indus provinces saw a renais-
sance of Philhellenism, drawing o n fresh Western sources, which brought a new
and strong Hellenistic influence o n the coinage, art and architecture of the
empire of Gondophares, which can be seen so clearly in the Pahlava period of
the excavations at Sirkap ( T a ~ i l a ) . ' ~

The monetary system and coinage79

The Sakas retained the monetary system of their immediate predecessors, the
Indo-Greek kings such as Apollodotus 11, with the standard denominations of
the silver tetradrachms and drachms struck t o the Indian weight standard, with
copper coins providing the smaller denominations. The high purity of the silver
remained unchanged until the end of the issues struck in the reign of Azes 11,
when the coinage suffered a great d e b a ~ e m e n tThe
. ~ ~ obverse and reverse types
of the silver tetradrachms remained the same, but the denominations were
struck in billon - a debased alloy of silver and copper in which silver forms
around 20 per cent only. There was a corresponding deterioration in the design
of the coinage and production of the copper denominations proper was sus-
pended. Marshall attributes this sudden change t o the introduction of lead and
billon currency by Rajiivula and expansion of trade with the Andhra Empire,
which had coins of those base metals; but the debasement coincided with the
breakup of the Azes dynasty's empire. It seems more likely that the dynasty
suffered an economic collapse and was deprived of access t o its sources of silver,
either from Panjshir o r the Arvelli mines. The coinage of succeeding rulers in

78. Marshall, 1951, Vol. I.

79. A study of Saka and Indo-Parthian coins has engaged the attention of many scholars and
numismatists: see Gardner, 1886; Whitehead, 1914; Jenkins, 1955; Narain, 1957. A good
summary is provided in Rayson, 1922, pp. 586 et seq., as also in ~ i l a k a n t h aSastri, 1957,
Val. 11, pp. 197 et seq.
80. Marshall, 1951, Vol. 1, pp. 53 et seq.
Taxih and Gandhira - the dynasty of the general Aipavarma and the Houseoi
Gondophares - was also struck in billon. There seems t o have been a compar-
able debasement in the silver coinage of Arachosia and the Paropamisadae,
where the silver tetradrachms of the successors of Hermaeus were struck there-
after with the same types in copper, and the denominations continued in this
form under Gondophares and his Kushan successors; and in east Panjab, where
the silver drachrns of the later Stratos and RZj8vula are succeeded by billon
drachms with the same types, the denominations struck by the Gondopharer
dynasty were struck in the same debased metal in this province, too. Only in
the silver drachms of Aria and Sistan modelled on the Parthian coinage and in
the provinces of the lower Indus, notably Sind, did the Pahlavas have a silver
coinage.
The mints and engravers of the Indo-Greeks remained at the service of the
new Saka rulers - Maues and the House of Azes - and they continued to use a
wide range of reverse types as their Indo-Greek predecessors had done. They
retained unchanged the bilingual character of the coinage of Gandhira and
Taxila with the titulature of the king in Greek on the obverse and in Kharosthi
on the reverse of each issue. But the Sakas made two major changes. Instead of
the Indo-Greek title of 'king' they styled the Saka rulers who struck coins
'Great King of Kings' - Basileus Basileon Megas in Greek and rajati~djamahdta
in Kharosyhi. Maues n o longer used the obverse type of the king's portrait on
the silver denominations, the normal Indo-Greek practice; and Azes I intro-
duced a new convention, the obverse type of the Saka king on horseback on the
silver denominations of tetradrachm and drachm.
The Pahlava rulers who succeeded the Sakas in the Indus provinces copied
these new conventions. In other territories they retained the pattern of the coin-
ages they found already current. In east Panjab, Gondophares and his succes-
sors copied the billon drachms of Rijuvula, with a stylized obverse portrait and
a crude representation of a thundering Pallas on the reverse. In Arachosia, Gon-
dophares and his successors issued copper tetradrachms with a distinctive
obverse portrait as the copper tetradrachms of Hermaeus had done, substituting
a new reverse type of Nike appropriate to the new conqueror. In Aria and
Drangiana, Gondophares and his successors issued Parthian-type silver drachms
retaining the obverse portrait and reverse type of the seated king that character-
ized the silver drachms of thc kings of Parthia.
 The Sakas and Indo-Parthians

Architecture and arts1

The Sakas at Taxila" followed Greek ideas in town planning, copied Greek
prototypes in their architecture and were inspired by Greek forms in minor arts
and crafts. Among Saka buildings at Taxila the temple at Jandial has a typical
Greek plan with classical moulding and Ionic columns. Greek in concept are the
small stupas dated by coins t o the period of Azes 1 and 11. Marshall's analysis
that the Sakas initially used Greek ornamental features only, and at Jan-
dial a plan typical of a Greek temple in mainland Greece with basically the same
methods of construction and mouldings; and stupa decoration used classical
acanthus leaves. Subsequently Indian influence becomes increasingly noticeable,
and there is a mingling of Greek and Indian motifs. The stupa of the double-
headed eagle has Corinthian pilasters and niches of three types - two with the
pediment characteristic of Greek buildings, some with Indian ogee arches and
others in the f o r m of early Indian toranas. Marshall sees the same creeping
Indianization in the minor arts. For example, the stone toilet trays retain Greek
models, but introduce Indian motifs. C u t off from the Greek world by hostile
Parthia, new inspiration only came from the Indian art centre of Mathura.
In the early years of the first century A.D., about the time of the Pahlava
conquest by Gondophares, Taxila suffered a great earthquake. When the city
was rebuilt, new building methods were introduced, including a strong form of
diaper masonry that was much more durable than the early rubble construction.
Buildings were reduced in height, and houses were limited to two storeys, the
lower one in the form of a basement. T h e legend of St Thomas, skilled in archi-
tecture and all types of work in wood and stone, being sold in Syria t o a mer-
chant called Habban and being brought to the court of Gondophares, is set
against the background of skilled craftsmen from Western Asia being commis-
sioned to d o w o r k for the Pahlavas.
A favourite theme in secular art was the drinking scene adopted for the
decoration of Buddhist stupas such as the drinking party of nine figures, prob-
ably from Swat, of ~ u r e l yHellenistic character." The men wear either a short
Greek chiton o r a simple himation; the women a long chiton and hiination. The
man in the centre holds a metal tankard, while others hold goblets of a dis-
tinctive type with horizontal flutings and a disproportionately small base of the
kind found in Pahlava levels at Taxila. Other reliefs now in the Peshawar

81. For a study of art under the Sakas and the Parthians, see Marshall, 1936, 1951, 1960;
Lohuizen-de Leeuw, 1949; Rosenfield, 1967; Rowland, 1953. The finds in Taxila at the
Saka-Parthian levels provide an index t o the material culture during this period. The
classical accounts - Greek and Rotnan - noticing trade contacts between India and the
Western world are helpful by recording items of import and export. Sec Marshall. 1951;
Warmington, 1928. Scc also Nilakantha Sastri, 1957, pp. 220-1.
82. Marshall, 1951, Vol. I.
83. Marshall, 1960, p. 33.
Museumx' retain Hellenistic characteristics; the form and posture of the figures
are Greek rather than local, though they wear local dress and all carry lotuser,
while a Corinthian pilaster frames the group o n each side.
The rebuilt Pahlava city of Taxila produced a rich range of finds, particu-
larly gold jewellery, silver plate and bronze vessels, probably buried when the
city was under immediate threat from the Kushans. Objects found among the
debris of buildings destroyed at this time include ornaments of personal use,
household utensils, implements and arms, many of strongly Hellenistic charac-
ter. Some objects seem to have been imported from the West such as a head of
Dionysus in silver repoussk, a cast bronze statuette of the Egyptian ~ h i l d - ~ ~ d ,
Harpocrates, and a buff-coloured terracotta female head with delicate and sensi-
tive modelling, all very fine specimens of Hellenistic art.85
The Pahlava conquest of Taxila clearly led to a major influx of articles
from the Graeco-Roman world, and to encouragement being given to artists
and craftsmen to imitate Western models. The reopening of trade routes across
Parthia and the development of sea routes t o the lower Indus facilitated this
process, but the revival of Hellenism generally came with the P a h l a ~ a s . ~ ~

Religious developmentsR7
Many of the inscriptions of the Saka and Pahlava period are religious dedica-
tions, often recording the deposit of Buddhist relics and foundations. Under
Maues, early under Saka rule, the Taxila copper plate of PatikaHarecords the
establishment of the relics of the Lord Sikyamuni and the founding of a
samghavama for the worship of all Buddhas. From the period of the Azes
dynasty we have a growing series of Kharoghi inscriptions from Buddhist reli-
quaries - the Bimaran vase,8" the reliquaries of Indravarma and Ra~naka,'~ of
Satrea," of Ajita~ena,"~etc. The establishment of Buddhist stupas, sometimes
including Indo-Parthian coins, continued into the Pahlava and Kushan periods.
The Sakas of Mathura were also patrons of Buddhism. The Mathura lion

84. Marshall, 1960, pp. 34-6.

85. Ibid., pp. 26-7.
86. Marshall, 1951, Vol. I, pp. 65-6.
87. F o r a study of the religious conditions in thc Saka-Pahlava pcriod, the dates provided by
inscriptions and coins arc thc authoritative source rnatcrials. Architectural finds of
Hindu divinities o n coins suggcst an atmosphere of catholicity in this period. The
secondary sources include Nilankantha Sastri, 1957; Rosenfield, 1967.
88. Konow, 1929, pp. 23-9.
89. Ibid., pp. 50-2.
90. Fussnian, 1980, pp. 1-43.
91. Fussman, 1985, pp. 29-34.
92. Fussnian, 1986, pp. 1-14.
 T h e Sakas and Indo-Parthians

records the family gift made by the Chief Queen of the mahik!atrapa
Rajula (i.e. ~ i i j ~ v u l aAyasia
), Kamuia, establishing a relic of the Buddha, found-
ing a stupa and samgharrima, and granting land t o Budhila, a monk from Nag-
,,,,The Briihmi inscriptions in western India show similar benefactions by the
western Saka satraps. T h e Nasik inscriptions of Rsabhadatta, son-in-law of
NahapPna the Ksaharita satrap,')' record the construction of caves and
benefactions t o the Buddhists and endowments to provide for the monks in the
rainy season.
There can be n o doubt that the prevailing religion was Buddhism. But the
Sakas also retained their o w n Iranian faith. The imposing temple of Jandial, in a
key position outside the gate of Taxila, shows the high regard in which Zoroas-
trianism was held b y the upper classes of Taxila. In general the Sakas seem t o
have had a sympathetic, tolerant attitude to all the religions of their subjects,
~uddhism,Jainism, Brahmanism, etc., and the position did not change under
the Pahlavas.
The reverse types of the coins of the Indo-Scythians and Indo-Parthians
remain essentially in the Greek tradition. The Greek gods Zeus, Artemis,
Apollo, Poseidon and Nike still dominate the types used by Maues, as well as
the same Greek gods with Hermes, Pallas, Demeter and Heracles under the
Azes dynasty, including, of course, animal types such as the lion, elephant and
Indian humped bull. T h e divinities externally were wholly Greek, though for
contemporary subjects they may have had a wider significance of other deities
disguised under the Interpretatio Graeca. The Pahlavas continued in the same
tradition. G o n d o ~ h a r e snaturally showed a reference for Nike in various
forms, but also used f o r the first time the figure of Siva on his coinage from the
middle Indus provinces.

93. Konow, 1929, pp. 30-49.

94. Senart, 1905, pp. 78 et seq.
 Ma Yong and Wang Binghua

The Xinjiang region

The Xinjiang region of China lies in eastern Central Asia t o the north and south
of the T'ien Shan mountain range. South of the T'ien Shan mountains lies the
Tarim basin, bounded o n the west by the Pamir plateau, 'the roof of the world',
and on the south by the Kunlun and Altyn Taq mountain ranges. These three
sides consist of high mountain peaks permanently covered with snow and ice,
with few passes by which they may be crossed. The east side of the Tarim basin,
however, is comparatively low and gives access t o the Gansu corridor. The
basin is rather like an inverted trapezium, with its centre occupied by the wide
Taklamakan Desert. T h e melting snows of the Kunlun, T'ien Shan and Pamir
mountain ranges flow from different directions towards the desert, forming
short inland streams. These give rise t o a number of oases, but each is relatively
small; and the desert that separates them makes communications difficult. From
antiquity it has been impossible f o r people t o live anywhere but in these oases
on its borders.
North of the T'ien Shan mountains lies the Dzungarian basin, shaped like
an isosceles triangle, with the Altai mountains t o the north-east, and a few low
mountain ranges t o the north-west, but n o major barrier t o the east o r west.
South of the basin in the T'ien Shan range there are many valleys with flourish-
ing vegetation. T h e lands near the Urungu and Irtish rivers, with the upper
reaches of the Ili, are all suitable for nomadic grazing.
The principal evidence for the history of Xinjiang in the centuries before
the Christian era comes from archaeology. There are few reports in Chinese writ-
ten records and these see111 to be largely derived from hearsay. It has been sug-
gested that the tribes living in the remotest east, mentioned by Herodotus (IV.13)

See Map 6.
 Ma Yong and Wang Hinyhua

quoting the poet Aristeas, were the ancient ~ e o p l e sof the Altai mountains, bur
this is mere conjecture.'

The Xiangbaobao graves, Tashkurgan

The grasslands in the valley of the River Tashkurgan, some 4,000 m above sea-
level, but still suitable for nomad grazing, have a large number of ancient graves
still visible on the surface covered with blocks of stone. Forty graves of this
type excavated at Xiangbaobao yielded Carbon-14 dates between the seventh
and fourth centuries B . c . ~ In spite of varied methods of burial - sometimes
inhumations in various positions with a wooden cover over the grave pit and
sometimes cremations with n o wooden cover - their funerary contents are bas-
ically uniform and belong t o a single cultural type. Four graves contain the
remains of human sacrificial victims. Most bones are decayed, but one well-
preserved skull is of a Europoid type. The tombs contain few funerary objects,
suggesting a poor, unostentatious life-style. Finds are mostly everyday utensils
with hand-made pottery made from clay containing coarse sand and mica
flakes, fired at a low temperature. Most of it is undecorated, with an uneven
red-brown o r grey-brown surface colour. Cooking implements and containers
- cauldrons, jars, dishes, bowls and cups - predominate. This pottery has dis-
tinct marks of use, even t o the extent of repair after breakage, and is found with
metal objects such as iron knives and bronze arrows. The ornaments found
include gold plaques, bronze or iron rings, earrings, buttons and terracotta,
stone, bone and agate beads. A few bronze ornaments shaped as a pair of
sheep's horns (Fig. 1) represent quite a high level of craftsmanship. Besides,
these the graves have also ~ i e l d e dcloth woven from sheep's wool, sheepskin
clothes, bones of animals (mostly sheep) and birds, and a wood-drill fire kindler
of the type commonplace at oases in the Tarim basin. It is clear that these local-
ities once supported fixed settlements of inhabitants w h o raised livestock, par-
ticularly sheep, as their principal economic support.
Hunting was an important supplementary occupation; the level of crafts-
manship was quite low, but the use of metal was developed. Distinctions be-
tween rich and poor had already appeared in this society. It is not clear whether
the variation in styles of burial was due t o racial differences or chronological
factors, but the coexistence of inhumation and cremation suggests that there
may have been an amalgam of cultures. This locality borders Saka territory to
the west and adjoins districts where the Ch'iang people lived to the south. The
Europoid skull from an intact burial may represcnt the Sakas. The re valence of
cre~nationamong the Ch'iang tribes may be sccn in written records of the

1. Sun, 1984.
2 . Chen, 1981.
 The culture of the Xinjiang region

FIG.1. Bronze ornaments: one is shaped like a pair of sheep's horns.

Xiangbaobao, Tashkurgan.

pre-Chin period.3 These graves of the Tashkurgan region may therefore repre-
sent a mixed culture of the Saka and Ch'iang tribe^.^

The Lou-lan graves at Lop N o r

In antiquity Lop N o r was a large salt lake at the hub of communications be-
tween the Gansu corridor and the Tarim basin, but changes in the course of riv-
ers have caused it t o dry up and become a salt marsh. The famous Lou-Ian site
lies on the north-west bank of the Lop N o r marsh, where the Kongque river
now flows into the marsh. In the first century B.C. it was the capital of the state
of Shan-shan. The graves found near by used to be considered as graves of the
Western Han period, but recent Chinese excavations have yielded material from
the seventh t o the first century B . c . ~ These graves have a coffin chamber of
wooden planks in a shallow pit. The corpse, laid in an extended position, was
wrapped in a woollen cloth, and wore hide boots and a ~ e a k e dbrown felt hat

3. Mo-tzu, Chapter 6 ; Chieh-tsang b; Hscrn-tzu, Chapter 19; Ta-lcreh Lu-shih ch'un-cb'lu,

Chapter 17; Hsias-lrsing-lacr, 1 .
4. Russian archaeologists have excavated ancient graves, similar to this culture, in the Pamir
region in the former Soviet Union. See Bernshtam, 1952. These scholars believe chat they
belong t o the Saka culture.
5. In the past there werc many reports of this culture type in the Lop N o r region. See Stein,
1974; Huang, 1948. In recent years the Archaeological Research Institute of the Xinjiang
Academy of Social Sciences has made three s u n e y s near the Lop Nor, and on the lower
reaches of the Kongque river, and has gathered a collection of fairly complete scientific
materials. The). are being collated and a report will soon be published.
 Ma Yong and Wring Binghcra

with bird feathers. The borders of the women's hats were often embroidered,
The woollen cloth was gathered into a pouch on the upper chest and filled
fragments of branches of the medicinal herb ephedra. They also contained a
small basket woven from hemp or grass containing up t o fifty grains of wheat,
Due to the extreme aridity, many of the corpses have been preserved.
Their physical features are very clear. They have golden hair slightly curled, a
high nose, deep-set eyes, rather long eyebrows and narrow cheeks.
ological measurements suggest Homo Alpinus features, similar to the ancient
peoples of the Pamirs and Hindu Kush regions. The graves also contain wooden
and stone female figurines, with long, round faces. They have clothes woven
from wool and wear a pointed hat, with long hair falling in plaits over their
shoulders. Their physique presents an interesting study, depicting the physique
and dress of the people of the time. Most of the funerary objects found are arti-
cles of everyday use and include ornaments. In early tombs there is no pottery,
and utensils are made of woven grass, wood, bone or horn. They used wild
hemp and tamarisk branches t o make cups, jars and baskets, utilizing varying
textures t o create patterns. Occasionally the exterior of a basket is daubed red.
Wooden basins, cups and spoons, as well as horn cups, are quite common
in finds. Felt was used widely for clothes and everyone had a felt hat. Woollen
clothes are almost all of plain pattern and are generally coarse, of relatively good
quality. The hides have been tanned, craftsmanship is quite high, and there are
many varieties of attractive leather boots. Beads, strung together and hung
around ankles and neck - some of amber, agate o r jade, but mostly made from
the bones of small animals o r birds - were used as ornaments. Groups of bone
tubes about 10 cm long were sometimes linked together and worn round the
waist. Among the funerary objects there were also large numbers of sheep and
ox bones.
It is easily seen that raising of domestic animals, fishing and hunting were
the principal economic activities of the inhabitants. Their life-style depended
primarily o n their animals but they also made use of local wild ~ l a n t s A
. small
quantity of wheat g a i n was found among the funerary objects but no agricultu-
ral tools have been found and there are no traces of agricultural fields or irriga-
tion works of the period in the vicinity. This clearly suggests that they did not
follow an agricultural economy and probably reccived the grains of wheat in
exchange from neighbouring tribes who were engaged in agriculture.
The Lou-lan sites lie within the ancient territory of Shan-shan, where the
soil is both sandy and saline. In describing the state of Shan-shan, the an-$AM
says that 'the earth is sandy and salty and its fields are few. It has to rely for
grain o n neighbouring states.' Grain was a particularly prccious commodity and
its scarcity, because there was no local production, accounts for the low living
standard of the area. The ornamentation employed, however, points to quite a
developed aesthetic taste in an inhospitable climate.
 The culture of the Xinjiang region

The pebble graves of Alagou (Turfan County)b

Alagou, below the southern slopes of the T'ien Shan mountains near thc
ordos gasslands, many ancient graves have been found. There arc two princi-
pal types, suggesting t w o different cultures. O n e group has a very distinctive
pebbled burial chamber. After a pit was dug, its perimeter walls were lined
pebbles to form a tomb chamber 2 m deep with a diameter of 2 t o 3 m. Tombs
of this type with similar contents have also been found at Lake Ayding-kbl and
subashi in Shan-shan County,' suggesting an cxtension of this culture illto the
Turfan basin. Carbon-14 testings date the graves between the eighth and second
~enturiesB.C. T h e early graves have multiple burials. Each gravc containcd be-
tween ten and twenty bodies of men and women, old and young, piled on top
of each other. They all lay face upwards with the head in a westward direction.
Funerary objects were placed near the head and at the waist. Below the waist
was placed a considerable quantity of bones of sheep, horses, oxen and camels,
apparently intended t o reflect the deceased's material prosperity.
In some instances, hair arrangements were preserved in a recognizable
style. They had all w o r n their hair long, parted centrally, with each half combed
into plaits and kept in place at the back of the head by a bone o r wooden hair-
pin. A delicately woven silk hairnet was then put over the top (Fig. 2). Many
short wooden combs were found in the graves which had presumably been used
for combing out the plaits. There were a great variety of woven materials, much
rather coarse and loosely woven, but some heavy woollen cloth was of a high
standard. Some cloths were plain woven, others interwoven. Light blue, light
red and deep black dyes were employed, and striped, criss-cross and triangular
patterns were used. All this is clear evidence of a developed weaving industry of
a high standard (Fig. 3). In fact several wooden and clay spinning whorls, used
for twisting woollen thread, were found.
The inhabitants clearly practised animal husbandry, but there is also evi-
dence of agriculture. Some pottery jars were found t o contain crop seeds such as
flax. There is also evidence of hunting in the wooden arrow-shafts and three-
edged arrow-heads filed from hard wood.
There were some bronze objects, but most implements for everyday use
were made of w o o d and potterv, often ~ a i n t e dwith a light black pattern o n a
grey-red body - vessels such as jars, dishes, bowls, jugs and cups all being made
by hand. They are decorated with triangular, net, whorl and ~ i n e - n e e d l epat-

6. The excavation w o r k o f the Alagou pebble-grave sites was directed by Wang Binghua o f
the Archaeological Research Institute of the Xinjiang Academy of Social Sciences, and
altogether eightv of these graves wel-e excavated. F o r a brief survey see Archaeological
Research Institute. 1979.
7. The excavation o f thc ancient graves of Lake Ayding-kol and Subashi was undertaken
the Cultural Obsel-vation U n i t of Turfan, Xinjiang. T h e materials have been kept in this
unit and are currently being arranged.
 Ma Yong and Wang Binghua

FIG.2. Silk hairnet. Alagou, Turfan County.

FIG.3. Woollcns, Alagou, Turfan County.

 FIG.4. Painted ceramic jar. Alagou, Turfan Count!

terns (Figs. 4 and 5). Wooden objects including trays, bowls, spoons, cups and
plates with engraved designs (Fig. 6) point to the skill of the artisans. There are a
considerable number of bronze objects, primarily round bronze plates and
small bronze knives; and wooden fire drills are found in virtuall!. ever!- grave. It
must have been the normal method for obtaining fire at the time (Fig. 7).
In craftsmanship the ornaments have great individuality. There are man!-
strings of bcads of bone, agate and jade, and both bronze and gold earrings.
There are also ornaments made of cowrie shell. Some bone ornaments arc
carved with the heads of animals such as wild boars o r bears, in a vigorous st!-le.
Graves of the later period arc rather different. The practice of multiple burials
was replaced by single o r double burials, suggesting a change in social structure.
Graves arc still pebble-lincd, but now contain a wooden bench supported b!.
four pillars. Funcral-v objects now include iron tools and weapons and craft
products such <IS silks (Fig 8), phoenix-pattern embroider!. and lacquer cups,
 FIG.5. Painted ceramic jar. Alagou, Turfan County.

Frc;. 6. Carved wooden plntc. Al'lgou, 'T'urfnn Count! .

 T h e culture of the Xinliang region

FIG.7. Wood-drill fire sticks. Alagou, Turfan County.

FIG.8. F~.agrnenrrof crnbroidered silk. Alagou, Turfan County.

 Ma Yong and Wang Ringhua

imports from China. The proportion of plain pottery (mainly of a red-brown

and grey-brown colour) greatly increased. Some new vessel forms are intro-
duced, particularly the high-footed pottery cup and wooden dish. All thesc
changes show that the inhabitants were increasingly subject to the influence of
the culture and economy of China. The early graves of the Chu-shih people
from the north of Yaerhu in Turfan County have funerary objects, other con-
tents, and pottery basically similar t o those of the Alagou pebble-grave culture,#
and this suggests that the ~ebble-gravesat Alagou also belong t o the Chii-shih.

The wooden chamber graves at Alagou (Turfan County)

N o t far from the pebble-grave sites at Alagou, ancient graves belonging to a dif-
ferent cultural type have been found.9 These graves appear as heaps of stones at
g o u n d level. Beneath there is quite a large vertical pit with a volume of some
200 m'. The pit is filled with sand and piled stones which seem to have been
brought from the nearby river bed. In the centre is a wooden coffin-chamber
constructed of pine, which abounds in T'ien Shan. The wood, cut to size, is
arranged in a criss-cross manner, vertically and horizontally, to line the four
walls of the pit, and the roof is also covered with wood to form a coffin-
chamber. The dead were buried singly or in pairs, always lying flat with the face
upwards. It appears that some sort of red pigment was applied before burial. In
most graves the chamber had caved in due t o the collapse of stones above, when
the wooden roof decayed. The skeletons are generally poorly preserved, but
enough survive t o suggest that they had a well-built physique. Carbon-14 dat-
ing places these graves between the fourth and second centuries B.C. The funer-
ary objects found are rich in quantity and equally refined in quality. Many are
gold luxury items such as beads. There are also round gold plaques, rather thick
and heavy, beaten into a tiger pattern by a hard moulding tool. The tiger's body
curves in a semi-circle in an active and expressive pose. These plaques, originally
fixed in leather, seem to have been used t o decorate belts. There are more than
100 gold-foil flakes, no thicker than a sheet of paper. Some are in the form of
animals, others are shaped as willow leaves, or formed into diamonds or circles.
The animals on these gold-foil flakes may be leaping lions o r a pair of tigers
crouching face t o face. There are holes at the edges and at the two ends to enable
them to be sewn as ornamentation onto clothes. Serving the same purpose were
similar plaques in silver, also pressed into a range of animal designs (Fig. 9).
There was also a remarkable bronze tray consisting of three separate parts - a
square stand, the main body and two lion-like animals standing in the centre of
the tray. The three sections had been cast separately, and then welded together
by pouring liquid bronze on t o the joints (Fig. 10). This unusual type of bronze

8. Huang, 1933.
9. Wang Binghua, 1981, pp. 18-22.
 FIG.. 9. Silver p l a q u e f r o m i \voodcn c11i1nbc1-g r r \ c. Alacou, T u r f a n Count!

tray with ,1nini,11 fiSul-e\ h,ls ~ l s obcen found in ancient g1-<1vesin the IIi \.alle!- in
Xinyuan Count\.. TI1cl.c. n.el-c also ,I few iron kni\-es and an-onrs intended for
domestic us?. T h c s r , ~ n d ~ lo- fd smelting of the gold, copper and iron ohiects is
good m d tllc 111c.t.11,c.\cept f o r objects in sil\.er. is I-elatirel!. pure.
 Mn Yong and Wang Binghua

FIG. 11. Lacquer tray in sztu in a w o o d e n chamber grave. Alagou, Turfan County.

Other luxury goods are agate beads, pearls, silk goods (such as a
diamond-pattern gauze), lacquerware, trays (Fig. 11) and cups which came from
the Yellow River region and provide evidence for trade links with China. Eve-
ryday household utensils include articles in a fine smooth pottery burnished to
a glossy surface. The pottery objects are generally hand-made of fine craftsman-
ship, and have been fired at quite a high temperature. They include bowls, dish-
es, trays and small cups. Some vessels have three flat feet affixed to the base of
the bowl - an unusual feature. These objects are very different from the pottery
vessels recovered from the pebble-gaves in the same locality.
Most of the funerary objects from the wooden chamber graves are house-
hold utensils and superior luxury goods, rarely production tools. It is clear that
the persons buried in these graves must have been the chief nobles, not ordinary
members of the nomad tribes.
The north-south orientation of these graves, the wooden coffin-chambers
and their contents suggest an intimate connection with the Wu-sun culture in
the Ili river basin.I0 Similar graves, also aligned from north to south, have been
found between Zhangye and Tun-huang in the Gansu corridor. The H a n - s h ~

10. The Archaeological Research Institute of the Xinjiang Academy of Social Sciences has
undertaken many excavations of the Ili river valley Wu-sun graves of the Han period,
the successive directors of this work being Yi Manbo, Wang Mingzhe and Wang
Binghua. Parts of the materials have already been published. See ArchaeologicaI Team,
1962; Ma and Wang, 1978, pp. 14-15; Archaeological Research Institute, 1979. The
majority of materials has not yet becn published and is stored in the ~rchaeological
Research Institute of the Xinjiang Academy of Social Sciences.
 The culture of the Xin~iungregion

[hat the Wu-sun tribe originally lived in the western part of the ('ransu cor-
,idor together with the y ~ e h - c h i h ,but escaped to the H s i ~ n wIlen
~ - ~[he
~
h i h their chief. A generation later, the Yijeh-chih were defeated by
y ~ ~ h - ~killed
[he ~ s i u n g - n uand forced t o migrate west to the Ili valley. T o avenge
ancientwrong, the Wu-sun attacked the Yiieh-chih, drove them out and occu-
pied their land. These events took place in the first half of the second century
B,C. ( ~ i a n - s h u61, Biography of Chang Ch'ien; Han-shu 96, Record on the
Western Regions.)
~t may be recorded that the years in which the Wu-sun tribes depended
on the Hsiung-nu, after they had been driven out of their homeland in the west-
ern part of the Gansu corridor by the Yiieh-chih, represented a period of great-
ness for the Hsiung-nu, with their power and influence possibly extending over
the Altai mountains t o the west. This was the time when the Alagou District of
Turfan was under the control of the Hsiung-nu; and it can therefore be sug-
gested that the Wu-sun tribes perhaps lived here under their protection, and the
wooden chamber graves might be attributed t o the Wu-sun. This context is con-
sistent with the Carbon-14 dating, the geographical situation and the period of
time covered by the wooden chamber-grave culture at Alagou.

The T'ien Shan settlements

Below the northern slopes of the T'ien-Shan mountains, where grass grows in
profusion, there is excellent grazing and land suitable for cultivation. Many sites
that were associated with settled cultures have been discovered. In the east, scat-
tered through Tuergan, are a series of sites usually between 1,000 and 2,000 m2,
while the largest is 7,000 m2." T h e remains of houses with thick walls of earthen
bricks and stone built t o resist the severe winters of the region are still visible.
The most interesting objects found are the large number of stone-mill travs. At
Kuisu thirty t o forty were found, several being worn from long use. At a num-
ber of sites painted pottery, stone axes, stone adzes, heavy bronze axes and
bronze knives have been found. These sites lie between Lake Tuer-kol in the
east and Lake Bar-kol (the famous Pu-li Sea of antiquity) to the west. The H o u
Hun-shu says that the state of Pu-li was called after a tribe of that name that
engaged in agriculture, while some of its members led a nomadic life. When the
Hsiung-nu controlled the western region, the King of Pu-li offended the
Hsiung-nu ruler, the Pu-li tribe were expelled and their land was occupied by a
people devoted t o animal husbandry w h o knew nothing of agriculture. It may
therefore be suggested that the sites of the ancient settled culture t o the east of
Lake Bar-ko1 may belong t o the early period of the Pu-li tribe.

11. For relevant material, see Li, 1959; Wu, 1964; and Archaeological Research Institute,
1979.
 Ma Yong and Wang Binghua

The ancient village site of Sidaogou, 10 km west of Mulei,lL has tracts of

habitation in the form of post-holes, kitchen ranges and ash pits. Carbon-lq
dating suggests an early occupation of around 1000 B.C. and a later one fronI
around the fourth century B.C. There are few funerary objects from the asso-
ciated graves, but the habitation site finds included pottery, stone, bone and
bronze articles. From the later period new articles have been found such as
cups, cauldrons, pallets and container vessels, and the proportion of painted
pottery increased, with vermilion colouring making its appearance. Most tools
are of stone in good craftsmanship - mill quarries, mortars and pestles, spindle
whorls, hoes, adzes, etc. There are bone spindle whorls, needles, combs and
arrows, as well as bronze knives, hairpins, rings and ornaments. The bones of
horses, oxen, sheep and dogs suggest the involvement of these animals in the
economic life of the time.
It is clear that the people living in these areas led a sedentary life and were
engaged in agriculture and domestic animal husbandry, with hunting as a sup-
plementary activity. The appearance of terracotta knife moulds suggests that
they had already g a s p e d the art of casting bronze objects, while their painted
pottery resembles that of the Shajing culture to the west of Gansu.

Grave sites on the slopes of the Altai mountains

The zone between the Altai mountains and the Dzungarian basin provides ideal
grazing throughout the year. The nomadic peoples w h o lived there have left
behind them many rock carvings of large-horned sheep, horses and deer, and
men with taut bows and flying arrows. But it is difficult to suggest either the
time or the people responsible for them.
Three principal types of grave have been found in the zone. The first type
is the large stone tumulus found near Huahaizi in Qinghe Province. The largest
example of this type of tumulus has a circumference of 230 m, encircled by a
ring of stones, with a square polished granite stela 300 m to the north. The stela
has carved running deer engraved on two sides. A second stela 10 m to the east
has a carved circular and rhomboid pattern, with the carving of a lamb and a
short sword. N o graves of this group have yet been excavated.')
The second type is the 'stone-warrior' grave, which has an engraved figure
of a man standing in front, sometimes represented by a human face carved on
the upper part of a large boulder. Some have merely a stone bar symbolizing the
human form. Behind the stone warrior there is a square stone coffin made from

12. Yang, 1982.

13. This type of large stonc mound was survcycd in 1965 by the archaeological team of the
Institute of Nationalities of the Xinjiang branch of the Chincsc Acadclny of Sciences.
F o r materials from the former Soviet U n i o n and Mongolia, see Savi~iovand Chlenova,
1978.
~ l c ; .12. Stone cup with handle in shape of c o n ' s head from n \ t o n c - \ \ , ~ ~ - rg1.11
~ o r c.
Kccrrnuqi, A l t a ~Count),.

Flc.. 13. Coupled 5tonc cup5 f r o m 1' \ronc-\\.drrlol T - ~ I C .h c c r m u q ~ .~ I t l C

l ount)
 Ma Yong and Wang Binghua

FIG.14. S~nallstone figurinc f r o m a stonc-w,lrrior gra\ c. Kccsmuqi, Altni County,

 The culture of the Xinjiung region

four enormous slabs of granite, capped by another slab. Inside the coffin the
bones were found in complete disorder. There were multiple burials; and the
objects found include stone arrows, stone CUPS, stone jars and pottery vcsscls
with some bronze and iron objects (Figs. 12-14).lJ
The third type is the earth-pit vertical shaft grave, a type often discovered
i, the same region as the stone-warrior graves, but there is no evidence for any
between the two.
The three types of grave clearly belong to three different cultures, but the
absence of full-scale excavations precludes any detailed discussion of their
and historical backgrounds; also the absence of written records
makes it impossible t o give any clear answers to questions of ethnic identity.
such a broad study of the different grave-culture peoples of Xinjiang would be
incomplete without reference t o classical Greek literature, which mentions
some of the tribes w h o might be associated with these areas. Literary works in
both China and the West seem t o focus primarily on the area to the north of
Xinjiang - the slopes of the Altai mountains and the grasslands north of the
Dzungarian Desert. This is probably because between the seventh and second
centuries B.C. the principal route across Eurasia ran north-west from the Inner
Mongolian grasslands near Hetao over the Altai mountains along the Irtish
river. Having crossed the south Siberian grasslands, it again went west t o reach
the land of the Scythians o n the northern shores of the Black Sea, as
archaeological finds seem t o confirm. The evidence from the epic Arimaspea,
referring t o the Issedones as neighbours of the Massagetae (Strabo XV.1.6),
speaks for their nomadic identity but it is difficult to identify them with the
tribes noted in the Chinese annals. Some Chinese scholars have identified the
geographical location of the terrin)l-y of the Issedones with the Yiieh-chih or
Wu-sun on the upper reaches of the Ili river and in Xinjiang, but this remains
uncertain. What is clear is that between the sixth and fourth centuries B.C. there
was a powerful confederacy of nomadic tribes under the name of Yiieh-chih liv-
ing on the steppes t o the south of the Altai mountains; and the graves excavated
in different areas of Xinjiang confirm the existence of several nomadic groups
and throw light o n some of their cultural and ~ o l i t i c a lrelations.

14. Wang Mingzhe, 1981, pp. 23-32.

 Ma Yong and Sun Yutang

The triangular clash

The second century B.C. had a profound influence on the history of Central
Asia.
This was the century in which the clash between two great powers, the
Hsiung-nu and the Han, for the mastery of the Western Regions took place.
The Hsiung-nu people originally lived in the northern part of present-day
Shaanxi in China. During the fifth and fourth centuries B.c., faced with the
northward expansion of the states of Ch'in and Chao, they were forced to
migrate to the grasslands of Inner Mongolia, where they became a strong tribal
confederation, and subsequently controlled many of the nomad peoples of
Mongolia. Early in the second century B.C. the Hsiung-nu drove out the Yiieh-
chih and took possession of their land. At that time there were several com-
munities in the oases round the Tarim basin leading a sedentary life, occupied in
agriculture, horticulture and animal husbandry. In the Turfan basin to the west
of Lop N o r and along the northern foot of the Kunlun and Altai ranges were
scattered communities carrying on a semi-agricultural and semi-nomadic way of
life. These groups had organized themselves around urban centres to form a
series of petty kingdoms. The Han historians refer to these as the 'Walled City-
States of the Western Regions'. There were some thirty-six of them, the more
important being Chii-shih (Turfan), Lou-lan, Yen-ch'i (Qarashahr), Ch'iu-tzti
(Kucha), Ku-mo (Aksu), Shu-16 (Kashgar), So-chii (Yarkand), Yii-t'ien (Kho-
tan) and Chii-mi (Khema). Because of the great distances between them they
were not integrated into one strong ~oliticalunit, and were unable to offer any
effective resistance t o the Hsiung-nu.
Under Mao-tun, the Hsiung-nu consolidated their ~ o s i t i o n .The land

:' See Map 6.

 Ma Yong and Sun Yutang

west of the Altai mountains, with the 'Walled City-States', was ruled by [he
Prince of Li-jhu. The Hsiung-nu under him were nomads living on the grassy
plain around Lake Pu-li (modern Lake Bar-kol). A resident with the title 'gen-
eral-in-charge of slaves' was installed to keep watch on the petty Walled City-
States, with the duty of levying taxes from them. The Hsiung-nu clearly con-
sidered all the peoples living in the Walled City-States as slaves, but left their
original political organization undisturbed. N o doubt their princes had been
sent as hostages to the Hsiung-nu. Taxes to the Hsiung-nu would naturally be
paid in agricultural produce such as grain and fruit. The tribal confederation of
the Hsiung-nu was prepared to challenge the H a n in China, which suffered
from political instability following the tyrannical rule of the Ch'in and the
widespread civil war. The Chinese tried t o appease the Hsiung-nu through
intermediaries, by paying indemnity and providing gifts for several decades,
Eventually, after forty years of peace during the reigns of Wen-ti and Ching-ti,
the Han had sufficient economic and military power to confront the
Hsiung-nu. Emperor Wu-ti, w h o ascended the throne in 140 B.c., proposed
joint action with the Yueh-chih against their common enemy, the Hsiung-nu,
and sent a mission under Chang Ch'ien to the Yueh-chih, in the Ili river basin,
seeking an alliance. About 117 B.C. Chang Ch'ien was again sent on a mission to
the Wu-sun, and dispatched deputy envoys to the countries of Ta-yuan, K'ang-
chii, Yueh-chih and Ta-hsia. The information he collected was recorded by the
great historian Szii-ma Ch'ien in the Shih-chi, the earliest reliable Chinese liter-
ary source for the history of Central Asia.
After Chang Ch'ien's first mission there was a long struggle for control of
the Western Regions. In 121 B.C. one of the Hsiung-nu leading nobles ruling
over the Gansu corridor surrendered to the Han, who set up the four ~refec-
tures of Wu-wei, Chang-i, Chiu-ch'uan and Tun-huang, the first step towards
extending H a n power over the Western Regions. After this the oasis kingdoms
of the Tarim basin, previously dependants of the Hsiung-nu, successfully trans-
ferred their allegiance t o the Han. The campaign by the Hail general Li
Kuang-li against Ta-yuan (part of modern Ferghana) in 101 B.C. further
increased the political prestige and influence of the Han.
The clash between the H a n and the Hsiung-nu now focused on Chii-shih
(the modern Turfan basin), an area of great strategic importance. During the
first half of the first century B.C. internal rivalries greatly weakened Hsiung-nu
power. In 60 B.c., the Hsiung-nu prince Jih-chu surrendered to the Han; the
Hsiung-nu post of 'general-in-charge' was abolished. The eventual victor in the
internal rivalries, s h a n - ~ i iH ~ - h a n - ~ e surrendered
h, to the H a n in 54 B.C. After
this event the thirty-six states of the Western Regions came under the direct rule
of the Han. They invested Ch'eng Chi as the first hsi-yii tu-hu (protector-gen-
era1 of the Western Regions), that is, the highest civil and military official in
charge of the area. From then until the end of the reign of Wang Mang (c. A.D.
23), there was an unbroken succession of protectors-general, eighteen in all, oi
 The Western Regions under the Hsiung-nu and thc Hnn

whom ten names have come d o w n t o US.'The last protector-general, Li Chung,

died at ~ u c h aand his personal seal has been discovered in Shaya County,
~injiang.~

China and the Western Regions

The large-scale peasant uprising in the last years of Wang Mang's reign and the
ensuing civil war left the Eastern H a n dynasty too exhausted t o regain control
over the Western Regions. After a fierce war between the petty kingdoms, the
state of So-chii (Yarkand) gained hegemony over the area for a few years. The
southern Hsiung-nu remained subject t o the Han, and stayed within the north-
ern prefectures; but the Western Hsiung-nu invaded the Western Regions, and
the Walled City-States of the Western Regions fell once more into their hands.
In the first years of the Eastern Han period, three fierce battles were
fought for the control of the Western Regions. Pan Ch'ao was the best-known
Han general. His military career of thirty-one years (A.D. 73-102) was crowned
with success, and he was popular with the people of the oasis states. H e success-
fully averted the invasion of the army sent by a Kushan king who attempted t o
interfere in the affairs of the Walled City-States, and sent his envoy, Kan Ying,
on a mission t o Tiao-chih (the Persian Gulf). Kan Ying was, however, prevented
from completing his journey by the Parthians.
In A.D. 126 Pan Ch'ao's son, Pan Yung, again defeated the Northern
Hsiung-nu and consolidated H a n rule in the Western Regions. Instead of the
protector-general the H a n court now appointed a hsi-yii chang-shih (secretary-
general), a post that was retained until the end of the reign of Ling-ti (c. A.D.
188). At least eight of these officials are recorded in documentary sources.
The protracted civil war in China towards the end of the Eastern H a n
severed links with the Western Regions, but the Northern Hsiung-nu also
declined. O n e part migrated t o the west, another part was dispersed over
Gansu, Shaanxi and Inner Mongolia and, like the Southern Hsiung-nu, began t o
merge with the local inhabitants of China. Meanwhile, a new nomadic confeder-
ation, the Hsien-pi, took over the Mongolian plateau and the lands west of the
Altai mountains. During the first half of the third century A.D., China proper
was divided into three independent states, the Wei, Shu and Wu. Both the Wei
and the Shu re-established relations with the Western Regions, the Wei reviving
the post of ~ e c r e t a r ~ - ~ e n e rAa lsubstantial
. number of wooden tablets with Chi-
nese inscriptions belonging t o the Wei and Ch'in periods have been found in the

1. Huang Wenbi, 1981, pp. 310-14.

2. Huang Wenbi, 1958, p. 1 13. The seal inscription is ' L i Cho?fg shi yi71 xin '. Huang
inexplicably rendered '.yinV as two separate characters and hence failed in its
decipherment. Sec MA, 1975, p. 29.
 Ma Yong and Sun Yutnng

ruins of Lou-lan north-west of Lop Nor.' They are official documellts

archives from the office of the secretary-general, governor of the Western
Regions.

The economy of the oasis states

The economic base of these oasis states was farming and horticulture, with irri-
gation and animal husbandry. Numerous dwelling sites have been discovered
recently in the Tarim basin dating from the Han, the most famous being at Niya
in the north of Minfeng County. The remains of houses are spread along both
banks of the now dry Niya river, covering an area of 10 km (south to north) by
25 km (east t o west). The houses had floors made from a mixture of wheat-
straw, cow-dung and mud; their walls were made of wattle woven with tama-
risk twigs daubed with mud, and fitted with fireplaces. Around the settlement
were extensive orchards. This site and the economic life it reflects are represen-
tative of the oasis states in this area (Fig. I ) . ~In laces where the land was poor
or sandy, as in Shan-shan, and it was impossible t o rely upon agriculture, people
reared animals and obtained cereals from the agricultural oases by barter.
The most outstanding handicraft product of the Western Regions under
H a n rule was probably textiles. Because herding was widespread, skins and
woollen garments were used for protection. Many pieces of woollen cloth,
brightly coloured and beautifully figured, have been found in ruins of the Han
~ e r i o d Fragments
. of woollen cloth with remarkably beautiful designs of men,
animals, vines, tortoiseshell and ~ e t a patterns
l (Fig. 2) were found in a husband-
and-wife tomb of the Eastern H a n ~ e r i o din Minfeng County. Tapestries from

FIG, 1. Ruins of ii v i f l ; ~Niya.

~~,

3. Luo and Wang, 1914; Huang Lic, 1981, pp. 58-63.

4. Shi, 1962.
 FIG.2. Woollen cloth with figurcs of mcn, animals and grapes.

the Western Regions used f o r bed coverings, carpets and hangings were held in
high esteem by the H a n people.' Clothing was also made from linen, cotton and
silk. Flax was extensively cultivated in the Western Regions and had quite a
long history. C o t t o n perhaps first entered the area from India. T w o pieces of
dyed cotton cloth of this period have been found at Niya (Fig. 3). Using a white
ground, the cloth is d y e d with wax and indigo t o form beautiful designs com-
parable in quality t o modern products using the same technique in China today.
One beautifully designed piece of dyed cotton has a head (perhaps of Buddha)
surrounded with a halo and the figure of a dragon. In t e r m of decorative st!.le
it has strong Indian characteristics, but the dragon is certainly Chinese in con-
ception. I t is n o t clear whether this material was imported o r a local product,
but it does s h o w that the inhabitants of the Western Regions were using I ~ i g h -
quality batik-dved cotton clothes during the Eastern H a n period. There is n o
doubt about the Chinese origin o f silks. T h e Emperor Wu-ti and Chang Ch'ien
had opened u p the Silk Route, and the H a n court regularl!. presented gifts of all
sorts of silk. M a n v pieces o f H a n silk have been found in Xinjiang in the \Vu-
sun tombs at Zhaosu , ~ n din the H a n sites at L o p N o r . A m o n g the best-pre-
served pieces o f silk are those f r o m the t o m b of a husband and wife at hllinfc~~g.
There is hard117 any iten1 o f clothing in the t o m b that is not of finel!, wtnren silk.
with quality elnbroidcr\r. O r ~ erobe is decorated with Chinese ch'~racterswish-
ing 'Good luck f o r 10,000 while a 'cron:i~lg-cock' pillow suggests
the idea of rising c ~ r l i~l l .the 1110l-nil1g.O t l l r r b c a u t i f u l l ~n-o1.m
~ Chinese chd-
 Ma Yong and Sun Yutang

FIG.3. Wax-resist dyed cotton cloth.

racters wish 'Long life and good fortune to you and your descendants'. They
were obviously in daily use in China proper, but the narrow sleeves of the bro-
cade robe suggest a local fashion that was possibly tailored locally (Fig. 4).6
The Western Regions had already developed mining and metal casting
techniques prior to the Han dynasty. The Han-shu description of the Shou-
ch'iang states that 'there is iron ore in the hills; and they produce their own
arms - bows, arrows, long knives, swords and armour'. The natives of Shan-
shan were also skilled at making arms and in the states west of Ch'ieh-mo 'the
arms made were like those of the Han'. So-chii (Yarkand) 'has mountains con-
taining iron ores' and in Ch'iu-tzti (Kucha) workmen were 'skilled at casting
and had reserves of lead'. It is clear that iron-casting and steel-making were
fairly widespread among the Walled City-States of the Western Regions, and
were used in making both tools and weapons.
Mining sites of the H a n period have been discovered at A'a Shan, Minfeng
and AqikO Shan. From A'a Shan the finds include crucibles, iron slag, ore and
pottery like the bellows air pipe found in Shaanxi (Fig. 5). At Niya and else-
where ore, sintered iron, slag, stone chisels, whetstones and fragments of iron
shovels have been found; and houses there contained iron adzes and sickles
with wooden handles. A copper-mining site has been located at ~a-ke-ma-ke.
'Wu-shu' coins of the Han period have been found at all these mining sites, sug-

6. Ma, 1975, pp. 29-30.

 7 . 1 7 ~ Western Regions under the N s i ~ n g - n uand fhr) I f a n

FIG.4. Brocade robe.

gesting that they were established after the beginning of H a n rule in the West-
ern Regions.' Froln thc Wu-sun t o m b site in Zhaosu C o u n t y dating f r o m
Han times comes an iron ploughshare, and tool marks found o n a wooden outer
coffin show that iron tools were in use."n the same context was found a strik-
ingly beautiful gold ring set with precious stones, but it is not yet ~ o s s i b l et o
determine whetliel- it was imported o r made locall!-.

7. Ibid., pp. 29-30.

8. Ma and Wang, 1978, pp. 11-12.
 Ma Yong and Sun Yutang

rr,,,.

FIG.6. Toilet box and toiletries.

The exchange of commodities between the Western Regions and the

heartland of China expanded considerably during the H a n period. At a number
of H a n sites in Xinjiang lacquerware, woven rattanware and bronze mirrors
have been discovered, all imported from China proper (Fig. 6). It is especially
worthy of mention that linen-made paper was discovered among Western Han
remains at Lop N o r and paper of a similar kind has been found in Eastern Han
tombs in Minfeng County.' Paper is one of the most striking inventions of
China, dating back t o the Western H a n period. Paper appeared in the Western
Regions soon after it was first invented in China, showing the speed at which
commodities were exchanged between the two regions. China imported from
the Western Regions fine breeds of horses, grapes, alfalfa and, of course, jade,
the famous product of Khotan. The name Yii-men Kuan (Jade Gate Pass) may
have been given by the Chinese merchants who imported jade from the Western
Regions. Some post-Han sources mention Chinese imports of hu paper (flax),
hu gourds, hu peaches and hu stallions. The term 'hu' means 'Western barbar-
ian' and points to the area from which they came. Fine woollen and cotton tex-
tiles from the Western Regions also flowed into China in large quantities.
There is hardly one Han site in Xinjiang that has failed to yield a great
number of H a n coins, including varieties of all ~ e r i o d s ,clear evidence that
under Han rule the coinage of China proper also circulated in the Western
Regions (Fig. 7). The Walled City-States did not issue their own coins. The only
exception was Khotan, which minted a small quantity of Sino-Kharosyhi bilin-
gual copper coins. They are round but, unlike Chinese cash, have no hole in the

9. Huang Wenbi, 1948, p. 168, Plates 23-5.

 The Westcrn Regions under the I-Jsiung-nu and thr /Inn

FIG.7. H a n 'Wu Shu' coins.

FIG.8. Sino-Kharosyhi bilingual coins.

centre or trimmed line round the rim. The obverse has a legend in Chinese seal
script indicating the value of the coin. O n the reverse is a horse o r camel sur-
rounded by a Prakrit legend in Kharosfhi script, giving the name and titles of
the king. T h e coins were struck in two denominations, the larger with the
legend 'bronze cash weight 24 rhu', the smaller with '6 shu cash'. Most of these
coins were found at the oasis of Khotan and were early recognized as issues of
the ancient state of Khotan. They were struck between A.D. 152 and 18(3, that is,
in the last few decades of the Eastern H a n period (Fig. 8), apparently not for
economic reasons so much as for prestige. Consequently the quantity issued
was small and circulation was limited. The use of Kharosfhi script shows the
influence of Kushan culture, but the honorific titles of the King of Khotan on
the coins are on an equal footing with those of the Kushan kings, making it
clear that at that time Khotan was not under Kushan rule. The fact that the
legend indicating the value of the coin and the unit of value is in Chinese shows
the strength of Chinese influence.
 Ma Yong and Sun Yutnng

Language, culture and religion

These Central Asian kingdoms had no script of their own; it seems that Chinese
was well known, though the large number of K h a r o ~ i h documents
i found Sug-
gests that Chinese may have been used only for official purposes. In a number
of Western Region states there were 'directors of interpreters', officials in
charge of the translators, who had full knowledge of spoken, and probably also
of written, Chinese because they were obliged to translate Chinese government
documents. Some of the wooden tablets inscribed in Chinese discovered at
Niya in Minfeng County are letters of the local nobility. The inscriptions on the
tablets were in classical Chinese characters with very fine calligraphy, reflecting
an excellent command of Chinese, which must have been their chief means of
communication (Fig. 9). During the last decades of the Eastern Han period, as a
result of Kushan influence, KharoSthi script seems to have dominated the West-
ern Regions; the Sino-Kharosthi bilingual coins struck by the King of Khotan
are perhaps the earliest instance of this.I0 Later, KharoSthi script reached the
state of Shan-shan," becoming the official script there for a period in the third
and fourth centuries A . D . ' ~

F I G . 9. Wooden tablets with Chinese text.

10. Ma, 1 9 8 3 ~Xia

; Nai, 1962.
1 1. Luo and Wang, 1914; Huang Lie, 1981, Tables 28-3 1 .
12. Ma, 1980b.
 T ~ LWestern
. Regions under the Hsiung-nu and thc fian

Besides using the Chinese script, some rulers of the oasis kingdoms also
,,pied the H a n court in other ways. Chiang-pin, King of Kucha during the late
Western Han, took instructions of the H a n court as his model in building his
palace, setting his guards of honour and adopting the ringing of bells and beat-
ing of drums f o r court ceremonies. Yen, King of So-chli, also and
the ceremonial rules of the H a n court.13
Our knowledge of burial customs is limited, except for Zhaosu County,
where a number of WU-sun graves have been excavated recently. The Wu-sun
with domed tumuli are found in groups, arranged in a row from north to
South. Most contain wooden outer coffins, with inner walls decorated with felt
hangings. Most are multi-chambered, with traces of wooden coffins, but some-
times the corpses were only wrapped in felt. The burials are in the extended
supine position, with the head pointing west. Most tomb furniture is poor and
simple, there being pottery vessels and iron objects such as small awls. In some
large tombs more luxurious grave furniture has been found, including such
beautiful objects as the gold ring discussed above.I4 Near the Wu-sun area, a
joint husband-and-wife burial has been found among the graves of the Eastern
Han period at Niya, with their box-like wooden coffin standing o n four feet,
placed in a rectangular shaft with mid-coated tamarisk wattle; t o judge from the
covering and clothing of the deceased, these would seem t o be the graves of the
local higher nobility.I5 Earlier it was suggested that the boat-shaped wooden-
coffin burials found in the L o p N o r area were the graves of the Lou-Ian people
during the H a n dynasty, but similar remains have now been found on the lower
reaches of the Kongque river, dated by Carbon-14 to around 1000 B.c., so the
so-called 'Lou-lan graves' can n o longer be ascribed to the Han period.
Many scholars believe that Buddhism was brought into China at the end
of the Western o r the beginning of the Eastern H a n ~ e r i o d hence
, the Western
Regions must have received Buddhism even earlier. The supposition is that if
Buddhism spread north-eastwards from India, it must naturally have been
accepted first by the peoples of the Western Regions before reaching central
China. The Biogvaphy of Pan Ch'ao, in the H o u Han-shu, however, mentions
that when he arrived in Khotan in A.D. 73, the local people believed in Shaman-
ism and a shaman tried t o kill his horse to practise sacrifices and incantations.
This could not have been Buddhism, and must have been a local religion prac-
tised among the oasis states. A greater part of the chapter on the Western
Regions in the H o u Han-shu was copied from Pan Yung's original written
record of the Western Regions, which was completed by the end of the reign of
Emperor An-ti, that is, around A.D. 124. Pan Yung spent the whole of his ~ o u t h

13. For Killg Chiang-pin of Ch'iu-tzG, see Han-shu 96; for King Yen of So-chii, see H o u
Han-shu, chapter on the Western Regions.
14. Archaeological Research Institute, 1979.
15. Wenwu, 1960.
 Ma Yong and Sun Yutang

in the Western Regions and was familiar with the local custorns. H e was also inter-
ested in the question of Buddhism. But it was only when he spoke about the state
of T'ien-chu (India) that he commented that the country 'practises the Buddhist
Way and does not take life, which has now become the order of the day'. From his
records we can find hardly any trace of Buddhism in the Western Regions under
Han rule. Moreover, from an art-historical point of view none of the Buddhist
caves with murals in Xinjiang can be dated before the beginning of the Eastern
Han. It may therefore be suggested that Buddhism reached the Tarim basin around
the middle of the second century A.D. There is no doubt that Buddhism reached
China proper a little earlier. This may be explained in two ways. The first hypothe-
sis is that the people who first brought Buddhism t o China were Buddhists from
the Kushan Empire. Although they took the road through the Western Regions,
the object of their mission was the H a n court. The introduction of Buddh'ism was
therefore not by a gadual expansion across the Western Regions as some scholars
envisaged. The alternative is that the spread of Buddhism into China during the
Eastern Han came another way, perhaps by the sea route.

Chinese administration
The Han-shu chapter on the Western Regions gives a fairly clear description of
the sphere of jurisdiction of the H a n protector-general. Wherever a state did not
come under his control, it was always indicated that this particular state 'was
not subject to the protector-general'. The official residence of the protector-
general was located at Wu-li (east of Lun-t'ai County, Xinjiang).
Among the inhabitants ruled by the protector-general was, first, the so-
called 'Tocharian' group.16 It is difficult to suggest a definite name for their lan-
guage, which is, however, named after the ethnic group using it. Spoken by the
Chii-shih, Yen-ch'i, Ch'iu-tzii and Lou-lan (Shan-shan) peoples, it was Indo-
European, though its relation with other languages of the family remains
unclear. We can distinguish three dialects: (a) the Kuchean dialect, spoken by
the Chii-shih and Ch'iu-tzii peoples (Tocharian B); (b) the Yen-ch'i dialect,
spoken by the Yen-ch'i people (Tocharian A); and (c) the Lou-lan dialect, spo-
ken by the Lou-lan This Tocharian group as a whole settled along the

16. T h e dispute a b o u t t h e term 'Tocharian language' is well k n o w n internationally so it is

unnecessary t o discuss it further here. I n recent years most Chinese scholars havc dis-
carded this term a n d used the 'Yen-ch'i-Ch'iu-tzii language' in its place. Nevertheless,
this language should also include the dialects of Chii-shih and Lou-Ian (Sh,~n-shan)area.
Accordingly, the term 'Yen-ch'i-Ch'iu-tzii language1 is n o t really appropriate. There-
fore, f o r o u r present purpose w e shall preserve the old usage and call it 'Tocharian'.
17. T h e K h a r o s ~ hdocuments
i found in the Kingdom of Shan-shan in Xinjiang are writtcn in
Giindhari Prakrit, b u t s o m e words borrowed f r o m the local langungc rcflcct the Lou-Ian
dialect, which should c o m e undcr the so-called 'Tochasian' categc;ry. See Burrow, 1937.
 The Western Regions under the Hsiung-nu and [he /Ian

nol-thernfringes of the Tarim basin and to its east in the Turfanbasin as as

the vicinity of Lop Nor. T h e second was the Ch'iang language group, of tribes
settledalong the northern foot of the Altyn, Kunlun and Karakorurn ranges, the
major tribe being the Yiieh-ch'iang. The Hsi-yeh, Pu-li, Yin-ai and Mo-lu-ti
Peoples, who lived o n the south-east of the Pamir plateau, also belonged to the
~ h ' i a n ggroup, but they may also have been intermingled with the Sakas.lxThe
language of the Ch'iang doubtless belonged to the Sino-Tibetan language
family.They had close relations with the Ku-yang people in the ~ ~ i - I moun- an
tain area and resent-day Qinghai Province. The third was the Saka language
group, spread over a broad area from south of Lake Balkhash to the south of
the Pamir ~ l a t e a u including
, such places as Shu-li, Yii-t'ien, etc., in the western
part of the Tarim basin. Except for the Wu-sun, who were mainly nomadic
herdsmen, most of the population had formed themselves into the Walled City-
States by the second century B.c., with agriculture, horticulture and animal hus-
bandry as the basis of their economy. Most of the city-states were very small.
According t o the population figures in the Han-shu, the largest was Ch'iu-tzij
(Kucha) with a population of 81,317 and the smallest, Shan-huan, with only
194. It is clear that the kings of these oasis states were only rulers of a town or
large village. For a time under the Eastern H a n the thirty-six Walled City-States
were divided further and became fifty-five. The H o u Han-shu records the
population of some of these states at a later date, but in far less detail. C o m -
paring the H o u Han-shu and the Han-shu we can see that the population in the
Eastern H a n period was about twice as much as in the Western H a n period, but
the total was still very small.
Under the rule of the protector-general and secretary-general of the West-
ern Regions, all the kings of these states, large and small, were subject t o the cen-
tral government of the Han, as previously they had been subject t o the central
government of the Hsiung-nu. They had t o send 'attendant sons' t o the H a n
court. Their royal status, and the status of certain of their officials, was dependent
on Han approval. According t o the Han-shu, the number of kings, aristocrats
and ministers of the Walled City-States w h o had received seals of authority from
the Han court exceeded 376 (Fig. 10). In certain states, the Han also set u p new
official posts and sometimes even appointed H a n ~ e o p l eas officials."
However, H a n rule in the Western Regions differed from that of the

18. The Han-shu 96 alleged that the H ~ i - ~ e Pu-li,

h, Yin-ai and Mo-lu-ti were of the same
nationality, 'differing f r o m the H u but similar t o the Ch'iang and Ti'. ' H u ' here seems t o
belong t o the east Iranian group. However, the archaeological remains w e find in this
area rcvealed that their culture had strong elements of the Sai (Saka) culture. (See C h a p -
ter 9 o f the present volume.) T h e Hun-shu notes that they were 'similar' to the Ch'iang
o r Ti, which shows that they were not of pure Ch'iang descent.
19. For example, 'marquis t o attack the Hu' and similar posts are obviously an-conferred
official titles. See Hm~-shu96.
 Mn Yong 'znd Sun Yulang

FIG.10. Official seal of the ' H a n Kuei I Ch'iang Chang'.

~ s i u n g - n u It
. only required the subject states to provide part of the military
force needed to withstand the Hsiung-nu in time of war, and t o accept responsi-
bility for protecting communication lines along the Silk Route during time of
peace.

Military and agricultural colonies

It was not enough t o rely exclusively on the military strength of these small
states for defence against Hsiung-nu incursions, therefore the Han government
stationed some armed forces in the Western Regions. T o avoid problems in sup-
plying remote areas, and transporting provisions across the Gobi Desert, the
t'un-t'ien (military agricultural colony) policy was adopted: the garrison troops
were required t o cultivate land on the spot. As early as the reign of Emperor
Wu-ti, t'un-t'ien colonies were set u p at Lun-t'ai and Ch'ii-li with an initial
complement of several hundred farming conscripts led by envoy-commanders.
During the reigns of Emperors Chao-ti and Hsiian-ti there were also t'un-t'ien
colonies at I-hsun City (east of modern Ruoqiang County), Chii-shih (west of
modern Turfan County), Chih-ku (near the Issik-ko1 in Kyrgyzstan) and on the
north-west bank of Lop Nor. The scale of the individual colonics increased and
the number of garrison conscripts of Lun-t'ai alone rose t o 1,500. During East-
ern Han rule there was a t'un-t'ien colony at I-wu (in the area of present-day
Hami) and a 'commander-in-charge of cereals' was posted there.1° Liu-jong in
the east of the Turfan basin was the long-term central location for the garrison
troops and farming colonies. These troops were under the command of an
officer of higher rank called the Wu-chi commandant.?' The site of his head-
quarters was called Kao-ch'ang-pi, this being thc origin for the use of the name
Kao-ch'ang (Qocho) for the Turfan basin as n whole. During the latter days of
the Eastern Han, the Wu-chi commandant had become the highest commanding

20. Ma and Wang, 1978, pp. 1 1-12.

21. Hou, 1980.
 The Westcm Regions undcr the H$iung-nu and the Ifan

,fficcr of the Western Regions garrisons, comparable to the secrrtary-gcnural,

who held authority over the inhabitants of the Western Regions; duties were
divided between the two.
Vestiges of the t'un-t'im cokmics of the Han period arc still to be seen i n
~U"-t'ai, Shaya and Ruoqiang Counties, and the region close to 1 . o ~Nor,
where traces of ancient irrigation channels and field ridges can be recoh~ n l z c don'

[he red clay banks of the Qizil river. From the Tu-yin site on the north bank of
Nor large numbers of wooden tablets (Fig. 1 I ) have been found, dating
from the second half of the first century B.c., lnostly official documents of the
thn-t'ien troops, reflecting their organization and original encampments, agri-
cultural products, tools used, methods of cultivation, granaries and their daily
life generally.22F r o m these tablets w e learn that these t'un-t'ien soldiers came
from all over China's inner prefectures, bringing their families with them and
living there for long periods of time, engaging on the one hand in agricultural
labour, and o n the other hand in fulfilling the task of military defencc. East of
LOP Nor and in the Turfan basin were important military granaries for storing
provisions. These t'un-t'ien districts gradually developed into fixed Chinese set-
tlements in the Western region^.^'

FIG. 1 I. Woodcn tablets relating t o a H a n t'un-t'len colon?.

2.2. For the Lou-Jan sires, see Huang Wmbi, n.d.B, pp. 1 8 1 4 . For the latest inre~tigationsof
the Lou-Ian sites, see Hou, 1981. For the Tu-!in site, see Huang Wcnbi, 1948, pp. 105-9-
As to the rern'~insof the ~ncielltcity of Kao-&'ang (Qocho), reports and materials are
abundant and so well documcl~tcdthat there is no rleccl to cite them hcrc.
23. Ma, 1975, pp. 27-30.
 Mu Yong and Sun Yutung

Closely related to the t'un-t'ien agricultural colonies was the work of irri-
gation. The Western Regions comprise a wide expanse of arid land
extremely low rainfall, where agriculture depends wholly o n irrigation channels
fed yearly by the melting snow. The construction of artificial irrigation systems is
absolutely essential. Long before the H a n dynasty, the local inhabitants lnust
have constructed some irrigation channels, but the Han introduced the tJun-t'ien
policy; there was a remarkable increase in irrigation construction due to the
adoption of advanced technology from China proper. In Shaya County there are
remains of an ancient H a n irrigation channel more than 100 km long. Closeto
this channel H a n coins and vestiges of cultivation were found. A Han irrigation
system has also been discovered at Miran, following the course of the ancient
Miran river, where main floodgates, bifurcation gates, trunk and branch canals
were constructed. H a n tombs and other objects have also been found near by.!4
Nowadays in the Turfan basin the most famous and remarkable type of irrigation
system is the so-called kahrez, consisting of an underground channel descending
the hill slope, which conducts meltwater on to the farmland. O n the surface,
walls are sunk at intervals, so that the underground channel passes through them.
Some scholars think that the kahrez system was introduced from Iran during the
Ch'in dynasty; on the other hand, a Han-shu chapter o n irrigation records that
during the H a n dynasty there were already 'irrigation channels with wells' in
Shaanxi, of a type found in the kahrez system. Perhaps we may therefore pres-
ume that the kahrez system of an 'irrigation channel with wells' was introduced
there by the t'un-t'ien agricultural colonies during the Western Han period.25
Another important measure implemented by the garrison troops in the
Western Regions during the H a n dynasty was the construction of a system of
fortifications and beacon towers - a warning system in case of a Hsiung-nu
attack. Each fort consisted of a small square encampment of rammed earth, sur-
rounded by walls, within which stood a small building. In one corner was a
rammed-earth watch-tower more than 10 m high. O n top was a lever device
holding a basket of firewood. The sentry on top of the watch-tower could
observe enemy movements and send a warning signal - smoke by day and open
fire by night. This signalling system could transmit messages in a very short
time and get news t o Chang-an, the capital, within the same day. Each sui (bea-
con fire) o r t'ing (watch-tower) had its own name and number. The beacon net-
work of the Western Regions began from Yii-men (the Jade Gate), west of Tun-
huang, and passed through the desert and along the northern bank of the Lop
N o r directly t o Kucha and Pai-cheng. Along this route remains of ancient bea-
cons can be found. The example at Qizil, west of Kucha County, is relatively
well preserved (Fig. 12).26 Carved on a cliff at the foot of the ~elatagh,

24. Ma, 1975, pp. 29-30.

25. H u a n g Wenbi, n.d.b, p. 89.
26. Ma, 1975, p. 37.
 The Westc~rnliegions under the t-lsiung-nu and ~ h tian
c

FIG.12. Remains of a H a n beacon tower.

north-east of Pai-cheng County, an inscription dating from the fourth year of

Yung-shou during the reign of Emperor Huan-ti of the Eastern H a n (A.D. 158)
records that Lieutenant-General Liu P'ing K'uo, o n assuming office at Kucha, led
six Chinese men t o construct a series of watch-towers (Fig. 13)." At strategically
important centres, there were also fortified barracks, such as the ruins of Lou-Ian
and T ~ - ~ oinnthe north-western bank of the Lop N o r at the site of the ancient
city of Kao-ch'ang in the Turfan basin and the site of Pochengzi at Banjiegou in
Jimsar, all strategic centres where H a n troops were gjarrisoned. The eaves-end
tiles with cloud pattern, excavated at Qitai, are typically Han in style (Fig. 14).

The Silk Route

The object of the H a n in posting garrison troops, establishing farming colonies,
and building fortifications and signal beacons was not simply to avert incursions

27. Ma, 1980a.

 FIG.13. Rock inscription of Liu P'ing K'uo.

b y the Hsiung-nu; they also had an important role in safeguarding traffic on the
Silk Route (see Map 5). This road played a key role in the history of civilization,
facilitating economic and cultural exchanges between East and West. Geograph-
ical factors had t o be overcome, especially the obstacles presented by the Takla-
makan Desert. As is well known, there were t w o routes circumventing this
impassable sea of sand. T h e southern route ran west from Tun-liuang, along the
southern bank of the Lop N o r t o Tashkurgan. Ascending the Palnir plateau, it
went through the Wakhan valley t o Balkh (ancient Bactra). A branch to the
west of Taslikurgan ~ a s s e dthrough Gandhira. As the Taklamakan Desert has
been spreading south, this section of ancient road, with the Walled City-States
along it, has been subtnerged in sand. O n l y in modern times have arcliaeologists
discovered the ruins of Niya and Enderc and their important remains.
T h e northcrn route,'' also starting from T u n - l ~ u a n g ,ran iisrtli-west through

. .
28. The new northern route was opened up a t the bcg~nn~ng of thc Chi-isti~n cr.1 .iccording
to the suggestion made by kisii P'LI,the \Vu-chi colonel (scc Htrlz-sb~.r96).
 I:lc,. 14. En\ cs-end tile n~ithcloud d c s ~ g n .

Sanlongsha t o Lou-lan o n the n o r t h bank of the L o p N o r , passing Qnraslialir

and Aksu before t u r n i n g south-west t o Kashgar. F r o m the F e r g l i a n ~basin it
made a d e t o u r a r o u n d K'ang-chii, a region covering the Tashkent o'isis ,lnd a
part of the territory between the A ~ n uDar\.a and tlie S\-r Dar1.a ri\.crs (see
Chapter 19) a n d t u r n i n g s o u t h to join the southern route in Bactria. There n,erc
branches f r o m K'ang-chii t o Parthia and t o the land betnreen the Aral and C a s -
pian Seas. A n o t h e r road w e n t north-west t o Wu-sun, !.et another going north
from Lou-la11 t h r o u g h Cliii-shih close t o Turfan, nn estrernel\- important stra-
tegic position. T h e r e was also a main line of comniunication connecting tlie
northern and s o u t h e r n routes; and at the end of the Western H a n period a n c n
. .
northern route was opened u p giirlng direct access t o \X1u-sun.
T h r o u g h o u t t h e t w o H a n d!-nasties, effective rne,lsul-es \\-ere t'lken t o
maintain and defend these imporrant routes, using posthnuscs, sentl-\. guclrds
and interpl-etcrs t o l~i,lint,lin c o m m u n i c , l t i ~ n sbetn.een C l i i n , ~and the \X'est.
When Eniperor \Y/u-ti filest sent C h a n g Ch'ien to the Yueh-cliih, the O ~ C C ~ ~ \ . C

was a militarj. one, b u t sul>sequentl\ political lid ecoiionlic f,lctors c.1111~to

play the ke\r role.
Before the 1 ~ J I Icstal>lislied the post of p r o t e c t n r - g e ~ ~ e ~o- ,fl lthe \Vcstcl-n
Regions, tlicl-e h,ld hNn e s c l l . ~ ~ i g cosf diflom.ltic cniro!-s n , h o ~ l s oscr\.cd .IS
~ n e r c h n ~ l~t ~~c. l ~ c , l s i l l c ~ ~rl-.lffic
, . l I t ~ ~1l1c
l g Silk R o u t e iln~11-is1icd.Cliillese silk
products I . c J ~ I I ~pdI-L1l;~l
~~ \.id [IlC Cit\.-St<~tcsof the \\;CS~CI-II Rcginns and
were sent 011 Konlc,. CIlincscsilk soon h e c , ~ ~ i the i c most cspcnsi\.e ~LIXUI.!.
 Ma Yong and Sun Yutang

item in the Roman Empire and commanded high prices. But Han merchants
were unable to overcome the monopoly of the Parthian middlemen and
little Roman currency flowed into China. Virtually no Roman coins before
the Byzantine period (c. A.D. 400) have been found in Xinjiaq or China
proper.29

29. Finally we should allude to the question of the relations between the Western Regions
and the Kushan kingdom during the H a n period. Many works by European scholars on
Central Asian history have given incorrect accounts of this episode, believing that, at the
end of the Eastern H a n period, Kanishka, King of the Kushans, had once conquered part
of the Tarim basin, at least as far as Shu-IC and Khotan; some scholars have even gone as
far as t o include Turfan and the Lop N o r region in the Kushan Empirc. This erroneous
conclusion resulted merely from the incorrect interpretation of the Chinese historical
records. In fact, thc Hou Hun-shu faithfully records events of the Western Regions fro111
the later pcriod of the Eastern H a n epoch and also mcntions contenlporary ~ino-Kushan
relations.
 B. N.Puri

T rise of the empire of the Kushans is an important landmark in the

HE
history of Central Asia. Known to Chinese historians as Kuei-shuang,l
they were one of the important tribes of the Great Yueh-chih who had
been driven out f r o m their original homeland by another warring tribe, the
Hsiung-nu (Huns) and had settled in northern Bactria (see Chapter 7).

The Early Kushans

The Hou Han-shu (Annals of the Later Han), compiled by Fan Yeh (c. A.D.
446), based mainly o n the report submitted t o the Chinese emperor by General
Pan Yung in o r before A.D. 125, describes their rise. Ch'iu-chiu-ch'ueh (Kujula
Kadphises), the yabghu of Kuei-shuang, attacked and destroyed the other four
yabghu and made himself King of the Yueh-chih. H e attacked An-hsi (Parthia)
and took the territory of Kao-fu (Kabul). H e also overthrew P'u-ta (Puskalii-
vati) and Chi-pin (Kashmir) and annexed these countries.' It was argued by Jit-
zuzo that the five yabghu already existed in Bactria when the Yueh-chih
arrived, and s o the Kushans could not have been the Yueh-chih. Some scholars,
therefore, refer t o the Saka-Kushans in the Yueh-chih hoard.' But Tarn4 regards
this theory as an unhappy offshoot of an elementary blunder that started the
belief in a Saka conquest of Graeco-Bactria; most scholars now agree that the
HOUHan-shu gives an authentic account that is trustworthy. The chronology,
however, of these events relating to the rise and consolidation of the Kingdom

:'. See May 4.

1. Pulleyblank, 1962, pp. 206 et seq.
2. Pulleyblank, 1968, pp. 247-58; Ziil-chcr, 1968, pp. 346-90.
3. Maenchen-Hclfen, 1945, pp. 71-81; Puri, 1965, pp. 1 et scq.
4. Tarn, 1951, p. 287.
of Kuei-shuang is disputed because it is closely related to the history of the
Great Kushans and the date of Kanishka. Excavations at Taxila and elsewhere
have conclusively settled the old argument as to whether the Kadphises Pre-
ceded the Kanishka group of kingsS as coins of the Kadphises group, but not of
Kanishka, Huvishka, etc., are found in the Early Kushan levels of Sirkap. The
H ~ U Han-shu further informs us that Ch'iu-chiu-ch'iieh (Kujula Kadphises)
died at an age of more than 80 and was succeeded by his son Yen-kao-&en
(Vima Kadphises), who in turn destroyed T'ien-chu (India) and placed a general
there to control it. The Chinese annals seem t o provide a terminus ante quem
for the Kadphises rulers of A.D. 125, the date of Pan Yung's report.
T w o series of dated inscriptions provide a more precise chronological
framework for the rise of the Early Kushans. The first series bears a sequence of
dates, some of which are qualified by Ayasa ('in the era of Azes') (see Chapter
8). The Takht-i Bahi inscription of the Indo-Parthian king Gondophares is
dated in the twenty-sixth year of his reign and Year 103 of the era.' Its reference
to erjhuna kapa suggests the presence of Kujula K a d ~ h i s e sas a prince at the
court of the Indo-Parthian king. The Panjtar stone inscription dated Year 122 of
the era, nineteen years later,' is dated in the reign of an unnamed king described
as the G u ~ a n amaharaja. This same term 'Gusana' occurs in the Manikyala
inscription of the time of Kanishkax which describes Lala as Gu;anava-
iasamvardhaka, 'the increaser of the Kushan race'. ' G u ~ a n a 'therefore stands
for 'Ku~ana'.The Taxila silver-scroll inscription of Year 136 Ayasa - of the era
of Azes - gives as ruler an unnamed king, 'the Great King, the King of Kings,
the Son of Heaven, the Kushan'. The nameless king with high titles has the same
context as the nameless king of the Early Kushan coins struck with the titles of
the King of Kings, the Great, the Saviour, which can now be placed after the
coinage of Gondophares and the local issues of Kujula, but before the standard
uniform coinage of Vima Kadphises.' It is clear that the prince of the Takht-i
Bahi inscription in Year 103 and the ruler of the Panjtar stone inscription in
Year 122 has extended his empire substantially by the time of the Taxila silver-
scroll inscription in Year 136 and adopted high-sounding titles. The sequence of
events clearly suggests that the three inscriptions refer to the same person, who
belongs to the period before the N e w Era was iiltroduced by Kanishka; and that
the nameless king of both the coins and these inscriptions represents the later
stages of the rule of Kujula Kadphises after he had captured P'u-ta (Puskalivati)
and Chi-pin (Kashmir). Kujula Kadphises is said to have lived for more than
eighty years. H e played the key role in establishing the Kushan Empire and his
coins are very numerous in the finds from the Early Kushan city of Sirkap.

5. Marshall, 1951.
6. Konow, 1929, pp. 57-62.
7. Ibid., pp. 67-70.
8. Ibid., pp. 145-50.
9. MacDowall, 19686, pp. 2 8 4 8 .
 The Kushans

~f the credibility of the Khalatse inscription12 is accepted, identifying

uvilna Kavthisa with Vima Kadphises, then the octogenarian father Kujula
Kadphises should be assigned a long reign of about fifty years, terminating
between Years 160 and 165 of this era, with a reign of twenty to
thirty years for Vima Kadphises, his son. It is now generally accepted that this
era of Azes (Ayasa) may well have begun at the same time as the Vikrama era of
58 8.c. (see Chapter 8). T h e dates assigned, then, to Kujula Kadphises would
include A.D. 45 (Takht-i Bahi), A.D. 64 (Panjtar) and A.D. 78 (Taxila silver
scroll),and the dates of Vima Kadphises would include A.D. 127 (Khalatse).
The second series of dated inscriptions includes the Taxila silver vase of
~ i h o ~ i kthe
a satrap dated Year 191." This used to be attributed to the series of
dates in the Azes era, but MacDowalllZ has shown that Jihonika's context falls
after the reign of Azes I1 and before Kujula Kadphises in the decade A.D. 30-40,
and the date must therefore be attributed to an Indo-Bactrian era." The tri-
lingual inscription at Dasht-i N a w u r of Vima Kadphises is dated Year 279.14The
unfinished inscription from Surkh Kotal of Kanishka is dated Year 27915 and
that of Vima Kadphises is dated Year 299.I"oth these inscriptions, as the Taxila
silver-vase inscription of Jihonika, belong to the same Graeco-Bactrian era,
probably the era of Eucratides beginning with his accession around 170 B.C. (see
Chapter 17). T h e dates then assigned t o Vima Kadphises would include A.D. 109
(Dasht-i Nawur) and 129 (Surkh Kotal).
Some scholars associate Kanishka with the Saka era of A.D. 78 and conse-
quently have t o place Kujula Kadphises and Vima Kadphises before that date.
To maintain consistency they have t o find earlier reference dates for the two
eras. For example, Fussman17 links Year 279 with a Graeco-Bactrian era of inde-
pendence from the Seleucids in 247 B.C. to give dates of A.D. 32 and 52 for Vima
Kadphises. T h e problems surrounding Kanishka's dating call for detailed con-
sideration.

The date of Kanishka

The date of Kanishka does not stand in isolation. In his time the Kushan Empire
covered a vast amount of territory from Bactria t o Benares and from Kashnlir to
Sind, and Kushan coins have also been found in recent excavations in Choras-

10. Konow, 1929, pp. 79-8 1.

11. Ibid., pp. 81-2.
12. MacDowall, 1973, pp. 215-30.
13. Tarn, 1951, pp. 494-502; Bivar, 1963, pp. 489 et seq
14. Fussman, 1974, pp. 8-22.
15. Bivar, 1963, pp. 498-502.
16. Harmatta, 1965. pp. 164-95.
17. Fussman, 1974.
 B. N.Puri

mia, Khotan and eastern Iran. There is now substantial agreement on most
points concerning the relative chronology of the Kushans, but the absolutedate
of the reference point for the era of Kanishka remains hotly disputed. It is now
agreed that it cannot have been the Vikrama era of 58 B.C. which was proposed
by Fleet and Kennedy.ln But the dates advocated still range from A.D. 78 (the
Saka era), which is still supported by many Indian scholars, to A.D. 278, once
proposed by Bhandarkar" and Majumdar2' and now supported by Zeimal.!l
The consideration of any of the dates proposed must be fully reconciled
with other established historical sequences of which the absolute dating is
firmly established, in particular the Guptas and Western Satraps. The establish-
ment of the Imperial Gupta dynasty by Candragupta in A.D. 319, and the inter-
vening kingdoms and republican states that came from the Kushan dynasty and
before the Guptas in India - the Nigas, Yaudheyas, Milavas, Arjunayanas,
Kunindas and Madras - provide a firm terminus ante quem for the Kushan
dynasty in Indian history. The context of the Western Satrap Rudradiman and
his occupation of Sind, Sauvira and Malwa before Saka Year 72 (A.D. 150) in the
Junagadh i n ~ c r i p t i o ncannot
~~ be disputed, nor can his independent status be
questioned. H e claims in this inscription that he had personally acquired the
status of mahdk;atrapa through his own prowess and strength.23If Kanishka is
taken to be the founder of the Saka era of A.D. 78, the dates of his successors
Huvishka and Visudeva would clash with those of Rudradiman, and it cannot
be proved that Rudradiman o r his family were ever subordinate to the Kushans.
Another fixed date that must be considered is the dispatch by Po-t'iao,
King of the Great Yiieh-chih, of an envoy with tribute to the Wei as a token of
his affection, on the day Kuei-mao (26 January) A.D. 230 (San-kuo-chih,
'Memoirs of the Three Kingdoms', 3.6a).
Po-t'iao has been identified with V i s u d e ~ a Advocates
.~~ of a date in the
second century for the era of Kanishka identify him with Visudeva I, while
those arguing for the A.D. 78 date regard him as a later ruler, Visudeva 11.
G h i r ~ h m a ndates
~ ~ the era of Kanishka to A.D. 144 because of his excavations at
Begram and the evidence of the trilingual inscription of the Sasanian emperor
Shapur I at Naqsh-i Rustam. The Begram excavations suggest three chronolog-
ical stages. The first phase predates Kanishka, yielding coins of ~ u j u l aKad-
phises and Vima Kadphises along with those of the Indo-Greek and ~cytho-
Parthian rulers. The second phase contains coins of Kanishka, Huvishka and

18. Vallke Poussin, 1930, pp. 346 et seq.

19. Ibid.
20. Majumdar, 1968, pp. 150 et seq.
21. Zeimal, 1974, pp. 292 et seq.
22. Kielhorn, 1905/06, pp. 36 et seq.
23. Raychaudhuri, 1953, pp. 424 et seq.
24. Ziircher, 1968, p. 371.
25. Ghirshman, 1946.
 The Kushans

vasudeva, and ends with a major destruction that Ghirshman associates with
the conquests of Shapur 1. H e argues that the conquests of Shapur I provide the
~ o i n of
t the second dynasty of the Kushans, and that ShapurPscon-
quest should be placed between his accession in A.D. 241 and his second war
againstthe Romans (A.D. 251-52). The latest coins found in the city of Begram
were those of Vasudeva, the Po-t'iao of the Chinese San-kuo-chih and the same
person as Vehsadjan, King of the Kushans, mentioned by the Armenian Moses
of Khorene. However interpreted, the Sasanian conquest of the western Kushan
provinces is a further fixed point which must be considered. Shapur I's inscrip-
tion on the Ka'be of Zoroaster at Naqsh-i Rustam claims to have incorporated
the Kingdom of the Kushans u p t o Peshawar in the Sasanian Empire.JLThe
inscription does not mention the date of the destruction of the Kushans leading
to this. In fact, it only records the inclusion of part of the Kushan Empire,
which could be the result of a conquest either by Ardashir o r by Shapur I and
which could have taken place at any time between A.D. 223 and 262. N a r a i ~ l ? ~
argues that Ghirshman's date f o r the destruction of Begram I1 (based on two
hypotheses - finds of eight poor coins of Visudeva I and Shapur's eastern cam-
paign) stands unproved; he claims the numismatic evidence goes clearly against
any classification of the Kushans into three dynasties, and argues for an inter-
mediate date of A.D. 103 for the accession of Kanishka.
Pulleyblankz%upports Ghirshman's date of A.D. 144 from other evidence.
Late Buddhist traditions connect Kanishka with Khotan and there is strong cir-
cumstantial evidence for Kushan penetration into the Tarim basin from the use
of north-west Indian Prakrit as an administrative language, and from the finds
of copper coins of Kanishka at Khotan. Pulleyblank argued that there could not
have been any Kushan invasion before A.D. 175. GoblZ9initially supporteA this
chronology of A.D. 144 with an analysis of Kushan coin types which, he argued,
were copied from Roman coins - Vima drawing from Trajan, Kanishka from
Hadrian and Huvishka f r o m Antonius Pius. But later Gobl'khanged his view
to A.D. 232 from a linkage he found between the Sasanian gold coinage of Sha-
pur I1 struck at Merv and the Kushano-Sasanian coinage of Hormizd I at the
beginning of the Kushano-Sasanian series. Majumdar" drew attention t o simi-
larities between Kushan and Early Gupta forms in iconography and palaeog-
raphy, and connected Kanishka's accession with the beginning of the well-
known era of A.D. 248149. ZeimaP? went further and suggested A.11. 278.
Endorsing Bhandarkar's 1899 suggestion that the beginning of the era should be

26. Maricq, 19586, pp. 295-360.

27. Narain, 1968, pp. 206-39.
28. Pulleyblank, 1968, pp. 247 et seq.
29. Gob], 1960, pp. 75-91; 1968, pp. 103-13.
30. Gobl, 1984, p. 52, 82.
31. Majumdar, 1968, pp. 150 et seq.
32. Zeimal, 1974, pp. 292-301.
 R. N.Puri

equated with the Saka era of A.D. 78, he regarded Kanishka's era as the
century, from A.D. 278. But any of these late dates placing the Great Kushans
(the dynasty of Kanishka) in the thirdjfourth centuries A.D. would involvea
clash not only with the Guptas but also with several other tribes ruling indepen-
dently between the Later Kushans and the Imperial G ~ p t a s . ~ )
Many scholars have identified the accession of Kanishka with the Saka era
of A.n. 78. Rapson3.' argued that the date on the coins and inscriptions of the
Western Satraps of Surashtra and Malwa should start in Kanishka's reign in A.D.
78, but because of its long use by the Saka Western Satraps it became known in
India as the Saka era, which effectively disguised its origin and perplexed mod-
ern scholars. T01stov~~ found an era of A.D. 78 used in Chorasmia. Bashamj6also
noted that the era of A.D. 78 was used by the Magha kings of KauSambi and was
equated with the Licchavi era used in Nepal; he argued that such wide use of an
era was only possible with the patronage of a great power, which could only be
the Kushans. But the difficulties in reconciling the presence of Rudradaman (the
powerful Western Satrap), who was independent of the Kushans, campaigning
against the Yaudheyas, in the lower Indus and Malwa between A.D. 130 and 150,
in territory that was part of the fully established Kushan Empire, led Puri3' to
suggest that the era of Kanishka might have started around A.D. 142. A date in
the early second century A.D. certainly seems to fit better the evidence of asso-
ciated Kushan and Roman coin finds3xand the careful analysis of events under
Shapur I by HarmattaY3'but the issue still remains open, awaiting new evidence
and an analytical reconstruction that adequately explains and takes full cogni-
zance of the fixed points of externally dated events.

The Great Kushans

The chronological framework of the dynasty of the Great Kushans is provided
by the series of inscriptions dated in the era of Kanishka. Inscriptions are
known of Kanishka dated Years 1-23, of Vasishka dated Years 24 and 28, of
Huvishka dated Years 28-60 and of Viisudeva dated Years 67-98."O There is
another inscription of Year 41 from Ara of a Kanishka, son of Vajheshka, with
the titles 'mahiraja rajatiraja devaputra' and 'Kaisara'. Year 41 falls in the

33. Flect, 1892, pp. 1 et seq.

34. Rapson, 1922, p. 585.
35. Tolstov, 1968, pp. 304-26.
36. Basham, 1968, pp. X I I - X I I I .
37. Puri, 1965.
38. MacDowall, 1968a, pp. 134-54.
39. Harmatta, 1965, pp. 186 ct seq.
40. Puri, 1965, 1977, pp. 101-61; Janert, 1961.
 The Kushans

middle of the reign of Huvishka. Smith, Puri and Banerji41 identified him with
the Great ~ a n i s h k aand suggested that with advancing years and pressure of
military affairs in Central Asia, Kanishka had left his son Vasishka as viceroy i n
~ ~ d~ ai s~i s .h k apredeceased his father and was replaced by h'IS brother Huv-
ishka. But it could as well be proposed that this Kanishka was another ruler
who held the western part of the Kushan Empire in Year 41, a brothcr
of ~ ~ v i s h kassociated
a with him in power o r a member of a collateral branch
,ho usurped power f o r a time in part of the empire. There are several other
possibilities such as the division of the empire between two brothers, Vasishka
and Huvishka, o n Kanishka's death, with a second Kanishka succeeding his
father and finally becoming sole Kushan emperor." But there is another possi-
bility, that both Vasishka and his son Kanishka belong to a separate group of
kings after the Great Kushans (Kanishka, Huvishka and Visudeva).
There is also a reference t o another Kushan ruler, Vaskushana, in an
inscription" dated Year 22 from Sanchi. H e could not have ruled independently
in this area when Kanishka was alive. It is, therefore, tempting t o identify this
Vaskushana with Vasishka. While a king called Vasishka is not known in the
coin series of the Great Kushans, a king of this name is known in the coinage of
the Later Kushans after V i i ~ u d e v a It
. ~ can
~ therefore be suggested that Vaskush-
ana, a Kushan mahariija in Year 22 and the Kanishka of the Ara inscription in
Year 41, belong t o the period after the century of the Great Kushans. These
Later Kushan rulers would include both Vasishka and his son Kanishka, and
perhaps another Kanishka known from the Mathura inscription of Year 14
which on palaeographic grounds comes closer to the Gupta period.45 Such a
chronological framework can cut the Gordian knot created by the Ara inscrip-
tion; the Kanishka in the Surkh Kotal inscription dated Year 31Jhseems to be
the same Late Kushan ruler.
In the light of these inscriptions, Table 1 sets out a chronological frame-
work of the Early, Great and Later Kushan rulers. The last ruler, Kanishka, may
then have been a contemporary of the later Indian dynasties receding the Early
Guptas. There is clearly a second era of the Later Kushans in the inscriptions
from Mathura, and evidence f o r a Later Kushan era starting in A.D. 234 and
used on coins of Tekin Shah, King of Udabhiindapura, and the Tochi valley
inscriptions. This has led some scholars (Harmatta, Humbach, MacDowall) to
place the beginning of the Kanishka era itself in A.D. 134, a century before the
commencement of the second Kushan era.

41. Smith, 1924, p. 286; Puri, 1977, pp. 159-60; Banerji, 1908, pp. 58 et seq.
42. Konow, 1929, p. 163.
43. Marshall and Foucher, 1947, Vol. I, p. 386; ~ o h u i z e n - d eLeeuw, 1949, p. 314.
44. Gob], 1984, pp. 58-78.
45. Puri, 1965, pp. 70 et seq.
46. Maricq, 19586, pp. 345 ct seq.
 B. N.Purt

TABLE1. Chronological framework of rulers

Era dates
Rulers --

Graeco-Bactrian Azes Kanishka Later Kushans

Sakas
Jihonika the satrap 191
Early Kushans
Kujula Kadphises 103
Nameless king 122 and 136
Vima Kadphises 279' 184 (7)

Great Kushans
Kanishka
Huvishka
Visudeva

Later Kushans
Kanishka I1
Vasishka
Kanishka 111

I . l'hc d ~ t eis read as 285 by Marien and 299 by Harniartn.

Relations with Iran

Kujula Kadphises is mentioned as a prince (eyjhuna Kapa) at the court of the
Indo-Parthian king Gondophares in the Takht-i Bahi inscription of Year 103
(A.D. 45). According t o the H o u Han-shu, Kujula is said t o have attacked An-
hsi (Parthia) and taken the territory of Kao-fu (Kabul). It is difficult to explain
the presence of a Kushan prince at the Indo-Parthian court in Taxila, but it is
clear that eventually Kujula Kadphises reconquered the province of Kipiia and
Kabul from the Indo-Parthians and then captured the Indus provinces of the
Indo-Parthians, including Taxila, from the successors of Gondophares." Vima
Kadphises (Fig. 1 ) seems t o have profited from the weakness of the Indo-Par-
thians t o seize all the Indus valley u p t o Sind. At the height of their power
under Kanishka, the Kushans did not seem t o be interested in territorial gains at
the expense of their neighbours, the Parthians. Buddhist tradition refers to a
war by Kanishka against the Parthians and according t o Ghirshman" it might
have taken place in the reign of Vologases 111, probably occasioned by a

47. Banerjea, 1957.

48. Ghirshman, 1978, p. 262.
 The Kushans

FIG.1. Statue of Vima Kadphises sitting on a lion throne. Mathura.

Parthian attempt t o recover some of the Iranian provinces captured by the

Kushans from the Indo-Parthians.
The ascendancy of the Kushans posed a continuing threat to Parthia's
eastern boundary. Eventually the founder of the Sasanian dynasty, Ardashir,
attacked the Kushans and conquered Margiana, Carmania and S i ~ t a n . ~Tabari '
says that the kings of the Kushans, of Turan and of Makran submitted without
battle to Ardashir and kept their territories as vassal^.'^ Ardashir's successor.
Shapur I, claims among his provinces Sind and the country of the Kushans up to
Peshawar in his inscription in the Ka'be of Z ~ r o a s t e r . The
~ ' Kushan dynasty of
Kanishka was deposed and replaced in the north by another line of Kushano-

49. Ghirshman, 1946, pp. I 00 et seq.; Narain, 1968, pp. 2 1 1-1 2.

50. Maricq, 1968, pp. 1 8 2 4 .
51. Maricq, 19586, pp. 295-360.
 R. N.Puri

Sasanian princes ruling a considerably reduced kingdom, and recognizing the

suzerainty of the Sasanians, at least for a time. There was a serious revolt in
eastern Sasanian provinces in the time of Bahram I1 (A.D. 276-93), when
king's brother, who was viceroy in Sistan, attempted t o seize the throne, and the
Kushall king supported him. Peace was restored with the marriage of Horlnizd
11, son and successor of Narseh (A.D.303-09), to a Kushan princess.52The death
of Hormizd I1 left a minor, Shapur 11, on the Sasanian throne. The Kushans
took advantage of this, and the internal disorders in Iran, to recover the lost ter-
ritory, but Shapur 11, on attaining his majority, waged a new war against the
Kushans and decisively defeated them.

Relations with China

The H o u Han-shu provides information only about the Kadphises rulers and
refers to the failure of a Kushan army sent against the Chinese general Pan
Ch'ao. The Chinese general's successful policy in Central Asia coincided with
the Kushan conquest of northern India and led t o a conflict of interest with the
~ o l i t i c a laspirations of Vima K a d ~ h i s e s(see Chapter 10).
Rivalry between the Kushans and the Chinese in Central Asia seems to
have continued u p t o the time of Visudeva. The Chinese work, the Sun-kuo-
chih, compiled by Ch'en Shou (A.D. 233-97), records that the King of the Great
Yueh-chih, Po-t'iao, sent an envoy with tribute to China and was given the
honorary title of 'King of the Yueh-chih who shows affection towards the Wei'.
Po-t'iao has been identified with Visudeva, either Visudeva I or Visudeva 11,
depending on the chronology favoured by the scholar concerned.

Relations with Rome

According to Dio Cassius5' many embassies came to Augustus, and the Indians,
having previously proclaimed a treaty of alliance, concluded it with the presen-
tation of gifts including tigers, animals that the Romans saw for the first time.
Florus, writing in the time of Trajan (A.D. 98-117)," refers to the arrival in
Rome of several embassies, especially from the Indians. Political relations, seen
in the dispatch of embassies, seem to be connected with trade contacts and com-
mercial transactions related t o the silk trade. Some of the copper coins of Kujula
Kadphises have an obverse head closely copied from the portraits on the Julio-
Claudian silver denarii of Augustus and Tiberius, and show the Kushan empe-

52. Ghirshman, 1978, p. 296.

53. McCrindlc, 1901, p. 212.
54. Ibid., p. 213.
Thc Kushans
 B. N.Puri

ror sitting on a curule chair which appears on the reverse of Roman coins of
Claudius and may well represent a gift from a Roman emperor. Roman aurei
and denarii were used extensively in Roman sea trade with India, which traded
in silk and spices. Pliny (Natural History XII.10.41) refers t o the serious drain
of Roman coins exported t o India. The gold coinage introduced by Vima Kad-
phises used a gold dinar that copied the weight standard of the Roman gold
aureus,55and the impact of Graeco-Roman art in Gandhira sheds light on the
and commercial relations between the Kushan Empire and the Roman
world.

Relations with north-eastern India

The extension of the Kushan Empire in northern India seems to have been the
achievement of Kanishka (Fig. 2), whose inscriptions are found at Mathura,
KauSambi and Sarnath. The distribution of copper Kushan coins of Kanishka
and Huvishka extends as far as Patna and Gaya in eastern India.56The RZjat-
arangini and the H o u Han-shu show Kanishka's hold over Kashmir and
parts of central and south-western India.57 The reference in the Suidhau-
mapitakanidznasitra to the defeat of the King of Pitaliputra, when Kanishka
demanded a large indemnity but agreed t o accept Aivaghosa, the Buddha's
alms bowl and a compassionate cock, confirms Kushan activities in north-east
India.
After Huvishka, the Kushans lost some more distant territories in east-
ern India, but Mathura long remained under Kushan rule. The long series of
inscriptions found there continues up t o Year 57 of the second Kushan era
under the Later K u s h a n ~ and
, ~ ~ it has been thought that Mathura was a second
capital of the Kushans for the eastern region (Fig. 3). The appointment of
satraps for Mathura, as at Sarnath, points to a determined control over the
region. Huvishka's reign was a period of political security and economic
prosperity. The extensive range of gold coins of Huvishka, retaining a good
weight standard and high gold purity, suggests economic stability closely
associated with political stability. Visudeva's long rule of more than thirty
years was equally characterized by political stability at home. After Visudeva,
the Kushans lost more territory t o a series of new dynasties and republican
states.

55. Sewell, 1904, p. 591; MacDowall, 1960, pp. 63 et seq.

56. Majurndar, 1932, pp. 127 et seq.; Banerji, 1951, pp. 107 et seq.; Gupta, 1953, pp. 185
etseq.
57. Thomas, 1935.
58. Rosenfield, 1967, pp. 270-3.
Relations with the Saka satraps
~t has been suggested by some scl~olars"that the Kushans had a radical affinity
with the Sakas and were a Saka clan. The term 'Saka' has been used in a very
imprecise way, and it is possible that the Kushans may have been the descen-
dants of some of the Sakas mentioned by Herodotus. According to the &lak;-
caryakathdnaka,"O the Sakas of the Indus conquered Surashtra and Malwa
shortly before the beginning of the Vikrama era (57 LC.), but were ousted by
Vikramiiditya. After a lapse of 135 years (c. A.D. 78) a new Saka came and re-
established the Saka dominion there. It has been suggested that the second con-
quest was associated with Vima Kadphises and his satraps ruled as the Saka
satraps of western India, without any regal appellation like mabirija. But they
also used the title of mahdk~atrapa,which could mean either the attainment of
independence or promotion in the administrative hierarchy. The expression
' ~ v a y a m a d h i ~ am
t a a h i k ~ a t ~ a pndmah'
a in the Junagadh inscription6' of Rudra-
diiman is especially significant. While there is n o specific evidence that the Saka
satraps of western India ever owed allegiance t o Vima Kadphises, circumstantial
evidence, as also that adduced by the H o u Han-shu, suggests his conquest of
Sind or the Indus region and his association with the Saka satraps who used the
Saka era, probably founded by their overlord, in their records.
The Kushans held the lower Indus valley. An inscription of Kanishka
Year 11 was found at Sui Vihar near B a h a w a l ~ u r and , there have been finds of
Later Kushan coins from the stupa site at Mohenjo-daro" and at Jhukar, about
30 km t o the north."' The find of potsherds with Kharosthi lettering at Tor
Dheri in the Loralai District of Baluchistanh4 may suggest an expansion of
Kushan power in that region. But Kushan rule in Sind and Sauvira (modern
Multan) seems t o conflict with the claims of the Western Satrap Rudradiman,
recorded in his Junagadh inscription of A.D. 150. This could be reconciled if we
presume that he was a satrap of Kanishka, for which there is no evidence, or
that he preceded Kanishka, which seems more probable (see discussion on the
date of Kanishka above).

The Kushan political system

The divinity of kingship seeins to have been the most conspicuous element in
the Kushan political system. Their kings were not only accorded the title of

59. Maenchen-Helfen, 1945, pp. 71 et seq.

60. Jacobi, 1880, pp. 247 et seq.; Konow, 1929, pp. XXVI-XXVIII.
61. Epigraphica Indica, p. 82.
62. Marshall, 1932, p. 127.
63. Majumdar, 1934, p. 7.
64. Konow, 1929, pp. 173-7.
- The Kushans

1dpyaprt~a'"5 (Son of God), corresponding to the Chinese imperial title 't9im-

(Son of Heaven), but were deified after death and their statues were set up
;,, a devakula (god house). Such statues of Kushan rulers have beerl recovered
from at Mat, near Mathura, and from Surkh Kotal in Afghanistan.
is probable that the statue of the deified Huvishka was erected i n the lifetime of
[he ruler.66The Kushan rulers were secularist in one sense, in that they depicted
divinities from different pantheons on their coins, but religion and polity were
interlinked. The Mat inscription of Huvishkah7refers to him as 'stcadfast in the
true law', a title also borne by the first Kushan king, Kujula Kadphises, on his
coins. It is further recorded that on account of his devotion, the kingdom was
conferred on the grandfather of Huvishka by Sarva (which is another name for
the god Siva) and Candavira (a name connected with the moon).
The Kushan kings assumed high-sounding titles"Qorrowed, like the
divinities on their coins, f r o m different regions and civilizations. They usc the
Indian titulature 'maharaja vcijatircija' (Great King, the King of Kings), its Ira-
nian counterpart ':aonano !do' and its Greek counterpart 'Basileus Basileon'
(Fig. 4). These titles, n o doubt, indicate Kushan paramountcy over areas where
lesser princes and feudal lords retained local power. In the Ara inscription, the
Later Kanishka also has the title 'Kaisava', the equivalent of 'Caesar' used by
Roman emperors, suggesting Kushan contact with Rome and a claim t o com-
parable status. Some titles were borrowed from their Bactrian, Saka and Indo-
Parthian predecessors. It has been suggested that they also inherited a system
of joint rule, but there is n o numismatic evidence for this. N o Kushan coin
portrays two rulers. T h e argument for supposed joint rule is based o n inscrip-
tions that seem t o show kings with overlapping dates: an inscription of
'Vasku~dna' (identified with Vasishka) from Sanchi with the title 'rcija' dated
Year 22 when Kanishka was king and the Ara inscription dated Year 41 when
Huvishka was king. But both these inscriptions are dated in the Later Kushan
era. There is consequently n o overlapping, and the dual kingship known in the
Indo-Parthian political system does not seem t o have been ~ r a c t i s e dunder the
Kus hans.b9

Kushan administration
The vast Kushan Empire, extending from Central Asia t o Bihar and from Kash-
mir to Sind, containing peoples of different nationalities and religions with a

65. Thomas, 1935, pp. 97 et seq.; Sharma, 1959, p. 177.

66. Janert, 1961, p. 145.
67. Ibid., p. 144.
68. Puri, 1939140, pp. 4 3 3 4 1 .
69. Puri, 1965, pp. 79-87.
 B. N.Puri

FIG.4. Coin of Kanishka I with Greek legend and the title 'Basileus Basileon'.

heterogeneous socio-economic background, was governed through an organ-

ized administrative system, probably in three tiers, at central, provincial and
local levels. The king seems t o have possessed unfettered powers, as we find no
reference in the Kushan records t o any advisory body or to councillors cor-
responding to amdtyas and sacbivas of the Mauryan period. The Kushans seem
t o have followed the earlier existing pattern of the Indo-Greeks and Parthians
by appointing k ~ a t ~ a p aand
s mahZkSatrapas for different units of the empire.
Inscriptions provide the names of some such ksatyapas, some foreign, like
Vanaspara, and the mabdkSat~apaKharapalliina at Varanasi, Namda at Mathura,
VeSpasi and Lala, a scion of the Kushan family, Liaka, and an unknown satrap,
son of the satrap Granavhryaka at KipiSa (Begram). Some inscriptions show
that certain appointments were hereditary.
They mention other officials both civil and military functions,
called 'dandanayaka' and 'mabadandaniyaka'. The two terms are found in
 The Kushans

inscriptions throughout India, suggesting the prevalence of [his feu-

dal element - as one might presume - in the administrative set-up of different
ruling families over a considerable period of time. They were cllarycd
..
and military responsibilities in different areas. The dandanayaka
was presumably the wielder of the rod (danda), acting both as commissioner of
police to prevent crime and as a judge or criminal magistrate administering jus-
,ice. He could also ~ e r f o r mmilitary functions although he is distinguished from
..
the renani, or real commander. H e is also differentiated from the dandapdiika
of the later records which probably signifies someone carrying fetters b a i a ) .
The laces where inscriptions mentioning satraps and other officials have
been found indicate localities f o r which they were responsible. Satraps are
known for KapiSa (Begram), Manikyala (near Rawalpindi), Und (west of the
Indus), Mathura, Varanasi, etc. There may have been satraps for other parts of
the empire, but the evidence o n this point is wanting. The relations between
ksatrapas and dandandyakas are n o longer defined, but it may be assumed that
ksatrapas were definitely at a higher administrative level than the dandaniiya-
kas. The use of foreigners alone at the higher level of political organization
ensured efficiency and minimized the chances of internal dissension and dis-
order, but this principle was not applied at local village level. The inscriptions
mention two terms - 'grdrnika' and 'padrapala' - both signifying 'village head-
man', who collected the king's dues and took cognizance of crimes in his area.
There is no information about the local government that we find later in the
Gupta period.
The scanty information available suggests that the Kushan rulers accepted
the prevalent Indian and Chinese concept of the divinity of kingship, and bor-
rowed the Achaemenid and subsequently Indo-Greek and Indo-Parthian sys-
tem of appointing satraps as ~ r o v i n c i a l governors, while the feudal lord
(dandanayaka) was their o w n creation. The title is n o doubt Indian, but all feu-
dal lords known t o have been associated with the Kushan administration were
foreigners.
 E C O N O M YA N D S O C I A L S Y S T E M
I N C E N T R AA LS I A
I N T H E K U S H A NAGE:'

A. R. Mukhamedjanov

D
URING the period of the Kushan Empire, great progress was made in
the social and economic life of the peoples of Central Asia. The econ-
omic prosperity they enjoyed was due to a number of factors: (a) the
unification of the greater part of Central Asia's ancient agricultural regions
under the authority of a single empire; (b) the maintenance of political stability
over long periods; (c) the rapid development of farming (with crop irrigation)
and handicrafts; and (d) the expansion and strengthening of trade relations with
India, China and the countries of the Near East. With the expansion of internal
and international trade, and the development of economic relations in Central
Asia, agriculture, which had already played a major role in the country's econ-
omic development, acquired even greater importance. In countries with inad-
equate rainfall, agriculture, the backbone of ancient civilizations, has always
depended o n artificial irrigation and many aspects of the social and economic
life of the peoples of Central Asia in the Kushan ~ e r i o dare closely linked with
irrigation as an element in agricultural production and general prosperity.

Irrigation
Archaeological evidence reveals intensive exploitation of new agricultural land
and the expansion of agricultural oases at the beginning of the Christian era in
the river valleys and ancient agricultural oasis areas of Central Asia, especially in
the southern regions, even though the best and most suitable croplands were by
that time already under cultivation. It has also been established that, with the
opening up of new regions and the extension of crop-farming to the northern
provinces of Central Asia o n the lower reaches of the Zerafshan, on the middle

:' See Maps 4, 5 and 6

reaches of the Syr Darya and in the Tashkent oasis, large numbers of nomadic
livestock-breeders switched to a settled way of life and new centres of urban
civilization were formed. As a result of the extensive development of irrigation
networks, practically all the main provinces of Central Asia were brought under
cultivation during this period and the establishment of the major crop-growing
oases was completed. The extent t o which northern Bactria was populated and
brought under cultivation at this time can be judged from the 1 1 7 archaeological
monuments of the Kushan period recorded in recent years in the territory of
Surkhan Darya province.' A major channel, the Zang canal, leading from the
Surkhan river, was constructed. In the zone irrigated by it a new oasis, the
Angor, was established around the town of Zar-tepe.* The founding of Dalver-
zin-tepe as a major urban centre also dates back t o this period. The Surkhan
Darya and Sherabad Darya valleys, with their flourishing agricultural oases, for-
tified towns and extensive grazing lands, were able t o provide a strong base for
unifying the domains of the Yueh-chih o n the right bank of the Amu Darya.
When they were unified by the ruler of Kuei-shuang, who subjugated the four
other Yueh-chih principalities, the nucleus of the Kushan Empire was formed.
This was the time when large-scale irrigation systems were developed in
the Zerafshan and Kashka Darya valleys and the Tashkent oasis. The major irri-
gation works constructed in the Samarkand oasis and which carried water from
the Zerafshan river were the Bulungur and Payarik canals on the right bank, the
Dargom and Narpai canals on the left bank, and the Ishtikhan and Naukinsk
systems in the Miyan-kala territory. Some of these extended over a distance of
more than 100 km. In the Bukharan part of the Zerafshan valley, the river fed
the Kanimekh (Kanimug), Kharkan Rud, Zandana and Ramitan Rud canals on
the right bank, and the main canal, the Shah Rud (Rud-i Zar) and many others
o n the left bank.j As a result of the development of irrigation in the Zerafshan
river valley, a vast area was supplied with water and brought under cultivation.
According t o our calculations, some 3,400-3,500 km2 of land along the lower
reaches of the Zerafshan alone were irrigated in the ~ e r i o dfrom the first to the
fourth century A.D. The western boundary of these ancient irrigated lands,
which today passes through the sands of Kyzyl Kum, was then at certain ~oints
situated some tens of kilometres beyond the k resent-day limits of the ~ukhara
o a ~ i sThus,
.~ l ~ entire flood- lain of the
during the Kushan period, ~ r a c t i c a l the
Zerafshan valley was brought under cultivation, and the two large agricultural
oases of Samarkand and Bukhara were established.
During the same period, a number of major irrigation systems - the
Rudaksa Kasan, Faizabad, Nasaf-Denau, Kamashi and many other canals -

I. Rtveladze and Pidaev, 198 1

2. Masson, 1981.
3. Mukhamedjanov, 1978.
4. Ibid.
 Economy and social system in Central Asia

were built along lower reaches of the Kashka Darya river. Many fortifications,
and farmsteads of the Late Kushan period were constructed i n the
of these canals, especially in the third and fourth centuries *.I,. The
establishment of ancient N d d ~ s h a boasis and its centre, the town of Er-kurgan,
was c ~ r n p l e t e d T
. ~h e oasis covered some 1,500-1,600 km2.
The construction of the Salar-Karasu-Dzhun irrigation system in the
second and first centuries B.C. gave impetus t o the development of the apricul-
tural oasis of ancient Tashkent. T h e origin of crop-raising on the territory of the
chirchik-Ahangaran basin dates back t o an earlier period. However, as the
~ ~ z ~ o n - t eTaukat-tepe,
pe, Kugait, Shash-tepe and other archaeological monu-
ments located in the irrigation zone of the Salar-Karasu-Dzhun system show,
[he intensive application of irrigation in that region and the urbanization of a
part of its settled area began at the dawn of the Christian era.' O n e characteristic
feature of the establishment of the Tashkent agricultural oasis is the fact that all
the lands comprised in it were not brought under cultivation at the same time.
Priority was given t o the use of water resources for irrigation areas which were
most favoured b y natural conditions and were, for the most part, situated in
regions adjacent t o the water supply.
Traces of irrigation systems of the Kushan period are found in the upper
Zerafshan, Kafirnigan and Vakhsh river valleys in Tajikistan. The northern and
western sectors of the Vakhsh valley were watered by the ancient Dzhuibar
canal, which was built in the second and third centuries A.D. Remains of this
canal, in the form of embankments 18 m wide and up t o 2.5 m high, have sur-
vived in the region of Urtaboz, extending over a distance of 12 km.'
In the Kushan period, in the Ferghana valley, prior t o the building of the
main canals leading off the Syr Darya, one of the two great rivers of Central
Asia, a complex of fan-shaped irrigation systems providing water for individual
agricultural oases was established at the base of the Isfara, Sokh, Shahimardan
(Margelan), Isfayram, Aravan and other mountain river gorges. At the head of
each system there was usually a large fortress, which provided a vantage point
from which the distribution of water could be strictly regulated. For example,
the Sari-kurgan fortress stood at the head of the Sokh river system. Archaeolog-
ical material indicates that the formation of complex multi-branch irrigation
systems, the rapid expansion of irrigated areas and the emergence of a large
number of fortified settlements in the Ferghana valley all took place in the first
centuries A.D."
The development of irrigation and the expansion of irrigated areas in
Central Asia during the Kushan period have been thoroughly investigated along

5. Kabanov, 1977, 198 1.

6. Buryakov and Filanovich, 1972.
7. Zeimal, 1971.
8. Gulyamov, 1974.
the lower reaches of the Amu Darya and Syr Darya where irrigation was prac-
tised in ancient times. During this period, entire networks of canals were built
and brought into operation in Chorasmia. For example, major canals such as the
Gaukhora, Toprak-kala (right-bank Chorasmia), Khaikhanik, Vadak and B~~~
canals, the left-bank canal originating in Daudan (left-bank Chorasmia) and
many others were all built during this period. It was a time of considerable
growth in the oases of the Bazar-kala and Guldursun canals, which were built as
early as the fourth and third centuries B.C. A large branch canal which was built
off the Toprak-kala canal irrigated the Sultan-Uizdag foothills. The fortress of
Ayaz-kala was erected on the edge of the newly cultivated lands. Following the
reinforcement of the Gaziabad-Chermenyab irrigation system, the Kandum-
kala and Kardarankhas fortresses were restored and new fortresses and towns
built. These included the Zamakhshar (Izmukshir) fortress, Khiva, the Devkes-
kan fortress on Chink of the Ustyurt and many others."
The discovery and detailed study of the remains of ancient irrigation sys-
tems along the lower reaches of the Amu Darya have shown that in the Kushan
period Chorasmia had the most highly developed of all the ancient irrigation
systems (Fig. 1). Progress in irrigation engineering took the form of improve-
ments in the systems of water supply, and made changes in the section of the
main canals. The archaic broad (20-40 m wide) and shallow canals were
replaced during the ~ e r i o dby narrower canals with deeper sections. At the same
time the canals were considerably lengthened, and extended by many kilo-
metres. The number of smaller local systems was reduced and these were amal-
gamated with much larger irrigation systems, shifting the main water intake
further upstream. The process of carrying water t o the fields was improved and
various water distribution devices were introduced. Irrigation was effected in
accordance with a specific flow pattern: main river, head, main canal, distribu-
tion canal, irrigation canal and fields. The total length of one of the largest
canals of the ~ e r i o d ,known as the ancient Kirkkiz canal (right-bank Choras-
mia), was 90 km. It watered numerous fields for cultivation purposes. The sur-
viving portions of a canal of the K'ang-chii period (fourth century B.C. to first
century A.D.) measure as much as 20 m from bank to bank; those dating from
the Kushan period (second and third centuries A.D.) measure only 10-1 1 m, but
have steep sides and are much deeper. The creation and maintenance of major
irrigation systems of this kind obviously called for extensive earth-moving
operations, the installation of sophisticated structures at the head of the system
and constant dredging t o prevent silting up. It has been calculated that over 222
million m of earth were removed in digging the Kirkkiz canal, a task which
took 15,000 labourers two months to complete. Some 6,000-7,000 labourers
were used annually to keep the canal clear of silt and maintain it in working

9. Gulyamov, 1957
 Economy and social system in Central AJra

i mRuins of dwelling-houses and serial

numbers of archaeological explorations -
0 100 200 300 rn

FIG. 1. Irrigation system of ~ n c i e n tChorasrnia. Canals in the environs of Dirgildzhe.

(Courtesy of B. V. Andrianov.)
 A. K. Mukhamedjanov

condition.10 Deep central canals extending over long distances proved more
helpful than the broad and shallow canals of ancient times. S. P. Tolstov, in his
observations on the ancient irrigation works of Chorasmia, concluded that by
late antiquity they had been completely rebuilt. The archaic and classical irriga-
tion systems of the K'ang-chu period were in many respects superior to those
that were fully developed in the Middle Ages."
In the K'ang-chu-Kushan period, when irrigation systems reached their
highest level of development, the area under irrigation along the lower reaches
of the Amu Darya and Syr Darya totalled 35,000-38,000 km2 (13,000 km2 on the
lower Amu Darya and 22,000-25,000 km2 on the lower Syr Darya).lz Thus, in
antiquity, the land area under irrigation along the lower reaches of the Amu
Darya and the Syr Darya was four times greater than it is today. It must,
however, be remembered that the land was not then as intensively irrigated as it
is today. Although the main canals were of considerable size and length, the
network of subsidiary irrigation canals was relatively small and, as a result, not
more than 10-15 per cent of the land area, the irrigation zone, was directly used
for crop-raising, in spite of the substantial supply of water."
In addition to the extensive development that occurred in the alluvial
zones of the major river valleys, the foothills and mountain regions of Central
Asia were also brought under cultivation during the Kushan period, as a result
of improvements in irrigation engineering and the accumulation of experience
in irrigation. Since the water flow in the gorges of these regions was not abun-
dant and the ~ o s s i b i l i of
t ~ expanding the total area of irrigated land was limited,
both groundwater and water from springs, which in those days were far more
numerous, were used for irrigation in addition t o the spring-thaw water from
mountain streams. Depending on the hydrographic and geomorphologic fea-
tures of each river valley and mountain region, different types of hydraulic
works were developed. T o store the limited water from mountain gorges and
springs, small covered reservoirs were built inside a ravine or at the point where
the gorge opened out from it. The techniques used for constructing these minia-
ture reservoirs were very simple. The structures were either rectangular or oval
in appearance, closely resembling the pens used for small livestock. Their sides
were built of boulders ~ a c k e dwith turf and they were located on the slopes of
terraces above the flood- la in. They measured 50 x 40 m at most; the walls were
up to 2 m high and 1-2 m wide. A reservoir usually had small openings in oppo-
site walls. The upper opening was the intake and the lower one was the outlet
for releasing the water into the irrigation network. The use of storage reservoirs
for irrigation was typical of terraced agriculture, and in the Kushan ~ e r i o dit

10. Andrianov, 1969.

1I . Tolstov, 1962.
12. Ibid.
13. Tolstov, 1969.
 Economy and social systcm in Central Asia

common practice in the upper Zerafshan valley and in the fc)()thil]s

Nuratau mountains. Along the northern slope of the Nuratau, a the p(lints
where streams emerge from their mountain gorges, fortified rural
have been identified and recorded, and around them remains of small
with traces of terraced farming have been found. Archaeological cvi-
dence shows that small reservoirs with an average capacity of 1,000-1,200 of
water, and terraced farming using those reservoirs, were introduced in the
mountain regions of Central Asia during the first centuries of the Christian era.14
In mountain valleys where there were no sources of surfacc water,
groundwater was widely used f o r irrigation. It was collected for this purpose in
kahrez, or underground reservoirs, consisting of horizontal water-bearing gal-
leries (which required a great deal of manpower to bore) and a large number of
vertical ventilation shafts. T h e remains of a number of abandoned ancient kah-
rez have been identified and studied in the region of Kopet Dag and Babadug, in
the Vakhsh river valley and along the upper Zerafshan. Archaeological investi-
gations have shown that underground irrigation reservoirs of this kind made it
possible to bring under cultivation a large area of land in the foothills of the
Nuratau region, and a small agricultural oasis was established at the edge of the
Kyzyl Kum Desert. In this period the whole of the upper Zerafshan valley, as
far as present-day Matcha, was converted to agricultural use.15
In rugged mountain terrain, it was especially difficult to select a site for
the head of a canal t o be fed b y a mountain river flowing down a deep gorge,
and to build a canal over land that was extensively broken by ravines. The major
achievements of Kushan irrigation engineering included the boring of tunnel-
like water-intake channels at the heads of main canals that emerged from the
sheer rock sides of a mountain river, and the construction of aqueducts across
ravines or gaps in mountain ridges. Remains of ancient engineering works of
this kind have been identified along the upper Zerafshan, particularl~lin the
locality of Ravatkhadzha, at the head of the Dargom canal, which was built out-
wards from the Zerafshan at the beginning of our era. In the Early Middle Ages
this locality was known as Vargsar, meaning 'head of a dam'. Sogdian irrigation
engineers chose this locality f o r the head of the Dargom canal for two reasons.
In the first place, the Zerafshan river narrows here and is not more than 200 m
wide, whereas upstream and downstream it is much wider - in some places even
2 km wide. Secondly, the river here has very hard banks and the left bank is a
mass of conglomerate rising 15 m. It was of course impossible to build the
opening section of the Dargorn canal through the high solid banks of the Zeraf-
shan, and so the ancient irrigation engineers chose instead t o bore a tunnel with
a number of water-intake openings and wells. O n e of the tunnel openings meas-
uring 1.5 m in diameter still survives at a point slightly above the resent-da).

14. Mukhamedjanov, 1975.

15. Staviskiy, 196 1 .
Pervomaisky hydro-electric power station. The ancient tunnel section of
Dargom canal probably ran almost parallel to the bank. At a later period this
section was eroded by the water passing through it and the ancient water-intake
of the canal merged with the water-meadow of the Zerafshan river.16 this
time also a small settlement was built in the locality of Vargsar, and it was prob-
ably here that the ancient superintendents of the headworks of the Dargom
canal used to live. According to written sources, in the Early Middle Ages the
inhabitants of Vargsar were required t o keep watch on the Dargom canal damas
a labour duty, in exchange for which they were exempted from land taxes.17
that time, about 40,000 people lived in Vargsar,l%hich was always of major
strategic importance as the main water-supply centre for the left-bank sector of
the Samarkand oasis and as a point commanding the approaches to Samarkand.
Whoever held Vargsar could deprive Samarkand of its water supply. In the pol-
itical history of Samarkand, there are numerous examples of attempts by foreign
invaders to destroy the Vargsar dam and so compel Samarkand to surrender.
The rulers of ancient Sogdiana therefore did all they could to strengthen its
defensive capacity, and always maintained large numbers of troops there.
According t o Nasafi, in the Early Middle Ages, Vargsar was defended by an
army of 4,000 men and by 12,000 ghazi o r warrior^.'^ Samarkand's municipal
canal was known as 'Juy-i arziz' (lead canal), since the bottom of the aqueduct
was lined with lead. Judging from the size of the bricksZ0discovered south of
Afrasiab near the Khasret-Khizr mosque, the aqueduct was an arched structure
about 3.8-4 m wide. The site where it was located in the Middle Ages was
known as 'Rasat-tok' or 'Sari-tok' (i.e. head of the arch). In ancient times, the
Samarkand authorities attached articular importance to this structure. Revenue
from land along the banks of the Juy-i arziz, in the locality of Sari-tok, was ear-
marked for the maintenance of the aqueduct and its bridge; and the Samarkand
magi (fire-worshippers) were required, as a labour duty, to keep the structure in
good repair and to guard it the whole year r ~ u n d . ~ '
The development of various types of water engineering works was
undoubtedly attributable t o the very wide ~ r a c t i c a lexperience of irrigation
accumulated over many centuries, t o the enormous expenditure of labour and
t o the application of special water engineering techniques by ancient irrigation
engineers. Tolstov, in his study of the remains of the ancient irrigation works in
Chorasmia, noted that it was re cis el^ during the ~ e r i o dof antiquity that a
school of irrigation engineers and high priests of science emerged at ~horasmia;

16. Mukhamedjanov, 1972.

17. Bartol'd, 1965, Vol. 3.
18. Nasafi, n.d.
19. Ibid.
20. The wedge-shaped bricks mcasurcd 48.5 x 59 - 26 x 8.5 cm; the rcctangulal- bricks
53 x 40 x 9 cm; and thc square bricks 42 x 42 x 9 cm.
2 1. Istoriya Samarkanda, 1969.
 Economy and social system in Centual Asia

it remained in existence until the time of Qutayba's campaign against Khwirizm

(ancient~ h o r a s m i a ) T
. h e school included experts in mathematics, water engi-
neering, cartography, astronomy and calendrical observations, which were of
great importance f o r an extensive irrigation economy.22The brunt of the task of
building irrigation works was, however, borne by the peasants, and many irri-
gation systems were d u g by Labourers from the rural communities, without any
particular expenditure of effort o r contribution by the authorities.
Thus, during the Kushan period, as farming developed and large areas of
land were brought under cultivation, an extensive irrigation economy was
created in the river valleys and agricultural oases of Central Asia, and this
played a major role in the socio-economic and cultural life of the ancient popu-
lation of the country.

Crop-raising and livestock-breeding

Agriculture attained a high level of development during the Kushan period. Its
growth was primarily due t o the rapid expansion of irrigation and t o the fact
that more land was supplied with water and brought under cultivation than at
any other time in the ancient history of Central Asia. In the oases crops were
grown on irrigated land, while in the foothills and mountain regions dry-land
farming was widespread. Also, in the natural wetlands along the river banks,
particularly o n the lower reaches of the Amu Darya, certain crops were grown
on semi-irrigated land.23T h e expansion of farming was, in turn, accompanied by
the development of agricultural equipment and improvements in methods of
cultivation. During the period, iron implements were widely used for the first
time and new types of implements introduced, the hoe being replaced t o an
increasing extent by the plough. The most important step forward in the devel-
opment of farm equipment was the introduction of the wooden plough with an
iron ploughshare, an extremely useful implement that is still used today in Cen-
tral Asia.24T h e magnitude of the total area of farmland, including arable land,
orchards, vineyards, etc., suggests the extensive use of the Such vast
stretches of irrigated land could not have been developed and cultivated with
the hoe alone.
Written sources and archaeological finds indicate that the crops ~ r o d u c e d
during the period under consideration were highly diversified. Different varie-
ties of grain, fruit stolles and other vegetable remains discovered in archaeo-
logical excavations show that the crops produced during the ~ e r i o dincluded

22. Tolstov, 1957.

23. The main crops grown on semi-irrigated land were melons, punlpkins and other gourds.
24. An iron plough-head was found during the excavation of the Tal-i Barzu sitc in
Samarkand.
 A. R. Mukhnmcdjanov

practically all the crops known in the Middle Ages: cereals (millet, barley and
wheat), fruit crops (apricots, peaches, plums, grapes, melons), industria] crops
(poppy seeds),?jfodder crops (lucerne), sesame seeds and pieces of cotton fabric
have been found.'"
Written sources dating from the end of the second century B.C. to [he
beginning of the first century A.D. provide extremely valuable information
about the ancient farming system of the Ferghana valley. They describe Ta-
yiian (Ferghana) as a province with a developed agriculture and specialized
horse-breeding farms. A Chinese ambassador w h o visited Ferghana in 128 n,c,
wrote that Ta-yiian comprised some seventy large and small settlements with a
population of 100,000 who tilled the land, sowed barley, rice and lucerne and
grew grapes.
As the result of a process of selection, transmitted from generation to gen-
eration, various high-yield crops adapted t o local conditions were developed.
should be noted that the Chinese copied the practice of growing lucerne, grapes
and walnuts from the farmers of Central Asia. Evidence of the increased diver-
sity of agricultural crops and of the great size of certain stretches of arable land
is provided both by archaeological finds and by the variety of the cultivation/
irrigation systems and the melon fields identified in the ancient irrigation zone
of Chorasmia. Of particular interest in this regard are the systems used for the
irrigation of vineyards and melon fields in farmsteads west of Dzhanbas-kala.
Here, the alternation of narrow (1.2-1.8 m) and wide (3.3-4.4 m) strips is clearly
visible from the colour of the soil and, in ~ l a c e s from
, the microrelief. At the
edges of the vineyard there are traces of a narrow rectangular building, with a
row of nine large Kushan clay vessels dug into the ground (Figs. 2 and 3). In
one of the buildings a ceramic figurine of a man with a bunch of grapes has been
found, and this, together with other evidence, proves that grapes were once
grown on these fields with alternating wide and narrow strips. A number of cul-
tivatiodirrigation layouts of this kind were brought t o light and investigated in
the neighbourhood of Koy-Krilgan-kala, and many grape-pips and graphic
representations of grape-pickers were found there.27N. M. Negrul, a ~alaeobo-
tanist, has ascertained that the pips came from a variety of grapes used for wine-
making and from large-size table grapes.2"ccording t o archaeological data,
wine-growing was also extensively developed during this ~ e r i o din other prov-
inces of Central Asia, in the Bukhara oasis, in the Ferghana and Merv valleys
and in Parthia. O n e document from Nisa even records the receipt of wine from
vineyards in eastern Parthia,2s and it is no wonder that the Chinese were struck

25. Poppy seeds were found during excavations of the Late K u s l ~ a nsettlelnent of Kzilkir
(Bukhara oasis).
26. Tolstov, 1962.
27. Koy-krilgan-kala, 1967.
28. Andi-ianov, 1969.
29. D'yakonov and Livshits, 1966.
 Economy and sock1 system in Central Asia

FIG.2. Plan of an ancient vineyard in Chorasmia.

by the development of wine-making in the Ferghana valley. Chinese chroniclers

noted the presence of flourishing vineyards and a wine industry in the Ferghana
valley, and recounted that rich Ferghanians stored large quantities of wine and
that old wine preserved its qualities over several decades.
It should be noted that the agricultural oases in the provinces of Central
Asia did not all reach the same level of development during the ~ e r i o dunder
consideration. The ancient agricultural oases, and especially their central areas
where there were irrigation systems with abundant water supplies, were the
most advanced from the agricultural standpoint. In these areas several types of
crops were grown. In areas where regular irrigation was not ~ossible,on the
periphery of the ancient Chorasmian oases and along the lower reaches of the
Syr Darya and the Zerafshan, especially in the north-eastern section of the
ancient Bukhara oasis, in the Karshi and Tashkent oases and in the Ferghana
valley, where there are vast foothills and forest-steppe pasture lands, the popu-
lation engaged in mixed farming. Crop-raising was combined with livestock-
 A. R. Mukhamedjanov

breeding, and only one type of crop was grown, usually barley, millet or
fodder known in Bukhara as alapi-gau.
Both before and during the Kushan period, livestock-breeding played
prominent role in the economic life of the ancient people of Central Asia,
provided draught animals for agriculture and transport, meat, milk and dairy
products for nutrition, and wool and hides for handicrafts. In this period,
according to the written sources and archaeological evidence, cattle, sheep,
goats, horses and camels were bred in Central Asia. In the oases, people kept
livestock in sheds and stables near their homes; in the steppes and foothills, ani-
mals were put out t o graze o n pasturelands; and in the mountainous regions
they g a z e d on mountain grass, a practice related t o the semi-nomadic way of
life of some of the population. Horse-breeding played an important role in the
life of Ferghana. This is clear from the frequent references made by Chinese
authors to large numbers of 'splendid horses' from their reports of the Fergha-
nians' 'prowess in shooting from horseback'. The Aravan petroglyphs of horses
were probably carved during the period under con~ideration.'~ Judging from the
evidence we have of the cultivation of lucerne, it may be assumed that the
inhabitants of the Ferghana valley not only drove their herds of horses out to
graze on mountain pasturelands but also kept them in stalls.

FIG.3, Traces of an anciene vineyard in Chorasrnin. (Gourtcsy of B. V. Andrianny.1

30. Bernshtam, 1952.

 Economy and social systcm in (:cntral Asra

cattle and horses accounted for a large proportion of the bred in

chorasmia; in the B ~ k h a r aoasis, sheep, goats and camels were common; and in
the Tashkent oases, both small and large livestock were raised. The KPang-&"
regardedthe ram as a noble animal. Farn, one of the Zoroastrian gods, was de-
Picted in the form of a ram, and the handles of vessels werc also shaped like
rams, Ferghana horses were especially prized and werc exported in large num-
bers beyond the borders of Ferghana. The two-humped Bactrian was
famousin the countries of the East as a strong pack animal, suitable for caravans
transporting merchandise over the difficult trade routes that crossed the arid
desert. Further evidence of the importance of livestock-breeding in the life of the
population of Central Asia in ancient times is provided by the numerous finds
of statuettes of camels, horses, rams, etc., during the excavation of archaeological
monuments. According t o the estimates of the p a l a e o ~ o o l o ~ i A.
s t B. Bashyrov,
61.6 per cent of the animal bones found during excavations at the Kushan site of
~ar-tepe(Surkhan Darya valley) were remains of sheep and goats, 21 per cent
were remains of cattle, 8.6 per cent were from asses, 4 per cent were from pigs,
2.6 per cent from horses and 2 per cent from camels. It must be noted, however,
that although the inhabitants of Tashkent and Ferghana at that time followed a
settled way of life and were engaged in crop-raising, livestock-breeding and
highly artistic handicraft work, careful study and analysis of written and mate-
rial sources indicate that ancient Ta-yiian (Ferghana) and Chach (Tashkent)
were less developed economically than Parthia, Bactria and Sogdiana.

Handicrafts and building

One characteristic feature of the economy of Central Asia in the first to the
third century A.D. was the considerable increase in handicraft production,
which came t o assume considerable importance in the life of the country. This
was to some extent due t o the development of irrigated agriculture, which pro-
vided the necessary raw materials, and t o the expansion of trade, which opened
up new markets f o r the sale of hand-crafted products. Another contributory
factor was, of course, the rise of the Kushan Empire.
The rich quality of the material culture remains of that period demon-
strates clearly that high levels of development were attained by different
branches of handicrafts such as ceramics, metal-working, iron-forging, weaving,
jewellery-making, etc. People in large towns and small settlements alike prac-
tised a wide variety of handicrafts. Pottery was especially well developed at this
time. Archaeological excavations have brought to light not only large quantities
of ceramic products but also the remains of a whole pottery works containing
several kinds of kilns. ~ ~ ceremonial
t h and table ware of various kinds and
shapes were produced in these kilns. The thin-sided goblets, bowls, cups and
other types of ceramic products from the sites of Afrasiab, Er-kurgan, ~ u k h a r a
and Dalverzin-tepe (Surkhan Darya), from the Tupkhan burial ground (in ~ i ~ -
sar) and from other such laces are notable for their high quality. Many Central
Asian ceramic products of the Kushan period are first-rate examples of the
potter's art.
Almost everywhere there were craftsmen producing metalware and
adornments for women (bronze vessels, candlesticks, mirrors, bracelets, ear-
rings, rings, etc.) and these were very finely made. Archaeological excavations
have brought to light moulds of various shapes for casting metal objects,~l
Judging from the large collection of objects found in the ancient burial
g o u n d s of Bukhara (Lavandak and Kuyumazar, Shuravul) and Hissar (Tupkh-
an), weapons were produced in large numbers. In Central Asia, during the first
few centuries of our era, the commonest type of weapon was the large (up to
1.2 m long) double-edged iron sword, without a tang but with a long, rod..
shaped hilt. Other types of weapons produced included daggers, spears, battle-
axes, slings and bows-and-arrows. O n e weapon extensively used at this time
was a special type of composite bow, pentagonal in shape, the parts fastened
together with strips of bone o r horn. In the Middle Ages, this type of bow was
known in the East as the 'kamin-i Saii' o r 'Sdi bow' (Sai is the Persian form of
the name Cat) and was noted for the distance it could propel an arrow and for
the accuracy attainable. The arrows were made of wood o r reed, the heads being
trihedral with a shank.
The ceramic o r marble bobbins and pieces of cotton fabric that are fre-
quently found at archaeological sites show that weaving was practised. The
written sources tell us that between the shahristan and the citadel of Bukhara at
the Guriyan gate there were large workshops ~ r o d u c i n gcotton fabrics, shawls
and curtains. From the jewellery of every imaginable kind discovered in many
different provinces, it is clear that the jeweller's art was highly developed.
With the growth of handicraft activities and the expansion of trade, the
extraction of minerals also increased considerably during the Kushan period.
Metal ores, semi-precious and precious stones and other minerals were regularly
mined. Mining developed rapidly, especially in the eastern regions of Central
Asia. It is known from the written sources that iron, gold, silver and nephrite
were mined in the mountains of Ferghana and Sogdiana, silver in Ilak, copper in
Karamazar, rubies in Badakhshan and lapis lazuli in Bactria. Some mining rod-
ucts and metal wares were exported.
In the Kushan period, building attained a high technical level. Many
towns such as Afrasiab, Kurgan-i Ramitan, Paikend (in the Zerafshan valley),
Er-kurgan (in the Kashka Darya valley), Termez, Dalverzin-tepe, Zar-tepe,
Khairabad-tepe (in the Surkhan Darya valley), Kanka (in the Tashkent oasis),
Toprak-kala, Kunya Uaz, Ayaz-kala (in Chorasmia), ~ u k h n a - k a l a and

31. D u r i n g the excavations at ancient Merv, traces of large-scale metal production were
found.
 Lconorny and social system in Central Asia

Kum-kala (in the Vakhsh valley) were enclosed by thick walls with rectangular
rowers, The towns and fortified settlements of the Kushan period were built
according t o a preconceived plan and had a very clear and systematic layout.
Many were the administrative and political centres of thc various Central Asian
and ~ r o v i n c e s and
, contained palaces, temples, workshops and dwelling
houses. Public buildings were frequently of monumental size. Palaces and cas-
tles were built o n high platforms and surrounded by strong fortifications. The
massive walls of large chambers with high ceilings were decorated with
and sculptures.
Central Asian fortification engineers were responsible for some major
achievements in building techniques. The strong fortification walls reinforced
by pojecting towers, and the intricate labyrinths with multi-tiered loopholes,
were some examples of major developments in the art of fortification at this
time. Many different building materials were used. Fortification walls and
monumental buildings were built of clay blocks and adobe bricks, which were
usually square. Baked bricks were seldom used. In Bactria stone components
(for example, base columns and capitals, the frieze from Ayrtam) were widely
used for load-bearing structures and decoration. Ceilings were usually sup-
ported by ~ i l l a r sand beams. Where the span was relatively small, arched roofs
were used. T h e largest Central Asian cities such as Bukhara, Samarkand, Ershi
and many others became centres for both handicraft ~ r o d u c t i o nand trade, and
were frequently visited by merchants coming with their caravans from the
countries of Western Asia, India and China.

The coinage and monetary system

The political map of Central Asia in the Kushan period was complex. It is clear
that northern Bactria and the regions along the Amu Darya as far as the middle
reaches of the river formed part of the Kushan Empire. The other provinces of
Central Asia constituted separate domains which, in the opinion of some histor-
ians, formed part of the Kushan state, while others have regarded them as
entirely independent. I t is probably nearer the truth t o say that they were
bound to the Kushan state by some kind of vassal relationship. It should be
noted that most of these territories had their own coinage.
In northern Bactria (south Uzbekistan and south Tajikistan), the appea-
rance of the specifically Kushan coinage was receded by issues of coins (Fig. 4)
that were copies of those minted by the Graeco-Bactrian kings Eucratides and
Heliocles, the commonest being imitations of those minted by Heliocles; they
were issued from about the end of the second century B.C. to the first half of the
first century A.D. O n the obverse was a bust of the king and o n the reverse the
figure of a deity with an inscription in Greek. In course of time the image of
Heliocles was replaced by that of the local ruler and the Greek legend became
 A. R. Mukhamedjanov

FIG.4. Tetradrachm of the Kushan 'Heraus'. (Courtesy of E. A. Davidovich.)

increasingly corrupt. Although these coins were issued in silver, the imitations
were struck in bronze. In size and weight they fell into four groups ranging
from 12-15 to 37 mm in diameter and 2.2-2.3 to 26.5 g in weight.I2
O n the earliest specifically Kushan coins struck by the nameless king
'Soter Megas', the deity was replaced by a horseman and a Greek legend reading
'King of Kings, the Great S a v i ~ u r ' . ~ '
In the reign of Vima Kadphises, a new type was introduced to the coinage
which remained in general use until the Kushan state stopped minting coins.
The obverse showed the ruler standing before an altar, while the reverse bore
the figure of some deity. The deities, however, were rarely of Greek origin;
representations of the Indian god Siva with the sacred bull Nandi are repeatedly
used; and on coins of Kanishka and Huvishka, eastern Iranian gods and god-
desses of fire, wind, sun, moon, etc., are common. Although there were many
Buddhists in the Kushan Empire, the image of Buddha is very rarely found on
coins. I n general, the representations of deities on Kushan coins seem to reflect
the diversity of religious beliefs throughout the vast territory of the Kushan
Empire.I4 Some Early Kushan coins of Kujula and Virna Kadphises had inscrip-
tions in Kharosthi, but once the regular series of Kushan coins was established,
each coin bore a legend in Bactrian only, using the so-called Kushan script
based on the Greek alphabet.

32. Masson, 1956; Rtveladze and Pidaev, 1981.

33. Masson, 1950.
34. Zeimal, 1965, 1967.
 Economy and social systcjm in Cenrral Asia

FIG.5. Coins of Kanishka. (Courtesy of E. V. Zeimal.)

Most Early Kushan coins were of bronze. After the reform introduced bv
~ a d ~ h i s 11,
e s the monetary system was based o n gold staters, o r dinars, which
usually weighed about 8 g, but there were also double, half and quarter coins
weighing 16, 4 and 2 g respectively, though these were more rarc. This was
practically the only example of a gold-based monetary system in the whole of
Central Asia and the neighbouring countries, where in almost every
right up to the Late Middle Ages, monetary systems were based on silver. Gold
coins, with their high purchasing power, were used for major transactions and
especially for international trade, and it was to meet the requirements of inter-
national trade that the gold coins were first produced, copper coins being used
for everyday transactions. They were issued in several denominations, but after
the reform of Kadphises I1 the commonest coin in circulation was the large
bronze 4 drachm (tetradrachm) that originally weighed about 16 or 17 g but
subsequently smaller denominations were also struck (Fig. 5). Large numbers of
bronze coins have been found in nearly every province of the Kushan Empire.
In northern Bactria, f o r example, Kushan copper coins have been found at the
sites of dozens of monuments, and there have been many finds of these coins
even in small rural settlements.'5 It is clear that large sections of the rural popu-
lation as well as towns people were involved in day-to-day commodity
exchanges involving money.
Unlike silver and gold coins, Kushan copper money did not generally cir-
culate outside the territory of the Kushan state, and the area in which copper-
coin finds have been recorded provides a clear indication of the line followed b p
the northern frontiers of the Kushan Empire. Copper coins have been found
not only in south Tajikistan and south Uzbekistan, but also along the Amu
Darya as far as Chorasmia. However, almost all the coins found in Chorasmia

35. Rtveladze and Pidacv, 1981.

 FIG. 6. Chorasmian coins. (Courtesy of V. M. Masson.)

itself had been countermarked, and in the opinion of modern historians, this
indicates that Chorasmia was not part of the Kushan state.jh
Chorasmia began minting its own coinage about the end of the second
century B.c., and for a long time it minted only silver. The first issues were imi-
tations of the Graeco-Bactrian tetradrachm coins of King Eucratides, but gradu-
ally Chorasmia developed its own types. The obverse bore a portrait of the
king, and the reverse the image of a horseman, the Chorasmian tamgha and a
Chorasmian legend (Fig. 6). The first copper coins were issued in Chorasmia at
about the end of the first century A.D., but it was not until the end of the third
century that they were minted in considerable numbers. The obverse portrayed
a horseman or the bust of a horseman, and the reverse normally a monogram.
N o t all coins bore legends. While silver coins had been minted primarily for
~ o l i t i c a lpurposes (proclamation pieces), the extensive issues of copper coins
were a sign that major advances were being made in the economic sphere. The
large numbcr of finds in many rural settlements shows that ordinary day-to-day
trading activity was already widespread. This last remark applies mainly to
right-bank Chorasmia and not Chorasmia as a whole."
Of all the provinces of south Turkmenistan, the most highly developed
from the economic standpoint was the province of Margiana. Parthian bronze

36. Masson, 1975; Va~nberg,1977; Zeimal, 1978.

37. Vaynberg, 1977.
 Economy and social system in Central Asia

FIG.7. Coins of Sanabares. (Courtesy of V. M. Masson.)

and silver coins circulated there before the third century A.D. O n both, the
obverse showed a bust of the king, and the reverse a royal archer seated (Fig. 7).
Early Parthian coins bore inscriptions in Greek which in time became more and
more corrupt, and from the first century A.D. local inscriptions in Pahlavi began
to appear. Although Margiana may have had its own silver coinage, the fact that
it issued its o w n bronze coins, which have been found in large numbers not
only in the ruins of cities but in many rural settlements, is of much greater im-
portance. In the development of day-to-day small-scale trading and commodr
ity-money relationships, Margiana closely resembled B a ~ t r i a . ? ~
In Parthia, another province of south Turkmenistan, the situation regard-
ing the circulation of money was quite different. Although excavations at Nisa
have brought t o light not only Arsacid silver, but also Graeco-Bactrian, Seleucid
Pontic and other silver coins, Parthia had n o copper coinage of its own. This
would seem t o indicate that Parthia was less advanced than Margiana in the
matter of trading and economic development in general.?'
In the Zerafshan valley, several domains issued their own coins. In the
first or second century n.n., Samarkand in Sogdiana began ~ r o d u c i n gsilver
coins with the bust of the king o n the obverse and the image of an archer on the
reverse. Originally these coins bore legends in both Sogdian and Greek, but

38. Masson, 19576.

39. Masson, 1955.
those in Greek gadually became corrupt and were eventually re~laccdentirely
by legends in local Sogdian. At the same time the weight of the d e ~ ~ o m i n ~ ~ i ~ , ,
was progressively reduced from 4 t o 1 g.
In the Bukhara oasis, silver coins were issued from the second century
A.D. They were modelled on the tetradrachm piece of the Graeco-Bactrian king
Euthydemus and bore his profile on the obverse and a seated Zeus on the
reverse. As was the case elsewhere, the Greek legends became increasingly cur-
rupt and were finally replaced by legends in Sogdian. They did not suffer any
significant reduction in weight but the purity of their silver was considerably
debased.
The so-called coins of Hyrcodes were probably minted in the north-west-
ern parts of the Bukhara oasis. The obverse bore a bust of the ruler and on the
reverse was the figure of a horse o r a standing deity. O n these coins, too, the
legends were changed, the weight reduced and the purity of the metal debased.40
There is n o evidence of the minting of coins in the Kashka Darya valley
during the Kushan period. The earliest issues of so-called 'Nakhshab' copper
coins were minted in the Karshi oasis, probably in the fourth century A . D . ~ ~
According to all the available evidence, Chach was the only province in
the Syr Darya region that minted its own coins, the so-called ancient Chach
copper coins with the head of the ruler on the obverse and a seal with a Sogdian
legend on the reverse. This group of coins, which dates from somewhere be-
tween the second and fourth centuries A.D.,.'~ has not yet been adequately stud-
ied. However, ancient Chach coins - and even hoards of them - have been
found at many early sites, showing a well-developed relationship between com-
modities and money."
In Central Asia, in the Kushan period generally, the minting and circula-
tion of money increased greatly, and in a number of provinces, local coins -
local with respect to their iconography and legends - came to replace the imita-
tions of Hellenistic coins. At the same time, the economic development of the
different provinces of Central Asia was very uneven. An analysis of the numis-
matic material indicates that northern Bactria and Margiana were the most
advanced provinces, while Chorasmia, the Zerafshan valley and Chach were
somewhat less advanced. Finally, there were some provinces, such as Ferghana,
that did not have their own coinage and where commodity-money relations
were still in their infancy. In general, however, it may be said that during the
Kushan period there was a developed monetary system with coins of various
denominations minted in large numbers. Copper coins accounted for the greater

40. Zeimal, 1978.

41. Kabanov, 1973.
42. Masson, 1953; Masson, 1966.
43. AS a result of extensive archaeological research in recent ycars, 1,000coins minted in
Chach in various denominations have now bcen found. P~-eviouslyonly a few specirncns
were available.
 Economy and social system in Ccntrul Asia

part of the ~ u s h a nissues, as these were evidently necessary for everyday

buying-and-~eIIing transactions.

Trade and commerce

~~~h internal and external trade and commerce flourished in the Kushan period.
The development of trade and the strengthening of economic tics
above all, from the consolidation of the supremacy of the Kushan Empire, the
of agriculture and the growth of handicraft production. As is
from the mass of archaeological material from various ancient sites of the
trade between the Central Asian provinces increased greatly. Items of
trade included products of handicrafts and agriculture, and both consumer
pods and luxury articles. Consumer goods such as cereals, fruit, textiles, pot-
tery, timber, etc. probably formed items of regular and extensive trade within
the country, which demanded the minting of local coinages in different regions
- Chorasmia, Margiana, Samarkand, Bukhara and Chach - serving as a medium
of exchange in retail transactions.
The agricultural regions of Central Asia were at this time conducting a
particularly vigorous trade with livestock-breeders of the nomadic steppe zone.
They were linked by a trade route that ran along the Syr Darya. This caravan
route, which linked the northern regions of Ferghana and ancient Chach with
the regions of the lower and middle Syr Darya and the Aral Sea area, served as a
kind of two-way transmission line for the agricultural areas.44Cereals, fruit,
handicraft products and weapons were transported along this route to the
nomads of the north; in exchange, furs and skins, meat and milk products, live-
stock and raw materials for weaving were accepted in the south by the seden-
tary peoples. It is not surprising, therefore, that this period witnessed the
growth of major cities in the Syr Darya basin, ruins of which have been found
at Akhsikent,'j Kanka and S h a h r u k h i ~ a , ~Otrar"
" and Dzheti-Asar." Foreign
trade also expanded considerably in this ~ e r i o d .The main trans-Asian trade
routes passing through Central Asia linked the Mediterranean countries with
India and the Far East. Substantial overland trade was conducted with India.
The most convenient route from India passed through the cities of Taxila and
Peshawar, and along the Kabul river valley into Bactria. From there merchants
travelled by boat d o w n the Arnu Darya, over the Caspian Sea and across Trans-
caucasia t o the Black Sea. They also made their way to southern Siberia. The

44. Litvinsky, 1972.

45. Bernshtam, 19.52.
46. Buryakov, 1975.
47. Akishev et al., 1972.
48. Levina, 1971.
Silk Route from China to the Mediterranean countries had a branch ]inking
Bactria to Barygaza (Broach), which had established regular maritime links
the countries of Western Asia. This branch acquired greater importance when
contact between Bactria and the West was suspended because of international
politics. In about 127 B.C. Chang Ch'ien discovered in Bactria some bamboo
articles and textile goods which had come from Szii-chuan via India.
The main exports from India were spices (pepper, ginger, saffron, betel),
perfumes and medicines (sandalwood oil, spikenard, musk, cinnamon, aloe,
bdellium), lacquers and dyes (indigo, cinnabar), silk, rice, sugar, vegetable oils
(sesame, coconut oils), cotton, precious woods (teak, sandalwood, ebony),
pearls, precious and semi-precious stones (diamonds, sapphires, rubies, jasper,
etc.), ivory, exotic animals and slaves.49
A t the same time, India imported precious metals (gold, silver) and non-
ferrous metals (copper, tin, lead, antimony), horses, purple dye, coral, wine,
slaves and artistic pottery and glassware. According t o a report by Pliny the
Elder (XII, 8) dating from the second half of the first century A.D., the value of
imports into India, East Turkestan and Arabia totalled 100 million sesterces.
Some of these imports undoubtedly came from the Central Asian provinces of
the Kushan Empire. Moreover, there is evidence of Bactrian merchants travel-
ling to the confines of the Roman Empire, particularly t o Alexandria in Egypt,
one of the leading commercial centres, and of Roman merchants visiting Central
Asia, where a fairly large number of Roman objects and swords have been
found,50 testifying t o the existence of trade links between the Roman Empire
and Central Asia.
Intensive trade was also conducted during this period with Han China,
which exported silk, nephrite, lacquerware, hides, iron and nickel. Central
Asian merchants exported glass, precious stones and ornaments to China. Lux-
ury goods were the main articles of trade, as was usually the case in ancient
times. The Sogdians played an important role in the development of trade links
with China. In Tun-huang (East Turkestan), letters in the Sogdian language
have been found, dating back t o the early fourth century A.D. (or to the end of
the second century A.D.). O n e of them notes that 100 freemen from ama ark and
were living in Tun-huang. W. B. Henning estimates that the number of Sag-
dians (including slaves and their families) in Tun-huang must have totalled
, The
1,000. Several letters contain information on merchandise, trade, ~ r i c e setc.
Sogdians living in East Turkestan maintained close contact with their home
town in Samarkand.5'
During the period under consideration, the rulers of different countries
and provinces played an active role in international trade and enjoyed a monop-

49. Pigulevskaya, 1951.

50. Staviskiy, 1964; Masson, 1966.
51. Hcnning, 1948.
 Economy and social system in Ccntral Asra

oly of trade in certain goods. They used to dispatch their ambassadors with
large quantities of merchandise and valuable gifts, and formed their own trading
guilds. For example, in Book 2 of the Mahabhirata (second to fourth century
h . ~ )there
, is a reference t o gifts brought t o Yudhklhira, the King of the Kurus,
at ~ ~ d r a p r a s t h(the
a site of modern Delhi) by emissaries of various peoples,
among them Central Asians. From Vahlika (Bactria) came 'woollen blankets, of
good proportions, beautifully dyed, pleasant to the touch', various fabrics,
weapons and precious stones, and the Sakas and Tocharians used to
bring horses 'capable of covering long distances' (Mahibhirata 11.47).

The Silk Route

A major role in the development of international trade during the Kushan
pied was played by the Silk Route, the main trans-Asian caravan route,
which, from the second century B.C. onwards, linked China, India and Ccntral
Asia with the countries of the Mediterranean. It owed its name t o the fact that
the principal commodity carried was Chinese silk. The Silk Route began at
Ch'ang-an, the capital of China at that time, and ran westward along the edge of
the Gobi Desert, passing through Lan-chou to Tun-huang. At Tun-huang, it
divided into two, one branch going south and the other north. The northern
route followed a straight line from Tun-huang to Turfan, crossing the sand-
dunes of the White Dragon salt desert, which at one time had been part of the
Lop Nor lake bed. That was the most difficult stretch of the Silk Route, and the
trade caravan guides - usually Sogdians or Bactrians - preferred to bypass the
sand-dunes of the White Dragon and make a large detour to the north o n the
way to Turfan. From Turfan the Silk Route went through Ch'iu-tzti into Aksu,
then from Kashgar t o Ferghana via Samarkand, and on to Antioch in Margiana.
The southern route f r o m Tun-huang went via Khotan and Yarkand to the capi-
tal of Bactria, and then t o Zariaspa and Antioch in Margiana, where the two
roads joined. F r o m Margiana the Silk Route ran west t o Hecatompylos, the
ancient capital of Parthia, and thence to Media, Ecbatana and Mesopotamia, and
across the Euphrates t o the ancient ports on the eastern coast of the Mediter-
ranean (see also Chapters 16 and 19).
There was a constant struggle between the Chinese and the Central
Asians, and between the Parthians and the Romans, t o establish control over the
Silk Route and so dominate international trade. As early as the first century
B.c., Han China took control of the eastern section and launched a military
campaign against Ferghana. From that time onwards, China had direct trade
relations with Bactria. According t o Szti-ma Ch'ien, from the years 115-1 14
B.c. onwards, more than ten missions a year were sent from Ferghana to the
West. Caravans made their way unimpeded t o Bactria, India and Sogdiana,
reached Parthia and penetrated even further west. The seizure of the silk Route,
which lnade it possible to maintain regular and direct contacts between H~~
China and the states of Central Asia and the West, laid the foundations for cul-
tural and trade exchanges. From Central Asia, China received grapes, lucerne
beans, pomegranates, saffron and nut trees; the acquisition from Ferghana of the
war-horses needed for the new Chinese cavalry was of particular importance,
Parthian merchants tried t o prevent the establishment of direct trade ]inks
between the Roman Empire and China. Merchants from the Kushan Empire
also competed with the Parthians and tried to become major intermediaries. The
basic means of transport in this international caravan trade was the camel. The
accounts of travellers suggest that some of the most difficult stretches of the Silk
Route were passable only because of the superior qualities of the Bactrian two-
humped camel.

Social structure
Very little is known of the social structure and types of land-ownership in Cen-
tral Asia under the Kushans. The Kushan Empire was one of the great powers
of the period. It comprised a large number of countries with different social
structures. It included fertile agricultural oases with many commercial and
handicrafts centres and rural settlements as well as vast steppelands and moun-
tain regions. In the towns, slave-owning systems existed, while in agricultural
regions freemen in communes preserved in their way of life many aspects of
tribal-clan relations. Such relations were particularly common among the live-
stock-breeders who lived in the steppe and foothill regions of Central Asia.
Before the establishment of their empire, the Kushans had been a relatively
small nomadic tribe and long preserved many of their own traditions even after
they had settled in Bactria, but once they had become rulers of a huge empire,
their patterns of social organization changed considerably. Detailed analysis of
archaeological material (especially the types of settlements and material
remains) shows that in the Kushan period there was considerable variety in
social status and property ownership, patterns which subsequently spread to
virtually the whole of the territory of Central Asia. O n the local coins minted in
Central Asia and in the 'Ancient Letters' and other Sogdian written documents,
a wide range of terms is used t o denote different social groups in the Kushan
period and the era immediately preceding it.
There is some direct, and a great deal of indirect, evidence to show that
the commune occupied an important place in the socio-economic life of central
Asia and in the ancient East as a whole.52 This seems to have continued until the
Early Middle Ages, for which evidence is available. Thus, the commune in Sag-
diana was known as ndfi it consisted of the aristocracy (dzdt, dzdtkdr), mer-
 Economy and social system in Cctltral Asra

(xvakar), and free peasants (who were members of the commune) and
craftsmen( k i ~ i k a r )O. ~f ~these three categories in the nif, the highest status was
by the i z a t , that is, persons of 'high and noble birthy, the i z i r k i r , or
free persons associated with the azat, and the 'children of the a ~ a of t aristo-
cratic,noble origin'.54 According t o the written sources, the i ~ a t the land
the villages and were the chief retainers of the local and provincial rulers.
Next came the xvakar, o r merchants, who constituted one of the proper-
tied classes. T h e third category consisted of the kririkir, who paid a and
were not regarded as noblemen. At that time there were certainly slaves and a
dependent, subject population. The Sogdian 'Ancient Letters' contain terms
such as 'bandak' (slave) and 'daya' (bondwoman). A fairly complete picture of
the composition of ancient Chorasmian classes and their use of slave labour in
the economy is provided by documents from the Toprak-kala palace archives.
These give the names of the heads of 'family households' and of 'house-owners',
their sons, sons-in-law and slaves. The roll of the 'House of Gaviimava' (Docu-
ment No. 8) listed a total of twenty-one males: the house-owner, his two sons,
his son-in-law and seventeen slaves - including twelve slaves serving the house-
owner, his sons and son-in-law, two in the service of their wives, two t o look
after the young grandchildren and one t o look after the concubine of the master
of the house. T h e 'House of VavanSira' (Document N o . 7) had seventeen males:
the master of the house, his son-in-law and fifteen slaves, including twelve
attached t o the master of the house himself. The families described in these doc-
uments were very prosperous, as is clear from the large number of slaves in pro-
portion to the number of free adult males.55
In spite of the very considerable number of slaves, slavery was not the
only, and probably not the predominant, form of labour. Little use was made of
slaves either in agriculture o r in handicraft work, as their labour was not ~ r o f i t -
able.56

Unfortunately, historians d o not yet have at their disposal concrete material o n

many of the most important aspects of the social and economic history of Cen-
tral Asia during the Kushan There is virtually no information from Cen-
tral Asia proper o n the different categories of land-ownership.
There is, however, some direct, and a great deal of indirect, evidence that
suggests that there were several different categories. The documents from the

53. Smirnova, 1970.

54. Henning, 1948.
55. Gudkova and Livshits, 1967.
56. Gafurov, 1972.
archives of O l d Nisa, which ~ r o v i d esome insight into the economy of southcr,,
Turkmenistan during the first century B.c., are particularly valuable. There was
one category known as uzbar land. The ~ z b a o~r ,levy, was already known in
the Achaemenid period as revenue directly received from royal land. A number
of estates consisting ~ a r t l yof vineyards belonged to this category. These estates
- about a score of them are known - were largely dastkivt, or royal estates. The
same estate might also contain patbazik land. In Achaemenid times the term
' p a t b ~ z i k 'meant the delivery t o the king of a contribution in kind, consistingof
fruit and types of produce.57 It is highly probable that a certain proportion of
irrigated lands in the Central Asian oases belonged t o temples. A special priestly
class, who is attested in the area long before the Kushans, also probably Pos-
sessed land during this ~ e r i o dThe . medieval name 'vagnze' was quite common
in Central Asia; it was linked with the Sogdian term 'payan' (temple) and rob-
ably meant land belonging to temples o r shrines. Besides the royal and temple
lands, there were private and communal lands. In all probability, there was
more land under communal ownership than any other type. There is some evi-
dence t o show that communes owned whole irrigation systems and the regions
irrigated by them, as well as settlements and grazing lands. Localities settled by
rural communes were called varzana, vardana o r gava, meaning village or rural
district, and it was precisely at this time that the fortified settlement of Var-
danze, in the northern part of the Bukhara oasis, was established. Unfortu-
nately, there is almost n o specific material on communal land-ownership; but it
seems probable that the commune during this ~ e r i o dwas intensively exploited
by the state and large land-holders, who tried t o attach members of the com-
mune to the land - a process that ultimately led t o the emergence of feudalism
in Central Asia.

57. D'yakonov and Livshits, 1960a.

 B. A. Litvinsky

The development of urban patterns

The evolution of urban life in the territories that formed part of the Kushan
state, or were subject t o its political o r cultural influence, can be traced back to
the Bronze Age. During the time of the Achaemenids, urban planning and
architecture were strongly influenced by West Asian styles. Hellenism had an
even greater impact o n town planning in Central Asia after the establishment of
Greek cities in the area. This was the time when cities began to amass so much
economic power that they became an important element in the power of the
state. In the Kushan period that followed (between the first century B.C. and the
fourth century A.D.), Central Asian, Hellenistic and Indian town planning
blended into a single form. During this period, the ancient cities grew faster
than ever before and urban life flourished.' Archaeological excavations provide
important evidence about the cities of the Kushan period - their layout, archi-
tecture and material culture. O n e of the most thoroughly studied sites is Taxila
in north Pakistan, where John Marshall carried out extensive excavations be-
tween 1913 and 1934.
The Mauryan city of Taxila o n the Bhir Mound was replaced under the
Graeco-Bactrians in the second century B.C. by a new city at Sirkap, which
remained in use u p t o the Early Kushan period. The city at Sirkap has the shape
of an irregular trapezium, stretching 1,300 m from north t o south and 900 m (at
its widest) from east t o west. Topographically, it is divided into two unequal
parts - a lower northern and an upper southern city. Remains of walls along the
dividing line between them still survive. The city was intersected from north to
south by the main street with side-streets running off at right angles. Each of

:' See Map 6.

1. Litvinsky, 1973.
 R. A. Litvinsky

FIG. 1. Sirkap. Location of shops ( 2 ) . (Photo: Music GuimetlTissot.)

the spaces between the side-streets (which were 36.5 m o r slightly more apart)
contained blocks of buildings, occasionally divided by small alleyways. Both
sides of the main street were lined with shops (Fig. I), as well as some shrines,
especially stupas. Behind the shops and the shrines were the dwelling houses.
East of the main street was the royal palace and, near by, some more opulent-
looking two-storey dwellings. I n the city and the surrounding areas, there were
Buddhist stupas (Figs. 2 and 3), monasteries and shrines. Some 650 m outside
the north gate was the non-Buddhist Jandial temple.2 Early under the Kushans,
the city was again transferred to a new site at Sirsukh (Fig. 4). This new Kushan
city, founded under the nameless king Soter Megas, covered an area of 1,370 x
1,000 m, but has not yet been excavated.
Shaikhan Dheri, the second city of Charsadda, was laid out in a similar
manner. The city was divided by a network of parallel streets some 36.5 m
apart. Between the two central thoroughfares in the city centre was a sanctuary,
probably a Buddhist stupa, and in between the streets were blocks of buildings.'
Subsequent excavations have established that this city was occupied from the
second century B.C. to the second century A.D.-'
Bhita is the modern name for the ruins located 16 km south-east of

2. Marshall, 1951, pp. 112 et seq., 139 ct scq., 1960, yp. 60 ct sccl.; A. Ghosh, 1948, pp. 41 et
seq.
3. Wheeler, 1962, pp. 16-17, Plates XV-XVI.
4. Dani, 1955156, pp. 17 et seq.
 C I I Z Land
' ~ urban lzfc rn thc Kuihan krnKrlom

FIG.2. Sirknp. Stupa IA betwren the second and third m e e t cart.

(Photo: hllus6c Guin-lcr/Tissot.)
 B. A. Litvinsky

FIG.4. Sirsukh. Wall running east t o south. (Photo: MusPe Guimet/Tissot.)

Allahabad. From seal inscriptions the settlement seems to have been known in
antiquity as Vichi. Excavations by Marshall in 1909-12 showed that the city
covered an area of about 26 ha, and was surrounded by a fortification wall 3.4 m
thick by 12 m high. The city area was traversed by straight parallel streets, one
of which, 9 m wide, the 'Main Street', began at the city gates and led to a sanctu-
ary in the centre of the town. Another, half its width, which Marshall called
'Bastion Street', ran directly parallel to the Main Street at a distance of 45 m.5
Although the houses on both streets had identical floor plans, those on the
Main Street were noticeably larger. In the spaces between the parallel streets,
there must have been two rows of two- o r three-storey houses each accommo-
dating between ten and twenty occupants, family members and servants
included. It has also been estimated that the city had about 940 such houses and
a population of between 10,000 and 20,000 ~ e r s o n s In
. ~ the block of buildings
on the south-west side of the High Street was a house (14 x 13.4 m), consisting
of a rectangular courtyard flanked by twelve rectangular square rooms. The
house had two entrances on opposite sides (north-east and south-west) each set

5. T h e foundations of the buildings in the city date from the Mauryan period, but many of
the surviving structures o n the Main Street and Bastion Street were built and existed
during the period from the first century n.c. t o the third century A.I).Tliis was the time
when the city had a network of parallel streets. The buildings constructed in the
post-Kushan period were not lined up on the same axis. Thus, Item 50, a Gupta temple in
the centre of the city, and items 43 and 45, f o u r t h - c e n t u r y - ~ . ~housing
). in the north-west
part of the city, are all oriented at an angle in the axis of the above-mentioned streets (set
Dani, 1955/56, pp. 40, 43).
6. Marshall, 1911, pp. 1 2 7 4 1 ; Schlingloff, 1970, pp. 24-7.
 Cities and urban life in the Kushan kingdom

"ear the longitudinal axis. In one corner room, there are the foundations for
four columns, and judging by the thickness of the walls, part of the building
may have had a second floor. Marshall was of the opinion that this house could
have been built in the Mauryan period. A seal found under the foundations
and, therefore, belonging t o an earlier period bears an inscription that Marshall
read as '~ahijitiyenlgamafa', prompting the suggestion that thc earlier house
could have served as the office (nigama) of a guild, though the reading has sub-
sequently been disputed.
From the seals found, it has been possible to identify the names of the
owners of different houses. O n e belonging t o Nigadeva and built in the first
century B.c., mostly of burnt bricks, had a section facing the Main Street which
Marshall thought was a shop. Flights of steps, flanked by platforms on both
rides, led from the street side t o the central rectangular room of the shop. O n
either side was a much smaller room forming a lateral wing, and all three were
built in a row along the street. Behind were the courtyards, on the farther sidc
of which were the living quarters (11.3 x 10.4 m), an inner courtyard sur-
rounded by structures mainly rectangular in design. The house was a self-con-
tained unit, separated from the surrounding buildings. It had its fa$ade on the
High Street, with lanes o n the other three sides and an additional doorwav
opening o n t o one of them.
Built in the first century B.c., this house remained in use throughout the
Kushan period (seventeen coins from the reigns of Kanishka and Huvishka
were found there). In a neighbouring house with a similar layout, an ivory seal
was discovered in the fourth-fifth-century stratum bearing the inscription
'Sreghi ~ a ~ a v a s u d awhich
', Marshall interprets as 'the banker Jayavasuda',- but
the person concerned, presumably the owner of the house, could very well have
been the elder of a guild. Like others in Bhita, this house was surrounded by
lanes. The same was true of houses in Vaiiili, Rijagha, Kolhapur, Sambhar and
other cities: each house was surrounded by narrow alleyways separating it from
neighbouring buildings. According to the written sources, these alleys were
three paces wide.8
In Sisupalgarh (ancient Kalinganagara), where the ruins of the ancient cit!~
cover an area of about 1.36 km2, the ramparts (10 m thick) enclose an area
almost perfectly rectangular, with a bastion at each corner. As La1 has noted,
this layout calls f o r a regular network of streets running from east to west and
north to south, intersecting each other inside the city.' The same system was
adopted in other cities in India, for example, Udegram.''
O n e of the most famous cities in the Kushan Empire was Begram, north

7. Marshall, 1915, pp. 36-48.

8. Schlingloff, 1970, pp. 27-8.
9. Lal, 1949, pp. 62-105.
10. Gullini, 1962, pp. 173 et seq.; Faccena, 1964, pp. 14-23.
 FIG.5. Begram. Decorated ivory plate, first/second centuries B.C.
(Photo: UNESCO/L. Hammerschmid.)

of Kabul, at the confluence of the Panjshir and Ghorband rivers. The city was
rectangular in shape, extending 800 m from north to south and 450 m from east
to west with a citadel in the north-east. The stone foundations (0.5-0.7 m high)
of the city walls were set into the subsoil, supporting the main section of the
wall constructed of square sun-dried clay bricks. Square towers were built along
the wall, at intervals of 17 m, and in front were two parallel ditches. A central
thoroughfare divided the city into two parts, and it is assumed that there was
another thoroughfare at right angles dividing the city into quarters. In the
palace in the southern part of the city a number of storerooms were discovered
containing hundreds of articles of carved ivory (Fig. 5) brought from India, and
Western objects of Roman date imported from the Mediterranean. The excava-
tions yielded a large collection of articles of material culture."
In south Uzbekistan, in Bactrian territory, a large city has been excavated
at Dalverzin-tepe. The main portion, tentatively called the 'lower city', formed
a rectangle 650 x 500 m. In the south corner, partly extending beyond the city
boundaries, is a citadel shaped like a rounded trapezium (maximum measure-
ment - 170 x 200 m). Outside the city walls were a Buddhist shrine and necro-
polis, and a Zoroastrian chapel (naur).The 'lower city' was surrounded by thick

11. Ghirshman, 1946; Hackin, 1954.

 Cities and urban fqe in the Kushan kingdom

ramparts with towers at 30-40 m intervals. Outside the fortifications, as a

furtherprecaution, were canals, a river-bed and a ditch. The only gate was in the
southern section near the citadel. In the Kushan period the city was dense]y
built with large blocks of buildings, urban thoroughfares and water reservoirs.
Houses belonging t o the aristocratic section of the population were situated in
the heart of the city, while those belonging to the poor were built on the out-
skirts. The southern quarters were inhabited by craftsmcn near kilns and pot-
tery workshops. It was there, too, on high ground that the temple of the Bac-
trian goddess was found. TWOpalatial dwellings (DT-5 and DT-6) had an
impressive structure decorated with columns with Attic-style bases. Con-
structed with vaults and arches of sun-dried clay-brick, their principal faqades
were embellished with a deep portico bounded in front by columns. Behind the
on the principal axis, was a large vestibule with a reception hall beyond.
The front part of the building, reserved for receiving guests, was separated from
the living quarters b y a corridor. All the dwelling-houses in Dalverzin-tepc
(ordinary as well as palatial) had one feature in common - a special room set
aside for household prayer, with a niche for kindling the holy fire. At Dalver-
zin-tepe some outstanding works of art were discovered, including many pieces
of secular and Buddhist sculpture and paintings (Fig. 6) and a most remarkable
treasure of 115 gold objects of jewellery, works of art and gold bars with
inscriptions in Kharoghi, indicating their weight.I2
At Toprak-kala in Chorasmia (Fig. 7), the rectangular site (2.5 km2 in
area) running from north t o south is surrounded by a wall with many square

2-
-\

FIG.6. Dalverzin-tepe. Fragment of wall painting showing the head of a goddess.

(After Pugachenkova and Rtveladze, 1978, p. 48.)
 FIG.7, Toprak-kala (Chorasmia). (Photo: O Vladi~nirTerebenin.)

towers. In the north-east corner was a huge castle for the ruler, with a large
courtyard and a triple-towered keep, the remains of which rise t o a height of 25
m. South-east of the castle was a building containing a large central area with a
corridor running round, probably a fire temple. The residential area was
bisected by a main thoroughfare running from north to south (where the city
gate had a huge protective structure in front of it). At right angles to this tho-
roughfare were streets that divided the city into ten symmetrical blocks. Al-
though the top stratum dates from the fourth-fifth centuries A.D. (and in a few
sections from the sixth-eighth), the city was originally laid out in the second-
third centuries. Each insula measures 40 x 100 m and the street widths are 4.5 m
and 10 m (in the case of the main artery). The buildings of the fourth and fifth
centuries A.D. formed part of a large building complex developed at one time. It
is not clear whether its large units formed part of a single architectural ensemble
or constituted separate households. Small groups of two or three buildings -
some of them craftsmen's dwelling houses - werc found in the outer blocks at
Toprak- kala.
The palace, situated in the castle inside the square formed by the outer
walls, had over 100 rooms on the ground floor, and there are rcmains of more
rooms on a first floor. Three stagcs in its existence from the second-third to the
 Cities and urban fqe in the Kushan kingdom

lourth-fifth centuries A.D. have been identified. The palace a large

number of works of art (paintings, sculptures, etc.), and the 'Hall of Kings'
alone contained 138 statues. This building, which dominated the who]e corn-
must be regarded as a holy palace because of its sanctuaries associated with
various aspects of the royal cult. Adjoining the north-west section of the
Toprak-kala site is the 'north complex' (250 x 400 rn), which has an amplified
layout and contains a number of imposing structures. Remains of bas-reliefs,
and wall-paintings have been found in the halls. The monumental
nature of the buildings, and their layout and decorations, suggest that it was an
open palace built at the same time as the palace in the citadel. O n the north-west
side was a rectangular undeveloped plot of land surrounded by an embankment
(perhaps a park o r a necropolis).I3
Mention should also be made of city-sites such as Zar-tepe, Kei-Kobad-
shah, Er-kurgan, Saksan-Okhur, etc. O f the various sites of the same period
which have been thoroughly investigated by archaeologists, the Kara-tepe and
Fayaz-tepe Buddhist complexes at Termez deserve special mention. The find-
ings resulting from excavations undertaken at Dilberjin (Fig. 8) by a joint
soviet-Afghan archaeological expedition14 are also of considerable importance
for studying the history of the Kushano-Bactrian cities.

City life in the Kushan period

From available evidence it is clear that life in the cities of Central Asia from the
first century B.C. t o the third-fourth centuries A.D. was incomparably more
intensive than that in the preceding period. This is illustrated by:
The quantitative growth of the network of urban settlements and the emergence
of new cities that had never existed before (at no time in the ancient his-
tory of Central Asia had there been so many cities).
The enlargement of urban areas in the old cities that had existed earlier and the
increased density of urban c o n s t i ~ c t i o n . ' ~
Further development of the division of cities into three parts: citadel, city
proper and suburbs (which, in addition to houses and workshops,
included religious buildings, especially Buddhist shrines, temples and
cemeteries). Side by side with the cities which had this tripartite pattern
there were also cities of other types, many of them lacking a citadel.
Fundamental internal socio-economic changes in the urban organism and the
increase in the importance of the city in the economic life of the country,

13. Tolstov, 1948, pp. 119, 123, Fig. 62, 1962, pp. 204-6; Rapoport, 1 9 8 1 ~ .
14. Kruglikova, 1982.
15. The reference here is to a general trend; in certain specific cases, this trend was not
evident.
1 Bench mark: height Excavation numbers

I Remains of construction Pottery ovens

- - - - --

FIG.8. Plan of Dilberjin-tepe. (After K r q l i k o v a , 1979, p. 121.) (Photo: V. N. Yagodin.)

- Cities and urban l f c in the Kushan kingdom

resulting ~ r i m a r i l yfrom the rapid growth of urban handicrafts. The

became centres f o r the production of commodities for sale, hence their
key importance in the city-village-nomadic-steppe system. With the con-
centration of religious buildings within cities the latter also m
increasingly important role as centres of ideological life.
These conclusions, based o n material relating to Central Asia, can - as is clear
from the available evidence - be applied also t o other territories of the Kushan
state. AS Dani and Khan note:

The urban centres increased t o a very large extent during the Kushan period. In thc
main valley of Peshawar all such cities lie to the north of the Kabul River along the
old route that came f r o m Taxila and across the Indus to Hund or Salature (present-
day Lahur in Swabi Tehsil) onward to Puskalivati (present-day Charsadda) at the
confluence of the Swat and Kabul Rivers. Here the routes diverged in various
directions. If the city mounds that exist today on these routes are countcd, it is not
surprising t o note that urbanization even in modern Pakistan has not rcached that
stage in the Peshawar region. This urbanization in the Kushan period was based on
industrial development and o n trade entrepots.'"

Ghosh, too, has noted that 'the Kushan Empire comprised many cities in the
Panjab and the Gangetic Basin'.)'
The role of the city in military operations can hardly be overestimated.
Cities were well fortified and some were virtually impregnable. The fortifica-
tions were designed t o make the best possible use of the characteristics of the
terrain, and were supplemented by deep ditches (one or two rows), forward
outposts and thick walls with rectangular (more rarely, round) towers, parapets,
etc.l"Together, the fortified cities formed the defensive backbone of individual
provinces and of the entire Kushan state. Thus, cities became vital components
of the whole infrastructure. T o mention only the case of Bactria, in Surkhan
Darya province, some 110 monuments have been recorded, most of them situ-
ated in river valleys. T w o o r three are of Achaemenid date, about twenty belong
to the Seleucid and Graeco-Bactrian periods and seventy or eighty belong t o
Kushan times."

City planning
In the Kushan period, both in Central Asia and in India, cities were still predo-
lninantly rectangular in shape,'O though a few had other shapes: trapezoidal,

16. Dani and Khan, 1974, p. 102.

17. A. Ghosh, 1975, p. 109.
18. Francfort, 1979.
19. Rtveladze, 1978, p. 114.
20. Filliozat, 1959, pp. 251-2; ~ c h l i n ~ l o f1970,
f , pp. 45-6.
 B. A. Litvinsky

semi-circular, circular o r polygonal. Several newly founded cities, and some dat-
ing from earlier periods, were extremely large; but there were also small- and
medium-sized towns. Some cities (the new ones in particular) had no citadels)
while others had large ones. Indian sources contained a highly developed termi-
nology for describing various types of cities." On the basis of archaeological
material, cities as organisms can be classified only in external and quantitative
terms, that is, in terms of their general layout, component parts, shape and size.
O n the basis of such material alone, it is extremely difficult t o describe the most
important features of urban life, ranging from the principles of urban planning
to details of municipal administration. When written sources are used, the situ-
ation is quite different. Although information about cities in Middle Asia is
scanty, for ancient northern India there are many epigraphical and literary
sources (the ArthaiZstra, the Milindapasha, the epics, the Jaina canon, the ]&a-
k a ~ special
, architectural treatises and others), dating back to the end of the first
millennium B.C. and t o the first half of the first millennium A.D. and containing
various kinds of information o n economic and social history and especially on
the history of the city.22In view of the parallel development of urban societies,
analysis of Indian sources is especially interesting2'
The Arthaiiistra states that, in selecting the place for building a fortress or
a settlement, it is important to take into account the features of the terrain, and
the final choice must be 'approved by architects'. 'The city must be strongly for-
tified: there must be three rows of moats filled with water, a rampart, walls with
square towers, etc. The city must be traversed by three roads running from
north t o south and three running from east t o west, and four of the twelve city
gates must be main gates. Within the city, the siting of various buildings - from
the palace and temples down t o the dwellings of craftsmen - is subject to strict
rules (Kautilya Arthaiastra 2.3.1-32; 2.4.1-32).2"he layout of streets and resi-
dential areas must be carefully planned - Suvibhata (Ramayana 1.5.8; 1.5.10:
V, 53, 20 etc.; Mahabharata 1.199.34).
The Milindapanlha2' gives a detailed description of the development of an
ideal city:

A city architect, when he wants t o build a city, first looks about for a district that
is level, not elevated, not low-lying, free f r o m gravel and stone, secure, irreproach-

21. A. Ghosh, 1973, pp. 45-6.

22. It was thought that sornc works from the p s t - K u s h a n period ,night also usefulls be
included here.
23. Litvinsky, 1979, 1981; Litvinsky and Scdov, 1983.
24. Kangle, 1972, pp. 66-72.
25. Horner, 1964, Vol. 11, pp. 170-1, cf. thc Manusmrti (Sanskrit code of laws), Val. VIL
pp. 69-74; the Ramayana 1.5.7; also the descriptiolls in the Jaina canon Uaina sutras),
Vol. I, pp. 252-3; and Schlingloff, 1970, p. 7.
 Cities and urban lifr in the> Kushan kingdom

able and delightful, and then when he has had made levcl there what was not levr]
and has had it cleared of stumps of trees and thorns, he might build a city therc.
Fine and regular [it would be], well-pla~incd,the moats and encircling walls dug
deep, the city gates, the watch-towers and the ramparts strong, the cross-roads,
squares, junctions and the places where three or four roads meet numerous, the
rnain-roads clean, level and even, and bazaar-shops well laid out, [the city] full of
parks, pleasances, lakes, lotus-pools and wells adorned with a wide variety 111
shrines to devas, the whole free from defects.

The description of the ideal city has much in common and in Inany ways is
identical with the description of Sagala (modern Sialkot). From the Milinda-
patiha, we also learn that the city gates had watch-towers. The city was encir-
cled by a deep moat and surrounded by walls. Among the urban roads, special
mention is made of the carriage-roads. The city had a large number of shops,
thousands of richly decorated buildings and 'hundreds of thousands' of dwell-
ing-houses.
The architect-builder 'plans the distribution of the carriage-roads, the
squares and the places where three o r four roads meet'. We learn that the city had
a special inspector w h o sat at a cross-roads in the middle of the city, from where
he could see anyone approaching from the eastern, southern, western o r north-
ern quarter of the city. From other ancient Indian sources (the Jitakas), it is
known that the city had a special official, the dovirika, t o shut the city gates at
night and also t o show the way t o stranger^.?^ The Milindapal?ha also provides a
vivid picture of the city and its streets swarming with 'elephants, horses, chariots
and pedestrians, with groups of handsome men and women; it was crowded with
ordinary people, warriors, nobles, brahmans, merchants and workers' and a
variety of ascetics." Alongside the carriages, riders on horseback moved along
the streets.2"here were many strangers in the cities - people from other prov-
inces of India and from Scythia (Saka), Bactria (Yavana) and China (Cina).2vThe
shops were overflowing with goods. Some sold Benares muslin and other fabrics.
From others came the sweet smells of flowers and ~ e r f u m e offered
s for sale. The
jewellery shops were filled with items of silver, bronze and stoneware, the store-
houses were full of goods of various kinds including foodstuff^.^^ The streets
swarmed with hawkers of herbs, fruits and roots, and meat, fish, cakes and other
different kinds of foods were offered for sale. Anyone with money could drop
into an eating house f o r a bite. Here and there street actors, conjurers and acro-
bats gave performances, o r professional wrestlers were locked in combat.'

26. Fick, 1920, p. 157.

27. Horner, 1964, Vol. I, pp. 1-2.
28. Ibid., pp. 171-2.
29. Ibid., Vol. 11, pp. 4-5.
30. Ibid., Vol. I, p. 2.
31. Ibid., Vol. 11, pp. 171-2.
 Other Indian literary works give an even more colourful and vivid picturc
of life in the cities of ancient India. The Umbhayibhisarika describes the city of
K ~ s u m a p u r awith its clean streets and canals enclosed between rows of houses,
Mountains of flowers (sacrificial offerings by devout city-dwellers) were heaped
along the streets which were lined with shops where various kinds of goods
were offered for sale. Occasionally, white-faced women were seen glancing out
on the streets, opening the windows of palaces as high as the clouds. Finely
dressed royal officers went about their business o n horseback, on elephants or
in carriages."
The Pidatdditaka recounts that the streets of the city resounded with
songs, the jingle of women's jewellery, the monotonous sing-song tones of pea-
reciting and studying the Veda, the chopping sound of axes in
shops, the clatter of dishes and the screeching of domestic birds. The city was
teeming with local townspeople, visitors from different provinces of India as
well as foreigners, including the Sakas, Yavanas and T u ~ a r a s(Tocharians). This
text also mentions that some inhabitants of Balkh (Bactra) had come to settle in
the city. It notes, too, that the entrances t o courtyards and the courtyards them-
selves were washed down regularly.33

Royal palaces and community walls

The residence of the ruler was located in the centre of the city, an area that also
contained the more fashionable and better-built multi-storey buildings, which
were not permitted to rise higher than the ruler's palace. Here, too, were many
public buildings, including several picture galleries (citraiili), open to the
public and visited regularly. Such buildings were well constructed, special care
being taken to ensure that the lighting was good. A picture gallery usually occu-
pied a number of rooms linked together by passage-ways and staircases. The
walls of the main gallery were covered with paintings of the heavenly world,
episodes from the epics o r astrological signs. Some galleries belonged to rich
city-dwellers and some even to prosperous courtesans. The royal palaces con-
tained magnificent picture galleries, far superior to those owned by private indi-
viduals. The Ratndvali, a seventh-century play by Harsa, mentioned the picture
gallery at the entrance t o the palace. The palace had a special music room and
many rooms were decorated with sculptures, carvings and paintings. Its park
had ponds with small islands, on some of which there were gazebos
(Padatiditaka $33).j4

32. M. Ghosh, 1975, pp. 4-5.

33. Padatdditaka (Russian translation by I. D. Serebryakov), §$22, 24, 30, 35, 104; cf.
M. Ghosh, 1975, p p 114-15, 119, 123, 153; see also p. 123 for a specific reference to an
inhabitant of Balkh, Hariicandra, of the Kankayana tribe.
34. M. Ghosll, 1975, p. 117; see also Serebryakov's translation of the P i d a t i d i t a k a .
 Cities and urban li/e in the Kushan kingdom

Bazaars and dwelling-houses

The liveliest part of the city was the bazaar, bustling with shopkeepers and pea-
selling their wares. 'Everywhere here men and women are clustering
the shops, and people are buying and selling' (Padatiditaka $26). 'From the
smithy comes the sound of the hammering; from the brazier's workshop comes
[he shrill whine of the lathe, and a hiss like a horse's breath as a sword is
plunged into its scabbard' ( P i d a ~ d i t a k a$29).
iconographic sources, supplemented by literature, provide a wealth of
information o n houses occupied by city-dwellers. P ~ r i notes ' ~ this in relation to
evidencefrom the sculptures of Gandhara and Mathura. In the construction of a
house, a raised terrace @rasida) preceded the setting up of the walls (kudyin)
and columns (stambha). T h e roof rafters (gopinasi) were of wood (diru). The
rooms had several windows ( g a v i i k ~ i and
) a balcony (harmya) was a usual fea-
ture in large houses. In the Mathura sculptures a projecting balcony with cou-
ples sitting o n it is usually portrayed, as are the d v i r a and torana - gate and
gateways. In rooms, partitions (bhitti) were set up for privacy and copings
(vedikas) ensured protection from rainwater. The highest apartment was given a
special name (kutagara); houses were ~ a i n t e d(varnita); the term 'sopanam' sug-
gests that the ground floor was connected with the top floor by stairs; and there
was a separate ladies' apartment (antahpuva) inside the house. There is further
interesting illustrative material in the frescoes of some of the caves at Ajanta,
especially Cave XVII.36 According to the literature, city houses (not only the
palace) often had a garden in the inner courtyard (Kimasutra IV.3); one such
garden was t o be found in the courtyard of a merchant's house." The Pallcatan-
tra (111.5.95) mentions a merchant's house, located in the main street of the city,
standing literally o n the 'royal road' (vijamirga). In the cities and towns, mer-
chants usually lived in a special quarter (Fig. 9). This is clear both from literary
and from archaeological sources:

These quarters were very similar t o those of today. Lines of small shops with
verandas that were raised slightly above street level. Opening right on the street,
they were crammed close together, separated by n o more than the thickness of a
post. T h e open fronts were closed at night with removable shutters. The merchant
lived with his family o n the floor above, in tiny rooms, o r else in living quarters
behind the shop o n the other side of an inner courtyard. Throughout the day, he
sat cross-legged o n the wooden

35. Puri, 1965, p. 98.

36. Yazdani, 1946, Plate XXIII.
37. Schlingloff, 1970, pp. 25-6.
38. Auboyer, 1965, p. 87.
 B. A. Litvinsky

FIG.9. Begrarn. Plan of the bazaar. (After Hackin et al., n.d.)

In Udegram, every block was divided into two areas, one containing dwelling-
houses, the other shops. In some cases, the shops were built in a row along the
street. They were rectangular in ground plan with a small room at the back.I9In
every section of the city there was a network of alleyways, which crossed one
another at right angles and divided the city into blocks (eighty-one blocks were
sometimes called a pada, a number which appears to have had ritual signif-
icance). According to Indian architectural treatises, each such block or pada was
associated with some deity, who was the patron of the block. Again, according
to the texts, each block was surrounded by a wall and enjoyed a certain degree
of autonomy; it had its own water reservoirs, holy trees and temples dedicated

39. Faccenna, 1981, p. 31.

 Cities and urban l f e in the Kushan kingdom

local deities." Outside the city walls were the suburbs, which often
over a very large u e a S 4 '
cities were centres of science and culture, especially fine arts and music.
Many of their inhabitants were literate, and it was precisely to [hem that the
monumentalinscriptions were addressed. Specimens of the written language on
metal,stone, fragments of earthenware and birch bark have survived and frc-
guent discoveries of inkwells (at Taxila and elsewhere) provide clear evidence of
the dissemination of literacy.

City administration
Indian cities were administered by a governor. Subordinate to him were the
three chief magistrates. T h e district inspector (gopah) was in charge of ten,
twenty or forty families. H e was expected to know the caste, names and occu-
pations of all the men and women living in his district, and even how much they
earned and spent. There was also a borough inspector (sthanikah) in charge of
each of the four city sections. Each of these inspectors managed the affairs of
one quarter of the fortified city (Kautilya Arthaiastra II.36.14).42The cities of
Sasanian Iran also had district inspectors4' and there is some evidence that dis-
tricts were enclosed by walls. In India the municipal authorities controlled the
activities of artisans and merchants. There are references to city councils and
some cities had a city seal. According to Megasthenes (Strabo XV.I.51), urban
life was administered by six committees, each of which consisted of five mem-
bers and had its o w n specific function^.^^
Information about the population of Central Asian cities during the
Kushan period is very scanty, but if certain adjustments are made, information
about the composition of the population of the Indian cities can probably be
extrapolated t o Central Asian cities as well. The documents from Nisa provide
no information o n the rank-and-file population of 'fortified settlements' (diz in
Parthian). In these documents the commandant of a diz is referred t o as a diz-
pat. It is clear from the material in the highly specialized Nisa archives" that
cities, especially larger ones, were inhabited by members of the aristocracy, the
clergy and officials of the complex administrative apparatus. The high three-
towered castle at Toprak-kala, the citadel of Bactra and the s ~ l e n d o u rof the

40. Auboyer, 1965, pp, 120-1.

41. Ibid., p. 125; A. Ghosh, 1973, pp. 53-6.
42. Kangle, 1972, p. 185.
43. Perikhanyan, 1973, pp. 393, 496.
44. Bongard-Levin, 1973, pp. 197-202.
45. D'yakonov anci Livshits, 1968, 1977.
 H. A. Litvinsky

in other cities constitute clear and unambiguous evidence of the import-

ance of the aristocracy in the life of the period.J0
The Kara-tepe inscriptions show that where Buddhism was widespread,
an important role was played by Buddhist monks and officials of the ~ ~ d d h i ~
religious community (sangha). An equal if not more important role was of
course played by the numerically larger Zoroastrian priesthood. The population
in many cities included a number of foreigners. Harmatta4' has calculated that
some 30 per cent of the names found in Kharosthi inscriptions in India were
Iranian, and Bactrian names predominate. Urasaka, a Bactrian from Noacha
who was an official of the Kushan administration, notes in an inscription from
Taxila that he built a Buddhist shrine there.JYThe Milindapanha (V.331)) tile
Mahdbhdrata (11.47.15-31) and other sources provide information about Bat-
trians and persons from other parts of Central Asia who reached or lived in
India. The situation was similar in the cities of Central Asia. Buddhist mis-
sionaries and pilgrims, merchants, representative~of the Kushan administration,
soldiers and possibly craftsmen from India settled in the Central Asian towns
and travelled beyond the borders of the Kushan state. In the finds at Termez we
see evidence of scribes w h o had an excellent knowledge of north-western Prak-
rit and a thorough command of written Kharoghi. It is not possible to say
whether they were Indians or Bactrians who had been well schooled in these
languages, but in view of the role played by north-western Prakrit in the admi-
nistration of the Kushan state and the life of the Buddhist communities, it seems
probable that they included both local inhabitants and Indians. In both Central
Asia and India, Sanskrit written in Brihmi script is found.49
There are noticeable similarities in architectural styles. Some of these are
due to the fact that in both India and Central Asia there was a Hellenistic ele-
ment in the substratum of Kushan culture. A ~articularlystriking example is
provided by the form of stone columns, even though during the Kushan period
the form of columns in India and Central Asia developed in substantially differ-
ent ways. Other similarities were due to the spread of Buddhism in Central Asia
and the adaptation to Buddhist religious architecture which, under the influence
of local architectural and building traditions, assumed new forms and incorpo-
rated design solutions that were unknown in India. The synthesis of Iranian,
Central Asian and Indian architectural and religious ideas gave fresh impetus to
the concept of a shrine with corridors, and we know that Indian experts went to
Central Asia to design and build Buddhist shrines.'O
The Sitralamkara (IV.21) tells the story of a ~ i o u artist
s from ~u~kalavati

46. Livshits, 1984, pp. 265-79.

47. Harmatta, 1764, pp. 387-8; cf. Livshits, 1969, p. 64.
48. Konow, 1729, pp. 74-5; Litvinsky, 1968, pp. 13-14.
49. Vorob'eva-Desyatovskqa, 1974, pp. 1 18-20.
50. Litvinsky and Zeimal, 1971, pp. 113, 145.
 Cities and urban life in the Kushan kingdom

who journeyed t o the land of Aimaka (land of stone) where he decorated a

Buddhist monastery. Tradition has it that the SitrZlamkira was the work of the
famous ~ i v a g h o p a a, contemporary of Kanishka or of Kumiralita, the fc,under
,f [he school of Sautrintika, which also dates to the second century h.1,. and
[his hould be the content of the story. The toponym 'land of stone7 to
somewherein the north-west, probably in Central Asia. Some elements of Ccn-
tral Asian architectural and building styles made their way into India,jl and dcc-
orativeart in architecture reflected the synthesis that was occurring between the
Indian, Bactrian and Hellenistic-Roman styles.52The construction of Buddhist
buildings greatly affected the appearance of Indian and Central Asian
cities, Buddhist stupas giving some of them a very characteristic vertical skyline.
Common features can be observed in the nature and design of municipal service
systems. Taxila, Dalverzin-tepe and the Chim-kurgan site all have the same tvpe
of underground sewerage system.
Cities were still political and administrative centres as hitherto, but their
role as the focal point of handicrafts and economic life in general increased con-
siderably. Merv, f o r example, possessed copper and bronze works, bone-carving
workshops, armouries, flour mills, textile, ceramics and other industries, as also
did Termez, Samarkand, Toprak-kala, Dalverzin-tepe and other cities of Cen-
tral Asia. In ancient India, according t o the written sources, various groups of
the population, including craftsmen, had their homes in strictly delimited areas
of the city. Their workshops were located in their houses. The streets of Taxila
were lined with rows of buildings whose lower floors contained ateliers or
shops facing the street. T h e same was true of Bhita and in Central Asian towns
such as Toprak-kala, Merv and Saksan-Okhur.

Craftsmen and guilds

Indian craftsmanship during this period was highly specialized. Among the
various categories of metal-workers, the Milindapanha mentions blacksmiths,
goldsmiths, silversmiths, lead-workers, tinsmiths, coppersmiths, iron-work-
err, metallurgical craftsmen, and even gold as sayer^.^' The Mahivastrr
(111.113.442-3) mentions tin-smelters, skilled lead-workers, copper-smelters,
etc. Crafts involved in the production of weapons were of special importance.
The sources d o not speak of 'armourers7 in general, but refer separately to mak-
ers of bows and makers of bow-strings." The sources of this period mention

51. The Pa'datditakn ($52) tells of the embellishment of the courtesans' district with a
'mobile sanctuarv from north Bactria', cf. M. Ghosll, 1975, p. 131.
52. Sharma, 1968, pp. 34-5.
53. Horner, 1964, VoI. 11, p. 171-2; see also Puri, 1965, pp. 110-1 I; A d h ~ a ,1966.
54. Horner, 1964, Vol. 11, pp. 171-2.
 B. A. Litvinsky

(and in some cases even list) a large number of trades. The

(111.1 13.442-3) refers to thirty-six types of craftsman. The Milindapa~halist,
seventy-four kinds of occupation, most of them in the productive category, ~ h ,
Jitakas mention the names of eighteen guilds (ireni) of craftsmen and mer-
c h a n t ~ The
. ~ ~ number eighteen is the traditional figure, but a comparison of
various sources indicates that there were as many as thirty guilds.
There is some reason to believe that all members of a guild lived in the
same area; for example, there are references t o an ivory-carvers' street Uitakas
1.320; II.197), a carpet-makers' village, a potters' village, a weavers' village and a
stone-polishers' illa age.^^ The Jitakas refer often t o the ~ a d d h a k i ~ a m(carpen-
a
ters' village). O n e of them had a population of 500 carpenters and another
1,000, in which there was one chief f o r 500 carpenters. They collected wood
from the forest t o make the wooden components for different types of build-
ings. When their work was completed, they went to the forest again to collect
more raw material5'
Professions were hereditary; thus, in the Pali texts, the word 'son of a
smith' is synonymous with the word 'smith'. This is also borne out by epi-
graphic materials. References t o the hereditary nature of the crafts are found in
the writings of Kilidiisa. The heads of the guilds were noted by a number of
terms: pramukha (chief), mahattama (head man), jyeghaka (senior). According
to the inscriptions, an elder was known as the irestbin (best one). In theory,
only a person w h o had achieved the highest level of skill in his trade could
become an elder. The p i l d heads had their own ~ e r s o n a lseals bearing their
name and the title of ire!thin; they were assisted by agents and a secretary
(kiyastha). The guild heads regulated working conditions and rates of pay. In
consultation with the heads of other g i l d s , they raised o r lowered their prices,
depending on circumstances. In many cases the elder managed the funds of the
local branch of the p i l d , the p i l d assuming aggregate liability for all its mem-
bers. The head was in charge of security and had a special armed detachment to
protect p i l d property and funds and to escort caravans. The guilds rob ably
had special premises for their administrators, and special banners and ceremo-
nial badges that members wore on festive occasions.
Some p i l d s were very rich and ~ o s s e s s e dreal-estate, including some
buildings and large temple-complexes. In the first century A.D., some skilled
ivory-carvers from VidiSii (near Bhopal) donated money for building the torand
of a stupa at Sanchi, one of the great masterpieces of ancient Indian sculpture. In
the fifth century A.D., silk-weavers from Daiapura had sufficient resources to

55. This number is cited in the Mahdvastu. The term ircvi, already present in the Vedic
literature, had the general meaning of 'group'. By the time of the Kautilya it meant
specifically 'corporation' o r 'guild' (Kane, 194 1, p. 66).
56. Geiger, 1960, p. 104.
57. Misra, 1975.
 Citics and urban life in the Kushan kingdom
-
build the Sun Temple there, and thirty-five years later paid for necessary
repairs, An inscription from Nasik mentions a potters' guild, an oil merchants'
and a water-carriers' guild, all of which had made large financial
donations. The head of the guild enjoyed high social status and was somctimcs a
dignitary of the royal court. T h e state supported the guilds and protected their
rights and property. In written sources rulers are warned not t o interfere with
customs of the guilds, and t o confirm their status. The ruler should only
interfere if their usages and procedures were violated.
The p i l d s in turn performed specific public duties. At the time of official
city ceremonies, craftsmen and the heads of their guilds stood alongside the
and the Brahmans (Mahavastu 111.442). In one of the fables of the
Paicatantra, it is said that in Vardhamana, 'royal and municipal affairs' were
directed by Dantila, 'chief of the merchants', who 'meted out punishment and
distributed awards'. F r o m the text of the Arthaiistra (X1.1.4) some scholars are
of the opinion that the guilds provided soldiers. It is clear that armed detach-
ments, who protected guild property in peace-time, were placed at the disposal
of the state during war. It is also known from the epics that the guilds were
regarded as one of the pillars of state author it^.^^ The few literary sources that
are available for Central Asia contain no information on craftsmen's guilds,
though it is known that they existed in Iran under the Sasanians, and excava-
tions in Central Asia show that the various groups of craftsmen, potters, millers
and smiths were each established in clearly demarcated quarters of the city. It is
possible that the organization of the guilds was not so formalized in Central
Asia as it was in India.
The 'Palamedes inscription' at Surkh Kotal, written in Bactrian, included
at the end the Greek name 'Palamedes' in the genitive. Harmattaw concludes
that the signature was deliberately added by the architect, who was anxious t o
receive credit for his work. Harmatta also notes three Kharoghi inscriptions of
the Kushan period from India t o which, in his view, the persons in charge of
construction had deliberately added their names. From all this he draws inter-
esting conclusions about the growth of social awareness among the artisan and
merchant classes in the Kushan state. The inscribing of architects' names on
buildings was a reflection of the high social standing they enjoyed at that time.
The output of handicraft wares was abundant, varied and of the highest

58. See Rhys Davids, 1901, pp. 862-7; Fick, 1920, pp. 275-80; Kane, 1941, pp. 66-9; PUI-i,
1965, pp. 106-7; Adhya, 1966, pp. 82-8; Upadhyaya, 1947, pp. 268-9; Chakraborti, 1966,
PP. 315-28; Auhoyer, 1965, pp. 102-5. Between the fifth and seventh centuries, legal
documents indicate that the guilds had writtcn statutes and were obliged to have their
own p~-crnise~ wIlere theil. ~nembcrscould meet. The sources of that time ~ r o v i d e
detailed descriptions of their functions, statutes and administl-ation (Cliakl-abnrti. 1966,
pp. 328-37).
59. Har~natta,1964, pp. 338-9.
 B. A. Litvinsky

qualitv. This was made possible because of the high standard and colnplex tech-
nology of the equipment and tools available. The metal-working industry Pro-
vided the city and rural areas with tools, household wares, ornaments and wea-
pons. O n e branch, the jewellery trade, produced gold, silver, bronze and brass
ornaments, with some decorative inlays. The jewellery and toreutics of the
Kushan period were noted for their high artistic standard and many were
genuine works of art. The textile, pottery, wood-working and other trades were
very highly developed; so, too, were the building trades and the related architec-
tural and decorative arts - carving in stone and alabaster, wood-carving, paint-
ing, etc. The extraction of minerals was also widely practised; handicraft pro-
duction in the various provinces of the Kushan Empire was very diversified and
individual provinces were noted for ~ r o d u c i n gspecific types of articles. Local
schools of craftsmen developed distinctive local styles, though some ware was
common to several provinces.

Trade and commerce

The high rate of marketable output of urban production, the need for exchange
of goods between cities and their agricultural environment and territorial dif-
ferences were the factors that led t o the extensive growth of trade within cities
and between the provinces of the Kushan state. According to Indian sources,
there were two types of merchants: the vanik (those who had regular shops)
and the sdrthavdha (caravan traders). The caravan traders also had their elders.
Because of poor roads and the dangers that might be encountered along them,
including attacks by bandits, the caravan trade found that large, well-equipped
and well-protected caravans were safest; the Milindapafiha mentions a merchant
who travelled to Pitaliputra with a train of 500 wagons. During the Kushan
period, according t o T h e Periplus of the Erythrean Sea and Indian sources, mer-
chant vessels also sailed the high seas and, taking advantage of the monsoon
winds, crossed the Indian Ocean.03
Trade between the different provinces of the Kushan state is well docu-
mented by archaeological finds in Central Asia. Articles imported from the
Indian provinces included ivoryware, precious stones, jewellery and other orna-
mental objects. But trade was not confined t o the provinces as its maritime and
overland routes linked the Kushan Empire t o the Mediterranean, the Far East,
the wooded steppes and South-East Asia. The movement of goods and cultural
treasures was a two-way process, creating opportunities for cultural cross-ferti-
lization in the areas of thought, art, architecture and material production.

60. Puri, 1965, pp. 107-8; Frisk, 1974; Warmington, 1974; Thorley, 1969; Schmitthenner,
1979.
 R E L I G I O N SI N THE K U S H A NE M P I R E : "
J. Harmatta, with the contributions of
B. N. Puri, L. Lelekov,
S. Humayun and D. C. Sircar

Religious life in Bactria before the Kushan conquest

On the eve of the nomadic invasions of the second century B.C. religious life in
Bactria and the adjacent territories was characterized by a number of religious
beliefs and cults of different origin. Zoroastrianism played an important role
among the Iranian-speaking population, and the teaching of Zoroaster had con-
quered the eastern Iranian territories before Alexander's conquest' even though
many remains of pre-Zoroastrian religious ideas and cults probably survived.
The traditions of pre-Zoroastrian Iranian religion, however, prevailed in the ter-
ritories north of the O x u s and t o a greater extent among the Iranian nomadic
tribes of the steppes. T h e Zoroastrian calendar had already been adopted in Per-
sia, Parthia, Bactria and C h o r a ~ m i a while
,~ the Sogdian system of month names
differs because the majority of pre-Zoroastrian month names were maintained
in Sogdiana. In any case there seem t o have been some major differences be-
tween original Zoroastrianism and later Zoroastrian orthodoxy. The latter, with
its iconoclastic tendency which appeared at the end of the Arsacid age, neb7er
became firmly established o n the territory of eastern Iran, though later, under
the influence of Kartir (Kirder), the mobed, great efforts were made t o streng-
then Zoroastrian ecclesiastical organization and orthodoxy - resulting in the
persecution of Christianity, Buddhism and other religions.
At the time of the nomadic invasions, however, Zoroastrian orthodoxy
did not yet exist in Bactria. O n the contrary, a considerable number of Greek
settlers living in the Graeco-Bactrian kingdom formed religious communities
worshipping their o w n gods and ~ r a c t i s i n gtheir own cults. ~rchaeological

:" See Map 4.

1. Gnoli, 1980, pp. 215 et seq., pp. 227 et seq.
2. Harmatta, 1969, pp. 369 et seq.
1, ij,,,.,,l'lfl~,
.~l,thIIJC
contributions of' H. N. Puri, L. Lelekov, S.Humayu,~and D. C, slrr,,,

finds and the coinage of the Graeco-Bactrian kings attest to worship of

~ n a j o rGreek divinities: Zeus, Poseidon, Apollo, Helios, Heracles, Dionysus,rh,
Dioscuri, Athena, Artemis, Hecate and Nike. Greek and Iranian religious ideas
and cults influenced each other, while Greek religious architecture and art
influenced the building of sanctuaries and created the iconography for Iranian
deities.
In eastern Iran, as everywhere, the Greeks attempted to understand local
religious ideas and t o identify local divinities with Greek ones (interpyetalio
Graeca). Zeus was identified with Ahura Mazda, Heracles with Verethragna,
Apollo and Helios with Mithra, Artemis with Nana. The Greek interpretation
of Iranian divinities t o a great extent determined their iconography. Apparently
Greek forms of religious worship even influenced Iranian cults. This can be seen
in the Greek votive inscription from the sanctuary of the god Oxus at Takht-i
Sangin. The inscription was incised o n a little stone altar with the bronze figure
of Marsyas playing a two-branched flute. The donor bears the Iranian name
*Atyosaukaj and dedicated his votive present to the god VaxSu (Oxus). Here,
both the believer and the god are Iranian, but the form of worship (the votive
altar with the bronze figure of Marsyas and the dedicatory inscription) is Greek,
excellent evidence for Graeco-Iranian syncretism in the religious life of Graeco-
Bactria.
The spread of Indian religions also began under Graeco-Bactrian rulers at
this period. According t o the testimony of the Greek and Aramaic versions of
the rock edicts of Aioka (see Chapter 16), the beginning of Buddhist missions
to Bactria dates back t o the third century B.C. The spread of Buddhism and
Brahmanism was due to Indian merchants and craftsmen emigrating to the great
centres of Graeco-Bactria (see Chapter 17 and the inscription of the potter
Punyamitra from Begram). The Greeks were also open t o Indian religious ideas
as is attested by the pillar inscription of Heliodoros, the ambassador of King
Antialcidas who became a worshipper of Vishnu and erected the Garuda pillar
with an inscription in Briihmi in honour of Vishnu at Besnagar.
Local cults like that of the Oxus played an important role among the Ira-
nian, Indian, Dardic and Kafiri population. The Graeco-Bactrian kings were all
aware of the importance of these local cults and sometimes represented an
important local god or goddess on their coins. The 'City-Goddess of KipiSa'
appears on the coins of a late Eucratides in the form of a female deity wearing a
turreted crown and seated on a throne. The representation of this city-deity can
be compared t o that of Zeus seated on the throne (i.e. it is Greek in character),
but the elephant to the left and the caitya t o the right clearly refer to an ~ndian
religious sphere. Thus, if the altar of Atrosauka furnishes an excellent case for
Graeco-Indian religious syncretism, then the figure of the city-deity of KipiSa

3. Litvinsky and Sedov, 1984, p. 61.

 I<eligions in the Kushan Empirc

first-class testimony f o r the amalgamation of Gracco-Indian religious

P,,ides
ideas. rania an, Greek and Indian religious cults existed side by side, influencing
each ~ t h e with
r their rich religious ideas and forms of worship and in
syncretism which continued to influence religious life after estab-
lishment of Kushan rule in Bactria.

The ancient religion of the Sakas and Kushans

When the Saka and Yiieh-chih tribes arrived in Bactria, they must have had
[heir own religious ideas and cults. For lack of relevant direct evidence, how-
ever, it is an arduous task t o form an idea of their ancient religion. There can be
hardly any doubt that the ancient religion of the Sakas and Kushans was not
~oroastrianism.In spite of some uncertainty in identifying the lands in the list
of lands in the first chapter of the Videvdit,' it is clear that neither the former
dwelling place of the Sakas nor the ancient home of the Yiieh-chih belonged to
the territories where Zoroastrianism spread at an early date. Thus, Saka
urmaysde (sun), going back t o O l d Iranian *Ahura-mazd2h-, cannot be used as
evidence for the Zoroastrianism of the Saka tribes. O n the contrary, it shows
that the name "Ahura-mazdih- is pre-Zoroastrian, and this is confirmed bv thc
occurrence in Assyrian sources of the form Asara Mazas, which reflects the
Proto-Iranian form "Asura mazdzs- of the name Ahura-mazdih-.
The Saka and the Assyrian evidence clearly supports the assumption that
"sura mazdis- was a pre-Zoroastrian divinity of the Iranian tribes with a
strong solar character which led to the semantic development Ahura-mazdah- >
urmaysde 'sun' in Saka.5 Consequently, there can hardly bc any doubt that the
Kushans already worshipped Ahura Mazda before their acquaintance with
Zoroastrianism in f o r m of a god of heaven with strong solar features. Beside
Ahura Mazda the pre-Zoroastrian worship of Mithra, Aryaman, Vayu, ASa,
Yama, Verethragna, Spenta Armaitij and the cult of Hauma are attested by lin-
guistic evidence. It is questionable, however, whether the same religious ideas,
gods and cults existed throughout all the extensive territory inhabited by the
Iranian tribes in pre-Zoroastrian times. It seems, for example, that thc worship
of Mithra, Aryaman, Vayu and Yama was not known among the Iranian
nomads of Central Asia and was consequently also unknown among the Sakas
and Kushans, while the cult of Ahura Mazda and Spenta ~ r m a i t can i ~ ~robabl!?
be assumed o n the basis of Saka urmaysde 'sun' and i i a n d ~ n z a t a'name of a
goddess'.
Differences in religion between the Iranian nomadic peoples are clearly
shown by Herodotus w h o attests the worship of Zeus, Ge, A ~ o l l o ~, ~ h r o d i t e

4. Gnoli, 1980, pp. 23 et seq.

5. Steblin-KamcnskiY, 1981, p. 238.
Urania, Poseidon, Heracles and Ares (according to the interpretatio Graeca)
among the Scythians (IV. 59), while he emphasizes (1.216) that the Massageta,
of Central Asia only worship the sun. Consequently the pre-Zoroastrian reli-
gion, which wc deduce from the evidence of common Indo-Iranian (Avestan
and Rigvedic) religious terminology, probably flourished only in the eastern
territories, adjacent to the area inhabited by the ancient Indian tribes, while the
religion of the northern Iranian nomads living in eastern Europe and Central
Asia may have had other peculiar features.
Thus, instead of the cult of Hauma, the cult of Hemp is attested among
the Scythian tribes. O n the basis of Pashto oma, 'name of a plant', Munji
yzimena, 'name of a ~ l a n t '(< "haumana-), Wakhi yimik, 'Ephedra' (<
maka-, cf. yir 'sun' < "bur) again we can perhaps count with the existence
among the eastern Iranian tribes, Sakas and Kushans of the Hauma cult in pre-
Zoroastrian times. Vayu is attested by Ossetian vayuk, 'giant, devil' (< *vayu-
ka-) and the Alanian personal name Vayuk (inscription of Ladinybene, fourth
century A.D. in runic script) for the north-western Iranian nomads. On the
other hand the north-eastern tribes preserved a rich pre-Zoroastrian religious
terminology, surviving in Khotan Saka, Southern Saka and in the language of
the Kushans (see Chapter 16). The most important are the following: Khotan
Saka vvuva-, 'god' < pays-, gyays-, jays-, 'to sacrifice', gyasta-, 'god'; Southern
Saka jasta-, 'god' < "yazata-; Khotan Saka dyu-, 'demon' < "daiva-, which also
was known according t o the testimony of Ossetic av-deu, 'evil spirit', among
the north-western Iranian nomads.
Essentially the Sakas and Kushans who invaded Graeco-Bactria may have
had similar religious ideas and cults t o the population of Sogdiana and Bactria in
pre-Zoroastrian times. They probably worshipped Ahuramazdih as 'God of
Heaven' with solar features and Svanta Armati as 'Goddess of Earth'. They
were acquainted with several categories of divine beings such as daivas, yazata~,
and bagas; and used the verb yaz- as a term for sacrifice and worship, and the
words bags-spasika- and bagana-pati- t o denote different categories of priests.

The epoch of the Kushan yabghus

When the Sakas and Kushans conquered the Graeco-Bactrian territories 110rth
of the Oxus, they found mailifestations of Greek religious life, religious archi-
tecture, sculpture, ideas and worship. If the first centre of the Kushan yabgbu
can really be identified with K l ~ a l c h a ~ ainn the Surkhan Darya valley, the finds
discovered there will enable us to follow step by step the formation and devel-
opment of the religious life and religious policy of the Kushan rulers. On the
coins of the first-known Kushan yabghu, Sanab, thc spelling HIAOY, pye-
viously read erroneously as Heraiou or Miaioa, is not the name, but the title of
the ruler, and should be read as hyau, representing the most archaic form of the
 Rtdigions in the Kushan Empirc~

title yau, y a v ~ yabgu-

, Nike, the Greek goddess of victory, already appears.
Apparently the Kushan aristocracy tried to adopt the royal ideology of
~ ~ ~ ~ c o - ~ a kings n its religious implications. It is therefore n o accident
c t r i aand
[hat in the sculptural decoration of the Kushan manor-house at Khalchayan the
enthroned ruler and his wife appear again with Nike." It is a renlarkablc fact that
~ i k ise represented in the company of a bearded god with a radiate halo
Hellenistic Mithra. Beside Nike and the bearded god, a sculpture of Athens was
discovered at Khalchayan. If the bearded god can be identified Zeus,
who also has a radiate halo o n the local copies of coins of Heliocles, then wc ob-
tain an interesting insight into Graeco-Bactrian-Kushan religious
Zeus with a radiate halo was obviously interpreted as the pre-Zoroastrian
~ushan-Saka Ahura Mazda of solar character, that is, the Greek sculptor created
an iconography that expressed both Graeco-Bactrian and Kushan religious ideas.
~t is noteworthy that beside the Greek winged Nike and the Graeco-Kushan
ZeusIAhura Mazda, the Hellenistic Mithra represents a third tvpe of Graeco-
Iranian syncretism. T h e Kushans may already have been acquainted with the cult
of Mithra in Sogdiana before they invaded Bactria proper, because the name of an
Iranian noble (Sisi-miera-, 'devoted t o Mithra') provides evidence for the cult of
Mithra on territory north of the Oxus in the time of Alexander the Great. The
figure of Mithra wearing a Phrygian cap from Khalchayan recalls the representa-
tion of the Western Mithra - the Graeco-Roman Mithras - and was the result of
Graeco-Iranian syncretism which the sculptor adapted to the Kushan Mithra.
The presence of N i k e o n coins of Sanab, the first Kushan yabghu, and in his
manor-house at Khalchayan is clear evidence for the existence of a goddess of vic-
tory who was called either Nike or Vanindo in the royal ideology of the Kushan
rulers before the rise of the Great Kushans.
The coinage of the next Kushan ruler, Kujula Kadphises, enables us to see
how the religious horizon of the Kushans was enlarged. His first issue, which
has the debased portrait and name of Hermaeus o n the obverse, shows Hera-
cles on the reverse, still following the Greek tradition, even though Heracles
may be the intevpretatio Gvaeca of the Iranian god Verethragna. O n the
reverse, however, the legend is already written in Kharosthi script: Kujula
Kasasa Kulana yavugasa dhramathidasa 'of Kujula Kasa, the K ~ s h a njlabghu,
who is steadfast in the Law'. The epithet dhramathida- < dharmasrbita- 'stead-
fast in the Law' of Kujula Kadphises occurs in fuller form in the legend of a
later issue, ilarnely sacadhramathita- < ~ a t ~ a d h a r m a s t h i t a'steadfast
- in the true
Law'. Contrary t o earlier assumptions, which regarded Kujula Kadphises as
Buddhist o n the basis of this epithet, it is now clear from the wording of a
Mathura inscription,7 in wllic11 Huvisllka bears the same epithet mtjudharmns-

6. Pugachenkova, 1966, p. 187.

7. Liiders, 1961, pp. 138 et seq.
j, H'rr,,l'qtrd, wirh thr contributions 01' B. N. Puri, L. Lrlrkov, 5. Humayun and 1). C, sirtar

thila, that the kingdom was conferred upon him by Sarva and Scamdavira
(Candavira), that is, he was a devotee of Siva. It is striking to see that ~ ~ j
e s already adopted the worship of Siva and the use of Kharosthi
~ a d ~ h i s has
.,
script at such an early date. We must not, however, forget that the spread of
Indian religious ideas and cults t o the north-west as well as the use of Gandhari
Prakrit and Kharosfhi script had already begun under the Graeco-Bactrians,
The Indo-Greeks from the time of Apollodotus I, Antimachus and Menander
had regularly struck coins with Gandhiri Prakrit legends and the later Eucra-
tides had used the city-goddess of KipiSa as a reverse coin type.
The spread of Indian religions, scripts and languages to Bactria presup-
poses the migration there of Indian merchants and craftsmen. They were
attracted by the quickly developing new Graeco-Bactrian cities and the favour-
able prospects of long-distance trade opened u p by the Greek kingdom of Bat-
tria and later by the Kushans. If the importance of trade between India and
Pontus was already clear to Antiochus I, the decisive significance of trade be-
tween India and China through Central Asia must have been even clearer for
the Graeco-Bactrian and Kushan rulers. This explains their ambition to acquire
and control the Silk Route. According t o the report of Aristobulos (quoted by
Strabo XI.7.3), the Oxus river was navigable and many Indian goods were
transported o n it as far as the Hyrcanian Sea, and from there to Albania and the
Pontic region. The importance of Indian trade down the Oxus river and the
activity of Indian merchants and craftsmen along this important trading route
gave the Kushan yabghus strong reason to refer Indian religious worship and
t o use Indian scripts and languages. The share of Indian merchants was also
important in the silk trade between India and China, which began to flourish
from just this period. From the first century B.c., corporations of Indian mer-
chants were formed in Xumdin, the Chinese capital, clear evidence of the close
trading relations between these two great and rich nations. When the Kushans
conquered Transoxania they became masters of the initial section of the Silk
Route, and it was almost a historical necessity that Kujula ~ a d ~ h i s ethe
s , foun-
der of the Kushan Empire, began t o refer the cult of Siva. None the less, Greek
religious ideas and Greek religious iconography remained important for the
Kushan dynasty. When, after his victories, Kujula K a d ~ h i s e sassumed the title
mahiiriija vijatirija (Great King, King of Kings) in his coin legends, he used the
winged Nike as the reverse type of the issue.

Religious life under Vima Kadphises

At the time when Vima Kadphises became Kushan emperor, religious life can be
characterized by two interesting features. O n e is the adoption of the forms of
Greek religious art and the Greek iconographic interpretation of Kushan divin-
ities. It is very likely that the Kushan gods O h r o ~ n a z d o ,Vanindo, Mihro and
 Religions in the Kushan Empire

0ilagno lie behind the Greek iconographical garb of Zeus, Nikc, Mithra and
Heracles- further evidence f o r the strong influence of Greek religic,us ideas and
forms before the rise of the Great Kushans. The other striking feature is
orientation towards Indian religions and the worship of Siva in partic-
ular.1t is, therefore, not surprising that the cult of Siva became even more pro-
minentunder Vima Kadphises, w h o conquered considerable territories i n India.
His coinage gives clear evidence of this, as Siva, or Siva with Nandi, appears 3s
[he reverse type of all his issues.
In some coin legends Vima has the epithet maheivara, which, being a typ-
ical name of Siva, can hardly be taken in the sense of 'the great lord' and refer to
Vima himself; but as Kharosrhi script did not indicate long vowels, it can more
probably be interpreted as miiheivara 'worshipper of Siva'. Archaeological and
finds also attest the leaning of Vima towards the cult of Siva. At Dil-
berjin, the temple of the Dioscuri, built in Graeco-Bactrian times, was transfor-
med by Vima Kadphises into a sanctuary of Siva and decorated with a wall-
painting representing Siva and Parvati (See Chapter 15, Fig. 9). According to the
fragmentary Bactrian inscription D 1 (see Chapter 17), Vima Kadphises prob-
ably had the wall-painting of O e ~ o(Siva) prepared, and gave orders that the
priest of the stronghold and the master of the hunt should take care of the sanc-
tuary and cult. It is clear from the long Bactrian inscription D 2 (see Chapter
17) that Vima Kadphises probably had craftsmen brought from Ujjayini (mod-
ern Ujjain) t o construct a water conduit to the sanctuary of Siva. According to
the Bactrian inscription DN 1, Vima Kadphises again ordered the town A n d ~ z o
(Lrapho = Qunduz) t o retain the tax it collected and use it for the sanctuary and
the warlike divinity (see Chapter 17). The fragmentary Brihmi inscription on
the pedestal of a statue from T6kri Tili, near Mat, speaks more precisely about
the relation between the god Siva and the Kushan king: . . . satyadharmastbit-
a ~ y a n a n a y a t s a r v a i c a m d a v i r a t i s ~ a r ~ j ~.a .s ~. a'who is steadfast in the true
Law, on whom, o n account of his devotion, the kingdom was conferred by
Sarva and Scarpdavira'.~~ h u s the , Kushan king ascribed his rise to power to
Sarva (= Siva) and C a ~ d a v i r a(who may be the same god as Candiivara, the god
of the Mahikila temple at Ujjain, probably a special form of Siva). Because the
circle of gods around Siva have a warlike character, it is very likely that the
phrase iazado i karisaro 'the warlike divinity' also denoted Siva.
This devotion of Vima K a d ~ h i s e sto Siva could have both ~ e r s o n a land
political reasons. T h e great commander and conqueror may himself have felt an
attraction towards the warlike god Siva and the ~ a r - ~ o in d shis ambience. It is
noteworthy that Vima bears the epithet mabozinigo of the moon
[god]' in his inscription DN 1 (see Chapter 17). and it seems that originally the
divine patron of the Kushan dynasty was the ancient Iranian moon god. In view

8. Liiders, 1961, py. 140 et seq.

/ HLrjmnrf.r,iilirb thr conrril.urio~zsof R. N Pun. L. Lelrkov, S.Humayu)7 and n, ,;( slso,
-
of the close connectio~ibetween Siva and the moon, dynastic religious id,,,
may have also suggested t o Vima the choice of Siva as his divine patron, In pol-
itical terms, both in the preparations for his Indian campaign and during the
campaign itself Vima Kadphises may have received valuable support from
groups of the Indian population who worshipped Siva. A number of Indian
tlements already existed on the territory of eastern Iran at the time, and the par-
thian ostraca from Nisa show that there was an Indian settlement called Hindu-
g j n in the neighbourhood of the ancient Parthian capital. Consequently the
support of the Indian population of his kingdom may have been importantfor
Vima both before and during his Indian campaign. The emergence of
Candavira/CandiSvara, the god of the Mahiikila temple at Ujjain, among the
divine patrons of Vima probably bears witness t o the assistance he received
from the priesthood of Siva there. The mention of 'Ujjayini' in his inscription
D 2 reflects the importance of relations with Ujjain and its cult of Siva, main-
tained by Vima Kadphises even after his Indian campaign.
A peculiar feature of the iconography of Siva adopted for the reverse on
coins of Vima Kadphises permits us t o think of some local factors in the spread of
the cult. O n some of Vima's coins Siva is shown with tongues of flame rising
from his head, a phenomenon otherwise unknown in Siva iconography. The male
figure with five-rayed head, on the reverse of the early issues of the Mithra kings
of Paficila, is clearly different and cannot have been a model for the flaming-head
Siva on coins of Vima Kadphises. In Indian mythology, it is the god Yama who
was imagined with flaming hair. Moreover, it should be remembered that Yama
(Imua < Y a m a ~ i j a is) the principal god of the Kafir tribes. We must therefore
reckon with the possibility that the iconography of Siva was also influenced by
local religious ideas, belonging originally t o Yama, worshipped by the local
Kafirs, and that this syncretism also contributed to the spread of his cult.
The Kushan kings derived their royal power from divine patrons, and so
they were charismatic kings, human incarnations of divine might and power. As
a consequence of their charisma, they also became objects of divine worship in
dynastic sanctuaries. Vima Kadphises began the construction of two such
centres of the royal cult, one at Mat, near Mathura, the other in Surkh Kotal.
The construction of the sanctuaryg at Mat was executed by ~ u m a ~ ~ athe la,
bagandpati (curator of the temple), according to the record incised between the
feet of a colossal seated figure of Vima K a d ~ h i s e s whose
, name appears in the
form Vema Takpisa (earlier reading, Vema Taksuma). The same form of his
name also occurs in the Bactrian inscription DH 1 ( O o ~ m T o a k ~ i s o )~. e s i d the
e
temple, a garden, a tank, a well, an assembly hall and a gateway were con-
structed. In Surkh Kotal, however, only the preparatory work began during the
reign of Vima. According to the unfinished inscriptioil (SK 2, see Chapter 17)l

9. Called devakula in the building inscription; sce Liiders, 1961, p. 135.

320
 Religions in the Kushan Empire
-
he had a canal dug there t o assure the water supply for building operations,
which were probably continued and finished by his successor Kanishka.

The religious policy of Kanishka I

The accession of Kanishka marked essential changes in the religious life of the
Kushan kingdom. While in the interest of his Indian conquest, Vima Kadphises
had given reference t o the worship of Siva in his religious policy, his successor
~anishkaput Bactria and its Iranian religious cults at the centre of his
policy. H e continued and finished building the dynastic sanctuary at Surkh
Kotal. If the restoration of the fragmentary building inscription (monumental
wall inscription SK 1) is correct, the construction of the stronghold and the
great staircase as well as Temple A was finished in four years. The sanctuary
bore the name 'Kanishka Oanindo-sanctuary' but according to the Bactrian
inscriptions SK 4A, 4B, 4M this name was only given later, in Year 31, when the
sanctuary was renovated and enlarged (see Chapter 17). So Temple A may origi-
nally have been used f o r the cult of the dynastic divinitics on the reverse of his
first coin issues, namely, Helios, Selene, Hephaistos and Nanaia.
While the first issues minted by Kanishka still bore Greek legends, they
were subsequently replaced by Bactrian legends. Correspondingly, instead of
the Greek gods the Iranian Mioro, Mao, A0so and Nana appeared.
In this phenomenon we d o not have a change in the religious cult of the
Kushan royal court, merely the omission of the Greek interpretation of their
dynastic gods. T h e representation of Mioro, Mao, A0so and Nana is identical
with the earlier forms of Helios, Selene, Hephaistos and Artemis Nanaia. The
coincidence is particularly striking in the case of Selene, who appears as a male
divinity, with the iconography of the Greek moon goddess applied to the male
Iranian moon god. Obviously, the Kushan Helios, Selene, Hephaistos and
Nanaia d o not represent the Greek deities Helios, Selene, Hephaistos and
Nanaia, but are the Iranian gods Mihro, Maho, ABso and Nana, divine patrons
of the Kushan dynasty, w h o appeared according to the interpretatio Graeca
bearing Greek divine names and in Greek iconography.
The first of them, Mihro, was already represented at Khalchayan as patron
god of the first Kushan yabghu, Sanab. O n the basis of the epithet mabozinigo,
borne by Vima Kadphises in the Bactrian inscription DN 1 'protege of the moon
god', Maho also belonged t o the group of the Kushan dynastic deities. As con-
cerns A0go-Hephaistos, probably the Kushans also had their dynastic fire as did
the Arsacids and Sasanians, and this was placed on the platform of Temple A at
Surkh Kotal. Possibly the dynastic fire cult was taken over by the Kushans from
the Arsacids in the same manner as the title 'King of Kings'."

10. Hasmatta, 1965. p. 171.

j, Hrrrmnttn, with the contributions of B. N.Puri, L. Lelekov, S. Humayun and D. C s , ~ ~

The origin of Nana worship points in the same direction. According

the evidence of the Parthian ostraca from Nisa, a Nana sanctuary also in
the ancient Parthian capital and royal residence. Very probably the cult of N~~~
arrived from Parthia. The evidence for the Nana sanctuary at Nisa is scanty and
does not throw any direct light on the relation of the Nana cult to the Arsacid
dynasty, but the existence of a Nana sanctuary in the Parthian royal residence
makes it likely that Nana was also one of the divine patrons of the Arsacid
dynasty. West Iranian religious influence can be seen among the Saka tribes
who borrowed some Zoroastrian terms from the Parthians (e.g. den 'religionJ,
ar;a < O l d Iranian areya-, Avestan &+a- 'pious', adu- < Old Iranian artavan-,
Avestan ajavan-). The phonetic form of these terms clearly supports a borrow-
ing from Parthian and excludes a local 'Bactrian' origin.
The emergence of the names, Mioro, Mao, A0so and Nana, instead of the
corresponding Greek names, Helios, Selene, Hephaistos, Nanaia, on the reverse
of the coins struck by Kanishka was made possible by the creation of the Bac-
trian writing system based on the Greek alphabet during the reign of Vima Kad-
phises. The possibility of writing Bactrian enabled Kanishka to replace Greek
with Bactrian legends on the coins, and t o set u p inscriptions written in Bac-
trian. As a consequence of this development, the Iranian gods removed the
Greek-language disguise and appeared with their Iranian names. Even the
names of the genuine Greek gods became slightly Bactrianized, the Greek
word-ending being replaced by a Bactrian one.
While the development of a Bactrian script made it possible to replace
Greek with Iranian names, it alone cannot explain the reference given by Kan-
ishka and his successors t o the Iranian divinities. Because the worship of the
Iranian gods prevailed first in the territory of Bactria, the re dominance of the
Bactrian cults in the religious policy of Kanishka I also indicates the increased
interest of the Kushan king in the western part of his empire - the home terri-
tory of Bactria. Behind this new orientation, we can note the strengthening of
the Parthian kingdom during the second century A.D. when Parthia became a
permanent threat t o the Kushans.
None the less, the importance of India and the Indian religions, especially
the worship of Siva, remained unchanged. Kanishka has a reverse type repre-
senting Siva with the name Oego < O l d Indian Vrla > Prakrit Vela, identifying
the god by an inscription for the first time. If Mioro, Mao, AOso and Nana were
the ancient divine patrons of the Kushan dynasty, then Siva had belonged to the
same group of gods since the reign of Vima Kad~hises.consequently, reverse
types of the coin issues of Kanishka represent ~ r i m a r i the
l ~ dynastic pantheon
of the Kushan king, to the worship of which the sanctuary of Surkh ~ o t a was l
dedicated.
 Religions in the Kushan Empire
-
Religious life under the 'triple' kingship
The successor of Kanishka seems t o have been his son Visishka, who, according
, the inscription of Kamra, was the great-grandson of Kujula Kadphises and
ruled jointly together with his first-born son, Kanishka II. At the same timc, on
[he basis of a fragmentary inscription from Mathura," wc can statc
grandfatherof Huvishka was Vima Kadphises. If, therefore, Kanishka I was
son of Vima and grandson of Kujula Kadphises, Vasishka and Huvishka rnust
have been brothers. According t o the inscription of Kamra, both Visishka and
~anishkaI1 were ruling in Year 30 of the era.
The Bactrian inscription of Ayrtam again attests the rule of Huvishka in
Year 30, and o n the basis of the Rijatarangini it could be argued that Hushka,
~ushkaand Kanishka all ruled at the same time. It seems, therefore, that Kanishka
I was followed by his son Vasishka w h o took his first-born son, Kanishka 11, and
his brother Huvishka as co-rulers. Taking into consideration that no coin issue of
Visishka is known so far in the dynasty of the Great Kushans, apparently it was
Kanishka I1 w h o minted coins and not his father Visishka. The coin issues bear-
ing the name of Kanishka can possibly be divided between Kanishka I and Kan-
ishka 11. In fact, w e can observe some striking changes in the Kushan pantheon,'?
represented o n the coin reverses, which make it possible to attribute coins
belonging t o the third emission, from section A 2 on13to Kanishka 11, who, o n his
coins, wears a hat-like crown with a broad, richly decorated brim.
O n the reverses of these coin issues ascribed to Kanishka 11, there appear
a series of divinities, w h o did not play any part earlier in the Kushan coinage.
They are Pharro, Manaobago, Ardoxso, Boddo, Oi-lagnoIOslagno, Lroaspo,
Mozdooano. Beside these deities, the ancient divine patrons of the Kushan
dynasty such as Mioro, Mao, Nana and Oeso are also represented. There must
have been some reason f o r the emergence of new gods in the ant he on of the
Kushan coins. Kanishka 11, the son of Vasishka, bearing the titles mahirija
yajatir2ja devaputra kaisara (Great King, King of Kings, Son of God, Caesar) in
the Ara inscription, is an enigmatic figure. His ~ersonality,however, appears in
a new light if w e recognize him as Chen-t'an Ki-ni-ch'a (*Candana Kanishka)
of the Buddhist work Sridharmapi~akanidinasitra,according t o which King
Candana Kanishka won a great victory over the King of P$aliputra and the
Parthian king. Candana Kanishka is also mentioned by the name of Sandanes in
the Periphr (Chapter 52) as a mighty ruler who conquered the most important
harbours o n the western shore of G d i a south of Barygaza (Broach). It follows

11. Cf. Liidcrs, 1961, pp. 138 et seq.

12. 'Zone of actuality'; see Giibl, 1983, pp. 85, 94.
13. According t o the system elaborated b!~ Gobi; ibid., pp. 85-7.

323
1, I-jdr,nilrrd, .xvith the, cmntributions of' B. N. Puri, L. Lelekov, S. Humaj~unand D, C: sirCdr

that the Buddhist legends woven around the figure of Kanishka belong not to
Kanishka I but to his grandson, Kanishka 11.
Kanishka I1 clearly recognized the importance of Buddhism in his king-
dom. There were some important Buddhist centres in Bactria, at Ternlezand
Ayrtam, where missionary work of both the Mahisarighika and the Sarvastivada
schools was active. Kanishka I1 was, without doubt, a great protector of ~ ~ d d -
hism and founded monasteries and built stupas according to the Buddhist tra-
dition. From the viewpoint of the history of Buddhism, however, his most im-
portant action was to convene the Buddhist synod in Kashmir, a decisive
turning-point in the life of the Buddhist schools. According to tradition, this
synod of the Sarvistivada school compiled the J6anaprasthdnam and entrusted
Aivaghosa, the famous poet, with providing for the correct language form of
the commentary written by Kityayana. Essentially, his charge was to rewrite
the Buddhist works in Sanskrit. Earlier both the Mahisinghika and the Sarvasti-
vida schools equally used Kharosthi and Brihmi t o write Gandhiri Prakrit.
After the synod of Kashmir, however, as a consequence of the literary activity
of Aivaghosa, the Sarvistivida preferred Sanskrit and Buddhist Hybrid Sanskrit
written in Brahmi script t o G i n d h i r i Prakrit written in Kharoghi script. So the
Buddhist Hybrid Sanskrit became the literary language of Buddhism, and in
this development the role of Kanishka I1 was decisive. It was not by chance that
around his figure a cycle of Buddhist legends came into being.
None the less, he did not neglect other religions and cults. O n the reverse
types of coins beside Boddo (Buddha) we find other Iranian divinities both
Zoroastrian and local. Among them Manao Bago roba ably represents the Bac-
trian name for Avestan Vohu Manah (Good Mind, Wisdom) which was in the
possession of xiaera- 'might, kingdom'; he bestows xiaera- for the righteous-
ness of man; and increased it t o triumph over the Druj and enlarge the realm of
Ahura Mazda. Thus the religious ideas belonging t o the figure of Vohu Manahl
Manao Bago excellently fitted Kushan royal ideology in the context of Kan-
ishka's victories and conquests. The function of Vohu Manah among the
Amesha Spentas, and his relationship t o Ahura Mazda, helped to introduce the
principal god of Zoroastrianism himself into the ambience of the divine patrons
of the Kushan king. Mazdah vana represents 'the victorious [Ahura] Mazda'
who triumphs over the Druj, like the Kushan king over his enemies. Of the
local divinities A r d o x ~ o ,Oflagno and Lroaspo were represented on the coin
reverses of Kanishka 11. Lroaspo was an ancient re-Zoroastrian god ensuring
the health of the horses of the Iranian equestrian nomads. H e was also war-
shipped among the north-western Iranian equestrian nomadic tribes and his
name was preserved in the form of the hydronym Dyrapses, reflecting the Ala-
nian outcome "'Dvuvafsa- of Old Iranian "'Druvaspa-. Lruvaspa occurring on
the Kushan coins may represent the Bactrian development of DruvZspa-
Obviously, the Bactrians, who had an excellent cavalry and a fa~nousrace ot
horses, worshipped ~'Druvaspa-> Lruvdspa- since their immigration to Bactrla.
 Religions in the Kushan Lmpire

0ilagnolOplagno was an ancient Indo-Iranian divinity, a warrior god

whose worship was broadly spread among both the western and the eastern Ira-
nians, The name Oflagno represents a local, Eastern Iranian development of
old lranian *VyBrayna-, namely, the consonant cluster -yo- developed into -i-
,, -r-. ~ h u s even
, though this warrior god is well known in the Avrsta, he was
included in the Kushan pantheon not as a Zoroastrian but as a local deity, who
was popular among the eastern Iranian warriors with the bird Viragna on his
helmet, armed with a spear and long sword.
~ r d o x s owas also a local divinity, as is clearly shown by the Manicharan
Persian text M 2, which relates the encounter of the Apostle M i r Ammo with
b g ) ~ dw'xi (Bay A r d spirit) who is named wymndb'n y hwr's'n 'Khorasan's
frontier guard'. T h e local character of Ardoxso is confirmed by the place-name
~ a ~ a r dmentioned
a, by Ptolemy (VI.18.5) in the description of the Paropamisa-
dae. Ardoxso can be identified with the Avestan goddess Ardvi14 who, accord-
ing to the Ardvisur Yar't, bestows the highest royal power over all lands t o her
worshippers. Thus the figure of Ardoxso also fitted the Kushan royal ideology
and enjoyed great popularity among the eastern Iranian population. From the
iconographic view-point Ardoxso was identified with the Hellenistic Tyche,
holding a cornucopia.
Lastly, Pharro, god of the royal splendour and glory, was probably of
Parthian origin. Old Persian farnah-, corresponding to Avestan hvardnah-
'royal splendour', was borrowed from Median farnah- which may perhaps go
back to a Scythian *farnab- < "hvarnah-.I5 In any case "farnab- > farr became a
firm element of Arsacid royal ideology. The adoption of Pharro in the Kushan
pantheon may have been connected with the idea seen in the K i m i m a k - i
Artaxfir-i Papakin, according t o which farr 'royal glory' always abandoned thc
defeated king and went over t o the triumphant one. O n the basis of this idea it
was a natural step o n the part of Kanishka I1 t o introduce the cult of the royal
glory into the religious life of the court, because it left the Parthian king Volo-
gases, who was defeated by Kanishka and went over to the conqueror.
An important moment in the religious activity of Kanishka I1 was the res-
toration of the dynastic sanctuary at Surkh Kotal. The renewal o r introduction
of the cult of the goddess Oanindo (Victory) was obviously connected with his
great victory over the Parthian king. H e sent an officer, Nokonzoko by name,
to the sanctuary in Year 31 (A.D. 165). By digging a well, Nokonzoko ensured
the water supply of the stronghold and the sanctuary, and leading back the sta-
tues of the gods, renewed the cult of the dynastic divinities there.
The coinage of Huvishka also provides rich evidence for the religions and
cults of the kingdom. T h e question must be raised, however: What chronolog-
ical relationship cdn be established between the coinage of Kanishka 11 and that

14. Harmatta, 1960, pp. 198 et seq.

15. Lecoq, 1987, p. 678.
of Huvishka? O n the basis of the epigraphic sources it is clear that the religious
activity of the two kings, being co-rulers for at least a decade, continued i,,
parallel. The religious activity of Huvishka was particularly intensive. ~ ~ ~
ing to the Bactrian inscription of Ayrtain (see Chapter 17), in Year 30 (A.u,
164), he sent his officer Sodila as treasurer t o the sanctuary there and had a
Pharro-Ardoxso image prepared and set up in the stronghold. Later, [he
river changed its course and the sanctuary became waterless, he had the divin-
ities and their cult transferred to another place. Then, by his officer Sodila, he
had a well and a water-conduit dug, and having ensured the water supply, he
resettled the cult of Pharro and Ardoxso into the sanctuary of Ayrtam. These
events may be dated between Years 30 and 40 (i.e. A.D. 164 to 174).
Similar activity by Huvishka can be seen at Mathura. The dynastic sanctu-
ary of the Kushan kings built by Vima Kadphises was in a ruinous state when
Huvishka sent a great general (mahadandaniyaka), w h o had the sanctuary reS-
tored and set up a statue of Huvishka in the devakula, ensuring regular hospi-
tality for the Brahmans in the assembly hall belonging to the sanctuary. Even
though the date of the inscription is not preserved o r it was not dated, the resto-
ration work can be dated t o a later period, perhaps after Year 40, when Huv-
ishka was already bearing the title r q a t i r i j a (King of Kings).
The parallel rule of Kanishka I1 and Huvishka also raises the question of
whether they minted coins in parallel. If this was the case, it would be easy to
explain why Huvishka used the device of Kanishka I1 on his first issues. Being
the brother of Visishka, Huvishka may have been substantially older than Kan-
ishka 11, and even though he apparently outlived him, it is improbable that he
would have still been alive up t o around Year 60. We probably have to reckon
with two Huvishkas, father and son, and t o divide the coinage between them.
O n e possibility for the division lies in the remarkable change in the Kushan
ant he on represented in Huvishka's reverse types.I6
It is very likely that the minting of Huvishka I began in ~arallelwith that
of Kanishka 11. At that time, the mints of Huvishka employed the device of
Kanishka I1 and used as reverse types the same divine patrons of the dynasty,
namely, Miiro, Pharro, Mao, Nanasao and Oeso with Manao Bago and Ardoxso
who also occur. Mozdooano is missing from the divinities represented on the
coin reverses. Instead of him, we find Serapis, the supreme deity of the Alex-
andrian pantheon whose name appears in the Bactrianized form of Sarapo. His
emergence seems to indicate the orientation of Huvishka towards Roman
Egypt, an important market for the wares imported from or through the
Kushan Empire. Also omitted is the ancient Iranian war god Oi-lagno, whose
place and function are occupied by a group of Indian war gods, Skando (Old
Indian Skanda), Komaro (Old Indian Kuiniira), Maaseno (Old Indian Maha-
sena), Bizago (Old Indian ViSikha), and even O m m o (Old Indian Umi), the
 Religions in the Kushan Empire

of Siva. Their use as reverse types of Huvishka I is clear evidunce for the
new in religious policy of the Kushan king, which war possiblv
by enlisting Indian warriors into the Kushan army during [he cam-
paign against Plraliputra.
~ l s interesting
o is the omission of Buddha from the reverse types of H ~ ~ -
iShka. This is surprising because according t o the R i j a t a r ~ t i ~ i nHuvishka
i, SUP-
ported ~ u d d h i s m and
, the existence at Mathura of 'the monasterv of the Great
King, the King of Kings, the Son of God, Huvishka'" proves beyond doubt that
the literary evidence corresponded with reality. The omission of Buddha from
[he coin types showing the divine helpers of the Kushan king requires an exp]a-
naion. The ~ r o b l e mis closely connected with the function and meaning of the
gods portrayed o n the coins. They could indicate worship by the kings of the
divinities represented, o r protection by the gods that the king worshipped.
Sometimes they may refer t o pious gifts, or represent statues set up in a sanctu-
ary, like the statue of Ardoxqo in Huvishka's issue" which had been set up in a
Buddhist sanctuary. This indicates the king's favour to Buddhism, and the ten-
dency of Buddhism t o absorb local cults. The omission of Buddha from his
pantheon of reverse types did not in itself mean that Huvishka neglected Budd-
hism, because his royal favour is seen in support of the local cults absorbed by
Buddhism.

New trends in the second phase of Huvishka

The second period of the coinage of Huvishka (perhaps Huvishka 11) reveals
some remarkable new trends. Beside the ancient divine patrons of the dynasty -
Miiro, Mao, Nana, Oeso, ABso 'the royal fire' and Pharro 'the royal splendour'
- further Zoroastrian and local deities appeared. Among the Zoroastrian divin-
ities, the emergence of Ooromozdo, the supreme god, is important. While
Mazdo oano 'Mazda the victorious' represented the Bactrian form of the
supreme god of Zoroastrianisln in the effigy of a Kushan horseman, the pho-
netic form of the name Ohromozdo clearly points to western Zoroastrianism.
The other Zoroastrian deities - Saoreoro 'best royal power', Asaeixso 'best
righteousness' (= Avestan XiaOrdm vairim and Avestan Aia vahiita), Ri~toIRisti
'uprightness' (= Avestan Ariti/Arit~?t)'~- represent important aspects of royal
ideology. Among them Saoreoro seems to be again of west Iranian origin, that
is, he was adopted into the Kushan ant he on from western Zoroastrianism. O n
the contrary, Asaeixso and Rigti are apparently local developments of Avestan
Aia VahiSta and ~ r & i Probably,
. the enigmatic legend Auabod also belongs to

17. Liiders, 1961, p. 68.

18. Harmatta, 1986, p. 136.
19. Grenet, 1987, p. 42.
 t a , rhr. contributions of B. N. Puri, L. Lelekov, 5. Humayun and D,
j. / ~ , ~ ~ n l a twith c ~j~~~~~

the ambience of Ohromozdo. In view of the fact that the name ends with a con-
sonant while in Bactrian each word has a final vowel, the spelling 'Auabod.
must represent an abridged form. Very likely the full form of the name can be
restored as *Ahu budano 'supreme lord of the creatures' (< Old Iranian * ~ h ,
b i t i n a m ) , being a Bactrian name for Mithra, parallel to his Avestan designation
ahu ratuita gaieanam 'supreme lord and judge of the living being'. Together
with the legend Ahubud(an0) the effigy of Mithra appears on the coin, i.e. ice-
nography and legend are in harmony with each other.
The other remarkable tendency is the emergence of the local divinities on
coin reverses. Beside Lroaspo, already introduced by Kanishka 11, Oaxso and
Iamso now appear. Oaxso was the well-known eastern Iranian god of waters
and rivers, in particular the deity of the Oxus river. His sanctuary was dis-
covered at Takht-i Sangin, o n the northern bank of the Amu Darya. His pop-
ularity and importance are best illustrated by the inscription on a seal: Oaxso j
iogo bayo 'Oaxso is the only god'. Iamso may again be identified with Imra
(< Yama r*), the supreme god of the Kafiri (or Nuristani) tribes.20The form
possibly reflects a popular dialect variant of the Bactrian *lam0 sao. The emer-
gence of the goddess Oanindo (Victory) on the coins of Huvishka I1 may have
completed the group of divine patrons of the dynasty and can perhaps be
brought into connection with the renewal of the Oanindo sanctuary at Surkh
Kotal.
In religious policy, as reflected in his coinage, the efforts of Huvishka
were obviously intended t o enlarge the social basis of his rule by religious ideol-
ogy, that supported all the local cults and Bactrian Zoroastrianism among the
population of eastern Iran. 'The divine figures on Kushan coin reverses reflect
the religious ideas and policy of the Kushan kings, but indirectly they also mir-
ror the general trends of religious life - a very complex hen omen on under the
Great Kushans, as we see at Mathura.
In the Kushan period there were numerous sanctuaries of different cults
in the environs of Mathura. The Buddhists had about fifteen monasteries, three
sanctuaries and numerous stupas; the Jains had three temples, and several stu-
pas, there were three niiga shrines, the sanctuary of the yaksa Miinibhadra and
the royal dynastic sanctuary of the Great Kushans. From the inscriptions, we
can follow the fortunes of particular sanctuaries and monasteries. Different
Buddhist schools, the Sarviistividins, the Mahisiinghikas, the Samitiyas and the
Mahopadeiakas, proclaimed their teaching at the same time. The golden age of
Mathura seems t o have been the time of Huvishka, from which the greatest
number of dedicatory inscriptions are reserved. Religious life in Mathura was
characterized by the co-existence of the great religions and their cults, mutuallv
influencing and enriching each other.

20. Grenet, 1984, p. 260.

 Rrligions in the Kushan Empirc

syncretism and absorption

In of the scanty evidence, fragmentary in many respects, we can draw
general conclusions about religious life throughout the territory of
~ ~ ~ h aItn was
s , highly developed and differentiated. The religious movements
of India - Sivaism, Vishnuism, Jainisln and Buddhism with their differcnt
- ~ e n e t r a t e dCentral Asia, as did Indian merchants when Kushan rule
facilitated long-distance international trade. In eastern Iran the Indian religions
met the Greek divinities, Zoroastrianism and many local pre-Zoroastrian forms
of worship, and encountered the ancient Iranian religious ideas of the northern
Iranian equestrian nomads. The Kushan kings selected for themselves from this
immense variety those religious elements, ideas and forms of cults which fitted
their ancient religious traditions and strengthened the religious ideology of their
royal power. So the 'Kushan pantheon' appearing on the coins represents only a
selection of the religious cults of their empire.
None the less, the Kushan kings were well aware of the current trends in
religious life and followed them. The most important was syncretism. The great
religions influenced one another and began slowly to absorb the local cults. In
Bactria the syncretic cult of Siva achieved great success; and on the coins of
Bazodeo (Viisudeva), the last Great Kushan king, Siva was the sole divinity
used, a figure that apparently combined Greek, Iranian and non-Sivaite Indian
elements.
In eastern Bactria and Gandhira the worship of Ardoxso became predo-
minant, absorbing some features of the local yaksi cults, of the worship of
Laksmi and other minor Indian female divinities with elements of the Hellenis-
tic Tyche. She became identified together with Pharro with the Indian couple
Kubera and HZriti, King and Queen of the yaksas and yaksis. Consequently
after the Sasanian conquest of Kushanshahr (the western part of the Kushan
kingdom), the independent eastern Kushan kings made use of Ardoxro for the
reverse of their coins. T h e syncretic character of the goddess is clearly shown by
the legend p k r i o n the coins of Gadahara.'l But while these two divinities, Sira
and Ardoxso, became predominant as divine patrons, their figures had absorbed
many features of other divinities and had a syncretic character. Syncretism and
absorption had finally rev ailed in the 'Kushan g ant he on'.

21. Cunningharn, 1971, No. 10.

 G. A. Pugachenkova, S. R. Dar, R. C. Sharma
and M. A. Joyenda,
in collaboration with H. Siddiqi

Kushan art in the north

Art was not uniform in style throughout the vast possessions of the Kushans.
Several local centres and distinctive schools have been identified, and the
Kushan Empire may be divided into four principal artistic regions: Bactria
(Tocharistan in the basin of the Oxus (Amu Darya) and its tributaries); Aracho-
sia and Nagarahira (Ninhar, n o w Kabulistan, and the Jalalabad Province of
Afghanistan); Gandhira in Pakistan; and Mathura in India. Bactria is rooted in
the Graeco-Bactrian traditions and Mathura in the Indian ones, while both geo-
graphically and in terms of history and culture, the second and third regions,
each of which had its o w n independent origins, held the middle ground. Despite
regional differences and variations, the fact remains that these areas were polit-
ically united under a single state, and this helped the pooling of ideas in various
fields of artistic culture, which finally led t o the shared stock of themes, images
and attitudes that make it ~ o s s i b l et o view the arts of the Kushans as a single
entity.

TOWN PLANNING A N D ARCHITECTURE

The rapid growth of towns in this age went hand in hand with an increasingly
vigorous urban culture. This process involved in the first instance the art of
building cities. Some settlements, such as Balkh (Bactra),' Dilberjin,? Termez'
and Dalverzin-tepe (Fig. 1)4 in Bactria, sprang u p around an original Graeco-

:' See Map 7.

1. Le Berre and Schlumbcrgcr, 1964, pp. 70 et seq.
2. Dolgorukov, 1984, pp. 55 et seq.
3. Shishkin, 1941, pp. 123 et seq., Plate 73.
4. Pugachenkova '~ndRtvelaJze, 1978, pp. 7 ct seq., Plate 2.
 a , R. Day, R. C. Sharma, M. A. Joyenda and H. Siddiqi
G, A. P ~ ~ a c h e n k o v S.

FIG. 1. Plan of the fortress of Dalverzin-tepe.

 Kushan art

Bactriancore which became the administrative and military centre or citadel.

These were generally laid out in the form of a rectangle, which left room for
subsequent development and was surrounded by fortress walls and a moat.
Later there were improvements in fortification t e ~ h n i ~ u eCitadels
s.~ were set
apart, and city walls were made extremely strong and were flanked by several
towers, mostly rectangular but some half-round. Passages and casemates were
built into the walls and bowmen's chambers into the towers. Both were pierced
by countless arrow-slits, real o r blind, while along the parapet lay passage walk-
ways for the defenders and mountings for balistas. The walls were 8-12 m thick
at the base and 15-20 m high. Within the walls, the towns consisted of close-
packed blocks of buildings in strict alignment with public squares, palaces and
temples.
The sheer size of the buildings with which the architects had to deal for-
ced them to devise new structural techniques. In Bactria, building materials
were largely of clay such as sun-baked brick and pakhsa (a kind of adobe), and
most structures were made of these materials. While beamed roofs were
employed, pit-head vaults were designed for elongated structures, and, in square
buildings, the densely patterned 'closed vault' was devised.
Exterior decoration was sober in the western part of the empire. The
smooth stucco of the walls was relieved only by the slit-like window openings
and the cornice edges of the flat roofs with their salient beam-ends, occasionally
surmounted by a battlemented parapet. But inside, the main rooms were dec-
orated with wall sculptures and paintings. The Bactrian Kushan architectural
order figured prominently, in sets of either free-standing columns or wall pilas-
ters. Columns were used in porticos o r aiwans on the front facade, and in large
halls to support the roof beams. They were of wood, but often rested on stone
bases the shape of which followed either the Old Iranian tradition of a massive
torus on a square ~ l i n t h or
, ~ the Attic style inherited from the Greeks.' The
pilasters were made of clay, stone or gypsum, and their capitals were variants of
the Corinthian order, generally squat in proportion and adorned with two or
three rows of heavy acanthus leaves. There were, however, different types for
which the term 'composite' might be more apt. In Buddhist buildings, the figure
of a Bodhisattva o r a gandharva might nestle among the acanthus leaves of the
capitals, as at Surkh Kotal o r T e r m e ~In
. ~Bactria, the capitals were highly indi-

Le Berre and Schlurnberger, 1964, Plates 15-18; Pugachenkova and Rtveladze, 1978, pp.
21 et seq.; Pugachcnkova, 1979, pp. 47 et seq., 1984, pp. 93 et seq.; Dolgorukov, 1984,
pp. 58 et seq.; S c l ~ l u r n b e r ~ et
e r al., 1983, p. 185, Plates 11, X, XXV; ~ t v e l a d z e ,1982.
Pugachenkova, 1966, p. 132, Plates 29, 79, 1973, Plate 7.
Pugachenkova, 1966, Plate 79, 1973, Platc 7, 1976, Plates 72-5; Pugachenkova and
Rtveladze, 1978, y . 199, Plate 131; Schlulnberger et al., 1983, Plates XX, XXII-XXIV-
XXX-XXXIV, LVIII-LX.
8. Schlun~bergeret a]., 1983, Plates LXVI-LXVIII; Pugachenkova, 1979, p. 55.
 G, A. ~ ~ ~ n c h c n k oS.- R.
~ a Dar,
, R. C. Shartnn, M . A. Joycnda and H . Siddiqi

vidualized; among the acanthus leaves could be seen two lion griffins back-to.
back, or a pair of zebu bulls with a fabulous bird-creature clawing at them in
between. Such examples can be seen at Termez (Sham-kala) (Fig. 2) and Shahr-i
Nau (Fig. 3).'
Like the public and religious buildings, the homes of the wealthy f0l-
lowed distinctive architectural designs. Some were patterned on the architectu-
ral norms of Bactria established in the previous period, while others revealed
new features. The palaces and homes of the urban aristocracy were laid out
either with a central hall and vestibule or with a courtyard - the whole being
surrounded by a corridor. Accommodation and auxiliary premises, as at Kha\-
&ayan, Dalverzin-tepe (Fig. 4) and Dilberjin (Fig. 5),1° were also provided for.
The same concept - a hall with a corridor and possibly outbuildings around it -
is typical of temples of the local cults in Kushan Bactria, whether Zoroastrian or

Flc;. 2. Fragment of a capitdl f r o m T c r m e z

9. Staviskiy, 1981, pp. 125 et scq., Plates 93-4; Dagens, 1960, pp. 38 et scq., 1968, pp. 36ct
seq.
10. P ~ ~ a c h e n k o v 1966,
a, Plate 23, 1976, p. 91; Pugachcnkova a n d Ktvcl~d;.c, 1978, Plates 15.
26.
 Kushan art

FIG. 3. Stone capital from Shahr-i Nau.

FIG.4. Dwelling-houses in Dalverzin-tepe (reconstrucrion).

 , , ~ , t ~ i l c h o i k a oSr , K Dar, K . C SJ~arrna.M . A. Joyymda and H , SiJJjrl,

FIG.5. Plan of a dwelling-house in Dilberjin.

dynastic, as at Takht-i Sangin (Fig. 6),l DilberjinI2 o r Surkh Kotal." Externally,

these residences, palaces and temples took their shape f r o m the rectangular mass
of the central hall o r sanctuary. O n the f a ~ a d e s nothing
, broke the smoothness
of the walls, the principal o n e being identified b y a colonnaded portico 01.
aiwan.

T E M P L E S ANL) BUI1L)HIST' B U I L L > I N G S

T h e temple buildings at Surkh Kotal (Fig. 7), dedicated t o the dynastic cult of
the Kushans, rnay be viewed as o f representative character. T h e y stand on a hill
f r o m the foot of which rise five flights of steps. High fortress walls flanked with
rectangular towers protect the square courtyard and the galleries running round
it, while in the centre, raised o n a platform, looms the chief temple built in Kan-
ishka's times by his official, N o k o n z o k o . T h e exterior is surrounded by a
colonnaded portico, while the four-columned shrine housing the altar is flanked
on three sides by an ambulatory passage. In course of time t w o further struc-
tures of a religious nature were built in the traditional Bactrian manner - a
square hall surrounded by a corridor divided by passageways - inside and out-
side the courtyard.

11. Litvinsky a n d Pichikyan, 1981, 12igs.2-3.

12. Kruglikova, 1982, Platc 0.
13. Schlumbcrgcr ct dl., 1983, Plates IX, XXXVI-XXXVIII
 Kushan rxrt

FIG.6. Colonnaded po~-tiro.Takhr-i Smgin. (Photo: D J'ladirnir Tcrebcnin.)

As Buddhism spread from India to the western parts of the Kushan

Empire, Buddhist buildings sucb as monasteries, stupas and shrines spring up
and their remains have been found at T e r ~ n e z(Fig. 8)," Ayrtam," Dilberjin'"
and Surkh Kotal." Architecturally, buildings in these areas are somewhat differ-
ent from Buddhist structures found in the Indian parts of the empire. They
accepted a blending of different architectural settings used for decoratilre pur-
poses. Wall sculptures and paintings were used as part and parcel of the dec-

14. Staviskiy, 196182; Pugachenkora, 1967, pp. 257 et seq.; Al'baum, 1982, pp. 56 et Seq.
15. Masson, 1976, pp. 81 ct seq.
16. Kruglikova and Pugacl1enko\la, 1977, pp. 61 et seq.
17. e r a!., 1983, pp. 75 et srq.. Plates XL\'II-XLVIII. LV.
S c h l ~ l n b e r ~ et
 a , R. Dar, R. C. Sharma, M. A. loyenda and H. Siddiqr
G. A. P ~ ~ a c h r n k o vS.

FIG.7. The sanctuary at Surkh Kotal (reconstruction).

FIG.8. Buddhist stupa, 'Bastion of Zurmal', at Tcrmcz.

oration of aiwans, main halls and shrines, while the facing of Buddhist stupas
was invariably relieved by representational sculptures.
Murals were constructed in a kind of glue-based tempcra laid on a thick
rendering of clay with vegetable additives (which was partly responsible for
crumbling when the additives decayed) either directly o r on a white ground.
The colour range was small with a predominance of red, black and white; ye]-
low, blue and green were rarely used. The additional hue of white, however,
enabled a whole spectrum of shades t o bc achieved.
 Kushan art

Only fragments of murals have survived but these command attention in

respectof the variety of themes and motifs. Paintings of people of different eth-
,ic groups are noticed, along with figures with real animals like the horse, or
imaginaryones like the griffin. So are scenes of court life and subjects taken
from the religious beliefs and current myths of the time in Bactria. In the Budd-
hist monasteries of Kara-tepe and Fayaz-tepe at Termez, for example, portraits
,f the ~ u d d h a a, monk and benefactors in typical Kushan costume have been
discovered.'nIn Dilberjin, after the Graeco-Bactrian temple of the Dioscuri had
been rededicated t o h a , a scene was added showing Siva, Parvati and the bull
~ a n d i(Fig. 9).Iy The temple of the Bactrian goddess in Dalverzin-tepe con-
tained a ~ a i n t i n gof her seated on a throne, and a representation of an unknown
ritual in which a priest and priestesses offer small children for her blessing (Fig.
Mural decoration also incorporated ornamental motifs, as the classical pal-
mettos in the temple of the Bactrian goddess at Dalverzin-tepe, reflecting the
impact of the Greek tradition. Others clearly represent patterned fabrics: the
Khalchayan palace had unconstrained shoots, foliage, clusters of grapes, violets
and rounded fruits painted white against a dark red background. Others again
consist of purely ornamental latticework interwoven with rings, as in Kara-tepe
at Termez.
All these fragments testify to great professional skill on the part of the
artists, and probably point t o the existence of special guilds of artists in large
cities. At the same time, they record traces both of Greek influence and of the
gadual assimilation of that influence with the emerging new style. Apart from
these paintings, it is the sculpture ~ r o v i d i n gdecoration for buildings that com-
mands great admiration. Among the artistic achievements of antiquity, the Cen-
tral Asian sculptures of the Kushan period now rank among the finest. In Bac-
tria, sculptures were usually of clay, finished in paint or laster, though some
were of white marble-like limestone.

M O N U M E N T A L SCULPTURE

The range of subjects and scenes in Bactrian monumental sculpture was un-
usually wide and varied, both secular and religious, dynastic and public. Of the
dynastic groups, particularly illuminating are the sculptures at Khalchayan dat-
ing from the beginning of the Christian era," Dalverzin-tepe from the first cen-
tury A . D . ~and
~ Surkh Kotal (second century A.D.)." In the palace at Khal-

18. Staviskiy, 1972, Tables IV and V; Al'bau111, 1975.

19. Kruglikova, 1974, p. 44, Plate 30.
20. Pugadlenkova and Rtveladze, 1978, pp. 79 et seq., Plates 50-5; Pugachenkova, 1979,
Plates 194, 196, 199, 200.
21. Pugachenkova, 1971, pp. 153 et seq.
22. Pugachcnkova, 1979, 1-7~.131 et seq., Figs. 1 2 3 4 .
e r al., 1983, Plates LVIII-LXXI.
23. S ~ h l u m b e r ~ et
 FIG.9. Siva, Parvati and t h e guard. Painting in the sanctuary of Dilberjin.

chayan, complete scenes were carved around the walls of the main hall and the
aiwan. In the hall, the centre-piece was a scene of imperial splendour - a
Kushan monarch of the lineage of Heraus, seated o n a throne with his lady.
Above them are Nike and Heracles, and on both sides are men and women of
the imperial household. T o the right is another dynastic scene - the head of the
family shown seated while others stand around in formal attitudes. T o the left is
a battle scene with mounted heroes in armour and helmets, and lightly armed
bowmen shooting as they advance (Figs. 11-14).
All the characters are portrayed in a completely individual manner and are
clearly taken from life. Their different ethnic origins are accentuated, Bactrian
and Parthian princes appear, and the majority of the figures are supposed to be
the Kushan clan of Heraus, the monarch well known from his image on coins.
The head compressed at the front and back, the eyes stretching lynx-like to the
temples, the straight nose, the finely drawn moustache and sideboards, the
straight hair caught u p in a fillet o r circlet - all these are tribal characteristics of
the whole of Heraus' lineage (Fig. 15).
This sculpture was executed in the expressively realistic style that the
 Kushan art

F I G . 10. Ritual scene. Fragment of a painting from Dalversin-tepc.

(Photo: O Vladirnir Terebenin.)

FIG.1 I . ljdll ot the p ~ l . ~ cinr KhJcha!*.~n (reconstruction).

 FIG. 12. Statue of a queen. Khalchayan. Painted clay.

Greeks brought t o Bactria, though the images thcrnselves are emphatically local.
A pronounced interest in human individuality is revealed in the portrayal not
only of the subject's physical features and age, but also of his emotions. The
same style may be seen in the sculpted head of the great goddess of Bactria in
the Dalverzin-tepe temple,2i which dates from the same period, and the image of
a crowned local ruler from the temple at Dilberjin.zi
The main scenes at Khalchayan are topped by a sculpted frieze showing
amorinis, naked o r in flowing tunics, holding garlands with the busts of actors,
musicians, satyrs and mummers inset.'" The Hellenistic inspiration for this is

24. Pugachenkova and Rtveladze, 1978, Platc 56; Pugachenkova, 1979, Plates 175-6.
25. Kruglikova, 1974, Table I .
26. Pugachenkova, 1971, pp. 22 et scq., 1979, Figs. 135-6.
 FIG.13. Statue of a Kushan Yueh-chih prince. Khalchayan. Painted cla!..

clearly evident, but the characters are all Asian, o r Inore specifically Bactrian
Kushan, in appearance. T h e identical motif of amorinis holding garlands is
carved on a second-century slab from Surkh Kotal.?'
The cult of Heracles, identified in Bactria with a local divinity o r demi-
god, continued until the very end of the Kushan period. A small first-century
statuette of this demigod was found in Takht-i Sangin2%hile his painted clay
statue of the second century was found in the garrison built along the gatewa!.
bastion at Dilberjin.2v
A considerable inlpact o n the development of sculpture throughout the

c r a]., 1983, Plate LV.

27. S ~ h l u m b c r ~ et
28. Litvinsky and Pichikvan, 1981, Fig. 12.
29. Pugachenkovn, 1977, pp. 77 et scq.
 FIG. 14. Kushan Yiiel~-chihhorseman. Khalchnyan. Pai~~ted
clay,

FIG.15. Head of a warrior. K l ~ n l c l i a ~ a nPainted

. clay. (Photo: 0 Vl,~dimirTercbenin.)
 Kushan art

~ ~ ~ region
h a nwas made by Buddhism, the basic images, subjects and topics of
which were developed in Gandhira and spread to the north-west of the empire.
Here they were assimilated with earlier indignous traditions. A case i n p i n t is
a recond-cent~rysculptured frieze from Ayrtam in which female figures half-
emerge from acanthus leaves.jOThese are the celestial musicians or gandharvas,
and girls bearing offerings of flowers, garlands or vessels. But in their ethnic
appeararlce, head-dresses and ornaments they differ from those typical of India,
and clearly portray local characteristics.
In ~ o r t r a y i n gthe image of the Buddha (Fig. 16), the Bactrian sculptors
followed the established canons of the beginning of the Christian era, but for
the secondary figures in the Buddhist pantheon - devatis (Figs. 17 and 18),
genies or gandharvas - they returned to earlier Graeco-Bactrian traditions.
These can be seen in figures such as the heads of devatas from the Buddhist
shrine at Dalverzin-tepe, whose softly modelled features and fleeting smiles are
reminiscent of the school of Pra~iteles.~'
The Bactrians also differed in their portrayal of lay devotees of Buddhism.
The Kushan prince in his pointed head-dress, great ladies with costly hairbands
and a magnate from the same Dalverzin shrine are notable for reflecting their
personal features. As distinct, however, from Khalchayan, sculptures from Dal-
verzin (Figs. 19 and 20) suggest a smoother moulding of the features without
any sign of age, and controlled, almost non-existent emotions. They correspond
not only t o the ethical standards and want of sensuality prescribed by Budd-
hism, but also t o the new tendencies in sculpture in which the emphasis in por-
traiture shifts from the individual to the formal. Even more generalized and for-
mal is the dynastic sculpture of Surkh Kotal, which represents the chief
emperors of the Great Kushan dynasty (Fig. 21), including the great emperor
Kanishka noted for the shape of his head-dress.

THE M I N O R ARTS

The hallmark of the various 'minor arts' in Kushan Bactria is seen from the arte-
facts made by professional craftsmen. Moulded o r hand-made terracotta sta-
tuettes became widespread, the former most commonly figures of the great god-
desses worshipped locally and bound up with ancient folk cults of the mother
goddess, the patron of fertility, childbirth and prosperity (Figs. 22 and 23).
They differed from region to region on the basis of features, head-dress and
clothing, apparentlv in response t o local variations in population." Less fre-

30. Trever, 1940, pp. 149 et seq., Tables 45-9.

31. Pugachenkova, 1979, p. 167, Tables 204-5.
32. Gardin, 1957, Plate X; Pugachenkova, 1979, Plates 178, 179, 181-9; Iirugliko\,a, 197.1,
Plates 26, 56, 68; Krugliko\ra and Pugaclienkova, 1977, Plates 32, 103; ~ e s h k e r i s ,1969,
Plates 23-5.
G, A, ~ ~ ~ a c h e n k oSv nR., Day, R. C. Sbarma, M. A. Joyrnda and H Stddlpl
 I.rc,. 17. I lecld of ' d c \ n t j . I>al\cr/ln-tepc. G J p j u n i o ~ (1.1).
1 i
( P h o t o : v? Vladimil- T c r c b e n i ~ ~ . )

quent are moulded inale statuettes, usuall\- in Kushan costu~ne,either as the

goddesses' male companion or as a demigod." But crudel!. f ~ s h i o n e dfigurines
of mounted h o r s e n ~ e nwere found far and wide, and are thou,]0 i t to relate to the
cult of the ancestor-god brought to the cities by the steppe tribes." The import
of ivory from India gave rise to the art of representational c a r ~ i n g(Fig. 24). 011
a plaque from 'l'akht-i Sangin dating from tlie beginning of tlie Christian era, for
example, there is a djrnamic hunting scene in which t w o horsemen are repre-
..
sented in a m,~nnei-ver\. reminiscent of tlie sculptured bowmen of Klialchayan:~

33. Gal-din, 1157, I'l~tc' X; P ~ ~ ~ . ~ ~ l l c1973. o v . 3~ 1.,

~ ~ kPlatc
34. Pugachcnko\,,l, 1005, pl3. 24s ct scq.
35. I,it\:insli\- J I I ~T'icliik\,.in, IC)SI. Figs. 15-16.
 G. A. Pugachenkova, S. R. Dar, R. C. Sharma, M. A. Joyenda and H. Siddiqi

FIG.18. Head of a devati. Dalverzin-tepe. Gypsum on clay.

The crafted metal jewellery of Kushan Bactria was shaped by the tastes of
the upper classes. Its outstanding objects were found in the tombs at Tillya-tepe
in northern Afghanistan,jhand include thousands of gold artefacts (Figs. 25 and
26), sometimes encrusted with precious stones - massive necklaces, bracelets,
sheaths, complex composite crowns, finely figured platelets for sewing on
clothing o r shrouds. Some of these are genuine masterpieces of the jeweller's
craft. Many incorporate representational motifs which place them in several sty-
listic groups.
Some of these motifs can be traced to ancient Oriental traditions in art,
such as the open bracelets finished at the ends with the horned heads of lion-
griffins. A statuette of a mountain goat in the round, a suite of rearing, horned
griffins with snarling dog-like heads o r a frontally portrayed scene in which a
royal hero grapples with horned, winged and fish-tailed dragons on either side
are equally interesting. All these images evoke the art of Achaemenid Iran,
though a number of elements in them d o appear alien to that tradition.
Some objects bear traces of Hellenistic influence. Among direct imitations
from Greek art is the image of Aphrodite. H e r Greek features are commonly
recast in the Asian mould, as in the case of naked cupids riding dolphins and the
figurines of a semi-nude winged goddess reclining on a throne (accompanied in
one case by a soaring Eros) o r the war-god Ares with a Greek cuirass and a Par-
tho-Bactrian helmet.

36. Sarianidi, 1985.

 Kushan art

Of p ~ r t i ~ intcl-est
~ l ~ r is the 'animal st!,le'. 'IS it was known. There are
slnall plates depicting rearing dragons o r a beast of prey pulling down a quad-
ruped, and p l ~ a l c ~ -cmbosscd
dc with beasts of prc! o r fabulous zoomorphic cl-ea-
tures rolled into a hall as if biting their o w n t'lils.
It should be noted that siinilal-
gold plialcme with turquoise insets, p r t r a ~ i n gabattle between three
have bcen found in Tnltllr-i S'111gin (Fig. 27).
But t]lc Til]\-a-tcl,ecol]cction is moi-c than J inere s\-ilthe~isof ancient
Oriental, HcllenirCd 'lnd srcppc u t : it contains n c , ~elrments. The faces of the
 G. A. ~ l ~ ~ d ~ . l j c l r kS.o vKn. ,Dilr, R. C'. J%armu, M . A. Joycndu and H . Slddirl,

FIG.20. Statue of a magnate. 1)alvcrzin-tcpe. Painted gypsutn on clay.

goddesses and cupids, the heroes and other characters, the clothing, weapons
and other details are of an intensely local kind. By their technique (known as
the encrustation style) and composition devices most of this collectiol~belongs
t o w h a t is called the Sarmatian culture, which flourished over a large part of
Eurasia from the first century H.<:. t o the first century A.I).
Archaeologists are inclined t o connect the Tillya-tepc burials with the
period of the Early Kushans. However, stylistic parallels are so close betweell
many of these items and the objects of Parthian Nisa and the Saka-Parthian
strata of Taxila that there is far Inore reason t o detect a link with these peoples'
advance into western Bactria, Sistan and then parts of ancient north-wester11
India, yielding t o the Kushans only in the first century A.I).
A hoard of first-century objects of gold f r o m Dalverzin-tcpe, hidden
beneath the doorstep of a wealthy home, illustrates the ncxt phase in the d e d -
.-
oplncnt of Central Asian ornamental metalware. Together with gold discs alld
 Flc;. 21. Statue of 3 ruler ( K ~ ~ n i s l ~ k , iSurkh
? ) . Kotnl. 1''lintc.d g!.psum.

bars, many marked with their weight in Kharosthi characters, the hoard pro-
vided a n u m b e r of crafted ornaments, supposed t o have been lnanufactured b!.
local Bactrian jewellers. These include bracelets, earrings and pectorals. Items o i
Indian craftsmanship irlclude a necklace that WAS probably owned b-a ~ n e n i b c r
of the upper Indian '~ristocrac!. while ,I pl~alel-adepicting a fabulous shaggy
beast rolled illto a ball seelns t o have been executed in the traditions o f Sc!.tho-
Sarmatian animal st)rlc.
Viewed as a wholc, Bactrian art of tlie Kuslian ~ c r i o dis a comples, c n ~ i i -
posite and e ~ . o l r i n gart. Tlic combination of ele~nelltst h ~ date t hack t o the p ~ s t .
Hellenistic ~ n o t i f s ,stcppc ' ~ n i n l n lstrlc' and Indo-Buddhist influences are all
grafted 011 t o 11'1ti1.eB ~ c t r i d ntr,iiiitions and tr,insfnrmed b!- the creati\.it! of the
artist, $ring l-isc t o a ne\\ dificrcnt ph,ise in tlie d c ~ e l o p i n e n tof B i c t r i ~ l i
Kushall .~rti+ticculture,.
 G. A. Pugachmkova, S R. Dar, R. C Sbarma. M.A. Joyenda and H Siddiqi

FIG.22. Statuette of a Bactrian goddess. Terracotta from Dalverzin-tepe.

FIG.23. Statuette of a Bactrian goddess. Tcrracotta from Dalverzin-tepe.

Kushan art in Bactria

SURKH K O T A L

Surkh Kotal, about 14 km north of Pul-i Khumri and 232 km north of Kabul, is
noted f o r dynastic temples that were set u p during the reign of the Great
Kushan king Kanishka and his successors in the second century A.D. In the
main temple a square fire altar was found, suggesting the existence of belief in
 Kfishan art

FIG.24. Comb with engraved figures. Ivory from Dalverzin-tepe.

Zoroastrianism. Schlumberger suggested that the sculptural decoration of the

temples showed a blending of different artistic trends, including steeped mer-
Ions of Oriental tradition. Noticing the stone frieze, the series of unbaked and
painted clay figures, and the damaged stone reliefs, he suggested that the statues
found there should be compared with the enthroned figures at Nimrud Dag of
the mid-first century B.C. representing rulers (or ~ o d s and
) the Kushan images
of Mathura.'"
The statues of a Kushan noble and the Kanishka statue exhibit the same
symmetry, dress and jewellery. Most of the sculpture from Surkh Kotal provides
evidence for indigenous Bactrian art and its relation with the art of Gandhiira.

TIL.LY A - T E P E

Tillya-tepe (Golden Hill), 5 km north of Sheberghan, was excavated b?, the

Afghan-soviet Arcli:leological Mission in 1978179 and )ielded 20.000 gold arte-
facts. It was thc site of a temple in the second millennium B.C. During the rule
of the Achaemcnids ,I magnificent palace was built but it was later damaged b ? ~

38. Rosentield, 1067, pp. 165-7.

 FIG.25. Sitting warrior. Tillya-tepc. (Courtcsy of V. Sarianidi.)

FIG.26. Goddess between t w o dragons. Tillya-tepc. (Courtesy of V. Sarianidi.)

 Kushan avt

1-'1(,. 27. Gold phnlcra. Takht-i Sangin. (Photo: '2, V l , ~ d i m i ~l+crcbcnin.)

-

fire, and c o m p l e t e l ~destroyed

~ b\r the Greek arm\. in the fourth century KC:.
Later, in the first centurv A.I)., the site was used as a gra\.evard for o n e o r t w o
generations in the time of the Kushans. T h e numerous objects of the site's ear-
lier period include G r e e k antiquities of Graeco-Bactrian date. Those from the
later period represent the production of Bactrian artists and can be c o n ~ p a r e d
with the Begram hoard providing evidence of contacts with India, Rome, China
and Parthia." T h e stvle of this group was influenced by local trends, and b!.
Graeco-Bactrian, R o m a n and Parthian art, a combination of styles that sheds
light o n some unsolved problems of the Kushan art of later centuries.

Thc temple o f Dilberjin Kazan, situated 4 k m north-west of Balkh, was es-

cavated b y the Afghan-Soviet Arcl~aeologicalMission o\.er sewral seasons, and
produced painted clay statues and wall paintings, which reveal traces of Bactrian
style influenced b ~Transoxania.
. T h e wall paintings are most important f o r the
study of art in nort11c1-n Afghanistan during the Iiushan period. They can be
cornpared t o thc f r o m B'11alik-tepe in southern U A e k i s t a n , dated b!.
. .
ex-Soviet a r c l ~ ~ i e c ~ l a ~t oi s tllc
t a cnd of the fifth and beglnnlng of the sixth cen-
tury A . I ) 'Thc d.~tingp~-oI>cficcl t ( ~thcsc
- of Dilhcrjin k a z ~ nis the first
 G. A. Pugachenkova, S. R. Dar, R. C. Sharrna, M. A. Joyenda and H. Siddiqi

half of the fifth century A.D. The paintings are not the earliest ones found at
Dilberjin Kazan, which belong t o the Graeco-Bactrian period, according
Kruglikova. The fagade of the temple is decorated with a painting representing
the Dioscuri with their horses, while another painting o n a later wall of
same temple depicts Siva and Parvati seated on the bull Nandi, surrounded by
devotees.

Kushan art in Nagarahara and Arachosia

BEGRAM

Begram, about 64 km north-east of Kabul, has been identified with KapiSa, the
summer capital of the famous Kushan king Kanishka. The DCICgation ArchPo-
logique F r a n ~ a i s een Afghanistan (DAFA) carried out several limited excava-
tions there between 1936 and 1946 but only a small part of the town area has
been dug. The ruins of Begram represent three stages in the history of this
famous city.1° The original foundations were laid out on a plan not markedly
different from Hellenistic cities such as Dura-Europos. This was the capital of
the last of the Graeco-Bactrian kings and the first rulers of the Kushan dynasty.
The second Begram, modified only by the construction of new palaces and for-
tifications, was the northern capital of Kanishka and his successors. It is clear
that the town was violently destroyed by fire, probably at the time of the dis-
astrous invasion of the Sasanians. The last town rose on the ruins and was
probably abandoned with the coming of the Hephthalites in the fifth century
A.D."
The fame of Begram rests o n the discovery of two rooms (probably part
of the 'palace'), which were filled with objects of enormous value - Chinese lac-
quer boxes, Graeco-Roman statuettes in bronze, a collection of fine Roman
glass, Graeco-Roman vessels of porphyry and alabaster and an extraordinary
group of plaster casts (Fig. 28) apparently taken from classical metalwork. In
addition, the treasure rooms yielded a large number of superb Indian ivory
carvings (Figs. 29-31), which had originally served as parts of various articles of
furniture. Many of these objects represent types and techniques otherwise un-
known to classical Graeco-Roman and Indian art. The discoveries at Begram
show that under the Kushans there were close relations at artistic and cultural
levels between Bactria, Gandhira and Rome. The find of such a treasure of clas-
sical luxury goods at KipiSa is thus an interesting heno omen on, which seems to
indicate a prevalent taste for Graeco-Roman art.42

40. MacDowall and Taddei, 1978, p. 257.

41. Rowland, 1976, pp. 24-5.
42. Ibid., p. 28.
- Kushan art

FIG.28. Plaster e~nblemof a slreping macnad. First cclirury n.11. Begrain

(diameter 17.5 cm),
 a , R. Dar, R. C. Sharma. M. A. Joyenda and H.
G. A. ~ u ~ a c h e n k o v S.

FIG.30. Ivory panel. Second century A.U. Begram (height 41 cm).

 Kushan art

HADDA

Had& was the magnificent Buddhist centre of Gandhira about 8 km south of

~ ~ l ~ l a bLike
a d . a golden mirage of towers, the thousand stupas of Nagarahira
,d ~ a d d adrew the Chinese pilgrims Fa-hsien and Hsuan-tsang to this pil-
grimagecentre in the fifth and seventh centuries A . D . respectivcIy. A~ a result of
archaeologicalexcavations by D A F A and the Afghan Archaeological Mission,
several sangharamas, stupas and other constructions such as porches decorated
with statues and paintings have been discovered. The numerous objects found
include statues of the Buddha, of Buddhist monks, donors, Vajrapinis, coins
and pottery. The evidence from the sanghdrimas and stupas of Hadda and the
surrounding areas shows the artistic importance of the region under the Kush-
ans and later rulers. The style of art is associated with the Gandhira school and
exhibits a blending of Bactrian, Graeco-Roman and Indian concepts. There are
sculptures in schist and limestone, but mostly in clay, stucco or lime plaster.
According to Marshall, the stucco sculpture represents a late Indo-Bactrian
renaissance, while Bachhofer considers it as a later development of the Gand-
hira style.43
The problem of the chronology of the stucco sculptures from Hadda
needs further elucidation. While it is certain that the majority belongs to the
Kushan period, it is clear that some of the material excavated should be assigned
to a date before the arrival of the Kushans. The outstanding feature is their
extraordinary skill in portraying human character and emotions, in a way rarely
seen in other parts of the Kushan Empire.
Some pieces from Hadda show the influence of local trends in style and
subject matter. Statues of local people, yaksis, donors with Kushan dress sug-
gest the mixing of the local and foreign elements in Gandhiira art, which,
according t o Schlumberger, had its base here. Finds of artistic material from
Begram, Ay Khanum, Surkh Kotal, Dilberjin Kazan and Tillya-tepe support
this and the excavations of the Afghan Archaeological Mission at Tepe Shotor
in Hadda throw new light on the theory advanced by Schlumberger.

P A I T A V A AND S H O T O R A K

The sanghiriimas at Paitava and Shotorak, near Begram, excavated by DAFA,

were decorated with sculptures that were mostly carvings in the familiar blue-
grey schist of Gandhira. The image of the Buddha and other Buddhist divinities
and representations from the JZtakas figure here. Some of these images have a
hieratic rigidity suggestive of the sculpture of Hatra and Palmyra, and this
resemblance extends to the treatment of the drapery as well. Other canrings

43. Roland, 1976, p. 28.

 , R. Dar, R. C. Sharma, M . A. Joyenda and H.
G. A. P u g a ~ h e n k o v aS.

from these sites, like the monumental reliefs of the Buddha and Kaiyapa,
gest the deeply carved reliefs of the Gandhira style of the second and third ten-
turies A.D. According to Rowland, 'these fragments of stone sculpture from the
region of Begram are of great importance, illustrating the final phase of the
Gandhiira style that was destined to exercise far greater influence on Buddhist
sculptures in Central Asia'.
The influence of a local element is traceable in Paitava carvings. The figure
of a standing donor carrying offerings under the Sila tree is a typical example,
He wears the characteristic Kushan mantle, baggy trousers and felt boots simi-
lar to the dress of the famous statues of King Kanishka from Surkh Kotal and
Mathura. Similar types of donors are ~ o r t r a ~ eatd~ a d d a .

TEPE SARDAR

The early layer of Tepe Sardar, south of Ghazni near the Kabul-Kandahar high-
way, excavated by the Italian Mission in 1959, yielded interesting material
mostly related to the time of the Later Kushans. The excavated layers are simply
a thick filling following the destruction of a rich decorative complex made of
unbaked-clay sculptures. These layers have similarities with those of the Kipiia,
Fondukistan, Tepe Maranjan and Jalalabad areas. The clay sculpture of the ear-
lier ~ h a s eof Tepe Sardar, in its manifold aspects, belongs to the tradition of
Bactrian Hellenism, and shows affinities with the clay images from Surkh Kotal
and Tepe M a r a ~ ~ j a n . ~ ~

The monastery of Tepe Maranjan, o n the eastern outskirts of Kabul, may be

dated to the late fourth century A.D. Its sculptures are executed in clay with a
thin veneer of lime plaster, and appear to be a later development of Hadda style,
anticipating the style of Cave G at Bamiyan and of Fondukistan, in a com-
bination of painting and sculpture. As in other areas of Gandhiiran art, jewellery
distinguishes a Bodhisattva from the Buddha." Similar material was found at
Tepe Khazana, north-west of Kabul town, reflecting the later Gandhira style of
the fifth o r sixth century A.D.

44. MacDowall and Taddei, 1978, pp. 278-9.

45. Snellgrove, 1978, p. 183.
 Kushan art

Kushan art of Gandhara

~ l ~ h the ~ ubeginning
~ h of Kushan art in Bactria can now be traced, as is shown
by [he excavations at Khalchayan4hnd Ay Khanurn," very little work has been
done so far t o establish its origin in Gandhira proper. This is because the
Kushan art of Gandhara has SO far been studied from a limited
Kushan material has been excavated from only three major cities within the cul-
tural periphery of Gandhira - at Begram (or K i ~ i S a ) ,Puskalivati
~" (or Peucelao-
t;s)49 and Taxilaso - and from the two sites outside the Indus region, that is,

Mathura5' and Surkh K0ta1.~*Takht-i Bahi, Jarnal Garhi, Sahr-i Bahlol, Shah-ji-
ki-Dheri (Peshawar), Tharelli, Mekha Sanda, Nathu, Sanghao, Hadda, Mani-
kyala, Rani Ghat, etc. have yielded material for study. Besides these, at least
twenty Kharoghi inscriptions and numerous gold and copper coins have been
found. But although we know about many Buddhist sculptures53of the Kushan
period from the region, Gandhara art has so far been studied only for the sake of
its sculptural wealth and Buddhist religion or t o detect Western influence, never
with a view to studying the civilization of which the sculptures were a part.54

GANDHARAN ART AND BUDDHISM

The Kushan sculptures from Gandhara are predominantly Buddhist. Although

the Buddha himself never visited Gandhara, with the passage of time the area
became a veritable holy land for his followers. Several spots were identified as
having an alleged association with the Buddha in his pre-natal existence, and
over these, stupas and monasteries were built. In early Buddhism, introduced
here by ASoka (third century B.c.), the Buddha was never represented in human
form. But constant exposure of the G a n d h ~ r aBuddhists to the art and pantheis-
tic religion of the Western world created a schism between the purist and the
more forward-looking Buddhists. Whether out of conviction or as an act of lib-
eral patronage, Kanishka is said t o have convened the fourth Buddhist Council
at Kuvana near Jullundur (or at Kandalavana in Kashmir), which finally put an
end to the dissensions that had distracted the Buddhist Church for nearly a cen-

46. Pugachenkova, 197 1.

47. Bernard, 1967; Bernard et al., 1973.
48. Ghirshman, 1946.
49. Dani, 1965166.
50. Marshall, 1951.
51. Vogel, 1910; Rosenfield, 1967, p. 41.
52. Schlu~nber~er, 1955, 1961, pp. 77-95.
53. Ingholt, 1957; Marshall, 1960, pp. 63-108.
54. Dani, 1969, p. 2.
 a , R. Dar, R. C. Sharmd, M. A. Joyenda and H, Siddiqi
G. A. P ~ ~ a c h c n k o v S.

tury.i As a result, Mahayanism - a liberal and progressive school of ~ ~ d d h i ~

- flourished in Gandhira and laid emphasis, among other things, on the transfor.
Ination of the Buddha into a great mythological, almost eternal, god, and on the
deification of future Buddhas as holding pro~idences.~"n the visual arts, [he
Buddha was permitted for the first time to be represented in human form. lt is a
moot point where and when the first image of the Buddha was made - in Gand-
hira o r at Mathura." Probably, it developed simultaneously at both places, one
developing directly out of indigenous Indian art and the other (Gandhira) bar-
rowing its type from the West. ~ o r m e r l yit was generally held that the earliest
Buddha images belonged to the time of Kanishka, but long ago Marshall wrote of
an Early G a n d h ~ r astyle developing during the first century A.D. New excavations
by the Italian Archaeological Mission in Swat and a reassessment of the Taxila evi-
dence have led several scholars to ~ u s back
h the date even to the first century B.C.

T H E DEVELOPMENT OF BUDDHIST AKT

The sudden liberalization of Buddhism was a signal for the development of

Buddhist art in Gandhiira and it soon reached its peak. However, it is wrong to
say that the profusion and popularity of Gandhira art owes its existence to the
state-owned Church o r that its distinctive appearance is indebted to the
mechanical product of higher craftsmanship from the West.5HAs a matter of
fact, Gandhiira art simply expresses the socio-religious fervour of its people. An
indigenous socio-economic stratum of cultivated taste - the merchants, bankers,
caravaners and minor officials - and not the Kushan nobility may have provided
the main impetus for its development.^') The Kushans themselves never extended
any official patronage t o it. They were eclectic in religion, fire-worshippers in
Bactria, Buddhists in Gandhira and Hindus in Mathura.'O

SCULPTUKI

The Kushan art of Gandhira is mainly known from the wealth of sculpture
recovered from the numerous Buddhist stupas and monasteries throughout
Gandhira. Standing and seated statues of the Buddha and the Bodhisattva Mai-
treya, and stone slabs depicting in low relief the legend of the Buddha's birth
stories, o r Jatakas, abound. Most of these statues and were carved out of
a locally available grey or blue slate stone called schist, but stucco was also used

55. Puri, 1965, p. 143; Wattcrs, 1904/05, pp. 273-6.

56. Puri, 1965, p. 145.
57. Coomaraswamy, 1926, yp. 165 ct scq.; Lohuizcn-dc Lccuw, 1949, pp. 170-1.
58. Rowland, 1970, pp. 121-5.
59. Roscnficld, 1967, p. 73.
60. Schlumbergcr, 1961, pp. 77-95.
 Kushan art

FIG.32. Miracle of Sriivasti. Lahore Museum.

 G, A. Pcrgachenkova, S. R. Dar, R. C. Sharma, M. A. Joyenda and H. Siddiyi

for making statues and reliefs (Fig. 32). Clay and terracotta were used relatively
sparingly. The change of medium (from schist to stucco) gave more freedom,
diversity and cheapness. Modelling in malleable material made it easier to Pre-
pare casts from moulds and made it more convenient and cheaper to increase
production. Whether o r not the use of stone and stucco for sculpturing finally
split u p in two different schools - the latter springing phoenix-like from the
ashes of the former but still separated from the other by a hiatus of a century
and a halfh' - does not concern us here.

FIGURE OF THE BUDDHA

The Gandhiira Buddha is an idealized figure having a delightful face unaffected

by age o r affliction. Standing barefoot o r seated cross-legged he is always shown
wearing an undergarment and a monk's robe. Among the signs of a maha-
p~rusalaksana(great man), the +ni!a, u ~ n dand dharmacakra are usually visible.
His k p o l l o Belvedere type of face, though just one among the numerous types
known, is n o doubt the earliest to ~ r o v i d ea model for others. The model of a
standing Buddha might have been copied from a Greek god or a hero or even
from a Roman emperor wearing ~ a l l i u mo r toga, as the Kushans definitely had
diplomatic and commercial relations with the contemporary Roman West. But
beyond this, borrowing ceases. The seated Buddha and the figure of the seated
o r standing Bodhisattva (Fig. 33), the latter a peculiar invention of Gandhiran
artists, have n o classical o r Indian ~ r e c e d e n t .Statues of Paiicika, the Comman-
der of the Lord's army, and his consort Hariti are also local figures pay excel-
lence. Furthermore, almost all Gandhiira statues were carved in the round.

CHARACTERISTIC FEATURES

The most characteristic feature of Gandhara sculptures is their frontality.

Figures normally stare fixedly into one's eyes o r are turned completely to right
o r left. There is seldom movement in their bodies. This can be understood in the
light of Kushan sculptures from Khalchayan. These are not strictly frontal but
turn slightly with restrained emotions. They are a step towards frontality and a
sharp contrast to the highly emotional images of the Hellenistic world and the
complete side views of Achaemenid sculpture. The fixed, unemotional frontality
of the Kushan art of Gandhira has obviously been brought from Khalchayan.
Attempts have been made t o identify portraits in Gandhara sculptures,
such as the two heads from Sahr-i Bahlol and Shpola ~ t u p a . But
~ ' these are the
heads of donors detached from their bodies. There may be statues of donors

61. Marshall, 1960, pp. 109-12.

62. Ingholt, 1957, p. 163, No. 423, p. 194, No. 563.
 Kushan arl

showing characteristic individual features and wearillg regional costumes but

they never represented specific individuals. N o doubt in Bactriab3and Matl~ura'~'
portraits of Kushan rulers and nobilit?. in characteristic Kushan dress are nume-
rous, but there is n o such presence of the Kushan n o b i l i t ~in Gandhira sculp-
ture, not even in the figures of the Buddha and Bodhisattva. In Buddhist art. an
individual - ecclesiastical o r telnporal monk, donor, king o r commoner - alnlays
remains anonvmous.

63. KowlanJ, 1970. 17. I 4 ( 3 , Fig. 86.

64. Roscnticld, 1967, pp. 138-53.
 c;. A. Pugn&cnkova, S. R. Dar, R. C. Sharma, M . A. J o y ~ n d aand H. Siddiqi

The Gandhira panels narrate the J ~ t a k a s ,o r birth stories of the Buddha, i n ,

simple, clear and lucid way which is in sharp contrast to the confused style of
earlier schools, seen at Bharhut and Sanchi. Although the number of events is
limited, the art of narration is simple and easy t o understand. Figures in a panel
stand out in correct relation t o one another with proper spacing between them,
Sometimes even perspective is emphasized. In this way a minimum number of
figures are needed to narrate a complete story. When more than one story is
required on a panel, each is separated from the other with the help of vertically
arranged columns, pilasters and recessed panels, o r horizontally arranged tor-
nices and mouldings. Normally, the actors of a story are arranged in single file
and move from right to left.

RELIGIOUS ARCHITECTURE

We are fortunate in having a better knowledge of the development of religious

architecture in Gandhiira. The Kushan contribution is substantial. Except for
Shrine 8 at Taxila, almost all Kushan monuments in Gandhira are Buddhist.
Shrine 8 is a square building within an enclosed wall measuring 37 x 37 m
relieved with buttresses o n each side and nine inter-communicating chambers.
O n e study has associated this shrine with the Kushan occupation of the area,
though its exact date and purpose cannot be a s ~ e r t a i n e d . ~ ~
Stupas and monasteries were the principal buildings of the period, as
Hsiian-tsang notices in his account of the Great Vihiira built by Kanishka at
Peshawar. It is said t o have been built as a thirteen-storey tower with a total
height of 213 m. Nothing of the monument survives except the famous
inscribed Kanishka casket, now in the Peshawar Museum, and a few sculp-
tures."" Among standing monuments, the remains at Takht-i Bahi, Jamal Garhi,
Tharelli and the sites of Jaulian, Mohra-Muradu, Kalawan, Pipala, etc. in the
Taxila valley are the best examples of the Kushan contribution to Gandhara
architecture.

G A N D H A R A STUPAS A N D M O N A S T E R I E S

The basis of the Gandhiira monastery is a court o r a series of courts open to the
sky and surrounded either by cells for monks o r by niches to place devotional
objects. Connected with the court are usually an assembly hall, refectories and
room for the chief priest on one side and the main stupa encircled by votive stu-
pas of varying sizes on the other. Gandhiira monastic establishments were built

65. Dar, 1980, pp. 91-106.

66. Spooner, 1908109, pp. 38-59; Dobbins, 1971, pp. 12-40.
 Kushan art

,xclusively in stone diaper masonry typical of the Kushan period. 'rhe entire
of the walls was covered with a thick coating of lime plaster. The roofs
were mainly vaulted and doors and windows were either trabeated or corbelled.
the first and second centuries A.D., this type and pattern of monastic archi-
tecture had emerged and became fixed all over northern India. Taxila call boast
of some of the earliest permanent m~nasteries.~' Although the development of
monasteries in Gandhsra antedates the arrival of the Kushans (cf. the Dharma-
rajika stupa), it was here that the pattern of monastic establishment was per-
fected and became popular. A large number of such establishments in Gand-
hira, their sizes, cultural content, elaborate facilities and architectural
embellishments, clearly point t o the economic prosperity of the period and the
patronage accorded t o the development of art.

ORNAMENTATION

The Gandhira stupa with its carved base, circular drum and spherical dome,
together with its sculptural embellishments, marks a development from the
primitive types known at Sanchi and Bharhut and at Dharmarijika (Taxila),
Shankaradara (Swat) and Manikyala. The Kushan contribution lies in the overall
sculptural ornamentation of the bases and drums of stupas. Apart from stone
reliefs fixed on the largest stupas, the smaller votive stupas were usually embel-
lished with stucco figures of the Buddha, Bodhisattvas and devotees set in
niches, and with figures of Atlantes, elephants, lions, caryatids, yaksis, etc.,
crouching under cornices and supporting the load of each receding terrace of the
stupa base. A variety of arches, pediments, debased Corinthian capitals, dentils,
merlons, rosettes and lotuses abound in both stone and stucco. The debased
example of an Ionic capital in stucco is known from the Pipala stupa at Taxila.
Sometimes figures of the Buddha and Bodhisattvas were set in the foliage of
Corinthian capitals carved in stone. Huge Corinthian capitals, set up on pilasters
or a round column, are known from Taxila and elsewhere. Most of the columns
and pilasters used as architectural ~ i e c e swere ~ r o b a b made
l ~ either of wood o r
of some other perishable material because except for a miniature fluted column
with a Corinthian capital, now in the Taxila Museum, and an Ionic column also
from Taxila, now in the Lahore Museum, no such column has survived.""

M I N O R ARTS AND COINAGE

There is not much evidence for the minor arts of the ~ e r i o dexcept for the
Kushan coinage. It is difficult t o allocate to periods the material from Scytho-

67. Dutt, 1962, pp. 24, 2 13.

68. Rowland. 1970, p. 145.
 G, A. ~ u ~ a c h e n k o v aS., R. Dar, R. C. Sharma, M . A. Joyenda and H. Siddiqi

Parthian and Kushan levels, ~ a r t l ybecause it comes from unscientific excava-

tions and partly because all these dynasties had the same geoethnic background
and were subject to the same sources of influence from the contemporary west.
The entire material represents the same mixture of classical, Iranian and local
forms and techniques that characterizes the art of G a n d h ~ r a . ~ '
The coins from Gandhara are the best evidence for the strangely syncretic
character of Kushan art, culture and religion. They show a portrait copied from
the bust of the Roman emperor Augustus, the first figure of Buddha and an
array of twenty-eight deities of Hellenistic Irano-Babylonian and Indian ori-
gins, all identified by legends in Bactrian Greek script.
Casting in bronze and copper was not as common in Gandhiira as sculpt-
ing in stone, stucco and terracotta. A few bronze statues are known from Sahr-i
Bahlol,70 Chinkolai (Swat) and other places.71These, however, appear to be later
than the Kushan period under review, but all the nine metal statuettes from the
Sirkap site of Taxila come from the two last strata and can conveniently be
dated to the first and second centuries A.D. With the exception of one thin
repoussk copper bust in a medallion, all the others are solid cast in open or
piece-moulds.72 Among these are figures of ~ u r e l yGraeco-Roman origin such
as Harpocrates, Cupid, Psyche and Aphrodite as well as Hindu and Buddhist
figures. Metal sculptures from other laces are mainly Buddhist.

TERRACOTTA FIGURINES

It is interesting to note that not a single terracotta figurine has been reported
from the limited excavations of the Kushan city of Sirsukh in Taxila. However,
the art of making terracotta and clay figures continued at Taxila, as elsewhere,
until the fifth century A.D. Despite some borrowings from Hellenistic motifs,
this terracotta art, 'in its own way, was just as original, forceful and independent
as the Gupta Art of Hindustan and more so than the contemporary Byzantine
Art in the West'.73Slip casting that is, the use of crushed stone and clay to make
terracotta figures, was introduced for the first time at Taxila and became pop-
ular throughout the Kushan period.74
The Kushan craftsmen inherited the art of making figurines in terracotta
o r clay from their predecessors, the Mauryas, Indo-Greeks, Scythians and Par-
t h i a n ~ . 'Their
~ figurines included human, animal and other toy models. The

69. Dar, 1977, pp. 61-89.

70. Rowland, 1970, pp. 185-6.
71. Hallade, 1968, p. 86, Plate 62, p. 168, Plate 128.
72. Marshall, 1951, Vol. 11, pp. 604-6.
73. Ibid., p.d442.
74. Mian, 1974, p. 206.
75. Marslia!l, 1951, Vol. 11, pp. 435-72; Wheeler, 1962; Dani, 1965/66, pp. 46-109.
 Kushan art

figurineswere either hand-made o r wheel-turned to make a hollow body that

(yas later moulded by hand t o a human or animal shape. Figures were also cast
in single as well as double moulds. Sometimes, faces were moulded and fixed to
hand-made bodies.76 Almost all these types began with the Indo-Greeks and
continued through the Parthian period up to the end of the Kushans." Accord-
ing to Dani, the real Gandhira terracotta human figurines, with well-formed
heads and beautiful faces, were developed during the Middle Kushan period.7n
Among animal figurines, bodies were either solid or hollow. Bulls, horses, ele-
phants, camels, monkeys, dogs, rams and goats were the most popular, but we
find tigers, snakes, crocodiles, pigs, birds, bird chariots and toy carts.

JEWELLERY

The inventory of specimens of Gandhara jewellery and ornaments that can defi-
nitely be dated t o the Kushan period is not very long. However, this scarcity of
material is amply compensated for by the profusion of jewellery represented in
Gandhara sculpture. For example, the figures of Bodhisattva, Hiriti and other
females are shown wearing gorgeous jewellery items that are not very different
from the specimens of the same or slightly earlier period. A collection of 180
items from the last stratum of S i r k a ~ , thirteen
'~ from Tor Dheri, three from the
Rawalpindi area and seventy-two from Palatu Dheri" and a few other gold
ornaments reportedly from the Taxila region, now preserved in the Victoria and
Albert Museum, L ~ n d o n , and
~ ' the Cleveland Museum of Arts" etc., is all that
we have of Kushan jewellery from Gandhara. T o this can perhaps be added the
famous bejewelled gold casket and thirty ornaments from Bimaran of slightly
earlier date.R3
The Gandhiira jewellery displays a variety of styles and techniques. The
representation of jewellery on statues throws light on the Kushan fashion of
bedecking individuals with ornaments. Men wore jewellery as much as women,
whereas before and after the Kushans, the wearing of ~ e r s o n a ornaments
l was
the prerogative of ladies alone. The richness of ornaments depended on the sta-
tus of the person wearing them. Bodhisattvas, kings and queens, men and
women of noble birth and even deities are always shown wearing heavy jewel-
lery. Commoners either wore samples of ornaments or none at all.

76. Mian, 1974, p. 206.

77. Dani, 1965/66, y. 47.
78. Ibid., pp. 43, 65.
79. Marshall, 1951, 11, pp. 616-37.
80. Marshall, 1902/03, yp. 172, 185-99.
81. Hallade, 1968, Plates IX and XI.
82. Anon., 1953, p. 200.
83. Wilson, 1841, p. 71.
The Taxila collection of jewellery is predominantly Graeco-Roman in chant-
ter," but Gandhira jewellery generally shows the same range of classical,
nian, Sarn~atianand Indian forms and techniques that we see in the art of Gand-
hira. These reflect current trends and taste, and show how the Gandllira types
developed out of earlier ones. The commonest types of Gandhira jewellery
include earrings of leech-and-pendant type, necklaces, neckbands, bangles,
bracelets and armlets, anklets, girdles, hairpins, amulets of great variety, etc,,
and are usually worn by Bodhisattvas. There is also a wide variety of finger-
rings with o r without encrusted gems and seal impressions. Decorative items,
such as broches, turban pins, and miniature gold figures of Cupid, Psyche, ani-
mals, birds, flowers and necklaces with fanciful designs are known from Taxila)
but waist-cords, bejewelled breast-chains and footwear are known only froln
statues. Quintus Curtius (History of Alexander 9.1.5) provides an interesting
reference to the golden staff set with beryls and jewelled golden sandals
Sopeithes, the King of the Salt Range and a contemporary of Alexander the
Great.
Gandhiira jewellery shows a fully developed stage of the crafts of gran-
ulation and filigree, which the Gandhara goldsmiths borrowed from Western
Asia, and the art of incrustation of gems, which the Orient lent to the West.
Technically, G a n d h ~ r aornaments were made with dies and by hammering,
casting, moulding (lost-wax method) and repoussk. The forty-two pieces of sil-
verware," all belonging t o the Late Parthian and Early Kushan periods, clearly
show that the crafts of the silversmith and coppersmith were equally developed.
Whereas much of the Taxila ware reflects Graeco-Roman culture in vessels such
as askoi, phialai, mesomphaloi, aryballoi, kantharoi, paterae and goblets, copper
and bronze vessels show a mixture of Western and Central Asian elements and
predominantly local influences. Also, almost all types are represented in pottery
forms, suggesting a uniform Gandhiira culture during the first and second cen-
turies A.D.
Much other material, such as gems, seals carved with a variety of designs,
bone, ivory, shell and glass objects and a wide range of beads, has been ex-
cavated from numerous cities and stupa sites. The best examples of ivory are
from TaxilaY6and Begram," all clearly showing how deeply this art was related
to local craftsmanship.
Precious and semi-precious stones werc used for making seals and jewel-
l e r ~ agate,
, amethyst, carnelian, chalcedony, crystal, garnet, jasper, lapis lazuli)

84. Marshall, 1951, 11, p. 616; Sharif, 1978, pp. 46-7.

85. Marshall, 1951, Vol. 111, pp. 564-606.
86. Ibid., Vol. 111, pp. 199 et scq.
87. Rowland, 1971.
 Kushan art

onyx and turquoise. Glass, marble, mother-of-pearl, shell, steatite, ivory and
bone were used f o r beads, which were made in a variety of shapes - domical,
and tubular - in animal and bird forms. They were from
one or both sides and ~ o l i s h e d .

A SUMMARY

In brief, the Kushan art of Gandhira was a living art open to influence from
within and without. It absorbed the earlier Graeco-Bactrian traditions current
in the area and was also receptive t o ideas and trends of the contemporary West
through international trade and commerce. Gandhira acted as the hub from
which Kushan art spread in many directions to places such as Mathura, Dev-
namori and Amaravati in India and towns in Afghanistan and Central Asia. As a
result of recent discoveries at Khalchayan, Ay Khanum and Surkh Kotal, there
is now no doubt that some form of Hellenistic school lay behind Gandhira art,
but it is wrong t o call it an example of Western art. It has its own individuality,
reflecting the socio-religious aspirations of its people. Economic prosperity and
peace remained the basis of its popularity and development, and when that was
no longer the case the art could not sustain itself: it languished and then dis-
appeared.

Kushan art in Mathura

AGE A N D T E C H N I Q U E

The art of Mathura both precedes and post-dates Kushan art over a total span of
about 1,000 years, but the Mathura workshops were most active and productive
during the rule of the Kushan emperors, especially Kanishka, Huvishka and
Visudeva (second and third centuries A.D.), which represent the golden age of
Mathura sculpture." The earliest dated specimen of Kushan a n at Mathura, the
statue of Bodhisattva now in the Sarnath Museum," was made in the third year
of Kanishka."" Mathura sculptures were carved from the spotted red sandstone
quarried at Sikri, near the city, and its craftsmen mastered the technique of carv-
ing stone that was liable to be marred by streaks or spots. Some scholars beliew
that they originallv covered the whole carved surface with a layer of pol!-
chrome or

88. Agrawala, 1965, p. 2.

89. Vogcl, 1930, p. 107, Plate XXVIIIa.
90. Chanda, 1936, pp. 1 1, 12, 16.
91. Rowland, 1970, p. 149.
 G A N D H A R A N I N F L U E N C E ANL) R O Y A L I'ORTRAI'TS

Mathura artists, by their central geographical position, were open to influences

from both Gandhara and Amaravati, and they sent their images to a wide range
of sacred sites. Typical Mathura sculptures of Kushan date have been discovered
at Varanasi, Gaya, S r a ~ a s t i , Taxilaw
"~ and Puskal~vati.''~
At first sight the style of Mathura seems to be a sequel to that of the stu-
pas at Bharhut and Sanchi, but it is related to two other traditions - the art of
Amaravati and the Graeco-Buddhist art of Gandhlra.')' The gallery of royal par-
) ~ given us portrait figures of
traits from the Devakula at Mat, near M a t h ~ r a , ' has
Vima Takruma, Kanishka and Cas~ana"set u p in the reign of Huvishka. They
are neither in the Gandhara nor in the Mathura style, but possess an autono-
mous stylistic character of their own. Vima's seated figure wears a short tunic
and heavy felt boots of Central Asian origin.98The headless statue of Kanishka
shows the Great King standing, wearing a Central Asian stiff mantle (caftan)
and heavy felt boots, his right hand resting o n a mace and his left holding a
broad sword.19 The posture is rigid, stiff, frontal and hieratic but conveys the
valour and kingly virtue of the dynasty (see Fig. 2, Chapter 11).
The Devakula portraits, almost in relief with no suggestion of any three-
dimensional form, show clear signs of Hellenistic and Parthian influence.1~3
They are the sole examples of portrait sculpture known from ancient India.
Apart from these portraits, the most striking examples of the Kushan art
of Mathura are the figures of yakras and yakris, nagas and naganis and female
(Silabhafijika) figures (Figs. 34 and 35), some wanton and sensual. The sculptu-
ral art of Mathura has many distinguishing features:
The material used is either red sandstone with creamish spots or buff sandstone,
which sometimes contains dull red patches. In certain sculptures efforts to
remove the adverse effect of spots o r ~ a t c h e sthrough the application of a
coat of colour can be seen.
Brahmanism, Buddhism and Jainism flourished simultaneously at Mathura
under the Kushans and icons and shrines of all the three sects were made
in large numbers. While Brahmanism continued to be the religion of the
masses, Buddhism received royal patronage, while Jainism had the follow-
ing of the rich merchant community.

92. Vogel, 1910, p. 28.

93. Khan, 1966.
94. Dani, 1965166, Plates XVIII, 4-5.
95. Agrawala, 1965, p. 3.
96. Vogel, 191 1112, pp. 120-7.
97. Rosenfield, 1967, pp. 135-53.
98. Vogel, 1930, p. 91, Plate V; Rosenfield, 1967, pp. 144-51
99. Vogel, 1930, Plate V; Rosenfield, 1967, pp. 144 ct seq.
100. Rowland, 1970, p. 149.
FIG.34. Silabhalijiki figure on the oh\,c~-scsidc of a p;llal.. S e c o ~ i dcentur! :\.I). ( F r o m
H. Zimnier, The A7.t of India71 Ajia, Val. 11. plate 743, Ncnr I'ork, 1958.)

In this period symbols representing the Buddha in earlier times were rep1'1ced
by the anthropomorphic representation.
The religious aspects of art had not obliterated its secular spirit. Skilled workers
and artisans w o r k e d for patrons of different creeds according to d e ~ n ~ l n d .
T h e spirit of secularism is seen in thc depiction of decorative motifs, social
and folk scenes c o m m o n t o all sects.
The assimilation o f different artistic forms and tl~eil-fusion into a distincti1.e
style \vet-e the important features of the Mat11u1-,1 school. T h e natural
reflection of contemporar\- social, religious and political mo\.emcnts h ~ s
t o be e\,aluatcd in proper p c ~ - s ~ c c t i vhlathura
e. '1st nctuallv scricd '1s 'I
 G. A. Pugathrnkova, S. R. Dar, R. C. Sharma, M. A. Joyenda and H. Siddiqi

FIG.35. Silabhaiijikn figure on a bracket. (From J. P. Vogel,

La Sc~lpturedc M a t h w ~ a platc
, XII, Paris/Brussels, 1930.)

bridge to correlate and unite indigenous and alien elements and success-
fully accomplished this through its adaptability, and its process of fusion,
amalgamation and interaction.
The preselltation of female beauty as a vehicle of art was a novel experiment oi
the Mathura school. In the earlier monuments of Bharhut and Sanchi
womenfolk seem unconcerned with this. Their function is either to W O ~ -
ship if represented in mundane form o r to receive worship if elevated to
the superhuman status of devati o r yakgi. But in the Kushan period the
sculptor at Mathura viewed feminine beauty from a different angle.
Arrested by a beautiful face, long hair, heavy hips, voluptuous breasts,
 Kushan art

movements, attractive looks and inviting gestures, he transformed

his subject into sculptural creations that blended the fervour of sensuous
emotions in a religious and spiritual environment.
The names of artists of the Mathura school have been recorded on the pedestals
of sculptures - R i m , Dharma, Sanghadeva, Joiisa, Disa, Sivara(k!ita),
Singha, Niyasa, Deyahu, Vishnu and Jayakula.1~1

VISHNU

Early Brahmanism o r Hinduism at Mathura was based on the Vishnu or Visu-

deva cult and Mathura artists made icons of Vishnu and his associates at an early
stage. Vishnu figures have been found in different forms. The two-armed repre-
sentation seems uncommon (Mathura Museum No. 1150). The four-armed
images hold a mace (gadi), disc (cakra), waterpot (kamandalu), and the fourth
hand either remains in protection (abhaya) or in boon-bestowing (varada) pose
(Mathura Museum Nos. 15.912, 15.948, 28.1729 and 34.2520, second deity).
Sometimes the mace is replaced by the conch (iankhu) (No. 15.4267). The lotus
(padma) does not appear in this period. The eight-armed figures of Vishnu have
hands that are broken so the attributes remain obscure (Nos. 15.1010 and
50.3550, and Lucknow Museum No. 49.247). In one sculpture the deity is seen
mounted on his vehicle Garuda in bird form (No. 39.2858).
The concept of the incarnation of Vishnu was still in its infancy. The
Lucknow Museum stela (No. J.610) probably shows the giant (trivikrama or
virZ!) form. The boar incarnation (variha) has been identified in another stela
(Mathura Museum No. 65.15) (Fig. 36). The deity is lifting the earth, which is
seen personified on his left s h ~ u l d e r . ' A
~ ' fragmentary sculpture in the Mathura
Museum (No. 17.1344) has been interpreted as Visudeva crossing the River
Jamuna with a basket over his head.lC3Another stela (No. 19.1563) is ~ r o b a b l v
to be identified as Rams and Siti.
The cosn~ic(caturvyiha) form of Vishnu is seen in an image of the Math-
ura Museum (No. 14.392-5) (Fig. 37). O u t of the central deity, Visudeva o r
Krishna, emerge other figures from his shoulders and head. Balarima can be
recognized from the snake canopy. The high crown and v a n a m i l i (garland
made of forest leaves and flowers) of Vishnu are noteworthy.

BALARAMA

The cult of Balariilna, the elder brother of Krishna, was already established at
Mathura before the Kushans (Lucknow Museum N o . G.215). In the Kushan

101. Sharma, 1984, p. 139.

102. Joshi, 1972, p. 7.
103. Ibid., p. 16.
 G, A. ~ u ~ a d ~ e n k oS.v aR., Dar, R. C. Sharma, M . A. Joyenda and H. Siddiqi

FIG.36. Boar incarnation. Buff sandstone. Second century A.I). (length 35 cm).
Mathura Museum N o . 65.15.

period, figures of Balarama have two o r four arms holding a cup in the left hand
with the right hand raised up in the protection pose (abhaya mudrd). Conceived
as the incarnation of the cosmic serpent Sesa, Balariima is shown with a snake
canopy (Mathura Museum N o . 14.406) (Fig. 38). Sometimes he carries a lion-
staff plough ( s i ~ h a l a n g a l abala). Rarely, between Balariima and Krishna, stands
a female deity, identified as EkinaliSa, sister of the two brothers (Mathura
Museum N o . 67.529).
Numerous Siva finds suggest that Mathura was also a seat of the Siva cult.
In the Kushan period Siva is represented in t w o forms: liiga, the phallic form
with the nut portion projecting from the shaft and fastened with a flat band
(Mathura Museum N o . 83.3) (Fig. 39), and purura, the anthropomorphic form.
Quite often a combined aspect is seen and in this case the liizga is shown with
one, two, four o r five faces. The beads known as t a t p u r u ~ a a, g h o ~ avamadeva,
,
sadyojata, iGna face east, south, west, north and upper direction respectively.
The Ardhaniiriivara form (a composite figure of Siva and his spouse Parvati) is
also from the Early Kushan period. In this form thc right half is generally repre-
sented as male with matted hair, a half vertical third eye and the organ in an
upward position (iirdhvaretas). The lcft female half is shown as graceful with
earrings and anklets (Mathura Museurn N o . 34.2520) (Fig. 40, first deity).
 Kushan art

FIG. 37. Bust of Visude~ain cosnllc (n)lGha) form. Buff sandcrone. Second ccntur!r A.I,.
(height 45 cm). From Saptasamudri ~vell.Mathura Museum No. 14.392-5.

Skanda o r Kirttikeya was also a favourite deity in Kushan Mathura. The texts
describe him as son of Siva (Matsyapul-ana, Chapter 158, and Kumarasambhava
of Kalidasa 10.57-60) and also of Agni (Mahabhamta, Vanaparva, Chapters
225-30 and Skandapurdna, Maheivarakhanda, Chapters 27-31). H e is known
as a god of war and commander of the divine army (dcvasenanl). In the Kushan
period he is shown as a two-armed young man, wearing a crown o r turban,
llolding a long spear in his left hand with his right hand in the protection pose
(Lucknow Museum No. 57.458). An image in the Mathura hSuseu111 ( N o .
42.2949) is dated Year I I . H e is sometimes shonrn with a cock o r J pe'lcock
(Mathura hluseum No. 33.2332). A bronze plaque from Sonkh has been identi-
fied as Kirttikcj.a.l-4 His nativit\. is sometimes shown nrith one o r more mother
 u aK.
G, /?. ~ u ~ , ~ c / , c n k o , S. , Dizr, R. C. S h a r m a , M . A. Jo)lendrr a n d H. Sjddjrli

FIG;. 38. Head of Balarama w i t h snakc canopy. Buff sandstone. Second century A.I).
(height 35.5 c n ~ ) F. r o m Kachahrighat well. Mathura Museum No. 14.406.

goddesses (matrkas), holding a child and with a jar (Lucknow Museum No.
0.250) (Fig. 41). GaneSa, the younger brother of Karttikeya, appears late in art; a
post-Kushan statuette represents him as an elephant-headed nude dwarf, wear-
ing a snake thread (vyala yajnopavita) and eating sweet balls (laddu) with his
trunk (Mathura Museum N o . 15.758).

SURYA

T h e sun god (Surya) in the Kushan period is shown squatting in a car drawn by
t w o horses, wearing an embroidered coat, trousers and turban, and holding a
stalked lotus bud in his right and a dagger in his left hand. T h e whole appea-
rance suggests an alien treatment as marked as o n the contemporary royal por-
traits (Mathura Museum N o . 12.269) (Fig. 42). These tight features are sub-
sequently relaxed, the number of horses increases t o four ( N o . D.46). and a
nimbus resembling the disc of the sun is added. T h e wings o n his shoulders
recall the early Vedic concept of the sun bird Garuda.I2'
 FI<,.39. S i v a l i ~ ? ~Na .e a r E;.lnk,~lr. Spotrcd I-cd s,~ndsronc.Iqil-st centul-! &\.I).

h l n t h u r , ~hluseuni No. S3.3.

The dominating yakZa cult o f an earlier period at Matliura was oversliadon~ed

by other popular deities under the Knslians. T h e !-aksas, n o w grotesque and
dwarfish with pot belly (No. C.3), served as a t t c n d ~ n t s .Their inulidanc nature
is depicted as excessive drinking ( N o . C.Z), a scene sometimes explained ,IS B.Ic-
chanalian, suggestillg a Grcck impact tlirougli Gandliira art."',

Niga (snake) \\;o~-sliip w . 1 ~p r c ~ a l e n tat Mathura. W e find initeycndcnt irn,~gcao i

the niga dcitics i n liuliian form hut surmounted by snalrc lioods. T h e site of
Sorikli h.1~rC\.~'lIc'JI . ~ I I ~ , Iof
~ ~.l~ lijga
S shrine ot Kushari date.

106. Smith, I SSc), T'JI-t I . pp. 140, 156.

 G, A. Pugachenkova, S. R. Dnr, R. C. Sharma, M. A. Joyenda and H. Siddiqi

FIG.40. Stela representing, from the left, Ardhaniriivara,

Vishnu, Gajalaksmi and Kubera. Spotted red sandstone (length 26.5 cm).
Mathura Museum No. 34.2520.

O n e lintel depicts the niigas and niigis with a snake scalp, while the devo-
tees o r visitors bear the turban (No. SOIV-36) (Fig. 43). A duel between the
nZga and Garuda (bird deity) has also been represented (No. 41.2915).

O T H E R FIGURES

A large number of female deities o r mother goddesses have been recovered from
Mathura. Important ones are Gajalaksmi (No. 34.2520) (Fig. 40 above, third
deity), Lakpmi, Hlriti, Yaksis, Mitrkis, Sasthi and nagis. A popular goddess of
the age was Durga (No. 33.2317). A few sculptures from Mathura suggest that
efforts were made t o avoid disharmony between different sects. An interesting
stela in the Mathura Museum (No. 34.2520) represents four deities together:
Ardhaniriivara, Vishnu, Gajalaksmi and Kubera (the lord of the yaksas).

J A l N A FIGURES

The excavations conducted by A. Fiihrer at the Kankali Mound, Mathura,

yielded hundreds of Jaina antiquities, most of which are housed in the State
 Kushan art

FIG.41. Stcla showing nativity of Kiirttikc!.a. Buff sandstonc. Second centul-\. A . I ) .

(lcrigtli 24 cm). Lucknow Musculn N o . 0.250.

Museum in luck no^.'^' These range from the second century 1 3 . ~ ~ t. o the
twelfth century A.D., but the majority belong to the Kushan period. The +a-
gapalas serve as the base for the dcvelop~nentof the Jaina pantheon; some of
them belong t o pre-Kushan times and one records the name of the
mahZkZatrapa Sodasa (Lucknow Museum N o . J.1).
The iiyiigapatas according t o the central theme ma!. be classified differ-
ently, that is, Cakrapata (representing the wheel, J.255), Svastikapata (repre-
senting auspicious cross J.252), Caityapata (showing the stupa o r c a i t ~ a5.255;
,
see Fig. 44), with a beautiful railing and gateway flanked by two female dancers
and an inscription recording that the stone tablet was set u p by the wife of the
dancer Phalguyasa f o r the worship of Arhata, and the nr/~atpa!a o r tirthditka-
rapafa when the Jina (main deit!.; see Fig. 45) is shonrn, replacing the s!-mbol
0.252). Some ot these tablets show a variety of beautiful motifs depicted as
eight auspicious s ~ - n ~ b o(astm2angalac~,Ihas).
ls

107. Smith, IC)O1.

 a , R. Dar, R. C. Sharma, M . A Joyenda and H Siddiqi
G. A. P ~ ~ a c h e n k o v S.

FIG.42. Surya clad in northern style. Buff sandstone. From Kankali Mound.
Late first century A.D. (height 48 cm). Mathura Muscum N o . 12.269.

FIG.43. Linrcl showing the scene of ngga family, Red san$o.tme. From Sonkh. A.n. 100
(length 121 cm). Mathura Museum No. SOIV-36,
 Kmhun art

FIG.44. Inscribed ayzgapata carved with stupa railing. Buff and red sandstone,
From Kankali M o u ~ i d(50 x 70 cm). Lucknsw Museum No, J.255.

Jina images of the Kushan period are generally broad-chested, stiff, with
shaven head o r little hair. T h e mark of irivatsa o n the chest of a Jina is an essen-
tial feature at Mathura. Palms and soles are usually marked with a triratna o r
cakra as the mark of a great man (mahipztrusalaksana). In the early period the
halo is plain, devoid of any concentric band but carved with a scalloped border
(hastinakha). With the passage of time the decoration o n the field of the halo
increases (Fig. 46).
It is not possible t o identify all the Jinas of the Kushan ~ e r i o d because
, we
do not know what developed subsequently. Rsabhaniitha can, however, be iden-
tified by the fall of hair o n the shoulders, and PiirSva o r SupiirSva b?. the snake
canopy. T h e depiction of life events of Jinas is rare, but a K u s h a n - ~ e r i o dstela
has been explained as illustrating the episode of the transfer of the enibr!m of
the last Jinn M A ~ ~ V T I - ~ tile
from I w o m b of Brihmaqi Devananandii t o that of
Ksatriyiil?~TriSali (No. 5.626). Neminitha, the twenty-second Jina, w h o is
described as cousin of K~-is]illa,w s represented flanked b!~ Balariini~and
Krishna.
 G A. Pugachenkova, S. R. Dar, R. C Sharma, M . A. Joyenda and H Siddiqi
-

FIG.45. Head of Erthankara. Second century A.L>. (From H. Zimmer,

The Art of I n d i a n Asia, Vo1. 11, platc 73, New York, 1968.)

BUDDHIST SCULPTURES

Kushan art of Mathura earned its real fame in producing hundreds of excellent
Buddhist images, which were both installed in Mathura region and exported.
Before the Buddha was represented in human form, his presence was conveyed
through different symbols: the elephant for his Birth; the horse for the Great
Renunciation; the tree for his Enlightenment; the wheel for turning the Wheel
of Law; and the stupa for his Death. The credit of introducing the Buddha
figure is sometimes given to Gandhira and sometimes to Mathura, but there is
some evidence that the evolution of the Buddha figure at Mathura came slightly
before the commencement of Kushan rule.lO"

108. Lohuizen-de Leeuw, 1949, p. 171.

 Kushan art

Pre-Kanishka Buddha figures are characterized by the snail shell (kaparda)

on the head, inconspicuous drapery marked by an incised line on the left shoul-
der, a deep navel, and a pedestal with three tiers or two lions supporting the seat.
The Buddha's corpulent body has a crude, primitive and archaic look. In
due course a set formula and an ideal form of the Buddha was developed. H e is
shown in high relief, with a nimbus bearing a scalloped border, a back slab
showing foliage of the bodhi tree, two celestials hovering in the sky and aco-
lytes flanking the deity. The top of his hair is shaped like a snail shell, the rest of
his head being smooth. H e has small earlobes, a circular (imi)mark on his
forehead, almond-shaped, wide-open eyes and a slightly smiling expres~ion."~ A
good example is the Buddha from the Katra Mound (Fig. 47).

109. Sharma, 1984, pp. 176-7.

Frc;. 47. B u d d h a inscribed as Bodhisattv,z. Spotted red sandstone. I,,~tefirst century /~JJ,

F r o m Katra M o u n d (height 71 cm). M a t I l u l . ~Museum No. A.1.

 Kgshan art

FIG. 48. Buddha with drapery covering both shoulders. Spotted sandstonc.
Second century A.D. (height 81 crn). Mathura Museum No. A.4.

During the reign of Huvishka the Mathura school introduced a notable

change in the Buddha's drapery, which now covers both shoulders (ubhayin-
sika ~akghatr)(Mathura Museum No. A.4) (Fig. 48), and then becomes thick
and stiff with broad pleats (Mathura Museum No. 76.17) (Fig. 49).'1°The dis-
tinction between the Buddha and the Bodhisattva is now made clear - the
Buddha being shown as a monk while the Bodhisattva is adorned with a crown
and the ornaments lending him a ~ r i n c e l ylook (Lucknow Museum No. B.26)
(Fig. 50). Beside the protection ( a b h a ~ a )pose, others are now introduced:
meditation (dhjj~na);earth touching (bhimirparia); reaching o r teaching pa-
deia o r u ~ ~ k h ~ ~ aand
n n )turning
; the wheel of law (dharmacakapravartama).
The Buddhist pantheon grew manifold.
 I:I(;. 49. Buddli,~~ v i t l i<;,lndlii~-,linfluence o n cloth. St>ottcd red ~ ~ ~ ~ i d s t o n c .
Second centur! : \ . I ) . I:rom ( ; o \ - i n d n , ~ ~(height
. ~ ~ - I IS cni).
Ail,~thul-,~hluscum No. 76.17.

The sculpturn1 ; ~ r tof M n t h u r n in t h e K u s h a n pcrioci W,IS cxtl-c~iiclycseati~c

(Figs. 51-56), a n d its p r o d u c t s wcl-c in licavv deliland t l i ~ - o u ~ l i ancient o ~ ~ r 110l.tll-
e r n I n d i a . M n t h u r a b o r r o w e d several f c a t u r c s fro111 (;;lnctlii~-,~,1,ur contt-ibut~d
m u c h i l l rct~1r11. T h e J n l L ~ knLa ~r r a t i o n , t h e l o t u s sc,lt, tlic d l - , ~ ~ c co\rc~-in$
~-j, 011~
~ - t h e B u d d h a and t h e lion t l i ~ x , n citlcl-i. sollic of t h e i r n p o ~ - t ~h~l ~~tll tl -
s l i i ) ~ i l d c of
urLl f e a t u r e s later ,ldoptcd it1 (;andliir,~. 'I'lic s e c o ~ i dp l i , ~ s eot d c \ ~ e l o p ~ n c n ~
 K ~ s h a nart

FIG. 50. Crowned Bodhisattva head. Spotted red sandstone. Second century A.D.
From Kankali Mound (height 50 cm). Lucknow Museum No. B.26.

Amaravati owes much to Mathura, notably the change from the aniconic to the
anthromorphic representation of the Buddha; the impact of Gandhira at Ama-
ravati came through the Mathura school;"' and the Sarnath school, which flour-
ished in the Gupta period, developed from the Late Kushan art of Mathura."'

111. Coomaraswamy, 1965, pp. 70, 71.

112. Saraswati, 1975, p. 135; Sharma, 1984, p. 241.

389
 S. R. Dar, R. C. Sharma, M. A. Joyenda and H. Siddiqi
G. A. P~~ac-/,enkova,

FIG.53. Railpost showing a woman fastening her garment.

Red sandstone. First-second ccntury A . I , . F r o m Bhuteswar (height 142 cm).
Mathura Museum No. J.4.
 Kwshan art

FIG.54. Bracket from a gate of the njga shrine with a woman under a n-cc.
Spotted red sandstone. Frmn Sonkll (height 77.4 cm).
Mathura Museum No. SOIV-27.
FIG,55- Railpost showing Rsyairinga in ecstay. Spottcxl red sandstme'
First-seeofid century A . n . From Chaubatra Mound (Iteight 89 emf.
Mathgra Museum No. j.7,
 Kushan art

I-.\<;. 5(1. Pal-asol car\.ed \\.it12 lotus .lnd c~uspiciousniotifs.

Spottcci red sandstone. FI-oln Maboli-ki-paur ( I I I x 105 r m )
Mathul-a hluscum No. 76.12.
 J. Harmatta

Old Persian, Imperial Aramaic, Old Bactrian

Script and writing appeared in eastern Iran long before the Yiieh-chih conquest
of Bactria. Under Darius I, O l d Persian administration and chancellery practice
had probably been introduced into the eastern Achaemenid satrapies. This
involved the use of the O l d Persian language and cuneiform script, and the
adoption of the Aramaic language and script as intermediary instruments of
communication between administrative centres. The royal weight inscribed
with an Old Persian cuneiform text from Bost (modern Qal'i-i Bist in Afghan-
istan) shows this development, even though it was prepared at the royal court in
western Iran; and the borrowing by the Prakrit languages of such important
terms as Old Persian dipi- (document), nipis- (to write) and nipistam (inscrip-
tion) clearly proves the use of O l d Persian in the Indus territories belonging to
the Achaemenid Empire at that time. The introduction and use of Aramaic
chancellery practice was, however, of much greater importance for the spread of
literacy in the Middle East. Although not a single Aramaic document from the
Achaemenid period has so far come to light in eastern Iran, indirect evidence
exists. Kharopihi script came into being under the influence of the Aramaic
alphabet. Consequently, the creation of Kharopihi indirectly attests t o the use of
Imperial Aramaic in the royal chancelleries of the eastern satrapies of the
Achaemenid Empire on the borders of India.
The use of Aramaic as an intermediary language did not come to an end
when the Old Persian Empire fell. As in other satrapies, the administration and
the Aramaic chai~celleriesstill continued t o function under the Hellenistic rulers
who succeeded them, and Greek could not immediately replace Aramaic as the
chancellery language in Bactria and Gandhira. There is abundant evidence for

:'. See Map 3.

 J. Harmatta

the survival of Aramaic in these territories in the rock and pillar inscriptions
up by ASoka, the Mauryan king. Six have been discovered so far: (a) the
edict of Taxila; (b) the stone inscription of Pul-i Darunta; (c) the rock edict of
Kandahar (Kandahar I); (d) the second inscription from Kandahar (Kandahar
11); (e) the first rock inscription (milestone) f r o m Laghman (Laghman I); and (f)
the second rock inscription (milestone) from Laghman (Laghman 11). As a Sam-
ple of these texts, the rock edict of Kandahar (Kandahar I) (Fig. 1) may be
quoted here (Iranian terms are italicized):

1. Snn X ptyty 'byd zy rnr'n prydrS rnlk' qSy~'rnhqSr

For ten years penitence was made by O u r Lord, Priyardari, the king, enforcing the truth.

2. rnn 'dyn z'yr rnr" lklhrn 'nSn wklhrn 'dbiy' hwbd

Since that time evil decreased for all men and he made disappear the quarrelsome.

3. wbkl 'rq' r'rn ity w'p zy znh brn'kl' Irnr'n rnlk' z'yr
A n d happiness arose o n the whole earth. A n d besides, this [is] concerning the food: for Our Lord,
the King, little

4. q ~ l nznh lrnhzh klhrn 'nSn 'thhsynn wzy zwny' 'hdn

is slaughtered. Sccing this all men have ceased [to d o it]. And those men w h o were catching living
beings,

5. 'Ik 'nin ptyzbt knrn zy prbst hwyn 'Ik 'thhsynn rnn

have been forbidden [to d o it]. Thus, w h o were bound [by their passions], those ceased to

6. prbsty whwptysty I'rnwhy wl'bwhy wlmzyity' 'nSn

be bound. A n d good obedience [is observed] t o his mother and to his father and to the elder men

7. 'yk 'srhy hlqwt' wl' 'yty dyn' lklhrn 'nSy' hsyn

as destiny imposed upon him. And lcgal ~ r o c e e d i ndoes
~ not exist against anyone who is pious
[literally: all pious men].

8. znh hwtyr lklhrn 'nSn w'wsp yhwtr

This benefited all men and will zn all benefit [them].

Without doubt, the language of this text is Aramaic, but it contains a number of
Iranian terms and some errors from the strict viewpoint of Aramaic linguistic
usage. This fact has led philological research to assume that what we are dealing
with here is an Iranian text written by Aramaic heterography.
As, however, the inscription contains inflected Aramaic nominal and ver-
bal forms, there can be little doubt that the author of the text still intended to
write Aramaic. Accordingly, the linguistic features, unusual from the viewpoint
of Aramaic, are to be explained by the character of Aramaic as an intermediary
language, permanently exposed to the interferences of the mediated languages.
 Langragrs and ayipts in Gracco-Bactria and rhc ,Taka klnyd(,mr

FIG.1. The bilingual inscription of ASoka from Kandahar (Kandahar I).

Third century B.C.

In the Aramaic of ASoka's rock and pillar edicts we must reckon with the
interferences of three languages - Old Persian, Old Bactrian and Prakrit. For
Old Persian influence on Aramaic, we have abundant evidence in the Ariama
letters and the Aramaic docu~nentsof Elephantine. In the Aramaic inscriptions
of ASoka O l d Persian interference is limited t o some important administrative
terms: hwnitwn - O l d Persian hu-niitdvan- 'good document', Biblical Aramaic
nkwn, Imperial Aramaic nittun' - and krpty = Old Persian kdra-paei- 'army
road' (from O l d Persia11 kara- 'army', unknown in Avestan, and Old Persian
Poi- 'road' as against Avestan paen-, pantay-, pa& 'road').
The intcrferencc with Aramaic of 'Old Bactrian', that is, the language of
ancient Bactria, was obviously very i n ~ ~ o r t a and
n t is therefore given a detailed
separate analysis in a subsequent part of this chapter.
The interfei-ence of Prakl-it is mainly felt on a semantic and syntactic level.
 J. Harmatta

The texts were translations of Prakrit originals, full of religious terms, which
had no exact equivalents in Aramaic. The translators had to resort to semantic
borrowings, as, for example, Prakrit dbamma- 'piety' = Aramaic qiyt' 'truth',
Prakrit piipa- 'evil' = Aramaic my" 'malady', Prakrit poriina pakiti 'according to
ancient rule' = Aramaic j k 'srhy hlqwt' 'as destiny imposed', etc., or to the reli-
gious vocabulary of another Iranian language, different from Old Persian, as,
for example, Prakrit sacce 'veracity' = Avestan arazuiii- (original meaning,
'veracity'), Prakrit guru- 'master' = Avestan maziita- 'the greatest', etc. There
are simple transcriptions in Aramaic letters of Prakrit passages and Prakrit
interference can be observed even o n a syntactic level. The word order b'lwl
m % 'in the month Elul' cannot be explained either by Aramaic or by Iranian
linguistic usage because the correct order of words would be ~ r 'lwl h in ha-
rnaic and m'h 'lwl in Iranian. In Sanskrit and Prakrit, however, the word order
is inverted, for example, Mdgbamdse 'in the month Magha', Pa~!arndsa- 'in the
month Pausa', etc. Consequently the phrase b'lwl m'b owes its word order to
Indian Sanskrit o r Prakrit interference.

The survival of Aramaic

The use of Aramaic script and language for administrative and economic
records apparently survived u p to the middle of the second century B.C. An
Aramaic ostracon found at Ay Khanum (hitherto undeciphered) reads:

Column I Column 2

Line x +1 IIIII Line x +1 zbyn X I I kwnywk I I

x +2 ] ibwk X IIIII x +2 kwrkln s' X I I hmwk
[XXJXX

Note: . = faint letters (except h, being the transcription of h a )

[I = disappeared letters.

The record is written in two columns of which the upper parts are missing. The
beginnings of lines x + 1 and x + 2 of Column 1 are also broken off but the con-
tents can be restored. The text consists of an enumeration of names and quanti-
ties of grain. The key-word of the record is abbreviated in the for111 i, which
must represent Aramaic i'rn 'barley'. The text of the ostracon can thus be inter-
preted:

Column I Column 2

Line x +1 [N.N.] I I I I Line x +1 Zbcn X I 1 Kav ~ i v a I1

k
x +2 [Uxliebovak X V I I I I x +2 Kur Kalan b(ar1ey) XI1
equal to [ X X I X X
 Languages and scripts in Graeco-Bactria and the Saka kingdoms

The measure for grain might have been the 'eqa (36.44 1) which was probably
mentioned in the lost first line. The indication i 'barleyJ in line x + 2 of Column
2 ~uggeststhat another sort of grain (wheat, millet?) was mentioned earlier. The
~srracontherefore represents a notice of the quantities of grain delivered by the
enumerated proprietors to the treasury of Ay Khanum. The Iranian word
hmwk *hamok 'equal, like' introduces the total of the delivered grain.
The names of the proprietors deserve particular attention. Zben may be
explained by Parthian zbyn 'attractive'. Kav Nevak means 'Lord Brave', kav
being the Eastern Iranian title kavi-, while the first component of the name Kur
Kahn may be the Eastern Iranian kur 'youth, boy' and the second one may be
compared to Parthian kaliin 'great, big'. The restored name [Ux]icbovak also
occurs on the Greek ostraca of Ay Khanum in the spelling Oxeboakos, going
back to Old Eastern Iranian "Uxiya-bavaka- 'grain growing' (cf. Parthian
b'wg, 'seed, grain, fruit'). With the help of these names and of the word hamok
we gain a modest insight into the language spoken by the Iranian population of
~y Khanum just before the Yiieh-chih invasion.

The language of ancient Bactria

The exact character of the Iranian language spoken by the ordinary population
of ancient Bactria has long been a tantalizing problem for linguistic research. In
the fifth and the thirteenth rock edicts of Aioka, the Yona Kamboja are men-
tioned as neighbours of Gandhara living within the borders of the Mauryan
Empire. As the name Yona denotes the Bactrian Greeks, linguistic research has
presumed that the Iranian elements in the Aramaic inscriptions found at Taxila,
Laghman and Kandahar must represent the language of the Kambojas. How-
ever, this logical conclusion only defers a solution of the problem, because the
language of the Kambojas is not known either. It is useless to guess that the lan-
guage of the Kambojas (i.e. the language spoken by the Iranian population of
ancient Bactria) might have been some minor Eastern Iranian language like
Ormu0 or Parati, because the language, which provided the Iranian basis for
Imperial Aramaic in the satrap's chancellery, could only be an important,
widely spread language of Bactria. For this language of ancient Bactria, we have
as evidence the following:

1. The testimony of YZska

iavatir 'to go' - Avestan iav- 'to go' as against O l d Persian iij1aa-y-, Old
Indian cyavate.

2. Iranian terms in the Aramaic inscriptions of Aioka

'dbGyJ *ad,%;- 'quarrelsome' - Avestan dabnei-, f b a ~ i - dvaki-
, 'to quarrel',
[baeiah-, dvnPial?- 'quarrel', tbiijjnnt- 'quarrelsome'.
 J. Harmatta

'rzwi *arzui - Avestan rrazav- (Nom. anzii) 'true, right', srazuG-

- original meaning: 'veracity, truthfulness'.
'wsp < *d vispai - Avestan d 'in', vispa- 'all'.
bg < Y'l?aga- - Old Persian baga-, Avestan baya- 'lord, god'.
dmydt-y < *ddmiddta- - Avestan dami.data- 'creature'.
bmwk < *hama-vak-/*hama-uk- 'equal, like < saying the same'.
hwnitwn < *hu-nistdvan- - Old Persian *hu-niitiivan- 'good document',
represented by Biblical Aramaic nitwn, Imperial Aramaic nitwn' 'docu-
ment'.
hwptysty l:h~-pa8yasti'good obedience' - Avestan paiti.a-stay- 'obedienceJ.
hwwrdh < *hu-varda8a- 'good growth' - Avestan varada8a-
increase'.
hww[yit-y'] < *hvoiita- 'elder' - Avestan hvoiita- 'supreme, first, eldest'.
m'h *mah- 'month' - O l d Persian, Avestan mah- 'moon, month'.
krpty *kdra-pa8t 'army road' - O l d Persian kara- 'armed people, army', old
Persian pat%- 'road'.
mzyit-y' *maziita- 'elder' - Avestan maziita- 'the greatest'.
prbst 4:pari-basta- 'bound' - Avestan band- 'to bind', Pahlavi parvastan, par-
vand- (< *pari-band-) 'to surround, enclose'.
prbsty *pari-bastay- 'constraint'; see above.
ptyty *patitay- - Avestan paititay- 'discharge, expiation'.
ptyzbt *pati-zbdta- 'forbidden' - Avestan zbd- 'to call', PPfP zbdta-, Old
Persian patty-zbd- 'to prohibit, forbid'.
shyty *sahyatai - Avestan sah-, Old Persian 8ah- Pass 'to be called'.
ity *idtay- - Avestan idtay- 'happiness, joy', O l d Persian Gydtay-.
trb'trh *8ri-hdOra- 'tripartite pinfold, caravanserai' - Avestan Oray-lBri-
'three', haera- 'section, pinfold'.
zwn-y' *iivana- 'living being' - Avestan jva- (Avestan spelling for j'iva- 'liv-
ing'), Avestan jvana- (Avestan spelling for jivana- 'living').

3. Bactrian names
(In Greek transcription: (e = eta ('I), 6 = Omega (to), y = ypsilon (u), ou =
omicron + ypsilon (ov)):
Aitates *Etdt, by haplology from +Aitatdt- 'glitter, lustre' - Avestan aita-
'glittering'.
Apama *Apama 'supreme' < jCUpamd-- Avestan upama- 'highest'.
Artabazos *Arta-bdzu- 'whose stay is Arta' - Avestan a n t a - 'right, law, holy
right' and Avestan bazu- 'arm, stay'.
Artanes (read Aryandes formerly) *Artand- 'righteous' (cf. Avestan arata-
above).
Artasouras :*Arts-sira- 'mighty by Arta' - Avestan a n t a - 'holy right' and
Avestan sura- 'mighty'.
Atrosokes *Atrd-~auka-- Avestan itra.saoka- 'firebrand'.
 L a n g ~ a g e sand scripts in Graeco-Bactria and the Saka kingdoms

Barzandis "Barzand- - Avestan barazant- 'highp.

~ ~ t a p h e r n *Data-farnab-
es 'who has glory by right' - Avestan, Old Persian
dzta 'right', O l d Persian (from Median) famah- 'glory'.
Itanis *Vitana- 'corpulent' - Avestan tan- 'to extend'.
Katanis *Katana- 'honourable' - Avestan k i t a - 'honoured'.
Mithroaxos *Mit)ra-vaxia- 'advancing by Mithra' - Avestan vaxi- 'to grow,
rise, advance'.
Oxyartis "Vaxiu-varta- 'chosen by VaxSu' - Avestan var- 'to choose'.
Oxiboakos < *Uxiya-bavaka- 'grain growing'; cf. above.
Oxydatis < * Vaxiu-data- 'begotten by VaxSu'.
Orsodatis < *Riva-dita- 'legitimately begotten' - Avestan ~~raiva- 'legal,
legitimate'.
Oumanis < * Vohu-manah- - Avestan Vohu-manah-.
Xaranos < *Xiarana- 'seedy' - Saka sara 'seed', Ossetic uxsar 'nut'.
Parzos *Parza- < *Parts- 'donor'; cf. O l d Indian parc- 'to mix, fill, increase,
etc.', parka- 'mixture, gift'.
RGxanF *Roxiana 'brightness'; cf. Avestan raoxina- 'light'.
Sinokratis (Hellenized from *Sinochrates) *Cina-xrattr- 'who has wish for
wisdom' - Avestan tinah- 'desire, wish', xratu- 'wisdom'.
S6chrakis (on a ~ o t s h e r dfound at Dushanbe) *Suxrak 'red' - Avestan suxra-
'red'.
Spitamenis *Spits-manah- 'of splendid intelligence'.
[Telirixares *Tiri-xiara- 'progeny [seed] of E r ' ; cf. above.
Hyspasinis *Hu-spas-ina- 'keen observer, spy' or *Hispas-ina- - Avestan
spas-, Pres. hispas- 'to watch'.

In Aramaic transcription:
'hwty *Ahuti- 'press-hill, offering-hill' - Avestan a + hutay- 'haoma-press-
ing'.
['hlibwk * Uxiebovak < * UxiYa-bavaka-; cf. above.
zbyn < "Zbayana- 'calling, inviting, attractive'; cf. Man. Parthian zben
'attractive'.
kw *Kav 'valiant, prince' - Avestan Kavi-.
kwr *Kur 'youth' - Eastern Iranian kuru- 'youth, boy'.
kln * K a l ~ n'big'; cf. Man. Parthian kalan 'big'.
nywk *Nevak 'valiant, brave'; cf. Middle Persian nev 'brave, valiant', nevak
'good'.
trmd "Tarmid < *Tars-rnada- 'oversized, great [hill]' - Avestan tnr6
'beyond, over', mad- 'to measure'.
w'iw *Wijava < :kVayta-van- 'charioteer' - Avestan vaia- 'chariot'.
whiwprt *Vaxiu-fiita- 'favoured by VaaEu' - Avestan fray-, PPfP frita- 'to
win the favour of a god'.
 J. Harmatta

From among the fifty-six items of linguistic data discussed above, there are only
some terms and names, namely: *hwniitavan- 'good document', *kzrapaoi-
'army road', *pati-zba- 'to prohibit, forbid' - being a 'Median' loan-word in
Old Persian itself - *NCvak (Old Iranian *naiba- 'good, brave' is only attested
in O l d Persian so far), "KalZn (Parthian) which seem to be borrowings from
Old Persian or Parthian, and a few others, such as *pari-basta- 'bound', *.purr-
bastay- 'constraint', *mah- 'month', which might also be of Old Persian origin,
The overwhelming majority (forty-eight out of fifty-six) reflect another
Iranian language, different from O l d Persian. O n the testimony of Yiska, this
Iranian language can be ascribed to the Kambojas living in the neighbourhood
of G a n d h ~ r a In
. Sanskrit and PZli literature, the Kambojas were one of the six-
teen great peoples of Indian geography. They killed worms, insects, moths,
snakes and frogs and thought that they acquired religious merit by this activity,
This clearly characterized them as Zoroastrians. They were ruled by kings and
became famous for their horse-breeding. Kaurilya mentions the K a ~ b o j ahorse
as one of the best breeds for war and speaks of the Kambojas' military organ-
ization and their warlike way of life. All elements of this description concerning
the Kambojas fit excellently the ancient Bactrians.
Moreover, among the linguistic data quoted above, there are a series of
special Avestan terms such as *arzuiZ- 'veracity', *damidZta- 'creature',
*hupaOyasti- 'good obedience', "huvarda6a- 'good growth', "hvoiita- 'elder',
"patitay- 'discharge, expiation', *sahyatai 'it is called', *SZti- 'happiness, joy',
*haera-, in *&i-haera-, 'pinfold', *iivana- 'living being'. The other part of the
linguistic evidence (terms and names) similarly coincides with the language of
the Avesta: iav- 'to go', *adpaiia- 'quarrelsome', :% vispai 'in all', *baga- 'lord,
god', "maziita- 'elder', *Etdt-, *Upama-, *Artabdzu-, *Artasura-, *Artdna-,
'"arzand-, *Datafarnab -, * Vitana-, *Katana-, *Mieravaxta-, * Vaxiuvarta-,
* Vaxiubiizu-, * VaxSud~ta-, * UxSjabavaka-, *Rivadata-, * Vohumannh-,
*RauxSand-, *Cinaxratu-, *Suxraka-, *Spitamanah-, "Huspasina-, *Ahtiti-,
:':Zbayana-, *Kavi-, "Kuru-, "Taramada-, :kViiiavan-, *Vaxiufrita-. Perhaps the
names Xiarana- and [7'Jirixs'dra- represent another Eastern Iranian language.
The testimony of this linguistic data is unambiguous. The overwhelming
majority of Iranian terms and names occurring in Aramaic and Greek docu-
ments of ancient Graeco-Bactria o r mentioned as Bactrian in ancient Greek Ilt-
erature represent a language essentially identical to Avestan. N o t even a dialectal
difference can be observed between this linguistic evidence and the language of
the Avesta; the two main dialects of Avestan are both reflected by the material.
O n the basis of the linguistic evidence available at present, it therefore
seems very likely that Avestan was the language spoken ( ~ e r h a in ~ sseveral var-
iants o r dialects) and used for administration in Graeco-Bactria and other east-
ern Iranian countries (such as later Sistan) and the Middle Iranian development
of Avestan must have been the most important language in eastern Iran on the
eve of the Saka and Yiieh-chih invasion.
 Languages and scripts in Graeco-Bacrrja and the Saka kingdoms

Greek language and script in Central Asia

Greek language and script had appeared in Central Asia long before the con-
quests of Alexander the Great. According to Herodotus, Darius I took the sur-
viving Milesians t o Susa after the capture of Miletus, and later settled them near
the mouth of the Tigris, while the inhabitants of Barke were transferred to Bac-
tria. Earlier historical research presumed that Hellenism was introduced into
central Asia by these earlier Greek settlers. The presence of Greeks in Central
Asia of the Achaemenid period can hardly be denied, but the general spread and
use of Greek script and language and the rise of Greek culture were only the
result of Hellenistic colonization. In the course of his military expedition
against Spitamenes, by order of Alexander, Hephaestion settled Greek and
Macedonian soldiers in the villages and cities of northern Bactria and Sogdiana
(Arrian IV.16.3; 17.4). As a consequence of this colonization large territories of
Central Asia became Hellenized and by way of Graeco-Iranian symbiosis the
use of Greek script and language spread among the Bactrian aristocracy.
All remains of the Greek language discovered so far in Central Asia date
from the Hellenistic age and represent the koine, the standard Hellenistic lan-
guage. The Greeks of Central Asia must therefore have maintained their close
relations with the Hellenistic kingdoms of the Near East until the rise of Parthia
and the Roman conquest. The Greek texts found in Central Asia certainly
reflect the same cultural level as was achieved in the great centres of Greek civil-
ization. The best evidence for the high culture of the Bactrian Greeks is seen in
the Greek translation of Rock Edicts 12 and 13 of ASoka:

[.eb]oifie~axai k Y x Q a ~ xara ~ i a niioaq t h q 6 i a r ~ ~ p b.kyx~urilq q lta)Llo~u

hi kor~v
ijq by yhho~lqkyxQarilS ill. Kai F L ~ ~ TkaurobS
E kna[i]vTooiv, ~ i i l trtov ~ nkhaq
~ i y o o J~C vE Q ~p116~voq'XEVOY y a kor~v. ~ xai nei~iio0ai~iiihhovrobs rrkhuq knut-
vgh xai ~ ) E V E L Yx a ~ navra
a t ~ o n o vT . a i ~ r a6'F x o i o k v r ~kauroilq
~ ai;50~ot
xui robS nEAaq hvax~Wvtai-n a ~ a p a i v o v r ~6;q railra hxh~i.are~oi TE yivovrut
xai t o i ~ nihaq h n 6 X 0 ~ v ~"Oi' . kauroirq krraivtooiv, rob2 b n ~ h a cq~iyt~otv
a ~b'ijlv
cp~hori~ior~ 6iarrebrovrat,
~ov f i ~ v h b ~ t ~nv ao t~ robc a holnoilq ~yhixyy~ut, ;roicil h'e
khhov fiACrrro~[~~] kauroilq. neirrt.1 6 i hhhilhouq 0azlcta~eivxai r a h ~ i ~ h r l ~ v
6~bb.yl~ara nu~ubi.~eaea[t]. T a i ~ r a 6'F ~ O I O ~ ~ V ~T O E ~~ ~ ' ~ ~ C L ~~o€oU ~ 'T~ E
u 1Q, O
n a ~ a b t 6 o v r hhhilhotq
~~ boa '(.xau~oqailrTov k n i o ~ a ~ u iKai . roiq rcli~rct
kJL[tK]oxo~!oi railrtx bxv~ivA ~ ~ E yya T I \ ~ ,'I~vT'OS F ~ I U F-~ O ~ ~ V
L V ~ E L ~ I I E ~ V ( I J ( bta
'Oybowi '(.TEL [~aoih~ilovroq niobc~ooozl x a r ~ o r ~ e x r arilv t K~hiyyilv. . , 'Hj'
k~wY@llIi~va xui k;ilyyi.vcx ~ ~ E T Oatolt~rcov E ~ I I Z I Q I ~ ~ ~E E~ X U ~ ~ V nut T E CLVCIL-
~i01louv'iwhhai I ~ ~ j Q i i l bkxu h ~ F ; xai ux~bbvFthhot ~ o u o i ~ r u~ ir ~ k ~ i l r i ~ 'A o:'a v .
Cxrivon roil xQb\~ozlBAroq xui o:xroc abrbv ~ A C K ~ E xai V . pa~leq 'i1v~yxt.v.61' oil
rgbnov ~ X ~ ~ F Ih~nFkVX ~ ~T ~ ~a klVC , ~ / I ~ l X t Oo \I~T O U ~ ~ ] TE V XCI'L U ~ ' V T CJIFITO~I~TCIL
I~L\'
n ~ ~ i~l o i~ p ~ i Kai a q . TOGTO ?TI ~ V ~ ~ F Q E ~ T E Qi,nt.ihilveOV b f i u u t ~ ~ i xai
~ l ; .iioot
k x ~(i~ ~ ' X O I IPQCICIEV(III
V UQ(L~IEVCLL i.1 xai ithhoi TL\;E~ oi nee; r i p E;'UE(IELC~V h t u r ~ i -
fiovr~q,roils kxei oixoi,v~ct~ 'i.6~1 t c t roil (Iuothi.to; o v l t y l ~ ~ o vvoei\~, ~ t r xai btbka-
 xuAov x n i n n t i e a x(& p q r ~ e ak n a ~ o ~ b v e o Ox u ~ i ~ ~ I J ~ L ~~ ~~ Ei L ~v krui-
xUi o, u ~
eovq ;lynnhv x a i llil ~ L ~ I ~ E ~ ~6 ocboh oO~ 5~x uLi pLoOwt0is , k5 X O U C ~ ~ T U Tx<u.
U
o o a l , robrwv hxei r b v t o t a i t a btnneaooopivwv EL n 5 rirHvilxey ij kEjXraL, Xui
ro;ro $1 n a ~ a 6 ~ 0 l qo;r A o ~ n o i l y ~ i v ~ abt , B'E [Plao~Aebgo o 0 6 e ~hni TO^^^^^
h b u o X ~ g a ~ v e~ L t o i s AOLJCO'~S k ~ v e o i vECOLV[
v . a~ iT kv

T r a n s l a t i o n of R o c k E d i c t 12:

. . . pietyand self-control in all philosophical schools. But mostly self-possessed is

that [man] w h o is master of his tongue. And they d o neither praise themselves nor
belittle their fellows in any respect. This is, namely, a vain thing. It is better to
praise their fellows and not t o belittle them in any way. Doing this they aggrandize
themselves and captivate their fellows; transgressing this, however, they will be
discredited and become odious in the eye of their fellows. W h o praise themselves
and belittle their fellows, those behave t o o ambitiously: wanting to excel more
than the others, the more they d o harm to themselves. I t is the correct thing to re-
spect one another and t o accept the lessons of each other. Doing this they enlarge
their knowledge as far as they share with one another that which anyone knows.
And one does not hesitate about saying these t o those w h o are practising these in
.order that they d o persist in exercising ~ i e t yat all times.

T r a n s l a t i o n of R o c k E d i c t 13:

In the eighth year of his reign Piodasses [= Priyadariin = ASoka] conquered Kal-
inga. A hundred and fifty thousand individuals were taken prisoner and deported
from there and another hundred thousand were killed and almost the same number
of individuals died. From that time o n pity and repentance overcame him and he
was heavily distressed. Therefore, he gave an order to abstain from [killing] the liv-
ing beings and made endeavour and effort t o exercise piety. And the King consi-
dered even more grievous the following: as many brahmanas and iramanas are Iiv-
ing there, they have to recognize what is useful t o the King and to respect and to
honour their master, their father and their mother, to like their friends and fellows
and not t o deceive them, t o use their slaves and servants as kindly as possible - if
anybody died o r was deported from among those w h o are living under such con-
ditions there, and the others regard this as a matter of secondary importance, the
King, however, was hotly angry with these. And that there are . . . with the other
peoples . . .

O n t h e basis of s o m e stylistic f e a t u r e s i t is e a s y t o s e e t h a t R o c k Edicts 12 and

13 of A l o k a w e r e t r a n s l a t e d i n t o G r e e k by t w o d i f f e r e n t translators. Both of
them were profoundly erudite and used Greek t e r m s in their
translations. T h u s , t h e p h r a s e 6nbxeoOu1. siuv lpcpCxo,v ' t o abstain f r o m [kill-
ing] t h e living beings' reflects P y t h a g o r e a n p h i l o s o p h y , w h i l e t h e terms f y x ~ h -
reia 'self-control' a n d yhiuoqs lyxearfis ' m a s t e r of his t o n g u c ' go back t o the
P l a t o n i c s c h o o l . O t h e r e x p r e s s i o n s s u c h as, f o r example, bibaypa 'lesson',
 Languages and scripts in Grarco-Hattria and rhe $aka kingdoms

n o h v l r ~ ~'polymathy',
i~ i.xhhpnetv 'to excel', i p n u e u h p o p i L fi.jFio~)uL
'to
consider a matter of secondary importance' characterize the vocabulary of
plate, ~ e n o ~ h o Isocrates
n, and Aristotle. The phrase onoubil xcil obvruoLs
(written owvtaEts) occurs in Plato's Symposium.
Being well educated and widely read and both writing in koine, the two
differed from each other in their stylistic ambitions. The translator
of ~ o c Edict
k 12 claimed t o be considered an erudite person, therefore he used
the Attic form b ~ a r c e h t o v t a l(they behave) because the Attic forms in a text
written in koine always reflected the writer's claim to erudition.
The other translator, trying to write pure koine, used the kozne form b ~ u -
~ l ~ a a o cinstead
o of the Attic b l a x ~ a t t c oand he created a verbal form x u ~ k -
ateentat, a hyperkoinism for xatkatgarctai. The Greek text of Rock Edicts 12
and 13 of Aioka shows the importance of the Greek population living in Cen-
tral Asia and permits a remarkable insight into their intellectual life, erudition
and literary ambitions.
The Greek stone inscriptions of A y Khanum and Takht-i Sangin are
another interesting group. The Takht-i Sangin inscription was set up by an Ira-
nian in honour of the god 0x0s. Its text runs: (1) Efifiv (2) &VEOI~XEV(3)
~ l ~" O ~ o'Atrosokes
' A t ~ o a h x (4) t dedicated his votive present to 0x0s'.
While Greek inscriptions can be taken for granted in the Greek cities of
Central Asia such as A y Khanum, the votive inscription from Takht-i Sangin is
surprising because the donor bears an Iranian name and dedicates his votive
present to an Iranian deity but does so in Greek script and language. The pecu-
liarity of this attitude becomes clear when we compare the dedication of Takht-i
Sangin with the Besnagar pillar inscription of Heliodorus, the ambassador of
King Antialcidas. Heliodorus was a Greek who became a worshipper of Vishnu
and obviously had a good knowledge of Brahmi script and Prakrit language as
he was sent to negotiate with King Kosiputra Bhigabhadra. Thus, as a half-
Indianized Greek, he erected the Garuda pillar inscribed with a text written in
Brihmi script and Prakrit language.
A Prakrit inscription in honour of an Indian deity prepared by a Greek
worshipper of the Indian god can be taken for ganted. But in Takht-i Sangin it
was an Iranian w h o used Greek script and language for the dedication of his
votive present t o the Iranian !god VaxSu. H e kept his own native religious ideas
and was not Hellenized in this respect; but he probably had some knowledge of
Greek and, for lack of a Bactrian written language, used Greek for his dedica-
tion, which must also have been understandable t o the priests of the sanctuary
of VaxSu. This single inscription clearly attests the spread of Greek script and
language a n ~ o n gthe Bactrian aristocracy and priesthood.
The use of Greek script and language was, however, not limited to the
public life of the Greek cities and to the needs of the Iranian sanctuaries. Greek
was also introduced into the administration as we can see from the ostraca
found in the treasury at Ay Khanum. Its occurrence on the ostraca of Iranian
 J. Harmatta

xrthov xcti n a r i p a x a i ~ q r i i~n aat o x b v e o O u ~x d O a ~ p h t ~ qihoug v, xui irui-

e o u c i y a n i l v x a i lCv irlayleubeo0a~,b o v h o y xai pro0u)rois ks x o v y l b r a ~xPh- ~
0 8 ~ 1 r, o b r a ~ vkxri r b v r o ~ a t r ab ~ u n e a o o o l i v w vEL TLS T E O V ~ X F i)~ ktinTuL,xui
roijro f l l naeaBeop$ o i h o ~ n o it y ~ i v r a b~ ,bk [ ~ ] ~ o ~ h t ocpobeu .ws kni ro(lrolc
k b v o X k Q ~ l v ~~v . abrti kv t o i s homois k8veoiv e;olv[

Translation of Rock Edict 12:

. . . piety and self-control in all philosophical S C ~ O O I S . But mostly self-possessed is

that [man] w h o is master of his tongue. And they d o neither praise themselves nor
belittle their fellows in any respect. This is, namely, a vain thing. It is better to
praise their fellows and not to belittle them in any way. Doing this they aggrandize
themselves and captivate their fellows; transgressing this, however, they will be
discredited and become odious in the eye of their fellows. W h o praise themselves
and belittle their fellows, those behave t o o ambitiously: wanting to excel more
than the others, the more they d o harm t o themselves. It is the correct thing to re-
spect one another and t o accept the lessons of each other. Doing this they enlarge
their knowledge as far as they share with one another that which anyone knows.
And one does not hesitate about saying these t o those w h o are practising these in
.order that they d o ~ e r s i s tin exercising piety at all times.

Translation of Rock Edict 13:

In the eighth year of his reign Piodasses [= Priyadariin = Aioka] conquered Kal-
inga. A hundred and fifty thousand individuals were taken prisoner and deported
from there and another hundred thousand were killed and almost the same number
of individuals died. From that time o n pity and repentance overcame him and he
was heavily distressed. Therefore, he gave an order t o abstain from [killing] the liv-
ing beings and made endeavour and effort t o exercise piety. And the King consi-
dered even more grievous the following: as many brdhmanas and iramanas are liv-
ing there, they have to recognize what is useful t o the King and to respect and to
honour their master, their father and their mother, t o like their friends and fellows
and not t o deceive them, t o use their slaves and servants as kindly as possible - if
anybody died o r was deported from among those w h o are living under such con-
ditions there, and the others regard this as a matter of secondary importance, the
King, however, was hotly angry with these. And that there are . . . with the other
peoples . . .

O n the basis of some stylistic features it is easy t o see that Rock Edicts I 2 and
13 of Aioka were translated into Greek by two different translators. Both of
them were profoundly erudite and used Greek philosophical terms in their
translations. Thus, the phrase ZnixeoHai siuv El~cpb~wv 'to abstain from [kill-
ing] the living beings' reflects Pythagorean philosophy, while the terms l y n ~ h -
t e i a 'self-control' and yhioo~lsC y x ~ a s 'master
i~ of his tongue' go back t o the
Platonic school. Other expressions such as, for exa,nple, bibaypa 'lesson',
 Languages and scripts in Graeco-Hactria and thc Saka kingdoms

n o h u ~ a 'polymathy',
~i~ dxhupnclv 'to excel', i.p nupuhgopill il+/cioOuL'to
consider a matter of secondary importance' characterize the vocabulary of
plate, ~ e n o ~ h o Isocrates
n, and Aristotle. The phrase ( ~ n o u b i lxrli n b v r u c r ~ ~
(written abvragi5) occurs in Plato's Symposium.
Being well educated and widely read and both writing in koine, the two
translators differed from each other in their stylistic ambitions. Thc translator
of Rock Edict 12 claimed t o be considered an erudite person, therefore he uscd
the Attic form G i a n e a t o v t a l (they behave) because the Attic forms in a text
written in koine always reflected the writer's claim to crudition.
The other translator, trying t o write pure koine, used the koine form btu-
n e a o a o instead of the Attic Glaneuttco and he created a verbal form xurk-
ateEntcx1, a hyperkoinism for x a t k a t ~ a n t a i The
. Greek text of Rock Edicts 12
and 13 of ASoka shows the importance of the Greek population living in Cen-
tral Asia and permits a remarkable insight into their intellectual life, erudition
and literary ambitions.
The Greek stone inscriptions of Ay Khanum and Takht-i Sangin are
another interesting group. The Takht-i Sangin inscription was set up bv an Ira-
nian in honour of the god Oxas. Its text runs: (1) E6Clv (2) &vkOqr,~v(3)
' A t ~ o a b x q(4)
~ "Occo~'Atrosokes dedicated his votive present to 0x0s'.
While Greek inscriptions can be taken for granted in the Greek cities of
Central Asia such as Ay Khanum, the votive inscription from Takht-i Sangin is
surprising because the donor bears an Iranian name and dedicates his votive
present t o an Iranian deity but does so in Greek script and language. The pecu-
liarity of this attitude becomes clear when we compare the dedication of Takht-i
Sangin with the Besnagar pillar inscription of Heliodorus, the ambassador of
King Antialcidas. Heliodorus was a Greek who became a worshipper of Vishnu
and obviously had a good knowledge of Brahmi script and Prakrit language as
he was sent t o negotiate with King Kosiputra Bhagabhadra. Thus, as a half-
Indianized Greek, he erected the Garuda pillar inscribed with a text written in
Briihmi script and Prakrit language.
A Prakrit inscription in honour of an Indian deity ~ r e ~ a r ebyd a Greek
worshipper of the Indian god can be taken for granted. But in Takht-i Sangin it
was an Iranian w h o used Greek script and language for the dedication of his
votive present t o the Iranian god VaxSu. H e kept his own native religious ideas
and was not Hellenized in this respect; but he ~ r o b a b had l ~ some knowledge of
Greek and, for lack of a Bactrian written language, used Greek for his dedica-
tion, which must also have been understandable to the priests of the sanctuary
of VaxSu. This single inscription clearly attests the spread of Greek script and
language among the Bactrian aristocracy and priesthood.
The use of Greek script and language was, however, not limited to the
public life of the Greek cities and to the needs of the Iranian sanctuaries. Greek
was also introduced into the administration as we can see from the ostraca
found in the treasurv at Ay Khanum. Its occurrence on the ostraca of Iranian
subaltern treasurers as, for example, Artanes, Barzandes, Oxeboakos and oxy-
bazos proves that Iranians working as officers in the Graeco-Bactrian admin-
istration were well acquainted with the Greek script and language. If the frag-
mentary name l i b w k on the Aramaic ostracon discussed above is correctly
restored as ['hlibwk < *Uxiya-bavaka-, then the person mentioned might have
been identical with Oxeboakos occurring on the Greek ostraca as one of the
subaltern treasurers at A y Khanum. In that case, we can assume that the Iranian
treasury officers were acquainted with both Aramaic and Greek chancellery
practice, and the two scripts and languages were used in parallel in the Graeco-
Bactrian administration.
Recent finds of inscribed potsherds at Birkot and Udegram (Swat, Pakis-
tan) prove that Greek language and script were still being used there in the
second and first centuries B.C. The Birkot inscription consists of two names
which may be restored as (1) [Eweu]Gfi[pw~](2) [ ' ~ p ] b v t a [ ~(1)
] T o [Euthylde-
[mos] - (2) Am]ynta[s], that is, it qualifies the pot as a present given by [Amlyn-
ta[s] to [Euthy]de[mos]. O n the Udegram potsherd the genitive case of the
Greek word voV5 has been reserved, forming the second part of a compound
name like ['Avt~,]voow and indicating the owner of the pot.
There are some indications that the use of Greek survived the fall of the
Graeco-Bactrians. At Dilberjin, two inscribed amphorae were found in Room
20. They belong to the fourth building ~ e r i o dof the room, which seems to
belong t o the age of the Great Kushans, since the abandonment of the room be-
tween the second and third building periods reflects the events of the Kushan
conquest. The inscriptions written in Greek run as follows:

They can only be interpreted o n the assumption that they consist of abbrevia-
tions:

'The load of the vessel: 50 diploun sesame [oil]'.

After the rise of the Kushan Empirc, the Greek scribes, masons and artisans
were working for the new Iranian aristocracy. A Greek architect or mason is
still mentioned with the Greek phrase hia IIuhnpfihov 'by Palamedes' in one of
the Bactrian inscriptions from Surkh Kotal - a phrase that provides evidence for
the survival of the Greeks and their language in the Kushan Empire up to the
end of the second century A.r).
 Languages and scripts in Graeco-Bactria and the Saka kingdoms

The language of the Southern Sakas

In the course of the ethnic movements caused by the rise of the H s i ~ n ~ - ~ "
nomadic empire, four Saka tribal groups settled on the territory of Parthia and
the Later Kushan Kingdom, namely, the Sakas of Sistan, the Sakas of Gandhira
and the Panjab, the Sakas ruling in Mathura and the Sakas of Surashtra and
Malwa. For the language of Saka groups, we have only the scattered evidence of
names and terms occurring in Kharosfhi and Brihmi inscriptions. Relatively
abundant are the data for the Sakas of Mathura, while for the Sakas of Sistan we
only have very scanty evidence.
The Sakas w h o invaded Bactria appear in the sources under different
names, namely, Indian Saka-murunda-, Chinese Sai-wang, Greek *Sakarau-
kai, Latin *Sa<ca>raucae. Of these both Indian Saka-murunda- and Chinese
Sai-wang mean 'Saka king' and 'Saka kings', respectively, in so far as
murunda- can be regarded as the Saka title for 'lord, king' and Chinese wang
as the translation of it. As both the Chinese and the Graeco-Latin sources
mention the same peoples as conquerors of Bactria, we have t o regard the
Sakaraukai as identical with the Saka-murunda- and the Sai-wang respec-
tively. Accordingly, the element -rauk- in the name Sakaraukai must have the
same meaning as Saka murunda- and Chinese wang. In fact, the word can be
compared t o Khotanese Saka r i k y a - 'commander, lord', going back t o "rau-
kya-. Saka murunda-, too, has an equivalent in Khotanese Saka: rrund- 'pos-
sessing power, lord, king'. As it is proved by Saka murunda-, both Khotanese
terms rrund- and r i k y a - derive from the root *mrav-l*mru- 'to declare, t o
order' as *mrav-ant- and *mrav-aka-/*mrau-ka- respectively. Old Iranian
%r- was reduced t o r- in Khotanese Saka, while in the language of the Sakas
of G a n d h ~ r athe initial mr- was preserved. It is a remarkable fact that the
outcome of Old Iranian *mrav-ant- and "mrav-aka- was different in the
Western and the Eastern Saka tribal groups. This phenomenon clearl~.sup-
ports the theory according t o which the *Saki m r a v a k i (> *Saki r a u k i ) and
the *Sakd mravantah (> *Saki murunda) - both meaning 'Saka lords' or
'Saka kings' - invaded Bactria and Gandhira separately. The name 'Saka
lords/Saka kings' originally denoted the Saka tribal aristocracy who were
alone able t o wander away from their territory while the common people
remained at home.
Beside the term *rauka- 'lord, king', the language of the Sakas, settled in
Sistan, seems to be represented by the following names, o r titles:

Aya ( G i n d h i r i form), AzPs (in Greek script) < "Azn- 'leader', Kroraina hinajha-
("hindza-) 'commander of army' title of the King of Khotan.
Ayiliia (Gindhari form), AzilisPs (in Greek script) < ",4znliia- 'cornn~ander-in-
chief; literally: leading commander'; "liia- represents a variant of riin-
'desiring, cornmanding'.
Spalagadama 'commander of army': spala (ipala is a Gindhari form) is
obviously borrowed from Parthian *spdha 'army', gadama < *katama-
'commander' from ka- 'to desire'.
Spalahora 'commander of army': spala 'army' < Parthian *sprida 'army', Khota-
nese Saka spdta 'military official' may also be an adoption of Parthian
"spida-pati- 'commander of army'. Hora < *haura is the same word as
Khotanese Saka haura- 'leader, commander'.
Spaliriia 'in command of army': riia spelling for riia- 'desiring, commanding',
Khotanese Saka rrii- 'to desire'; for the meaning cf. Sogdian ryz-krJk
'sovereign'.

For the language of the Sakas of Gandhara and the Panjab the following names
and terms may be quoted:

Murunda 'lord, king'.

Ksahayata is not a name but a title as is proved by its joint use together with
ksatrapa: Liaka Kusulaka is styled as ksaharatasa cuksasa ca kZatrapasain
the Taxila copper plate, Bhzinzaka is named ksaharatasa kSatrapasa on his
coins, similarly Nahapana bears the titles rdfio ksahardtasa kSatrapasa in
the Nasik inscriptions. Ksahardta may go back to Old Iranian *xiaOra-
pati- 'lord of the country', the phonetic development of which was similar
to that of Khotanese Saka spata 'military official' < *spdda-pati-. Both ele-
ments of the term survived in Khotanese Saka: k r d ~ a -'power, dominion'
(< "ksahra- < ;fc~iaOra-) and -vata- 'lord' in phar~avata-'judge' < *fraia-
pati-. K~ahayZta- (< "ksahra-vata-) may be the Saka synonym of Old
Indian ksatrapa- 'protector of the country'. This would best explain the
joint use of the two terms o n coins and in inscriptions.
Moga, Moa (in Prakrit inscriptions), Maues (in Greek script) < "mava-,
'fcmauka-'tiger, hero', Khotanese Saka mauya-, muyi- 'tiger', Sogdian
m y w 'tiger' < "mavya-.
Liaka < *rya-ka- 'youth', Khotanese Saka rya- 'youth'.
Kusulaka < *Kuzula-ka- 'striving, ambitious, energetic', Khotanese Saka k@s-
'to seek, search'.
Patika, Gandhari spelling for Saka *Padika 'leader', Khotanese Saka pat4
patana 'before, in front of'.
Jihonika < :*Jihonyaka- 'benefactor', Khotanese Saka jehufin-, g~ehdiia-future
participle, to jeh-, jih- 'to heal'.
Manigula < "Manya-kula- 'delightful progeny', Khotanese Saka rndn~la
'delighted' and -kula- 'progeny' in ysarkula-.
Arajhamda < *ara-zata- 'noble-born', Khotanese Saka ara- 'worthy, noble',
ysdta 'born'.
Denipa < "Daina-pava 'protector of the religion', Khotanese Saka pa- 'to pro-
tect', Avestail daind- 'religion'.
 Languages and scripts in Grarco-Ractria and thc Saka kingdoms

Horarada 'rejoicing, rich in gift', Khotanese Saka hora-, harra- 'gift', tsira-
'rich'. Gandhara Saka ~ a d a'rejoicing, rich' may also be a borrowing from
Parthian o r Bactrian iad. The joint use of the two terms also occurs in
Khotanese Saka: haura rsdttdfia yandka 'maker of gifts, riches'.
~ h a ~ d a n a m<a "Zits-ndman- 'famous by birth', Khotanese Saka ysa- (<
*ysdta- cf. Sanskrit jdtam 'birth, origin'), and ndma- 'name', ndma-tsuta-
'famous'.
Damijada < *Ddmi-zdta- 'begotten by the creator', Avestan dimay- 'creator',
Khotanese Saka ysdta- 'born'.
Veipasi < *Vayaspasi < *Baga-spasa- 'servant of the Lord/God'; Khotanese Saka
wwiiwa- < *baga- in wwiiwayau 'of royal origin', spaii- 'to look', spaia
'observer'.
Veqasia < "Vayaspasika- < *Baga-spas-ika- 'servant of the LordIGod'.
Hiye (former reading hipe) 'master, swamin-', Khotanese Saka hiye 'master, S W ~ -
min-'.
Achia (former reading Adhia) < *Atayaka- 'observant, reverent', Khotanese
Saka dcyawamda 'observant, reverent' (< "dcaya-vant-).
Horamurta 'lord of gifts, danapati-', Khotanese Saka hora, haura 'gift', murta <
*mrautd < *mrau-tar- 'lord, ruler', Khotanese Saka rautci (< *mrauta-
tauna- 'commandant, ruler').

Recently Saka names and words have been discovered in the Kharoghi inscrip-
tions of Chilas, which may be connected with the linguistic remains of the Sakas
of Gandhara. In this new material we find the name of the Saka king Moga (see
above) and the terms iaa (or iao 'king') < *;aha- with the palatalization charac-
teristic of Saka, going back t o O l d Iranian *xiayaOiya-, and iae 'royal'
(< "iihiya-).
The language of the third Saka group ruling at Mathura can be character-
ized by the following linguistic data:

Sayasthana- 'Saka-land' < *$aka- ethnic name and Gandhari sthana- 'home,
land'.
Horamurndaga < *Hora-mrunda-ka- 'lord of gifts, dznapati-', Khotanese Saka
hora-, haura- 'gift', rrund- (< "mrund-) 'lord, king'.
Horaka-, shortened form of *hora-mrunda 'lord of gifts, dznapati-', Khotanese
Saka haurdka 'donor, giving'.
K~ahardta-< *xiaO~a-pati-'lord of the country'.
Viivasika, viivaiika (a title, only borne by persons of foreign, non-Indian, des-
cent) < *Bags-spas-ika- 'servant of the LordIGod'; cf. Ves;Pasi above. Vii-
wasika represents a hvper-Sanskritized form.
U h n a 'high-born', Khotanese Saka ula 'up' + the suffix -ins-.
~ i j u v u l a ,Sanskritized form of Rnjriln < *Raznla < * R i m - y a m - 'ruling king',
" ~ a z a -'directi~lg, ruling, ruler', Khotanese Saka rrajls- 'to direct, rule',
 J. Harmatta

rraysaa- 'official title', *razaka- 'ruling, ruler', vara- 'excellent, strongB;

for the meaning cf. Tocharian B walo 'ruler'.
Sudasa, S o d ~ s a'who kept the good acts in memory' < *iuda- < *bratati- 'good-
ness, good acts', Khotanese Saka iiuru, iuru 'good', beside ii;ira-
i G d e 'goodness, good acts' (< *iiaratati-) and risa 'kept in memory'.
~ h a r a ~ a l l ~Charobalano
na, (in Greek script) < *xira-vahna- 'splendid youthp,
Khotanese Saka *khara- (in kharava- 'shining'), vala-ka 'young'.
H a g i n a < *Frakana- 'leader, chief', Avestan fraka- 'forward, adjective' + -ana-,
Khotanese Saka h a < *frdk 'forward'.
Hagiimara < *Fraka-amaxia- 'whose chariot proceeds in front', Hags- <
*fraka- 'forward, adjective', Khotanese Saka h i < *frik 'forward',
-amara- < *amaxfa- 'chariot', Khotanese Saka maipa 'road, wayJ <
*amaxiya-pdda- 'cart-path'.
Arta < *ArOya- 'pious', Avestan aiya- (< *arOya-) 'pious'.
Piipasri w. < *VCspasi- < *Baga-spasi- 'servant-girl of the LordIGod'; cf. VeSpasi.
Abuhola w. < *Abva-hora- 'giving abundant gift', Khotanese Saka abvdtta
'abundant', hora-, haura- 'gift'.
Kharaosta, Charahostei (in Greek script, genitive) < *XZra-hcista- 'splendid,
mighty', hosta- < *hauvasta-, Khotanese Saka hauvasta- 'powerful'.
Kamuia w. < *Kama-uya- 'who looks d o w n upon the lust', Khotanese Saka uy-
'to survey, look down upon'.
Khalamasa < * X ~ r a - m a s a h - 'splendid greatness', Khotanese Saka muse <
*masah 'greatness'.
Maja 'delightful', Khotanese Saka m i j a 'delightfuly.
Hayuara < *Hadabara- 'helper, companion' (literally: 'riding together'), Khota-
nese Saka bar- 'to ride'.
H a n a w. < *:Xana- 'smiling', Khotanese Saka khan- 'to laugh, smile'.
Ayasi (Kamuia) w. < "Azazi 'lady', Khotanese Saka aysim/e 'girl', eysija
'daughter of a minister'.
N a d a 'leader', Khotanese Saka nada- 'leader'.
Diaka < 8'Diydka- 'supervisor', Khotanese Saka dai-ldi- 'to see', inf. diyana-.
Nauluda < "'Nama-rauda- 'worship-heightening', Khotanese Saka nauda 'obei-
sance, worship' < *nama-ta-, rvuv- 'to grow' (< "raud-).
Kalui < PKaluka- 'noble', Khotanese Saka kala- 'excellcnt'.
Kusulua Patika cf. above.
Mevaki < *Mavya-ka- 'tiger, hero', Khotanese Saka mauya-, muyi 'tiger'.
Miyika < *Mayaka- 'prosperous', Khotanese Saka ggumai 'at will' < * v i - m i ~ ~ i -
'pleased'.
Khardaa < *Xrata-ka- 'wise', O l d Iranian xratu- 'wisdom', Khotanese Saka
grata 'instruction, advice' < : k ~ r a t a(replacing
- xratu-).
Khalaiamuia < 1:Xira-iima-ta- 'of shining face', Khotancse Saka *khara- 'in
kharava- 'shining'), is'aman- 'face'.
Muki < 8:Mava-ka- 'tiger, hero', Khotanese Saka mauya-, muyi 'tiger'.
 Languages and scripts in Gracco-Bacrria and the Saka kingdoms

The linguistic evidence for the Sakas of Malwa (Ujjayini) is again modest:

~ ~ ~ h a r<a *xiaOra-pati-
ta 'lord of the country'.
~ ~ a m o t i k<a*Zama-pati-ka- 'landlord', Khotanese Saka ysama- 'earthp.
~ i m a y s a d a< *Dimi-zata- 'begotten by the Creator', Avestan dimay- 'Crea-
tor', Khotanese Saka ysata- 'born'.
Nahapina 'protector of the clan', Khotanese Saka *nriha- 'navel' (genitive sin-
gular neha), -pdna 'keeper', pa- 'to protect'.
Cagana, Tiastanes (in Greek script) 'master', Khotanese Saka cage? (<
*ca![ana-) 'master as a proper name'.
UsavadZta- < *Riva-dita- 'rightly created', Avestan araiva- 'right9, Khotanese
Saka data- 'placed, established; law'.
Aduthuma < "Artava-tauxma- 'offspring of a righteous man', Khotanese Saka
ttima- (< "tauxman-) 'seed'.

Scanty as it is, this evidence does permit us to form some idea about the lan-
guage of the Saka groups settled o n the territory between Sistan and Malwa.
The language of the Sakas of Sistan was apparently influenced by Parthian as the
borrowing of the Parthian term spa& 'army' testifies. During the reigns of Azes
I, Azilises and Azes I1 the power and influence of this Saka group extended
eastwards as far as Taxila, and there were obviously some linguistic contacts be-
tween the Saka groups of Sistan and Gandhira. Notwithstanding, the diffe-
rences between them are also apparent. Such terms as murunda, ksaharita, hiye,
veS;basika, horamurta, horamurndaga, casrana, etc. never appear t o the west of
Gandhira.
O n the other hand, from the evidence of the names and terms used on
the territory of both Gandhira and Mathura, the close ties between the Saka
groups of these two countries are evident. Even though linguistic evidence for
the Sakas of Malwa is very limited, it nevertheless becomes clear that they were
related to the Saka groups of Gandhira and Mathura. It follows that even if the
four Saka groups of Sistan, Gandhara, Mathura and Malwa were in contact with
each other, they represented two different Saka dialects - the dialect of Sistan
and that of G a n d h ~ r a Mathura
, and Malwa. Established on the basis of names
and terms, this division corresponds exactly to the distribution of the titles
Vauka- (in Saka rauka) and murunda- (in S a k i mrunda) and probably reflects
two waves of Saka immigration, proceeding by two different routes t o the
south.
Similarly significant is the comparison of Southern Saka vocabular!- with
northern, Khotanese Saka (S = Sistan, G = Gandhira, M = Mathura, U = Ujja!.-
ini and Malwa):
 Southern Saka Khotanese Saka
achia (4:Zayaka-)'observant, rev- acya-vamda 'observant, revere~~t'
erent'
abu (< *abva-) 'abundant' abvatta 'abundant'
amaZa 'chariot' mai- (< * a m a ~ y a -'chariot'
)
aya (< *aza-) 'leader' *-aza (Kharosthi -ajha) 'leader'
ayasi (< *azazi) 'lady' ~lysamje,eysGa- 'noble girl'
ayiliia (< *aziliia-) 'commander- -

in-chief'
ara (< *dra) 'noble' ara 'worthy, noble'
arta (< *arOya-) 'pious' -

ddu- (< *artava-) 'righteous'

-uia 'looking down upon' uy- 'to look down upon'
ulina 'high-born' ula 'up'
u ~ a v a -'right'
*-hosta 'mighty' hauvasta 'powerful'
k a m - (< "kdma-) 'lust' -

kalui (< *kaluka-) 'noble' kala- 'excellent'

kusuluka, kusulua (*kuzuluka) kUYs- 'to seek, search'
'energetic'
GMU k ~ a h a r -(*xiahr-) 'country' k ~ a r a -'power, dominion'
GMU k~ahardta-(< *xSldhra-vata-) 'lord -

of the country'
M khara- (*xdra-) 'splendid' khdrava- 'shining'
M khardaa- (*xrataka-) 'wise' grata (*xrata-) 'wisdom'
M khala- ("xala-) 'splendid' khdrdva- 'shining'
G -gula 'progeny' -kula 'progeny'
S -gadama (l'kitama-) 'commander'
cagana 'master' cagem 'master'
-jada (#:zdta-)'begotten' ysdta- 'born'
jihonika 'benefactor' jeh-, jib-, jehuna- 'to heal'
Ijhamda (:!:zdta-)'born' ysdta- 'born'
2jhamda- (*zata-) 'birth' ysd- (< 4:ysdta-)'birth'
-thuma (#:thuma-< tauxma-) ttlma- (< "tumn- < *tauxrna-)
'offspring' 'seed'
dami- 'Creator' -

-data 'created' data- 'established'

diima- 'Creator' -

diaka 'supervisor' dai-ldi-, diyina- 'to see'

deni- 'religion'
nada 'leader' nada- 'leader'
nama ('\nama) 'name' nama 'name'
 Languages and scripts in Graeco-Raclria and tho Saka kingdoms

naha- ("niha-) 'navel, clan' *niha- 'navel'

nau- 'worship' nau-da 'obcisance, worship'
-pa (< * p a v i ) 'protector' pa- 'to protect'
patika- 'leader' pata, patana 'before, in front of'
-pallina, -balun0 ("valana-) valaka 'young'
'youth'
-pinu 'protector' -pins 'keeper'; pa- 'to protect'
maja ( * m i j a ) 'delightful' mija 'delightful'
mani (< "manya-) 'delightful' minya 'delighted'
mava- 'tiger, hero' mauya- 'tiger'
-masa 'greatness' mase 'greatness'
muki (< *mava-ka-) 'tiger, hero' muyi (< *mavya-) 'tiger'
murunda, murndaga 'lord, king' rrund- 'lord, king'
-murta (< " m r a u t i ) 'lord, ruler' rautci (< "'mrauta-ta-una-)'corn-
mandant, ruler'
moga, moa (< *mavaka-), 'tiger, muyi (< *mavya-) 'tiger'
hero'
miyika (< *mayaka-) 'prosperous' ggu-mai 'at will'
mevaki (< "mavyaka-) 'tiger, muyi (< *mavya-) 'tiger'
hero'
-ysada (< * z i t a - ) 'begotten; cf. G ysata 'born'
-jada'
ysam- 'earth' ysama- 'earth'
raj- (5-Zza-)'ruling, ruler' rraysaa- 'official title'
-riia- (*riia-) 'commanding' rrii- 'to desire'
-rauka- 'lord, ruler' rukya- 'commander, lord'
liaka 'youth' rya 'youth'
-1iia (*liia-) 'commanding' rris- 'to desire'
-1uda (*luda- < "rauda-) 'height- rruv- 'to grow'
ening'
-vula (< *vara-) 'ruler, king' vara- 'excellent, string'
ve- (< *vaya- "bags-) 'lord, god' v v i v a - (< "vava- < *6aga-) 'lord'
-famu- (< " i i m a - ) 'face' iiiman- 'face'
iud-liod- 'goodness' iiade 'goodness'
ipalagadama 'commander of -

army'
ipalahora 'commander of army'
ipaliriia 'in command of army' -
~ a d a'rejoicing, rich' tsita- 'rich'
-spasika 'servant' spaia 'observer'
haga- 'forward' ha (< " h a k ) 'forward'
hagina- 'leader' -

hana (*xana) 's~niling' khan- 'to laugh, smile'

 hayuara 'helper, companion' -
hiye 'master' hiye 'master'
'hora- 'gift' hora-, haura- 'giftJ
Jhora- 'lord, commander' haura- 'commander'
horaka 'lord of gifts' haurdka 'donor, giving'
horamurndaga 'lord of gifts' -
horamurta 'lord of gifts' -
-bola 'gift' hora- 'gift'

O n the basis of this comparison it becomes clear that the Saka groups of Sisran,
Gandhiira, Mathura and Malwa o n the one hand, and the population of Khotan
(and TumSuq) o n the other hand, spoke closely related dialects of the same Ian-
guage. The Southern Sakas preserved the common Saka social terminology well:
thuma 'offspring', hora 'lord', -gala 'progeny', ndha 'clan', nada 'leader', *aZaZi
'lady', hiye 'lord', raza 'ruler', CaStana 'master', kalui 'noble', aza 'leader',
murunda. . 'lord, king', rauka- 'lord, ruler', murta 'lord', xiahr 'country' and vr-
'lord, god'. It is a remarkable fact that they also maintained the ancient religious
vocabulary as, for example, arta 'pious', a d u 'righteous', u ~ a v a'right', dam,
'Creator', den 'religion', we 'god', which was replaced by Buddhist terminology
in Khotanese Saka. Other archaic features in Gandhiira and Mathura Saka may
be the retaining of initial mr- in contrast t o the Saka dialects of Sistan and Kho-
tan where initial m of this consonant cluster disappeared as well as the develop-
ment -dux- > - i - , while in Khotanese Saka a further shift u > i took place.
However, there are also some illnovations in sound system and vocabu-
lary of the Southern Saka dialects. They borrowed some important terms like
spala 'army', spasa 'servant' and ~ a d a'happy' from Parthian or Bactrian and
created a new terminology to denote the ruling aristocracy as, for example,
kSaharbta, hayuara, horaka, horamurta, h o r a m u ~ n d a ~ aziliia,
a, spalahora, spa-
lagadama, spaliriia. A further striking heno omen on is the change r > 1
which often occurs (cf. -luda, liaka, -liia, -bola, khala-, -vula). Perhaps the
strengthening of the role of 1 in the ~ h o n e m i csystem of Southern Saka is due to
the influence of Bactrian and Indian languages in which the hone me 1 was
rather frequent. Thus, however scanty they may be, the linguistic remains of the
Sakas settled in Sistan, Gandhara, Mathura and Malwa furnish valuable
information about the Southern Saka dialects and their relations to Khotanese
Saka.
 J. Harmatta

An unknown language in an unknown script

Since 1954 a striking series of linguistic documents written in an unknown lan-
guage and in an unknown script have come to light in the territory of Central
Asia of the Graeco-Bactrian and the Kushan periods. The following documents
are known:
1. Surkh Kotaf, three lines, written with black ink on a small fragment of stone.
2. Dasht-i Nawur, stone inscription, nine lines.
3. K h a l ~ h a ~ a one
n , inscription on a potsherd, another on a tile.
4. Kara-tepe, three fragmentary inscriptions on potsherds.
5. Ay Khanum, inscription o n a silver ingot.
6. Issik (50 km t o the east of Alma Ata), inscription on a silver cup.
7. Khatin-Rabat (in southern Tajikistan), fragmentary inscription on a
potsherd.
8. Tekkuz-tepe (in southern Tajikistan), inscription o n a potsherd, unpub-
lished.
9. Old Merv, inscription(s?) o n a potsherd, unpublished.
10. Fayaz-tepe (near Termez), several inscriptions on earthenware, unpub-
lished.
1 1 . Kafirnzgan-tepe (40 kin to the south of Dushanbe), fragment of a wall
inscription (?), unpublished.
Consequently, the spread of this unknown script and language covers a vast ter-
ritory fro111 Alma Ata up to Mew, Dasht-i Nawur and Ay Khanum.
There have been speculations about the character and ethnic background
of the script, but only one suggestion really deserves consideration - the theory
that the script goes back t o the Kharos!hi alphabet and the language written in

:' See Map 4.

this script may be a Saka dialect, perhaps also spoken by the Kushans. In fact, in
spite of the similarity of several letters to the characters of the Orkhon-yeniscv
Tiirk runic script, it is clear that the number and shape of the letters, the
of vowel mitris and the presence of compound akgaras prove without anp
doubt the Kharoglhi origin of the alphabet. The coincidence of some aksara;
with runic characters is restricted t o the cases where the Aramaic prototypesof
both the Kharosthi and the Sogdian letters (the latter serving as models for the
Tiirk runic signs) were similar.
If we tentatively substitute the syllabic values of the KharoZlhialphabet,
the resulting text has a Saka character. So one of the two inscriptions from
Khalchayan, containing only one compound akgara, can be read as lya, This
reading can be interpreted as a personal name and compared to the well-known
Saka name Liaka (cf. Khotanese Saka rya 'young'). The reading of the other
inscription from Khalchayan is more uncertain because it is not clear whether it
is to be read in the position given by the publication o r upside down. In the first
case, its reading may be jha-yi-ka (i.e. *Zayika, a name to be compared with the
Middle Iranian name Zik); in the second, it can be read as ya(m)-1 mi(m)-pa(?)
(i.e. *Zamipa, similarly a name, representing the same type as Denipa). Both
names could be, however, equally of Saka origin.
O n e fragment from Kara-tepe can be read as ]ii(m)-mii[ and connected
with Khotanese Saka iimja 'the thorny jujube' used for preparing juice in Kho-
tan. The other fragment from Kara-tepe may be read as ]nd(m)-sd(m) ksa[, i.e.
]ndsd kZa[ ']portion six[' and ndsd may be the same word as Khotanese Saka
nasd- 'portion', while ksa[ can be compared to Khotanese Saka k ~ a k~asg , 'six'.
N o r is the fragmentary text from Khatin Rabat longer: e I yo[sa 'whole [is] I
musk', e being equal t o Khotanese Saka i (one, whole), and the spelling yo[sa
representing the same word as Khotanese Saka yausa 'musk'.
The texts of the inscriptions from Dasht-i Nawur and Surkh ~ o t a are l
rather long and reading them presents great difficulties because of their being
poorly preserved. Line 1 of the inscription of Dasht-i Nawur ( D N 111) can ten-
tatively be read as follows: sa-[lil mi pam-ja-sa [bra]-ka-iim mi ma-ste +am-
juJ-sa [ha]-& 'The year [is] now 50, BrakaSi [is] now the month, 15 days'. To
illustrate the character of the language, we may compare the same text in ~ h o t a -
nese Saka (in Brihmi orthography) with it:

Dasht-i Nawur: sali mi pamjasa byakaiip mi maste parfllusa ha&

Khotanese Saka: sali m i p a m j s i s i byamkhaysji m i miisti pamjsisa hnda.

The similarity is obvious and if the proposed reading of the date proves to be
correct, it follows that the Southern Sakas (or the Kushans) had a knowledge of
the month names used also in Khotan and of the time reckoning by cycles of
sixty years o r by another era, different from the one used in the Bactrian
inscription ( D N I) of Dasht-i Nawur.
 Languages and literature in the Kushan Empirc

The text of lines 2-9 of the inscription DN 111 runs as follows:

2. ye rva-da-ti ri a-ba]-ti vi(m)-ja-rka ka-ltvi-sa] [ k ~ ] - ~ a - ~ a

3. m i mri pa(m)-ra-mmi-na sta-nam pa(m)-ri-vim ii-da va-[ril k a m h i m
4. sa gra-vam ti-rma da-bha sa-di pa ka(m)-pi-sa(m) ia-di-ria
5. ha-rnri(m)-ja kam-ld'a vam-yi-riim k a m - j u - v i m ii-ksa-ii dha-kam
6. jham-sam ka-[dla ta-rma pa a-jam n a m - v i m ha-mri-ka sa-na ii-jha
7. mri-kam ii kam-vul-vam mi-[ita ha-ra]-lstal ha-mi ha-mi ha-ya-da ja-sta ha-sa
8. he-ko mri(m)-ka mi ho-kam jyom pa-pam-sa vam-ta ham-mi-ga-ia
9. mla ka-ria e-ii ham-da-Idalpam-mripu-[daltam-ka u-da Ida-ri jal-rmi ljal-sta l a .

On the basis of the far-reaching agreement of the language of this inscription

with Khotanese Saka and with the aid of its Bactrian version (see later) its text
can be interpreted in the following way:

2. Behold! w e ] King of Kings, the noble, great Katvisa, the Kusina,

3. now, here, w e order to erect the commanded text for the welfare as heroic
words:
4. H e [Katvisa] mounted o n the mountains, [he] was able to cross the high region.
H e inspected KZpiSa.
5. [He] put relief to [his] advancing domestics, moved forward [his] forces,
6. fought a battle, crossed the region, pursued, captured the crushed Sanas [=
Avestan Saini-1, destroyed [them].
7. Graciously he rested [his] servants, he offe[red] pres[ents] to all of them. H e
celebrated a feast for the god,
8. being devoted and gracious. Then he held feastings for the officers and the war-
riors altogether.
9. H e ordered to engrave on the rock the favourable report [that] he removed the
tax and contribution from [the sanctuary of] the supreme god.

The content of this inscription coincides in all essential details with that of the
Bactrian version (discussed below) of the epigraphic monument at Dasht-i
Nawur. However, a remarkable phenomenon is that the relation of this inscrip-
tion is much more detailed than the Bactrian text. Obviously, the most import-
ant version of the report about the campaign led by Vima Kadphises t o the
region of Dasht-i Nawur was represented precisely by this text. From the
repeated mentions of the domestics, their rewards, and the festive banquet given
in honour of the officers and warriors, it follows that this was the language spo-
ken and understood in the royal court of Vima Kadphises and among his reti-
nue and army, whether this was some Saka dialect adopted by the Kushans or
the original language of the Kushans themselves. The central position and the
detailed text of this inscription clearly speak in favour of the latter assumption.
Another interesting document, written in the same language and with the
same script, is represented by the inscription from Surkh Kotal. The character
of the record is striking. It was written in black ink on a stone fragment, meas-
 J. Harrnatta

uring 22.5 x I 1 x 4.9 cm. This fact excludes the possibility of an official doc"-
lnent and renders the assumption of an occasional record probable. The text of
the inscription, also coming very likely from the Kushan age, can tentatively b,
read as follows:

1, hi-yo e-se h o ta-na: mva-ra ha-mu-di a-ja hi-rya pa-ii da-pa va-rya
2. k a - v i - g y o ja-rya da-ja ho-la cha-da gyo-rmi va-gyo dha-na cha-ka mo-ia
3. v a - h i da-hu dam-na.

Contrary to the inscription of Dasht-i Nawur, here we have no support for the
understanding of this text. In spite of this apparent difficulty, however, the
interpretation is not impossible because some terms and phrases can clearly be
identified again with the aid of Khotanese Saka. Thus, the inscription can be
interpreted in the following way:

1. The lord gives orders so: The procedure happened. It is possible to release the
non-perished wealth: the mantle,
2. the coat of mail, the armour, the flame[-coloured] covering, the miler excellent
racehorse, the grain, the goat will you quickly carry away!
3. The house is given to the man [or to Dahu].

This text obviously represents a report on a judgement about the division of

property either in the case of divorce o r by way of inheritance: one party
obtained the movable wealth (the things enumerated in the report), the other
one kept the immovable property (the house). This report was apparently sent
by a person w h o belonged to the retinue of the 'lord' exercising the jurisdiction
and who was personally acquainted with at least one of the parties. The use of
the stone fragment for the purpose of this information is probably due to the
lack of other writing materials in Surkh Kotal at that time.
All the records written in this variant of Kharosrhi script and Saka
language discussed so far date back to the Kushan age. T w o inscriptions of this
type, however, represent an earlier period. The inscription of Ay Khanum,
engraved o n a silver ingot, comes probably from the second half of the second
century B.c., while the inscribed silver cup from Issik was dated to the sixth-
fourth centuries B.C. Nevertheless, there can be hardly any doubt that the ]at-
ter dating is too early. Taking into consideration the fact that the inscription
from Issik cannot be separated from other inscriptions of this type and that it
clearly presents the characteristics of the Kharoprhi script, it cannot be dated
before the second half o r the end of the third century B.C. In any case, these
two inscriptions present more archaic, more angular, simpler letter forms than
the other.
Even though some of these features may be ascribed to the writillg tech-
nique (engraving), they still indicate an earlier date.
 Languages and literature in the Kushan Empire

The text of the silver ingot from Ay Khanum can bc read as follows:

a-l-za-t~mi-pa-zam-na pa-ya a-mi-zam-na pe I pa-ya-di-na I [ . . .

Silver: smelt sort, mixed, greenish [?] I examined I [weight . . .

The text is probably incomplete as the end of the record is broken off.
The term alzato (silver) exactly coincides with Khotanese Saka iljrata-
(silver) but except amizamna (< Old Iranian *imaiEa-nu-, Middle Persian
amextan 'to mix' all words o r stems also occur in Khotanese Saka.
The inscription on the silver cup from Issik can tentatively be transcribed
again in the following way:

I. za(m)-ri ko-la(m) mi(m)-vam vam-va pa-zam pa-na de-ka mi(m)-ri-to

The vessel sliould hold wine of grapes, added cooked food, so much, to the n ~ o r ~ a l ,

2. iia-ka mi pa-zam vam-va va-za(m)-na vam

then added cooked fresh butter on.

The vocabulary of this inscription, too, has quite exact parallels in Khotanese
Saka: za(m)ri 'vessel' - Khotanese Saka jsard 'receptacle', kola 'grapes' - Khota-
nese Saka k i r a 'grapes', Vedic kola 'jujube', mi(m)va- 'wine' - Khotanese Saka
meva, miya- 'intoxicant drink', vamva 'added' (< *ava-nava-) - Khotanese
Saka pu?zvdna- 'to be inserted' (< *pati-nava-nya-), pazam 'cooked' - Khota-
nese Saka pajs- 'to cook', pa<m>na 'food' - Khotanese Saka pamna- 'food',
deka 'so much' - Khotanese Saka deka 'so much', mi(m)rita 'mortal' (< " m ~ y -
ata-) - Khotanese Saka mar- (< *myya-) 'to die', Gaka 'fresh butter' - Khota-
nese Saka niyaka- 'fresh butter', mi 'then, now' - Khotanese Saka mi 'now,
then', vaz- 'to hold' - Khotanese Saka vaj-lvij- 'to hold', va(m) 'to, on, for' -
Khotanese Saka va 'for'.
O n the basis of these texts and of the close ~arallelsbetween them and
Khotanese Saka linguistic data, it is easy to recognize the close relationship of
the two languages. In spite of some uncertainties in the reading and inter-
pretation of these texts, written in a variant of the Kharoghi script, there can be
hardly any doubt about the essential features of their language. They clearly
represent a language of Saka type with some ~ e c u l i a rfeatures. The question
remains, however, whether the language of these texts was a Southern Saka dia-
lect also adopted f o r their chancelleries by the Kushans o r whether it represents
the original language of the Kushans, which was closely related to the Saka
dialects.
The Bactrian language in Greek script
The importance of Hellenism in Central Asia may be best illustrated by the fact
that the Greek alphabet was adopted to write the Bactrian language. Earlier,ir
was generally assumed that Bactrian literacy came into existence under
Kushan king Kanishka 1, because it was under his rule that the Kushan mintr
struck coins with partly Greek and partly Bactrian legends, written using Greek
characters. In 1967, however, a trilingual inscription of Vima Kadphises was
discovered at Dasht-i Nawur, one version of which was written in the Bactrian
language using the Greek alphabet. It thus became clear that Bactrian literaci
dates back t o the time of Vima Kadphises o r perhaps even earlier.
The Greek alphabet of Bactria was adapted with its contemporary sound
values to the phonetic system of Bactrian. Thus, the Greek spellings EL and ou
were used to denote Bactrian t and ii respectively. The differences between the
Greek and Bactrian phonetic systems, however, necessitated some changes in
the sound values of the Greek letters, for example sigma (a, S ) denoted beside 5
also i and Greek zeta (5) had the sound values z and .; In Bactrian Greek the
consonant cluster ks became assimilated to ss, s. Consequently, the Greek letter
xi (5)was not suitable t o represent Bactrian xi. Therefore, the consonant khi (x)
and the newly created s (1)) were introduced to denote this Bactrian consonant
cluster. A striking peculiarity of the Bactrian alphabet is the new sign 1) for Bac-
trian $ and the use of Greek upsilon (w) for Bactrian h.
Bactrian writing was widely used throughout the Kushan Empire both for
official purposes and for everyday life. Accordingly, there are several types of
records in Bactrian writing: (a) stone inscriptions; (b) wall inscriptions; (c)
inscriptions o n objects; (d) coin legends; and (e) seal inscriptions. The most
important sites of Bactrian inscriptions are: (a) Surkh Kotal with six stone
inscriptions; (b) Kara-tepe with inscriptions on ~ o t s h e r d s(the short wall ..
inscriptions, numbering about 100, were scratched on the corridor walls by VISI-
tors to the sanctuary during the Sasanian age; (c) Dasht-i Nawur with two stone
inscriptions; (d) Dilberjin with two stone inscriptions from the ~ u s h a n~eriod
(some wall inscriptions and ostraca are of post-Kushan date); and (e) Ayrtam
with one stone inscription.
According to the evidence of the Bactrian inscriptions known so far, it
was the Kushan king Vima K a d ~ h i s e swho first had monumental rock or stone
inscriptions prepared. Of his inscriptions, that of Dasht-i Nawur (DN I) seems
to be the earliest. Consisting of thirteen lines containing 246 letters, the inscrlp-
tion was engraved o n a rock at a height of 4,320 ~nin the mountains. Its text call
be read and interpreted in the following way:

1. z00 ~oQ7ClCtl0~
LE
[Ern-ycar] 279, 15th [day o f t h c m o n t h ] C ; O I . ~ I , I I O \ .
 Languages and literature in the Kushan f:'mpive

2. I)uovavo I)uo i c ~ t u b o
King of Kings, the noble,

3. oul;o~xoOoqpo T a x n ~ o o
great O o e m o Takpiso,

4. xobuvo L ~ U U O ~ , L V LL ~hubo-
O
the Kusina protCgC of the moon [god], the right-

5. yo L ljayo ol;ohabo el60

eous, the Majesty had this prepared,

6. xoGoyaQyo upo Laxqao

he, the benefactor for the welfare.

7. Ooqpo bao aoo Av61lco uto

King O o e m o came both here from

8. paho aya60 atqo Caviye

Andezo and the Sanigos

9. vopo~6av60060 paho
were destroyed by him. And here

10. cpgopabo Av6qco JCOQ~O

he ordered: 'Be the tax paid by Andezo

11. Poo~liox~ljboaBo ljuyuvo

its o w n for the sanctuary

12. 060 ~al;a6oi xaeLoueo uljo apelyo

and the warlike divinity for ever!'

13. a t o otavo poho xoav6o

For that hecausc he was called b!. them herc.

As can be seen, the content of the Bactrian inscription ( D N I) agrees with the
Kushan version ( D N 111) discussed above in all essential ~ o i n t sThe
. epigraphic
record was to commemorate the crossing of the high mountains and
the victory won by Vima K a d ~ h i s e swhen he came from Andezo (Qunduz)
over the Siinis (Sanige in the Bactrian text, Sana in the Kushan version) dwell-
ing in the region. The dare of the inscription is -ear] 279, 15th [day of the
month] Gorpiaios'. Very likely, the era concerned is the so-called Graeco-Bac-
trian or Eucratides era, beginning with the accessioll of Eucratides about 170
B.C. The last epigraphic record of Vima K a d ~ h i s e sdating from the same era is
 1. Harmatta

the unfinished inscription of Surkh Kotal (discussed below) from Year 299
Obviously, this inscription was left unfinished because of the death of Vima, so
that Year 299 may correspond to the year before the accession of Kanishka.
Accordingly, the date of DH I would approximately correspond to a day in
September A.D. 113 and the accession of Eucratides would be in 166 n.c.
The date of the Kushan inscription of Dasht-i Nawur ( D N 111) is consist-
tent with this: like Gorpiaios, Brakaii is an autumn month and if Year 50 rep-
resents the fiftieth year of a sixty-year cycle, it would fall in A.D. 113 according
to the Chinese sixty-year cycle time-reckoning and in A.D. 117 according to the
Indian one. The former conversion exactly corresponds with the date of the
Bactrian inscription DN I. Hence, the Kushans probably became acquainted
with the Chinese sixty-year cycle while they were still in their ancient horne in
Gansu.
The other Bactrian inscription of Dasht-i Nawur is hardly legible and is
still t o be deciphered, but all five inscriptions of this site were probablv
engraved at the same time and can be ascribed to Vima Kad~hises.
At the Dilberjin site several epigraphic fragments were found which
belong to two inscriptions. Their texts are rather fragmentary: in inscription 1
only one complete word has been reserved, while in inscription 2 no complete
sentence can be found. In spite of the fragmentary state of both inscriptions,
their texts can tentatively be restored and their contents roughly understood.
The name of Vima can probably be recognized in both records.
Consisting of at least ten lines and of 200-220 letters, the tentatively res-
tored text of inscription D 1 runs as follows:

l ..[. . . . . . . . . . . . . . . . . . . . -1
[Era-ycar . . . , . . . [day of month] . . .]

2. [l)aovuvo 1)ao L acubo]

[King of Kings, the noblc,]

3. [oacogno Ooilll~iol[Tanniuo]
[great O o e l ~ n o[Takpiso,]

4. [nol)ovo i] h u b ~ [ ~ yi opayo]
[the Kusina, the] right[cous, the Lord]

5. [ E L ~niboya~o]
O uc~ybo[apo Orlbo]
[had this imagc] prcparcd [to OesoJ

6. [obo cy~opuboi]Ou u[-cl[uvoxibo]

[and hc ordercdl thu\ that [by thcln who]
 Languages and literature in the Kushan Empire

8. [x~bopuho] V ~ X ~ E ~ L JPayo]
[QO~~L~O
[who is here master of] the hunt [, care]

9. [hayyo n~bo~~llxolqo
ob[o JCOQOOUQ]
[should be] taken [for the sanctuary] and

10. [oqo nlbo L ~ ] [ o l ~ y o[bo

l o J L hqvo]
[the cult should be performed according to the] rite [and the religion].

The inscription was discovered in the sanctuary lying in the north-eastern cor-
ner of the Dilberjin fortress and decorated with a wall-painting representing
Siva and Parvati. The wall-painting was prepared in the reign of Vima Kad-
phises.
The other inscription from Dilberjin consists of at least twenty-four lines
comprising about fifty letters each. Thus, it must have had altogether about
1,200 letters and represented the most considerable Bactrian epigraphic text
known so far. Unfortunately, in the three fragments discovered only 442 letters,
that is, about a third of the original text, have been preserved. Happily, import-
ant terms such as cpago, a(3[o], oab[o], ape L wgao[vo] and [wga]ovo po i a(3yo
'abundant water', 'well', 'waterflow' clearly reveal the main topics of the
inscription: the water supply of the Dilberjin stronghold and sanctuary. It
seems that the stronghold was at first provided with water from a source lying
outside the walls where later a sardoba was built. When the water of the source
began to fail, a well was dug in the bastion flanking the gate and the use of the
water was strictly regulated. These and other measures were apparently taken
by order of King Vima Kadphises. In view of the rather fragmentary state of the
inscription, its text can only partly and tentatively be restored.
The conjecturally completed text of the inscription runs as follows:

1. [xl)ovo . . . . . . . . . . . . . Payo baovavo bao i aLabo oaLo~xo]

[Era-year . . . , . . . [day of the month] . . . King of Kings, thc noble, great]

2. [Ooillpo [Tal[xn~oo
xo1)avo I ~lauot~vtyo
I haboyo ~ t b o
(3uyohayyol
lOoelmlo Tafkpiso, the K u s i ~ ~proreg6
a, of the moon [god], rhe lord dedicated this sancruar!.]

3. apo O~ll,o[L~ O Q ~ C I O U V~~aOt a b o. . . . . . . . .I

to OSso, [the exalted divinity . . . . . . . . . . . . . . . . . . . .I

4. oLona clva[ygo . . . . . . . . . . . . . . .I
thc eter[nal lord of] the univcrsc [. . . . . . . . . . . . . . . . . . . . . ..]

5. oLono o a [ o ~ a ~hubo
o ~a xclhbo ~ ~ c t hqqoyt~bo
l~o ~clbilioxaeavo 01
Ma[stel. of] all beings. [At that timc, nrhcn tllc fortrcss \\,as conlplctcd, thcre was no pure]
 6. bo y,aeo aP[o vLmo xoro T * ~ Laoo 6~aAlLoapaflyo (PQOXOQTO 0 1 ~ 1 1 - 1
alld a b u ~ l d . ~ nwater
t [ ~ na t o drink. Then, the god OC!o wanted to leave thc watcrlcss funrnLl

7. o Payo orlb~apl[o payohayyo aoo avo xalvo U [ ~ oaoruqio

O tab( U O D ]
I n order [to conduct the water from the old sprling t o [the sanctuary, then]

8. 1~~51lvI11
1 ) a o ~ o[aeL~ye
o obo x ~ e w yoaortlvbo
~ xa[h6~po Lao Oo~lyo]
[from] the land Ujjayini ~ [ o r k e r sand artisans] were led here. When [King Ooi.mo]

9. [To~]1polbav~
a[paepayyo paho 5~60rablqo oab[o po av6aeo yjeooueo]
[sent Toxlrnodane as ~ ~ [ p c r i n t e n d e here,
nt then] he [had] a well [dug in the bastion]

10. [xavbo] 060 t [ ~ b po

o weaovo i apyo aoo] avo avo apo pah~t;oLOU art
and [he had tlie running water] con[ducted from the old spring t o the fortress so that]

[the abundant and pure water should not be missing in the fortress and then the god] Oqso
should]

12. [aoo payohayyo pa cpeoxoabq~o060 x a h 6 ~ELQO pa nalI6qI0 a[ti X ~ U - ]

[not want to leave the sanctuary and even when the watcrflow] would [not be streamling, [thcn
from the well pure]

13. [vo 060 rpago apo aoo oa6o apo pah&o (30oq~lowh6a a n cp~ag[aovo]
[and abundant water shall be for the sanctuary] there. But the right[eous]

14. [bao Ooqpo wcav6o o ~ xaeavo

6 ~ apo] lolPo6eayya n160 avo avo tub-]
[King Ooerno learned that the pure water] is scanty in the old sp[ring. Therefore,]

15. [qio A~iallyol[yallhol a[(3a~pay]yo

haso tab1 ay60 aPae[pavbo a(30 aubo]
lie appointed [Liialgo to ~ u [ ~ e r i n t c n d ] e n[he]rc.
t H c received tlie supervisory [authority over thc
well]

16. [obo ~ a v o ] l o ~ t Lqpoe q o a ~ oPo&0 1 ELQO oaeilh~( ~ o I a ~ ~ ] [ o vrbai b

1 ~~ 0 A-]
[and the spring so] that it should be his decision that thc donlestics of the fortress [should] cover
the drinking water.

17. [a60 a o ~ biOa

] ~ crl. ELO pavo Ko(3~~eilo
A ~ ~ a yaho
o [ ~ L ~ O Q L X (~X TI L~ha-]
I O
[Then it was also ordercd] s o that Liiago sliould continually [take care] for tlie Kubcrean house.
[Then King]

18. [o Oo~lplohaobqo L ahoyba habo U L ~1"L aoo uc~'Qbo[ycl aho pooqio]

] ~ tlic verbal Instructloll that 'From my possegslons water-condutt [never should be
[ O O ~ I I Igave
made!] Because o t h c r w ~ s c

19. [oea]ovo110 L aPyo tab1 pa aho avo

E L ~ O aPo i COQUOIVIO [at1 ( J C X Y O L V O ~ ~ ~ - 1
this never will be a watcr-flow!' [Tlicn to pric\t]
 Languages and literature in the Kushan Empire

20. [bo T ] o x ~ o ~ (hub0

I v ~ or1 ELILOxocxbqo VE xtht c~(.lulel[l~ct-r/o
obo 7 t ~ h ~ ~ ]
[~]oxrnodaniwas appointed. Thus it 1s our krng who exercise\ thc super[t,rs~onand] should [t.~kc
care] of us.

21. [~xolqorub1 1 puvo v ~ v b ~ e u ut oa ~ 10u

b ~ clyb~vb~
~ Q I ~ ' o ~urtrvo
[F i.cllu-]
Then the house was assigned and at that thcy obtained the dutics [ s o that they prec[cntcd

22. [o ha160 xahbi up0 ~o q)~yu-(yobuo Oo[qp]o [x~bt~(JCIQCI(IV~O]

OUT~LO
[a sift] when King O o e m o turns to the rnastcr [of thc mcrchants?]

23. [LOU]aybo utuvo v o n a ~ t eupo vupo UL ulfY[o] qqo<cl[po obo q~ol)oyl~bo]

[who] received [the privilege so] that the dutics of them are pledged for the cult which [shtruld
be] up to the end of timc and eternity.

24. [Po&]o ~ [ poloq~o

i O~lboooeo olanoouvavo Y , L ~po
[ L ]~ [ o u b q oVE]
Then be the chosen of O t s o , who is [our] k[ing], victorious over all!

In spite of its fragmentary state, the Bactrian inscription D2 of Dilberjin gives

us an interesting insight into the religious policy and the organizational work of
Virna Kadphises. The propagation of the Siva cult at Dilberjin and elsewhere
presupposes the conquest of the north-western part of the Indian subcontinent
by Vima, and this might have happened soon after his accession to the throne.
Similarly, the crossing of Mount Qarabayu rising to a height of 4,500 m and the
victory over the Siiinis as well as the preparation of the inscriptions at Dasht-i
Nawur could only take place after the campaign he had led into the Indian sub-
continent. The crossing of the high mountains is commemorated on his gold
coins with Siva and Nandi o n their reverse, that is, the event was preceded by
the spread and the propagation of the Siva cult in Bactria. Thus, the building
activity of Vima Kadphises at Dilberjin and the preparation of inscriptions D l
and D2 can be dated t o the period between A.D. 110 and 120.
It seems that the religious policy of Vima underwent some modification
towards the end of his reign. According t o the testimony of the so-called unfin-
ished inscription from Surkh Kotal (SK 2) he also extended his building activity
to that region but apparently his intention was to build a sanctuary for a Bac-
trian or Kushan deity there. The text of the unfinished inscription from Surkh
Kotal can be read in the following way:

xl)ovo ax0 6~01,le I~cro~vavol)ao IOoqll~o Tlaxln~oo 1(3luyo Ixiobovo ILeol1

v[oyov60 ~iuho]
Era-ycar 299, o n the 9th [cia!,] of [month] Dros. King of Kings C3oF1no Takpiso, the hlaicst!, the
Kusir?n, had thc carinl d[ug herc].

Very likely, Vima Kadphises died after the completion of the canal and before
the finishing of the inscription. Thus, he assured the water supplv for the build-
ing operations which were probably continued by his successor Kanishka with-
 J. Harmarta

out interruption. Therefore, the inscription witnessing the building activityof

Vima Kadphises at Surkh Kotal was never finished.
None of the Bactrian inscriptions set up during the reign of Kanishka
(Years 1-23 of the Kanishka era = A.D. 134-56) was preserved completely. A[
Surkh Kotal, the monumental wall inscription (SK 1) must have been prepared
at the time of the first Great Kushan king. Unfortunately, however, only on,
fifth of the whole inscription (124 letters altogether) was preserved.
~ u the
t fragments permit us to form an idea about the contents of this
important Bactrian record, which might originally have been composed of some
700 letters.
At the beginning of the inscription, the names and titles of the Kushan
king were probably mentioned:

](3ay[o Gaovavo b]ao 0(3[ooa~oKavqLxo . . .]

the lord, Ki[ng of Kings], the mi[ghty Kane,sko . . .] (Fragment 1 + b)

The context is not clear; perhaps the passage can be restored in the following
way: 'The lord, Ki[ng of Kings], the mi[ghty Kanesko, the Kusina, had this
stronghold built]'. Then, very likely, a date followed (Fragment k + t + v):

[nib0 L ~ ] o y o[xbovlo T[. . . . . . . . . . . . . . .] ~ i h oa[yabo.. .]

[in the] first [era yelar T [an officer of the king] c[ame] here.

Apparently, the next section of the inscription described the building of the
stronghold (Fragment m + c + g + a):

o t q [ ~ oE L ~ pah~Z,o
O obo (3ayohuyyo n~b]ooa(3[a~o]oa@h[o]avb~[L)to]
Then [this stronghold and the sanctuary] were built by him in four years.

It seems that further building operations were mentioned in the following pas-
sage (Fragment p + w + aa + u + s + y + q + n + j + x + f + r):

[obo X ] E ~IO~l[ahlS]o( P Q [ O ~ L~Qa ~bOq ~ELLO

o PO ~ ~ a ] b [obo]
to rcuy[bo I w]h~
a[uyoy~x ~ g b oot~lionlbo a o a y y ~heovo v a ~ o ] v y o~ [ L A L ~Q 0~ ~aOT I ~ xaelavo
L O
($0 [ X L ~
~OQ O V U] O
( ~ O[(~UYUVOV O ~ ) C ~ ](PQOOC(]OTO
~ C L [ O [ C L T ~( ~LO O(YO~(LYY ~O] O Q O -
[Yat01
[And] when the st[rongl~o]ldwas com[plctcd, then this fa]c;adc [and] thc stairs I[cading thlerc [were
built by him. Moreover, thc canal was wh]olly bulttresscd with ctoncs so that plurc water was [pro-
v ~ d l e d[by h ~ min the canlal for the ahlode of the T h u s he] to[ok care of thc sanctuary].

The last passage of the inscription obviously summarized the activity of the
royal officer o r of his attendants and gave information about the preparatioll of
the record. (The end of the inscription was preserved in situ):
 Languages and literature in the Kushan Empire

[oto ELLOPO PULLSO h e ~ u oSo-and-so sc~ebonLbo L ~otrbvoy , ~ o p c ~ voo ~ So-

o
and-SO]V O ~ ~ L X TPO
O FC(~)TOO U ~ EPO X U Y ~ O1 d~(IUY(IJYI

[Moreover, this strorlghold and the canal were built by So-and-so by the order of the kinK]. Thcn
So-and-so inscribed the fasade and the stairs leading there.

Thus, on the basis of the preserved fragments about three-fifths of the inscrip-
tion (altogether about 400 letters) can be restored, while Fragrncnts d, e, h, i, o,
z = 23 letters were not used for the restoration. The missing passages, consisting
of some 270 letters, might have mentioned the preparatory work and earlier
building operations of Vima Kadphises and perhaps the intended purpose of the
and the consecration of the sanctuary.
The third inscription of Surkh Kotal (SK 4) was prepared in three versions
(SK 4A, SK 4B and SK 4M; see Fig. 1) shortly after Year 31 of the Kanishka era,
probably under the joint rule of the Kushan kings Visishka, Kanishka I1 and
Huvishka, as Huvishka is already mentioned in Year 28 of the Kanishka era
while the two former kings are jointly attested in the inscription from Kamra
dated from Year 30 of the same era.
The three versions of the inscription differ from one another in both lan-
guage and content. Version A describes the earlier fate of the stronghold and the
arrival of Nokonzoko, the karalrango, who had a well dug to provide drinking
water for the stronghold. Besides this officer, nobody else is mentioned; even
the scribe and the mason, preparing the record, are only indicated by their per-
sonal devices (Device I and Device 2). The language of the inscription is correct
Bactrian.
Version B was prepared by another scribe and mason who are both indi-
cated by Device 3 and Device 4 and also mentioned by name - Liiago and
Adego - w h o can be regarded as Kushans o r Sakas on the basis of their names.
This version already mentions the name of the architect who dug the well.
Apart from this, the text of Version B coincides with that of Version A. From a
linguistic viewpoint, however, there is an important difference. In Version B,
some verbal forms, the particles, the relative pronouns and some nouns termi-
nate in -i instead of -0. This striking phenomenon cannot be explained by
orthographic variation o r instability because it only occurs in one and the same
Version B, while Version A and Version M offer no instances of it. In view of
the fact that the scribe and mason of Version B were probably of Kushan o r
Saka origin and in their language the outcome of Old Iranian -ah was -i instead
of -0 in Bactrian, this linguistic feature of SK 4B can ~robabl! be regarded as
the interference of the Kushan or Saka language. If, therefore, the term Kush-
ano-Bactrian or Sako-Bactrian had a real linguistic background, it could best be
applied t o the language of the inscription SK 4B.
'The reason for the prepration of Version B can only have been the lack
of any reference in Version A to the architect and to the order of the king hy
which he had the well dug. However, it seeins that further essential building
 J. Harmatta

FIG.1. Bactrian inscription SKM from Surkh KotaI.

operations were executed later on. Another architect, Xirgomano by name, had
the lower falade of the sanctuary built. T o commemorate this event, the scribe of
Version A, indicated by Device 2, and a third mason represented by Device 5,
were again commissioned t o prepare a new inscription -Version M. They copied
the text of Version A but added two passages, one mentioning the building of the
faFade by Xirgomano, the other indicating the names of the scribe and mason.
The text of SK 4 (A, B, M) runs:

1. (M) ~ 1 6 0~ C I A L L OPO Kavy1,no O C I Y L V

flayokayyo
~O 0160 [B: oibi] 1 bay0 ha0
[B: Gpyo] K a v ~ $ ) x[B:
~ Kav111~nI)ni]v a p o P a ~ y o~ 1 ~ [B:
6 0~ ~ ~ 6 1 1 .
This stronghold is the 'Kanesko' O a n i n d o sanctuary which the lord king ~ n a d cthe naniebearer of
Kanesko.
 Languages and literature in the Kushan Ernpirc

2. tabloo X E ~ O[A: xlbo, B: xebi] ~ o ~ b a p u

pakco
o cyyoyqho t a b y ~ op n v b a ~ o
arjo vioto [B: v ~ a t xoto
~ l aoibo [B: cxoibt] p(~hiSo[B: ~UALSU]uPaPyo 17tcxh0.
obo xahbo aoo heouo [B: heou] pivuvo L EQO otabo, tubo [B: t a h ~ ]L P U ~ F
aoo i vollahpo [B: la v~l~ahpo] cyeoxoetivbo [B: cyeo~ogrivbt]tabo clpo
Aeacpo oaot~vbo[B: oaot~vbl]aPo AvbqSo oro [B: ott] ~UALSO niboe~ybo[B:
x~bo~tyb~].
At that time when the stronghold was first completed, then its inner water to d r ~ n kwa5 rn~sring,
therefore the stronghold was without water. And when the water-flow d~sappcarcdfrom the
canal, then the gods wished themselves away from the abode. Then they were led to Lrafo,
[namely] t o Andezo. Afterwards the stronghold became abandoned.

3. ta xahbo [B: xahbi] NoxovSoxo [B: Noxovc~xo]L xaeaheuyyo L qeeLy,-

oabqayo x ~ 6 o[B: nib!] cygeLotaeo a(30bao t [A: !)avo] Payonou~o[B: payo-
nooeo] LAO [B: ah11 1 xopooago L btcoyaeyo [B: I)LSOYUQYE] L uhoby,aho [A:
ahaxbaho] xi60 [B: (paeo oioxoavo po oabo rja~yavooooybopayyo
n ~ b oi ioyo 060 v l ~ o o[A: loyo <o>bo, B: ~ o ~ y u o~ o o xeovo
] N~loavopa0 [A:
paw01 paho aya6o apo [B: a00 PO]rjayohayyo rabqio pah~conoeoyuTo [B:
no~ya[to]].tab1110 ELLO[B: &LO] oabo xavbo otqio [B: a t q ~ o arjo
] otooao~o
[A: aSooaoto, B: ~ o o a o t i o] t q ~ oxibo aoayye 100 [B: iOa] O L ~ L QaTavo
~ O crflo
paA~Socpaeo xaeavo a(3o pa yaoillo obo xahbavo aoo heovo [B: hgov]
pLvavo L &Leo(30011~0~ a b a v oL flay€ [A: (3[ay]o] aoo i vopahpo [B: L a vd)a).po]
pa ( P Q O X O ~ ~ O V ~[B:
I ~ (OP Q O X W ~ ) ~otavo
V ~ T ~pah~So
L O ] pa n t b o ~ t ~ o q i o .
Then, when Nokonzoko, the karalrango, the king's favourite w h o is most devoted towards the
king, the Son o f God, the patron, the benefactor, the merciful as wcll, w h o wishes glory, all-
winning strength from pure heart, came here t o the sanctuary in the 31st Era-year, in the month
NisZn, then he took care of the stronghold. Then he had a well dug, thus he ~ r o v i d e dwater.
Thereafter, he buttressed [the well] with stones so that the fine, pure water should not be missing
for the stronghold. And when for them the water-flow would disappear from the canal, even then
the gods should not wish themselves away from their abode, thus the stronghold should not
become abandoned b y them.

4. otqio aoaoxo po oabo axbrelyo x ~ e b oahrjaeyo o o ~ a b oLOO [AB: i(3a] aro

[B: at11 ~cib~ivo
[B: J C L ~ Eoabo
~V~ [B:O ~ L ~ E L
~ L] ~ E L V ~ XV~ L
' T]Q Lvaeovyo
~O [A:
<ua>~ovyo]ltahico xovto JcoeooaTo.
Moreover, he appointed an inspector over the nrcll, he placed a helper there, so that a separate [in-
spector] took good care of the well and a separate inspector of the whole stronghold.

5. oto ELLO po oabo 060 yalxo X ~ e y o ~ ~ aXvLoQ ~ Oapo BOQIO~IL~QO alto KO<-
yal)xLnoveo apo A o ~ i h o y a v o ~apo
t y ~ Noxov j t x ~xaeahgayye pagilyo ntdo t
xoa6qo qgopavo [A: -, B: O T E~ ~ Ooabo BO~SOCLLOQO x ~ b iK
, O~~~~XLJI[O]UQ
Yao~~hoyavi;~tyo,Noxovc~x~ xaeahgayyt paply1 xtbo xoabilo q~opavo].
Moreover, this well and the facade were made by Xirgo~iianoand Borzonlihro, the son of
Kozgasko, the c ~ t i z e no f Ast~logan,the attendant of Nokonzoko, the knralrango, by the order o f
the king. [B: Morcobcr, t h ~ swell was made by Borzomioro, son of Kozgasko, citizen of Hastilo-
gan, attendant of Nolconziko, the knvalvnngo, by the order of the k ~ n g . ]

6. oto Et~oyavoVOPLXTO all0 Mtuea~tuvoalto B O ~ ~ O ~ L ~ U Device

Q O ~ ~5,Oall-
~QO,
Lveap-xvo, Device 2 [A: Device 1 , alltoeapavo, Device 2, B: huayo, Device 3,
Abllyo Dcvice 41.
 Moreover, E i i o m a ~ l oinscrihcd [this] together wirh Mihramano, the son of Borzomihro I L ) ~ , 51
,~~~
jointly [Device 2 ) . ( A : Dcvice 1 jointly, Device 2, U: Liiago, Dcvice 3, Adego. Dcvicc 4).

In the historical context of inscription SK 4 of Surkh Kotal, the question ,nay bc

raised: Which of the Kushan kings is mentioned by the modest titles Payoljao
in this record? According to the testimony of the Kharosihi inscription froln
Kamra, in Year 30 of the Kanishka era, it was V ~ s i s h k awho bore among others
the titles maharaja r2jatirZja while his son Kanishka was probably styled only
w h a r i j a . Similarly, Huvishka only bore the title maharaja in Brihmi inscrip-
tions between Years 23 and 40 of the same era. Corresponding with the Brihmi
inscriptions, o n the inscription of Ayrtam, written in Bactrian and dated Year
30 of the Kanishka era (see below), he is styled P a o and Payo Pao which appar-
ently correspond to the title mahZrZja o n the one hand, and coincide with the
title Payo Lao used in inscription SK 4 of Surkh Kotal o n the other. Thus in
Year 31 of the Kanishka era (A.D. 164) three Kushan kings, namely Vasishka I
with the Indian titles mahiraja vdjdtiriija (- Bactrian Puyo baovuvo baa), Kan-
ishka I1 bearing the Indian title maharaja (- Bactrian payo kuo), and Huvishka
I with the same Indian title maharija and with the Bactrian title Payo l ~ ore- ,
spectively, were ruling. Obviously, the king styled Payo l ~ in~inscripion
o SK 4
of Surkh Kotal could only be either Kanishka I1 o r Huvishka (I). In view of the
fact that according to the text of the inscription 'the lord king made [the sanctu-
ary] name-bearer of Kanesko', it is ~ e r h a p smore likely that 'the lord king' was
Kanishka 11, w h o was able t o revive the cult of Oanindo/Victory in Surkh Kotal
with good reason after his victory over the Parthians about A.D. 162, attested by
the Sridhavmapi~akanidanasutra.
An important inscription in the Bactrian language was discovered in 1979
at Ayrtam, 18 km east of Termez o n the northern bank of the Amu Darya. The
inscription was engraved o n the front side of a square base of a monumental
relief representing the deities Farro and Ardoxso. Its text runs as follows:

1. [Lalo ooql)no L xl~oyoA xa[Ab]! L p[ebo]xeo cpaee[o x ~ b o y a ~pa)i[i

a ] L] pu[y]o
buo (3aybo obo oatab[o]
K ~ n g[IS] Ooe!ko, the Era-year [I?] 30 whcn the lord king presented and had thc Ardoxso-Farr~
image set u p hcre.

2. [ t a x]ah[b~]cpeoyt[eb]o paA\La otabo 1)obLAa [. . . . . . . . .] i ~ ( x v ~ c x & ~ Q ~ c (

payohayyo L~boat1
d Sodila [. . . . . . . . .] thc treasurer nus sent to
At that timc whcn thc stronghold was c o ~ n ~ l e t cthe11
the sanctuary. Thereupon

3. [ ~ i b ox~G]oyagtr1 ) o ~ ~ A
x qab o avi <t>u crpo 11aAil;twcocr~abocxy nphbt
[alp0 cpeoWaeho
Sodila had this image PI-eparcd, then he [is] w h o had [it] set u p in thc stronghold. ~frcrwnrds
when the watcr moved farther awav,
 Languages and literature in the Kushan Empirc

then thc divinitics werc led away from the waterless stronghold. Just thcrcforc, Sodila had a wcll
dug, then

5. I.)obt.Aaupo pahtSu aCjoyuvbo Q L ~OTL

~ Lo f i ~LiL U ~ U p~thj
~ E !iflo pu[-~]oi.uy[-~]o

4[a-I
Sodila had a water-conduit dug in the stronghold. Thereupon both divinities returned b ~ c khcrc

6. [o]o y o a t ~ v b oOTL E ~ L L Q O L U V~ LU~ ~ L X Tmtbo

L ~ O O LU I)obliLuy,~opuvcl
to the sanctuary. This was written by Miirozada by the order of Sodila.

The Bactrian inscription of Ayrtam allows us an interesting insight into the

inner organization and religious policy of the Kushan kingdom. The Kushan
gods represented o n the coins were for a long time shadowy figures. The situ-
ation changed when the sanctuary of Oanindo was discovered at Surkh Kotal,
and the sanctuary of Oaxso was found at Takht-i Sangin. N o w the cult of Farro
and Ardoxso is firmly attested by the relief and inscription from Ayrtam.
The Bactrian script and language were used for a long time after the
Kushan age but only small fragments of Bactrian literary works have been dis-
covered so far. The latest known examples of Bactrian script date from the end
of the ninth century A.D. and were found in the Tochi valley in Pakistan.

Sanskrit and Prakrit

The territory of the Kushan Empire included important parts of modern Pakis-
tan and India with a large population speaking Indian languages. Long before
the Kushan age two scripts - Brihmi and Kharosthi - and several literary lan-
guages - Sanskrit and different Prakrits - came into being and were highly
developed in the Indian subcontinent. Of the two scripts, Kharosrhi was used in
the north-west, its eastern limit running across the Panjab with only exceptional
examples further east, for example, in Mathura. Variants of Brihmi spread in the
other parts of the subcontinent. The language, written in the Kharos~hiscript,
was the G i n d h i r i Prakrit spoken in Gandhira and adjacent regions; Brihmi was
used for Sanskrit and, except for Gindhiri, for the other Prakrit languages.
The use of Kharosthi had already reached Bactria during the time of the
Graeco-Bactrian kingdom. The Graeco-Bactrian kings used Kharosfhi and
Gindhiri Prakrit as well as Greek for their coin inscriptions. This can be
explained partly by the fact that the Graeco-Bactrian kingdom included Gan-
dhira, a territory where Gandhiri Prakrit and Kharoghi script were used, partly
by their spread towards Central Asia across Bactria. Evidence of such a process
can be seen in the coins with the Gindhari legend in KharoslhT: Kaviiiye Nag-
nradevata (*KZpiiikZ Nagaradevata, city-goddess of Kipiia). There is also
. . inscription on the smoothing knob of a potter from the Graeco-
Kharosthi
Bactrian level of Begram (Klpiia): pn-8a-mi-tra-sa '[property] of Punyamitraj,
The name Pnnyamitra has a clear, Buddhist character and so this inscription
attests not only the spread of the Kharogrhi script and Glndhiiri Prakrit, but
also the appearance of Indian Buddhists in Graeco-Bactria.
Another early trace of Kharosthi can be seen at A y Khanum, where on a
potsherd a Kharoghi record came t o light: [sa x+]I dam I11 dha 111 '[stater X+]I
dramma I11 dhana 111'. It is likely that Kharosthi script and Giindhiri Prakrit
were brought by Indian merchants and artisans t o Transoxanian Bactria in the
Graeco-Bactrian period if the Kushan script (the 'unknown script', see above)
can really be derived from the K h a r o s ~ h ialphabet, and if the dating of the
inscription from Issik (see above) t o the end of the third century B.C.proves to
be correct. In any case, the use of Kharosrhi and Giindhari became more and
more extensive in the Saka and Indo-Parthian periods. The Kharosrhi inscrip-
tions on the gold ingots of the hoard from Dalverzin-tepe in northern Bactria
bear witness to this development.
The reasons for the quick spread of Kharoghi and Giindhari Prakrit in
Bactria and Central Asia are easy t o see. The first was that literacy was widely
spread among both Buddhist monks and Brahmans, and it was much easier to
find Indian scribes acquainted with Kharosrhi than experts in other scripts. So
Saka and Indo-Parthian and later Kushan administration became based, to a cer-
tain extent, on Indian scribes. Then, from the beginning of thc silk trade about
100 B.c., Indian merchants travelled t o China across Central Asia and contrib-
uted to the spread of Kharosthi in the Saka and Indo-Parthian kingdoms and
later in the Kushan Empire. As a trace of their travels in the western Pamirs, the
Kharosthi inscription of Dayr-Asan, dated t o the beginning of the first century
B.c., may be mentioned. Last but not least, Buddhism appeared in Central Asia,
and Buddhist monks also followed the Silk Route in the tracks of the mer-
chants, did active missionary work, found patrons and established monasteries.
The growth of the silk trade, the spread of Kharoghi script and Gindhiri Prak-
rit and the propagation of Buddhism reached a peak under the Kushans.
As a result of this development, Kharosthi script and Giindhiri Prakrit
conquered new territories in northern Bactria in the region of Termez, Chilas
and Gilgit as well as in Chinese Turkestan. According t o Hsiian-tsang, there
were ten Buddhist monasteries in the neighbourhood of Termez in the first half
of the seventh century A.I). Some of them must have been founded in the
Kushan age, and among them the cave monastery of Kara-tepe (excavated dur-
ing the last twenty years) was the most important. The nulnerous ~haro!!hi
inscriptions found there mostly represent records of donors written on earthen-
ware vessels. O n the basis of the letter forms, they can be datcd to the Kushan
period.
The Kharo!ihi rock inscriptions from Chilas and Gilgit, discovered as the
 1-anguages and literature in the Kushan 1:'rnplrc

of explorations since 1979, can similarly be dated to the Kushan period.

They are of three types: (a) records of pious donations (the image of a stupa o r
, carved o n the 'Sacred Rock of Hunza'); (b) records of personal
the ~ u d d h a etc.
names followed by the good-wish formula subralu (with bra instcad of bhya
like dra instead of dhra in the Kharosthi inscription of Kamra; thus < *su-
bhratrr < *su-bhartu < Old Indian su-bharatrr o r ru-bharatin ' S C > - ~ ~may ~-S~
be well!'); and (c) personal names. These are of great importance from both the
historical and cultural points of view. They bear witness t o Saka and Kushan
suzerainty in Gilgit, and provide clear evidence of both the penetration of
~ u d d h i s mand the spread of Kharosihi script and G i n d h i r i Prakrit into the
northernmost Indus valley.
The third region, that is Chinese Turkestan, was penetrated by Kharos!hi
and Gindhiiri Prakrit in the Late Kushan period. The numerous Kharosthi ..
administrative documents (about 800), written o n wood, leather and paper, were
found mainly at Niya and Lou-lan. Earlier researchers thought that they were
introduced into the administration of the Kingdom of Shan-shan as a result of
Kushan rule there. Later, however, it became clear that the Tarim basin had
never been subject t o the Kushans and the emergence of Kharosthi script there
cannot be explained by that theory. Kushan chronology also makes any such
connection impossible because the western part of the Kushan Empire was
annexed by the Sasanians in A.D. 234, while Kharoghi script was introduced into
the administration of the Kingdom of Shan-shan about A.D. 245. This can prob-
ably be explained by the assumption that when the Sasanians conquered Balkh,
many Indian staff w h o had worked in the Kushan administration escaped by the
Silk Route t o the Kingdom of Shan-shan, entered the service of King Tajaka who
in about A.D. 245 was reigning there, and played an important role in creating its
state organization, introducing Gandhari chancellery practice.
Compared with the Kharosthi script of Gandhira, the alphabet of the
Kharoghi documents from Niya and Lou-lan has some ~ e c u l i a rfeatures, of
which the most striking is the indication of long vowels by a short stroke writ-
ten below the line at Niya. T h e same hen omen on can only be observed in the
KharoSthi inscriptions of Kara-tepe and Fayaz-tepe near Termez. However, the
origin of this sign is explained, as its earlier emergence in northern Bactria
proves that it was from there that Kharos!hi script spread t o Shan-shan by the
Silk Route, that is, it did not reach Niya directly from Gandhira via Gilgit and
the Karakorum.
T h e indication of the length of vowels is fully developed in the Brihmi
script which was used t o write Sanskrit and Buddhist Hybrid Sanskrit. It there-
fore seems obvious that the indication of vowel length in Kharo?!hi developed
under the influence of the B r i h m i script in a religious or administrative centre,
where the t w o scripts were used side by side. T h e spread of Brihmi towards the
north-west had already begun in the Saka ~ e r i o d .Indian merchants using
Brihmi script for G i n d h i r i Prakrit had already reached China about the middle
of the first century B.c., as their presence is attested by the Brihmi inscriptioll
on a silk strip found on the Chinese limes at Tun-huang.
The role played by Buddhist monks in the spread of Brihmi was even
greater. The decisive turning-point was the synod of the S a r v i i s t i ~ i dschool
~
held in Kashmir during the reign of Kanishka, which, according to the tradition,
compiled the Jiiinaprasthrinam, and entrusted Aivaghopa, the famous poet from
Siketa, with providing for the correct language form of the commentary written
by Kityiiyana. In view of the fact that Aivaghopa wrote his works in standard
Sanskrit, his commission obviously meant the preference of Sanskrit to Prakrit,
which was also used earlier by the Sarvistividins. Earlier, both the
Mahisilighika and the Sarvistivida schools used Kharopthi and Brihmi equally
in the territories where the two scripts spread. Thus, in Mathura, both the
Mahisinghikas and the Sarviistiviidins used Brihmi script for their inscriptions,
while both schools adopted Kharosthi f o r their epigraphic monuments in
Gandhira.
After the synod of Kashmir, however, the Sarvistividins preferred Sari-
skrit o r Buddhist Hybrid Sanskrit and Brihmi script, and when they penetrated
Bactria on the tracks of the Mahisiinghikas, Brihmi also appeared in the Budd-
hist monasteries. This development can be seen clearly at Kara-tepe, where
inscriptions written in both Kharopthi and Brahmi occur on earthenware ves-
sels. The Kharopthi inscriptions belonged to the Mahasiinghika school as is
proved by the texts themselves. Therefore, the inscriptions written in Briihmi
probably represent the Sarvistiviidins. This connection between script and sect
after the synod of Kashmir is further proved by the fact that the first wave of
Buddhism brought the Mahasiinghika school together with Kharos~hiand
G i n d h i r i to Khotan, while the second transferred the Sarvastividins there
together with Buddhist Hybrid Sanskrit and Brahmi script.
There can be n o doubt that the indication of vowel length in Kharos~hi
script came into being under the influence of Brihmi script in the Buddhist
monasteries of northern Bactria, especially in the region of Termez, where
Mahisiinghikas and Sarviistiviidins lived side by side, and Kharosthi and Briihmi
were used side by side in the Kushan ~ e r i o d Thus,. at Kara-tepe, the spellings
kaii 'cup' and [ma]hasamghikanam 'of the M a h i s ~ n ~ h i k a occur s' while in
Fayaz-tepe the spelling sarvasatvana 'of all beings' is attested.
G i n d h i r i Prakrit, the language spoken in Gandhara and used for admill-
istrative and economic purposes by the Kushans, was also one of the literary
languages of Buddhism, and before the synod of Kashmir it had produced a
relatively rich Buddhist literature which was later thrust into the background b)
Buddhist works written in Buddhist Hybrid Sanskrit. Of Buddhist works in
G i n d h i r i Prakrit, only the Kharopihi Dhammapada has been preserved, and
this was discovered in Khotan, far to the east of ancient Bactria. The fate of the
Dhammapada shows what happened to Buddhist Gindhiiri Prakrit literature. It
was slowly driven out by the Buddhist Hybrid Sanskrit works written i l l
 Languages and literature in ~ h Kushan
c Empire.

Brihmi, and only survived to a limited extent in the city-states of the 'rarim
basin, while even there the local languages, Khotanese, Agnean and Kuchean,
used ~ r i h m instead
i of Kharopthi. KharoSthi was only retained for administra-
tive purposes in Kucha, where the latest documents are dated between A . I I . 618
and 647.
~ c c o r d i n gt o Buddhist tradition preserved in the Pili canon, monks of
Brahmanic origin proposed t o the Buddha that his words should be put into
~anskrit;and even though the Buddha ordained that everyone should use his
own language in reciting the sacred texts, the Sanskritization of Buddhist tcxts
began at an early date. T h e language, which came into being gradually bu the
, ,
increasing Sanskritization of Buddhist texts fixed in a Middle Indian dialect
(Prakrit), became Buddhist Hybrid Sanskrit.
Some Buddhist Hybrid Sanskrit works already existed as early as the
first century B.c., and the 'nucleus' of the Mahivastu, written with the aim of
describing the life of the Buddha, may go back t o the first century B.c., even
though it was successively expanded by additions, the latest of which can be
dated t o the fourth century A.D. While the growth of Buddhist Hybrid Sari-
skrit literature covers half a millennium, its golden age was the period of the
Great Kushans. T h e most important Buddhist Hybrid Sanskrit works were
compiled o r given their definitive form during this period. These include the
Mahivastu, the Lalitavistara (a Vinaya text of the Lokottaravidins, a school of
the Mahasanghikas, originally a work of the Sarviistiviida school giving a biog-
raphy of the Buddha), the Avadinas (tales of great acts o r of the fruits of
man's actions, the oldest of which may be the Avadinaiataka), the D i v y i v a -
dina (a collection of Buddhist legends), and the Saddharma-Pundarika (pro-
pagating the ideal and the worship of the Bodhisattva and glorifying the
Buddha as a being of inconceivable might).
T h e perfection of Buddhist Hybrid Sanskrit literature could hardly have
taken place without the personality and activity of the great Indian poet
Aivaghosa. According t o Buddhist tradition he lived at the court of the Kushan
king Candana Kanishka, w h o is t o be regarded as Kanishka 11, ruling from
Years 30 t o 42 of the Kanishka era (i.e. A.D. 164-76).
H e wrote the t w o kavYa epics, the Saundarananda (the legend of the con-
version of Nanda, the half-brother of the Buddha) and the Buddhacarita (the
story of the life of the Buddha himself). Unfortunatelv, the greater part of
Aivaghosa's poetic w o r k has been lost o r is only reserved in fragments, but it
is clear from his t w o epics that he was one of the most important poets of San-
skrit literature, w h o exercised an influence even o n Kiilidiisa. The style of
Aivaghoqa is relatively simple and obviously represents the so-called aida arb ha
style, but it is still impressive, sensuous and daintily elaborated. T o illustrate
this w e may quote two verses from the Buddhacrlrita depicting a sleeping
beauty of the harem:
 vibal~hauk a ~ a l a ~ n a v e n uanyii: r stanavisrastasitamiuk~iayina
~ j ~ ~ a t p a d a p a n k t i j u ~ ~ a p a~alaphrnaprahasatta@
drnd: nadiva.
One was glca~iing,holding a flute in her hand: she w . 1 ~lying with a white gnrnicnt slipping fro,,, lIcl.
bosom
lilic tIlc I.iver in whose lotuses whole s w a r m s of bccs delight: w h o s e banlts laugh with tlic fodm of lie,
w.Itcrs.

The importance and the popularity of ASvaghosa's poetic works are best shown
by their influence on Kiilidiisa and their spread beyond the borders of the
Kushan Empire to the Tarim basin, and to China in Chinese translations. Gjnd-
hhri Prakrit literature could not set anything of equal literary value against
them, and it was not therefore by chance that the fragments of the $iriputra-
prakarava, a drama of ASvaghosa, came to light in Turfan.

Sogdian
The territory of Sogdiana (the Zerafshan valley) did not belong to the Kushan
Empire, but Sogdian merchants engaged in the silk trade often visited both Bac-
tria and Gandhiira. In some periods they used the route across the Karakorum
range to Gilgit, and left many hundreds of Sogdian inscriptions on the rocks at
Thor and Shatial Bridge. These Sogdian records were written in the same alpha-
bet as the Sogdian 'Ancient Letters' found o n the Chinese limes at Tun-huang
from the end of the second century A.D., s o the bulk of the Sogdian inscriptions
at Thor and Shatial Bridge should belong to the Kushan, or at most to the Late
Kushan, period. They mostly consist of the proper name of an individual
together with that of his father with some indication of his origin and the
circumstances of his journey. Inscriptions with a longer text scarcely occur. It is
interesting to note that some of the Sogdian names mentioned in the 'Ancient
Letters' as Nanipandak, Nane0/3av, Aruvasppandak, Taxsicpandak also occur
in the inscriptions of Thor and Shatial. As most of the Sogdian names at Thor
and Shatial have no parallel in the 'Ancient Letters', the occurrence of the
quoted names may have particular importance. Perhaps Taxsicpandak, father of
Nanepandak, may be identical with Taxsicpandak, son of Nanepandak, men-
tioned in Letter 2; and Aruviisppandak, father of Farnc, may be the same as
Aruvasppandak, who is also mentioned in Letter 2. In this case the rock inscrip-
tions of Thor and Shatial would be dated to the end of the Kushan and the
beginning of the Late Kushan period in the third century A.D.
The same date can be proposed for the Parthian and Middle Persian
inscriptions carved on the rock among the Sogdian records. Both the Parthian
inscription (wryhm ihypwbrn < Varihrin S ~ h i ~ u h r i i and
n ) the Middle Persian
one (ipyh *Sap& or *Sipih) are written in the Pahlavik and Piirsik alphabets of
Early Sasanian date, that is, they can also be dated to about n.1). 230-60. The
 Languages and lircrature in thc Kushan L;rnl)irc

position of these inscriptions enables us t o elucidate the historical

background o f their emergence in T h o r and Shatial. Obviously the conqucst of
[he western part of the Kushan kingdom by the Sasanians interfered with traffic
and trade between Sogdiana and Kuiiniahr (now belonging to Iran), and be-
tween Sasanian KuSiniahr and the north-western part of the Indian subconti-
nent. T o keep away f r o m Sasanian KuSiniahr, Sogdian merchants took the
route through Gilgit and across the Karakorum range. Later, when political
relations between Iran, Sogdiana and the Indian Kushan kingdom were consoli-
dated, the difficult route across the Karakorum was abandoned.
The indications of origin in the Sogdian inscriptions deserve special attcn-
tion because they considerably enlarge our understanding of trade relations in
Central Asia. W e may quote the following inscriptions:

4a. pnit pysk dprt/?y 'n BR Y n 'pc 'Pisak, son of 0(3artfiayin, citizen of Nafi,
perished'. N a p can be identical with Nawa of the Arab geographers, a vil-
lage 2 - 3 farsakhs from Samarkand.
4b. wnnysrd ZK nrck BRY wrdnc 'Vananisarb, son of Narcak, citizen of War-
ban'. Warban may be identified with Wardina of the Arab geographers, an
important village in the district of Bukhara.
45. . . .]p'c BRY ixyp'yc '[So-and-so], son of [ . . . ] p'c, citizen of Sihfiay'. T h e
- - of Arab geography, a district in
latter name may be compared t o Sihbahi
the area of Bukhara.
51. pwxr'kk ZK wnxrk BRY p'ykndc 'Boxsak, son of Vanxarak, citizen of
Paykand'. T h e town Paykand lay 5 farsakhs from Bukhara.
57c. n'wrp' ZK rwd j n c 'Nawrapa, citizen of Roben'. T h e toponym Roben
'Copper [Fort]' may be another name for P a ~ k a n d the , 'Copper Fort'.
135. xwt'wz'mk Z K ki'ykndc 'Xwatiwzimak, citizen of Kaiekanb'. The latter
toponym may be the forerunner of Kiyikan or Kiikan of the Arab geog-
raphers (< KaSikanb), a village in the neighbourhood of Bukhara.

Most of the indications of origin refer to the tel-ritory of Bukhara and Samar-
kand. Besides, there are some remarkable indications:

9c. xnsc dwyt'kk cyn'nch 'Xansacbuytik, daughter of Xansac, citizen of

Cininc'. T h e fuller form of this toponym was Cininckanb; it was the
Sogdian name for Turfan.
64b. This is the record of wr/?'kk ZK 'kwc'k 'Warpik, the citizen of Kuta'.
Warflak seems t o be a name of Kuchean origin (cf. Kuchcan wlim.-, t o
stimulate).
122b. This mentions pysk ZK rrwtc 'Pisak, citizen of Raxwat'. Raxwat is the
Middle Iranian name for Arachosia.
 J. Harmatta

TIlus the settlements of the Sogdians were already spread throughout the whole
of Central Asia. From Bukhara and S a l ~ ~ a r k a nt do Turfan and from Arachosia
to Kucha, they played an important intermediary role in the mutual exchange of
both material and intellectual culture between Iran, India and China in the
Kushan age.
 N . N. Negmatov

The transition to sedentary culture

The Aral and Syr Darya region of northern Central Asia has, in almost every
~ e r i o dof its history, been the junction at which the advanced sedentary civil-
ization of the south met the nomadic peoples of the boundless steppes to the
north. It was the area where successive waves of sedentary farming people from
the Indus valley, Bactria, Parthia, Margiana, Sogdiana and neighbouring lands
met and intermingled with similar movements from the pastoral societies of the
Eurasian steppes. This far-flung Iranian-speaking population gave rise to the
civilizations of Central Asia's sedentary and nomadic peoples.
There were a number of social, economic, political, ethnic and cultural
processes underway in the Aral and Syr Darya region at this time, the most
important being the gradual transition from a ~ r i m e v a tribal
l society to the for-
mation of classes and early forms of statehood, which involved the introduction
and intensification of a farming economy, urbanization, the consolidation of
ethnic communities and the emergence of the historical regions with a sedentary
culture.

Ethnic history
In all these processes an important role was played by the changes that followed
the introduction of iron. In Central Asia this occurred during the first half of
the first millennium B.C. The process of learning to produce iron was facilitated
by previous experience of working bronze, but the new technologv was assimi-
lated slowly and the obvious advantages of the new metal were not imrnediatel!.
recognized. At the start of the Early Iron Age tools and weapons made partly of
bronze and partly of iron - daggers with an iron blade and a bronze handle -
were quite widespread. When, however, iron came into full use, it provided
great opportunities for socio-economic progress.
This transitional phase, in which Central Asia advanced from its primeval
condition to the formation of classes and early states on the basis of an agricul-
tural and pastoral economy, is reflected in the oldest texts of the Avesta. ~t lists
among 'the best of regions and countries' a i r y a n m vat?$, the Iranian territory
probably lying to the north of Gava (Sogdiana), Mouru (Margiana or Merv) and
Bix6i (Bactria). This was the first country in which Zoroaster's teachings
spread. It lay around the River Daitya and its winter lasted for ten months,
Many scholars have long equated it with K h w ~ r i z m(Chorasmia) and the River
DaityZ with the Amu Darya, while some have sought t o identify Chorasmia as
the Avestan dahyu o r confederation of lands and link it with the work of Vi;-
taspa, in whom they see the chief of the confederation.' It may, however, be
suggested that the lands of the Avesta most probably correspond to the ethno-
geological connotations of its name - the Aryan territory. Its socio-cultural cha-
racter includes the entire Aral and Syr Darya belt of northern Central Asia,
which was then a zone of sedentary farming and nomadic pastoral people,
stretching from Chorasmia through Cat (Chach) and Usrushana to Ferghana in
the east. In this context we can refer t o a 'Greater Chorasmia' as the supposed
equivalent to a i r y a n m v a ~ j k . 'The Chorasmians are incidentally mentioned by
Herodotus and Hecataeus of Miletus. This suggestion is also borne out by the
written and archaeological evidence for Chorasmia and the Syr Darya basin,
which were closely linked in many aspects of social, cultural and econoinic life.
They are located on the northern periphery of the sedentary farming cultures
and have many common elements of material culture - far more than the ele-
ments that Chorasmia may have in common with regions of Merv and Herat.
The Early Iron Age in Central Asia was marked by the establishment, be-
tween the seventh and fifth centuries B.c., of a federation uniting the Iranian-
speaking sedentary and semi-nomadic tribes of the region under the ~olitical
hegemony of the Chorasmians. In Chorasmia, Usrushana and Ferghana there
was rapid urbanization - a rapid growth of towns and their fortification, the
construction of city citadels and the development of agriculture by artificial irri-
gation works. These mark the transition to a class society and the emergence of
a strong central authority which undertook public works, and eventually devel-
oped new forms of economy based on sedentary farming and urban culture. At
the height of its power under Darius I, thc Achaemenid Empire included

1. F o r a suinlnary of information on n i , y a n ~ mv a ~ i oand its location in Chorasmia, see

Istol-iya Tadzhikskogo nnroda, 1963, pp. 151-4, 509-10; Isto,-l)la U z b c k s k o ) ~SSR. 1967.
pp. 84-5; Gafurov, 1972, pp. 37-8, 50, 52.
2. Gnfurov, 1972, pp. 58-9; Vorob'cva, 1979, pp. 38-42. Sec C h a p t e r 2 '~bove.
 States in north-western Ccntral Asia

, Saki Tigraxaudi (or Massagctae), Sogdiana and the 'Sakas who

~ h o r a s m i a the
are beyond Sogdiana'. The first three are recorded in the list of countries sub-
ordinate t o the next king, Xerxes (486-465 B.c.), and in the inscriptions (DNa,
DSe, XPh) and reliefs at Persepolis, Susa and Naqsh-i Rustam. a a siR-
nificant ~ r o p o r t i o n of the sedentary zone of northern Central Asia was
included in the Achaemenid Empire at the height of its prosperity. Only Fergh-
ana lay beyond its borders. Chorasmia, with Sogdiana, Parthia and Aria, made
up the sixteenth satrapy, which paid 300 silver talents in tribute to the imperial
treasury. The fifth satrapy of the Sakas paid 250 silver talents ( I silver talent in
this case amounted t o 34 kg).
Chorasmia, Sogdiana and the Sakas played a considerable role in the econ-
omy and ~ o l i t i c sof the Achaemenid Empire. Sources record their presence in
the capital, and the presence of Chorasmians on Elephantine Island (on the
Nubian border) and at Memphis, and both Chorasmians and Sakas in Nippur
and other Babylonian cities. Chorasmian and Saka names are known from Sip-
par in Babylonia.' A Saka detachment fought in Darius' army at the battle of
Marathon (490 B.c.); Saka cavalry and Chorasmian and Sogdian warriors were
included in Xerxes' ilite units during his campaign against Greece. Saka infantry
took part in the battles of Thermopylae (480 B.c.) and Plataea (479 H.c.). Terra-
cotta figurines of Saka warriors in pointed caps have been found in cxcavations
at various places in the empire, ranging from Egypt to Central Asia. By the
second half of the fourth century B.c., however, Chorasmia and the Sakas had
become independent states - rather earlier than the other, more southerly, parts
of Central Asia. Ethnic history developed from the tribal distribution of the
Late Bronze Age. The Iranian-speaking population in the valleys and oases of
northern Central Asia became more localized and stable. The Chorasmian peo-
ple emerged o n the southern Aral coast and in the Amu Darya delta; a Saka
community arose between the lower Amu Darya and the Syr Darya and in the
hill country of the upper Syr Darya; and a northern branch of the Sogdians was
settled on the left-bank flatlands of the middle Syr Darya. Somewhat later, the
Saka and Sogdian populations of the middle and upper Syr Darya established
Ferghana and Usrushana. The formation of these ethnic groupings was pro-
moted by two complementary processes - the emergence of sedentary regions
in which the dominant occupation was agriculture, and the beginnings of
urbanization. The Chorasmians, Sogdians and Sakas shared many ethnic and
cultural traits with other peoples and tribes of Central Asia. valuable evidence
for their anthropological and ethnic characteristics come from the Bisutun, Per-
sepolis and Naqsh-i Rustam reliefs. T o judge from this evidence, the central
Asian peoples tribes were vil-tually indistinguishable in clothing, head-gear

3. Further details will be f o u n d in Istol-ijn Tlldzhikskogo,~'~rodn,

1963, pp. 208-9; Gafuro\..
1972, p. 83.
and armaments. Their dress was much the same, consisting of short tunics, a
broad belt and narrow trousers; only the Saki Tigraxaudi were set apart by
their ~ h a r ~ - ~ o i n cap.
t e d Ancient Greek and Persian accounts d o not distinguish
between the oasis peoples and the steppe tribes. O n the contrary, they claim
that 'the Massagetae and Sakas also include Attasians and Chorasmians' and that
'in ancient times the Sogdians and Bactrians differed little from the nomads in
life-style and customs' (Strabo). All this points t o close ethnic, linguistic and
cultural bonds between the Central Asian peoples and the tribes speaking East-
ern Iranian languages of the Indo-European group, while anthropological aria-
lysis of excavation material shows that they were all of Europoid type. Between
the seventh and fifth centuries B.c., all these ~ e o p l e swere in the final stages of
the transition t o a sedentary farming and pastoral life-style.

Irrigation and agriculture

The development and success of the agricultural economy was closely depen-
dent on the progress made in artificial irrigation. This had been introduced in
the Late Bronze Age, and agricultural and pastoral techniques were still rela-
tively primitive in the Early Iron Age, but advanced irrigation networks were
developed in Chorasinia and Ferghana between the sixth and fourth centuries
B.C. (see Chapter 12). Different types of irrigation were developed to meet dif-
ferent geographical circumstances. O n the plains of Chorasmia and Ferghana, in
the Amu Darya deltas and on the upper and middle Syr Darya, spring and sum-
mer floods were used t o store water, which was then diverted into artificial
canals.
In foothill areas, tunnel irrigation was the commonest form: groundwater
was diverted via underground catchment tunnels. In the mountains meltwater
streams were used. Artificial irrigation was intensively developed, particularly
in the lower reaches of the Amu Darya and along the delta arms, where the
steady flow was easier t o control. Chorasmia had grand canals 20 to 40 m wide
but not very deep, extending for a distance of 60 t o 70 km. In the Aral and Syr
Darya region, the irrigation system was complex. Farming o n mountain terraces
and in river basins and basin (estuary) farming depended on natural rain with
the water supply controlled by dykes and dams.' In the valleys of the Syr D a r ~ ~ a
and lower Amu Darya, the development of artificial irrigation made agriculture
the basis of the state economy. The agricultural products included millet, bar-
ley, sesame and wheat. In addition to cereals and oil crops, pumpkins, melons,
fruit such as apricots, peaches and pluins, and vineyards were cultivated in the
oases, while orchards of apricots, walnuts and almonds were grown in the hills.
 States in north-western Central Asia

Mining and manufacture

~ r c h a e o l o ~ i c astudies
l and written sources indicate that the popu]ation was
engaged in various occupations - in mining and smelting copper and iron, min-
ing precious stones, manufacturing tools, arms and pottery, and in weaving and
. ..
building actlvmes. Internal trade and commerce flourished among the popula-
tion of the oases and steppes in Chorasmia, Ferghana and Usrushana. The
region was involved in trade with countries to the south of Central Asia, India,
the northern Aral steppes and possibly the Volga region. Gold, copper, silver
and iron were mined in the Kyzyl Kum, the Nuratau mountains, the Naukat
deposit in the Ferghana valley, the Khojand hills, the Kurama (Kara-Mazar
mountains) and Chatkal ranges, the Ahangaran valley, the Almalvk district and
the Karatau mountains. Many places where metals were smelted have been iden-
tified, complete with fragments of slag, in settlements in the Kayrak Kums.
These roba ably drew their raw materials from deposits at Naukat, Uchkatli
Miskon, Dzhidargamirsay, Chakadambulak, Aktashkan, Kochbulak and Koni
Mansur in the Kara-Mazar. Antimony workings from the middle of the first
millennium B.C. are known at Bagashiny in the Sokh river valley in south-west-
ern Ferghana. Piles of copper slag have been recorded in the southern Bukantau
hills and along the Kerichetau range near the Aral Sea. Greenstone hammers and
a large stone pick weighing as much as 8 kg have been found, used by ancient
miners chiefly for open-cast ore-mining. Hammers, rounded or elongated, were
made of rectangular pebbles with the sharp edges struck off for ease of attach-
ment t o the wooden shaft.
T h e quantity of the metals and the range of artefacts made from them,
particularly arms, harness, armour, sickles, knives, needles, jewellery and perso-
nal ornaments, are seen in the objects found in archaeological excavations.
According t o Herodotus (1.215), the Massagetae used bows and lances, but their
favourite weapon was the battle-axe. They used copper for their battle-axes,
spear-points and arrow-heads, and adorned their head-gear, belts and girdles
with gold. Their horses had breast-~latesof copper, but they inlaid their horse
trappings with gold. Both gold and copper were abundant, but not silver o r
iron. Precious stones were also found there. Turquoise was mined in the land of
the Sakas (Pliny), possibly in the Nuratau, Khojand and Isfara lnountaills and in
the Kyzyl K u n ~ According . t o the DSf inscription of Darius, when his palace in
Susa was built, k i s a k a hara axiaina (turquoise) was brought from Chorasmia
and k i r a k a bays kapautaka (blue stone o r lapis lazuli) and sinknbrui (carnelian
or cinnabar) f r o m Sogdiana. During this period the bulk of those w h o ~ r o d u c e d
material goods in the region's sedentary farlning and pastoral societies lived in
kinship-bascd rural, o r ldter urban, communes and were collectively engaged in
agriculture o r in dolllestic crafts. Thev paid rent t o the st.~te,the tribal military
aristocracy and the clergy, and also had other social obligations. The class of
slaves included those made captive in war and impoverished commune mem-
hers, who had to work on the farms of landlords, help in building and maintain-
ing irrigation systems, o r join the private armies of the rulers and the aristoc-
racy.
111 the steppes and mountains, cattle-breeding and horticulture dolninated
econolllic activities. The rapid growth of cattle-breeding, particularly in the
steppes, played an increasing role in the economy as a whole, changed the life-
style of the pastoral population and added t o the importance of these areas as
oases with their craftsmen and farmers. The two economies supple-
mented each other in a way that increased general prosperity, but eventually led
t o the emergence of a class structure and the exploitation of slave labour.

The development of cities and urbanization

The process of urbanization began earlier and o n a greater scale in Choraslnia
and on the left bank of the middle Syr Darya, localities which were 1nol-e
advanced in economic and cultural terms. They were geographically closer to
the ancient urban centres of south-western and southern Central Asia and were
open t o their influence through Margiana and Sogdiana. They were later incor-
porated as provinces of the Achaemenid Empire and came into its socio-econ-
omic orbit for a time. In the southern Aral region, the sedentary farmers and
pastoralists of the Chorasmian oasis represent the Late Bronze Age Amirabad
cultural pattern seen in the Dzhanbas and Yakka-Parsan settlements. At that
time they had master craftsmen (the 'house of the caster') with settled houses
and social gradation^.^
The oldest Chorasmian city, and the key monument of this period, was
Kyuzeli-gir, dating from the fifth and fourth centuries B.C. It lay on the left
bank of the Amu Darya in the Sarikamish region of the delta. Standing on a
natural elevation, roughly triangular in ground-plan, it occupied an area of 25
ha. The city was surrounded by a powerful defensive wall with oval bastions. Its
residential district was densely packed with buildings of rectangular unbaked
brick and pakhsa. It had an advanced pottery industry, based on the wheel, and
art objects of a type comrnon in Saka burial coinplexes of the ~ e r i o dhave been
found.' Another early city of the same date, Kalali-gir, was surrounded by triple
walls with bastions and had four gates with entrance barbcans and a hill-top
palace, but it was never completed.
In Chorasmia, in the eastern part of the south Akcha Darya delta, the
agricultural oasis of Dingildzhe dates from the fifth century B.C. The eastern

5. Itina, 1977n.
6. T o l s t o v , 1962, pp. 96-104; V i s l ~ n e v s k a ~ n1972,
, pp. 533-4, 1978, pp. 544-5; Vorob'eva,
1979, p. 39.
 States in north-western Crnlr~llAsia
-
part of the site was occupied by a large house with many rooms and out-bui]d-
ings, the wester11 part by a large courtyard. The whole was surrounded by an
outer wall 2 m thick, built of large rectangular unbaked bricks of archaic type.
The archaeologist w h o studied the farmstead suggests that it might llave been a
~ommunalhome f o r the extended family of the district governor. Its inhabitants
were engaged in agriculture and cattle-breeding within the oasis, pottery, metal-
work and making farm products. Dingildzhe points to the relatively high stan-
dard of architecture, building techniques and design that prevailed in
Chorasmia.'
Between the fourth and second centuries B.C. Chorasmia had a series of
walled cities with strong moats, complex fortifications and gateway barbicans.
They defended farming districts that lay along the caravan routes, and served as
centres for crafts, trade and culture. They include Dzhanbas-kala and Bazar-
kala, with precise and regular ground-plans, on the right bank and Hazarasp
(Sauvar) and Dzhingirbent o n the left bank of the Amu Darya.# Chorasmia's
ancient cities and fortresses have several characteristic features. Thev were either
built o n marshland where the farming population met the steppe, or stood on
the major trade routes. Fortresses on the plains, such as Hazarasp, had a regular
rectangular ground-plan; smaller fortresses stood o n high ground, such as
Kalali-gir 11, Lesser Kirkkiz, Burli-kala; and at the foot lay undefendcd secon-
dary settlements, such as Guldursun, Akcha Gelin, Kunya-Uaz and Toprak-
kala. Their size and strength contrasted sharply with the mass of small unfor-
tified settlements in the farming oases.y
T h e outstanding structure of ancient Chorasnlia was the great religious
centre of this far-flung region, the fortified sanctuary of Kay-Krilgan-kala
(fourth century B.C. t o fourth century A.D.) in the southern Akcha Darva delta,
on the right bank of the Amu Darya. Circular in g r o u n d - ~ l a nwith a diameter
of about 90 m (Fig. I), it consists of a large cylindrical building surrounded at a
distance of 15 m by a fortress wall. The space between the building and the wall
is occupied by several tiers of buildings. The site has two ~ e r i o d sof occupation
and three phases of building. The central two-storey building is a round, monu-
mental structure with a single row of arrow-slits over a row of trapezoid win-
dows. In the lower storey there were eight rooms roofed with twin domes,
communicating with the central chamber by arched passages. It is thought that
the central building may have been a temple, whose lower storey had religious
functions, while the upper part served as a store. T h e lower floor was divided
into two identical lialves, suggesting that the temple may have been used as an
astronomical observatory, as is suggested by its alignment. shortly after the cen-

7. Vorob'cva, 1973.
8. Tolstov, 1948, pp. 113 et scq.; Itiua, 1977b, pp. 193-201; Itina, 1951, p. 15.
9. Nerazik, 1972, pp. 34-7, 1981. pp. 219-27.
 FIG. 1. Koy-Krilgan-kala. Drawing froni an aerial photograph. E. J. Brill, Leiden.

tral structure was completed, the open space u p t o the fortress walls began to be
filled with houses and storage premises. In the second period of construction
the ruined central building was partially adapted for living quarters; some
rooins were cleared and repaired and the whole area bctween the ccntral build-
ing and the fortress wall was filled with clearly separated blocks of buildings.
T h e third construction period provided a nulnber of similar houses which util-
izcd the oldcr buildings. Each house was probably an independent economic
unit. T h e finds, especially the pottery, reinforce this interpretation. The bottom
stratum contains an assemblage of well-formed vesscls with ,I red slip - a censer,
 States in north-w~stcrnCcntral Asia

small-stepped altars and krracotta figurines representing deities of the Choras-

mian pantheon. The upper strata, however, yielded a completely new series of
pottery with a light-coloured slip, unusual in Chorasmia, with colllmonly
found ~ u s h a n - s t y l evessels and coins. Koy-Krilgan-kalalbay be
with an astral cult.
The system of Chorasmian fortification reached a high standard in this
period. walled cities usually had a severely regular rectangular ground-plan
divided by an axial street. Koy-Krilgan-kala, the ollly example of a fortified
sanctuary, takes the form of a central building surrounded by an external wall
strengthened with nine bastions. Frontier fortresses were built at strategic
~ o i n t son the borders of oases, defended by many-tiered projecting bastions
spaced t o command all approaches. Fortified houses had a dual role, for resi-
dence and defence. Chorasmian fortifications were developed particularly be-
tween the fourth century B.C. and the first century A.D. Eventually privatc cas-
tles proliferated and fortresses of new design developed on the edge of the
oases. The walls of cities, fortresses and settlements in ancient Chorasmia were
built on inclined pakhsa o r raw-brick socles. They were 10-20 m high and 5-8
m thick. T o strengthen the defences, supplementary outer walls were built 5-20
m from the fortress walls, with open ground intervening. Arrow slits were
generally arrow-shaped and to widen their fire they were grouped in a chess-
board pattern. Particular attention was paid to the gates, which sometimes had
additional projecting bastions at the entry, on the corners and along the sides."
There is evidence of cattle-breeding in the Sarikamish area of the Amu
Darya delta. Here archaeologists have discovered the Kuyusay culture, repre-
sented by settlements and burial-grounds (between the seventh and fourth cen-
turies B.c.). Its'culture has a local Saka substratum in close contact with an
external component, reflected in imported wheel-thrown pottery from south-
western and southern Central Asia.Iz There is a striking variety of burial rites in
Chorasmia at this period and a growing influence of the Zoroastrian rite with its
ossuary type of burial is clearly discernible.
The economic and cultural pattern of the semi-sedentary Saka-Massagetae
pastoralists and farmers in the lower Syr Darya lain is illustrated by a series of
sites: the extremely rich mausoleum complexes for tribal leaders in the northern
Tagisken country from the late second and early first millennia B.C. with their
monumental raw-brick architecture and pottery, made by skilled craftsmen; the
~ i x t h - t o - f i f t h - c e n t u r ~ -barrows
~.~. at Uygarak and in southern Tagisken;" the
f ~ ~ r t h - t ~ - ~ ~ ~ walled ~ ~ dsettlements
- ~ ~ ~ ~ of uC hri r~i k ~
- ~ a. bca tand
. ~abish-

10. Koy-Krilgaw-knla, 1967.

11. K h o d ~ h a n i ~ a z o 1981,
v, pp. 43-56.
12. V a ~ n b e r g ,1979, pp. 1-76.
13. Visl~nevslta~a, 1973; Tolstov dnd Itina, 1966, pp. 1 S 1-75; Vishncvska!,a and Itina, 1971,
pp. 197-208.
 N.N.Negmatov

Molls, together with their farming land and its irrigation works on the zhani
Darya" to the west of the Syr Darya; and finally, the monuments of agricultural
oases at Dzheti-Asar along the tributaries of the Zhani Darya and Kuvan Darya,
which lasted from the first century B.C. to the Early Middle Ages.15 ~ 1 tllese 1
suggest a distinctive, complex culture with an advanced pastoral economy
alongside agriculture. There were large cities, smaller settlements, a system of
fortress-type strongholds with thick walls and towers and enormous burial
grounds. The early phases are marked by monumental raw-brick architecture,
wheel-thrown vessels and ornaments that incorporate elements from the south-
ern farming cultures (the northern Tagisken cemeteries imported Yaz-II-type
pottery, beads and animal figure art, drawn from the worlds of Central and
Western Asia (from Uygarak and south Tagisken) and elements of the Scythe-
Siberian animal style (from southern Tagisken and Uygarak).
O n the left bank of the middle Syr Darya, urbanization was intensive and
impressive in scale. Here in Usrushana the earliest urban centres were formed
between the seventh and fifth centuries B.C. The fortress-city of Khvatak (now
the site of Nur-tepe), stands o n a high chain of hills on the north-western edge
of the Ura-Tyube oasis. Some 16 ha in area, it consists of a citadel and town sur-
rounded by a system of fortifications. The southern defence of the city con-
sisted of a natural dyke of loess and a wall of pakhsa, 8 m thick, built between
the sixth and fifth centuries B.C. Within the town and citadel, excavations have
revealed thick cultural strata containing the remnants of the original under-
ground dwellings and of later well-designed pakhsa and unbaked-brick struc-
tures. The finds include an assortment of wheel-turned and hand-luted pots for
storage, cooking and table use, querns and spindle whorls. The irrigation sys-
tem, using mountain streams and springs, helped develop the agricultural econ-
omy. Craftsmen also played an important role in city life. The powerful
defences of both town and citadel, its complex architecture and planning struc-
ture and the fine pottery are evidence of the advanced social life and cultural
level of the city.16
Kurukata (Cyropolis in classical sources) is known to have been the
second oldest and largest city of Usrushana, with ~ o w e r f udefensive
l walls and a
specially strengthened inner fort. Many opinions have been expressed about its
location. Although the name 'Kurukata' has been preserved in the name of the
modern settlement of Kurkat in northern Tajikistan (and the ruins of an ancient
city lie nearby) it is preferable to identify it with Ura-Tpube. The heavily for-
tified citadel (Mug-tepe), here in its centre, 6 ha in area, is surrounded by high,
thick walls. Excavations have revealed deep cultural strata with architectural

14. Tolstov, 1962, pp. 136-86.

15. Levina, 1971, Vol. 7.
16. Negmatov et al., 1982, pp. 89-1 1 1 .
 States in north-western Central Asia

from the sixth t o the second centuries s.c.17Written also

of the old Usrushanian fortress-cities of Gaza, Baga and four that are
unnamed. The city walls at Gaza are, however, 'earthen and low'.1& The remain-
ing cities of the province are being explored.
~ e s i d e sthe above-mentioned sites, archaeologists have discovered another
major urban centre, on the Khojand plain on the left bank of the Syr Darya
where Usrushana meets Ferghana. It was founded in the sixth to fifth centuries
B.C. and was fortified initially with loess dykes and later with defensive walls of
large unbaked bricks of an archaic rectangular type. It occupied a square site,
20 ha in area, and consisted of two parts: the town itself and the citadel. I t was
inhabited continuously up t o the early centuries of our era. The earliest strata
wheel-thrown and hand-made pottery for storage (pitchers and jars),
for cooking (hand-made cauldrons) and for the table (bowls, cups, vase-like ves-
sels and goblets), including carefully made, technically advanced examples
which borrowed pottery techniques from southern Central Asia and from the
central Ferghana's Eylatan culture (seventh to fourth century B.c.). Large grain-
stores using pitchers, stone querns, pestles and terracotta spindle whorls have
been found, indicating cereal farming, milling, the storage of foodstuffs and
other craft products and suggesting that the local economy was based on agri-
culture and local crafts.I9

Development in Ferghana
In Ferghana an early farming life-style was established at oases of the Chust cul-
ture (e.g. Dalverzin and Chust) which had underground structures dating from
the tenth t o the eighth centuries B.c.'O In the subsequent (Eylatan) period
(seventh t o fourth century B.c.) agricultural settlements with traces of ground-
level architecture and defensive dykes appear. They provide a clear picture of
early towns in the Shurabashat and Markhamat (the Ta-yuan and Kushan)
periods between the fourth century B.C. and the fourth century A.D. Cultural
strata from these two periods have been recorded at the sites of Eylatan, Shura-
bashat, Kara Darya, Markhamat, Minga-tepe, Dzhelandi, Turtkul', Severo-Kur-
shab-I, Kurgan-tepe, Yangibazar, Uchkurgan and Kaynovat.?'
The Eylatan culture of Ferghana is known from a number of excavated
sites. Eylatan, in eastern Ferghana, between the Naryn and the Kara Darya

17. Negmatov, 1957, pp. 16-20; Ranov and Snltorska)-a, 1961, pp. 1 17-29; Neglnatov and
Saltovskn!.a, 1962, pp. 71-7.
18. Negmatov, 1957. pp. 21-3.
19. Negmatov, 1962, 1973, 1975, 19SOb.
20. Zadneprovslti!., 1962, pp. 1 1-1 07, 1966, pp. 193-207, 1978.
21. Z a d n e p r o ~ s k i ~1962,
, pp. 108-99; G o r h u n o v ~ ,1977, pp. 107-20, 1979, pp. 16-34.
rivers, consists of an inner walled village 20 ha in area and an outside dyked area
of 200 ha that might have been used for cattle-pens. The residential settlement
had ground structures but no citadel. Finds include pottery, hand-made on a
cloth mould (a sand-filled bag), and wheel-thrown pottery of the same shapes,
the commonest types being round-based CUPS and bowls. The cups were often
painted with horizontal lines and sometimes with decorative designs. Eylatan-
type pottery is found in the lower strata at Shurabashat and other sites in east-
ern Ferghana.
T o this sedentary farming culture belong the Aktam, Niyazbatir and
Kungai burial-grounds of central Ferghana, all with much the same burial
inventory of pottery, iron bracelets, pins, beads, earrings, iron knives and
bronze arrow-heads. There are, however, different grave arrangements: 'small'
mounds with a single burial pit and long ones with three to nine pits in Aktam,
'long' graves 30 m from north to south, with three to eleven burials, and 'small'
round ones with one to three burials, all resembling cairns of cobble-stones, in
Kungai. The differences can probably be explained by membership of different
ethnic groups. In the Shurabashat and Markhamat periods, the development of
agriculture and the growth of urban centres were matched by the construction
of an irrigation network, the establishment of a separate Ferghana state (Ta-
yiian) and the emergence of ancient Ferghana's Iranian-speaking people, the
Parikanians, from local Saka and Sogdian stock.22
In the Chirchik and Ahangaran District of cat
(Chach), small early urban
centres sprang up around the settlement of the predominantly farming cultures
of the very Late Burgulyuk ~ e r i o d(ninth to fourth century B.c.) and through-
out the Kaunchi ~ e r i o d(third century B.C. to eighth century A.D.). One such
centre has been identified at Shash-tepe in the city of Tashkent. A Late Burgu-
lyuk cultural stratum o n continental loess was found to contain the remains of
sixth-to-mid-fourth-century-B.C. semi-subterranean houses, while above was
the first defensive wall of the Shash-tepe fort, made of raw brick andpakhsa and
dating from the mid-third century B.C. By that date the farming settlement of
the mid-first millennium had developed into a small town with an area of about
1 ha with a defensive wall.23A small urban centre dating from the first centuries
B.C.has been identified beneath the site of Kanka, and in the Kaunchi-I1 ~eriod
(fourth to fifth century A.D.) a host of urban-type settlements arose. Tashkent
too was developing as a town, with urban components of a town fort, a ruler's
palace and a khristdn. Archaeologists conclude that the ancient cultural sub-
stratum underlying the urban and rural culture of C i ? , or that part of the Chir-
chik valley that was incorporated into 'Greater Tashkent', was the ~ a u n c h i

22. Zadneprovskiy, 1956, pp. 3 9 4 4 , 1978, pp. 47-9.

23. Filanovich, 1980, pp. 459-60.
 States in north-western Central Asia

culture.24 The evidence of the largest site, Kaunchi-tepe, dates the beginning
irrigated agriculture in 'Greater Tashkent' t o the third and second centuries
, . ~ . 2 j and the construction of the first artificial canal, that of Baz-,u, to first
century B . c . ~ T~ h e agricultural use of the Chirchik basin and the construction of
canalsdeveloped during the early centuries A.D.'' The initial pllaseof the urban-
ization of C ~ can E then be dated significantly later than that of other parts of the
regionsZR
A review of the urbanization process in the Aral and Syr Darya zone of
"~rthernCentral Asia and the descriptions given above of the rarly urban
centres of Chorasmia, Khojand, Usrushana, Ferghana and Cii- consequently
enable us t o state that the earliest of these centres, the most impressive in terms
of area, topography and planning and the most advanced in economic and cul-
tural respects, were those of Usrushana and Khojand on the left bank of the
middle Syr Darya. These centres witnessed the transformation of cultural tra-
ditions coming from other parts of northern Central Asia.
The towns and urban culture of Chorasmia and the upper and middle Syr
Darya seem t o have based their development on predominantly local traditions
and on those drawn from the centre (Sarazm), south (Sapalli, Dashli and
Kuchuk-tepe) and south-west (Altyn-tepe) of Central Asia.zy W e see a com-
bination of local traditions, the use of cultural standards established in other
centres and cultural interaction and cross-fertilization.

Subsequent developments
In the last centuries B.C. and the first centuries of the Christian era the agricultu-
ral areas in the Aral and Syr Darya zones developed chiefly within the indepen-
dent states of Chorasmia (Khwlrizm) and Ta-yiian (Ferghana), though for brief
periods they were partially included as provinces of the empires of Alexander the
Great, the Seleucids, Graeco-Bactrians and Kushans. Between 290 and 160 B.c.,
Usrushana and Khojand appear t o have been part of the Graeco-Bactrian king-
dom." These political changes influenced their material culture. Archaeological
investigations have revealed Hellenistic elements in Usrushana, Khojand

24. Buryakov, 1975, pp. 186-7; Filanovich and Abdullaev, 1975, p. 515; ~ b d u l l a e vet at.,
1977, p. 522; Buryakov et al., 1979, p. 546.
25. D r e v n i j Tdskkent, 1973, p. 141; Drevnostj Tashkenta, 1976, pp. 49, 115-16, 124, 130;
z a , pp. 112, 158.
Drevnosti T u ~ ~ n b u ~ u 1978,
26. Drevnij, Tashkent, 1973, pp. 102-3.
27. Drevnosti Charvrlkd, 1976, p p 8 3 4 .
28. Rtveladze, 1980, p. 29.
29. Negmatov, 19826, pp. 61, 63, 198217, pp. 51-3, 1982c, FP. 72-88.
30. Negmatov, 1957, p. 287.
 N. N. Negrnatov

and western Ferghana" and finds from Khojand clearly betray Hellenistic
influence in the formation of Central Asia's ~ u l t u r e . ' ~

Around 160 u.c. Usrushana and Khojand became independent of the Graeco-
Bactrians: Ferghana, it seems, had never been part of it, and Greek ascendancy
never extended beyond the regions conquered by Alexander." A remark by
Strabo (XI.11.1), however, has led many scholars to assume that Ferghanawas
included in Graeco-Bactria. Coins of the Graeco-Bactrian monarchs have been
found there, but possibly as a result of commerce." In the mid-second century
B.C. the Yiieh-chih tribes passed southwards through Ferghana and Usrushana,
and subsequently conquered Bactria. It seems likely that the far-flung wealthy
and densely populated state of Ta-yuan arose about the same time. ~~~h
detailed information about this state is given by the Chinese chronicler SzG-rna
Ch'ien, who passed through Ta-yuan in the latter half of the second century
B.C.~' The name Ta-yuan was used until the second century A.D., when it was
replaced by Pu-han and Pa-han-na (fifth century A.D.) - the Chinese trans-
criptions of the name 'Ferghana'. The identification of Ta-yuan with Ferghana
is firmly established in historical 1 i t e r a t ~ r e . j ~
According to the Chinese sources, the country had many large and small
towns and settlements, numbering over seventy. The population was 300,000
and the inhabitants had deep-set eyes and thick beards; they were skilled mer-
chants and held women in high esteem. The country's army numbered 60,000
fighting men armed with bows and spears, skilled in shooting from horseback.
It was a land of highly developed agriculture; both wheat and rice were grown;
there were large vineyards, wine was made and stored for dozens of years, and
much mu-su (lucerne) was sown. Particularly famous were the Ferghana horses,
highly prized in neighbowing lands and especially in China. They were said to
'sweat blood' and were considered 'heavenly'. Emperor Wu-ti was
keen to have these blood-sweating horses. At one time they were worshipped in
China and poets wrote odes to them.
Ta-yuan also included Khojand and Usrushana." T o the north and west it
bordered o n K'ang and to the south o n the Yueh-chi11 or Kushan possessions.
Its capital was the city of Ershi, identifiable either with the ancient site of Mark-

3 1. Negmatov, 1980a, pp. 54-6.

32. Belyaeva, 1980, pp. 11-12.
33. Bartol'd, 1964, p. 456.
34. Gafurov, 1972, p. 123.
35. Bichurin, 1950, Vol. 11, pp. 161-2, 186-8.
36. Gafurov, 1972, p. 132.
37. Bicliurin, 1950, Vol. 11, p. 312.
 Stares i n north-wcstc~rnCenrral Asia

hamat in Andizhan District or with Khojand o r Ura-Tyubc.'m lts rulers also had
a in the city of Yu-then, possibly present-day Uzgcn. As it was rich i n
horses, Ta-yiian attracted the attention of the Han emperors, who =ttempted to
subdue Ta-yuan. Many years of warfare ensued against these invaders. In 104
l Kuang-li moved against Ta-yuan with a 100,000-strong army -
8.c. ~ e n e r a Li
60,000 cavalry and 'several dozen thousands of young warriors from China. T h c
war lasted f o r f o u r years. T h e entire H a n Empire was set in motion. More than
fifty chieftains were sent t o Ta-yuan . . .'. The campaign was carefully
the Chinese army even included craftsmen to divert the water from Fcrghanays
cities and horse marshals t o select 'heavenly' horses in Ferghana; it carricd
'dried food provisions'. Headed by their king, Mu-ku-a, the people of Ferghana
fought the invaders boldly. Both sides suffered heavy losses and Mu-ku-a died
through treachery. Particularly ferocious was the forty-day siege of the capital
of Ferghana. According t o SzG-ma Ch'ien, 'Regardless of all this, the Chinese
could not enter the city and took the road back', contenting themselves with
'several dozen of the renowned horses' and elevating t o the throne, instead of
Mu-ku-a, the old Ferghana magnate Mo-tsai. However, they did not succeed in
entrenching themselves in Ferghana. After their army had withdrawn, the pup-
pet ruler was killed by the people, and the younger brother of the valiant Mu-
ku-a, known in Chinese as Chang-fun, was placed on the throne. It is inter-
esting t o note that the Chinese took from Ferghana the art of cultivating vines,
lucerne and possibly also pomegranates, cucumbers, walnuts and figs.

Chorasmia had secured its independence from Persia before the end of the
Achaemenid Empire. According t o Arrian (IV.15.4-6) Pharasmanes, King of
Chorasmia, visited Alexander during his sojourn in Bactria in 329-320 B.C. and
promised t o guide him and provide supplies for his army if he chose t o proceed
against Colchis and the Amazons, subduing all the races that lived in these
regions u p t o the Euxine Sea. T h e arrival of an embassy from the European
Scythians in this context probably refers t o the Saka-Massagetae, their neigh-
bours from the Aral regions. A friendly alliance was concluded with the head of
the Chorasmian state. Between the fourth and first centuries B.c., Chorasmia
was a powerful independent state. Although there is virtually no written
information, there is incontrovertible evidence about the growth of its ancient
cities and settlements with strong and structurally developed fortifications
which have been discovered in recent times. Little is known about the country's
political history in the early centuries of the Christian era.
It has been suggested that Chorasmia formed part of the Kushan Empire"

38. Bei-nsht~m,1951, p. 1 I; Aristov, 1903, p. 93.

39. Tolstov, 1948, p. 151, 1962, p. 224; Gulsarno\., 1974, p. 119.
 N.N.Negmatov
-
and that the latter stretched from the Aral Sea to the Indian ocean.'^ Although
there is as yet no firm evidence to substantiate the claims concerning 'Chorar-
mia in the Kushan Empire' that are frequently found in the literature41and some
scholars locate the northern borders of the Kushans far to the south of Choras-
rnia,42 the Aral region and the lands between the lower Amu Darya and Syr
Darya - in fact the boundary areas - were still held by semi-sedentary and
nomadic Saka-Massagetae tribes under the influence of Chorasmian culture,

40. Gafurov, 1974, p. 61.

41. Gafurov, 1972, p. 151.
42. Masson, 1975, pp. 42-9.
 T H EN O M A D S
O F NORTHERN C E N T R AA
LSIA
A F T E R THE I N V A S I O N O F ALEXANDER:'

Y. A. Zadneprovskiy

Transoxania
A small part of Transoxania came under the rule of Alexander the Great after
his eastern campaign, but in Transoxania and in the steppes the nomads who
belonged t o various tribes of Sakas and Massagetae played the dominant polit-
ical role. Several important changes had occurred on the steppes of Eurasia. In
the west the Scythians were succeeded by the Sarmatians, while in the east a
strong nomadic power had emerged - the Hsiung-nu (Huns). This important
epoch in the history of the nomads of Eurasia is aptly named the Hunno-Sar-
matian period.
These events affected the historical fortunes of the nomads of Central
Asia. The third and second centuries B.C. were a long transitional phase for
these nomads, marked by major migrations and by the consequent emergence
of new groupings of tribes on the historical scene. The movement of the
nomads was a constant threat to the security of the Graeco-Bactrian kingdom.
The middle of the third century B.C. saw the rise to power of a group of tribes
consisting of the Parni ( A ~ a r n i and
) the Dahae, descendants of the Massagetae
of the Aral Sea region. They invaded Parthia, the older Achaemenid satrapy,
froin the north and took advantage of the weakness of the Seleucids to establish,
in 250 B.c., an independent Parthian state under the Arsacid dynasty (see Chap-
ter 5). This powerful nomad state, which lasted from the middle of the third
century B.C. t o the beginning of the third century A D . , became a dangerous
rival of the Graeco-Bactrian kingdom, the Seleucid state, the Romans and the
Kushan Empire. The Arsacids often recruited mercenaries from other nomadic
groups, which resulted in a constant replenishment and growth of the nomadic
population of the country. Nomad burial grounds investigated at a number of

:" See Map 9.

 Y. A. Zadneprovskiy

sites in the valleys of the Kopet Dag and Greater Balkhan mountains as well as
in the lower hills1 provide solid evidence of the permanent presence of llomadr
in northern Parthia, and of the major role they played in the life of that state,
They also show that the nomads had maintained their distinctive way of life and
culture.
Classical sources knew about the conquest of Bactria by nomadic tribes
(see Chapters 4 and 5), referring to them as peoples who came from beyond the
Jaxartes (now the Syr Darya). Strabo (X1.8.2) mentions four tribes: the Asii, the
Pasiani, the Tochari and the Sacaraucae. Writing of the event which followed,
Pompeius Trogus notes briefly that the Asiani, kings of the Tochari, laid waste
the Sacaraucae (Justin. Pvologus, XI.1).
Chinese chronicles merely recount the conquest of Bactria by the Yiieh-
chih from Central Asia, whereas other ancient sources mention several invading
tribes from beyond the Jaxartes. The juxtaposition of these two contradictory
accounts gives rise to the notion of a two-pronged invasion from the north and
west. But all attempts to identify the actual invaders have been disputed. Only
one thing is beyond doubt concerning this major event in world history. The
defeat of the rulers of Bactria was the work of the local nomadic tribes of Trans-
oxania as well as of tribes from northern Central Asia.
The principal sources for the history of these nomads are the Shih-chi
(Historical Records), by the Han-dynasty court historiographer Szi-ma Ch'ien,
the Han-shu (Annals of the Former Han), the H o u Han-shu (Annals of the
Later Han) and the Pei-shih (Annals of the Wei D y n a ~ t y ) Valuable
.~ but very
brief references are to be found in the works of ancient historians and geog-
raphers, such as Strabo, Ptolemy, Pomponius Mela, etc. The interpretation of
these sources continues to be difficult, but the growing volume of archaeolog-
ical evidence helps to clarify some issues.

The
The Chinese sources tell us of four major groups of nomads: the Wu-sun, the
K'ang-chii, the Yen-ts'ai and the Yueh-chih. The most detailed information
available concerns the Wu-sun. The second century u.c. saw the formation of
the Wu-sun, a tribal confederation in the north-eastern part of Turkestan (the
T'ien Shan mountains) and Semirechye. According to the Chinese sources, the
Wu-sun originally lived in Central Asia, together with the tribes of ~ueh-chih
and Hsiung-nu. We have the semi-legendary account that when the WU-sun
were defeated by the Yueh-chih, their leader was killed and some of the Wu-

1. Mandel'shtam, 197 1; Marushchenko, 1959.

2. Bichurin, 1950; Kyuner, 1961; Chavannes, 1905, p. 519, 1907, p. 1.19; HulsewC, 1979.
 The nomads of northcrn Central Arla

run, with the new-born son of this leader, obtained the of [he
~ ~ i u n g - n uLater,
. the WU-sun, now allied to the H ~ i u n ~invaded- ~ ~ , T'ien
shan around 160 B.C. and settled down in their newly conquered territories,
which became their second homeland. In course of time, when the Wu-sun had
become sufficiently powerful, they ceased to obey the Hsiung-nu. In 125 H.<:.
the Wu-sun were visited by Chang Ch'ien, the famous Chinese traveller and
diplomat, who had been sent to establish a coalition against the Hsiung-nu.
Chang Ch'ien recommended a plan to deal with the Hsiung-nu. The Chinese
were to make peace with the WU-sun, sealing the compact by marriage. The
plan was approved and Chang Ch'ien was sent as ambassador to the Wu-sun in
115 B.c., with the proposal that the Wu-sun should return to their original
homeland and attack the Hsiung-nu jointly with the Han. The Wu-sun ruler
was offered an alliance and the hand of a Han princess in marriage, but the Wu-
sun, wary of the Hsiung-nu, refrained from giving a final answer. After the Han
state had achieved considerable military victories in East Turkestan, the Wu-sun
finally concluded the marriage alliance with the Han princess; but the
Hsiung-nu also sent a princess to marry the lord of the Wu-sun, and she was
declared his senior consort, while the Han princess was only his junior wife.
The Han bride complained in verse of her destiny, mourning her enforced
union with a Wu-sun king whose abode was made out of felt, who ate meat and
whose drink was sour milk. In spite of this treatment the next ruler of the Wu-
sun was given another Han princess in marriage. After their defeat by the Yiieh-
chih and subsequent revenge, the Wu-sun had settled into their new homeland
and had become so strong that the Han state felt obliged to win their friendship
in an alliance, based on a royal marriage.
In the T'ien Shan region the Wu-sun were the first tribal group about
which substantial evidence is available. The Chinese sources refer to the Wu-
sun or nomad state. The Wu-sun were bounded by the Hsiung-nu to the east,
by the settled peoples of East Turkestan to the south, by Ta-yiian (Ferghana) t o
the south-west and by K'ang-chii to the west. Their federation included locally
conquered Saka tribesmen, as well as some Yiieh-chih. The question of the eth-
nic origin of the Wu-sun themselves remains debatable, and contradictory
hypotheses have been advanced. The one thing that is clear is that the majority
of the population consisted of linguistically Iranian Saka tribes.
The administrative and ~ o l i t i c acentre
l of the Wu-sun state was the walled
city of Ch'ih-ku, 'the City of the Red Valley', situated in the basin of the Issik-
kbl. Lying on one of the branches of the Silk Route, it was also an important
trade centre, but its exact location has not yet been established. The ~rincipal
activity of the Wu-sun was cattle-raising. They freelv wandered with their live-
stock seeking pasture and water, but the geographical conditions in Semirechye
and T'ien Shan did not allow constant wandering, and the economy of the Wu-
sun remained semi-nomadic, with the population moving from one climatic
zone to anorher with each change of season. They combined cattle-breeding
 Y. A. Zadneprovskiy

with agriculture, as is evident from archaeological finds of the Wu-Sun

from settlements in the C h u valley, the Issik-k01 basin and in eastern semi-
rechye. These contained the remains of pis6 dwellings, some with mud floors
and other built on stone foundations. Numerous querns and agricultural imple-
ments as well as bones of domesticated animals have been found, suggesting a
semi-nomadic pastoral economy.
The social structure of the Wu-sun followed the Hsiung-nu pattern. Their
ruler was the Great K'un-mo, whose power was hereditary. There was a fairly
developed administrative apparatus, consisting of sixteen officials. The ruler was
assisted by a council of elders, a body which t o some degree limited his power,
The Great K'un-mo and his two sons, the rulers of the left and right domains,
each commanded a personal force of 10,000 horsemen. There was also a regular
army and each freeman was considered as a warrior. The administrators and
members of the ruling nobility maintained themselves on the tribute paid to
them by conquered tribes, war booty and profits from trading activities. Tro-
phies acquired in wars, which were a frequent occurrence, were at times of quite
considerable value. They included large herds of cattle, abundant goods and
many prisoners. Most of this booty was shared by the ruling Clite and by the
privileged warriors of the king's guard, w h o amassed enormous riches.
The inequitable ownership of livestock and pasture inevitably resulted in
the concentration of power in the hands of the wealthiest and largest family
within the tribe. This in turn led t o social inequality, which is evident from both
archaeological finds and written records. The richest of the Wu-sun owned as
many as 4,000 o r 5,000 horses, and there is evidence pointing to the privileged
use of certain pastures. These factors created dissatisfaction, disputes and popu-
lar unrest, especially in the lower strata of Wu-sun society. The accumulation of
wealth by the dominant stratum led t o social stratification and to relations typ-
ical of early class societies, in which the patriarchal-clan order played a major
role. A manifestation of this was the widespread application of the custom of
the levirate, by which a widow was obliged t o remain within the late husband's
family, becoming wife t o one of his relatives.
Wu-sun society included slaves, most of them prisoners of war. One
report says that 10,000 persons were captured in one campaign against the
Hsiung-nu. Most slaves laboured as household servants though some worked as
craftsmen, but the principal producer was the freeman. The socio-economic
structure was similar t o that of the Hsiung-nu as an organized community of
nomads.
The Wu-sun played the part of a third force between the Hsiung-nu and
the H a n state. They were attacked around 80 B.C. by the Hsiung-nu and were
badly worsted. Their ruler, the reigning K'un-mo, Wu-ku-chi-mi, turned to the
H a n emperor for assistance and their joint force defeated the Hsiung-nu In
72 B.c., taking numerous prisoners and capturing thousands of horses, cattle!
camels and asses. This rich booty greatly strengthened the Wu-sun and gave
 The nomads of'north~rnCentral Asia

them great influellce over the political life of the settled oases of East Turkcstan.
The son of Wu-ku-chimi became the ruler of Yarkand, while his daughter was
given in marriage t o the lord of Kucha. The first century a,<:. was a period of
success and prosperity for them.
Little is known of the WU-sun during the early centuries of the Christian
era. u n d e r pressure from the Ju-jan, a new group of nomadic tribes from Ccn-
tral Asia, the Wu-sun were obliged to abandon Semirechye and seek refuge in
the T'ien Shan mountains. The last reference to the Wu-sun in the historical
sources is in A.D. 436, when a Chinese diplomatic mission was dispatched to
their country and the Wu-sun reciprocated. It is probable that by the middle of
the fifth century A.D., the Wu-sun, with other neighbouring peoples, had suc-
cumbed t o the Hephthalites.
The archaeological sites of the Wu-sun period (Fig. I ) which have been
explored in the regions of Semirechye and T'ien Shan are very varied and reflect
the ethnic heterogeneity of the population. Most cemeteries are burial grounds
with the dead interred in pit-graves3 of the Chil'pek group. They belong to the
local Saka population, which formed part of the Wu-sun federation4 and prc-
served the traditions, funeral rites and material culture of the earlier Sakas. A
second group consists of kurgans with burials in lined and 'catacomb' chamber
graves. Sites of this Aygirdzhal group are widely found from the first century
A.D. and are not basically a local type.5 For a long time, between the second cen-
tury B.C. and the fifth century A.D., these graves co-existed with burials of the
Chil'pek group, often within the same area, inside the same burial ground. The
lined graves and shaft chamber tombs were probably those of the immigrant
population, and there is reason to associate lined graves with the Yueh-chih.
It is clear that the tribesmen who migrated into Turkestan and their descen-
dants spent some 600 years living side by side with the indigenous population,
mingling with them to form some kind of unity. Typical of the time are burials in
simple grave-pits, which were sometimes covered with logs. The deceased were
supplied with a large quantity of utensils, roba ably containing milk and ~ i e c e of
s
mutton. They were buried with their personal adornments and articles of every-
day use. Some graves contained gold ornaments, but those of ordinary members
of the community were usually poor, with a uniform assortment of grave goods.',
The graves of the Aygirdzhal group often contain weapons.
Of particular interest is the Kargali burial of a female shaman discovered
in a gorge at an altitude of 2,300 m, near Alma-Ata, which contained many
items of jewellery, clothing and head-dress - a total of nearly 300 gold objects

3. Akishev and Kushacv, 1963; Bernshtam, 1949, 1952; Voevodski~.and Gryaznov, 1938.
4. Zadncprovskiy, 1971, 19756.
5. Akisliev and Kushaev, 1963; Zadneprovski!,. 1971. 19756; Kibirov, 1959.
6. Akishev and Kushaev, 1963; Bernshtam, 1952; Vaynberg, 1981.
 Fr<;. 1. Nomad culture of the Wu-sun period.

with turquoise inlay. A unique find was a diadem depicting animals, birds and
human beings, embellished with settings of carnelian, almandine and turquoise,
testifying t o the high degree of artistic skill of the ancient jewellers.' A rich
burial of Wu-sun times at Tenlik in eastern S e ~ n i r e c h ~contained
e the grave of a
high-ranking warrior whose clothing had been decorated with about 100 ski]-
fully wrought golden bosses. T h e wide distribution of such rich burials suggests
that Wu-sun society was stratified o n the basis of property ownership.
 Thc nomads of northern Ccntral Asia

The K'ang-chii
The nomadic federation of the K'ang-chli was the second great power
yueh-chih in Transoxania. According t o the Chinese sources, KPang-ch" lay
north-west of Ta-yuan and west of the Wu-sun, bordering upon the Yiieh-&ih
to the south. T h e territory of the K'ang-chu, therefore, covered the region of
the ~ a s h k e n oasis
t and part of the territory between the Amu Darya and Syr
Darya rivers, with its heartland along the middle Syr Darya. It seems t o have
emerged as a powerful state in the second century B.c:. As the historians of
~ l e x a n d e rd o not refer t o the existence of any political confederation on the
Jaxartes (Syr Darya) except Chorasmia, the K'ang-chu must have appeared a lit-
tle later. They united a number of regions which had sedentary, agricultural and
nomadic populations.
T h e K'ang-chu were inevitably affected by the events of the mid-second
century B.c., when the Central Asian tribes invaded Graeco-Bactria. The migra-
tion of the nomadic peoples (the Asii, Tochari, etc.) to the south altered the
balance of power in the valley of the Syr Darya. Taking advantage of these
circumstances, as the Hou Han-shu suggests, the K'ang-chu subjugated Yen-
ts'ai in the region of the Aral Sea, and the still more remote land of the Yen in
the southern Urals. Yen-ts'ai is identified with the large confederation of Sar-
matian tribes led by the A o r s i RThus, K'ang-chu established direct contact with
the Sarmatian world t o the north-west. The expansion of K'ang-chu in this
direction in the first and second centuries A.D. was occasioned by the rise of the
~ o w e r f u lYueh-chi11 confederacy (subsequently the Kushan Empire) t o the
south and b y the presence in the east of the formidable Wu-sun state allied with
the Hsiung-nu and the H a n Empire. The Chinese sources inform us that K'ang-
chu was tributary t o the Yueh-chih in the south and t o the Hsiung-nu in the
east. T h e north-west advance of K'ang-chu and its conquest of Yen-ts'ai appa-
rently obliged some tribes of the Aorsi, and later of the Alans, t o move west; it
may, therefore, be concluded that K'ang-chu played a major historical role in
the initial stages of the Great Migration of Peoples, which was such an impor-
tant event in world history. In this way, K'ang-chu gained control over the
northern sector of the international trade route known as the Northern Route.
In endeavouring t o maintain its influence over the southern portion of
this route, K'ang-chii pursued an active policy in the east and south-east, allying
itself in 101 B.C. with the Ta-yuan, and helping them t o preserve their indepen-
dence against the Han. During the course of its continued struggles against the
W U - ~ K'ang-chii
~ ~ , sought assistance from Chill-chih, ruler of the northern
Hsiung-nu, in the illiddle of the first century B.C. Initially ~ h i h - c h i h ' sarmy
penetrated deep into the country of the Wu-sun and besieged their capital in

8. Matsulevicli, 1947.
42 s.c-:But the H:I~state intervened and defeated and killed Chih-chih at Talas
in 36 B.C. The K'ang-chu ruler was ohligcd to send his son as a hostage to the
court of the Han emperor as a token of his submission. All these events in the
campaigll against Chih-chih are related in colourfbl terms in The Ltfe History
Ch 'eng-t'ang.'
Undaunted, K'ang-chu continued to pursue an independent policy.
maintained its independence up to the end of the third century A.D. and con-
tinued to send embassies to the Chinese court. Convincing evidence of its
independent status may be ieen in the coinage it issued in the second and third
centuries. During this period the K'ang-chu rulers at Chach (the Tashkent
oasis) began to issue their own currency,Io similar t o some of the early coin
issues of ancient ~ h b r a s m i a . Soon afterwards the fortunes of K'ang-&~
declined and it was absorbed into the Hephthalite state - a fate which it shared
with the other states of Transoxania.
The Han-shu describes the typically nomadic way of life of the K'ang-chii
ilite and particularly of its sovereign, who spent his winters in the capital, the
city of Pi-t'ien, and his summers at his steppe headquarters, situated seven days'
journey away o n horseback. The ruling nucleus of K'ang-chu consisted of
nomadic tribes whose customs resembled the Yueh-chih. Excavations at archae-
ological sites (Fig. 2) associated with the KYang-chunomads reveal their role in
the state. In the burials at Berk-kara and Tamdi, the dead were placed in pit-
graves, sometimes covered over with logs, under kurgan mounds. The graves
contain hand-made pots, iron swords, arrow-heads and some simple jewellery,
and belong to KYang-chiitribes of the early period, their traditional culture
exhibiting traits characteristic of the Saka tribes as a whole.
From the beginning of the Christian era, burials in 'catacomb graves' (in
shaft- and chamber-tombs) became widespread, as we see in the burial grounds
of the Kaunchi and Dzhun cultures covering the period from the first to the
fourth century and accepted in scholarly literature as the remains of the K'ang-
chu tribes.
A considerable level of sophistication distinguished the culture of the set-
tled agriculturalists of K'ang-chu, as we see from brief references in written
sources and in archaeological finds such as the burial sites of the Kaunchi and
Dzhun cultures of the Tashkent oasis and the middle Syr Darya, of which some
seem to belong to the sedentary farming population.

9. Taskin, 1973.
10. Buryakov, 1982.
 The nomads of northern Cenlral Asia

FIG.2. Nomad culture of south Jiazakl~stan(third to first century B.c.).

The Yen-ts'ai, Aorsi and Siraci

The third major nomadic state, that of the Yen-ts'ai, was situated in north-
western Central Asia in the steppe around the Aral Sea and the northern shores
of the Caspian, where it was in contact with the world of the Sarmatians. The
nomadic population of this region belonged to the Sarmatian group of tribes
whicjl replaced the Scythians around the turn of the third century a.c.ll During
the second century B.C. a new major grouping of Sarmatian tribes, of which the
chief were the Siraci and Aorsi, appeared on the steppes between the Caspian
and the Tanais (the River Don), as Strabo describes. Abeacus, King of the Siraci,
could mobilizc 20,000 horsemen (at the time when Pharnaces was lord of the
Bosporus), while Spadinus, King of the Aorsi, commanded as many as 200,000
and the Upper Aorsi had even more. That explains their camel caravan trade in
Indian and Babylonian goods which they procured by barter from the Arme-
nians and the Medes (Strabo XI.5.8).
It is evident from this text that the Aorsi and their kinsmen, the Upper
Aorsi, were tribes of Sarmatian origin and were masters of the lands lying along
the coast of the Caspian Sea. The precise eastern boundaries of the Aorsi are un-
known, but their influence probably extended to the Aral Sea. They were a
great military power and for almost three centuries, until the arrival of the
Alans, they played a major role in events of the northern Pontic region. King
Eunonus of this tribe was an ally of Mithradates VIII (A.D. 40-44) in his strug-
gle against Rome, and offered him asylum after his defeat.
Strabo refers to the established international trade links of the Aorsi with
the states of the Caucasus. They also controlled trade routes leading from the
Bosporus and other Black Sea states to Transoxania and China. According to
Chinese sources, one of the branches of the Silk Route - the Northern Route -
passed through East Turkestan, Ta-yuan and K'ang-chu, ending in the country
of Yen-ts'ai. Chinese artefacts from archaeological excavations ~ r o v i d econcrete
evidence of the use of this route during the first few centuries A.D.
Scholars generally identify the Aorsi mentioned by classical writers with
the Yen-ts'ai state of the Chinese sources.

The Shih-chi states that Yen-ts'ai lies almost 2,000 1i north-west of K'ang-chii, and
it is a nomadic country whose custoins are like those of K'ang-chii. Its army num-
bers over 100,000. It lies on a large lake that does not have high banks - the North-
ern Sea.'?

This independent nomadic state played a role of some significance in the history
of Transoxania and the neighbouring localities along the international trade
routc. It is not, therefore, surprising that the H a n Empire should have sent
embassies there and fostered trade relations. Eventually, in the first century B.c.,
Yen-ts'ai lost its independence and became a dependency of K'ang-chi].
According to the HOUHan-rhu: 'The domain of Yen-ts'ai was renamed
A-lan-ya, over which K'ang-chu held sway.'13 Another country to lose its inde-

1 1. Harmatta, 1950.
12. Bichurin, 1950; Kyuner, 1961.
13. Bichurin, 1950; Kyuner, 1961; Hulsewi., 1979.
 The nomads of northern Central Asia

pendence was Yen, which paid tribute in furs. Many scholars seek to identify
A-Ian-ya (or A-lan-Lao) with the Aorsi and Alans of thc ancient sourccs. It
should be noted that the appearance of the name A-lan-ya in the How Han-rhu
coincides with the emergence of the Alan tribes on the political stage.

The Alans
At the beginning of the first century A.D. the Alans secured a dominant position
among the Sarmatians living between the Caspian Sea and the River Lion.
According t o Ammianus Marcellinus, they were descended from the Massage-
tae. The ~ e o p l eof the Alani in the first to third centuries A . D . represented a
powerful force with which the Roman Empire was obliged to reckon. They fre-
quently threatened Rome's more remote possessions along the Danubc and in
Asia Minor, and were successful in penetrating the Caucasus. They also waged
successful warfare against Parthia. Historical and archaeological evidence ena-
bles us t o link Yen-ts'ai (the Aorsi), A-lan-ya (the Alans) and K'ang-chii with
the Iranian tribes with whom, as the Chinese chronicles state, they had tics.
They had similar dress and identical customs. This cultural affinity can also be
traced in burial sites that have been excavated along the lower Volga, in the
southern Urals, in the Tashkent oasis and along the middle Syr Darya.
In the Aral region a considerable number of heterogeneous burial sites of
the nomads have been discovered. Distinctive circular-plan n~ausoleumswith
cruciform interior layouts along the lower reaches of the Syr Darya were places
of multiple interconnected burials. The sites in question, Chirik-Rabat, Babish-
Molla and Balanda, date from the fourth to second century B.C. and were built
by the Apasiacae tribes.'"he Dzheti-Asar burials in the basin of the Kuvan
Darya, a tributary of the Syr Darya, unusual kurgans with round and rectangu-
lar ground-level chambers built of raw brick, are attributed to Strabo's Tocha-
roi.15 Lined kurgans were used for burials by nomads on the left bank of the
Amu Darya from the fourth century B.C. The number of such burials increased
during the last centuries B.C. and the first two centuries A . D . ' ~ At a certain stage,
a change occurred and the dead were buried with the head ~ o i n t i n gsouth, as in
the contemporary Sarmatian burials in the southern Urals. These sites have been
tentatively attributed to the Yiieh-chih group of tribes. Similar lined kurgans of
the fourth t o second centuries B.C. have survived in the south-east part of the
Ustyurt plateau in the Aral region, where sites similar t o the Late Sarmatian
complexcs have been discovered.

14. Tolstov, 1962.

15. Ibid.
16. VaynL>erg, 198 1; Lokhvits and Kliazanov, 1979.
 The complicateci palaeo-ethnographic character of nomadic settlementin
the Aral Sea region during the period under consideration is thus reflected in
the archaeological finds, which show the successive replacement of one group of
nolnads by another. What needs to be stressed again, however, is the tentative
character of all the ethnic (tribal) identifications.

The Hsiung-nu (Hunni, Huns)

An important role in the political history of Central Asia was played by the
Hsiung-nu (Hunni o r Huns) at the turn of the first century B.C.They were first
involved with the affairs of the Wu-sun and the K'ang-chu, but there is hardly
any reliable evidence of their presence in the lands where the Wu-sun and
K'ang-chu lived. Of particular interest are reports about the small country of
Wu-shan-mu," which lay between the lands of these two tribes. Wu-shan-mu
had close ties with the Hsiung-nu. Hu-lu-ku, ruler of the Hsiung-nu (96-85
B.c.), arranged a marriage with the family of the ruler of the Wu-shan-mu,
establishing blood ties between the two states. In 60 B.C. Ch'i-hou-shan, son of
the ruler of the Hsiung-nu, having failed to inherit the throne, fled to the court
of his father-in-law in Wu-shan-mu, who played a decisive role in the election
of Ch'i-hou-shan to the position of ruler of the Southern Hsiung-nu in 58 B.c.,
at the time of the division of the Hsiung-nu into two mutually hostile king-
doms. Wu-shan-mu, a minor power, could have played such a role only with
the support of a Hsiung-nu military force. As we have seen, the Hsiung-nu first
emerged into historical prominence at the beginning of the first century B.C. It
was in the year 60 B.c., in connection with the election of Hu-han-yeh as their
ruler, that the armed detachment of Chih-chih, leader of the Northern Hsi-
ung-nu (enemies of the Southern Hsiung-nu), appeared in K'ang-chu. Chih-
chih entered into an alliance with the K'ang-chu against Wu-sun, the lord of
K'ang-chu, giving his daughter in marriage to Chih-chih while he himself mar-
ried Chih-chih's daughter. Their joint forces then attacked the Wu-sun, reach-
ing the Wu-sun capital in 42 B.C. After so much success, Chih-chih began to
demand tribute from Ta-yuan and other kingdoms. But later, when the allies
had quarrelled, Chih-chih was attacked and killed by a Chinese force.Ix The
Northern Hsiung-nu continued to move south, and their numbers increased
considerably following their defeat by H s i e n - ~ i .It is with this mass migration
that the emergence of the new Kingdom of Yueh- an in the Lake ~alkhash
region is associated.

17. Bichurin, 1950.

18. Taskin. 1973.
 The nomads of northern C r n t m l Asia

The Yueh-pan
The Chinese chronicle Pei-shih informs us that the territory of the Y o ~ h - ~ ~ n
was formerly a possession of the Hsiung-nu, crushingly defeated by the Chi-
nese. The Northern Hsiung-nu retreated west to K'ang-chi, while part of the
local population (some 200,000) remained to form the Kingdom of Y i i ~ h - ~ a n . ~ ~
In the course of the first century ~ . c . / f i r s tcentury A.D. the Hsiung-nu
graduillly became masters of the steppe regions north of the Syr Darya. Unlike
the Southern Hsiung-nu, who became subjects of the Han emperors, the
Northern Hsiung-nu remained independent and grew so powerful that even-
tually, under Prince Hu-yen (A.D. 123-35), they could establish a vast domain
of their own.

Later history
The long occupation of parts of the Central Asian region by numerous North-
ern Hsiung-nu tribes has left archaeological evidence behind. The finds at
Dzheti-Asar o n the Syr Darya include clay pots whose shape closely resembles
that of the typical bronze cauldrons of the H s i u n g - n ~ suggesting,
,~~ with other
remains, the influence of Hsiung-nu culture on the population of Dzheti-Asar.
It could equally be suggested that these tribes formed part of the peoples known
as the White Huns, o r Hephthalite~.~' 'Catacomb' (shaft-and-chamber) tombs
dated between the first and fourth centuries A.D. have been excavated at the
burial ground of Kenkol (Fig. 3) in the Talas valley.22The dead were placed in
wooden coffins and their grave goods included weapons and a bow of dis-
tinctive Hsiung-nu type with bone arrow-heads and wooden vessels. All these
material objects were typical of the Hsiung-nu2?while the earthen pots and arti-
cles of everyday use were the work of local tribes. The actual form of the grave
structure was different from the common Hsiung-nu types and the ~ e o p l e
buried at Kenkol were of two distinct racial types. The majority turned out to
be of the Pamir-Ferghana Europoid type. Others were E u r o ~ o i d salbeit
, with a
significant Mongoloid admixture.?' The ethnic (tribal) identification of Kenkol
is a matter of controversy. Some scholars attribute it to the Hsiung-nu and
others t o local tribes.'5 Since Kenkol is situated in the Talas valley, in the eastern

19. Bichurin, 1950.

20. Levina, 1966.
21. Tolstov, 1962.
22. Bernshtam, 1951; Kozhomberdiev, 1963; Sorokin, 19566.
23. Bernshtam, 1951; Mandel'shtam, 1975n.
24. Ginzburg and Trofimova, 1972.
25. Bernshtam, 1951; Kozhomberdiev, 1963; Harmatta, 1952.
 Y. A. Zadnrprovskiy

FIG.3. The Kenkol culture.

part of what was once K'ang-chii, there are reasons for taking it t o be a K'ang-
chii site which reflects Hsiung-nu influence on the local K'ang-chii populace.
A similar pattern emerges with the kurgans (Fig. 4) in the area of Char-
dara along the middle Syr Darya. O n e of the local graves contained a bronze
cauldron and earthenware pots similar to those of the Hsiung-nu.zf1Here too,
however, the majority of the people were of K'ang-chii stock. It may be noted,

26. Maksimova et al., 1968.

 Thc nomads oj'northert~Ccntval A,ia

FIG.4. N o m a d culture of south Kazakhstan and the Tashkent oasis

(first t o fourth century A.D.).

in conclusion, that the nomad tribes of the Yiieh-chih, w h o constituted the

most powerful force in Transoxania in the second and first centuries B.c.,
played a specific and decisive role in the emergence of one of the most for-
midable powers of the ancient world, the Kushan Empire. The burial grounds
excavated in northern Bactria - Tulkhar, Aruktau, Kokkum and Babashi2- -
 Y. A. Zadncprovskiy

seem to have belonged to the Yiieh-chih, and similar nomad burial sites
explored in T'ien Shan,'$ Semirechye,?' Sogdiana30and Ferghana" can reasonably
be used as evidence for the southward migration of the Yueh-chih tribes
towards Bactria and India.'?
These historical and archaeological data shed interesting light on the role
of the Transoxanian nomads in the history of both Central Asia itself and the
world at large. It was as a direct result of their enterprise and warlike activities
that the new strong states of Parthia and the Kushans, the major powers
K'ang-chii, Wu-sun and later the H e ~ h t h a l i t eEmpire emerged. The nomads of
Central Asia also played a key role in the Great Migration of Peoples. They
contributed much to the interchange of cultural achievements between the civil-
izations of the ancient world and equally exercised considerable influence on
the development of the sedentary cultures of the East, especially in matters of
warfare and the arts. The Central Asian nomads of antiquity did indeed leave
their indelible mark on history.

28. Bernshtam, 1952; Kibirov, 1959.

29. Akishev and Kushaev, 1963; Beriishtan~,1951.
30. Obel'chenko, 1961.
3 1. Baruzdin, 1962; Kozhombel-diev, 1977; Lokhvits and Khazanov, 1979.
32. Z a d n c p r ~ v s k i ~1975a.
,
 B. A. Litvinsky,
with the contributions of M. Hussain Shah
and R. Shabani Samghabadi

The decline of Parthia

Parthia was one of the great empires of the ancient world, a rival of the Roman
Empire and the Kushan kingdom. But by the beginning of the second century
A.D. all that was a thing of the past. The last hundred years in the life of Parthia
was a period appropriately described as the 'downfall of the Parthian Empire'.'
The period began with the reign of Vologases 11, who ruled until A.D. 146/7.' At
that time, the Parthian state was anything but unified; in particular, to judge
from the numismatic evidence, Iran was at that time under the rule of his rival,
Mithradates IV (c. A.D. 13047). Parthia was also weakened by its conflict with
the Alani (c. 136). During the war between the two peoples, a Parthian army of
20,000 foot-soldiers had fallen into a trap and were only saved by heroic fight-
ing. Nevertheless, the Parthians were forced to retreat and this left the way into
Mesopotamia open for the Romans. Had the Alani not been attacked simul-
taneously by another people, the Parthians would have been annihilated."
Then began the long reign of Vologases I11 (A.D. 14718-190/1). At first,
relations with Rome under Antoninus Pius were comparatively peaceful. I t was
only after the emperor Marcus Aurelius came to power that a new round of
wars began between Parthia and Rome. Initially, good fortune was on the side
of the Parthians. They captured Edessa and invaded Syria, where their detach-
ments wreaked havoc and destruction, but then the Romans took the initiative.
Their army invaded Armenia and Mesopotamia. After a bloody siege, Dura-
Europos fell. This was followed by the capture of Seleucia, which was burnt to

1. Debevoise, 1938, pp. 240 et seq.

2. Here and elsewhere, LitvinskY adheres to the syitern of Late Parthian genealogy and
chronology suggested by Bivar, 1983b, pp. 92-9; see also Selln~ood,1967.
3. Debevoisc, 1938, pp. 242-3.
 of' A4. I-lussnin Shclh nnd
k J ~ ,rhe contributio)~~
8, ,-\, l , i r ~ ~ l i ~ l j,zith H. Shabcrni Samghdbndi

the ground during an uprising of its inhabitants. The Parthian capital Ctesiphon
was taken, and the palace of Vologases was destroyed (Dio Cassius LX~1.2.3;
Lucian, Bis accusatus 2). In the view of present-day scholars, the defeats of the
Parthians were to a large extent due to an epidemic of smallpox that was raging
at that time. It seems to have begun in Southern Asia, spread to China, and then
taken hold of the Kushan Empire, possibly during the reign of Kanishka. After-
wards, it spread along the Silk Route to the Persian Gulf and Red Sea. In A.D.
165 Roman detachments entering Ctesiphon picked up the disease, which then
spread into the Roman Empire..' Evidence of the scale of the epidemic can be
seen in the fact that in some parts of the Roman Empire over a quarter of the
urban population was wiped
This, however, did not prevent the Roman armies from continuing their
victorious campaign against Parthia. Osroes, the Parthian general, was saved by
swimming across the Tigris. Vologases I11 was replaced by Vologases IV (A.D.
191), who was in turn succeeded by his son, Vologases V (A.D. 20718-22112).
The Romans controlled a large part of Mesopotamia; Armenia and Osroene
were turned into Roman provinces; the territory west of the Khabur river
remained a permanent part of the Roman Empire; and Carrhae and Edessa were
also in the Roman sphere of influence. There were repeated uprisings in Parthia
by local rulers who were bent on achieving independence. A major uprising
occurred in A.D. 196 when the royal forces were surrounded by the insurgents
in a valley in Khorasan and forced to flee to the mountains, where they lost
many soldiers and a considerable amount of equipment. Reorganizing their for-
ces, the royal troops put down the insurgents but paid a heavy price. The Par-
thian kingdom was now rent by internal divisions: the king's brother, Artaba-
nus V (Ardavin in Middle Persian), emerged as an independent ruler between
approximately 213 and 224. His rule was centred on Media, and from there he
mounted his campaigns into the neighbouring provinces. His ~ r i n c i ~ success
al
was the capture of Susa where, in 215, he ruled as a king assisted by Xwisak as
governor." In 216 the Roman emperor Caracalla perfidiously attacked the
troops of Artabanus V (who managed to escape capture), destroyed a large part
of Media, and had the graves of the Parthian kings dug u p and their bones scat-
tered (Dio Cassius LXXVIII, pp. 1 et seq.). And although two years later the
Parthians in turn were victorious, the years of the Parthian kingdom were num-
bered.
Under Antoninus Pius (A.D. 138-61), H ~ r c a n i a nand Bactrian ambassa-
dors arrived in Rome and probably engaged in negotiations directed against
Parthia. It is ~ o s s i b l ethat by that time Hyrcania was no longer dependent on

4. Bivar, 1970, pp. 19-2 1 .

5. Bivar, 198317, p. 94.
6. Henning, 1952, p. 176; Bivar, 1983b, p. 95.
 Thc risc of .Sasanian Iran

the central government of Parthia (Sextus Aurelius Victor, Epitome 15.7). Iran-
~ h a h then
r consisted of small domains, only nominally subordinate (and some-
times not even that) t o a central ruler. At the head of the state stood the 'King of
Kings'. His personal domain covered only a very small part of the Parthian state
- ancient ~ e d i and
a the adjacent lands. Over these he reigned supreme. Within
his domain there were small provinces or patk6s (administered by a patk6rpdn)
and towns with surrounding districts administered by officials with the title of
= xiahrap. T h e capital was also located here. Beside the royal domain there
was a whole system of semi-independent or virtually independent domains (birr
= xiahr), each ruled over by a king (MLK' or hitrdr) who was usually the repre-
sentative of a local hereditary dynasty. These domains, in turn, were made u p of
smaller estates.'
T h e structural instability of the Parthian kingdom became even more pro-
nounced in the second and early third centuries A.II. and led to the development
of centrifugal tendencies that were exacerbated by the relcntless attacks of
Rome in the West and the pressure of the Kushan Empire in the East. All this
led t o economic stagnation. Parthia's ~ u b l i cand social life also suffered from
the same sort of problems - the crisis of a slaveholding society - that beset the
countries of the Mediterranean in the second and third centuries A.D.; the imba-
lance in the social structure of its society was ~ r o b a b al ~major element in the
ultimate fall of the Parthian Empire. At the beginning of the third century, the
state had largely disintegrated and the downfall of Parthia was imminent.

The new Sasanian dynasty

The birth of the new Sasanian dynasty and, more important, the beginning of a
new epoch in the history of Iran were linked with Pars (called Persis by the
Greek author^),^ the small state in the south-west of the Iranian uplands
(= modern Fars). There are several versions of the events surrounding the origin
of the S a ~ a n i a n sT. ~h e sources tell us almost nothing about the history of Pars
itself during the first centuries of the Christian era. King Papak, who usurped
the crown of the Pars rulers, played a major role in unifying the land." H e
apparently had t o wage a difficult struggle against the central Parthian govern-
ment. Shortlv after Payak's death, his adopted son Ardashir became King of
Pars (c. A.D. 216). According t o one group of sources, he was a descendant of
Sasan, and this gave the Sasanian dynasty its name.
Ardashir organized a coalition that annihilated the Parthian army at the

7. Widengren, 1956; H ~ r m a t t a ,1957. 1958; Lukonin, 1961.

8. D'vakonov, 1961, pp. 230-1.
9. Lukonin, 1961, pp. 12-1 5; Christensen, 1944; Taqizadeh, 1943-46.
10. Tab.~ri,1879, pp. 6-7.
 R, .4, /2;r~z~i,lskj~,
,;l,th rhc co,ztvibutions of' M . Hussain .Shah and K. Shabani Samghabadi

FIG.1. Naqsh-i Rustam. The investiture of Ardasliir I. (Photo: UNESCO/Almasy.)

battle of Hormizdagan, where thc last Parthian ruler, Artabanus V (Ardavin),

was killed. T h e exact date is not known. Depending o n the computation system
used, Artabanus' death occurred in either 223 o r 226, and Ardashir ascended to
the throne in either 224 o r 227 (Fig. 1). But that was only the beginning.
Ardashir then had t o overcome the resistance of the rulers of many provinces
before he finally united the whole of Iran under his authority, and placed mem-
bers of the royal family in charge of different parts of the country. O n e of his
sons, called Ardashir, was installed in the province of K e r n ~ a n At
. the same time
many of the former leading aristocratic families retained their power, as the
royal Sasanian inscriptions show. King Ardashir brought undcr his control a
large number of Mesopotamian principalitics in the wcst u p t o the borders of
the Roman Empirc in Syria and Asia Minor. H e seenls t o havc cstablishcd the
Arabian principality of the L a k h ~ n i d sat Al-Hira, which protected Mesopotamia
f r o m the raids of thc nomadic Arabs.
 The rise of Sasanian Iran

Tabari describes Ardashir's conquests in the east as follows:

He then left Sawid for Istakhr, thence went first to Sistan and then to Gurgnn,
~barshahr,Merv, Balkh, and Khwirizm, right up to the very borders of the coun-
try of Khorasan, whence he again returned to Merv. After he had killed many pco-
and sent their heads to the Anahid Temple of Fire, he returned to I'irs from
Merv and settled there again. The ambassadors of the kings of Kushan and of
Turin and Mukrin came there and paid homage to him.

olde eke, w h o published this text, suggested that it contained many exaggcr-
ations," but Herzfeld produced evidence substantiating its reliability and auth-
enticity." This tallies with the conclusions reached by other scholars, such as
Maricq," GhirshmanI4 and Harmatta.I5 Reviewing the events that occurred in
the second half of Ardashir's reign, around A.D. 232, Harmatta concluded that
Tabari's report reviewed the campaigns of Ardashir from the political and stra-
tegic points of view in such a convincing order that its correctness as a whole
can hardly be doubted.
O n the other hand, serious doubts have been expressed about the reliabil-
ity of Tabari's account by other scholars.Ih Tabari's facts about the campaigns,
which are corroborated by Moses of Khorene (Movses Xorenac'i), apparently
merit credence, but we d o not know how serious the consequences were. There
is no doubt that Merv was captured and that Ardashir, the King of Kings, in-
stalled either his brother o r son, also named Ardashir, on the throne there.
Fryel7 concludes that we can only speculate that Merv was the outpost of the
empire in the north-east since neither Sogdiana nor Khwarizm (ancient Choras-
mia) are mentioned in any source as ruled by Ardashir. It should be added, as
Harmatta noted, that the Kushan kingdom could have also recognized ( ~ e r h a ~ s
nominally) the suzerainty of the ~ o w e r f u lSasanian King of Kings.

The campaigns of Shapur I

After Ardashir's death, his son Shapur I ascended the throne; his reign began in
'
the year A.D. 239 o r 241 and ended in 270 o r 273 (the period of his rei,en is
usually given as 242-72). Under Shapur, there was a serious dispute with the
Roman Empire. Rome viewed the new Sasanian state as a dangerous enemy and

11. Tabari, 1879, pp. 17-1 8, nn. 3-5.

12. Herzfcld, 1924, pp. 36 et seq.
13. Maricq, 1953, pp. 106 ct seq.
14. Ghirshman, 1947.
15. Harmatta, 1965, p. 190.
16. Tabnl-i, 1879; Lukonin, 19691, pp. 22-7.
17. Frye, 1983, p. 124.
 s k ~the, contributions of M. Hwssain Shah and R. Shabani Samghabadl
R. A, ~ ~ ~ v i nwith

FIG.2. Naqsh-i Kustam. The victory of Shapur f over the Roman emperor Valerian.
(Phota: UNESCOIMalval.)
tried to take the initiative, but in 244 Shapur routed the Roman army on the
Euphrates. The Roman emperor Gordian fell in battle or was killed by h'is own
troops; the town near the scene of the battle was given the name of Per6z-
Sapfir, or 'Victorious is Shapur'; and the Romans paid a tribute of 500,000 gold
dinars. Other wars with the Romans ensued and resulted in the Sasanian capture
of Syria and part of Asia Minor. In the battle at Edessa (260), the Roman empe-
ror Valerian was taken prisoner along with his army (Fig. 2). The details of the
battle are unknown, but in its wake the Iranian army captured thirty-six towns
. ~ ~ before had the Roman army suffered such a defeat. Sha-
and f o r t r e s s e ~Never
pur's victories in the west demonstrated the power and stability of the young
state. In battle, Shapur I showed himself to be a brilliant strategist, an intelligent
and bold statesman.lYIn honour of his victory, temples were built and rock

18. Noldcke, 1887, p, 93.

19. Lukonin, 1969a, p. 59.
 T ~ rise
P of Sasanian Iran

carved. By his defeat of the Romans, Shapur I was able to gain a firm
foothold in Armenia and Georgia. He also conducted an active i n the
east, winning a number of victories there (see below), and took the title of 'King
of Kings of Iran and non-Iran'.
O n his death in A.D. 270, Shapur was succeeded in turn by Hormizd
(270-71), Bahram I (271-74), Bahram I1 (274-93), Bahram 111 (293) and Narseh
(293-302).~' By then the position had changed. Iran had suffered setbacks in its
foreign ~ o l i c yin the west, and Rome had consolidated its power and influence
in the east. Shapur I was an outstanding ruler. Apart from his personal qualities
as diplomat and general, he played a major role in strengthening and consolidat-
ing the new Sasanian state. H e initiated a series of appropriate changes in social
and economic structures, which still bore a feudalistic imprint. There was a
marked expansion of the royal domains and many local dynasties were sup-
planted b y members of the Sasanian royal family. The state was substantially
centralized, a process that considerably enhanced its economic and military
power. A t this period Zoroastrianism was made the state religion and its priests
and temples became the mainstays of state a u t h ~ r i t y . ~ '

Kushanshahr
Shapur's conquests in the east are worthy of special mention. They followed u p
the military activities of Ardashir discussed above. The Arbela Chronicle
describes one of Shapur's campaigns against the Chorasmians, the mountain
tribes of Media and Atropatene (Azerbaijan) and other eastern tribes (the Gila-
nians, the Dailamits) in the very first year of his reign. Shapur is credited with
the founding of the town of ~ ~ v - S i p (medieval
ur Nishapur).?' But what is most
significant is the passage in the inscription o n the KaLbeof Zoroaster at Naqsh-i
Rustam which relates that among other ~ r o v i n c e s(iahr) Eriniahr included
'Kuiiniahr forward u p t o Pikbwr and to the border of Kai, Sugd and t i t ' .
Pikbwr, o r P u r u ~ a p u r a(Peshawar), was the capital of Gandhira." Preceding the
enumeration of the provinces (including Kushanshahr) in the inscription are the
words 'I possess', and following the enumeration the sentence 'The) all paid us
tribute and were subject t o us.' This gave rise t o a major debate," with several

20. W e use the chronologJ. proposed by Frye, 1983, p. 178. O n the histor!-. see Lukonin,
1979.
21. Lukonin, 1969a, pp. 62-90.
22. Markwart, 193 1, pp. 12, 52.
23. Harmatta, 1969.
24. Lukonin vigorously contested the authenticity of the information about Shapur's
inscription o n the Knlbc of Zoroaster. Howercr, his argulnents were convincingly
refuted by Harlnatta (1969, pp. 385, 486-90). See also Livshits. 1969, p. 56 and Gafuror.
1972, pp. 153-4.
 R, A. [-;rv;nskJ,, z i r h rhe conrributions of' M. Hussain Shtzh a n d R. Shabani Samghabadi

scholars, for a variety of reasons, refuting o r expressing doubt about the clailn
that Shapur I had conquered K ~ s h a n s h a h r . 'The
~ campaign itself possibly took
place between 245 and 248. The inscription of Narseh in Paikuli (c. 293) men-
tions 'Kuiiniah'. Apparently between the years 330 and 340 the Kushan lands
in the northern part of the Kushan Empire were under the suzerainty of the
Sasanians, but we know nothing about the form of this dependence. Henning
was inclined to the view that one of the Middle Persian inscriptions of visitors
at ancient Termez dated back to the year 264/5.2"
O n the eve of Ardashir's conquest (230), the Kushan kingdom, according
to Chinese sources, covered a vast territory. The Wei Lio (History of the Wei
Dynasty) informs us that the Kingdom of Kabul (Kao-fu) and the Kingdom of
India (T'ien-chou) were both dependencies of the Ta-Yiieh-chih.*' What then
became of the Kushan Empire after the Sasanian conquests in the east? Accord-
ing t o one hypothesis, formulated in an exceedingly rigorous manner by Har-
inatta, Ardashir mounted his campaign at a date (which Harmatta calculated as
A.D. 233) when the Kushan kingdom was weakened by being divided into two
parts, one ruled by Visudeva 11, the other by Kanishka 111. This marks the
beginning of the Late Kushan era, which was also used in the Tochi valley
inscriptions. For a number of reasons, HarmattaZ8suggests that it was Vasudeva
I1 w h o submitted following Ardashir's successful invasion, and that, conse-
quently, it was Viisudeva I1 w h o ruled the western part of the Kushan Empire
(later Kushanshahr) while Kanishka I11 ruled the eastern part (Gandhira and
the Panjab). Harmatta's position is highly consistent and carefully argued. It is
shared by other scholars, among them B i ~ a r . ~However,
' the situation is not as
simple as this might suggest. First the question of terminology must be clarified.
The existence of a separate ruler (Visudeva 11) is called into question by several
scholars. O n e view is that there was only one Visudeva, whose coin type was
changed in successive issues.jO Kanishka I11 is called Kanishka I1 by other
researchers." Zeimal also recognized that the supposed synchronism between
the coins of Visudeva I1 and Kanishka I11 presupposes that they ruled at the
same time, although apparently over a different territory; in other words, at
some point during the first half of Visudeva's reign the Kushan kingdom was
divided in two.'' However, Zeimal, like Gobl, dates the reign of these kings to a
much later period, a century o r more after the campaign of Ardashir I. So long

25. F o r a new rendering of this passage of the inscription by P. 0. Skjaervoe, see Gob],
1984, p. 173, n. 131.
26. Quoted in Livshits, 1969, p. 46, n. 24.
27. Chavannes, 1905, pp. 538 et seq.
28. Harmatta, 1969, pp. 365-87.
29. Bivar, 19836, pp. 203-4.
30. Zeimal, 1983, pp. 215-17.
31. Gob], 1984, pp. 75 et scq.
32. Zeimal, 1983, p. 225.
 The rise of Sasanian Iran

as there is no scientific certainty about the exact dates of the Kushan kings, this,
too, will remain an open question, even though Harmatta's position is very
attractive. Numismatic maps and statistics d o not provide sufficient evidence to
us t o settle the question of the boundaries between these two parts of the
Kushan state, o r two Kushan states (see also Volume 111, Chapter 7).

Margiana
Margiana (the Merv oasis)" and Bactria occupied a special place in the conflict
between the Kushans and the Sasanians. By the beginning of the third century
A.D., the states of southern Mesopotamia and the provinces of eastern Iran -
Margiana, Segistan (Sistan) and Kerman - were virtually independent states,
governed by local dynasties which only formally recognized their dependence
on the Arsacids.j4 Already as early as the first century A.D. the rulers of Mar-
giana minted their own bronze coins, copying the types of the Arsacid silver
drachms. A few of these coins bore the name of the local ruler, King Sana-
baresSJ5
In the second decade of the third century, when the new Sasanian dynasty
of Persis marched against the Arsacids, the rulers of the eastern Iranian prov-
inces, including the dynasty of Margiana, apparently supported Ardashir I in his
struggle against the last Arsacid, Artabanus V.j6 By A.D. 230, Ardashir con-
trolled a great part of the former Parthian territories. The rulers of Merv volun-
tarily recognized the suzerainty of the Sasanians while preserving for a time a
certain degree of autonomy. In the list of court officials of Ardashir I forming
part of the inscription on the KaLbe of Zoroaster, Ardashir's name and the
names of the King of Merv and other kings of the eastern Iranian provinces
headed the list.37 Between 240 and 260, the Merv ruler minted in his own name a
bronze coin with the figure of a horseman and the Pahlavi inscription mlu3'
MLK'.j8
Around 260, during the reign of Shapur I (A.D. 243-72), the dynasty of
the M e w kings was abolished. The King of Merv is no longer mentioned on the
list of Shapurys court officials of this iihaniih found in the inscription on the
KaLbeof Zoroaster.39 Margiana became part of the administrative province that

33. Numismatic and archaeological material (including unpublished material) on the klcn.
oasis has been presented by A. B. Nikitin.
34. Lukonin, 1969a, p. 36.
35. Pilipko, 1980, p. 117.
36. Lukonin, 1969'1, p. 38.
37. Ibid., p. 39.
38. Loginov and Nikitin, 1986.
39. Lukonin, 1969'1, p. 62.
 8, A. ~ , t ~ , ~ ~.xlith
s k ~the, , contl-ibutions of M. Hussain Shah and R. Shabani .%zmghabndi

FIG.3. Old Merv. T h e citadel. (Courtesy of I. Iskender-Mochiri.)

was given the name of 'Hind, Sagistiin and Turistiin to the sea coast' and was
ruled by members of the Sasanian family, sons and brothers of the Sasanian
king. The first ruler of this province was Narseh, the son of Shapur I. The suc-
cession of rulers of the eastern Iranian provinces can be traced up to the begin-
ning of the fourth century.J0Merv was an integral part of the Sasanian state;
during this period it issued copper coins and, to a lesser extent, silver of the
same types as those minted by the Sasanian state.
Between the third and fifth centuries, the capital of the Merv oasis was
still the ancient Antioch in Margiana, the present-day site of Gyaur-kala, near
Old Merv. At the fortified site of the ancient town, which covers an area of over
4 km2, studies have been carried out o n the citadel (Erk-kala) of which the
oldest part belongs to the third-century keep (Fig. 3)," the fortifications, the liv-
ing and workshop quarters in the northern section,42 the Buddhist religious
building complex,43 a Christian monastery which was probably functioning
from the third century,.14 and finally a necropolis located outside the limits,
which was used from the second century until the end of the Sasanian period.4s
Under the first Sasanians the oasis fortresses that had been built in
the previous period - Chil'burdzh, Durnali, Chichanlik, Kirk-depc - were

40. Lukonin, 1979, p. 13.

41. Usmanova, 1963, pp. 33 et seq.
42. Katsuris and Buryakov, 1963.
43. U s ~ n a n o v aand Dresvyanskaya, 1976, p. 553; Papakhristu and U s ~ i ~ n n o v a1978,
, p. 548.
44. Dresvy anskaya, 1974.
45. Ershov, 1959; Obel'chcnko, 1969.
 The rise of Sasanian Iran

strengthened and r e ~ o n s t r u c t e d .As

~ ~a rule, no buildings were located within
fortresses. Settlements grew up round the walls, some remaining until Mu,-
lim times. Each fortress was used to quarter the troops who defended a partic-
"jar sector of the oasis. The construction of most of these fortifications (dcye or
tepe) dates back t o the Early Sasanian period. They were erected on high adobe
alongside the settlements. Of unusual design was the Gebekli-depe
fortress, guarding the approach to the north-west limit of the oasis. The citadel
platform was enclosed by a second fortification with towers at each corner.
Excavations revealed the presence of earlier fortifications of the Parthian period,
though the coins found show that most of the work was erected during the
reign of Shapur I.
Margiana was an ancient agricultural oasis in the delta of the Murghab and
was irrigated by its waters. The location of monuments from the Bronze and
~ a r Iron l ~ Ages shows that the boundaries of the lands under irrigation g-adu-
ally shifted southward. In antiquity and the Early Middle Ages, Margiana was
the site of a complex system of canals fed by tributaries of the Murghab. It did
not possess its own ore deposits, but imported the raw material for its local
metallurgical production from mines in northern Iran4' Iron blooms were dis-
covered in many settlements and an arms workshop dating to Early Sasanian
times was found in O l d Merv itself. Plutarch mentions Margiana's steel, which
was used t o make armour for Parthian soldiers. Pottery production was highly
developed, and potters occupied a whole quarter in Old Merv. Kilns of the Par-
thian period were discovered in D z h i n - d e ~ e and , ~ ~ pottery workshops operated
in many towns and settlements. Pottery of the Late Parthian and Early Sasanian
period differed little in shape, although there were some changes in production
techniques." Among other objects found were spindle whorls and loom
weights. From ancient times the inhabitants of Margiana were Mazdeans. In the
third and fourth centuries, their religious beliefs were gradually transformed
under the influence of the orthodox Zoroastrianism of the Sasanian Empire, and
this was reflected in a change of burial rites. The burial of bodies, which was the
normal practice in the Parthian ~ e r i o d ,was supplanted by the burying of
b o n e s . ~Local
~ Mazdean practices are undoubtedly responsible for the terracotta
statuettes of the Margiana goddess, the goddess of fertilit);, which have been
found in great profusion in Merv." According to al-Biruni, Christianity had
reached Merv within 200 years of the birth of Christ and the first reference to a
Merv bishopric dates to the year 334.'? From the middle of the third century a

46. Koshelenko, 1977, pp. 42-3; Pugachenko\la, 1958.

47. Koshelenko, 1977, p. 35.
48. Merezhin, 1962.
49. Rutkovskaya, 1962, pp. 67 et seq.
50. Koshelenko, 1977.
51. Pugaclienko\~a,1962, pp. 1 18 et seq.
52. al-Biruni, 1957, p. 330; Nikitin, 1984, p. 123.
 H. A. l-itvinsky, with thc contributions of M. Hussain Shah a n d R. Shabani Samshabadi

Manichaean community existed there.53Still earlier, possibly in the second ten-

tury, Buddhism appeared,54and the third and fourth centuries witnessed the dc-
velopment of a complex of Buddhist buildings o n the site of the ancient town of
Gyaur- kala.55
At Merv, the beginning of the Sasanian period can be regarded as a time of
relative economic advance, compared with the Late Parthian period (contrary to
the view previously held). Economic g o w t h was fostered by the oasis having
become firmly part of Sasanian territory and by Merv's increasing importance
as a military outpost and trade centre resulting from the conquests in the east
and the development of international trade.

Bactria and the Silk Route

The second and third centuries were also a time of noteworthy achievement for
the culture of Bactria. During this period the country consisted of towns,
including the important cities of Balkh and Termez, and rural settlements
(roughly in the ratio of one t o seven). The rural settlements also included the
nomadic population. Bactria had a wide range of rich temples and other places
of worship for various religions. Its art and architecture had a distinctive charac-
ter and had reached a high stage of d e ~ e l o p m e n t . ~ ~
The Silk Route (see Map 5 ) continued to be extensively used, as evidenced
by the written source^,^' as well as archaeological discoveries in China proper,
East Turkestan, Central Asia and Afghanistan. But the conquest of the western
part of the Kushan Empire by Ardashir, the founder of the Sasanian dynasty,
caused some difficulties in the traffic and in the silk trade in the north between
Sogdiana and Kushanshahr, as a vassal kingdom now belonging to the Sasanian
Empire, and in the south between Iran and Gandhara.
Consequently Indian and Sogdian merchants, keeping away from Sasa-
nian Kushanshahr, made efforts to reorganize the silk trade and looked for a
new route leading from the eastern Kushan kingdom across Gilgit and the
Karakorum range to the Pamirs, Sogdiana and Chinese Turkestan. Interesting
testimony for the use of this route for the silk trade is seen in the inscriptions
engraved by Sogdian merchants o n the rocks at Thor and the Shatial Bridge (see
Chapter 17).

53. Lukonin, 1969a, p. 77.

54. Koshelenko, 1966, p. 175.
55. Usmanova, 1975, p. 530.
56. Staviskiy, 1977; Pugaclienkova et al., 1978; Pugachcnkova, 1979; Kruglikova, 1974;
Kruglikova and Pugaclienkova, 1977; Litvinslciy and Sedov, 1983, 1984; Litvinskiy, 1968;
Schlumberger et al., 1983.
57. Herrmann, 1938, p. 2.
 J. Harmatta

Between 700 B.C. and A.D. 250, the development of the civilizations of Central
Asia was mainly determined by two factors. O n e was the rise of nomadic ani-
mal husbandry and the appearance of equestrian nomads; the other was repre-
sented by the growth of agriculture, craftsmanship and urbanization. Orig-
inally, the Iranian tribes pursued both stock-breeding and tillage. Before the
invention of the two- and four-wheeled vehicle and the war-chariot, that is,
during the first wave of their mass migration towards the east and the south,
their culture was fairly uniform. But the invention of riding made it possible to
drive great herds of cattle and horses and immensely facilitated long-distance
traffic. The northern Iranian tribes took possession of the vast steppe zone of
Asia. Abandoning sedentary life, they became nomadic herdsmen. Looking for
grass and water, they constantly wandered over the pastures - as both Greek
and Chinese authors agree when characterizing their way of life.
These Iranian equestrian nomads created a highly developed nomadic
civilization, all the elements of which were adapted to nomadic stock-breeding.
Stable settlements were replaced by mobile nomadic camps. Instead of adobe
houses they constructed yurts - tents with a light wooden structure, covered
with skin o r felt, which could easily be assembled and taken down. The Iranian
nomads developed the high-wheeled light nomadic vehicle and perfected the
crafts connected with horse trappings, harness, saddles and arms, in particular
the bow-and-arrow and both human and equestrian mail. Their diet was rich in
animal protein, but they were permanently short of cereals. The driving and
grazing of great herds of cattle, horse and sheep required a strong social organ-
ization of military type. The nomadic rotation of pastures between two climatic
zones and the distribution of g a z i n g grounds between tribes necessitated the
creation of broader political units. Thus, great tribal confederations of Iranian
nomads were fonned, which in some cases developed into nomadic states. Dur-
ing the seventh and sixth centuries B.C. several nomadic states of northern Ira-
nian tribes came into being in Central Asia. In the west some Saka tribal confed-
 J. Harrnatta

erations are mentioned in ancient Greek literature and O l d Persian inscriptions

while in the east the Hsien-yun, and later the Yueh-chih and the H s i ~ q - ~ u ,triI
bal confederations are attested by the Chinese sources.
The extremely rich tombs of nomad kings excavated at Arzhan and Pazi-
rik clearly point to the accumulation of great wealth and the concentration of
power in their hands. Lively contacts and easy communications promoted the
rise and spread of a fairly uniform nomadic culture in the steppe zone. The same
types of horse-harness (bridle, bit, cheek-piece, saddle, trappings), arms (bow,
bow-case, arrow and quiver, sword, battle-axe, mail) and garments (trousers,
caftan, waist-girdle, boots, pointed cap) were used in the steppe zone from Cen-
tral Europe to Korea. The Iranian nomads had plenty of livestock, transport
and animal products but lacked agricultural products and handicrafts. They
constantly tried to obtain the articles they lacked by barter o r by force from
neighbouring sedentary peoples, and this caused recurrent instability on their
borders.
The Iranian tribes w h o advanced towards the south met a well-deve-
loped sedentary civilization. It had a state organization and fortified cities with
handicrafts at a high level, productive agriculture based on irrigation, long-
distance trade in mineral resources and agricultural products. It made extensive
use of written records and had a rich oral and written literature. The immigrat-
ing Iranians had already formed tribal confederations. Adopting many ele-
ments of the sedentary civilization, they quickly developed their own state
organizations and founded kingdoms in Media, Persia, Bactria, Chorasmia, and
perhaps Arachosia. They retained their independence of each other for a cen-
tury, but eventually Media gained the ascendancy, and almost all the southern
Iranian tribes were united for the first time under the Medes. In the middle of
the sixth century B.C. the Medes were replaced by the Persians, who enlarged
their territory to include nearly all the well-developed lands of ancient seden-
tary civilizations from the Aegean t o the River Indus and from Ethiopia to the
River Syr Darya.
In extending its frontiers up to the Syr Darya, the Persian Empire
annexed part of the steppe zone with several nomadic tribal confederations
(Saki Haumavargii, Saki Tigraxaudii). There was, therefore, permanent contact
between the Iranian nomads and the sedentary Iranian population during the
whole of the Old Persian period. In Central Asia, the Achaemenid kings estab-
lished a well-organized administration, built fortified cities and satrapal centres.
They constructed military roads and ~ r o l n o t e dagriculture with irrigation. They
also employed Iranian nomads as cavalrymen in the army and settled them as
military colonists. Saka armament, harness and warfare had a clear effect on the
O l d Persian army. At the same time, O l d Persian culture, incorporating the best
elements of earlier civilizations in the ancient Near East, had an influence fa].
beyond the northern frontiers of the empire, as the finds in the burials of the
nomad kings at Pazirik show. After the wars of Cyrus and Darius I against the
 Conclusion

Iranian nomads, contacts were peaceful on the northern frontiers until the inva-
sion of Alexander the Great.
The Achaemenids also extended their supremacy over Asia Minor, the
Greek islands and the eastern coast of the Mediterranean. his was a very
important event because it now became possible for the first time, at least
within the limits of the Old Persian Empire, for the peoples of CentralAsia to
come into contact with Greek Mediterranean cultures.
c o m m o d i t y production and money economy were better developed i n
the reek city-states. Greek coins consequently circulated widely in the eastern
satra~iesof the Achaemenid kings. From an economic viewpoint, the Greeks
had conquered Iran long before the army of Alexander the Great crossed the
Bosporus. The process was accelerated when the Persians began to hire Greek
mercenaries. It was only a matter of time before Greek economic and military
superiority came into operation, and the final result was never in doubt.
After the collapse of the Old Persian military resistance in 329 B.c., the
Macedonian and Greek troops of Alexander arrived at the River Svr Darya, the
frontier between the sedentary and the nomadic civilizations. The encounter of
the Greeks with the Iranian nomads of Central Asia was of decisive importance
for future developments. The Greeks brought a well-developed urban culture,
with a great tradition of handicrafts and arts, a rich literature and many religious
cults. They had a new art of warfare, an efficient system of land-ownership for
city-states, a developed commodity production and money economy, and they
were interested in long-distance trade. All these elements of Greek culture made
a strong impact on the Iranians living inside the former Persian Empire, and
exercised considerable influence on the nomads.
When the Seleucid Empire disintegrated, the formation of a syncretic
Graeco-Iranian culture intensified. In Parthia, the nomadic tribe of the Iranian
Parni became masters of a slightly Hellenized land. They retained many ele-
ments from their nomadic culture, particularly in military organization, but
adopted both key elements of Old Persian tradition (such as the royal chancel-
leries) and the most important achievements of Hellenism. It was not accidental
that many Parthian kings incorporated 'Philhellene' in their titles and that
Greek dramas were performed at their court.
In Graeco-Bactria the Greek ethnic element was more numerous than in
other Iranian countries and the impact of Greek culture was more effective.
Greek cities were built where Greek arts and handicrafts flourished, and the
forms of Greek religious cults were adopted even by many Iranians. Greek Ian-
guage and literacy was widespread and the Greek theatre stimulated lndian dra-
matic art. Due t o the landed-property area (chora) of the reek cities, the sys-
tem of double landownership came into being in both ~ a r t h i aand act ria. he
supreme of the land was the king. while the village ~ ~ m m u n i t i of
es
half-free peasants, who cultivated the fields. were only secondary landowners.
Beside the royal land, there were also landed-~ropert?areas of the reek cities
 J. Harmatta

and sanctuaries. These three categories of landed property were adopted by

other countries of Central Asia. Perhaps the most important effect produced by
Hellenism, however, was the large-scale development of long-distance trade.
The Graeco-Bactrian kings led military expeditions to the Phryni and Seres to
secure trading contacts with China, and for the first time an exchange of mate-
rial goods and culture between East and West across Central Asia became pas-
sible. The rise of the Graeco-Bactrian kingdom as a Central Asian power
upgraded the importance of the region.
The impact of urban development was also felt in the Tarim basin. The
wide deserts of this region were unsuitable for nomadic horse-breeding but
agriculture was possible in the oases along the rivers. The population at that
time consisted of diverse Saka tribes. The nomadic Yueh-chih tribal confeder-
ation controlled the vast territory from the Altai mountains to the Huang-ho
and held sway over the Hsiung-nu tribes living t o the east. Between 203 and
177176 B.c., however, the Hsiung-nu defeated the Yueh-chih tribes, who
migrated to the west, to the region of the Ili, C h u and Naryn rivers.
Urbanization in the Tarim basin had begun in the third century B.C. when
the population still consisted of Saka tribes, as is proved by the evidence of
place-names that are exclusively of Saka origin. But the Saka population was
driven towards the west, and the eastern part of the Tarim basin was occupied
by an Indo-European people speaking centum-language. At an earlier stage they
may have belonged to the Yiieh-chih confederation, as this would explain the
name 'four Tuyrak' for Qarashahr, Kocho and Bishbaliq.
Both the ancient Saka population and the later Indo-European immi-
grants continued to develop cities and formed small city-states in Kashgar, Yar-
kand, Khotan, Kucha, Qarashahr and Lou-lan. The wealth and prosperity of the
petty kingdoms in the Tarim basin from agriculture, handicrafts and transit
trade aroused the interest of the Hsiung-nu, w h o took over the territory.
The Hsiung-nu confederation consisted of twenty-four tribes, controlling
a nomadic empire with a strong military organization. They were excellent
stock-breeders but they were well acquainted with agriculture and had several
permanent fortified settlements where handicrafts flourished. Their economy
made use of the forests to ~ r o v i d efodder for their animals and raw material for
handicrafts. Their royal tribes and kings (shan-YU)bore Iranian names and all
the Hsiung-nu words noted by the Chinese can be explained from an Iranian
language of Saka type. It is therefore clear that the majority of Hsiung-nu tribes
spoke an Eastern Iranian language.
While the Hsiung-nu ruled the Tarim basin and the trading routes leading
across it, the Yiieh-chih (known as Tochari in Greek and Latin sources) invaded
Bactria between 133 and 129 B.C. This move of the Yiieh-chih was occasioned
by the aggression of the Wu-sun (Asiani in classical sources), a nomadic Iranian
people under Hsiung-nu sovereignty. It also set in motion the Saka tribes again.
After several successful wars against the Parthians, the western branch of the
 Conclusion

Sakas settled in Sistan where a Saka kingdom was ~b~~~ the same
time, the Sakas using the Karakorum route established a Saka kingdom in Gan-
dhira. They eventually conquered Mathura and, advancing south, later became
masters of Surashtra and Malwa.
The Sakas quickly abandoned their nomadic way of life and adapted
themselves t o local social and economic customs. They adopted many elements
of Parthian, Greek and Indian cultures and became Hellenized or Indianized.
They issued coins modelled on the Graeco-Bactrian coinage, took part in the
worship of their Indian subjects and patronized both Brahmanic sanctuaries and
Buddhist monasteries. But they retained their armaments and art of warfare -
the deployment of armoured cavalry which began to spread from them
throughout the ancient world.
In Bactria the Yueh-chih (Tochari) remained to the north of the River
Amu Darya, and lived in a confederation of five tribes, headed by a yabghu.
O n e of them was the Kushan tribe who played an important role later. The
Kushan yabghu settled in the Surkhan Darya valley where his residence might
have been at Khalchayan. Finds excavated there reveal a syncretic culture, unit-
ing their former nomadic civilization with Parthian and Hellenistic elements.
The Kushans in northern Bactria controlled the starting point and an important
section of the Silk Route as well as the crossing places on the Amu Darya. They
benefited from the transit trade and kept away from the struggle between the
Chinese and the Hsiung-nu. After more than a century they were able to extend
their hegemony over the other four Yueh-chih tribes and when the Indo-
Parthian kingdom of Gondophares declined and Parthia was preoccupied in
internal strife, they occupied southern Bactria and G a n d h ~ r ain the mid-first
century A.D.
The organization of the Kushan Empire must be attributed to Kujula
Kadphises (c. A.D. 50-100). O n his first coin issues he had the title yavuga (yab-
ghrr), but after his great conquests he assumed the titles of m a b i r q a (Great
King), rqatinija (King of Kings) and devaputra (Son of God), claiming equalitv
with the Indian, Parthian and Chinese sovereigns. H e was succeeded by Vima
Kadphises (c. A.D. 101-133), Kanishka I (A.D. 134-156), Vasishka (A.D. 157 to
c. 164), Kanishka I1 (c. A.D. 158-176), Huvishka I (c. A.D. 158-182). ~ u v i s h k aI1
(c. A.D. 183-193) and Vasudeva I (c. A.D. 194-233) who became known as 'Great
Kushan' in historiography. The two Kadphises still used the Eucratides era,
beginning around 166 B.c., and the Old Saka era, beginning about 66 B.C. re-
spectively; but Kanishka introduced his own era, beginrling in A.D. 134, which
was also used by other Great Kushans.
The period of the Great Kushans (c. A.D. 50 to c. 233) was the golden age
of ancient Central Asia. They substantially enlarged their territory, led success-
ful expeditions as far as Pitaliputra, conquered India up to Kauiambi, con-
trolled the most inlportant ports on the western coast of India, and were mas-
ters of the area up to K'ang in the north and t~ the Tarin1 basin in the
 J. Harrnattn

11orth-east. Their territories covered all the trading routes on which Chinese
goods (silk and other textiles, spices, precious stones and metal wares) brought
across the Tarim basin could be transported onwards t o Parthia, Egypt, Europe
and the Roman Empire; and they could consequently control the whole East-
West trade crossing Central Asia.
The Kushans made n o serious attempt t o conquer the Tarirn basin, even
during periods when the Chinese withdrew their garrisons, though Kushan pol-
itical, economic and cultural influence was considerable in the city-states of the
region. The Chinese had extended their control over the Tarim basin between
108 and 65 B.C. t o protect caravans, travelling o n the Silk Route from China,
against robbery by the Hsiung-nu. They stationed garrison troops in some stra-
tegic cities and tried t o install kings loyal t o the Celestial Empire and to control
the nomination of principal officers, imposing o n them the rules of Chinese
protocol in diplomatic contact. Otherwise local administration, customs, social
and economic structure remained untouched b y Chinese influence. Kushan cul-
ture, at its zenith at that time but not associated with foreign political oppres-
sion, made a greater impact.
The prosperity of the Kushan Empire was based upon its highly devel-
oped agriculture. The Kushan kings made great efforts t o secure a good water
supply, to improve and extend irrigation systems and t o enlarge cultivated terri-
tory. The extent of irrigated fields was even greater than today and agricultural
techniques were improved. It was the period when the wooden plough with an
iron ploughshare first appeared in Central Asia.
The growth of agricultural production accelerated the process of urban-
ization. Many new cities arose surrounded b y landed-property districts. They
had a Hellenistic character with a quadrangular ground-plan and massive adobe
walls. In them great dwelling houses, sanctuaries and ~ a l a c e s barracks,
, arsenals
and royal treasuries were constructed. Urbanization promoted and required the
development of arts and handicrafts. In the Kushan cities architecture, sculpture
and painting, textile art and manufacture, metallurgy, pottery, metal work,
jewellery, the manufacture of harness and arms and hydraulic engineering
became highly developed.
The prosperity of handicraft ~ r o d u c t i o nlaid the basis for an intensive
exchange of goods both internally and abroad. Vima K a d ~ h i s e sintroduced a
new monetary system with a standard gold dinar designed for foreign trade and
a range of copper denominations for internal use.
The Kushans benefited substantially from trade with China, and contro-
lled all the routes by which Chinese silk and other wares could be transported
to the West. The Silk Route had one branch leading across Karakoruln to
Gandhira and the ports of western India. The main route ended in the ~ a k h s h
valley from where one branch led t o southern Bactria, another through Merv to
Parthia and a third t o the Caspian and Transcaucasia. Roman and Chinese
coins, Egyptian and Syrian paste and glass articlcs, Chincse mirrors and
 Conclusion

east ~ e d i t e r r a n e a nceramics from excavations in Kushan Central Asia provide

telling evidence f o r the lively trade o n the Silk Route and its branches.
~ u s h a nstate administration, based o n royal officers and chancelleries,
had its antecedents in the O l d Persian Empire and in Greek Bactria. Under the
~ c h a e m e n i d sthe royal chancelleries used Aramaic; the Graeco-Bactrian admi-
nistration retained Aramaic, but also introduced Greek. At first, the Kushan
kings retained Greek but also introduced G i n d h i r i Prakrit. Under Vima Kad-
phises a writing system f o r the Bactrian language using the Greek alphabet was
created and Bactrian became one of the official languages of the administration.
It has recently become clear that the Kushans had a third writing system using
Kharoghi script f o r their o w n language. T w o of the three official languages
(Bactrian and Gindhiiri) were extensively used for Buddhist literature.
T h e creation of new writing systems for several languages bears witness t o
Kushan intellectual life. O n its territory several peoples, languages and religions
mutually influenced and enriched one another. The pantheon o n the Kushan
coins, with gods of pre-Zoroastrian, Zoroastrian, Greek, Egyptian, Mesopota-
mian and Indian origin, clearly reflects the syncretic character of Kushan cul-
ture. But from the historical point of view most important is the role ~ l a y e dby
the Kushan king Kanishka I1 (Candana Kanishka) in the spread of Buddhism in
Central Asia. H e convened the Buddhist synod of the Sarvistivada school in
Kashmir which compiled the Jfidnaprasthdnam and entrusted ASvaghosa, the
famous Indian poet, with providing for the correct Sanskrit language form of
the commentary written by Katyiyana. After the synod, Buddhist Hybrid San-
skrit replaced G i n d h a r i Prakrit as the literary language of Buddhism and in this
development the role of Kanishka I1 was decisive. The Kushan Empire also
played an important role in the spread of Buddhism in the Tarim basin. Budd-
hist religious teaching had already appeared in Bactria as early as the third cen-
tury B.C. Later, under Kushan rule, many Indian merchants and craftsmen
migrated into the Kushan cities of Central Asia. Buddhist missionaries followed
them and many Buddhist religious centres were established.
A t first the Mahisinghika school advanced t o northern Bactria and from
there t o Khotan and other cities of the Tarim basin, bringing G i n d h i r i Prakrit
and the Kharopthi script, which became the official language and script of the
Kushan administration and were also adopted in the kingdom of Lou-lan. The
Buddhist mission also arrived in the Tarim basin o n the Karakorum route and
brought along a different variant of G i n d h i r i Prakrit. By the middle of the first
century A.D., Buddhism had already arrived in China, and the earliest transla-
tions into Chinese of Buddhist works were made from G i n d h i r i Prakrit. Later,
the Sarvistiviida school, using Buddhist Hybrid Sanskrit and B r i h n ~ iscript,
eclipsed the Mahisihghika school in Central Asia. An important role in its suc-
cess was played by Aivagho!a whose works were also translated into Chinese.
While the second century represented a golden age there was a radical
change in Central Asia in the third century. When the delegation of Po-t'iao
(V~sudevaI) arrived at the court of the Wei dynasty in A . D . 230, the Kushans
still possessed Chi-pin (Gandhara), Ta-hsia (Bactria), Kao-fu (Kabul) and T'ien-
chu (India) but the conditions for East-West trade were no longer favourable.
In A.D. 220, the Han Empire fell and its administration collapsed. Parthia
suffered from the war against Rome and the dynastic struggle there led to the
rise of the Sasanians in A.D. 224. North of the Syr Darya loss of control over the
Silk Route across Sogdiana seriously affected the economic and political posi-
tion of the Kushans. Although the Indian provinces of their empire remained
, Kushans now faced a new powerful enemy in
unaffected by these ~ r o b l e m sthe
Sasanian Iran.
The disintegration of the Kushan Empire happened very quickly. Visu-
deva 1's kingdom was divided between Vasudeva 11, ruling in Bactria, and Kan-
ishka 111, reigning in Gandhara and Mathura. A t this historical moment,
Ardashir I, the Sasanian king, invaded the Kushan Empire and took Balkh.
Unable to offer serious resistance, Vasudeva I1 acknowledged Persian supre-
macy and his kingdom became Kushanshahr, one of the vassal states of Sasanian
Iran. The eastern part of the former Kushan Empire preserved its independence,
and its king, Kanishka 111, introduced a new era, the Late Kushan era, beginning
in A.D. 234, which survived in use u p to the ninth century A.D. This collapse of
the Kushan Empire created a new situation. The central power that had con-
trolled East-West trade was now missing and the reorganization of long-
distance trade between China and the Mediterranean only became ~ossiblea
century later.
MAP 1. The Achaemenid Empire and the Iranian nomadic tribes of Central Asia.
M A P2. The campaigns of Alexander the Great in Central Asia.
 - Frontiers of Craeco-Bactrian
kingdom a1 the time of
Eucratides and Menander

-I

---
Fronliers of Parlh~aat the death
of Mithradates II?

Saka kingdoms
II
I Sakastan
II Gandhlra
Ill Mathura
IV Ujjayini

- Indo-Parlhian kingdom
of Gondophares

Sites

1. Pitaliputra
I
2. Ujjayini
3. Barygaza (Broach Land)
4. Sialkot
5. Taxila (Sirkap)
6. Und
7. Pu~kaljvati(Peucelaotis)
8. Takht-i Bahi
9. Nagarahira
10. KlpiiaIBegram
11. Chilas
12. Cilgit
13. Ay Khanum
14. Takht-i Sangin
IS. Tepe-i Dinistan
17.
16. Tepe
Kobadian
Nimlik

18. Emshi-tepe
19. Dilberjin
20. Tillya-lepe
21. Khojand 1 Antioch
22. Afrasiab
23. Erk-kala ICyaur-kala
24. Nisa
25. Shami
26. Spasinu Charax
27. Uruk (Warka)
28. Seleucia
29. Ctesiphon
30. Dura-Europos
31. Nisibis
32. Gindara
33. Carrhae

MAP3. Parthia, Graeco-Bactria, Indo-Parthia and the Saka kingdoms.

 ,
, Frontier of the Kushan Empire
in the time of Kanishka II

Tribes

Sites
II
1. Noin-Ula
2. Sudzhinsk
3. Derestui
4. lvolginsk
5. Khunui-go1
6. Pazirik
7. Lou-Ian
0. Khotan
9. lssik
10. Kafirnigan-tepe
1 1. Tepe-i Dinistan
12. Tekkuz-tepe
13. Takht-i Sangin
14. Kafyr-kala
15. Tepe-i Shah
16. Aruktau, Kokkum, Tulkhar
17. Ay Khanum
10. Qunduz
19. Surkh Kotal
20. Zar-tepe
21. Ayrtarn
22. Fayaz-tepe
23. Kara-tepe
24. Zang-tepe
25. Khalchayan
26. Dilberjin
27. Bactra (Balkh)
20. Tillya-tepe
29. Bargada
30. Mew
31. K3piialBegrarn
32. Dasht-i Nawur
33. Kuh-i Khwaja
34. Nisa
35. Tor Dheri
36. Ara
37. Nagarahara
39. Takht-i
38. Taxila (Sirkap)
Bahi

40. Shatial Bridge

41. Manikyala
42. Sakala
43. Bhakkat
44. Sui Vihar
45. lhukar
46. Mohenjodaro
47. Mathura
48. Kauihbi
49. Sarnath
50. Beshnagar
51. Sanchi
52. Ujjain
53. Barygaza
54. Junagadh

1000 krn I

MAP4. The Hsiung-nu, the Yiieh-chih and the Kushan kingdom.

MAP 5. The Silk Route.
 1 Tun-huang
2 Bar-kol
3
4 Tu-yin
Lou-Ian

5 Sidaogou
6 Turfan
7 Alagou
8 Ayding-kol
9 Shan-Shan
10 Keermuqi
11 Qararhahr
12 Lun-t'ai
13 Kucha
15
14 Aksu
Niya

16 Kherna
17 Minleng
18 Khotan
19 Yarkand
20 Subashi
21 Tashkurgan
22 Kashgar
23 Karasu
24 Margelan
25 Sokh
26 Isfara
27 Aravan
28 Ravat
29 Matcha
30 Hissar (Tupkhan)
31 Zar-tepe
32 Aynam
33 Bactra (Balkh)
34 Dilberjin
35 Fayaz-tepe
36 Kara-tepe
37 Dalverzin-tepe
38 Er-kurgan
39 Nakhsab
40 Ishtikhan
41 Afrasiab
42 Chach
43 Otrar
44 Dingildzhe
45 Koy-Krilgan-kala
46 Toprak-kala
47 Ayaz-kala
48 Kunya-Uaz
49 Zamakhshar
50 Dzhanbas-kala
51 Nisa
52 Merv (Antiochia)
53 Tillya-tepe
54 Saksan-Okhur
55 Sarazrn
56 Dzheti-Asar
57 Surkh Kotal
58 ~egram/Kipiia
59 Udegrarn
60 PuskalSvati
61 Shaikhan Dheri
62 HundIUnd
63 Taaila (Sirkap)
64 Bhita
65 Barygaza
66 Ajanla
67 Sisupalgarh
68 Sanchi
69 Nagarahsra
70 Kauiimbi

500 krn

MAP 6. T h e nomadic and urban cultures in Central Asia.

 Sites 1 Khalchayan 7 Dilberjin 12 Barniyan 17 Tepe Sardar 22 Taxila (Sirkap)
2 Dalverzin-lepe 8 Bactra(0alkh) 13 Paitva 18 Hadda 23 Manikyala
3 Fayaz-tepe 9 Ay Khanurn 14 Shotorak 19 Nagarahlra 24 Kalra M o u n d
4 Kara-tepe 10 Takht-i Sangin 15 BegrarnIKipiia 20 Puskallvati 25 Kankali M o u n d
S Ayrtam 11 Surkh Kotal 16 Tepe Maranjan 21 Takht-i Bahi 26 Sanchi
6 Tillya-tepe

MAP 7. Sites of Kushan art.

 Sites 1 Tagisken Yakka-Parsan 17 Ura-tyube ZS Eylatan
2 Dzheti-Arar Dzhanbas-kala 18 Caza 26 Uzgen
3 Babish Molla Koy-Krilgan-kala 19 Khojand 27 Chust
4 Uygarak Toprak-kala 20 lsfara 28 Kara-Mazar
5 Chirik Rabat Dingildzhe 21 Kokand 29 Chirchik
6 Kalali-gir Hazarasp 22 Sokh 30 Sarazm
7 Kyuzeli-gir Dzhigirbent 23 Naukat 31 Kunya-Uaz
8 Bazar-kala Kanka 24 Markhamat

MAP 8. States in north-western Central Asia.

MAP9. T h e nomads of northern Central Asia after the invasion of Alexander.
A A = Arts asiatiques, Paris
A A H = Acta Archaeologica Academiae Scientiarum Hungaricae, Budapest
Acta Ant. Hung. = Acta Antiqua Academiae Scientiarum Hungaricae, Budapest
A M = Asia Major, London
A 0 = A r k h e ~ l o ~ i c h e s k iotkritiya,
e Moscow
A O H = Acta Orientalia Academiae Scientiarum Hungaricae, Budapest
AP = Ancient Pakistan, Peshawar
A R T = Arkheologicheskie raboti v Tadzhikistane, Dushanbe
AS = Afghan Studies, Kabul
A S I A R = Archaeological Survey of India Annual Report, New Delhi
B C H = Bulletin de cowespondance hellenique, Paris
BEFEO = Bulletin de 1'~colefrancaise d'Extri2me-Orient, Paris
BSOAS = Bulletin of the School of Oriental and African Studies, London
C A H = The Cambridge Ancient History, Cambridge
C H M = Cahiers d'histoire mondiale/Journal of World History, Neuchitel
C I I = Corpus Inscriptionwm Indicarum, Calcutta
C R A I = Comptes rendus de /'Academic des Inscriptions et Belles lettres, Paris
EI = Epigraphia Indica, N e w Delhi
E V = Epigrafika Vostoka, Moscow/Leningrad
EW = East and West, Rome
I A N T S S R = Izvestiya Akademii nauk TSSR, Ashkhabad
I I A = Institut po izucheniyu arkheologii
IIJ = Indo-Iranian Journal, The Hague
I M K U = Istorij/a m a t e r i a l ' n ~kul'tlrui
~ Uzbekistana, Tashkent
I O O N Tadzh. SSR = Izvestij~n Otdelentya obsbchestvennikh nauk Akademii nauk
Tadzhikskoy SSR, Dushanbe
IsMeo.R.M. = Istituto italiano per if Medio ed Estremo Oriente. Rqorts and Memoirs.
Rome
J A = Journal asiatique, Paris
J A O S = Journal of the American Oriental Society, New Ha\len, Conn.
 Bibliography and Refc~rences

JBBRAS = Journal of the Bengal Branch of the Royal Asiatic Society

J C A = Journal of Central Asia, Islamabad
J N S I = Journal of the Numismatic Society of India, Varanasi
JRAS = Journal of the Royal Asiatic Society, London
KhAEE = Khorezmskiye arkheologo-Ctnograficheskiye Ekspeditsii, Moscow
K S I A = Kratkie soobshcheniya Instituta arkheologii, Kiev
K S I I M K = Kratkie soobshcheniya Instituta Istorii material'noy kul'turi, Moscow
M A I K T S A = M e z h d ~ n a r o d n a ~assotsiatsiya
a po izucheniyu Kul'tur Tsentral'noy Azii
M D A F A = Memoires de la Delegation archiologique franqaise en Afghanistan, Paris
MDAFI = Me'moires de la De'legation archeologique franqaise en Iran, Paris
M E = Materiali PO itnografii, Moscow/Leningrad
M I A = Materiali i issledovaniya po arkheologii, Moscow/Leningrad
M I F A O C = Memoires de lJInstitutfranqais d'archhologie orientale du Caire, Cairo
MKhe = Materiali Khorezmskoy ikspeditsii, Moscow
M U = Materiali Uzkomstarisa, Tashkent
M Y u T A K E = Materiali Yuzhno-Turkmenistanskoy arkheologicheskoy k o m p l e k s n ~ ~ ,
ekspeditsii, Leningrad
N C = Numismatic Chronicle, London
O N U = 0bshchestvennie nauki v U z bekistane, Tashkent
O S = Orientalia Suecana, Uppsala
PIIE = Polevie issledovaniya Instituta Etnografii, Moscow
R A = Revue archkologique, Paris
R N = Revue numismatique, Paris
S A = Sovetskaya arkheologiya, Moscow
S A I = Svod arkheologicheskikh istochnikov, Moscow/Leningrad
SE = Sovetskaya Etnografiya, Moscow/Leningrad
S N V = Strani i narodi Vostoka, Moscow
T A N T a d z S S R = Trudi Akademii nauk Tadzhiskoy SSR, Dushanbe
T I I A N T a d z h S S R = Trudi Instituta istorii Akademii nauk Tadzhiskoy SSR, Dushanbe
T M K I A TSAKe = Trudi m e z h d ~ n a r o d n okonferentsii
~ po istorii arkheologii i kul'turi
Tsentral'noy Azii v Kushanskuyu 2pokhu, Moscow
T N I I Y a L I = Tuvinskiy nauchno-issledovatel'skiy institut yazika, literaturi i istorii,
K Y ~ Y ~
T P = T'oung-Pao, Leiden
TrGE = Trudi Gosudarstvennogo Ermitazha, Leningrad
T r G I M = Trudi Gosudarstvennogo Istoricheskogo muzeya, Moscow
T r S A G U = Trudi Sredneaziatskogo Gosudarstvennogo Universiteta, Tashkent
Trudi I I A A N U z b . SSR = Trudi Instituta istorii, arkheologii Akademii nauk Uzbeks-
koy SSR, Tashkent
Trudi I I A E A N Turkm. SSR = Trudi Instituta istorii i arkheologii i itnografii Akademii
nauk Turkmenskoy SSR, Ashkhabad
Trudi K h A EE = Trudi Khorezmskoy arkheologo-etnografi~hesko~ekspeditsii, MOSCOW
Trudi Y u T A K E = Trudi Y u z h n o - T u ~ k m e n i s t a n s k oa~r k h e o l o g i ~ h e s k o kompleksnoy
~
Ekspeditsii, Leningrad
U S A = Uspekhi Sredneaziatskoy Arkheologii, Leningrad
V D I = Vestnik Drevney Istorii, Moscow
VKF A N UzSSR = Vestnik K a r a k a l p a k ~ k o ~filiala
o Akademii nauk Uzbckskoy SSR
 Bibliography and References

zDMG = Zeitschrift der deutschen morgenlundischen Geselfschaft, Wiesbaden

ZDO = Zemli drevnego orosheniya, Moscow

ABAEV,V. 1. 1949. Osetinskiy yazik i f~l'klor,Vol. 1. Moscow/Leningrad.

ABU'L-GAZI.1958. Rodoslovnaya turkmen. (Ed, by A. N. Kononov), Moscow/Lenin-
grad.
AKISHEV,K. A. 1978. Kurgan Issik. Iskusstvo Sakov Kazakhstana. Moscow.
AKISHEV,K. A*; KUSHAEV,G. A. 1963. Drevnyaya kul'tuya sakov i usuney do/ini
Ili. Alma-Ata.
ARTAMONOV,M. I. 1971. Skifo-sibirskoe iskusstvo zverinogo stilya (osnovnie itapi i
na~ravleniya).Sbornik: Problemi skifskoy arkheologii Moscow.
- 1973. Sokrovishcha sakov. Moscow.
BARTOL'D, V . V. 1965. Svedeniya ob Aral'skom more i nizov'yakh AmudarBis drev-
neyshikh vremen d o XVII v. Sochineniya (Moscow), Vol. 3.
BAZHENOV,L. V. 1937. Srednyaya Aziya v drevneyshiy period (mezhdu IV-II v v . do
n. P.). Tashkent.
BERNSHTAM,A. N. 1949a. Iz itogov arkheologicheskikh rabot na Tyan1-Shanei Pamiro-
Alae. KSIIMK, 28, pp. 54-66.
-. 19496. Osnovnie Etapi istorii kul'turi Semirech'ya i Tyan'-Shanya. SA, 11, pp.
337-84.
-. 1950. Trudi Semirechenskoy arkheologicheskoy i?kspeditsii 'Chuyskaya dolina'.
M I A , 14.
-. 1952. Istoriko-arkheologicheskie ocherki Tsentral'nogo Tyan'-Shanya i Pamiro-
Alaya. M I A , 26.
-. 1956. Saki Pamira. V D I , 1, pp. 121-33.
CHERNIKOV,S. S. 1965. Zagadka zolotogo kurgana. Moscow.
CHLENOVA,N . D. 1967. Proiskhozhdenie i rannyaya istoriya plemen tagarskoy kul'turi.
Moscow.
DANDAMAEV, M. A. 1958. Sotsial'naya sushchnost' perevorota Gaumati. V D I , 4, pp.
36-50.
-. 1963a. Iran pri pervikh Akhemenidakh (VI v. do n.;). Moscow.
-. 19636. Pokhod Dariya protiv skifskogo plemeni Tigrakhauda. KSIA, 61.
FRYE,R. N. 1962. The Heritage of Persia. London.
GAFUROV,B. G . 1972. Tadzhiki. Drevneyshaya, drevnyaya i s r e d n ~ v e k o v a ~istoriya a
(ed. by B. A. Litvinsky). Moscow.
GRACH,A. D. 1971. Novie dannie o drevney istorii Tuvi. Uchenie z a ~ i s k i .T N I I Y a L i ,
Vol. 15, pp. 93-106.
GRANTOVSKIY, E. A. 1960. Indo-Iranskie kasti u skifov. X X V M e z h d ~ n a r o d n kongress
i~
vostokovPdov [Proceedings of the Twenty-Fifth ~nternationalCongress of Orien-
talists]. Papers submitted by the delegation of the USSR. Moscow.
- 1975. 0 v o s t o c ~ ~ n o i ~ a ~ ~plemenakh
skikh kushanskogo areala. Tsentral'm~laAziya v
kusl?anskuyu tpokhu, Vol. 2. Moscow.
G R I G O I ~ ' E V.
V , V. 1871. 0 skifskom narode sakakh. St ~ e t e r s b u r ~ .
G ~ y n z ~ o \M. r , p. 1956. Severniy Kazakhstan v +okhu rannikh kochevnikou- ~ I J J ~ K ,
61, pp. 8-16.
 Bibliography and References

V ,P. 1958. Drevnyee iskusstvo Altaya. Leningrad.

G K ~ . A Z N C >M.
-. 1975. K khronologii drcvneyshikh parnyatnikov Cpokhi rannikh kochevnikov.
USA, 3, pp. 9-12.
-. 1978. K voprosu o slozhe~liikul'tur skifo-sibirskogo tipa v svyazi s otkritienl kur-
galla Arzhan. KSIA, 154, pp. 9-18.
-. 1980. Arzhan. Tsarskiy kurgan ranneskifskogo vremeni. Leningrad.
HERODOTUS.1979. Histories (ed. by H. H. Huxley). Letchworth, Bradda Books.
Istoriya Turkrnenskoy SSR. 1957. Vol. 1, Book 1. Ashkhabad.
ITINA,M. A. 1981. Khorezrnskaya Ckspeditsiya - osnovnie itogi i perspektivi issledova-
niy. Kul'tura i iskusstvo drevnego Khorezma. Moscow.
K A D ~ R B A EM.V , K. 1966. Parnyatniki tasmolinskoy kul'turi. In: A. K. Margulan (ed.),
Drevnyaya kul'tura Tsentral'nogo Kazakhstana. Alma-Ata.
KHAZANOV,A. M. 1966. Slozhnie luki evraziyskikh stepey i Irana v skifo-sarmatskuyu
Cpokhu. Material'naya kul'tura narodov Sredney Azii i Kazakhstana. Moscow.
-. 1971. Ocherki voennogo dela Sarmatov. Moscow.
-. 1975. SotsialJnaya istoriya skifov. Moscow, Nauka.
LITVINSKY, B. A. 1959. Izuchenie kurganov v severo-vostochnoy chasti Leninabadskoy
oblasti v 1957. TANTadzhSSR, 103, pp. 109-29.
-. 1960a. Arkheologicheskie otkritiya na Vostochnorn Parnire i problema svyaziey
rnezhdu Sredney Aziey, Kitayem i Indiey v drevnosti. Proceedings of the Twenty-
fifth International Congress of Orientalists. Papers submitted by the delegation of
the USSR. Moscow.
-. 1960b. Saki, kotorie za Sogdom. TANTadzhSSR, 120, pp. 91-6.
-. 1963. Saka Haumavarga v svete sovetskikh arkheologicheskikh issledovaniy. Bei-
trage zur Alten Geschichte und deren Nachleben. Festschrift fur F. Altheim, Vol. 1.
Berlin.
-. 1965. Sredneaziatskie zheleznie nakonechniki strel. SA, 2, pp. 75-91.
- 1972. Drevnie kochevniki 'Krishi mira'. Moscow.
LITVINSKY, B. A.; P'YANKOV,I. V. 1966. Voennoe delo u narodov Sredney Azii v VI-IV
w. d o n.;. V D I , 3, pp. 36-52.
LOKHOVITS,V. A. 1963. R a s k o ~ k ikvadratnogo pogrebal'nogo sooruzheniya na gorod-
ishche Chirik-rabat. MKhE, 6, pp. 214-20.
-. 1980. Podboyno-katakombnie i kollektivnie ~ o g r e b e n i ~kurgannogo a rnogil'nika
Turnek-kichizhik. In: M. A. Itina (ed.), Kochevniki na granitsakh Khorezma, pp.
134-50. Moscow, Nauka.
LUKONIN,V. G . 1977. Iskusstvo drevnego Irana. Moscow.
MANDEL'SHTAM, A. M. 1966. Kochevniki na puti v Indiyu. M I A , 136.
-. 1975. P a m ~ a t n i k ikochevnikov kushanskogo vremeni v Severnoy Baktrii. Lenin-
grad, Nauka.
-. 1978. 0 nekotorikh ~ r o b l e r n a k hizucheniya istori rannikh kochevnikov Sredney
Azii i Kazakhstana. KSIA.
MARGULAN, A. H. 1966. Drevnyaya Kul'tura Tsentral'nogo Kazakhstana. Alma-Ata.
MARGULAN, A. H.; AKISHEV,K. A.; K A D ~ R B A EM. V , K.; O R A Z R A E V
A., M. 1965. Drev-
nyaya kul'tura Tsentral'nogo Kazakhstana. Alma-Ata.
MARKOV,G. E. 1970. Nekotorie problemi o b s h c h e s t v e n n ~organizatsii ~ kochevnikov
Azii. SE, 6, pp. 74-89.
-. 1973. Problemi razvitiya o b s h c l ~ e s t v e n n ostrukturi
~ kochevnikov Azii. ~roceedings
 Bibliography and References

of the ~ i n t hInternational Congress of Archaeological and Elhnographic Soencer.

Payers submitted by the delegation of the USSR. Moscow.
-. 1976. Kochevniki Azii. Struktura khozyaystva i obsh~hestvenno~ organizatsii. Mos-
cow.
MARUSHCHENKO,A. A. 1959. Kurgannie pogrebeniya sarmatskoRo vremeni v pod-
gornoy polose Yuzhnogo Turkmenistana. Trudi 1IA A N Turkrn SSR, 5 ,
pp. 110-22.
MASSON,V. M.; ALEKSHIN,V. A.; BOKOVENKO,N. A. 1978. Vsesoyuznoe soveshcha-
nie 'Rannie kochevniki Sredney Azii i Kazakhstana'. KSIA, 154, pp. 3-9.
MASSON,V. M.; ROMODIN,V. A. 1964. Istoriya Afganisrana, Vol. 1. Moscow.
MOSHKOVA,M. G . 1963. Pamyatniki prokhorovskoy kul'turi. SAI, 1-10.
-. 1972. Savromatskie pamyatniki Severo-Vostochnogo Orenburzh'ya. Pamyatniki
Yuzhnogo Priural j a i Zapadnoy Sibiri sarmatskogo vremeni. Moscow.
OBEL'CHENKO, 0 . V. 1956. Kup-Mazarskiy mogil'nik. Trudi I I A A N Uzb. SSR, 8,
pp. 205-27.
-. 1965. I z istorii i z u c h e n i ~ akurgannikh p o g r e b e n i ~v S r e d n e ~Azii. IMKU, 6,
pp. 185-200.
POLYBIUS.1922. Histories (trans. W. R. Paton), Books 1-111. London, Heinemann. 6 vols.
P'YANKOV,I. V. 1964. K voprosu o marshrute ~ o k h o d aKira I1 na massagetov. V D I , 3,
pp. 115-30.
-. 1968. 'Saki' (Soderzhanie ponyatiya). TIIANTadzhSSR, N o . 3 (53), pp. 12-18.
-. 1972. Khorasmi Geroteya Miletskogo. V D I , 2, pp. 3-21.
-. 1975. Massageti Gerodota. V D I , 2, pp. 46-70.
RUDENKO,S. I. 1953. Kul'tura naseleniya Gornogo Altaya v skifskoe vremya. Moscow1
Leningrad.
-. 1961. Iskusstvo Altaya i Peredney Azii. Moscow.
-. 1968. Drevneyshie v mire khudozhestvennie kovri i tkani. Moscow.
Skifi i Sarmati. 1972. Kiev.
SMIRNOV,K. F. 1961. Vooruzhenie savromatov. MIA, 101.
-. 1964. Savromati. Moscow.
SOROKIN, S. S. 1956. Sredneaziatskie podboynie i katakombnie zakhoroneniya kak
pamyatniki mestnoy kul'turi. SA, 26, pp. 97-1 17.
STRABO. 1923-32. Geography (trans. H. 1. Jones). London, Heinemann. 8 vols.
STRUVE,V. V. 1945. Vosstanie v Margiane pri Darii. V D I , 12.
-. 1968. Pokhod Dariya I na sakov-massagetov. Etyudipo istorii Severnogo Pncher-
nomor j a , Kavkaza i Sredney Azii. Leningrad.
TERENOZHKIN, A. I. 1976. Kimmeriytsi. Kiev.
, P. 1948a. Drevniy Khorezm. Moscow.
T o ~ s r o v S.
-. 19486. Po sledam d r e v n e k h ~ r e z m s k otsivilizatsii.
~ MoscowILeningrad.
- 1961. Priaral'skie skifi i Khorezm. (K istorii zaseleniya i osvoeniya drevney del't
Sirdar'i). SE, 4, pp. 1 1 4 4 6 .
- 1962. Po drevnim del'tam Oksa i Yaksarta. Moscow.
- 1963. Sredneaziatskie skifi svete n ~ v e ~ s h i k a rhk h e ~ l o ~ i c h e s k i kotkritiv.
h V D I , 2,
pp. 2 3 4 5 .
TOLSI-ov, S. P.; I - ~ I N AM., A. 1966. Saki nizov'ev Sirdar'i. SA, 2, pp. 151-75.
TREVER,K. V. 1947. Etnicheskiy sostav naseleni)ra Sredney Azii v VI-V vv. d o n.6. SE,
6-7, p p 307-8.
 Bibliography and References

TRUIINOVSKAYA, S. A. 1963. Krugloe pogrebal'noe sooruzhenie na gorodishche Chirik-

rabat. MKhE, 6.
-. 1980. Rannie pogrebeniya Yugo-vostochnoy kurgannoy gruppi mogil'nika Guz-
gir. In: M. A. Itina (ed.), Kochevniki na granitsakh Khorezma, pp. 101-10. M ~ ~ -
COW.
VAYNBERG, V. I. 1980. Pamyatniki kuyusayskoy kul'turi. In: M. A. Itina (ed.), Kul'tura i
iskusstvo drevnego Khorezma. Moscow.
-. 1981. Skotovodcheskie plernena v drevnern Khorezme. In: M. A. Itina (ed.),
Kochevniki na granitsakh Khorezma, pp. 7-76. Moscow.
VISHNEVSKAYA, 0. A. 1973. Kul'tura sakskikh plemen nizov'ev Sirdar'i v VII-V vv, do
n.2. Trudi KhAEE, 8.
VISHNEVSKAYA, 0. A.; ITINA, M. A. 1971. Rannie saki Priaral'ya. Problemi sk$skoy
arkheologii. Moscow.
VOEVODSKIY, M. V.; GRYAZNOV,M. P. 1938. Usun'skie rnogil'niki na territorii Kirgiz-
koy SSR. K istorii usuney. V D I , 3.
, 1975. Priuzboyskie turkmenskie plemena X I V - X V vv. Ashkhabad.
Y u s u ~ o v K.
-. 1976. Uzboy v antichnoe vrelnya v svete novikh a r k h e ~ l o ~ i c h e s k i kdannikh.
h
I A N T S S R , 6, pp. 38-47.

ABAEV,V. I. 1969. Skifo-evropeyskie isoglossi. Moscow.

DANDAMAEV, M. A.; LUKONIN, V. G . 1980. Kul'tura i ikonomika drevnego Irana.
Moscow.
DUCHESNE-GUILLEMIN, J. 1962. La religion de /'Iran ancien. Paris.
D'YAKONOV,I. M. 1971. Vostochniy Iran d o Kira. Istoriya iranskogo gosudarstva i
kul'turi. Moscow.
D'YAKONOV,M. M. 1961. Ocherk istorii drevnego Irana. Moscow.
FRYE,R. N. 1962. The Heritage of Persia. London.
GERSHEVITCH, I. 1959. The Avestan H y m n to Mithra. Cambridge, Cambridge Univer-
sity Press.
GHIRSHMAN, R. 1977. L'Iran et la migration des Indo-aryens et des Iraniens. Leiden.
G N O L I ,G . 1975. Zur Sistan-Hypothese. Acta Iranica, 4, pp. 277-90.
GRANTOVSKIY, E. A. 1970. Rannyaya istoriya iranskikh plemen Peredney Azii. Moscow.
HARMATTA,J . 1979. Darius' Expedition Against the Saki Tigraxaudi. In: J. Harmatta
(ed.), Studies in the Sources on the History of Pre-Islamic Central Asia, pp. 19-28.
Budapest.
H E N N I N G W.
, B. 1951. Zoroaster, Politician or Witch-doctor? London.
HERZFELD,E. 1968. The Persian Empire. Studies in Geography and Ethnography of the
Ancient Near-East. Wiesbaden.
HINZ, W. 1976, 1979. Darius und die Perser. Einc Kulturgeschichte der Achameniden,
Vols. 1-11. Baden-Baden.
KENT, R. G . 1953. Old Persian. Grammar. Texts. Lexicon, 2nd rev. ed. New Haven,
Conn., American Oriental Society.
LEVINE,L. D . 1974. Geographical Studies in the Neo-Assyrian Zuguos. ~ o r o n t o / L o n d o n .
MASSON,V. M. 1959. Dre~nezemledel'cheska~a kul'tura Murgiuni. ~~~~~~~~~~~~~~ad.
MULLERUS, C . 1849. Fragmenta historicorum Graecorum, Vol. 2. Paris.
 Bibliography and References

NYLANDER,C . 1970. Ionians in Pasargadae. Studies on Old Persian Architecture. Upp-

sala.
OLMSTEAD,A. T. 1948. History of the Persian Empire, Achaemenid Period. Chicago,
Chicago University Press.
PORADA,E. 1965. The Art of Ancient Iran: Pre-Islamic Cultures. New York.
PRASEK,J. 1906-10. Geschichte der Meder und Perser bis zur makedonischen Eroberung.
Gotha.
WIDENGREN,G. 1965. Die Religionen Irans. Stuttgart.

Afghan Studies. 1978, 1979, 1982.

BERNARD,P., et al. 1973. In: Fouilles d'Ai' Khanoum, Vol. I. Paris. (MDAFA, 21.)
-. 198 1. Kabul. Problkmes d'histoire coloniale grecque H travers l'urbanisme d'une
citi hellinistique d7Asie Centrale. 1 J 0 Jahre. Deutsches Ar~hiiolo~isches Institut
1829-1979, pp. 108-20. Mainz.
-. 1985. Les monnaies hors trisors. Questions d'histoire grico-bactrienne. Fouilles
d'Ai' Khanoum, Vol. 4. Paris. ( M D A F A , 28.)
BLOCH, J . 1950. Les inscriptions dJAioka. Paris.
BONGARD-LEVIN,G . M. 1971. Studies in Ancient India and Central Asia. Calcutta.
C R A I . 1965-80.
DAGENS, B.; LE BERRE, M.; SCHLUMBEKGER, D. 1964. Monuments preislamiques
d'Afghanistan, pp. 61-104. Paris. ( M D A F A , 19.)
DANI,A. H. 1980. In the Footsteps of Alexander the Great. Megas Alexandros, pp.
117-30. Thessalonika.
DERKETT,J. D. M. 1979. Megasthenes. Der Kleine Pauly, Vol. 3.
EGGERMONT,P. H. D. 1970. Alexander's Campaign in Gandhira. Orientalia Lovenien-
sia Periodica, Vol. 1, pp. 63-123.
-. 1975. Alexander's Campaigns in Sind and Baluchistan and the Siege of the Brahmin
Town of Harmatelio. Louvain.
FILANOVICH, M . I. 1974. Gyaur-Kala. Trudi YuTAKE, 15.
FOUCHER, A. C. 1947. La vieille route de lJlnde de Bactres a Taxila. Paris. (MDAFA, 1.)
FRASER,P. M. 1979. Afghan Studies, 2, pp. 9-21.
FRYE,R. N . 1963. The Heritage of Persia. London.
GHIRSHMAN, R. 1954. Iran from the Earliest Times to the Islamic Conquest. Harmonds-
worth, Penguin Books.
KOSHELENKO,G. A. 1979. Grecheskiy polis na illinisticheskom Vostoke. MOSCOW.
LEZINE,A. 1963164. Hkrat. Notes de voyage (1). Bulletin d'etudes orientales, Vol. 18,
pp. 1 2 7 4 5 .
MAJUMDAR,R. C. 1960. The Classical Accounts of India. Calcutta, Firma K. L. Muk-
hopadhy ay.
MEREZHIN,L. I. 1978. Obsledovanie rayonnoy steni I11 v. d o n.;. vokrug Margiani.
Trudi YuTAKE, 16, pp. 1 1-1 5.
MITCHINER,M. 1975. Indo-Greek and Indo-Sc~thian Coinage, Vol. I . London,
Hawkins.
N A R A I NA. , IS. 1957. Thc Indo-Greeks. Oxford, Oxford University Press.
 Bibliography and References

NEW~LL E., T. 1978. The Coinage of the Eastern Seleucid Mints from Seleucus I to Antio-
chus III, 2nd ed., pp. 231-6. N e w York, American Numismatic Society.
OLMSTEALI, A. T. E. 1948. History of the Persian Empire, Achaemenid Period. Chicago,
Chicago University Press.
PUGACHENKOVA, G . A. 1976. Drevnyaya Baktriya. Materiali Sovetsko-Afganskoy ;ks-
peditsii 1969-1973. Moscow.
ROBERT,L. 1973. In: Fouilles dJAi' Khanoum, Vol. I, pp. 211-37. Paris. (MDAFA, 21.)
SCHLUMBEIIGER, D . 1964. Une nouvelle inscription grecque dlAqoka. C R A I , pp. 126-40.
SCHLUMBERGEK, D.; ROBERT,L.; DUPONT-SOMMER,A.; BENVENISTE, E. 1958. Une
bilingue greco-aramienne d'Aioka. /A, 246, 1, pp. 1-48.
SCHOBER,L. 198 1. Untersuchungen zur Geschichte Babiloniens und der oberen Satrapien
v o n 323-303 v . Chr. Frankfurt.
SCHWANBECK, E. A. 1846. Megasthenis Indica. Bonn.
SCHWARZ,F. 1970. Die Griechen und die Maurya-Dynastie. In: F. Altheim and R. Stieh]
(eds.), Geschichte Mittelasiens i m Alterum, pp. 267-316. Berlin.
SHISHKINA, G . V. 1969. 0 mestonakhozhdenii Marakandi. SA, 1, pp. 62-75.
-. 1975. Ellinisticheskaya keramika Afrasiaba. SA, 2, pp. 60-78.
STEIN,A. 1929. O n Alexander's Track to the Indus; Personal Narrative of Explorations
on the North-west Frontier of India, Carried O u t Under the Order of H.M.
Indian Government. London.
-. 1937. Archaeological Reconnaissance in North-western India and South-eastern
Iran Carried O u t and Recorded with the Support of Harvard University and the
British Museum. London.
TARN,W. W. 1951. Alexander the Great. Cambridge, Cambridge University Press.
TSCHERIKOVER,
V. 1927. Die hellenistischen Stadtegrundungen von Alexander dem
Grossen bis auf die Romerzeit. Philologus Supplementband, XIX, 1. Leipzig.
VIAZIGIN,S. A. 1949. Stena Antiokha Sotera vokrug drevney Margiani. TrudE YuTAKE,
1, pp. 260-75.
WILL,E. 1966, 1967, 1979. Histoire politique d u monde helle'nistique, Vols. 1-11. Nancy.
WOLSKI,Y. 1947. L'effondrement de la domination des Sileucides en Iran au IIIe sikcle
av. J.-C. Bulletin international de I'Acade'mie polonaise des Sciences et Lettres, cl.
Philologie, pp. 13-70. Cracow. (Supplement, 5.)

AS. 1978, 1979.

BELYAEVA,T. V. 1979. Raskopochnie raboti na Khodzhentskoln arke. AO, 1978,
pp. 575-6.
-. 1980. Issledovaniya v Leninabadskoy oblasti. A O , 1979, pp. 470-1.
BERNARD,P. 1965. Fouilles d'Ai Khanourn. C R A I , pp. 636-9.
-. 1967. Ai Khanourn on the Oxus: A Hellenistic City in Central Asia. Proceedings of
the British Academy, Vol. 53, pp. 71-95, Plates I-XX.
-. e Asie centrale. Syria, 47,
1970. Sieges et lits en ivoire d'Cpoque h e l l i ~ l i s t i ~ uen
pp. 3 2 7 4 3 .
-. 1976. Les traditions orientales dans I'architecture gri-co-bactrienne. ]A, 264,
pp. 245-75.
-. 1981. Problkmes d'histoire coloniale grecque i travers I'urbanisme d'unc citi
 Bibliography and References

hellinistique d'Asie Centrale. l l i O Jahre. Deursches Arch~olo~isches Instirut 1829-

1979, pp. 108-20. Mainz.
-. 1985. Les monnaies hors trisors. Fouilles d'Ai'Khanoum, Vol. IV. Paris. (MDAFA,
28.)
ARD,P.; AUDOUIN,R. 1973. TrCsors de monnaies indicnnes et i n d ~ - ~ r e c q u ed'Ai' s
Khanoum (Afghanistan). RN, 6th series, Vol. 16, pp. 238-89.
1974. Trksors de monnaies indiennes et indo-grecques d'Ai' Khanoum (Afghan-
istan). R N , 6th series, Vol. 17, pp. 7-41.
BERNARD,P., et al. 1973. Campagnes 1965, 1966, 1967, 1968. Fouilles d'Ai' Khanoum,
Vol. I. Paris. (MDAFA, 21.)
-. 1976. Fouilles d'Ai' Khanoum (Afghanistan): Campagne de 1974. BEFEO, 63,
pp. 5-58.
-. 1980. Campagne d e fouille de 1978 a Ai Khanoum. BEFEO, 68, pp. 1-103.
BOPEARACHCHI,0. 1991. Monnaies gre'co-bactriennes et i n d ~ - ~ r e c q u e Catalogue s. rai-
sonne'. Paris, Bibliothtque Nationale.
B U R ~ YV., P. 1976. Tekhnika zhivopisi. Drevnyaya Baktriya, Vol. 1, pp. 110-24. (French
summary on pp. 187-8.)
CII. 1929, 2, 1, 1.
CRAI. 1965-72, 1974-76, 1978, 1980.
CURIEL, R.; FUSSMAN,G. 1965. Le trhsor mone'taire de Qunduz, Vols. 1-3. Paris.
(MDAFA, 20.)
DADASHEVA, S. A. 1976. In Baktriyskie drevnosti, pp. 106-9.
DAGENS, B.; LE BERRE,M.; SCHLUMBERGER, D. 1964. Monuments preislamiques d'Af-
ghanistan. Paris. (MDAFA, 19.)
DANI, A. H. 1965166. Shaikhan Dheri Excavations, 1963-64 Seasons: In Search of the
Second City of Pushkalavati. AP, 2, pp. 17-214.
DENISOV,E. P. 1975. ART, 15, pp. 100-9.
DIHLE, A. 1965. Umstrittene Daten. Untersuchungen zum Auftreten der Griechen a m
Roten Meer. Cologne.
-. 1978. Die entdeckungsgeschichtlichen Voraussetzungen des Indienhandels der
romischen Kaiserzeit. In: H . T e m ~ o r i n (ed.), i Aufsteig und Niedergang der romis-
chen Welt, Vol. 11, N o . 9, Part 2, pp. 546-80. Berlin.
D'YAKONOV, M. M. 1953. A r k h e o l ~ ~ i c h e s k raboti
ie v nizhnem techenii reki Kafirnigana
(Kobadian) (1950-1951 gg.). MIA, 37, pp. 253-93.
FILANOVICH, M. I. 1973. K istorii slozheniya gorodskikh u k r e p l e n i ~Afrasiaba. Afrasiab,
Vol. 2, pp. 85-94.
-. 1974. Gyaur-Kala. Trudi YuTAKE, 15, pp. 15-139.
FOUCHER,A. 1951. In: Me'moires de I'Academie des Inscriptions et Belles Lettres, VOI. 43,
N o . 2, pp. 289-95.
FRANCFORT,H.-P. 1976. Les modkles grico-bactriens de quelques reliquaires et ~ a l e t t e s
ifards 'grCco-bouddhiques'. AA, 32, pp. 91-8.
-. 1979. Les fortifications en Asie Centrale de /'Age du. Bronze a l'epoque Kouchane.
Paris. (Trav. URA, N o . 10, I.)
-. 1984. Le sanctuaire du temple i niches indentees. 2: Les trouvailles. Fouilles d'Ai
Kbanoum, Vol. 111. Paris. (MDAFA, 27.)
FRASEK,P. M. 1979. AS, 2, pp. 9-21.
GARL)IN,J.-C. 1957. CPrarniqcres dc Bacrres. Paris. (MDA FA, 15.)
 Bibliography and References
-

G A K D I NJ.-C.
, 1973. In: Fouilles d'Ai'Khanoum,Vol. I, pp. 121-88. Paris. (MDAFA, 21.1
G A R D I N ,J.-C.; LYONNET, B. 1978179. La prospection archCologique de la B~~
riane orientale (1 974-1 978): premiers risultats. Mesopotamia, Vols. 13-14, pp, 99-
154. Turin.
GENTLLLE,P. 1978. ~ t u d geographique
e de la plaine d'Ai Khanoum et de son irrigation
depuis les temps antiques. Paris, ~ d i t i o n sdu CNRS.
GHIRSHMAN, R. 1946. Begram, Recherches arche'ologiques et historiques sur les K ~ ~ . .
chans. ParislCairo. (MDAFA, 12; MIFAOC, 79.)
GHOSH, A. 1947148. Taxila (Sirkap). Ancient India, Vol. 4, pp. 41-83. New Delhi.
GRENET,F. 1983. L'Onornastique iranienne i Ai'Khanourn. BCH, 107, pp. 373-81.
GUILLAUME, 0 . ; ROUGEULLE,A. 1987. Les petits objets. Fouilles d'Ai' Khanoum, vol.
VII. Paris. (MDAFA, 31.)
HOLT, F. 1981. The Euthydernid Coinage of Bactria: Further Hoard Evidence from ~ i .
Khanoum. RN, 23, pp. 7 4 4 .
HOUGHTON,A. A.; LE RIDER,G. 1966. U n trisor de rnonnaies h e l l i n i ~ t i ~ u etrouvi s
pris de Suse. RN, 6, pp. 11 1-27.
KOZLOVSKIY, V. A.; NEKRASOVA, E. G . 1976. Stratigraficheskiy shurf na tsitadelli Drev-
nego Terrneza. Baktri~skiedrevnosti (Leningrad), pp. 30-9.
KRUGLIKOVA, I. T. 1974. Dil'berdzhin: Raskopki 1970-1972 gg., Vol. 1. Moscow.
-. 1976. Nastennie rospisi Dil'berdzhina. Drevnyaya Baktriya, Vol. 1, pp. 87-110.
-. 1977. Les fouilles de la Mission archiologique soviito-afghane sur le site grico-
kushan de Dilberdjin en Bactriane (Afghanistan). CRAI, pp. 407-27.
-. 1979. Nastennie rospisi v porneshchenii 16 severo-vostochnogo kul'tovogo kom-
pleksa Dil'berdzhina. Drevnyaya Baktriya, Vo1. 2, pp. 120-45.
KRUGLIKOVA, I. T.; PUGACHENKOVA, G. A. 1977. Dil'berdzhin (Raskopki 1970-1973
gg.), Vol. 2. Moscow.
KRUGLIKOVA,I. T.; SARIANIDI,V. I. 1971. Drevnyaya Baktriya v svete novikh ar-
kheologicheskikh otkritiy. SA, 4, pp. 154-77.
LAHIRI,A. N. 1965. Corpus of Indo-Greek Coins. Calcutta.
LERICHE,P. 1986. Les rernparts et les monuments associis. Fouilles dJAi'Khanoum, Vol.
V. Paris. (MDAFA, 29.)
LERICHE,P.; T H O R A V A J. ~ ~1979.
, La fontaine du rempart de l'oxus i Ai' Khanoum.
Syria, 56, pp. 171-205.
LE RIDER,G. 1965. Suse sous les SPleucides et les Parthes: Les trouvailles monetaires et
l'histoire de la ville. Paris. (MDAFI, 38.)
LEZINE,A. 1963164. Hkrat. Notes de voyage (1). Bulletin d'itudes orientales, Vol. 18,
pp. 127-45.
LITVINSKY, B. A.; PICHIKYAN, I. R. 1979. Kushanskie Croti. VDI, 2, pp. 89-109.
-. 19816. Dicouvertes dans un sanctuaire du dieu Oxus de la Bactriane septentrionale.
RA, 2, pp. 195-216.
-. 1981b. Nozhni akinaka iz Baktrii. VDI, 3, pp. 87-110.
MACDONALD,G . 1922. Note to Chapter XIV: Ancient Persian Coins in India; Note to
Chapter XV: Ancient Greek Coins in India; Chapter XVII: The Hellenistic King-
doms of Syria, I. In: E. J. Rapson (cd.), Ancient India, pp. 342-4, 386-90, 427-66,
London.
M A N A N D I A NH. , A. 1965. The Trade a n d Cities of Armenia in Relation to Ancient
World Trade, pp. 47-52. Lisbon.
 Bibliography and Referentrs

J. H . 1951. Taxila, A n Illustrated Account of Archaeological Excavations

MARSHAI-I-,
Carried O u t at Taxila under the Orders of the Govcrnment of India between tht#
Years 1913 and 1934, Vols. 1-111. Cambridge, Cambridge University Press.
MASSON,M. E.; PUGACHENKOVA, G . A. 1959. Parfyanskie ritoni Nisi. Trudi Y&TAKE,
4.
MINORSKY, V. 1967. A Greek Crossing on the Oxus. BSOAS, 30, pp. 45-53.
MITCHINER,M. 1975. Indo-Greek and Indo-Scythian Coinage, Vols. 1-111. London,
Hawkins.
MQRKHOLM,0. 1965. A Greek Coin Hoard from Susiana. Acta A~chaeolo~ica, Vol. 36,
pp. 127-56. Copenhagen.
NAKAIN,A. K. 1957. The Indo-Greeks. Oxford, Oxford University Press.
NEGMATOV,N . N.; BELYAEVA, T. V. 1977. Raskopki na tsitadeli Leninabada i lokaliza-
tsiya Aleksandrii Eskhati. A O , 1976, p. 569.
PETITOT-BIEHLER, C.-Y. 1975. TrCsor de monnaies grecques et grico-bactriennes trouvi
i Ai' Khanoum (Afghanistan). R N , 6th series, Vol. 17, pp. 23-57.
PICHIKYAN,I. R. 1980. Nozhni ksifosov i makhayr v Severnoy Baktrii. SA, 4,
pp. 202-12.
-. 1991. Kul'tura Baktrii. Akhemenidskiy i allinisticheskzy Period. Moscow.
PUGACHENKOVA, G. A. 1979. Zhiga-tepe (raskopki 1974 g.) Drevnyaya Baktriya, Vol.
2, pp. 63-94.
RAPIN,C. 1992. La trisorerie d u palais hellinistique d'Ai Khanoum. L'apogie et la chute
du royaume grec de Bactriane. Fouilles dJAi'Khanoum, Vol. VIII. Paris. (MDAFA,
33.)
RAPIN,C.; GRENET,F. 1983. Les inscriptions Cconomiques de la trisorerie hellenistique
d'Ai' Khanoum (Afghanistan). BCH, 107, pp. 315-72.
RAPSON,E. J. 1922. Chapter 11: A. Peoples and Languages, B. Sources of History; Chap-
ter XIII: T h e Puranas; Chapter XXI: Indian Native States after the Period of the
Mauryan Empire; Chapter XXII: The Successors of Alexander the Great; Chapter
XXIII: The Scythian and Parthian Invaders. In: E. J. Rapson (ed.), Cambridge His-
tory of India. Vol. I: Ancient India, pp. 37-64, 296-318, 514-92. Cambridge, Cam-
bridge University Press.
RAYCHAUDHURI, H. 1920. Materials for the Study of the Early History ofthe Vaishnava
Sect. Calcutta.
ROBERT, L. 1973. Les inscriptions. Fouilles d'Ai' Khanoum, Vol. 1, pp. 207-37. Paris.
(MDAFA, 21.)
RTVEI-ADZE, E. V. 1977. K lokalizatsii 'grecheskoy' pereyravi na okse. V D I , 4, pp. 182-8.
SCHLUMBERGER, D. 1947. In: C R A I , pp. 241-3.
-. 1972. In: C R A I , pp. 188-99.
SCHLUMBEKGER, D.; BERNARD,P. 1965. Ai Khanourn. BCH, 89, 2, pp. 590-657.
SELTMAN,C. T . 1955. Greek Coins: A History of Metallic C u ~ r e down n ~ ~ to the Fall of
the Hellenistic Kingdoms, 2nd ed. London.
SEYRIG,H. 1973. Tresors du Levant anciens et nouveaux. Paris.
SHARMA,G. R. 1980. Reh Inscription of Menandel* and the Indo-Greek Invasion of the
Ganga Valley. Allahabad.
SIHISHKINA, G. V. 1969. 0 mestonakhozhdenii Marakandi. SA, 1, pp. 62-75.
-. 1975. E l l i n i s t i ~ h e s k a ~kcranlika
a Afrasiaba. SA, 2, pp. 60-78.
-. 1985. Les remparts de Samarcande iI'Cpoque hellCnistique. In: P. ~ e r i c h e(ed.), La
 Bibliography and Rcfcrc~ncc,

fortif;'cation et sa place dans l'histoire politique, sociale et culturclle du mondf grec,

pp. 71-8. Paris, CNRS.
STKAUSS,P. 1971. U n trcsor de monnaies hell6nistiques trouvk prks d e Suse: 2. RN, 13,
pp. 109-40.
T A R N ,W. W. 1951. The Greeks in Bactria and India, 2nd ed. Cambridge, Cambridge
Univcrsity Press.
, 1976. Hadda i la lumikre des trois dernikres camyagnes d e fouilles de Taps-6-
T A K Z I 2.
Shotor (1974-1 976). C R A I , 1976, pp. 381-410.
VEUVE,S. 1982. Cadrans solaires grCco-bactriens i A i Khanoum (Afghanistan). BCH,
106, 1, pp. 23-51.
-. 1987. Le gymnase. Architecture, ciramique, sculpture. Fouilles d'Ai' Khanoum,
Vol. VI. Paris. ( M D A F A , 30.)
WHEELER,R. E. M. 1962. Charsada, A Metropolis on the North-west frontier. Being a
Report on the Excavations of 1958. London, British Academy.

Debevoise, N. C . 1938. A Political History of Parthia. Chicago, University of Chicago

Press.
D'YAKONOV,M. M. 1961. Ocherk istorii drevnego Irana. Moscow.
D'YAKONOV,I. M.; LIVSHITS,V. A. 1960. Dokumenti iz Nisi I v, do n.;. Moscow.
FISCHEII,T. 1970. Untersuchungen z u m Partherkrieg Antiochus V I I i m Rahmen der
Seleukidengeschichte. Munich.
G H I R S H M A NR., 1962. Iran, Parther und Sasaniden. Munich, C . H. Beck.
-. 1976. Terrasses sacrkes de Bard-; Ne'chandeh et Masjid-i Solaiman, Vols. 1-11. Paris.
( M D A F I , 45.)
KOSHELENKO,G . A. 1963. V n ~ t r i ~ o l i t i c h e s k a ybor'ba
a v Parfii. V D I , 3, pp. 56-68.
-. 19666. T h e Beginning of Buddhism in Margiana. Acta Ant. Hung., 14, pp. 175-83.
-. 1966b. Kul'tura Parfii. Moscow.
-. 1971a. Charskaya vlast' i ee obosnovanie v ranney Parfii. In: B. G . Gafurov, E. A.
Grantovskiy and M. S. Ivanov (eds.), Istoriya iranskogo gosudarstva i kul'turi k
2500-letiyu Iranskogo gosudarstva, pp. 212-18. Moscow. (English summary, p.
343.)
-. 1971b. La politique commerciale des Arsacides et les villes grecques. Studi in Onore
di Edoardo Volterra, Vol. I, pp. 761-5. Milan.
-. 1972. Monetnoe delo Parfii pri Mitridate I. Numismatika i ;pigrafika, Vo1. I. MOS-
COW.
-. 1976. Genealogiya pervikh Arshakidov (eshche raz o nisiyskorn ostrake). Istoriya i
kul'tura narodov Sredney Azii, pp. 31-7. Moscow.
-. 1977. Rodina Parfyan. Moscow.
-. 1979. Grecheskiy polis na ~llinisticheskomVostoke. Moscow.
-. 1980. Les cavaliers parthes. Aspects d e la structure sociale d e la Parthie. Dialogues
d'histoire ancienne, Vol. 6, pp. 177-99. Paris.
KOSHELENKO,G. A.; NOVIKOV,S. V. 1979. Manumissii Selevkii na 1:vIcc. V D I , 2, pp.
41-54.
LEPPER,F. A. 1979. Trajan's Parthian War. London, Greenwood Press.
 Bibliography and References

LE RIDER,G . 1965. Suse sous les Sileucides et les Parthrs: Les trouvaillcs monktaircs de
l'histoire de la ville. Paris. (MDAFI, 38.)
, G . 1961. Iran v ipokhu pervikh Sasanidov. Leningrad.
L U K O N I NV.
MCDOWELL,R. H. 1935a. The Coins from Seleucia on the Tigris. Ann Arbor.
-. 19356. Stamped and Inscribed Objects from Seleucia on the Tigris. Ann Arbor.
PILIPKO,V. N. 1980. Parfyanskie bronzovie moneti so znakom P pod lukom. VDI, 4,
pp. 105-24.
PUGACHENKOVA, G. A. 1967. Iskusstvo Turkmenistana. Moscow.
ROSTOVTZEFF,M . 1938. Dura-Europos and its Art. Oxford, Oxford University Press.
SARIANIDI, V. I.; KOSHEI-ENKO, G . A. 1982. Moneti iz raskopok nekropolya, raspolo-
zhennogo na gorodishche Tillya-tepe (Severniy Afganistan). Drevnyaya Indrya.
Istoriko-kul'turnie svyazi. Moscow.
SARKISYAN, G . K. 1952. Samoupravlyayushchiysya gorod Selevkidskoy Vavilonii. VDI,
1, pp. 68-83.
SELLWOOD,D. 1980. An Introduction to the Coinage of Parthia. London, Spink & Son.
WELLES,C . B.; FINK,R. 0 . ; G I L L I A MJ., F. 1959. The Parchments and Papyri (The Exca-
vations at Dura-Europos. Final Report V, Part 1). New Haven, Conn.
WENKE,R. J. 1975/76. Imperial Investments and Agricultural Developments in Parthian
and Sasanian Khuzestan: 150 B.C. to A.D. 640. Mesopotamia, Vols. 10-1 1, pp. 3 1-
221.

ANON. 1979. A Short History of the Peoples Inhabiting Northern China in Ancient
Times. Huhehot. (In Chinese.)
BAI SHOUYI. 1980. Outline History of China. Shanghai. (In Chinese.)
BAN G u . Ch'ien Han Shu. Bonaben. (Chapters on the Huns, the Xianbi and the Ukhan;
in Chinese.)
BICHURIN,N. Y. 1950. Sobranie svedeniy o narodakh obitavshikh v Sredney Azii v dre-
nie vremena, Vol. I. Moscow/Leningrad.
BIRA, S. 1977. Problems of History, Culture and Historiography of the Mongoliun
People's Republic. Ulan Bator.
BIRA, S.; WATSAGDONJ,S.; PERLEE,IS.,et al. (eds.). 1984. History of the Mongolian
People's Republic. Ulan Bator. (In Mongolian.)
CHENSHOU. San Guo Zhi. Bonaben. (Chapters on the nomadic peoples of Central Asia
and their relations with China and other countries; in Chinese.)
DEBEC,G . F. 1941. Paleoantropologiya USSR. Moscow/Leningrad.
DE GUIGNES,J. 1756-58. Histoire gPnPrale des Huns, des Turcs, des Mongols et des autres
Tartares Occidentaux, Vols. I-IV. Paris.
DORZF-ISUREN, T. 1958. The Archaeological Excavations Conducted in Arkhangdi Aimak
of the MPR in 191i6-1957. Ulan Bator. (In Mongolian.)
-. 1961. The Northern Huns. Ulan Bator. (In Mongolian.)
FANYE. Hou Hun Shu. Bonaben. (Chapters o n the Huns, the Xianbi and the Ukhan; in
Chinese.)
GUMIL-EV L., N. 1960. Khunnu. Moscow.
HAHMATTA, J. 1952. The Dirsulution of the H u n Empire: I . The H u n State. AAH, VOI.
2, pp. 277-304.
 Bibliography and References

History of the Mongolian People's Republic. 1966. V O ~ S1-111.

. MOSCOW. (In Mongolian.)
ISHJAMTS, N . 1955. The Ethnic Affiliation of the H u n s and their Social Structure. works
of the MPR Committee for the Sciences (History and Philology), No. 1. Ulan
Bator. (In Mongolian.)
-. 1965a. The Formation of the Mongolian People. Studia Ethn~~raphica, Val. 2,
Ulan Bator. (In Mongolian.)
-. 1965b. The Question of the Ethnic Affiliation of the Tribes Inhabiting Mongolia in
Ancient Times. Studia Ethnographica, Vol. 2. Ulan Bator. (In Mongolian.)
-. 1974. The Formation of a Unified Mongolian State and the Establishment of Feu-
dalism in Mongolia. Ulan Bator. (In Mongolian.)
KISELEV,S. V. 1951. Drevnyaya istoriya Yuzhnoy Sibiri. Moscow.
KONOVALOV,I. B. 1976. Khunnu v Zabaykal'e. Ulan Ude.
LATTIMORE,0. 1951. Inner Asian Frontiers of China. New York.
LAUFER,B. 1914. Chinese Clay Figures. Part I: Prolegomena on the History of Defensive
Armour. Chicago, Field Museum. (Field Museum of Natural History Publication,
177, Anthropological series, 111, N o . 2.)
LIN GAN. 1979. History of the Huns. Huhehot. (In Chinese.)
MA CHANGSHOU.1972. The Northern Di and Hunnu. Beijing. (In Chinese.)
MAIDAR,D.; DAR'SUREN,L. 1976. The Yurt. Ulan Bator. (In Mongolian.)
NOVGORODOVA,E. A. 1984. Mir petroglifov Mongolii. Moscow.
OKLADNIKOV, A. P. 1955. Neolith i bronzoviy vek Zabaykal'ya. MoscowILeningrad.
Outline History of China. 1981. Beijing. (In Chinese.)
PERLEE,H. 1961. A Short History of Ancient and Medieval Towns and Settlements of the
MPR. Ulan Bator. (In Mongolian.)
ROSTOVTZEFF,M. I. 1929. The Animal Style in South Russia and China. Princeton1
London, Princeton University Press. (Monographs in Art and Archaeology,
14.)
RUDENKO,S. I. 1962. Kul'tura khunnov i Noin-ulinskie kurgani. MoscowILeningrad.
SERODZHAV, N . 1977. The Ancient History of Mongolia. Ulan Bator. (In Mongolian.)
SHIRATORI,K. 1900. Uher die Sprache des Hsiung-nu-Stammes und der Tung-hu-
Stamme. Tokyo.
SIMAGUANG. 1926. Zizhi Tongjian. Beijing. (Chapters and paragraphs on the ancient
nomads; in Chinese.)
SIMAQIAN(SZU-MACH'IEN)(ed.) 1931. Shiji. (Chapters on the history of the Huns, the
Xianbi and the Ukhan; in Chinese. A Russian translation of Vol. I was published
in Moscow in 1972.)
-. 1972. Istoricheskie zapiski [Shiczi]. Moscow.
SUKHBAATAR, G. 1980. Ancestors of the Mongols: the Huns. Ulan Bator.
TASKIN,V. S. 1968. Materiali po istorii Syunnu (po k i t a ~ s k i mistochnikam). Moscow.
(Foreword, translation and commentary.)
TREVEK,K. V. 1940. Pamyatniki greko-baktriyskogo iskusstva. Leningrad.
UMEHAKASUEJI.1960. Studies of Noin-Ula Finds in North-Mongolia. Tokyo. (TOYO
Bunko Publication, 27.)
URAY-KOHALMI, K. 1974. La ~Criodisationde I'histoire des arrnements des nomades des
steppes. ~ t u d e smongoles, Vol. 5, pp. 145-55. Paris.
W A N G GUOWEI.1927. Studies on the Kuei-fang, P'i-i and H s i e n - ~ i i ntribes. collected
Works, Vol. I , pp. 1-10. (In Chinese.)
 Bibliography and References

YUE YUANAN;YUE BAYAN.1958. An Outline of the History of Inner Mongolia. Chap-

ter I. Shanghai. (In Chinese.)

BERZINA,S. Y. 1982. D r e v n y a ~ aIndiya i Afrika. Drevnyaya Indiya. Istoriko-kul'tuynie

svyazi. Moscow.
, Y. 1950. Sobranie svedeniy o narodakh, obitavshikh v Sredney Azii v
B I C H U R ~ NN.
drevnie vremena. Vol. 2. Moscow/Leningrad.
COLLEDGE, M. A. R. 1967. The Parthians. London.
Dal'verzintepa. Kushanskiy gorod nu yuge Uzbekistana. 1978. Tashkent.
DAVIDOVICH, E. A. 1976. Perviy klad tetradrakhm kushantsa 'Geraya'. V D l , 4,
pp. 56-78.
DEBEVOISE, N. C. 1938. A Political History of Parthia. Chicago.
D'YAKONOV, M. M. 1961. Ocherk istorii drevnego Irana. Moscow.
EGAMI,N . 1948. Azia-Minzoku to bunka no keisei. Tokyo.
-. 1951. Gesshi n o minzokumei ni tsuite. Oriental Studies Presented to Sei Wadu, pp.
123-32. Tokyo. (English abstract, pp. 19-21 .)
GULLINI, G . 1964. Architettura iranica dagli Achemenidi ai Sasanidi. I1 'palazzo' di
Kuh-i Khwagia (Seistan). Turin.
GUMILEV,L. N. 1960. Khunnu. Moscow.
HALOUN, G . 1937. Zur U e - t $ - ~ r a ~ eZDMG,
. 91, pp. 243-318.
HUDSON,G . F. 1931. Europe and China: A Survey of their Relations from the Earliest
Times to 1800. London.
IL'IN,G . F. 1958. Drevniy indiyskiy gorod Taksila. Moscow.
KISELEV,S. V. 1951. Drevnyaya istoriya yuzhnoy Sibiri. MIA, 9.
KOSHELENKO,G . A. 1971. La politique commerciale des Arsacides et les villes grecques.
Studi in onore di Edoardo Volterra, Vol. I, pp. 761-5. Milan.
-. 1974. Literatura p o arkheologii Sredney Azii i Kazakhstana v 1971 godu. (Criticism
of the Look by E. A. Pugachenkova, Skul'ptura Khalchayana, 1971.) SA, 2, pp.
304-5.
-. 1980. Les cavaliers parthes. Aspects de la structure sociale de la Parthie. Dialogues
d'histoire ancienne, Vol. 6. Paris.
KUWABARA, J. 1939. Tbzai kbshbshu ronsb. Tokyo.
-. 1940. Tbzai kotsushi rons6, 5th ed. Kyoto.
LITVINSKY, B. A. 1972. Drevnie kochevniki 'krishi mira'. Moscow.
MA FEI-1'0. 1979. Kuan-tzu Ch'ing-chung-pien hsin-ch'iian, Vols. 1-11. Beijing.
-. 1982. Ch'in chi-shih. Beijing.
MANDEL'SHTAM, A. M. 1966. Kochevniki nu puti v Indiyu. ~ o s c o w / L e n i n ~ r a d .
-. 1975. Pamyatniki kochcvnikov kushanskogo vremeni v Severnoy Baktrii. Leningrad.
MARSHALL,J . H. 1951. Taxila, An Illustrated Account of Archaeological Excavations
Carried Out at Taxila under the Orders of the Government of India between the
Years 1913 and 1934, Vols. 1-111. Cambridge, Cambridge University Press.
MASSON,V. M. 1974. Problerna drevnego goroda i a r k h e ~ l o ~ i c h e s k~i ea m ~ a t n i Sever-
ki
noy Baktrii (Perspektivi i s s l e d o v a ~ ~ i ~ aDrevnyaya
). Baktriya ~redvaritel'nie
Soohsbchcnij~a ob arkhe~lo~icheskikh rahotakh nn yuge Uzbekistana, V O ~ .4.
Leningrad.
 Bibliography and References

MASSON,V. M. 1976. Kushanskie poseleniya i kushanskaya arkheologiya. Baktyiyskie

drevnosti, pp. 3-17. Leningrad.
MASSON,V. M.; ROMODIN,V. A. 1964. Istoriya Afganistana. Moscow.
MATSUDA,H . 1939. Gushi n o gyoku to kBkan n o shu. T6zai k6sh6shi rons6, pp. 157-86.
Tokyo.
MIA. 1949.
N A R A I N ,A. K. 1962. The Indo-Greeks, 2nd ed. Oxford, Oxford University Press.
OBEL'CHENKO,0.V. 1974. Kurgannie rnogil'niki tpokhi kushan v Bukharskom oazise.
Tsent~al'na~a Aziya v Kushanskuyu ipokhu. Moscow.
PELLIOT, P. 1929. ~ ' ~ d i t i ocollective
n des oeuvres de Wang Kouo-wei. TP, 26,
pp. 113-82.
POTAPOV, L. P. 1953. Ocherki po istorii alta~tsev.Moscow/Leningrad.
PUGACHENKOVA, G . A. 1966. Khalchayan. K probleme k h ~ d o z h e s t v e n n okul'tuyi
~
Severnoy Baktrii. Tashkent.
-. 1971. Skul'ptura K h a l ~ h a ~ a n aMOSCOW.
.
PULLEYBLANK, E. G . 1966. Chinese and Indo-Europeans. JRAS, 1, pp. 9-39.
-. 1968. Chinese Evidence for the Date of Kaniska. In: A. L. Basham (ed.), Papers on
the Date of Kani~ka,pp. 247-58. Leiden.
RICE, T. T. 1957. The Scythians. London.
RUDENKO,S. I. 1951. Der zweite Kurgan von Pasyryk. Berlin, Verlag Kultur und For-
schritt.
-. 1953. Kul'tura naseleniya Gornogo Altaya v skifrkoe vremya. Moscow/Leningrad.
-. 1960. Kul'tura naseleniya Tsentral'nogo Altaya v skifSkoe vremya. Moscow/Lenin-
grad.
SARIANIDI,V. I. 1984. Afganistan: sokrovishcha bezimyannikh tsarey. Moscow.
SARIANIDI, V. I.; KOSHELENKO,G . A. 1982. Moneti iz raskopok nekropolya, raspoloz-
hennogo na gorodishche Tillya-tepe. Drevnyaya Indiya. Istoriko-kul'turnie
svyazi. Moscow.
SELLWOOD,D. 1980. A n Introduction to the Coinage of Parthia, 2nd ed. London, Spink
& Son.
SERDIT~KH,K. V. 1983. (Review of Sharma, 1980). Informatsionniy B~ulleten'
MAIKTSA, N o . 5, pp. 84-8.
SHARMA,G . R. 1980. Reh Inscription of Menander and the Indo-Greek Invasion of the
Ganga Valley. Allahabad.
T A R N ,W. W . 1938. The Greeks in Bactria and India. Cambridge, Cambridge University
Press.
T6zai kosh6shi no kenkyti, Saiiki-hsu. 1943. 2nd ed. Tokyo.
WADA,S. 1939. ManmBshi n o taisei [A Survey of the History of Manchuria and Mongo-
lia]. Azia mondai k&a, Vo1. 7. Tokyo.
-. 1942. Toashi rons6. Tokyo.
W A N G KUO-WEI.1927. O n the Place of the Yiieh-chih Beforc Their Migration to Ta-
hsia. Collected Works. (Supplement to the Kuan-t'ang ~ i e h - c h iPU-i, 10; in Chi-
nese.)
WHEELER,R. E. M.; GHOSH, A.; DEVA,K. 1946. Arikamedu: An Indo-Roman Trading
Station o n the East Coast of India. Ancient India, N o . 2, pp. 17-124.
WHITKOMB,D . S.; JOHNSON,J. H. 1980. Qossier el-Qadim und der Rote-Meer Handel.
Altertum, Vol. 26, N o . 2, pp. 103-12.
 Bibliography and References

YADAVA,B. N . S. 1978179. The Accounts of the Kali Age and the Social Transition from
Antiquity t o the Middlc Ages. Indian Historical Review, Vol. 5, Nos. 1-2.
YANG HSIEN-I.1983a. I-yii yii-shih m o t e s Obtained by Chance]. Beijing.
-. 1983b. Ling-mo hsin-chien [Notes Fragmentary but New]. Hong Kong.
ZEIMAL,E. V. 1978. Politicheskaya istoriya drevney Transoksiani po nurnizrnaticheskim
dannim. Kul 'tura Vostoka. Drevnost ' i rannee srednevekov'e. Leningrad.
ZURCHEK,E. 1968. T h e Yiieh-chih and Kaniska in Chinese Sources. In: A. L. Basham
(ed.), Papers on the Date of Kani!ka, pp. 346-90. Leiden.

ALLAN,J. 1936. Coins of Ancient India. London, British Museum.

ALRAM,M. 1983. Pahlavas: Hybouzanes ein neuer Konig und eine neuc Lesung fiir Gon-
dophares. Literal Numismatical Vindobonenses, Vol. 2, pp. 68-74.
BAGCHI,P. C. 1955. India and Central Asia. Calcutta.
BANERJEA,J . N . 1941. The Development of Hindu Iconography. Calcutta.
BHANDARKAR,D. R. 1902. A Kushana Stone Inscription, and the Question about the
Origin of the Saka Era. JBBRAS, 20, pp. 269-302.
-. 1941. Indo-Parthian Dynasty of Vonones. Indian Culture, Vol. 7, pp. 133 et seq.
BIVAR,A. D. H. 1984. Maues at Taxila. J C A , 7, pp. 5-15.
C A H . 1966. Vol. IX. (Eds. S. A. Cook, F. E. Adcock and M. P. Charlesworth.) Cam-
bridge, Cambridge University Press.
CAMERON, G . G . 1937. History of Early Iran. Chicago.
CRIBB,J . 1985. N e w Evidence of Indo-Parthian Political History. Coin Hoards, Vol. 7,
pp. 282-300.
DANI, A. H. 1983. Chilas. The City of Nanga Parvat (Dyamar). Islamabad.
-. 1985. T h e Sacred Rock of Hunza. J C A , 8, pp. 5-131.
DEBEVOISE,N . C. 1938. A Political History of Parthia. Chicago.
Encyclopadia Britannica. 1973.
FLEET,J . F. 1907. Inscription on the Sohqaura Plate. JRAS.
FUSSMAN,G . 1980. Nouvelles inscriptions Saka. BEFEO, 67, pp. 1 4 3 .
-. 1985. Deux didicaces kharoslhi. BEFEO, 74, pp. 29-34.
-. 1986. Documents epigraphiques kouchans IV. BEFEO, 75, pp. 1-14.
GARDNER,P. 1886. The Coins of the Greek and S9rthic Kings of Bactria and India in the
British Museum. London, British Museum.
GHIRSHMAN, R. 1961. Iran from the Earliest Times to the Islczrnic Conquest, 2nd. ed. Bal-
timore, Md., Penguin Books.
HELMS,S. W. 1982. Excavations at the City and the Famous Fortress of Kandahar. AS,
3/4, pp. 1-24.
HIRTH,F. 1917. The Story of Chang-K'ien, China's Pioneer in Western Asia. J A O S , 37,
pp. 89-152.
HUART,C . 1927. A n b e n t Persia and the Iranian Civilization. London, Kegan Paul.
JACOHI,H . 1880. Das Kilakicirya-Kathinakam. Z D M G , 34, pp. 247-318.
JAMES,M. R. 1924. Tl7c Apocryphal New Testament. Oxford, Oxford University Press.
JENKINS,G. K. 1955. Indo-Scythic Mints. JNSI, 17, pp. 1-26.
KII;I.HOI<N, F. 1905/06. Junagadh Rock-Inscription of Rudradaman. EI, 8, pp. 3 6 4 9 .
 Bibliography and References

K o ~ o w S., 1929. Kharoshthi Inscriptions with the Exception of Those of ASoka. CII,
Vol. 2, Part 1. Calcutta.
-. 1933. Notes on Indo-Scythian Chronology. Journal of Indian History, Vo]. 12,
pp. 1-46.
J. E. V A N . 1949. The 'Scythian' Period: A n Approach to the His-
L O I < U I Z E N - ~LEEUW,
)E
tory, Art, Epigraphy and Paleography of North India from the 1st Century B,C to
the 3rd Century A.D. Leiden.
LOZINSKI,B. P. 1959. The Original Homeland of the Parthians. The Hague.
MACDOWALL,D. W. 1965. The Dynasty of the Later Indo-Parthians. N C , 1965,
pp. 13748.
-. 1968. Soter Megas, The King of Kings, the Kushana. JNSI, 30, pp. 28-48.
-. 1973. The Azes Hoard from Shaikhan-Dheri: Fresh Evidence for the Context of
Jihonika. South Asian Archaeology, pp. 215-30. London.
-. 1985. Azes successors of the Indo-Greek at Begram, Afghanistan. South Asian
Archaeology, Vol. 11. Naples.
-. 1989. The Copper Coinage of Demetrius, The Son of Euthydemus. South Asian
Studies, Vol. 5, pp. 29-33.
MAJUMDAR,R. C. (ed.). 1951. The Age of Imperial Unity. Bombay.
MAJUMDAR, R. C.; PUSALKAR, A. D. (eds.). 1960. The History and Culture of the Indian
People. Bombay. (See Chapter IX on the age of imperial unity.)
MARSHALL, J. H . 1936. A Guide to Taxila, 3rd ed. Delhi.
-. 1947. Greeks and Sakas in India. JRAS, 1947, pp. 3-32.
-. 1951. Taxila, A n Illustrated Account of Archaeological Excavations Carried Out at
Taxila under the Orders of the Government of India between the Years 1913 and
1934, Vols. 1-111. Cambridge, Cambridge University Press.
-. 1960. Buddhist Art of Gandhara: The Story of the Early School, Its Birth, Growth,
and Decline. Cambridge, Cambridge University Press, for the Department of
Archaeology in Pakistan.
MCGOVERN,W. M. 1939. The Early Empires of Central Asia, A Study of the Scythians
and the Huns, and the Part They Played in World History, with Special Reference
to the Chinese Sources. Chapel Hill, N.C., University of North Carolina Press.
MINNS,E. H. 1908-25, 1961. Scythians. In: J. Hastings (ed.), Encyclopedia of Religion
and Ethics, Vol. 11, pp. 276a-279a. Edinburgh/New York, T. & T. Clark.
-. 1913. Scythians and Greeks: A Survey of Ancient History and Archaeology on the
North Coast of the Euxine from the Danube to the Caucasus. Cambridge, Cam-
bridge University Press.
-. 1925. The Scythians and Northern Nomads. In: J. B. Bury, S. A. Cook and F. E.
Adcock (eds.), Cambridge Ancient History. Vol. 111: The Assyrian Empire,
pp. 187-205. Cambridge, Cambridge University Press.
MITCHINER, M. 1976. Indo-Greek and Indo-Scythlan Coinage, Vol. VIII. London, Hawkins.
, K. 1957. The Indo-Greeks. Oxford, Oxford University Press.
N A R A I NA.
NILAKANTHA SASTRI,K. A. (ed.). 1957. A Comprehensive History of India. Bombay.
PULLEYBLANK, E. G. 1968. Chinese Evidence for the Date of Kaniska. In: A. L. Basham
(ed.), Papers on the Date of K a n i ~ k a ,pp. 247-58. Leiden.
PURI,B. N . 1963. India in Classical Greek Writings. Ahmadabad.
-. 1965. India under the Kusdnas. Bombay.
-. 1968a. History of Indian Administration. Bombay.
 Bibliography and References

PURI,B. N. 1968b. India in the Time of Paratijali, 2nd. ed. Bombay.

RAPSON,E. J. 1908. Catalogue ofthe Coins of the Andhra Dj,narty, the Weslern Ksalya-
pas, the Traikutaka Dynasty, and the 'Bodhi' Dynasty. London, British Museum.
-. (ed.) 1922. Cambridge History of India. Vol. I : Ancient India. Cambridge, Cam-
bridge University Press.
ROSENFIELD,J. M. 1967. The Dynastic Arts of the Kushans. Berkeley/Los Angcles, Uni-
versity of California Prcss.
ROWLAND, B. 1953. The Art and Architecture of India: Buddhisi, Hindu, jain.
Harmondsworth, Penguin Books.
SELLWOOD,D. 1980. An Introduction to the Coinage of Parthia, 2nd cd. London, Spink
& Son.
SENART,E. 1905. The Inscriptions in the Caves at Nasik. EI, 8, pp. 56-96.
SIMONETTA, A. M. 1957. An Essay on the So-called Indo-Greek Coinage. EW, 8,
pp. 44-66.
-. 1958. A N e w Essay o n the Indo-Greeks. The Sakas and the Pahlavas. EW, 9,
pp. 154-83.
SMITH, V. A. 1906. T h e Indo-Parthian Dynasties from about 120 B.C. to 100 A.U.
ZDMG, 60, pp. 49-72.
-. 1907. T h e Sakas in Northern India. ZDMG, 61, pp. 403-21.
-. 1924. Early History of India, 4th rev. ed. Oxford, Oxford University Press.
SYKES,Sir P. M. 1922. Persia. Oxford, Oxford University Press.
TARN,W . W. 1930. The Invaders of Bactria. Seleucid-Parthian Studies, pp. 105-
35.ondon. (Proceedings of the British Academy, 16.)
-. 1951. The Greeks in Bactria and India, 2nd ed. Cambridge, Cambridge University
Press.
THOMAS,F. W . 1906. Sakastana, Where Dwelt the Sakas Named by Darius and Herodo-
tus? JRAS, pp. 181-216, 460-4.
V A L L ~ POUSSIN,
E L. DE LA. 1930. L'Inde aux temps des Mauryas. Paris.
WAKMINGTON,E. H. 1928. The Commerce between the Roman Empire and India.
Cambridge, Cambridge University Press.
WHITEHEAD,R. B. 1914. Catalogue of Coins in the Panjab Museum, Lahore. ~ o l I:.
Indo-Greek Coins. Oxford, Oxford University Press.
-. 1923. Notes on Indo-Greek Numismatics, NC.

ARCHAEOLOGICALRESEARCHINSTITUTEOF THE XINJIANGACADEMY01' SOCIAL

SCIENCES.1979. Jianguo yilai Xinjiang kaogu de zhuyao shouhuo [Major Archae-
ological Finds in Xinjiang since 19491. Wenwu kaogu gongzuo sanshi nian (1949-
1979) lJhirty Years of Archaeology 1949-791, pp. 169-83. Beijing, Wenwu Press.
ARCHAEOLOGICAL TEAM,INSTITUTEOF ETHNOLOGY,X I N J I A N GBRANCHOF ACA-
D E M I A SINICA. 1962. Zhaosu xian gudai muzang shijue jianbao [Prelinlinary
Report of the Excavation of the Ancient Tombs in Zhaosu County]. Wenwu,
N o . 7/8, pp. 98-102.
B E R N S ~ ~ I ' AA.MN, . 1952. Istoriko-arkheulogicl~eskieocherki Tsentral'nogo Tyan'-hanva
i Parnit-o-Alaya. M1,4, 26.
 Bibliography and References

CHIN GI-. 1981. Pami'er gaoyuan gulnu [Ancient Graves of the Pamir Plateau]. Kaogu
xuebao, N o . 2, pp. 199-215.
HUANGWI:.NRI.1933. Gaochang Pottery. Xibei kexue kaocha tuan congkan.
-. 1948. Luobunao'er kaoguji [An Archaeological Record of Lop Nor]. Beijing.
HULSEWE, A. F. P. 1979. China in Central Asia. The Early Stage: 125 B.C. - A.D. 23. (An
annotated translation of Chapters 61 and 96 of The History of the Former Hun
Dynasty, with an introduction by M. A. N . Lowe.) Leiden, E. J. Brill.
LI YUCHUN.1959. Xinjiang faxian de caitao [Painted Pottery Found in Xinjiang]. Kaogu,
No. 3, pp. 153-4.
MA YONG; WANG BINGHUA.1978. Xinjiang lishi wenwu [Historical Relics of Xin-
jiang]. Ed. by the Museum of the Uighur Autonomous Region of Xinjiang. Beij-
ing, Wenwu Press.
SAVINOV,D.; CHLENOVA,N. 1978. The Western Limits of the Distribution of the
'Reindeer Stone' and the Question of its Ethnic Nature. Mongolian Archaeology
and Ethnology. Novosibirsk, Soviet Union Scientific Press.
STEIN,A. 1974. O n Ancient Central Asian Tracts. 3rd ed. University of Chicago Press.
SUN PEI-LIANG.1984. The Trade Route of the Scythians and Stories of Ancient Central
Asia, Collected Essays on the History of Relations between China and the Rest of
the World, Vol. 1. Beijing, Zhonghua Publishers.
W A N GBINGHUA.1981. Xinjiang Alagou shukeng muguo mu fajue jianbao [Excavation
Report of the Vertical-pit Wooden Coffin Chamber Tombs of Alagou in Xin-
jiang]. Wenwu, N o . 2, pp. 18-22.
WANG MINGZHE.1981. Xinjiang Ke'ermuqi gu muqun fajue jianbao [Brief Excavation
Report of Ancient Graves at Ke'ermuqi in Xinjiang]. Wenwu, N o . 1, pp. 23-32.
W u ZHEN. 1964. Xinjiang dongbu de jichu xinshiqi shidai yizhi [Some Neolithic Sites in
Eastern Xinjiang]. Kaogu, N o . 7, pp. 333-41.
YANG YIYONG.1982. Xinjiang Mulei xian Sidaogou yizhi [The Sidaogou Site in Mulei
County, Xinjiang]. Kaogu, N o . 2, pp. 113-20.

ANON. 1960. Xinjiang Minfeng xian bei da shamo zhong gu yizhi muzang qu Dong Han
hezangmu qingli jianbao [Preliminary Report of the Conjoint Burial Tomb of the
Eastern Han Period in the Desert to the North of Minfeng County, Xinjiang].
Wenwu, N o . 6, pp. 13-21.
ARCHAEOLOGICAL RESEARCHINSTITUTE0 1 ' T H E X I N J I A N GACADEMY0 1 ; SOCIAL
SCIENCES.1979. Jianguo yilai Xinjiang kaogu de zhuyao shouhuo [Major Archae-
ological Finds in Xinjiang since 19491. Wenwu kaogu gongzuo sanshi nian (1949-
1979) [Thirty Years of Archaeology 1949-791, pp. 169-83. Beijing, Wenwu Press.
ARCHAEOL.OGICAL TEAM,INSTITUTEOF ETHNOLOGY,X I N J I A N GBRANCHOF ACA-
DEMIA SINICA.1962. Zhaosu xian gudai lnuzang shijue jianbao [Preliminary
Report of the Excavation of the Ancient Tombs in Zhaosu County]. Wenwu,
No. 718, pp. 98-102.
BERNSHTAM, A. H. 1948. Aravanskie naskal'nie izobrazheniya i davan'skaya (Fergan-
skaya) stolitsa irshi. SE, 4, pp. 155-61.
BUIIROW,T. 1937. Introduction. The Language of the Kharosthi. . Documents from Chi-
nese Turkestan. Cambridge, Cambridge University Press.
 Bibliography and References

H 0 u CAN.1980. Shilun Xiyu de Wuji xiaowei [An Investigation of the Wuji Colonel of
the Western Regions]. Xinjiang shixue, NO. 1.
-. 1981. Loulan yizhi kaocha jianbao [A Brief Keport on Investigations of the IAoulan
Sites]. Lishi dili, NO. I . Shanghai, People's Publishing Company.
HUANGLIE. 1981. Wei Jin Nanbeichao shiqi Xiyu yu neidi de p a n x i [Relations between
the Western Regions and the Interior of China during the Wei, Jin and Nanbci-
chao Periods]. Wei Jin Sui Tang shi lunji, Vol. 111. Beijing, Zhongguo shChui kexuc
chubanshe.
HUANGWENBI. 1948. Luobunao'er kaogu ji [Archaeology in the Lop-Nor Area].
Beijing.
-. 1958. Talimu pendi kaogu ji [Archaeology of the Tarim Basin]. Beijing, Kexue chu-
banshe.
-. 1981. Luobunao'er Hanjian kaoshi [Commentaries on the Han Tablets from I,op-
Nor]. Shanghai.
-. n.d.a. G u Loulan guo lishi jiqi zai Xiyu jiaotong shang zhi diwei n h e History of
the Ancient Kingdom of Loulan and its Position in the Communications in the
Western Regions]. Xibei shidi luncong [Collected Papers on the History and
Geography of the North-western Regions].
-. n.d.b. H a n tong Xiyu hou dui Xiyu zhi yinxiang [The Han Influence o n the West-
ern Regions after the Opening U p of Communications]. Xibei shidi lwncong [Col-
lected Papers o n History and Geography of the North-western Regions].
KOTANINAKOA. 1970. Sai to Rouzhi [The Sai and the Yueh-chih]. Toyoshi kenkyu, Vol.
28.
KUSHAEV,G. A. 1963. Drevnyaya kul'tura sakov i usuney dolini reki Ili. Alma Ata, Izda-
tel'stvo A N Kaz. SSR.
L u o ZHENYU;WANG GUOWEI.1914. Liusha zhuijian [Surviving Documents from the
Desert].
MA YONG (MENG CHI). 1975. Cong Xinjiang lishi wenwu kan Handai Xiyu de
zhengzhi cuoshi he jingji jianshi [Observations on Han Political Measures and
Economic Activities in the Western Regions Based on Historical Relics from Xin-
jiang]. Wenwu, N o . 7-8, pp. 26-34.
-. 1980a. H a n Qiuci Zuojiangjun Liu Pingguo zuo ting song ji shi kaoding Pextual
Criticism and Annotation of a Series of Watch-towers by Liu Pingguo the Left
General of Qiuci of Han]. W e n w u jikan, N o . 2, pp. 45-58.
-. 1980b. Xinjiang suochu Quelu wenshu de duandai wenti [On the Dates of the
Kharosthi Documents Unearthed in Xinjiang]. Wenshi (Beijing, Zhonghua), No. 7.
-. 1981. T o n g Han C a o Q u a n Bei zhong youguan Xiyu de zhongyao shiliao [Impor-
tant Historical Materials concerning the Western Regions in the Cao Quan Stele of
the Eastern H a n Period]. Wenshi (Beijing, Zhonghua), No. 11.
-. 1983a. Gudai Shanshan, Yutian diqu Quelu wenzi ziliao zongkao [A C o m ~ r e h e n -
sive Survey of the Kharos~hi Documents of Ancient Khotan and ~hanshan].
Zhongguo minzu guwenzi yanjiu [Studies on the Ancient Scripts of China's
Nationalities]. Beijing, Chinese Social Sciences Publishing C o .
-. 19836. Xinjiang Quelu wenshu zhong de kosava ji qrrshu kao [A Study of the Term
kosavn in the K h a r o s ~ h Docu~nents
i being Equivalent to the Chinese Term qushu].
Zhongguo minzrr guwenzi yanjiu [Studies on the Ancient Scripts of China's
Nationalities]. Beijing, Chinese Social Sciences Publishing CO.
 Ribliography and References

MA Y ~ N GW ; A N G B I N G H U A1978.
. Xinjiang lishi wcnwu [Historical Relics of Xin-
jiang]. Ed. by the Museum of the Uighur Autonomous Region of Xinjiang. Bei;-
ing, Wenwu Press.
MAKSIMOVA, A. G . 1959. Usun'skie kurgani levoberezh'ya reki Ili. Izvestiya A N Kaz.
SSR. (Historical, Archaeological and Ethnographical series.)
SHI SHUQINC;.1962. Tan Xinjiang Minfeng Niya Yizhi [On the Remains of Niya, Min-
feng County, Xinjiang]. Wenwu, N o . 7-8, pp. 20-7.
X I A N A I . 1962. Hetian maqian ko [ O n the Horse Coins of Khotan]. Wenwu, No. 7-8,
pp. 60-3.
ZUEV,Y.A. 1957. K voprosu o yazike drevnikh Usuney. Vcstnik A N Kaz. SSR, No. 5,
pp. 61-74.
ZURCHER,E. 1968. The Yiieh-chih and Kaniska in Chinese Sources. In: A. L. Basham
(ed.), Papers on the Date of Kaniska, pp. 346-90. Leiden.

A. S. 1950. Buxar Hoard of Kushana Coins. JNSI, 12, pp. 121 et seq.
ALTEKAK,
ASIAR.
BANERJEA, J. N . 1957. The Satraps of Northern and Western India. In: K. A. Nilakantha
Sastri (ed.), A Comprehensive History of India, Vol. 11, pp. 263-92. Bornbay.
BANERJI,A. 1951. Kushanas in Eastern India. JNSI, 13, pp. 107-9.
BANERJI,R. D . 1908. The Scythian Period of Indian History. Indian Antiquary, Vol. 37,
pp. 25-75.
BASHAM,A. L. 1953. A N e w Study of the Saka-Kusina Period. BSOAS, 15, pp. 80-97.
-(ed.). 1968. Papers on the Date of Kaniska Submitted to the Conference on the Date
of Kaniska, London, 20-22 April 1960. Leiden, E. J. Brill.
BIVAR,A. D. H. 1963. The Kaniska Dating from Surkh Kotal. BSOAS, 26, pp. 488-502.
CHAVANNES,E. 1907. Le pays d'occident d'apres le Heou Han Chou. TP, 8, pp. 149-
234.
CUNNINGHAM, A. 1894. Coins of Medieval India from the Seventh Century down to the
Muhammadan Conquests. London.
DEYDIER,R. 1950. Contribution a l'etude de Part du Gandhara. Paris.
Epigraphia Indica, Vol. 8.
FLEET,J. F. 1892. Gupta Inscriptions. CII.
FUSSMAN,G . 1974. Documents Cpigraphiques kouchans. BEFEO, 61, pp. 1-77.
GHIRSHMAN,R. 1946. BPgram, Recherches archPologiques et historiques sur les Kou-
chans. Paris/Cairo. (MDAFA, 12; MIFA O C , 79.)
-. 1957. Le problkme de la chronologie des Kouchans. C H M , 3, pp. 689-722.
-. 1978. Iran. London, Pelican Books.
GOBL,R. 1960. Roman Patterns for the Kushana Coins. JNSI, 22, pp. 75-96.
-. 1968. Numismatic Evidence Relating t o the Date of Kaniska. In: A. L. Basham
(ed.), Papers on the Date of Kaniska, pp. 103-13. Leiden.
-. 1984. System und Chronologie der Miinzpragung des Kwiinreiches. Vienna.
GUPTA, P. L. 1953. The Date of Kushana Currency in Eastern India. JNSI, 15,
pp. 185-92.
HARMATTA, J. 1965. Minor Bactrian Inscriptions. Acta Ant. Hung., 13, pp. 149-205.
India antiqua (Vogel volume). 1947. Leiden.
 Bibliography and References

Indian Antiquary. Bombay.

Indian Culture.
Indian Museum Bulletin.
JACOBI,H . 1880. Das Kilakicirya-Kathinakam. Z D M G , 34, pp. 247-318.
JANEKT, K. L. (ed.). 1961. Mathuri Inscriptions (Unpublished papers of Heinrich
Luders). Gottingen.
Journal of Indian History.
KIELHORN,F. 1905106. Junagadh Rock-Inscription of Rudradaman. El, 8, pp. 36-49.
KONOW, S. 1929. K h a r o s h ~ h Inscriptions
i with the Exception of 'Those of ASoka. C l l ,
Vol. 2, Part 1.
LEVI,S. 1896. Notes sur les Indo-Scythes: I. Les contes. J A , 8, pp. 444-84. (9th series.)
-. 1897. Notes sur les Indo-Scythes: 11. Les textes hstoriques; 111. Saint Thomas,
Gondopharks et MazdCo. J A , 9, pp. 5-26, 2 7 4 2 . (9th series.)
-. 1927. La DyStinta-Pankti et son auteur. J A , 111, pp. 95-127.
LOHUIZEN-DELEEUW,J. E. VAN. 1949. The Scythian Period: A n Approach to the His-
tory, Art, Epigraphy and Paleography of North India from the 1st century KC.. to
the 3rd century A.D. Leiden.
LUDERS,H. 1926. Bruchstiicke der Kalpanamanditika des Kumiralara. Leipzig.
MACDOWALL,D. W. 1960. The Weight Standards of the Gold and Copper Coinages of
the Kushina Dynasty. JNSI, 22, pp. 63-74.
-. 1968a. Numismatic Evidence for the Date of Kaniska. In: A. L. Basham (ed.), Pap-
ers on the Date of Kaniska, pp. 134-54. Leiden.
-. 19686. Soter Megas, the King of Kings, the Kushina. JNSI, 30, pp. 2 8 4 8 .
-. 1973. T h e Azes Hoard from Shaikhan-Dheri: Fresh Evidence for the Context of
Jihonika. South Asian Archaeology, pp. 2 15-30. London.
MAENCHEN-HELFEN,0 . J. 1945. The Yueh-chih Problem Re-examined. JAOS, 65,
pp. 71-81.
MAJUMDAR,N. G. 1932. Three Kushan Coins from North Bengal. Journal of the Asiatic
Society of Bengal, Vol. 28 (new series), pp. 127-9.
-. 1934. Explorations in Sind. Delhi.
MAJUMDAR,R. C. 1968. T h e Date of Kaniska. In: A. L. Basham (ed.), Papers on the Date
of Kaniska, pp. 150-4. Leiden.
MARICQ, A. 1958a. La grande inscription de Kaniska et I'etto-tokharien, I'ancienne
langue d e la Bactriane. JA, 246, pp. 345440.
-. 1958b. Classica et Orientalia: 5, Res Gestas Divi Saponis. Syria, pp. 295-360.
-. 1968. T h e Date of Kaniska. In: A. L. Basham (ed.), Papers on the Date of Kaniska,
pp. 155-99. Leiden.
MAKSHALL, J. H. 1912/13. Excavations at Taxila. ASIAR, pp. 1-52.
-. 1932. Mohenjodaro and the Indus Civilization, Vol. I. London.
-. 1947. Greeks and Sakas in India. JRAS, pp. 3-32.
-. 1951. Taxila, A n Illustrated Account of Archaeological Excavations Carried O u t at
Taxila under the Orders of the Government of India between the Years 1913 and
1934, Vols. 1-111. Cambridge, Cambridge University Press.
MARSHALL, J. H.; FOUCHER,A. 1947. The Monuments of Sanchi, ~ 0 1 s 1-111.. Calcutta.
MCCKINDLE, J. W. 1901. Ancient India as Described in Classical Literature. Being a Col-
lection of Greek and Latin Texts Relating to India, ~xrractedfrom ~ e r o d o t u s ,
Strabo . . . and Other Works. Lolidon.
 Bibliography '2nd References

, K. 1966. Cambridge Ancient History, VoI. 11.

N A K A I NA.
-. 1968. The Date of Kaniska. In: A. L. Basham (ed.), Papers on the Date of Kaniska,
pp. 206-39. Leiden.
N e w Indian Antiquary.
PULLEYBLANK, E. G . 1962. The Consonantal System of O l d Chinese. A M , 9, pp. 58-144,
206-65.
-. 1968. Chinese Evidence for the Date of Kaniska. In: A. L. Basham (ed.), Papers on
the Date of KaniZka, pp. 247-58. Leiden.
PURI,B. N. 1939/40. An Examination of the Titles of the Imperial Kusanas. Proceedings
of the India History Congress, Vol. 111, pp. 433-41. Calcutta.
-. 1941. Some Dates of the Kushina Kharoshthi Records and Their Bearing on
Kanishka's Era. Indian Culture, Vol. 7, pp. 490 et seq.
-. 1965. India under the Kushanas. Bombay.
-. 1977. K u ~ d n aBibliography. Calcutta.
RAPSON,E. J. 1908. Catalogue of the Coins of the Andhra Dynasty, the Western Ksatra-
pas, the Traikutaka Dynasty, and the 'Bodhi' Dynasty. London.
-(ed.). 1922. Cambridge History of India. Vol. I : Ancient India. Cambridge, Cam-
bridge University Press.
RAYCHAUDHURI, H. 1953. Political History of Ancient India, 6th ed. Calcutta.
ROSENFIELD, J. M. 1967. The Dynastic Arts of the Kushans. Berkeley/Los Angeles, Uni-
versity of California Press.
SEWELL,R. 1904. Roman Coins Found in India. JRAS, pp. 591 et seq.
SHARMA,R. S. 1959. Aspects of Political Ideas and Institutions in Ancient India. Delhi.
SIRCAR,D. C . 1960. The Kushinas. In: R. C . Majumdar; A. D . Pusalkar (eds.), The His-
tory and Culture of the Indian People, Chap. 9. Bombay.
SMITH,V. A. 1924. The Early History of India, from 600 B.C. to the Muhammadan Con-
quest, Including the Invasion of Alexander the Great, 4th rev. ed. Oxford, Oxford
University Press.
SPECHT,E. 1897. Les Indo-Scythes et l'Cpoque du rkgne de Kanichka, d'aprks les sources
chinoises. JA, 10, pp. 152-93. (9th series.)
TARN,W. W. 1951. The Greeks in Bactria and India, 2nd ed. Cambridge, Cambridge
University Press.
THOMAS,F. W. 1935. Tibetan Literary Texts and Documents Concerning Chineseurkes-
tan. London.
TOLSTOV,S. P. 1968. Dated Documents from the Toprak-kala Palace and the Problem of
the 'Saka Era' and the 'Kaniska Era'. In: A. L. Basham (ed.), Papers on the Date of
Kani~ka,pp. 304-26. Leiden.
V A L L ~ E - P O U S S IL.N ,DE L.A 1930. L'Inde aux temps des Mauryas et des Barbares Grecs,
Scythes, Parthes et Yue-tchi. L'Histoire du Monde, Vol. VIA. Paris.
WATTEKS,T. 1904/05. O n Yuan Chwang's Travels in India, 629-645 A.D., 1-11. Trans-
lation in: T. W. Rhys Davids (ed.), Oriental Translation Fund, Vol. XIV (new
series), London.
ZEIMAL,E. V. 1974. Nachal'naya data Kanishki - 278 g. n.2. Central Asia in the Kushan
Period, Vol. 1, pp. 292-301. Moscow.
ZUKCHER,E. 1968. The Yiieh-chih and Kaniska in Chinese Sources. In: A. L. Basham
(ed.), Papers on the Date of Kaniska, pp. 346-90. Leiden.
 Bibliography and References

AKISHEV,K. A*; BAYPAKOV, K. M.; EKLAKOVICH, L. B. 1972. Drevniy Orrar Alma-

Ata.
ANDRIANOV,B. V. 1969. Drevnie orositelhie sistemi Priaralka y svyazi s isloriry vor-
niknoveniya i razvitiya oroshaemogo zemledeliya, M ~ ~ ~ ~
BARTOL'D, V. V. 1965. Sochineniya, Vols. 2 and 3. Moscow.
BERNSHTAM,A. N. 1952. Istoriko-arkheologicheskie ocherki T ~ e n t r ~TyanP-hanya l ' ~ ~ ~ ~
i Parniro-Alaya. M I A , 26.
BURYAKOV,Y. F. 1975. Istoricheskaya topografiya drevnikh gorodov Tashkrntskogo
oazisa. Tashkent. (Istoriko-arkhe~lo~icheskiy ocherk Chacha i Ilaka.)
BURYAKOV,Y. F.; FILANOVICH,M. I. 1972. Stanovlenic gorodskoy kul'tury i itapi ee
razvitiya na territorri Tashkenta. O N U , 3.
Dal'verzintepa. Kushanskiy gorod na yuge Uzbekistana. 1978. Tashkent.
D'YAKONOV, I. M. 1967. Problema sobstvennosti. 0 strukture obshchestva Blizhnego
Vostoka d o seredini I1 tis. d o n.i. V D I , 4, pp. 13-35.
-. 1968. Obshchina na drevnern Vostoke v rabotakh sovetskikh issledovatelcy. V D I ,
1.
D'YAKONOV, I. M.; D'YAKONOV,M. M.; LIVSHITS,V. A. 1951. Dokumenti iz drevney
Nisi. M Y u T A K E , 2.
D'YAKONOV, I. M.; LIVSHITS,V. A. 1960a. Dokumenti i t Nisi I v. do n.6. Predvaritel'nie
itogi raboti. Moscow.
-. 19606. Parfyanskoe tsarskoe khozyaystvo v Nise I veka d o n.6. V D I , 2, pp. 14-38.
-. 1966. Novie nakhodki dokumentov v Staroy Nise. Peredneazlatskiy sbornik,
Vol. 2. Moscow.
GAFUROV,B. G. 1972. Tadzhiki - Drevneyshaya, drevnyaya i srednevekovaya istoriya
(ed. by B. A. Litvinsky). Moscow.
GUDKOVA,A. V.; LIVSHITS,V. A. 1967. Novie Khorezrnskie nadpisi iz Nekropolya
Tok-kala i ~ r o b l e r n a'Khorezrniyskoy Ori'. VKF A N UzSSR, 1.
GULYAMOV, Y. G . 1957. Istoriya orosheni~aKhorezma s drevneyshikh vremen do nash-
i k h d m y . Tashkent.
-. 1974. Kushanskoe tsarstvo i drevnyaya irrigatsya Sredney Azii. TMKIATSAKe, 2.
H E N N I N G ,W . B. 1948. The Date of the Sogdian Ancient Letters. BSOAS, 12, 3 4 ,
pp. 601-15.
Istoriya Samarkanda. 1969. Vol. I . Tashkent.
Istoriya tadzhikskogo naroda. 1963. Vol. I . MOSCOW.
KABANOV,S. K. 1973. Pozdnie kushani v Nakhshebe. V D I , 3, pp. 159-71.
-. 1977. Nakhsheb na rubezhe drevnosti i srednevekov'ya (III-VII vv). Tashkent.
-. 1981. Kul'tura selikikh poseleniy y ~ z h n o g oSogda III-VI v v . (po materialam issle-
dovaniy v zone C h i m k ~ r ~ a n s k o gvodokhranilishcha).
o Tashkent.
KO),-krilgan-kala -pamyatnik kul'turi Drevnego Khorezma IV v . do n.6. 1967. Moscow.
LEVINA,L. M. 1971. Keramika nizhney i sredney Sirdar'i -J I risyacheletii. n.h. Moscow.
LITVINSKY, ~ v d r e v n e ~Sredney Azii (k ~ r o b l e m eivo-
B. A. 1966. S l o z l ~ n o s o s t a v n oluk
lyutsii luka na Vostoke). SA, 2, pp. 51-69.
- 1972. Drevnie kochevniki 'Krishi Mira '. MOSCOW.
MASSON,M. E. 1933. Monetnie nakliodki, zaregistriro\rannie v Sredncy ~ z i vi 1930 i
1931 gg MU, 5.
 Bibliography and References

MASSON,M . E. 1950. Proiskhozhdenie bezimennogo 'tsarya tsarei - velikogo spasitelya3,

TrSAGU, 11 (new series).
-. 1953. Akhangaran - Arkheologo-topograficheskiy ocherk. Tashkent.
- 1975. K voprosu o severnikh granitsakh gosudarstva (Velikiy Kushan),
TMKIA TSAKe.
MASSON, V. M. 1955. Denezhnoe khozyaystvo Sredney Azii p o numizmaticheskiIn
annim. VDI, 2.
-. 1956. Drevnebaktriyskie moneti, chekanennie p o tipu tetradrakhm Geliokla. EV,
11.
-. 1957a. K voprosu o chekane yuCzhiyskoy Baktrii. I O O N Tadzh. SSR, 14.
-. 1957b. Vostochno-parfyanskiy pravitel' Sanabar. Numizmaticheskiy sbornik, Part
2. TrGIM, 26.
-. 1966. Khorezm i Kushani (Nekotorie voprosi Khorezmiyskoy numizmatiki). EV,
17.
-. 1981. Kushanskaya i p o k h a v drevney istorii Uzbekistana (Voprosi periodizatsii i
kronologii). O N U , 4.
MUKHAMEDJANOV,A. R. 1972. Dargomskaya plotina i krepost' v Ravatkhodzhe.
IMKU, 9.
-. 1975. K istorii irrigatsii v Kushanskuyu i p o k h u . Tsentral'naya Aziya v Kushan-
skuyu ipokhu. TMKIAKTSAKe, 2.
-. 1978. Istoriya orosheniya Bukharskogo oazisa (S drevneyshikh vremen do nachala
X X v). Tashkent.
NASAFI. n.d. 'Kandiya' Manuscript. Institut Vostokovedeniya Akademii N a u k Uzbeks-
koy SSR. Inventory N o . 5405.
Ocherki istorii zemledeliya v agrarnikh otnosheniy v Turkmenistane (s drevneyshikh vre-
men do prisoedineniya k Rossii). 1971. Ashkhabad.
PIGULEVSKAYA, N. V. 1951. Vizantiya na putyakh v Indiyu. Moscow/Leningrad.
RTVELADZE,E. V.; PIDAEV,S. R. 1981. Katalog drevnikh monet yuzhnogo Uzbekistana.
Tashkent.
SMIRNOVA,0. I. 1970. Ocherki iz istorii Sogda. Moscow.
STAVISKIY, B. Y. 1961. Osnovnie itapi osvoeniya zemledl'cheskim naseleniem gornikh
rayonov verkhnego Zerafshana (Kukhistana). ME, 1.
-. 1964/82. Kara-tepe, I-IV. Moscow.
TERENOZHKIN,A. I. 1974. Voprosi istoriko-arkhe~lo~icheskoy yeriodizatsii drevnego
Samarkanda. V D I , 4.
TOLSTOV, S. P. 1957. Biruni i ego 'Pamyatniki minuvshikh ~ o k o l e n i y ' .Aburayhan
Biruni (973-1048). Izbrannie proizvedeniya, Vol. I . Moscow.
-. 1962. Po drevnim del'tam Oksa i Yaksarta. MOSCOW.
-. 1969. O b ispol'zovanii dannikh istoricheskikh nauk dlya ~ r a k t i k inarodnogoho-
zyaystsva. Z D O .
VAYNBERG,B. I. 1977. Moneti drevnego Khorezma. Moscow.
ZEIMAL,E. V. 1965. Kushanskoe tsarstvo p o numizmaticheskim dannim. ~ e n i n ~ r a d .
(Dissertation.)
-. 1967. Moneti Velikikh Kushan v Gosudarstvennom Ermitazhe. TrGe, 9.
-. 1978. Politicheskaya istoriya drevney Transoksiani p o numizmaticheski~ndanninl.
Kul'tura Vostoka. Drevnee i ranee srednevekov'e. Moscow.
ZLIMAI-,T . I. 1971. Drevnie i Srednevekovie kanali Vakhchskoy dolini, SNV, 10.
 Bibliography and References

ADHYA,G . L. 1966. Early Indian Economics. Studies in the Economic o/Northeyn

and Western India c. 200 B.C. - 300 A.D. Bombay/New York.
ANON. 1981. Gorodishche Toprak-kala (raskopki 1965-1975 gg.) Trudi KhAEE, 12.
AUBOYER,J . 1965. Daily Life in Ancient India from Approximately 200 n.c. to 700 n.n.
London.
AUBOYER,J.; ENAULT,J. F. 1955. La vie publique et privee dans l'lnde ancienne. Vol. 1:
L'Architecture civile et religieuse. Paris.
BONGARD-LEVIN,G . M. 1973. Indiya Cpokhi Maur'ev. Moscow, Nauka.
CHAKRABORTI,0. H. 1966. Trade and Commerce of Ancient India (c. 200 1j.c.450
A. D.). Calcutta.
DANI,A. H. 1955156. Shaikhan Dheri Excavation. AP, 2, pp. 17-214.
DANI,A. H.; KHAN, F. 1974. Kushan Civilization in Pakistan. Tsentr. Aziya v kushan.
Cpokhu, Vol. 1, pp. 95-106. Moscow.
D'YAKONOV, I. M.; LIVSHITS,V. A. 1968. Iz materialov parfyanskoy kantselyarii 'Staroy
Nisi'. Issledovaniya po istorii kul'turi narodov Vostoka. Sbornik v chest' I. A.
Orbeli. MoscowILeningrad.
-. 1977. Corpus Inscriptionurn Iranicarum. Part 11: Inscriptions of the Seleucid and
Parthian Periods of Eastern Iran and Central Asia. Parthian Economic Documents
from Nisa. Vol. 11: Parthian. London.
FACCENNA,D. 1964. A Guide to the Excavations in Swat (Pakistan - 1956-1962). Rome.
-. 1981. Butkara I (Swat, Pakistan - 1956-1962). IsMe0.R.M.' 3, 1-5.
FICK, R. 1920. The Social Organizations in North- East India in Buddha's Time. Calcutta.
FILLIOZAT,J. 1959. Gopura 'porte de ville'. JA, 247, 2, pp. 251-5.
FRANCFORT, H.-P. 1979. Les fortifications en Asie Centrale de I'Age du Bronze a
l'kpoque kouchane. Paris.
FRISK,H. 1927. Le Periple de la Mer Erythrke suivi d'une etude sur sa tradition et sa lan-
gue. Giiteborg. (GGteborg Hiigskolas ~ r s s k r i f t 30.)
,
FRUMKIN,G . 1970. Archaeology in Soviet Central Asia. LeidenICologne, E. J. Brill.
GEIGER,W . 1960. Culture of Ceylon in Mediaeval Times. Wiesbaden.
GHIRSHMAN,R. 1946. Begram, Recherches archeologiques et historiques sur les Kou-
chans. ParisICairo. ( M D A F A , 12; MIFA O C , 79.)
GHOSH, A. 1948. Taxila (Sirkap), 1944-1945. Ancient India, N o . 4, pp. 41-84. N e w
Del hi.
-. 1973. The City in Early Historical Indta. Simla.
- 1975. T h e Kushan Levels at Some Excavated Sites in North India. Tsentr. Aziya v
kushan. Cpokhu, Vol. 2, pp. 108-12. Moscow.
GHOSH, M. 1975. Glimpses of Sexual Life in Nanda-Maurya India. ~ a l c u t t a .
GULL.INI,G . 1962. Udegram. IsMeo.R.M., 1.
H A C K I NJ,. 1954. Nouvelles rechevches ar~heologi~ucs a Begram (ancienne Kripib), 1939-
1940; vencontre dc trois civilisatio)~~,Indc, Gricc, Chine. Paris.
H A C K I NJ.;
, CARL,J.; M E U N I J.~ ,n.d. Diverscs recberches archeologiques en ~fghanistan
(1933-1940). (MDAFA.)
HARMATTA, J. 1964. The Great Bactrian Inscription. Actn Ant. Hung., 13, 3 4 .
HORNEK, I. B. (trans.). 1964. Milinda's Q~estions.London. 2 vols. r ran slated from
Pali.)
 Bibliography and References

I S M E O ( I - 1 . ~ 1 . 1AKCI-IAEOLOGICAL
~~ MISSION).1982. Pakistan, Swat - 1956-1981.
Pcshawar. (Documentary cxhibition.)
JACORI,H. (trans.). 1968. Jaina Sutras, Vols. 1-2. N e w York. (Translated from Prakrit.)
KANE,P. V. 1941. History of Dharmaizstra, Part 2 , 2 n d ed. Bombay. (English translation
with critical and explanatory notes.)
KANGLE,R. P. 1972. The Kauyiliya Arthaiistra, Part 2, 2nd ed. Bombay. (English trans-
lation with critical and explanatory notes.)
KHVOSTOV, M. M. 1907. Istoriya vostochnoy torgovli greko-rimskogo Egipta.
Kazan.
KONOW, S. 1929. Kharoshrhi Inscriptions with the Exception of Those of ASoka, C I I ,
Vol. 11, N o . 1.
KRUGLIKOVA, I. T. 1979. La Bactriane ancienne - mate'riaux de l'expe'dition a r ~ h e ' ~ l ~ -
gique afghano-sovie'tique. Moscow, ~ d i t i o n sSciences.
-. 1982. Dil'berdzhin - kushanskiy gorod v Severnom Afganistane. Arkheologiya sta-
rogo i novogo sveta. Moscow.
LAL, B. B. 1949. S i ~ u ~ a l g a r1948:
h A n Early Historical Fort in Eastern India. Ancient
India, Vol. 5, pp. 62-105. N e w Delhi.
LITVINSKY, B. A. 1968. Outline History of Buddhism in Central Asia. Moscow.
-. 1973. Drevniy sredneaziatskiy gorod (Mestnie traditsii i inozemnie modeli). Drev-
niy Vostok. Goroda i torgovlya (111-I tis. do n.6.). Yerevan.
-. 1979. Kushanskiy gorod v Sredney Azii i Indii (paralleli). Narodi Azii i Afriki,
N o . 1.
-. 1981. Kushanskiy gorod. Indo-sredneaziatskie paralleli. Drevniy Vostok i mirovaya
kul'tura. Moscow.
LITVINSKY,B. A.; SEDOV, V. A. 1983. Tepay-shakh. KulJtura i svyazi kushanskoy
Baktrii. MOSCOW.
LITVINSKY,B. A.; ZEIMAL,T. I. 1971. Adzhina-tepa: Arkhitektura. Zhivopis'. Skul'ptura.
Moscow.
LIVSHITS,V. A. 1969. K otkritiyu baktriyskikh nadpisey na Kara-tepe. Buddi~skie
peshcheri Kara-tepe v Starom Termeze. Moscow.
-. 1984. Dokumenti. In: Y. A. Rapoport and E. E. Nerazik (eds.), Toprak-kala
Dvorets. Moscow.
MARSHALL,J. H. 1911. Archaeological Exploration in India - 1909-1910. JRAS,
pp. 127-58.
-. 1915. Excavations at Bhita, A R A S H I , 1911-1912. Calcutta.
-. 1951. Taxila, A n Illustrated Account of Archaeological Excavations Carried Out at
Taxila under the Orders of the Government of India between the Years 1913 and
1934, Vols. 1-111. Cambridge, Cambridge University Press.
-. 1960. A Guide to Taxila. Cambridge, Cambridge University Press.
MISRA,R. N . 1975. Ancient Artists and Art Activity. Simla.
PERIKHANYAN, A. G. 1973. Sasanidski~sudebnik 'Kniga tisyachi sudebnikh resheniy'.
Yerevan.
PIGGOTT,S. 1945. Some Ancient Cities of India. Oxford, Oxford University Press.
PUGACHENKOVA, G. A. 1976. K otkritiyu nadpisey kkharoshtkhi na zolotikh predme-
takh Dal'vcrzinskogo klada. V D I , 1, pp. 64-71.
PUGACHENKOVA, G. A.; RTVELADZE,E. V., et al. 1978. Dal'vcrzin-tcpe. ~ u s h a n s k i y
gorod na yuge Uzbekistana. Tashkent.
 Bibliography and References

PURI,B. N. 1965. India under the K ~ s h i ~ a Bombay.

s.
-. 1966. Cities of Ancient India. MeerutIDelhilCal~ut~a.
RAPOPORT,Y. A. 1981a. Gorodishche Toprak-kala. Moscow.
-. 19816. Nekotorie itogi izucheniya dvortsa na gorodishche Toprak-kala. Kul'tuya i
isskustvo Drevnego Khorezma. Moscow.
RAY,A. 1964. Villages, Towns and Secular Buildings in Ancient India, c. 150 s.c. - c. 350
A. D. Calcutta.
RHYS DAVIDS,C. F. 1901. Notes on Early Economic Conditions in Northern India.
JRAS.
RTVELADZE,E. V. 1978. 0 genezise kushanskikh poseleniy Severnoy Baktrii. V D I , 4,
pp. 108-16.
SCHLINGLOFF,D. 1970. Die altindische Stadt. Eine vergleichende U n t e r ~ u c h u nMainz.
~.
SCHMITTHENNER, W . 1979. Rome and India: Aspects of Universal History During the
Principate. Journal of Roman Studies, Vol. 69, pp. 90-106. London.
SCHOFF, W. H. (trans., ed.). 1974. The Periplus of the Erythrean Sea. Travel and Trade in
the Indian Ocean b y a Merchant of the First Century, 2nd ed. New Dclhi. (Trans-
lated from Greek.)
SHARMA,G. R. 1968. Khusina Architecture with Special Reference t o KauSimbi (India).
In: G . R. Sharma (ed.), Kusana Studies, pp. 1 4 1 . Allahabad.
THORLEY,J. 1969. The Development of Trade between the Roman Empire and the East
under Augustus. Greece and Rome (2nd series), Vol. 16, No. 12. Oxford.
TOLSTOV,S. P. 1948. Drevniy Khorezm. Moscow.
- 1962. Po drevnim del'tam Oksa i Yaksarta. Moscow.
UPAHDHYAYA, B. S. 1947. India in K ~ l i d ~ s Allahabad,
a. Kitabistan.
VOROB'EVA-DESYATOVSKAYA, M. I. 1974. Novie nadpisi pis'mom kkharoshtkhi iz Ter-
meza. V D I , 1, pp. 116-26.
-. 1976. Nadpisi pis'mom kkharoshtkhi na zolotikh predmetakh iz Dal'verzin-tepe.
V D I , 1, pp. 72-9.
WARMINGTON,E. H. 1974. The Commerce between the Roman Empire and India,
2nd ed. London, Curzon Press.
WHEELER,R. E. M. 1962. Charsada, A Metropolis of the North-West Frontier. Being a
Report on the Excavations of 1958. London, British Academy.
YAZDANI,G. 1946. Ajanta, Vol. IV. LondonINew Y o r k l ~ o m b a ~ .

BACHHOFER,L. 1937. Piiicika und H i r i t i - Pharo und Ardoxsho. Ostasiatiscl~eZeit-

schrift, Vol. 13, pp. 6-15.
BIVAR, A. D. H. 1979. Mithraic Images of Bactria: Are They Related to Roman
Mithraism? Mysteria Mithrae, pp. 741-5 1. Leiden/Rome.
BOYCE, M. 1975. O n Mithra, Lord of Fire. Acta Iranica, 4, pp. 69-76.
C U N N I N G H A MA. , 1971. Coins of the Indo-Scythians, Sakas and Kushnns. (Reprint.)
N e w Delhi.
DAVIDOVICH, E. A. 1980. The First Hoard of Tetradracllrnas of the Kusina 'Hcraios'.
Acta Ant. Hung., 28, pp. 147-77.
D U C I ~ E S N I : - G U I I ~ L J. ~ I N ,Ln religion de /'Iran nncien, pp. 2 3 8 4 3 . Paris.
E ~1962.
FRAUWAI,I.NI:R, E. 1952. Die Buddliistische~lKonzile. Z D M G , 102, pp. 240-61.
 Bibliography and References

GHIRSI] M A N ,R. 1946. Begram, Recherches archeologiques et historiques sur les K ~ ~ -

chans. ParisICairo. (MDAFA, 12; MIFA O C , 79.)
GNC)LI,G . 1963. La religiosita Kuiana. Naples.
-. 1980. Zoroaster's Time and Homeland. Naples.
GOBI-, R. 1983. Der veruntreute Himmel: Die Gotterwelt des Kuiinreiches. Litterae
Numismaticae Vindobonensis, Vol. 2, pp. 79-97.
GRENET,F. 1984. Notes sur le panthion iranien des Kouchans. Studia Iranica, vol. 13,
pp. 253-62.
-. 1986. L'art zoroastrien en Sogdiane. Mesopotamia, Vol. 21, pp. 97-130. Turin.
-. 1987. L'Athina d e Dil'berdiin. Cultes et monuments religieux duns 1'Asie Centrale
preislamique, Vol. 2, pp. 41-5. Paris, ~ d i t i o n sdu CNRS.
HARMATTA,J. 1960. Cusanica. Acta Orient. Hung., 11, pp. 191-220.
-. 1964. Sino-Indica. Acta Ant. Hung., 12, pp. 3-1 1.
-. 1965. Minor Bactrian Inscriptions. Acta Ant. Hung., 13, pp. 149-205.
-. 1969. Late Bactrian Inscriptions. Acta Ant. Hung., 17, pp. 297-432.
-. 1975. Der Ostliche Hintergrund der Partherkriege unter Marcus Aurelius. Actes de
la XIIe Confhence Internationale d ' ~ t u d e s Classiques 'Eirene' Cluj-Napoca,
2-7 octobre 1972. Bucharest/Amsterdam.
-. 1986. T h e Bactrian Inscription of Ayrtam. Studia Grammatica Iranica. pp. 131-46.
Munich. (Festschrift fiir Helmut Humbach.)
HUMBACH, H. 1975a. Vayu, Siva u n d der Spiritus Viveris im ost-iranischen Synkretis-
mus. Acta Iranica, 4, pp. 3 9 7 4 0 8 .
-. 1975b. Mithra in the Kusina Period. Mithraic Studies, Vol. I, pp. 135-41,353. Man-
chester.
-. 1978. Mithra in India and the Hinduized Magi. Acta Iranica, Vol. 17, pp. 229-53.
-. 1988. Herrscher, G o t t und Gottessohn in Iran und in angrenzenden Landern.
Menschwerdung Gottes - Vergottlichung von Menschen, pp. 91-114. Freiburg/
Gottingen.
KELLENS,J. 1980. Yima, magicien entre les dieux et les hommes. Orientalia J. Duchesne-
Guillemin emerito oblata. Acta Iranica, 24, pp. 267-81.
-. 1988. Yima et la mort. Languages and Cultures. Studies in Honor of Edgar Polome,
pp. 329-34. BerlinINew YorkIAmsterdam.
K o ~ o w S. , 1929. Kharosl~yhiInscriptions with the Exception of Those of ASoka. CII,
2.1.
L ~ c o q P., 1987. Le mot farnah- et les Scythes. C R A I , pp. 671-82.
LITVINSKY, B. A. 1970. Outline of Buddhism in Central Asia. Kushan Studies in USSR.
Calcutta.
LITVINSKY, B. A.; SEDOV,V. A. 1984. Kul'ti i rituali Kushanskoy Baktrii pogrebal'niy
obryad. Moscow.
LITVINSKY,B. A.; PICHIKYAN,I. R. 1980. Monuments of Art from the Sanctuary of
O x u s (Northern Bactria). Acta Ant. Hung., 28, pp. 25-83.
LUDERS,H. 1961. Mathurd Inscriptions. Gottingen.
MACDOWALL, D. W. 1978. Mithra's Planetary Setting in the Coinage of the Great Kush-
ans. Acta Iranica, Vo1. 17, pp. 305-15.
N A K A I NA.
, K. 1957. The Indo-Greeks. Oxford, Oxford University Press.
PUGACHENKOVA, G. A. 1966. Khalchayan. K probleme khudozhestvennoy kul'tuyi
Severnoy Baktrii. Tashkent.
 Bibliography and References

PURI,B. N . 1965. India under the Kushinas, pp. 136-56. Bombay.

-. 1986. Buddhism in Central Asia. Delhi.
RosENFIELD, J . M. 1967. The Dynastic Arts of the Kushans. Berkclcy/Los Angeles, Uni-
versity of California Press.
SCHLUMBERGER,D. 1964. Le temple de Surkh Kotal en Bactriane (IV). j A , 252,
pp. 303-26.
STAVISKIY, B. Y. 1980. Kara Tepe in Old Termez. A Buddhist Religious Ccnter of the
Kushan Period on the Bank of the Oxus. Acta Ant. Hung., 28, pp. 95-135.
STEBLIN-KAMENSKIY, I. M. 1981. Pamirskie yaziki o mitologii drevnikh irantsev. Ethnic
Problems of the History of Central Asia in the Early Period (Second Millennium
B. c.),pp. 2 3 8 4 1 . Moscow.
STEIN,A. 1887. Zoroastrian Deities on Indo-Scythian Coins. Oriental and Babyloniun
Record, August, pp. 1-12.
STEIN.M. A. (trans., ed.). 1961. Kalhanai RZjatanginl. A Chronicle of the Kings of Kas-
mir. Delhi/Patna/Varanasi.

AGRAWAL.A, V. S. 1965. lndiun Art. Varanasi.

AL'BAUM, L. I. 1975. Zhivopis' Afrasiaba. Tashkent.
-. 1982. 0 tolkovanii karatepinskikh kompleksov (v svete raskopok Fayaz-tepe).
Buddiyskie pamyatniki Kara-tepe v Starom Termeze, pp. 56-63. Moscow.
ALLCHIN,F. R. 1957. The Cultural Sequence of Bactria. Antiquity, Vol. 31, pp. 1 3 1 4 1 .
Cambridge, Heffer & Sons Ltd.
ANON. 1953. Art and Archaeology at the Cleveland Museum of Art. Archaeology, Vol. 6,
pp. 195-202. Boston, Archaeological Institute of America.
BANERJEA,J. N. 1956. The Development of Hindu Iconography, 2nd ed. Calcutta.
BERNARD,P. 1967. Ai' Khanoum on the Oxus: A Hellenistic City in Central Asia. Pro-
ceedings of the British Academy, Vol. 53, pp. 71-95, Plates I-XX.
BERNARD,P., et al. 1973. Campagnes 1965, 1966, 1967, 1968. Fouilles dJAi' Khanoum,
Vol. I. Paris. ( M D A F A , 21.)
CHANDA,R. 1936. Medieval Indian Sculpture. London.
COOMARASWAMY, A. 1926. The Indian Origin of the Buddha Image. JAOS, 46,
pp. 165-70.
COOMARASWAMY, A. K. 1965. History of Indian and lndonesinn Art. Magnolia, Mass.,
Peter Smith.
DAGENS, B. 1960. Fragments de sculpture inidits. Paris. (MDAFA, 19.)
DANI,A. H. 1965/66. Shaikhan Dheri Excavations, 1963-64 Seasons: In Search of the
Second City of Pushkalavati. AP, 2, pp. 17-214.
-. 1969. Kushana Civilization in Pakistan. Journal of the Asiutic Society o f Pakistan,
Vol. 14, N o . 1, pp. 1-20.
D A R ,S. R. 1973. Ta ellCnistika stoikheia stin arkhitekroniki ton Taxilon. Thessalonika.
- 1976. Taxila and Hellrnisnz: Architectural Evidence. Lahore.
-. 1977. T h e Question of Roman Influence in the Gandhara Art: Numismatic Evi-
dence. Rivista italiana di numismarica, Vol. 79, pp. 61-89. Rome.
-. 1979. Toilet Trays from G a n d h a r ~and Beginning of Hellet~isrnin Pakistan. JCA, 2,
pp. 141-84.
 Bibliography and References

DAR,S. K. 1980. Toilet Trays from Gandhara. Megas Alexandros. 2300 khronia, pp. 91-
106. Thessalonika.
DOBBINS,G . W. 1971. The Stupa a n d Vihara of K a n i ~ k aI. Calcutta. (Asiatic Society
Monographs, 18.)
DOLGORUKOV,V. S. 1984. Oboronitel'nie sooruzheniya Dil'berdzhina. Drevnyaya
Baktriya, Vol. 3, pp. 58-92. Leningrad.
DUTT, S. 1962. Buddhist Monks a n d Monasteries of India. London.
GARDIN,J.-C. 1957. Ceramiques de Bactres. Paris. (MDAFA, 15.)
GHIRSHMAN, R. 1946. Bigram, Recherches arche'ologiques et historiques sur les K ~ ~ -
chans. ParisICairo. (MDAFA, 12; MIFAOC, 79.)
HALLADE, M. 1968. The Gandhara Style a n d the Evolution of Buddhist ~ r t .
London.
HARTEL,H. 1976. Some Results of the Excavations of Sonkh. German Scholars of India.
Bombay.
INGHOLT,H. 1957. Gandharan Art in Pakistan. N e w York.
JOSHI,N . P. 1972. Catalogue of the Brahmanical Sculptures in the State Museum, Luck-
now. Lucknow.
KHAN, M. I. 1966. Mathura Objects in Taxila Museum. Journal of the Asiatic Societ3, of
Pakistan, Vol. 2, pp. 44-9.
KRUGLIKOVA, I. T. 1974. Dil'berdzhin (Raskopki 1970-1972 gg.), Part 1. Moscow.
-. 1982. Dil'berdzhin - kushanskiy gorod v Severnom Afganistane. Arkheologiya Sta-
rogo i Novogo sveta. Moscow.
KRUGLIKOVA, I. T.; PUGACHENKOVA, G. A. 1977. Dil'berdzhin (Raskopki 1970-1973
gg.), Vol. 2. Moscow.
LE BERKE,M.; SCHLUMBERGER, D. 1964. Observations sur les remparts de Bactres. Paris.
(MDAFA, 19.)
LITVINSKY, B. A.; PICHIKYAN, I. R. 1981. Dicouvertes dans u n sanctuaire du dieu Oxus
de la Bactriane septentrionale. RA, 2, pp. 195-216.
LOHUIZEN-DELEEUW,J. E. VAN.1949. The 'Scythian' Period: An Approach to the His-
tory, Art, Epigraphy a n d Paleography of North India from the 1st century B.C. to
the 3rd century A. D. Leiden.
MACDOWALL,D. W.; TADL)EI,M. 1978. Archaeology of Afghanistan. London.
MARSHALL, J. H. 1902103. Buddhist Gold Jewellery. ASIAR, pp. 185-94.
-. 1951. Taxila, An Illustrated Account of Archaeological Excavations Carried Out at
Taxila Under the Orders of the Government of India Between the Years 1913 and
1934, Vols. 1-111. Cambridge, Cambridge University Press.
-. 1960. The Buddhist Art of Gandhara. Memoirs of the Department of Archaeology
in Pakistan, Vol. I. Cambridge, Cambridge University Press.
MASSON,M. E. 1976. I z vospominaniy sredneaziatskogo arkheologa. Tashkent.
MESHKERIS, V. A. 1969. Terrakoti iz Kara-tepe. Buddiyskie peshcheri Kara-tepe v Starom
Termeze. Moscow.
MIAN,A. K. M. 1974. P210plastik$ ton Taxilon kai ellznistika tEs stoikheia. Thessalonika,
Thessalonika University. (Ph.D. dissertation.)
..
MUKHTAROV, A. 1968. Novie nakhodki kalnennikh kapitclcy kushanskogo vremenl lz
Shakhrinau (Yuzhniy Tadzhikistan). IOONTadzh.SSR.
PUGACHENKOVA, G . A. 1965. O b odnoy gruppe lepnikh terrakotovikh statuetok Tokh-
aristana. Novoe v sovetskoy arkheologii. Moscow.
 Bibliography and Rcferencer

-. 1966. Khalchayan. Kprobleme k h ~ d o r h e s t v e n n okul'tuyi

~ .$evcynoy Bakryii. Tash-
kent.
-. 1967. Dve stupa na Yuge Uzbekistana. SA, 3, pp. 257-64.
-. 1971. Skul'ptura Khalchayana. Moscow.
1973. N o v k dannie 0 k h ~ d o z h e s t v e n n okul'ture
~ Baktrii. lZ istoyii antichnoy
kul'turi Uzbekistana, pp. 78-128. Tashkent.
-a 1976. K poznaniyu antichnoy i r a n n e s r e d n e v e k o ~ o ~arkhitekturi Severnogo
~ f g a n i s t a n a Drevnyaya
. Baktriya, Vol. 1, pp. 125-62. Moscow.
-. 1977. Gerakl v Baktrii. V D I , 2.
-. 1978. Les trisors de Dalverzin-tipi. Leningrad.
-. 1979. Isskustvo Baktrii Ppokhi kushan. Moscow.
-. 1984. ~ a s k o ~ ~k ui z h n i k hgorodskikh vorot Dil'berdzhina. Drevnyaya BaktriYa,
Vol. 3, pp. 93-1 11. Moscow.
PUGACHENKOVA,G . A.; RTVELADZE,E. V., et al. 1978. Dal'verzin-tepe, kushanskiy
gorod na yuge Uzbekistana. Tashkent.
PURI, B. N. 1965. India under the KushZnas. Bombay.
ROSENFIELD,J . M. 1967. The Dynastic Arts of the Kushans. Berkeley/Los Angeles, Uni-
versity of California Press.
ROWLAND,B. 1970. The Art and Architecture of India: Hindu, Buddhist, lain, 3rd ed.
Harmondsworth, Penguin Books.
-. 1971. Art in Afghanistan: Objects from the Kabul Museum. London.
-. 1976. Ancient Art from Afghanistan: Treasures from the Kabul Museum. New
York, A r n o Press.
RTVELADZE,E. V. 1982. Bactrian Fortifications: Genesis and Evolution in Their Social
and Cultural Aspect. Symposium: La fortification et sa place dans I'histoire poli-
tique, sociale et culturelle du monde Grec. Valbonne.
SARASWATI, S. K. 1975. A Survey of Indian Sculptures. Delhi.
SARIANIDI,V. I. 1985. Bactrian Gold from the Excavations of the Tilljla-tepe Necropolis
in Northern Afghanistan. Leningrad.
SCHLUMBERGER, D . 1955. Surkh Kotal in Bactria. Archaeology, Vol. 8, No. 2, pp. 82-7.
Boston, Archaeological Institute of America.
- 1961. T h e Excavations at Surkh Kotal and the Problem of Hellenism in Bactria and
India. Proceedings of the British Academy, Vol. 48, pp. 77-95.
SCHLUMBERGER, D.; LE BERRE,M.; FUSSMAN,G. 1983. Surkh Kotal en Bactriane,
Vol. I. Paris. ( M D A F A , 25.)
SHARIF,M. 1978. Ta ellinistika stoikheia sta kosmimata t i s Gandaras. Thessalonika,
Thessalonika University. (Ph.D. dissertation.)
SHARMA,R. C. 1984. Buddhist Art of Mathura. Delhi.
SHISHKIN, V. A. K. 1941. K istoricheskoy t o ~ o g r a f i Starogo
i Termeza. Termezska~aark-
hcologicheskaya kompleksnaya Pkspeditsiya 1936 g., pp. 123 et seq. Tashkent.
SMI'TH,V. A. 1889. Graeco-Roman Influence in India. journal of the Asiatic Soclety of
Bcngal, Vol. 58. Calcutta.
-. 1901. The Jainn Stupa and other Antiquities of ~ a t h u r a Allahabad.
.
SNELI.C;ROVE,D . L. 1978. TIw Image of the Buddl7a. London, Serindia Publications.
SPOONER,D. B. 1908109. Excavations at Shah-ji-ki Dheri. A S I A R , pp. 38-59.
STAVISKII',B. Y. (ed.). 1964-82. Knra-tepc I-IV. Moscow.
-. 1972. Buddiyski' kul'roviy tsentr Kara-tepc v starom Termcze. MOSCOW.
 Bibliography and References

STAVISKIY,B. Y. 1981. Novaya kamennaya kapitel' iz Kara-tepe v Starom Termezc.

Drcvniy Vostok i mirovaya kul'tura. Moscow.
TREVE,R,K. V. 1940. Pamyatniki greko-baktriyskogo iskusstva. Leningrad.
VOGEL,J . P. 1910. Catalogue of the Archaeological Museum at Mathura. Allahabad.
-. 191 1/12. Explorations at Mathura. Annual Report, Archaeological Survey of India,
pp. 120-33. N e w Delhi.
-. 1930. La Sculpture d e Mathura. Ars Asiatica, Vol. 15. Paris/Brussels.
WATTERS, T . 1904/05. O n Yuan Chwang's Travels in India, 629-641 A.D., Vols.1-11.
London, Royal Asiatic Society.
WHEELER,SIR R. E. M. 1954. Rome Beyond the Imperial Frontiers. H a r m o n d ~ w o r ~ h ,
Penguin Books.
-. 1962. Charsada, A Metropolis of the North-West Frontier. Being a Report on the
Excavations of 19j8. London, British Academy.
WHITEHEAD,R. B. 1914. Catalogue of Coins in the Panjab Museum, Lahore. Vol. I:
Indo-Greek Coins. Oxford, Oxford University Press.
WILSON, H. H . 1841. Ariana Antiqua. London.

ALRAM.M. 1986. Nomina propria Iranica in nummis. In: M. Mayrhofer and R. Schrnitt
(eds.), Iranisches Namenbuch, Vol. 4. Vienna.
BAILEY,H. W. 1958. Languages of the Saka. Handbuch der Orientalistik, Section I, Vol.
4: Iranstik. I. A bscsch Linguistik, pp. 131-54. Leiden/Cologne, E. J. Brill.
-. 1979. Dictionary of Khotan Saka. Cambridge, Cambridge University Press.
BENVENISTE,E.; DUPONT-SOMMER,A. 1958. U n e bilingue grico-aramienne d'ASoka.
JA, 246, pp. 45-6 (on the Kambojas).
- 1966. U n e inscription indo-aramienne d'ASoka provenant de Kandahar
(Afghanistan). JA, 254, pp. 437-65.
CALLIERI,P. 1984. A Potsherd with a Greek Inscription from Bir-kot (Swat). JCA, 7,
pp. 49-53.
DAVARY,G . D . 1981. Epigraphische Forschungen in Afghanistan. Studia Iranica, Vol.
10, pp. 55-6 (Inscription Laghman 11.)
DAVARY,G. D.; HUMBACH, H. 1974. Eine weitere aramao-iranische Inschrift der
Periode des Aioka aus Afghanistan. Mainz.
DELAUNAY, J. A. 1974. L'aramCen d ' e m ~ i r eet les d i b u t s de l'icriture en Asie Centrale.
Acta Iranica, Vo1. 2, pp. 219-36.
DUPONT-SOMMEK, A. 1970. U n e nouvelle inscription d'ASoka, trouvie dans la vallie du
Laghman (Afghanistan). C R A I , pp. 158-73.
FUSSMAN,G. 1974. Documents ipigraphiques kouchans. BEFEO, 61, pp. 1-77.
HARMATTA,J. 1966. Zu den griechischen Inschriften des ASoka. Acta Ant. Hung., 14,
pp. 77-85.
-. 1969. Late Bactrian Inscriptions. Acta Ant. Hung., 17, pp. 297-432. (See pp. 349-52
on Avestan as the language of ancient Bactria.)
HENNINC;,W . B. 1949. T h e Aramaic Inscription of ASoka Found in Lampika. BSOAS,
13, pp. 80-8.
HUMBACH,H . 1969. Die aramaische Inschrift v o n Taxila. Mainz.
 Bibliography and Refcrenccs

- 1973. Die aramhische Aioka-Inschrift von Laghman-Pluss. Indologcn-Tarrgung

1971, pp. 161-9. Wiesbaden.
-. 1974. Aramaeo-Iranian and Pahlavi. Acta Iranica, 2, pp. 2 3 7 4 3 .
-. 1976. T h e Aramaic ASoka Inscription from Taxila. German Scholars of India. Con-
tributions to Indian Studies, Vol. 2, pp. 118-30. Bombay.
-. 1978. T h e Aramaic Aioka Inscription from Taxila. JCA, 1, pp. 87-98.
K o ~ o w S., 1929. K h a r o s h ~ h Inscriptions
i with the Exception of Those of Aioka. C I I ,
2, 1. Calcutta.
KUIPER,F. B. J. 1976. Old East Iranian Dialects. IIJ, 18, pp. 241-53.
LIVSHITS,V. A.; SHIFMAN,I. S. 1977. K tolkovaniyu novikh aramcyskikh nadpisey
Ashoki. V D I , 2, pp. 7-24.
LUDEKS, H. 1940. Die Sakas und die 'nordarische' Sprache. Philologica Indica,
pp. 236-55. Gottingen.
MUKHERJEE,B. N. 1984. The Taxila Edict of Aioka. JCA, 7, pp. 17-27.
PUGLIESE CARRATELLI, G . 1965. Greek Inscriptions of the Middle East. EW, 16 (new
series), pp. 35-6 (on the Greek inscription of Udegram).
RAPIN, C. 1983. Les inscriptions Cconomiques de la tresorerie hellenistique d'Ai Kha-
noum (Afghanistan). B C H , 107, 1, pp. 315-81. (The Aramaic ostracon is repro-
duced in Fig. 34, p. 348.)
ROBERT, L. 1973. In: Fouilles d'Ai' Khanoum, Vol. I, pp. 211-37. Paris. ( M D A F A , 21.)
SCHLUMBERGER, D . 1964. Une nouvelle inscription grecque d ' A ~ o k a .C R A I , pp. 1 2 6 4 0 .
SCHLUMBERGER, D.; ROBERT, L.; DUPONT-SOMMER,A.; BENVENISTE,E. 1958. Une
bilingue grtco-aramienne d'Adoka. JA, 246, 1, pp. 1 4 8 .
TROUSDALE,W . 1968. A n Achaemenian Stone Weight from Afghanistan. EW, 18,
pp. 277-80.
Tucc~,G.; SCERRATO,U.; PUGLIESECARRATELLI, G.; LEV]DELLA VIDA,G . 1958. U n
editto bilingue Greco-Aramaico di Aioka. Rome.
T u c c ~ G.;
, SCERRATO,U.; PUGLIESECARRATELI-I, G.; GARBINI,G . 1964. A ~ i l i n ~ u a l
Graeco-Aramaic Edict b y Aioka. Rome.

AKISHEV,K. A. 1974. Kurgan Issik. V glub' vekov. Alma-Ata.

-. 1978. Kurgan Issik. Iskusstvo Sakov Kazakhstana. Moscow.
BAILEY,H . W . 1946. Gindhiri. BSOAS, 11, pp. 764-97.
-. 1955. Buddhist Sanskrit. JRAS, pp. 13-24.
BENVENISTE,E. 1961. Inscriptions de Bactriane. ]A, 249, pp. 113-52.
BERNARD,P. 1980. Campagne de fouilles 1978 i Ai Khanoum (Afghanistan). C R A I ,
pp. 29-30.
BH A'ITACHARYA, B. 1976. As'vagho~a. Santiniketan.
BOYER,A. M.; RAPSON,E. J.; SENART,E.; NOBLE,P. S. 1920-29. Kharo~thiInsc~iptions
Discovered b y Sir Aurel Stein in Chinese Turkestan. Oxford, Oxford University
Press.
BROUGH,J. 1962. The G a n d h a n Dharmapada. London.
BUHLER,G . 1896. Indische Paleographic. Strasbourg.
BURROW,T . 1937. The Langwage of the Kharo~!hiDocuments from Chinese Turkestan.
Cambridge, Cambridge University Press.
 Bibliography and References

C U K I E LK.
, 1954. Inscriptions de Surkh Kotal. JA, 242, pp. 189-205.
D A N I ,A. H. 1963. Indian Palaeography. Oxford, Oxford University Press.
-. 1983. Chilas. The City of Nanga Parvat (Dyamar). Islamabad.
-. 1985. The Sacred Rock of Hunza. JCA, 8, pp. 5-131.
DAVARY,G . D. 1981. Epigraphische Forschungen in Afghanistan. Studia Iranica,
Vol. 10, pp. 53-5. (Tiravharna inscription.)
-. 1982. Baktrisch. Ein Worterbuch auf Grund der Inschriften, Handschriften, Mcn-
zen und Siegelsteine. Heidelberg.
DAVARY,G. D.; HUMBACH,H . 1976. Die baktrische Inschrift I D N 1 von dash^-^
Nawur (Afghanistan). Mainz.
EDGERTON,F. (ed.) 1953. Buddhist Hybrid Sanskrit Reader. Northford, Conn.,
Books.
-. 1975. Buddhist Hybrid Sanskrit. Grammar a n d Dictionary. Mystic, Conn., Verry,
Lawrence. 2 vols.
FUSSMAN,G. 1969. Une inscription kharosthi i Hadda. BEFEO, 56, pp. 5-9.
-. 1970. Inscription kharosthi du Musie de Caboul. BEFEO, 57, pp. 43-55.
-. 1974. Documents Cpigraphiques kouchans (I). BEFEO, 61, pp. 1-66.
-. 1978. Inscriptions de Gilgit. BEFEO, 65, pp. 1-64.
-. 1980a. Documents ipigraphiques kouchans (11). BEFEO, 67, pp. 45-58.
-. 1980b. Nouvelles inscriptions Saka (I): ;re d'Eucratide, &red'Azks, itre Vikrama, Pre
de Kaniska. BEFEO, 67, pp. 1 4 3 .
-. 1984. Nouvelles inscriptions Saka (11). BEFEO, 73, pp. 3 1 4 6 .
FUSSMAN,G.; SCHLUMBERGER, D.; LE BERRE,M. 1983. Surkh Kotal en Bactriane,
Vol. I. Paris. (MDAFA, 25.)
GERSHEVITCH, I. 1966. The Well of Baghlan. AM, XI1 (new series), 1, pp. 90-109.
-. 1980. The Colophon of the Nokonzok Inscription. Acta Ant. Hung., 28,
pp. 179-84.
GREK, T. V. 1964. Indiyskie nadpisi na keramike iz Kara-tepe. Buddiyskiy pesherniy
monastir' v Starom Termeze, pp. 62-81. Moscow.
-. 1972. Novie indiyskie nadpisi iz raskopok Kara-tepe. Buddiyskiy kul'toviy tsentr
Kara-tepe v Starom Termeze, pp. 114-17. Moscow.
HARMATTA, J. 1960. Cusanica. Acta Orient. Hung, 11, pp. 191-220.
-. 1964a. The Great Bactrian Inscription. Acta Ant. Hung., 12, 3-4, pp. 373-
471.
-. 1964b. Sino-Indica. Acta Ant. Hung., 12, pp. 3-21.
-. 1965. Minor Bactrian Inscriptions. Acta Ant. Hung., 13, 1-2, pp. 149-205.
-. 1966. The Oldest Kharosthi Inscription in Inner Asia. A O H , 19, pp. 1-12.
-. 1967. The Oldest Brahrni Inscription in Innermost Asia. A O H , 20, pp. 1-
32.
-. 1969a. Late Bactrian Inscriptions. Acta Ant. Hung., 17, 3-4, pp. 297432.
-. 1969b. K interpretatsii nadpisey na keramike iz Kara-tepe. Buddi~skiepeshcheri
Kara-tepe v Starom Termeze. Moscow.
H E N N I N GW. , B. 1960. The Bactrian Inscription. BSOAS, 23, pp. 47-55.
HUMBACH,H . 1966-67. Baktrische Sprachdenkmaler, Vols. 1-11. Wiesbaden.
-. 1980. Die sogdischen Inschriftenfunde vom oberen Indus (Pakistan). Allgemeine
und vergleichende Archaologie-Beitrage, Vol. 2, pp. 2.01-28.
-. 1985. The Sogdian Inscriptions of Thor-Shatial. JCA, 8, pp. 51-7.
 Bibliography and References

KEITH,A. B. 1920. A History of Sanskrit Literature. London. pp. 5 5 4 4 (on Aivaghosa),

491-6 (on Buddhist Hybrid Sanskrit literature).
LAZARD,G.; GRENET,F.; D E LAMBERTERIE, C. 1984. Notes bactriennes. Srudia Iranica,
Vol. 13, pp. 199-232.
LIVSHITS,V. A. 1976. Nadpisi iz Dil'berdzhina. Drevnyaya Baktriya, p. 165, n. 14.
LIVSHITS,V. A.; KRUGLIKOVA, I. T. 1979. Fragmenti baktriyskoy monumental'noy nad-
pisi iz Dil'berdzhina. Drevnyaya Baktriya, Vol. 2, pp. 98-1 12.
LUDERS,H. 1909/10. A List of Brihmi Inscriptions from the Earliest Times to About
A.D. 400 with the Exception of Those of ASoka. Epigraphica Indica, Vol. 10,
Appendix, pp. 1-226.
-. 1961. M a t h u r i Inscriptions. Gottingen.
MARICQ, A. 1958. Inscriptions de Surkh Kotal (Baghlin). La grande inscription de
Kaniska et 1'CtCo-tokharien, l'ancienne langue de la Bactriane. JA, 246, pp. 345-
440.
PANDEY,R. B. 1957. Indian Palaeography, Vol. 2. Varanasi.
PUGACHENKOVA,G . A. 1966. Khalchayan. K probleme khudozhestvennoy kul'turi
Severnoy Baktrii. Tashkent.
PURI,B. N. 1965. India under the Kushinas, pp. 161-86. Bombay.
SHARMA,G . R.; N E G I ,J. S. 1968. The S a k a - ~ u ~ i n in
a sthe Central Gangi Valley. Kusina
Studies, pp. 43-73. Allahabad.
TURGUNOV,B. A.; LIVSHITS,V. A.; RTVELADZE, E. V. 1981. Otkritie baktriyskoy mon-
urnental'noy nadpisi v Ayrtame. O N U , 3, pp. 38-48.
VERTOGRADOVA,V. V. 1975. O b odnoy indiyskoy nadpisi na sosude iz Kara-tepe.
Novie nakhodki nu Kara-tepe v Starom Termeze, pp. 70-81. Moscow.
-. 1982a. Nakhodka nadpisi neizvestnim pis'mom na Kara-Tepe. In: B. Y. Staviskiy
(ed.), Buddiyskie Pamyatniki Kara-Tepe v Starom Termeze, pp. 160-7. Moscow.
-. 19826. Indiyskie nadpisi na kerarnike iz r a s k o ~ o k70-kh godov na Kara-tepe. Bud-
diyskie pamyatniki Kara-tepe v Starom Termeze, pp. 134-59. Moscow.
VOROB'EVA-DESYATOVSKAYA, M. I. 1974. Novie nadpisi pis'mom kkharoshtkhi iz Ter-
meza. VDI, 1, pp. 116-26.
-. 1976. Nadpisi pis'rnom kkharoshtkhi na zolotikh ~ r e d m e t a k hi t Dal'verzin-tepe.
VDI, 1, pp. 72-9.

ABDULLAEV,K. A.; BURYAKOV, Y. F.; BURYAKOVA, E. Y. 1977. Raskopki gorodishcha

Kanka v Tashkentskoy oblasti. AO, 1976, pp. 521-2.
ANDRIANOV,B. V. 1969. Drevnie orositel'nie sistemi Priaral'ya (V svyazi s istoriey voz-
niknoveniya i razvitiya oroshaemogo zemledeliya). Moscow.
ARISTOV,N. A. 1903. Etnicheskie otnosheniya na Pamire i v prilegayushchikh stranakh
p o drevnim, preimushchestvenno kitayskim istoricheskim izvestiyam. Russkiy
antropologicheskiy zhurnal, No. 1.
A R R I A N .1962. Pokhod Aleksandra. Moscow/Leningrad.
BARTOI_'D,V. V. 1927. Istoriya kul'turnoy zhizni Turkestana. Leningrad.
- 1964. Greko-Baktriyskoe gosudarstvo i ego rasprostranenie na severo-vostok.
Sochinenzja, Vol. 2. Moscow.
 Bibliography and References

BELYAEVA, T. V. 1978. K otkritiyu drevnego poseleniya na territorii Leninabada. lstoriya

i arkheologiya Sredney Azii, pp. 4 1-8. Ashkhabad.
-. 1980. O b Sllinisticheskikh Slementakh v kul'ture Khodzhenta. Vsesoyuzniy simpo-
zium po problemam t?llinisticheskoy kul'turi na Vostoke. Tezisi dokladov. Yerevan.
BEKNSHTAM, A. N . 1951. Drevnyaya Fergana. Tashkent.
BICHURIN,N . Y. 1950. Sobranie svedeniy o narodakh, obitavshikh v Sredney Azii
drevnie vremena, Vols. 1-11. Moscow/Leningrad.
BILALOV,A. I. 1980. I t istorii irrigatsii Ustrushani. Dushanbe.
BILALOV,A. I.; BELYAEVA, T. V. 1980. Issledovanie tsitadeli Khodzhenta v 1975 g. ART,
15.
BURYAKOV, Y. F. 1975. Istoricheskaya topografiya drevnikh gorodov Tashkentskogo
oazisa (istoriko-arkheologicheskiyocherk Chacha i Ilaka). Tashkent.
BURYAKOV,Y. F.; GAI-IEVA, 2. S.; NABOKOV,V. I.; TIKHONIN,M. P.; USOVA,E. S.
1979. Rabota Shashilakskoy Skspeditsii. A O , 1978, pp. 545-6.
Drevniy Tashkent. 1973. Tashkent.
Drevnosti Charvaka. 1976. Tashkent.
Drevnosti Tashkenta. 1976. Tashkent.
Drevnosti Tuyabuguza. 1978. Tashkent.
DROYZEN, I. G . 1981. Istoriya illinizma, Vol. I. Moscow.
FILANOVICH,M. I. 1980. Raboti v Tashkente. A O , 1979, pp. 459-61.
FILANOVICH, M. I.; ABDULLAEV, K. A. 1975. Raboti na gorodishche Kanka v Tashkent-
skoy oblasti. A O , 1974, p. 515.
GAFUROV,B. G. 1972. Tadzhiki - Drevneyshaya, drevnyaya i srednevekovaya istoriya
(ed. by B. A. Litvinsky). Moscow.
-. 1974. Kushanskaya tpokha i mirovaya tsivilizatsiya. Tsentrallnaya Aziya v kushan-
skuyu ipokhu, Vol. 1 , pp. 61-80. Moscow.
GORBUNOVA,N. G . 1975. Fergana v kushanskoe vremya. Tsentral'naya Aziya v kushan-
skuyu t?pokhu, Vol. 2, pp. 50-5. Moscow.
-. 1977. Poseleniya Fergani pervikh vekov nashey 2ri (nekotorie itogi i issledovaniya).
SA, 3, pp. 107-20.
-. 1979. Itogi issledovaniya arkheologicheskikh pamyatnikov Ferganskoy oblasti
(k istorii kul'turi Fergani). SA, 3, pp. 16-34.
GRIGOR'EV,V. V. 1871. 0 skifskom narode sakakh. St Petersburg.
-. 188 1. Pokhod Aleksandra Velikogo v Zapadniy Turkestan. St Petersburg.
GULYAMOV, Y. G . 1957. Istoriya orosheniya Khorezma s drevneyshikh vremen do nash-
ikh dney. Tashkent.
-. 1974. Kushanskoe tsarstvo i drevnyaya irrigatsiya Sredney Azii. Tsentral'naya
Aziya v kushankuyu ipokhu, Vol. 1, pp. 118-22. Moscow.
HERODOTUS.1979. Histories (ed. by H . H. Huxley). Letchworth, Bradda Books.
Istoriya Tadzhikskogo naroda. 1963. Vol. 1. Moscow.
Istoriya Uzbekskoy SSR. 1967. Vol. 2. Tashkent.
ITINA,M. A. 1977a. Istoriya stepnikh ylemEn Yuzhnogo Priaral'ya (I1 - nachalo I. tis. do
n.6.). Trudi KhAEE, 10.
-. 19776. Osnovnic itogi rabot Khorezmskoy Eks~editsiiv minuvshern pyatiletii.
PIIE, 1975, pp. 187-210.
-. 1981. Khorezmkaya t k s ~ e d i t s i y a- osnovnie itogi i pcrspcktivi isslcdova~liy.
Kul'tura i iskusstvo drevnego Khorezma. Moscow.
 Bibliography and References

K H O ~ H A N I Y A ~ G. U V1981.
, Istoriya razvitiya fortifikatsii antichnogo Khorczma. SA,
2, pp. 43-56.
Koy-krilgan-kala - pamyatnik kul'turi Khorezma IV v. do n.6. 1967. M ~ ~ ~
LATININ, B. A. 1962. Voprosi istorii irrigatsii i oroshaemogo zemledeliya Drcvney
Fergani. Leningrad.
LEVINA,L. M. 1971. Keramika Nizhney i sredney Sirdar'i v I tisyacheletii ".i. Trudi
KhA EE, 7, pp. 243-9.
LITVINSKY, B. A. 1960. Saki, kotorie za Sogdom. Trudi AN TadzhSSR, 120, pp. 9 1 4 .
-. 1968. Sarmatsko-kangyuyskiy farn. Dushanbe.
LITV~NSKY, B. A.; OKLADNIKOV, A. P.; RANOV,V. A. 1962. Drevnosti Kayrak-Kumov
(Drevneyska~aistoriya Severnogo Tadzhikistana). Dushanbe.
LUKONIN,V. G. 1969. Zavoevaniya Sasanidov na Vostoke i problema kushanskoy abso-
lyutnoy khronologii. VDI, 2, pp. 2 0 4 4 .
MASSON,M. E. 1975. K voprosu o severnikh granitsakh gosudarstva 'Velikikh Kushan'.
Tsentral'na~aAziya v kushanskuyu tpokhu, Vol. 2, pp. 42-9. Moscow.
NEGMATOV,N. N . 1957. Ustrushana v drevnosti i v rannem srednevekov'e. Stalinabad.
-. 1962. K istorii khodzhentskoy tsitadeli (nekotorie itogi istoriko-arkheologiches-
kogo izucheniya). I O O N A N Tadzh.SSR, Vol. 1, pp. 51-65.
-. 1973. Issledovaniya v Severnom Tadzhikistane v 1970 g. ART, 10 (1970), pp. 79-
105.
-. 1975. Issledovaniya v Severnom Tadzhikistane v 1971 g. ART, 11 (1971),
pp. 143-68.
-. 1977. Issledovaniya Severo-Tadzhiki~tansko~o arkheologicheskogo otryada v 1973
g. ART, 13 (1973), pp. 112-36.
-. 1980a. O b Zllinisticheskikh Glementakh v kul'ture ustrushano-khodzhentsko-
zapadnoferganskogo regiona Sredney Azii. Vsesoyuzniy simpozium po problemam
illinisti~hesko~ kul'turi na Vostoke. Tezisi dokladov. Yerevan.
-. 1980b. O b i s s l e d ~ v a n i ~ a kSTAKE
h v 1975 g. ART, 15 (1975).
-. 1982a. T h e Development of Urbanized Cultures in Tadjikistan and Problems of
Cultural Contacts. Information Bulletin, N o . 2. International Association for the
Study of the Cultures of Central Asia. Moscow.
-. 19826. Drevnyaya Baktriya i istoriko-kul'turnie oblasti Sredneaziatskogo mezh-
durech'ya. Drevneychie kul'turi Baktrii: sreda, razvitie, svyazi, pp. 51-3. Dush-
anbe.
-. 1982c. Tadzhikistan na puti k urbanizatsii. Kul'tura pervobitnoy tpokhi Tadzhikis-
tana (or mezolita do bronzi), pp. 72-88. Dushanbe.
NEGMATOV, N . N.; BELYAEVA, T. V. 1977. Raskopki na tsitadeli Leninabada i lokalizat-
siya Aleksandrii Eskhati. A O , 1976, p. 569.
NEGMATOV,N . N.; BELYAEVA, T. V.; MIRBABAEV, A. K. 1982. K o t k r i t i ~ ugoroda
pozdney bronzi i rannego zheleza - Nurtepa. Kul'tura pervobirno~rZpokhi Tadz-
hikistana (or mezolita do bronzi), pp. 89-1 11. Dushanbe.
NEGMATOV,N. N.; SALTOVSKAYA, E. D. 1962. 0 rabotakh ~hodzhentsko- str rush an-
skogo otryada v 1960 g. TIIANTadzh.SSR, 34, pp. 69-89.
-. 1975. MateriaI'naya kul'tura kushanskogo vremeni v ~ s t r u s h a n ei zapadnoy Fer-
gane. Tsentral'naya Azijla v kushanskuyu ipokhu, Vol. 2, pp. 258-66. MOSCOW.
NEKAZIK,E. E. 1972. Problemi issledovaniya Khorezmskikh gorodov antichnogo
per-iod~.Drevnij~gorod Sredney Azii, pp. 34-7. ~ e n i n ~ r a d .
 Bibliographjl and References

N E K A Z I KE.
, E. 1981. K probleme razvitiya gorodov drevnego Khorezma. Kul'tura i
iskusstvo drevnego Khorezma. Moscow.
NISSEN,H. 1866. Griechische und romische Metrologie. Handbuch der klassischen Alter-
tums Wissenschaft, Vol. I. Nordlingen.
PETRUSHEVSKIY, F. P. 1849. Obshchaya metrologiya, Part 1. St Petersburg.
RANOV,V. A.; SALTOVSKAYA, E. D . 1 9 6 1 . 0 rabotakh Ura-Tyubinskogo otryada v 1959
g. TIIANTadzhSSR, 31, py. 108-29.
RTEVELADZE, E. V. 1980. Kiropol' i vremya vozniknoveniya gorodov v Tashkentskom
oazise. Stroitel'stvo i arkhitektura Uzbekistana, N o . 3.
STAVISKIY,B. Y. 1961. 0 severnikh granitsakh Kushanskogo gosudarstva. VDI, 1,
pp. 108-14.
TOLSTOV, S. P. 1948. Po sledam drevnekhorezmiyskoy tsivilizatsii. M o s ~ o w l L e n i n ~ r a d .
- 1962. Po drevnim del'tam Oksa i Yaksarta. MOSCOW.
TOLSTOV, S. P.; ITINA, M. A. 1966. Saki nizov'ev Sir-dar'i (po materialam Tagiskena).
SA, 2, pp. 151-75.
TOMASCHEK, W. 1887. Central-asiatische Studien. Vol. I: Sogdiana. Vienna.
VAYNBERG, B. I. 1979. Pamyatniki kuyusayskoy kul'turi. In: M. A. Itina (ed.), Kochev-
niki na granitsakh Khorezma. Trudi KhAEE, 11, pp. 1-76.
VISHNEVSKAYA, 0 . A. 1972. Raskopki na gorodishche Kyuzeli-gir. AO, 1971, pp. 533-4.
-. 1973. Kul'tura sakskikh plemen nizov'ev SFrdar'i v VII-V vv. d o n.2. (Po materia-
lam Uygaraka). Trudi KhAEE, 8.
-. 1978. Raskopki na gorodishche Kyuzeli-gir. AO, 1977, pp. 544-5.
VISHNEVSKAYA, 0. A.; ITINA, M. A. 1971. Rannie saki Priaral'ya. MIA, 177, pp. 197-
208.
VOROB'EVA,M. G . 1973. Dingil'dzhe. Usad'ba seredini I tis. do n.6. v drevnem Kho-
rezme. Moscow.
-. 1979. Problema 'Bol'shogo Khorezma' i arkheologiya. Etnografiya i arkheologiya
Sredney Azii. Moscow.
ZADNEPROVSKIY, Y. A. 1956. O b Ztnicheskom sostave naseleniya drevney Fergani.
KSIIMK, 61, pp. 39-44.
-. 1962. Drevnezemledel'cheskaya kul'tura Fergani. MIA, 118.
-. 1966. Chustskaya kul'tura v Ferganskoy doline. S r e d n a ~ aAziya v tpokhu kamnya
i bronzi, pp. 193-207. MoscowILeningrad.
-. 1978. Chustskaya kul'tura Fergani i p a m ~ a t n i k ir a n n e z h e l e ~ n o ~veka
o Sredney
Azii. Moscow. (Doctoral dissertation.)
ZEIMAL,E. V. 1962. Klad rimskikh denariev iz Tadzhikistana. Numizmatika i tpigrafika,
Vol. 3, pp. 141-6.

ABETEKOV,A. K. 1970. Novie a r k h e ~ l o ~ i c h e s k dannie

ie o khozyaystve drevnikh USU-
ney. KSIA, 122, pp. 67-70.
AKISHEV,K. A. 1969. Zimovki-poseleniya i zhilishcha drevnikh usuney. Izvestiya Aka-
demii Nauk KazakhskoY SSR, seriya Obshchestvennikh nauk, No. 1, pp. 29-47.
Alma-Ata.
AKISHEV,K. A.; KUSHAEV,G. A. 1963. Drevnyaya kul'ttrra sakov i usuncy dolini reki
Ili. Alma-Ata.
 Bibliography and Refirences

BABANSKAYA, G . G . 1956. Berkkarinskiy mogil'nik. (K nekotorirn itogam izucheniya

~arnyatnika.) Trudi Instituta istorii, arkhcologii i itnografii Akademii NaHk
Kazakhskoy SSR, Vol. 1 , pp. 189-206. Alma-Ata.
BARUZDIN,Y. D. 1962. Pamyatniki kochevnikov rubezha i pervikh vekov n.i. In: Y. I).
Baruzdin and G. A. Brikina (eds.), Arkheologicheskie pamyatniki Satkena i Lyay-
lyaka (Yugo-Zapadnaya Kirgiziya). Frunze.
BEKNSHTAM,A. N. 1940. Zolotaya diadema iz shamanskogo pogrcbeniya na r. Karga-
linke. KSIIMK, 5.
-. 1949. Osnovnie itapi istorii kul'turi Semirech'ya i TyanB-Shanya. SA, 1 1 ,
pp. 337-84.
-. 1950. Problemi drevney istorii i Stnogcneza yuzhnogo Kazakhstana. Izvcs
iya Akademii Nauk Kazakhskoy SSR, seriya arkhcologicheskaya, No. 2. Alrna-
Ata.
-. 195 1 . Ocherk istorii gunnov. Leningrad.
-. 1952. Istoriko-arkheologicheskie ocherki Tsentral'nogo Tyan'-Shanya i Pamiro-
Alaya. MIA, 26.
BICHURIN,N. Y. 1950. Sobranie svedeniy o narodakh obitavshikh v Sredney Azii v drc-
nie vremena, Vols. 1-111. Moscow/Leningrad.
BURYAKOV,Y. F. 1982. Genezis i Ztapi razvitiya gorodskoy kul'turi Tashkenrskogo
oazisa. Tashkent.
CHAVANNES, E. 1905. Le pays d'occident d'apres le Wei Lio. TP, 6, pp. 519-71.
-. 1907. Le pays d'occident d'aprks le Hou Han Chou. TP, 8, pp. 149-234.
GINZBURG,V. V.; TKOFIMOVA,T. A. 1972. Paleoantropologiya Sredney Azii. Moscow.
HARMATTA, J . 1950. Studies on the History of the Sarmatians. Budapest.
-. 1952. T h e Dissolution of the H u n Empire: 3. Acta Archeologica, Vol. 11, fasc. 4.
Budapest.
HULSEWE, A. F. P. 1979. China in Central Asia. The Early Stage: 125 K C . - A.D. 23. (An
annotated translation of Chapters 61 and 96 of the History of the Former Han
Dynasty, with an introduction by M. A. N. Lowe.) Leiden, E. J. Brill.
KIBIROV,A. K. 1959. Arkheologicheskie raboti v Tsentral'nom Tyan'-Shane 1953-1955
gg. Trudi Kirgizskoy arkheologichesko-itnograficheskoy ikspeditsii, Vol. 2, pp. 63-
138. Frunze.
KLYASHTOKNIY, S. G . 1964. D r e ~ n e t ~ u k r s k runicheskie
ie parn~atnikikak istochnik po
istori Sredney A z i i Moscow.
KOZHOMBERDIEV, I. K. 1963. Katakombnie pamyatniki Talasskoy dolini. ~ r k h c o l o g -
icbeskie pamyatniki Talassko~dolini. Frunze.
-. 1977. Osnovnie etapi istorii kul'turi Ketmen'-Tyube. ~ e t m e n ' - T ~ l u b pp. e , 9-24.
Prunze.
KYUNER,N. V. 1961. Kitayskie izvestiya o narodakh Yuzhnoy Sibiri, Tscntral'no~~ Azii i
Dal'nego Vostoka. Moscow.
LEVINA,L. M. 1966. Keramika i voprosi khronologii pamyatnikov Dzheti-Asarskoy
kul'turi. Material'naya kul'tura narodov Sredney Azii i Kazakhstana. Moscow.
LITVINSKY, B. A. 1967. Dzhunskiy mogil'nik i nekotorie aspekti kangyuyskoy problemi.
SA, 2 , pp. 29-37.
s , A.; KHAZANOV,A. M. 1979. P o d b o p i e i katakombnie pogrebeniya
L ~ K l r v l ~ V.
mogil'nika Tuz-Gir. In: M. A. Itina (ed.), Kochcvniki na gratritsakh Khorezma.
Moscow.
 Bibliography and References

MAKSIMOVA, A. G.; MERSHCHIEV, M. S.; VAYNBERG, B. I.; LEVINA,L. M. 1968. Drev-

nosti Chardari. Alma-Ata.
MANDEL'SHTAM, A. M. 1966. Kochevniki na puti v Indiyu. MIA, 136.
-. 1971. Meshre~itakhtinkiymogil'nik. KSIA, 1.28, pp. 66-73.
-. 1 9 7 5 ~ .K gunnskoy probleme. Sootnoshenie drevnikh kul'tur Sibiri s kul'turami
sopredel'nikh territoriy. Novosibirsk.
-. 1975b. Pamyatniki kochevnikov kushanskogo vremeni v Severnoy Baktrii. Lenin-
grad.
MARUSHCHENKO, A. A. 1959. Kurgannie pogrebeniya sarmatskogo vremeni v podgor-
noy polose Yuzhnogo Turkmenistana. Trudi I I A E A N Turkm. SSR, 5, pp. 110-22.
MATSULEVICH, L. A. 1947. Alanskaya problema i Etnogenez Sredney Azii. SE, 6-7,
pp. 1 2 5 4 7 .
OBEL'CHENKO, 0 . V. 1961. Lyavandakskiy mogil'nik. IMKU, No. 2, pp. 97-176.
SOROKIN,S. S. 1956a. 0 datirovke i tolkovanii Kenkol'skogo mogil'nika. KSIIMK, 64,
pp. 3-14.
-. 1956b. Sredneaziatskie podboynie i katakombnie zakhoroneniya kak pamyatniki
mestnoy kul'turi. SA, 26, pp. 97-1 17.
TASKIN, V. S. 1973. Materialipo istorii Syunnu (po kitayskim istochnikarn), Vol. 2. Mos-
cow.
TOLSTOV,S. P. 1962. Po drevnim del'tam Oksa i Yaksarta. MOSCOW.
VAYNBERG,B. I. 1981. Skotovodcheskie plemena v drevnem Khorezme. Kul'tura i
iskusstvo drevnego Khorezma. Moscow.
VOEVODSKIY, M. V.; GRYAZNOV, M. P. 1938. U-sun'skiy mogil'nik na territorii Kirgiz-
skoy SSR. K istorii U-suney. VDI, 3.
YAGODIN,V. N . 1970. Pamyatniki kochevnikh plemen drevnosti i srednevekov'ya.
Drevnyaya i ~ r e d n e v e k o v a ~kul'tura
a yugo-vostochnogo Ustyurta. Tashkent.
ZADNEPROVSKIY, Y. A. 1971. O b Etnicheskoy prinadlezhnosti ~ a m y a t n i k o vkochevni-
kov Semirech'ya usun'skogo periods I1 v. d o n.E. - V v. n.;. Strani i narodi Vos-
toka, Vo1. 10.
-. 1975a. K istorii kochevnikov Sredney Azii Kushanskogo perioda. Tsentral'naya
Aziya v kushanskuyu ipokhu, Vol. 2, pp. 293-6. Moscow.
-. 1975b. Opit regional'noy klassifikatsii pogrebal'nikh pamayatnikov kochevnikov
Sredney Azii drevnego periods (I1 v. d o n.E. - VI v. n.:.). Stranitsi istorii i mate-
rialJnoy kul'turi Kirgizstana, pp. 159-67. Frunze.

AL-BIRUNI,A. M. 1957. Pamyatniki minuvshikh pokoleniy (translation). Tashkent.

BIVAR,A. D. H . 1970. Hiiriti and the Chronology of the Kusiinas. BSOAS, 33, pp. 10-21.
-. 1983a. The History of Eastern Iran. Cambridge History of Iran, Vol. 3 (1). Cam-
bridge, Cambridge University Press.
-. 1983b. The Political History of Iran under the Arsacids. Cambridge History of
Iran, Vol. 3 (1). Cambridge, Cambridge University Press.
CHAVANNES, E. 1905. Les pays d'occident d'aprks le Wei Lio. TP, 2nd series, Vol. 6,
pp. 519-71.
CHRISTENSEN, A. 1944. L'Iran sous les Sassanides. Copenhagen.
 Bibliography and References

DEREVOISE,N . C. 1938. A Political History of Parthia. Chicago, University of Chicago

Press.
DRESVYANSKAYA, G . Y. 1974. 'Oval'niy' dom khristianskoy obshchini v Starom Merve.
Trudi YuTAKE, 15, pp. 155-81.
D'YAKONOV, M. M. 1961. Ocherk istorii drevnego Irana. Moscow.
ERSHOV, S. A. 1959. Nekotorie itogi arkheologicheskogo izucheniya nekropolya s
ossuarnimi zakhoroneniyami v rayone goroda Bayram-Ali. Trudi IIAE A N
Turkm. SSR, 5, pp. 160-204.
FRYE,R. N. 1983. T h e Political History of Iran under the Sasanians. Cambridge History
of Iran, Vol. 3 (1). Cambridge, Cambridge University Press.
GAFUROV,B. G . 1972. Tadzhiki - Drevneyshaya, drevnyaya i ~ r e d n e v e k o v a ~istoriya
a
(ed. by B. A. Litvinsky). Moscow.
GHIRSHMAN,R. 1947. Le problkme de la chronologie des Kouchans. CHM, 3, pp. 689-
722.
GOBL, R. 1984. System und Chronologie des Miinzpragung des Kuianreiches. Vienna.
HARMATTA,J . 1957. T h e Parthian Parchment from Dura-Europos. Acta Ant. Hung., 5,
pp. 261-308.
-. 1958. Die ~ a r t i s c h e nOstraka aus Dura-Europos. Acta Ant. Hung., 6 , pp. 87-175.
-. 1965. Minor Bactrian Inscriptions. Acta Ant. Hung., 13, pp. 149-205.
-. 1969. Late Bactrian Inscriptions. Acta Ant. Hung., 17, pp. 297432.
HENNING,W. B. 1952. T h e Monuments and Inscriptions of Tang-i Sarvak. A M (new
series), Vol. 11, Part 2, pp. 151-78.
HERRMANN,A. 1938. Das Land der Seide und Tibet im Lichte der Antike. Quellennd
Forschungen zuv Geschichte der Geographie und ~ o l k e r k u n d e , No. 1 .
Leipzig.
HERZFELD,E. 1924. Paikuli, Monument and Inscription of the Early History of the Sassa-
nlan Empire. Berlin.
KATSURIS,N.; BURYAKOV,Y. 1963. Izuchenie remeslennogo kvartala antichnogo Merva
u severnikh vorot Gyaur-kali. Trudi YuTAKE, 12.
KOSHELENKO,G . A. 1966. The Beginning of Buddhism in Margiana. Acta Ant. Hung.,
14, pp. 175-83.
-. 1977. Rodina Parfyan. Moscow.
KRUGLIKOVA, I. T . 1974. Dil'berdzhin (Raskopki 1070-1972 gg.), Vol. 1. Moscow.
~ R U G L I K O V A ,I. T.; PUGACHENKOVA, G . A. 1977. Dil'berdzhin (Raskopki 1970-1973
gg.), Vol. 2. Moscow.
LITVINSKY,B. A. 1968. Outline History of Buddhism in Central Asia. Dushanbe.
LITVINSKY,B. A.; SEDOV, V. A. 1983. Tepay-shakh, Kul'tura i svyazi kushansko~1
Baktrii. MOSCOW.
-. 1984. Kuf'ti i rituali kushanskoy Baktrii. Progrebal'niy obryad. Moscow.
LJVSHITS,V. A. 1969. K otkritiyu baktriyskikh n a d ~ i s e y na Kara-tepe. Buddi~lskie
peshcheri Kara-tepe v Starom Termeze. Moscow.
LOGINOV,S. D.; NIKITIN,A. B. 1986. Moneti s vsadnikom iz Menfa. SA, 3, pp. 243-9.
LUKONIN,V. G . 1961. Iran v 6pokhu pervikh Sasanidov. Leningrad.
- 1969a. Kul'tura sasanidskogo Irana. Iran v Ill-V vv. Ocherki po istorii kul'tu~i.
Moscow.
-. 19696. Zavoevaniya Sasanidov na Vostoke i problerna kushanskoy absolyutnoy
khronologii. V D I , 2, pp. 20-44.
 Bibliography and References

L U K O N I NV., G. 1979. Iran v III veke. Novie materiali i opit istoricheskoy rekonstruktsii.
Moscow.
MARICQ,A. 1953. Recherches sur les Res Gestae divi Saporis. Memoires de I'Acadbmie
Roya/e de Belgique. Classe des lettres et des sciences morales et politiques (Brussels),
Vol. 47, No. 4.
MARKWART, J. 1931. A Catalogue of the Provincial Capitals of Eranjahr. Rome.
MEREZHIN,L. N. 1962. K kharakteristike keramicheskikh pechey perioda rabovladeniya
i rannego srednevekov'ya v Mervskom oazise. Trudi YuTAKE, 11, pp. 12-40.
NIKITIN,A. B. 1984. Khristianstvo v Tsentral'noy Azii (drevnost' i srednevekovPe).
Vostochniy Turkestan i Srednyaya Aziya. Moscow.
NOLDEKE,T. 1887. Aufsatze zur persischen Geschichte. Leipzig.
OBEL'CHENKO, 0 . V. 1969. Nekropol' drevnego Merva. (Material? raskopok 1955g.).
Trudi YuTAKE, 14, pp. 86-99.
PAPAKHRISTU, S. A.; USMANOVA, Z. I. 1978. Buddiyskiy pamyatnik v Merve. A O , 1977,
pp. 548-9.
PILIPKO,V. N . 1980. Parfyanskie bronzovie moneti so znakom P pod lukom. VDI, 4,
pp. 105-24.
PUGACHENKOVA, G. A. 1958. Puti razvitiya arkhitekturi Yuzhnogo Turkmenistana pori
rabovladeniya i feodalizma. Trudi YuTAKE, 6.
-. 1962. Koroplastika drevnego Merva. Trudi YuTAKE, 11, pp. 117-73.
-. 1979. Iskusstvo Baktrii gpokhi Kushan. Moscow.
PUGACHENKOVA, G. A.; RTVELADZE,E. V., et al. 1978. Dal'verzin-tepe. Kushanskiy
gorod na Yuge Uzbekistana. Tashkent.
RASCHKE,M. 1978. New Studies in Roman Commerce with the East. Aufstieg und
Niedergang der romischen Welt, Vol. 2, No. 9. Berlin.
RUTKOVSKAYA, L. N . 1962. Antichnaya keramika drevnego Merva. Trudi YuTAKE, 11,
pp. 41-1 16.
SCHLUMBERGER, D.; LE BERRE,M.; FUSSMAN, G. 1983. Surkh Kotal en Bactriane. Paris.
(MDAFA, 25.)
SELLWOOD,D . G. 1967. A Die-engraver Sequence for Later Parthian Drachms. NC.,
1967, pp. 13-28.
STAVISKIY, B. Y. 1977. K ~ s h a n s k a ~Baktriya:
a problemi istorii i kul'turi. Moscow.
TABARI (trans. by T. Noldeke). 1879. Geschichte der Perser und Araben zur Zeit der
Sassaniden. Leiden.
TAQIZADEH, S. H. 1943-46. The Early Sasanians. Some Chronological Points which POS-
sibly Call for Revision. BSOAS, 11, pp. 6-51.
USMANOVA,Z. I. 1963. Erk-kala (po materialam YuTAKE 1955-1959 gg.). Trudi
YuTAKE, 12.
-. 1975. Prodolzhenie raskopok v Starom Merve. A O , 1974, pp. 530-1.
USMANOVA, Z. I.; DRESVYANSKAYA, G. Y. 1976. Raskopki na gorodishchakh Starogo
Merva. A O , 1975, pp. 553-4.
WIDENGREN, G. 1956. Recherches sur le fiodalisme iranien. OS, 5, pp. 79-182.
ZEIMAL,E. V. 1983. Drevnie moneti Tadzhikistana. Dushanbe.
 INDEX
Note: Page numbers in italics refer to illusirations

Abdagases 199 lacquer tray 220

Abeacus, King of the Siraci 466 metalwork 218, 219
Abisares 79, 81, 82, 84 pebble graves 213-18
Achaemenid art 31, 39, 45, 50-65, 53, silk 215, 217
55-7 wooden chamber graves 2 18-21
Achaemenid Empire 494-5 (map) wooden plate 215, 216
and Alexander 67-8, 69, 82, 84, 87 Alans 136, 467-8
culture 55-7 Alexander the Great 20, 67-88, 496-7
economy 46-7, 53-5 (map)
extent 442-4 Aornos captured 78
language 55 Aspasians 74-5
monetary system 53-4 Assacenians 75-8
nomadism, transition to 19-20 death and aftermath 87-9
postal service 52-3 founding cities 70
state administration 52-3 Indus region 73-4
wars 44-6 motivation 68
as world power 43-6 against Persia 46
Achaemenid kings 43-4, 486-7 and Porus 81-2
Agalassi 84-5 resistance in Bactria 71-2
Agathocles 115, 127 retreat 84-7
agriculture sculpture 83
Hsiung-nu 158 and Taxila 79-81
Kushan 273-7 Transoxania 72-3, 457-8
military agricultural colonies 240-3 Alexandria ad Caucasum (Begram) 70, 73,
and nomads 182-4 90, 105, 108
Parthian 137 Alexandria in Arachosia (Ghazni) 70,
terracing 270- 1 108
Ahura Mazda 49, 58-9, 315, 31 7, 324 Alexandria in Aria (Herat) 70, 91
Akkadian translations 48-9 Alexandria Eschate (Khojand) 71, 91, 100,
Alagou 105
ceramic jar 215, 216 Alexandria in Margiana 91, 482; see also
fire sticks 21 5, 217 Erk-kala; Gyaur-kala
 Alexandria Prophthasia (Phrada) 70 Sakas 205-6
Alexandropolis (Kandahar) 70 Ardashir 136, 255, 475-6, 480
Altai mountains, grave sites 222-5 Ardhaniriivara ( ~ i v aform) 376
Alyattes, ruler of Lydia 38 Ardoxso 325, 326, 329, 432, 433
Ammianus Marcellinus 467 Aria 40, 42, .69, 70, 87, 90
Amorges, King of Scythia 43 Armenia 135, 136, 473, 479
Amu Darya (Oxus), irrigation 268, 269, Arrian
2 70 Alexander the Great 73, 74, 77-8, 84,
ancestor gods 347 85-6
Andragoras 131 Anabasis 40, 67
Angra Mainyu 59 Aornos 78
animal style 28, 31, 32, 32, 33, 189, armour of nomads 30
349-50, 354, 355 Cyrus 40
Antialcidas 102 Massaga 76-7
Antioch 91, 108, 482 Arsaces 131, 141
Antioch of Scythia 100, 103; see also Arsacid dynasty 134-6, 145, 146-7, 457-8,
Alexandria Eschate 48 1
Antiochus 1 72, 90-1, 92 ArSima letters 399
Antiochus I1 92, 95, 96, 100 Artabanus I 132, 141, 181
Antiochus I11 95-6, 99, 100, 107, 131-2 Artabanus I1 134-5
Antiochus VII 132 Artabanus V 136, 474, 476, 481
Antoninus Pius 474-5 Artacoana 70, 91
Aornos, captured 78 Artaxemes I, inscription 55
Aorsi 463, 465-7 Artaxerxes 11, cylinder seal 52
Apama 72, 91 Arthaiastra 80, 302
Apasiacae 131 Aruktau burial ground 182, 471
Aphrodite 348 Arzhan burial mound 26, 27, 29, 32
Apollodorus of Artemita 99, 100 ASoka 96, 107, 117, 398-400
Apollodotus I1 102 ASokan rock edicts 78, 106, 314, 398,
Ara inscription 261 399, 401-2, 405-7
Arachosia 40, 53, 55, 69, 70, 87, 192 Aspasians 74-5
Arachosians 60, 69, 70, 86 Assacenians 75-8
Aral region 441-4, 457, 467 Assagates 77
irrigation 444 Assyria 37-8, 41
Aramaic chancellery practice 397-8 Assyrian texts 36
Aramaic language 48, 52, 55, 397-9, Astis 73-4
400- 1 astral cult 449
Aramaic script ostracon 107, 400-1, 407 astronomy 129
Aramaic transcriptions 403 Astyages, King of Media 38, 39, 40
architecture Aivaghosa 257, 309, 324, 436, 437
Buddhist 333-4, 336-9 Attic system of weights 187
funerary 113 Augustus, Emperor 134
Gandhiira 366-7 Avadanas 437
Graeco-Bactrian 94, 110-14, 128 Avesta 23, 41-2, 57-8, 442
Greek domestic 1 13-14 Avestan language 47, 404
Kushan 278-9, 308, 333-6 Avestan society 42
Parthian 138, 147 Ay Khanum 92-5, 108-13
 abandoned 103 Bactrian language, Greek script 127, 322,
Aramaic ostracon 107, 400-1, 407 402-3, 422-32
coins 95, 100 Bactrian people, at Persepolis 57
courtyard 111 Bactrian-style head-drcss, in clay
inscriptions 417, 420, 421 and alabaster 123
excavations 105, 106 Balarima cult 1 15, 375-6, 378
Greek colonies 104-5 Baluchistan, Alexander the Great 85-7
Greek inscriptions 407 barrows 26, 29, 32; see also kurgans
gymnasium 111-12, 112 Barsaentes 69, 70, 80
Hellenism 107 batik-dying 231, 232
Kharosthi inscription 434 bazaars 305-7
palace 94, 109-1 1 Bazar-kala 447
plan 93 Bazira 76-7
statuette 119 beacon towers, Han 242-3
stone stele 107 Begrarn 295-6
temple 1 15 bazaar 306
terracotta moulds 120 bellows pipe 233
theatre 112-13 coins 250-1
i y i g a p a ~ a s381, 383 excavations 250-1
Aygirdzhal burial group 461 ivory 296,357,358
Ayrtam inscription 326, 422, 432-3 Kushan art 356-8
Azes 194-6, 200, 202, 203, 248, 249 plasterwork 357
Azilises 195 see also Alexandria ad Caucasum
AOso-Hephaistos 321 Berossus, Cyrus 24, 40, 43
Besshatir burial ground 26
Babylonia, captured 43 Bessus 67, 69, 70, 71
Bactria Bhita 292, 294
archaeological remains of nomads 41, billon coins 203-4
182-4, 353-4 Bindusira, Emperor 96
burial grounds 471-2 Birkot, inscribed potsherd 408
cities 92, 95, 108- 10 Bisutun inscription 40, 41, 44, 45, 48-9,
coins 187-8 50, 443
dark ages 185-9 bits (horse) 152
independence 100 boar incarnation 376
invaded 17 Bodhisattva statues 362, 365, 389
jewellery 351 bow, composite 45, 163, 278
Kushan art 352-6 Brahmanism 314, 372
language 401 -4 Brihrni script 435-6
pastoralism 41 brocade robe 233
political instability 185 bronze casting 368
religion, pre-Kushan conquest B u c e ~ h a l a82
313-15 Buddha, statues 362-5, 363, 386, 387, 389
resisting Alexander 71 -2 Buddhism
Silk Route 484 ASoka 96
Yiieh-chi11 179-80 China 237-8, 491
see also Ta-hsia; Graeco-Bactria Gandhira 359
Bactrian camels, at Perscpolis 58 Graeco-Bactria 3 14
 and Greek 106 Tashkurgan 2 10- 1 1
Kanishka 25 1, 323-4 Tenlik 462
Kushan period 308, 328-9, 345 Tillya-tepe 350
and local cults 116-17 Tulkhar 182, 471
Mathura 372 Western Region 237
Merv 484 Wu-sun 461
Parthia 149-50 see also barrows; graves; kurgans
Sakas 206-7
Buddhist architecture, Kushan period Caityapata 3 8 1
333-4, 336-9 Cakrapata 381
Buddhist art 361-2, 373 Cambyses I, King of Persia 38
Buddhist Hybrid Sanskrit 324, 437 Cambyses 11, King of Persia 40, 43-4
Buddhist sculpture 259, 346, 362-4, 384-9 camel breeding 151
Bukhara oasis 266 Candana Kanishka 323-4, 491
burial grounds 278 Candragupta 80-1, 89-90, 250
coins 284 capitals 334, 335
burial grounds Caracalla 136, 474
Alagou 213-21 caravan trade 186-7
Altai mountains 222-5 carpets
Aral region 467 Noin-Ula 160, 161, 162
Aruktau 182 Parthia 137
Arzhan 26-7, 29, 32 Carrhae, battle of 134
Aygirdzhal 461 Casia, and Yueh-chih 171-2
Bactria 471 -2 cattle-breeding
Berk-kara 464 Chorasmia 446-7
Besshatir 26 nomadic 23, 156, 157
Bukhara 278 Wu-sun 459-60
Chil'pek 461 ceramics 30, 122; see also pottery
Chorasmia 449 Chach (&i?) 277, 284, 452-3, 464; see also
Derestui 159 Tashkent oasis
Dzheti-Asar 467 Chang Ch'ien
D z h u n 464 Bactria 103, 125, 180, 182
Ferghana 452 captured 177
Hsiung-nu 159-63 Parthia 187
K'ang-chu 464 to Wu-sun 228, 459
Kargali 461 -2 t o Yueh-chih 171, 228, 245
Kaunchi 464 Ch'iang language group 239
Kenkol 469-70 Ch'iang tribes, cremation 21 0- 11
Khunui-gal 159 Chilas
Kokkum 182, 471 K h a r o s ~ h inscriptions
i 41 1, 434-5
Lou-Ian 211-12 petroglyphs 192
Noin-Ula 159-63 Chilik barrow 29
nomad 26, 32, 182, 457-8, 486 Chil'pek burials 461
Scythian 26 Ch'in dynasty 153
Sudzhinsk 159 China
Tagisgen 26, 33 Buddhism 237-8, 491
Tamdi 464 Ch'in dynasty 153
 - - -

colonies 240-3 Graeco-Bactrian 99, 125-7

Han, emperors 172, 228, 469 Han 234-5
and Hsiung-nu 153-5, 160, 221, 227 Huvishka 257, 325-8
and Kushans 256, 286 Indo-Greek 102, 124, 126, 187, 203-4
and nomads 153 Indo-Parthian 197-8
official seal 240 joint rule 199-200
and Parthia 187 Kanishka 262, 281
Silk Route 243-6 Kujula Kadphises 196, 3 17-1 8
trade 20, 187, 286 Kushan 279-85, 367-8
and Western Regions 229-30, 232-4, Margiana 282-3, 481
238-40 Maues 193-4
Xinjiang 209, 225 Pahlavas 204
Chinese bone inscriptions 152 Parthia 138, 139-43, 141-3, 181, 283
Chinese text o n wooden tablets 236 and portraiture 127
Chionites 21 Sanabares 283
Ch'iu-chiu-ch'iieh: see Kujula Sino-Kharoslhi 235
Kadphises Sogdian 188
Chorasmia (Khwirizm) 42, 455-6 Tillya-tepe 183
cattle-breeding 446-7 tetradrachms 280
coins 282 and trade routes 124
ethnic history 442-4 Vima Kadphises 248, 280, 490
irrigation 268, 269, 272-3, 444 comb, ivory 353
Koy-Krilgan-kala 447-9 Cophaeus 77
mining and manufacture 445-6 cotton 231, 232
origins 42 cowries 152
Toprak-kala 297-9 cremation, Ch'iang tribes 2 10-11
urbanization 446-5 1 Croesus, King of Lydia 39
vineyard 27J, 276 Ctesias
Christianity, Merv 483-4 Cyrus' death 43
Chii-chih 172-3 Cyrus' war 24
Cimmerians 37 Median rule 40
CiSpiS (Tei'spes) 37 Ctesiphon 135, 187
city deities 314-15, 434 cults
Clearchus of Soli 92, 107 astral 449
Coenos 72 BalarZma 375-6, 378
coins fertility 116, 345
A y Khanum 95, 100 fire 116, 321
Azes 194-6 Hauma 315
Bactrian 187-8 H e m p 316
Begram 250-1 Heracles 343
billon 203-4 Kirttikeya 377-8, 381
Chach 284, 464 local 62, 116, 127-8, 314-15
Chorasmia 282 Mat, royal 320
copper 281, 284-5 Nana worship 322
Demetrius 187 OXUS 1 16, 118, 3 14, 407; see also
Diodotus 95 VaxSu
drachrns 141-3 ~ i v a322, 329, 376, 427
 index

sun worship 165, 3 16 devati sculpture 347, 348

Surkh Kotal, roayl 320 Dhammapada 436-7
Vishnu 375, 407 Dilberjin
yak$ 329 Bactrian writing 422, 424-7
cuneiform inscriptions, O l d Persian 47-8, dwelling house 336
397 inscribed amphorae 408
Cyaxares, King of Medes 37, 38 mural 339, 340
Cyropolis 43, 71, 450; see also Ura-tyube plan 300
Cyrus 1 24, 37 temple 319, 355-6
Cyrus I1 38-40, 43 Dingildzhe oasis 446-7
Dio Cassius 256
Dahae 24, 25, 31, 43, 457 Dio Chrysostomus 186
Daiva inscription 49 Diodo rus
Dalverzin-tepe Alexander the Great 86
devati sculpture 347, 348 Alexander's death 88
dwelling houses 335-6 Cyrus 40
excavation 296-7 Massaga 76
fortress 322 Ninus 40-1
metalwork 350-1 Diodotus 1 95, 100, 131
origin 184 Diodotus I1 100, 101
painting 297, 341 divinity of kings 260-1, 263
sculpture 339-40, 349, 350 double kingship 200
Darius 1 44, 45, 69-70 drachms 140, 141-3
administrative reforms 46-7 Dura-Europos, excavations 147
Bisutun rock relief 45, 50 dwelling houses
cylinder seal 53 Dalverzin-tepe 335-6
inscriptions 24, 49, 50, 52 Dilberjin 336
Milesians 405 Greek 113-14
monetary system 53-4 Kushan 305-7
Sakas 24-5, 44, 191 Dzhanbas-kala 446, 447
statue 51 Dzheti-Asar burials 285, 450, 467, 469
taxation 53 Dzhigirbent 446, 447
Darius I11 67 Dzhun burials 464
dark ages 185-9
Dasht-i Nawur Early Iron Age 442
Bactrian inscription 249, 3 19, 41 8-20, earth-pit vertical shaft grave 225
422-4 East Turkestan 37, 103, 286, 435, 459,
inscription, unknown script 41 7, 41 8, 461, 466; see also Xinjiang Region
419 Ectabana, capital of Media 37, 40
Dayr-Asan, Kharosthi inscription 434 Egypt, attacked by Persia 44
Demetrius I 101, 187 Elam 36, 39, 44
Demetrius I1 132 Elamite texts 48, 49, 52-3
Demodamas 90, 95 Elephantine, Aramaic documents 48, 49,
Derbices 24, 43 399
Derestui burial ground 159 Elymais 132, 135, 144
Detached Scythians, and Yueh-chih 178 embroideries, Hsiung-nu 167-8
Devakula portraits 372 encrustation style, sculpturc 350
 Index

equine armour 30 Gebekli-dcpc 485

Eratosthenes, /ran 36 Gedrosia 69
Erk-kala 91, 108 Ghazni (Alexandria in Arachosia) 70, 108
Ershi 279, 454-5 Gilgit, Kharoghi inscription 434-5
Esarhaddon, King of Assyria 37 Gindara, battle of 134
Eucratides 101, 103, 1 10, 279, 282 gold, from Rome 2 1
Eunonus, King of Aorsi 466 Gondophares 196, 197, 199, 248
Euthydemus I 100, 101, 128, 179 Gordian, emperor 478
Eylatan culture 45 1-2 Gotarzes 135
Graeco-Bactria 498-9 (map)
felt work 160-2, 212 architecture 1 10-13
Ferghana (Ta-yiian) 275-7, 451-5 coinage 99, 125-7
burial grounds 452 culture 405-7
China 176 Greek ethnic elelnent 487-8
Chinese texts on 454 Hellenism 107
irrigation 444 kingdom 132, 457
mining and manufacture 445-6 kings 20, 100
written sources 274, 277 sites 103-4
fertility cults 116, 345 see also Greek kingdoms
fire cult 116, 321 Graeco-Buddhist style 128
Florus 256 Graeco-Iranian, symbiosis 405
fortifications 108-9, 482 Granicus, battle at 67, 68
Chorasmia 449 graves
H a n 242-3 Alagou pebble 213-18
for irrigation 268, 272 Alagou wooden chamber 21 8-21
Kushan 333 Altai mountains 30-2, 222-5
Margiana 482 earth pit 225
oases 482-3 Lou-Ian 211-12, 237
funerary architecture, Ay Khanum 113 Mongolian 151-2
funerary objects Noin-Ula tombs 159-60
pebble graves 215-1 8 nomadic 182
wooden chamber graves 218-20 stone-slab 151-2
stone-warrior 222-4
Gandhira, religious architecture 366-7 Tillya-tepe 183
Gandharan art tumulus 222
and Buddhism 359-60, 361-2, 364 Wu-sun 237
foreign influences 370-1 Xianbaobao 2 10-11
jewellery 369-71 Greek coins, in India 102
Kushan 361-71 Greek colonists 45-6, 101-2, 104-5,
sculpture 362-5 117-23, 313-14
Gandhiirans, at Persepolis 60 Greek influence 20, 92, 127-9, 487-8; see
Gandhari Prakrit 436-7 also Hellenism; Philhellenism
GaneSa 378 Greek inscriptions 106
Gansu, Yiieh-chih 171-4 Greek kingdoms 92, 99-129
Gaugamela, battle of 46, 67, 80 arts 117-23 -
Gaumata 44, 49 Buddhist influence 1 16- 17
GaOas 42, 58 coinage 125-7
 Indian campaigns 101-3 Heraus 188, 189
sites 103-5 hereditary professions 3 I 0
urbanization 108- 16 Hermaeus 103, 126, 200, 3 17
see also Graeco-Bactria Herodotus
Greek language Chorasmia 42
Attic form 407 Cyrus 24, 40, 43
koine 405, 407 Darius 1 25, 405
as lingua franca 203 Detached Scythians 177-8
and local languages 105-8 Massagetae 30-1, 445
as script for Bactrian 127, 322, 402-3, nomadic armour 30
422-32 Queen Tomyris 28
Greek religion 114- 15, 207 religion of nomads 31 5-16
Greek religious art 31 7, 31 8-19 Sakas 37, 191
Greek style temples 1 15 Scythians 28, 29, 30-1, 37
Greek technology 122 taxation 53
Guptas 21, 250 Xinjiang 209
Gyaur-kala 91, 109, 482 Hieronymus of Cardia 89
Hormizdagan, battle of 136
Hadda, Kushan art 359 horse, depicted in art 31-2
Halys river, site of battles 38, 39 horse breeding
H a n Empire Ferghana 454
and Chih-chih 464 Hsiung-nu 156
coins 234-5 Kushan 276-7
fall 492 Mongolia 151
fortifications 242-3 Persia 38
and Hsiung-nu 153-5, 227-9 H o u Han-shu
t'un-t'ien colonies 240-3 K'ang-chu 463
Western Regions 229-35 Kushans 247, 248
and Yen-ts'ai 466 nomads 458
Han-shu Western Regions 221, 239
K'ang-chu 464 Yen ts'ai 466
Shan-shan 2 12 Yueh-chih 103
Western Regions 221, 232, 238-40 Hsien-pi state 155-6, 168-9
Wu-sun 220-1 Hsiung-nu (Huns) 153-9, 468, 469-70,
Yueh-chih 103, 174 J00-1 (map)
Hauma cult 315 agriculture 158
Heliocles 103, 124, 178 burials 159-63
Heliodorus 1 16- 17, 3 14, 407 cattle-breeding 156-7
Hellenism 92, 96, 97, 107, 112, 128, 348, Chih-chih, ruler 463-4, 468
488; see also Greek influence; confederations 488
Philhellenism culturc 160-2, 163-8
H e m p cult 316 economy and organization 156-9
Hephaestion 74, 405 and Ha11 227-9
Hephthalites 21, 461, 469, 472 influencc on China 163
Heracles cult 118, 343 legal system 159
Herat 91-2, 181; see also Alexandria in nomadic migrations 20, 457
Asia religion 164-5
 Index

settlements 163 role in development 441-2

and Wu-sun 459 irrigation 4 1
and Yiieh-chih 175 Amu Darya 268, 269, 270
Hsiung-nu-Hsien-pi script 166 Chorasrnia 268, 269, 272-3, 444
Hsiian-tsang 366, 434 Ferghana 444
H u script 165 fortified 268, 272
Hu-han-yeh, shun-yii 155 Graeco-Bactrians 100
Hu-yen 155 Han 242
Hunno-Sarmatian period 457-8 K'ang-chu 270
Hunza, Sacred Rock 192, 435 Kashka Darya 266-7
Huvishka Kushan 265-73
coins 258, 325-8 Salar-Karasu-Dzhun 267
coming t o throne 253 Samarkand 272
and Egypt 326-7 Syr Darya 267-8, 270
and Kanishka I1 326 Tashkent 266, 267
religious policy 327-8 Vakhsh 267
Hyrcania 41, 136 Zerafshan 265, 266, 271-2
Hyspaosines, King of Characene 133 Isidore of Charax 125, 181-2
Issik, inscription on silver cup 41 7, 420,
Iamso 328 42 1
Ichianli 30 Issus, battle at 67, 68
incrustation, jewellery 370 Ivolginsk 163
India ivory artefacts 118, 148-9, 356-8
astronomy 129
and Egypt 186 jade 172, 234
Greek campaigns in 101-2 Jaina figures 380-3, 385
and Kushans 257, 286, 304, 307-8, Jaina writing, Sakas 192-3
309- 10 Jainisrn 328, 372
and Parthia 138 Jandial temple 205, 292
religious movements 329 Jataka stories 366
and Sakas 192-6, 202-3 city official 303
Indian scribes 434 craftsmen 3 10
Indians, at Persepolis 61 jewellery
Indo-Greek coins 124, 126, 187, 203-4 Gandhira 369-71
Indo-Iranian tribes 19 Hsiung-nu 160
Indo-Parthian coins 197-8 incrustation technique 370
Indo-Parthian kingdom 20, 196-9 Kushan 278, 348
ink-wells 123 in Taxila 370
Iranian nomads in Tillya-tepe 348
anthropological studies 24 Wu-sun 462
culture 28-33, 55-7 Jina images 381, 383-4
and husbandmen 19, 20, 486-7 joint rule 261
literary sources 23-5 Ju-jan 461
society and economy 25-8 Junagadh inscription 260
stock-breeding 485-7 Justin
iron Alexander the Great 87
Parthia 137 Candragupta 80- 1
 Ninus, Assyrian monarch 41 A y Khanum 434
Chilas 41 1, 434-5
Kalali-gir 42, 446 coins 235, 433-4
Kalinganagara (modern Sisupalgarh) 295 Dayr-Asan 434
Kambojas 401, 404 early date 318
Kandahar: see Alexandropolis; ASokan Gilgit 434-5
rock edicts Kara-tepe 308, 434, 436
K'ang-chu 270, 463-4, 470, 472 Lou-lan 435
Kanishka dynasty Niya 435
coinage 280, 281 unknown language 41 8
dating 248, 249-52 Western Regions 236
Great Kushans 252-3 Khatin Rabat fragment, unknown script
triple kingship 323 417, 418
Kanishka I 249-58 Khojand 453; see also Alexandria Eschate
coin 262 Khotan
inscription 261 and Kanishka 251
political system 260-1 month names 418
religious policy 321 -2 Khotanese Saka 413-16, 418-21
statue 257 Khunui-gal 159
Kanishka I1 323, 324, 325, 326 Khvatak (modern Nur-tepe) 450
Kankali Mound excavations 380-3, 385 Khwirizm: see Chorasmia
KZpiSa (modern Begram) 318, 356-8 kingship
city deity 314-15, 434 divinity of 260-1, 263
Karakorum route 439 joint 199-200, 261
Kara-tepe Kushan 260-1, 263
Bactrian writing 422 triple 323-7
Brihmi inscriptions 436 Kirkkiz canal 268, 270
Kharosthi inscriptions 308, 434, 436 koine (Greek language) 405, 407
unknown script 417, 418 Kokkum burial ground 182, 471
Kargali burial 461 -2 Koy-Krilgan-kala, Chorasmia 447-9, 448
Kirttikeya cult 377-8, 381 Kuan-tzG, Book of 172
Kashka Darya valley, irrigation 266-7 Kuh-i Khwaja 184
Kashmir, synod of 436 Kujula Kadphises 247, 248, 489
Katra Mound, Buddha sculpture 385, 386 coins 196, 31 7-1 8
Kaunchi burials 464 H o u Han-shu 254-5
Kaunchi-tepe 452-3 Kushan Empire 20, 489
Kau~ilya80, 302, 307, 404 rise of 247, 248
kivya epics 437-8 Kunya-Uaz 278, 447
Kazakhstan, nomad culture 465, 471 kurgans 25, 28, 32, 461, 467, 470-1
Kenkol burials 469-70 Kurukata (ancient Cyropolis) 450-1
Khalchayan KuSinSahr: see Kushanshahr
monumental art 189 Kushan art 506 (map)
reconstructions 184 architecture 278-9, 308, 333-6
sculpture 339-40, 341, 342-4 in Bactria 352-6
Kharahostes, satrap 201 Gandhira 361 -71
Kharosrhi script Mathura 371-95
and Aramaic 397 minor arts 345-52
 Nagarahira and Arachosia 356-60 and Romc 256-8
in north 331-51 Saka satraps 260
Paitava and Shotorak 359-60 and ~ i v a319-20
pottery 277-8 and Yiieh-chih 174
sculpture 339-45 Kushanshahr 329, 439, 479-81, 492
temples 336-9 Ku-tu-hou 158-9
town planning 331-6 Kuyusay culture 449
Kushan Empire 500-1 (map) Kyuzeli-gir 446
administration 261-3
agriculture 273-7 Laghman, inscriptions 398
architecture 278-9 Laksmi 329
bazaars 305-7 Lalitavistara 437
city administration 307-9 Liu P'ing K'uo, inscription 244
city life 299-301 livestock-breeding 276-7
city planning 301-4 Lou-Ian (Shan-shan) 175
coinage 279-85, 317-18, 367-8 graves 21 1-12, 237
craftsmen and guilds 309-12 Kharos!hi inscription 435
crop-raising and livestock breeding Lroaspo, god 324
273-7 Lucian, Scythians 28
disintegration 492 Lung-chcng 154
dwelling houses 305-7 Lydia 38, 39
handicrafts 277-8
and India 258 Macedonians 46, 76, 89, 90, 114, 405; see
jewellery 348 also Alexander the Great
land-ownership 289-90 Magnesia, battle of 132
political system 260-1 Mahisinghika school 436, 437
royal palaces 304 Mahavastu 310, 437
rulers, chronologically 254 Mahayanism 362
and Sakas 260 Malli 84, 85
social structure 288-9 Manao Bago 324
trade and commerce 285-7, 312 Mao-tun 154, 163, 174, 175, 179, 227-8
urban patterns 291-9 Maracanda (Samarkand) 71, 72, 91, 108,
yabghus 3 16-1 8 109
Kushan languages and scripts 280, 319, Marcus Aurelius 473
417-40, 491 Margiana
Kushans Alexander 91
and China 256, 286 coins 282-3
Early 247-9 C y m s 40
Great 252-4, 489-90 political unrest 90
Greek heritage 127-9 religion 483-4
and Iran 254-6 Sasanians 48 1-4
irrigation 265-73 urbanization 91
land-ownership 289-90 see also Merv
literacy 434 Mark Antony 134
and north-eastern India 258 Massaga 75-7
origins 20-1, 171 Massagetae 28, 30-1, 71, 457, 467
religion 308, 315-16, 31 8-27, 328, 345 Alexander 72
 Index
-- -

confederations 25-6 city planning 302-3

Cyrus against 43 craftsmen 309, 31 0
religion 30-1 traders 3 12
Saka 446 military agricultural colonies 240-3
weapons 445 millstones 122
see also Saka-Massagetae Mithra 31, 315, 317, 328
Mat, royal cult 261, 320 Mithradates I 107, 132
Mathura Mithradates I1 133, 181, 199
Huvishka 258, 326 monasteries 366-7
inscription 3 17 Mongolia
Kushan art 371-95 Early 151-3
religion 328, 372 Hsien-pi 155, 156
Mathura lion capital 201, 206-7 Hsiung-nu 153, 156, 158, 159-62
Mathuran art mosaics 1 17
Buddhist sculptures 384-9 Moses of Khorene 477
feminine beauty 374-5 mother goddesses 380
Gandhiran influence 372-5, 388-9 murals, Kushan art 338-9
Maues 192, 193-4
Mauryas 89-90, 96-7, 117 Nagas 372, 379-80, 382, 393
Medes 37-8, 486 Nakhshab oasis 267
Media 37-8 Nana worship 322
Astyages, King of 39 Naqsh-i Rustam
Ecbatana, capital 48 Ardashir I 476, 478
immigration 35-7 Darius 1 49
language 47 inscription 24-5, 443, 479-80
Megasthenes 97 Shapur I 251, 478
Menander 101, 113, 114, 117, 126 Narseh 482
meridarchos 202 N i s i k inscription 410
Merv naturalism, Greek 128
aerial view 139 Nero, emperor 135
citadel 482 Nev-Sapur (Nishapur) 479
development 91, 481 -4 Nicaea 82
plan 140 Nicanor 78
Sasanians 481 Nisa 137, 138, 147, 149, 1J0, 283, 307
see also Margiana Niya 230, 435
metalwork Noin-Ula tombs 159-60
Achaemenid 56, 63-5 nomadic t o sedentary culture transition
Kushan 277-8, 350-1 44 1-4
Middle Persian/Middle Iranian 438, 404 nomads 19-20, 28-33, fi00-1 (map), fi04-fi
migrations (map), 508 (map)
Hsiung-nu 20, 457 after Alexander's invasion 457-72
Indo-Iranians 19 and agricultural land 182-3
Media 35-7 archaeological remains, Bactria 182-4
nomadic 457-8 armour 30
Mihro 321 burial grounds 26-9, 182, 457-8, 486
Milesians 405 cattle-breeding 23, 156, 157
Milindapaeha confederations 25-6
 Index

Early Mongolia 151-3 and China 187

equestrian 485-6 coins 138, 139-43, 181, 283
and Parthia 180-2 culture 147-50
religion 30-1 decline of 136, 473-5
see also individual groupings and Greeks 107
numismatics 99, 125-7; see also coins king and council 145
and nomads 180-2
Oanindo 328,433
religion 149
oases
and Rome 133-6, 473-5
economy 230-5
sedentary population 144
fortified 482-3
and Seleucids 131-3
and nomads 183
socio-economic fabric 136-7, 144-5
strategic importance 91
trade and commerce 137-9, 187
Oaxso 328, 433
Parthian language 47, 438
O l d Bactrian 399
Parthian prince, temple of Shami 148
O l d Iranian 47, 57-62, 409-1 0
Parthians, at Persepolis 59
O l d Persian inscriptions 47-52, 397
Pasargadae 38, 39, 55, 56
oral literature 167
Patala 85
Oilagno 323, 325
Pazirik 30, 32, 32
Orobatis 78
Pazirik, and Yueh-chih 177-8
O x u s River, in local cults 116, 118, 314,
~ e b b l egraves, Alagou (Turfan County)
407
213-18
Oxyartes 72-3
Pei-shih 458, 469
Padataditaka, city life 304 Peloponnesian war 46
Pahlavas (Indo-Parthians) 498-9 (map) Persepolis, reliefs 56, 57, 58, 59, 60, 61,
administration 199-202 62, 443
art 205-6 Persian Empire 486-7, 494-5 (map)
coins 204 Babylonia captured 43
in India 196-9 extent 40
religion 206-7 immigration 35-7
Paitava and Shotorak, Kushan art 359-60 rise of 38-43
Palamedes inscription 3 1 1 sedentary tribes 38
Pali texts, hereditary professions 310 social organization 38
Pamirs, excavations 183 Persian guardsman, at Persepolis 62
Pan Ch'ao 229, 237 Pescennius Niger 136
Pan Yung 229, 237-8 Pharro 325
Panjtar stone inscription 248 Philhellenism 107, 202-3, 205-6, 487;
paper 234 see also Greek influence;
carved 395 Hellenism
Parni 131, 144, 457 Philostratus 3 1, 197
Pars (Fars) 36, 475-6 Phraates I1 132, 181
Parsua 36 Phraates IV 142
Parthia 131, 487, 498-9 (map) Phrada (Alexandria Prophthasia) 70
administration 146-7 Pliny
agricultural 137 Cyrus 40
architecture 147 Kushans 257
and Buddhism 149-50 Plutarch, Taxiles 80
 Index
--

Polybius, Antiochus I11 99 V i ~ n aKadphises 3 18-2 1

Pompeius Trogus see also Buddhism; cults;
Asiani 173, 180 Zoroastrianism
Greeks 99 religious architecture 366-7
Scythians 180 reservoirs 270-1
portraiture rhytons 56, 63
Greek 118, 127 Nisa 124-5, 149, 150
Mathura 372-5 Rome
Porus 67, 79, 81 -2 and Kushans 256-8
postal service, Achaemenids 52-3 and Parthia 133-6, 473-5
Po-t'iao 250, 491-2 and Sasanian state 477-9
pottery, Kushan 277-8 and Seleucids 132
Prakrit 397, 399-400, 433-8 roof tile 245
pre-Achaemenid states 40-3 Roxane 72-3
protector-general 228-9 royal way 186-7
Pul-i Darunta, stone inscription 398 Runic inscription, possible 108

Quintus Curtius Saddharma-Puarika 437

Abisares 79 Sai 180
Alexandria ad Caucasum 105 Saka, religion 31, 206-7
armour of nomads 30 Saka language 239
Barsaentes 80 borrowed terms 41 6
Cyrus' war 24 Buddhist terminology 416
jewellery 370 in Gandhira and Panjab 410-11
Massaga 75-6 in Malwa 413, 414-16
Patala 85 in Mathura 411-12
Samaxus 80 Parthian influence 4 13
in Sistan 409-10, 413
railposts, sculpted 390-2, 394 Southern and Khotanese compared
reindeer stones 151, 152 413-16
religions Sakd T i g r a x a ~ d z44-5, 49, 5 6
Bactrian, pre-Kushan 313-1 5 Saka tribes 24-5, 31, 181, 182, 485-6, 488,
distribution 21 489
Greek 114-16 Saka-Massagetae 24, 28, 31, 33, 449-50,
Hellenized 127-8 456, 457
Hsiung-nu 164-5 Saka-Pahlava kingdom 194
Huvishka 327-8 Sakas 498-9 (map)
India 329 Achaemenids 443
interpretatio Graeca 3 14, 32 1 administration 199-202
Kanishka I 321-2 art and architecture 205-6
Kushans 308, 315-16, 328, 345 Chorasmia 443
Kushan yabghus 3 16- 18 classical authors on 193
O l d Iranian 57-62 ethnic history 20, 23, 24-5, 442-4, 488-9
Parthia 148-9 Greek influence 202, 205-6
Sakas 30-1, 206-7, 315-16 in India 192-6
syncretized 329 invading Bactria 409
triple kingship 323-7 and Kanishka 252
 Index

Khotanese 41 8-19, 421 encrustation style 350

and Kushans 420 G a n d h ~ r a359, 362-5
migrations 180 Kalchayan 339-40
monetary system 203-4 Kushan period 339-45
and Persians 44-5 Mathura 373-6, 380, 382
religion 31, 206-7, 315-16 Scylax of Caryanda, explorer 54
Sanskrit literature 192-3 Scythian bow 45, 163, 278
satraps 260 Scythian language 47
Southern, language and script Scythians 37, 44
409- 16 aiding Persians 43
warriors 443 burials 26, 27, 28, 32
see also Scythians culture types 29
~iilabhaiijikiifigures 372, 373, 374 Greek sources 24, 30, 31
Salar-Karasu-Dzhun irrigation system 267 social differentiations 28, 29
Samarkand oasis 266, 272; see also and Yiieh-chih 174
Maracanda see also Sakas; Saka Tigrauxad2
Samaxus 80 sea-routes 185
Sangaeus 73, 78 Seleucid king, head in clay and alabaster
Sanskrit 192-3, 433-8 121
Sargon I1 36 Seleucids 87-8, 487, 498-9 (map)
Sarmatians in Central Asia 87-98
culture 29, 33, 350 and Parthians 131-3
and K'ang-chii 463 and Rome 132
tribal group 465-7 urbanization 91 -5
Sasan 197, 198,-1.99 waning of power 95-7
Sasanians see also Seleucus; Antiochus
and Arsacids 481 Seleucus I 90-1
coronation fires 148 Seleucus I1 95, 131
kings 251, 255, 256, 475-7, 479 Seleucus Nicator 72, 87-8, 89
and Kushans 21, 255, 492 Shaikhan Dheri 108, 109, 292; see also
and Margiana 48 1-2 Peucelaotis
Naqsh-i Rustam 476, 478 Shalmaneser I11 36
oasis fortresses 482-3 Shamanism 164, 237, 461-2
and Rome 477-9 shun-yu 153, 154, 155, 158, 164, 165, 167,
Zoroastrianism 483, 492 177
Satibarzanes 69, 70 Shapur I
satrapies campaigns 477-9
Achaemenid 42, 46-7, 53 conquests 251, 255, 479-81
Central Asia 88-9 Shapur I1 256
Indo-Greek 200, 201 Shash-tepe 452
Kushan 257, 262, 263 Shih-chi 152, 458
Sakas 260-1 Bactria 176
Seleucids 90- 1 Hsiung-nu 165, 228
Yiieh-chill 185 Yueh-chih 172, 177
sculpture Sibi 84-5
Buddhist 259, 346, 362-4, 384-9 Sidaogou 222
devata 347, 348 silk 245-6
 Silk Route 43-6, 287-8, J02-3 (map) Aorsi 466
Bactria 484 Bactria 100, 180
Hsiung-nu 154-5 Bactrian kingdom invaded 173
international trade 138-9 Cyrus' war 24
Karakorum branch 192, 490-1 Egypt 186
origins 20, 21 Geography 74, 99
and royal way 187 Sind 102
silk trade 21, 231-2 strategos 202
Western Regions 231 Strato 1 102
Sino-Kharos~hicoins 235 Strato I1 102, 103, 126
Siraci 465-7 stupas 205, 338, 366-7
Sirkap (Taxila) 291-2, 293, 294 Sudracae, tribe 84, 85
Sirsukh 294 sun worship 165, 316
Sisicottus 67, 73, 74, 78, 82 sundials 123
Sistan sanctuary 184 Surkh Kotal
Sisupalgarh (ancient Kalinganagara) 295 Bactrian language 422, 424, 427,
~ i v a31 8-20, 329, 376, 427 429-32
sixty-year cycle 424 inscription 32 1, 427-32, 430
slavery Kushan art 352-3
Kushan 289 restoration 325
Parthia 144, 145 sanctuary 338
Wu-sun 460 statue of ruler 3 J 1
Sogdian language 47, 288-9, 290 temples 336, 338, 339
Sogdiana (Zerafshan valley) 438-40 unknown script 417, 418, 419-20
coins 188 Vima Kadphises 320, 321
ethnic history 442-4 Surya (sun god) 378, 382
system of months 313 ~ i t r d l a ~ k d 308-9
ra
Sophagasenus 95 Svastikapata 3 8 1
Soteira 92 Syr Darya (Oxus), irrigation 71, 267-8,
Spadinus, king of Aorsi 466 270, 441-4
spalagadama 194 SzO-ma Ch'ien
spaliriia 194 and Chang Ch'ien 228
Spalyrises 194 Ferghana 454, 455
Spenta ArmaitiS 315 Mongolia 152
Spitamenes 71-2, 91 Silk Route 287-8
~ r a v a s t i ,miracle of 363 Ta-hsia 183
~r~dharma~i;akaniddnasitra 323 Yiieh-chih 173, 174, 175, 176, 178
steppe tribes: see Iranian nomads; see also Shih-chi
Hsuing-nu; Sakas; Scythians;
Wu-sun; Yiieh-chih Tacitus 3 1
stirrups 163 Tagar culture 30
stone artefacts 1 17- 18, 223, 224, 225 Ta-hsia (Bactria) 171, 173, 176, 178, 492
stone-mill trays 221 Takht-i Bahi inscription 196, 248
stone-slab graves 151-2 Takht-i Sangin 116, 118, 121, 123
stone-warrior grave 222-4 Greek inscriptions 407
Strabo hunting plaque 347
Alexander the Great 86 O x u s 1 16
 Index

phalera 3 5 j T'icn Shan settlements 213, 218, 221 -2

portico 337 Tillya-tepe 183
Tamdi 464 art 188-9, 3 j 4
T'an-shih-huai 156 burials 350
Tarim basin oases 21 1, 227, 228, 230, excavations 183, 353-5
488 jewellery 347-8
Tashkent local art 349-50
coins 284, 464 time reckoning (Hsiung-nu) 165
early urban centre 452-3 tirthankara 381, 384
irrigation 266, 267 Tochari 20, 173, 180, 238-9
nomad culture 471 toilet box 234
see also Chach Toprak-kala 297-9
Tashkurgan, Xianbaobao graves 2 10-11 T'ou-man 153-4
Tasmol culture 30 town plans, Greek 202, 205
taxation trade
Achaemenid Empire 53 caravan 186-7
Parthia 144, 147 Egypt and India 185-6
Taxila (Sirkap) Kushan 285-7, 312
and Alexander 79-81 oases and steppes 445-6
Apollonius 197 sea-routes 185
captured 192 Sogdian inscriptions 439-40
excavations 184, 248, 291 -2 transmission of ideas 21
Indo-Parthian coins 197-8 Western Regions and China 234
jewellery 370 trade routes 124-5
monasteries 367 Achaemenid Empire 54-5
origins 108 and Aorsi 466
pillar edict 398 Karakorum mountains 192, 490-1
rebuilt 206 see also Silk Route
silver-scroll inscription 248 Trajan 135
terracotta figurines 368 Transoxania, history 72-3, 457-8
Taxiles 73, 79 tribal federations 153, 442, 457-8, 485-6
Ta-yiian Triparadisus 88
nomads 463, 468; see also Ferghana triple kingship, religious life 323-7
Tenlik burials 462 Tso-chuan 172-3
Tepe Maranjan 360 Tulkhar burial ground 182, 471
Tepe Sardar 360 tumulus 222; see also graves; kurgans
Termez Tung H u 154
Buddha sculpture 346 t'un-t'ien (military agricultural colony)
fragment from capital 334 240-3
stupa 338 Tiira 23
terraced agriculture 270-1 Turfan basin oases 227, 228, 242
tcrracotta figurines 119, 120, 368-9 Tiirk dynasties 21
terracotta statuettes Tuva barrow 29, 32
Kushan 345 Tu-yin site 241
Margiana 483
Parthia 147-8
textiles, Parthia 137
 Index

inscribed ~ o t s h e r d408 metals 232-3

Ujjain (Ujjayini) 192, 319-20, 413, 414-16 fortifications 242-3
Umbhaydbhisarika, city life 304 Silk Route 243-6
unknown language and script 417-21 textiles 230-1
Ura-tyube 43, 450, 455 wine presses 123
Urartian building techniques 56 women
Urartian writing 48 Hsiung-nu 164
urbanization 108- 10, 184, 33 1-6, 442, Saka-Massagetae 28
446-51, 488, 490 wooden chamber graves 27, 32, 218-21
Usrushana 442, 445-6, 453 wooden tablets, H a n colony 241
woollen cloth 213, 214, 231
Vakhsh valley irrigation 267 writing systems 165-7, 308, 491
Vardanes 135 Wu-sun 458-62
Vasishka 252-4, 323, 326, 432 archaeological sites 461 -2
Vaskushana 253 graves 237
Visudeva 117, 250, 252, 253, 254, 258, Hsiung-nu 221
377 jewellery 462
Vayu 316 origins 175
VaxSu 3 14, 407; see also O x u s Yiieh-chih 488
Vima Kadphises Wu-ti, emperor
coins 248, 280, 490 Chang Ch'ien 171, 245
Great Kushans 21 Ferghanese horses 454
religion 318-21, 322, 323 H a n emperor 155
rock inscriptions 422-4, 427-9 against Hsiung-nu 228
Saka satraps 260 Wang H u i 177
statue 21iJ
vineyards 274-5, 276, 455 Xerxes, inscriptions 49
Vishnu cult 329, 375 Xianbaobao graves 210-1 1
Vohu Manah 324 Xinjiang region 209-10
Vologases I 135, 140, 149 Alagou graves 2 13-2 1
Vologases I1 473 Altai mountains 222-5
Vologases I11 136, 473-4 Lou-Ian graves 2 1 1- 12
Vologases IV 136, 474 T'ien Shan settlements 221 -2
Vologases V 136, 143, 474 Xianbaobao graves 210-1 1
Vonones 134, 194 see also East Turkestan
vowel length indications 435-6
yabghus (governor-generalships) 171, 185,
Walled City-States 227-8, 229, 239, 244, 247, 316-18, 489
447 yaksas, figurines 372, 373, 374
Wang Mang 228, 229 yaksi cults 329
Western Regions Yiska, testimony of 401, 404
agriculture 242 YaJt 23
China 229-30, 232-5 Ycn-kao-chen: see Vima Kadphises
Chinese administration 238-40 Yen-ts'ai 465-7
Chinese influence 236-8 Yona Kamboja 401
colonies 240-3 Yiieh-chi11 JOO-1 (map)
culture and religion 236-8 and Bactria 179-80, 488-9
 Index

in Gansu 171-4 Zerafshan valley (Sogdiana) 438-40

Mao-Tun 154 irrigation 265, 266, 271-2
migration 175-6 nomadic gravcs 182
and Pazirik 177-8 Zoroastrianism
and Scythians 174 and Huvishka 327-8
see also Tochari Kushan period 308, 313, 352-3,
Yiieh-pan 469 491
yurt 32, 164 origins 42, 57-9, 62
Parthia 149
Sakas 31, 207