(Routledge Indian Ocean Series) Ruth Barnes (Editor), David Parkin (Editor) - Ships and The Development of Maritime Technology On The Indian Ocean-Routledge (2002)

S h ips a n d the D evelopm ent
of M a r it im e T echnology
in the I n d ia n O cean
I n d ia n O cean S e r ie s
Editors: David Parkin & Ruth Barnes, University o f O xford
T h ere is a need to understand the Indian O cean area as a

cultural com plex which should be analysed beyond the
geographical divisions o f A frica, the M iddle East, the Indian
subcontinent, and South-East Asia, as its coastal populations
have interm ingled constantly. T h e movement o f people, goods
and technology make it imperative that spatial concepts and the
role o f material culture be central in the study o f the region by
archaeologists, historians, ethnographers and anthropologists.
I slam ic P rayer A cross the I n d ia n O cean
E dited by David Parkin and Stephen C . Headley

S h ips and t h e D evelopment
of M a ritim e T echnology
in t h e I nd ia n O cean
E DITED B Y
D avid Parkin and Ruth Barnes
O Routledge
Taylor & Francis Croup
LONDON AND NEW YORK
First Published in 2002
by RoutledgeCurzon
Published 2016 by Routledge

2 Park Square, M ilton Park, Abingdon, O xon 0 X 1 4 4 R N
Simultaneously published in the USA and C anada by Routledge

711 Third Avenue, N ew York, N Y , 10017, USA
Routledge is an imprint o f the Taylor & Francis Group, an informa business
(' 2002 David Parkin and Ruth Barnes
All rights reserved. N o part o f this book may be reprinted or reproduced or

utilised in any form o r by any electronic, mechanical, or other means, now
known or hereafter invented, including photocopying and recording, o r in any
inform ation storage or retrieval system, without perm ission in writing from the
publishers.
British Library Cataloguing in Publication Data

A catalogue record o f this book is available from the British Library
Library o f Congress Cataloging in Publication Data

A catalog record for this book has been requested
ISBN 13: 978-0-700-71235-9 (hbk)

ISBN 13: 978-1-138-98184-3 (pbk)
C ontents
List o f Figures Vll

Contributors ix
Preface — David Parkin xv
1 S h i p p i n g i n the I n d i a n O c e a n : A n O v e r v ie w
Himanshu Prabha Ray I
2 S e a fa r in g s, S h ip s a n d S h ip O w n e r s : In d ia a n d the
I n d i a n O c e a n (a d 700~ l 5 ° 0 )
Ranabir Chakravarti 28
3 S e a f a r i n g i n P e n i n s u l a r I n d ia i n the A n c i e n t P e r i o d :
O f W a te r c r a ft a n d M a r it i m e C o m m u n i t i e s
Himanshu Prabha Ray 62
4 A r c h a e o l o g i c a l E v id e n c e o n S h i p p i n g C o m m u n i t i e s
o f S ri Lanka
Osmund Bopeararchchi 92
5 P r e - M o d e r n S r i L a n k a n S h ip s
Somasiri Devendra 12 8
6 C l a s s i f y i n g V e s s e l- T y p e s in I b n B a tt u ta ’ s Rihla
Dionisius Albertus Agius I 74
7 N a v ig a tio n a l M e th o d s in th e In d ia n O c e a n
Abdul Sheriff 20 9
8 T h e Rise a n d Fall o f H a d h r a m i S h i p p i n g in the

I n d i a n O c e a n , c. 17 5 ° — c - J9 4 °
William Gervase Clarence-Smith 22 7
9 C o n t i n u i t y a n d A d a p t a t i o n by C o n t e m p o r a r y S w ahili
B oatbu ild ers in K en y a
Susan Beckerleg 259
v
VI C ontents
10 T h e H i s t o r i c a l C o n t e x t o f the C o n s t r u c t i o n o f the Vattai

F is h in g B o a t a n d R e la te d F r a m e - F i r s t Vessels o f T a m il
N ad u and Beyond
Lucy Blue 278
11 Y a m i B oats a n d B o a t B u i l d i n g i n a W i d e r P erspectiv e
R.H. Barnes 291
Bibliography 3 15
Index 347
L ist of F igures
2.1 T r a p y a g a vessel o n an i n s c r i b e d seal 38

2.2 A n i n s c r i b e d seal s h o w in g a ship w ith an o a r 38
2.3 S h ip o n a te rr a c o tta seal 4°
2.4 S h ip o n a te rr a c o tta seal sh o w in g a stylized stalk o f
g r a i n i n a basket a n d a s te e r in g m e c h a n i s m fitte d
to the ste rn 4°
2.5 A sto n e sc u lp tu r e sh o w in g a b r o a d vessel w ith a
sharp ste rn a n d a s t e r n - p o s t r u d d e r 44
3.1 S titc h e d craft o n the A n d h r a coast 65
3.2 C lo s e - u p o f th e stitc h in g 65
3.3 B o a t r e p a i r o n O r is s a coast 68
3.4 C l a y sealings f r o m C h a n d r a k e t u g a r h 77
3.5 M e m o r i a l sto n e , G o a M u s e u m 77
3.6 S h o e - d h o n i a lo n g A n d h r a coast 88
3.7 C o n s t r u c t io n o f fis h in g c ra ft 88
4.1 A n c ie n t p o rts o f S ri L an k a 96
4.2 T r a d e C e n t r e s a n d a n c i e n t p o r ts i n S o u t h I n d ia 97
5.1 S h ip b u i l d i n g t e c h n o l o g y z o n e s 145
5.2 E v o lu tio n o f con stru ction al techn iqu es 150
5.3 T y p e s o f w ater t r a n s p o r t 1 52
5.4 Jaffna T h o n i 154
5.5 J a f f n a T h o n i , views o f b o w f r o m m id s h ip s a n d views
o f ca r g o stowage a r r a n g e m e n t s 157
5.6 Y athra D h o n i M o d e l fro m D od and u w a 158
5.7 Y athra D h o n i in p o rt 160
5.8 Y a t h r a D h o n i S tr u c tu r a l details 1 62
5.9 B attal 165
5.10 V attai/V atta 1 67
6.1 Ib n B a ttu ta ’ s voyage 1 77
vii
Vlll L ist o f Fig u r e s
6.2 V e sse l-ty p e s in Ib n B a ttu ta ’ s Rihla 178

7 -i T h e kamal a n d th e c r o s s - s ta ff 218
9 -i F u n d i u s in g th e b o w d r ill 266
9*2 S izi p o ses w h ile F u n d i w orks 268
9-3 F u n d i’ s assistant awaits in s tr u c tio n s 270
9-4 F u n d i p r e p a r e s to fit th e fin a l p la n k 272
IO. I C o a sta l sites in s o u th e r n T a m il N a d u 280
10.2 A thoni at a fo r e s h o r e b u ild in g site at T u t ic o r in 282
10.3 A vattai o n th e fo r e s h o r e at A d ir a m p a ttin a m 282
10.4 R e c tifie d p la n s o f a vattai r e c o r d e d o n the
E r ip u r a k a r a i fo r e s h o r e , J a n u a ry 1997 283
10.5 D ia g ra m to illu s tra te th e d e sig n o f vattai fra m e s
u s in g a scrieve b o a r d a n d a sin g le m o u ld 286
11.1 S m a ll tatara 293
1 1.2 A te n -m a n boat 294
n -3 T a p p in g th e b a r o k b e tw e e n th e seam pegs 296
11.4 A la rg e b o a t h a lf a ssem b led 297
1 1-5 I n te r io r view 298
00
cr
jo
11.6 V ie w fr o m ab ove o f th e m e th o d o f a tta ch in g th e
11.7 T h e h u ll is p r o p p e d u p w ith rib s 30 0
11.8 G iv in g th e fin a l to u c h e s to th e h u ll 30 0
11 -9 P r o je c tin g lu gs in a p a rtly fin is h e d tatara 301
11.10 I n te r io r , sh o w in g a r r a n g e m e n t o f o a rs, la te ra l
c r o s s - m e m b e r s , a n d rib s 301
11.11 H a lf s e c tio n (tran sv erse) o f th e b o a t f r o m B o te l
Tobago 30 2
11.12 A r r a n g e m e n ts o f th e p la n k s 304
11.13 T h e p la n k p a tte r n o f L a m a le ra b o ats 305
11.14 I n te r n a l c o m p r e s s io n str u c tu re o f th e h u ll o f a
L a m a le ra b o a t 306
11.15 R o u g h s h a p in g a tree 30 8
11.16 S h a p in g th e p r o w 30 9
11.17 W o r k -s h e d u se d in b u ild in g a la rg e b o a t 310
11.18 T h e p a ta k a ra n b e in g j o i n e d 311
C o n tr ib u to r s
Dionisius Albertus Agius is a reader in Arabic and the

Medieval M editerranean at the University o f Leeds. He was a
m em ber o f the Jesuit O rd er in Malta and Lebanon, studied
Classical A rabic at the Universite St-Joseph in Beirut and the
Pontificio Istituto di Studi A rabi in Rom e. He read A rabic,
Turkish and Farsi at the University o f T oron to, where he
received his PhD . His research interests include A rabic
language and linguistics; semantics o f the material culture;
travel literature o f Medieval Islam; the Medieval M editerranean
and Indian O cean . His current research is 'T raditional
seafaring in the Arabian G u lf and O m an: historical and
linguistic links’ . A m on g his books is the forthcom ing In the Wake
o f the Dhow: The Arabian G u lf and Oman (G arn et-Ithaca); author o f
Arabic Literary Works as a Source o f Documentation fo r Technical Terms o f the
Material Culture (1984); Siculo Arabic (1996); editor (with Ian R
Netton) o f Across the Mediterranean Frontiers: Trade, Politics and Religion
6 5 0 — 1 4 5 0 ( l 997 ) and A l-M asaq: Islam and the M edieval Mediterranean,
Volum e 13 (2 0 0 l).
Robert Barnes is a Faculty Fellow o f St. A n ton y’ s C ollege and

Professor o f Social A nthropology at the University o f O xford.
His books include Kedang: a study o f the collective thought o f an eastern
Indonesian people ( l 974 )> Indigenous Peoples o f Asia ( l 995 )>
Andrew Gray and Benedict Kingsbury, and Sea Hunters o f
Indonesia: Fishers and Weavers o f Lamalera (1996).
Ruth Barnes has carried out extensive field research in eastern

Indonesia and is the author o f The Ikat Textiles o f Lamalera. A Study o f
an Eastern Indonesian Weaving Tradition (Leiden, E.J. B rill 1989)- She
also has written num erous articles on textiles and related art
ix
X C o n tr ibu to r s
form s. Together with Joanne B. Eicher, she edited Dress and

Gender: M aking and Meaning (J993 )- She retains a particular
interest in Indonesia and continues to carry out research in the
region, although for the last ten years she has also focused on
issues o f Indian O cean trade and cultural contacts, in
particular for the pre-E urop ean era; as part o f this investiga
tion she has recently published Indian Block-Printed Textiles in Egypt.
The Newberry Collection in the Ashmolean Museum (O xford: C larendon
Press). She is now cataloguing the Ashm olean M useum ’ s
second part o f the Newberry C ollection , which is made up o f
Islamic em broideries.
Susan Beckerleg is a Social A nthropologist who has been

carrying out research on the Kenya coast since the m id-1980s.
Apart from the local boatbuilding industry, she has a wide
range o f research interests, inclu ding Swahili m edicine,
ethnicity and the effects o f international mass tourism on
Swahili com m unities. Recently she has set up the O m ari
Project, an N G O to assist h eroin users living in Coastal Kenya.
She works as an international developm ent consultant, mostly
on health projects in A frica and Asia.
Lucy Blue com pleted her D .P h il. at the Institute o f A rch ae
ology, O x fo rd , in M aritim e Archaeology (1996), specialising in
Bronze Age harbours and trade in the eastern M editerranean.
Shortly afterwards she had the opportunity to becom e involved
in a project with her form er supervisor Prof. Sean M cG rail
addressing ethnographic boats in South Asia. She is currently a
researcher/lecturer at Southam pton University, C entre o f
M aritim e Archaeology, D epartm ent o f Archaeology, where
she teaches Pre-Classical Seafaring to undergraduates and
Masters students, as well as m aritime ethnography. She is co
director o f the excavations at Q useir al-Q ad im , a Rom an and
Islamic harbour site involved with Indian O cean trade on the
Red Sea coast o f Egypt, a University o f Southam pton project.
Osmund Bopeararchchi specialises in Central Asian, Indian

and Sri Lankan numismatics and history, and is D irector o f
Research at the C entre National de Recherches Scientifiques
C o n tr ibu to r s XI
(G .N .R .S .), Paris. He is also, since I9 9 4 > the director o f the

French M ission o f Archaeological C o -o p eratio n in Sri Lanka,
as well as professor (habilite) o f P aris-Sorbonne University (Paris
IV). He holds a B .A . from the University o f Kalaniya (Sri
Lanka), and a Licence, Maitrise, D .E .A ., and doctorate from
the Paris Sorbonne University (Paris I). His publications
include six books and m ore than one hundred articles in
international journals, for which he has received two awards
from the French Academy o f Inscriptions and Belles Lettres’
and two silver medals from the G reek Numismatic Society
(Athens) and from the Indian C o in Society (Nagpur).
Ranabir Chakravarti, with M .A. and Ph.D . degrees from the

University o f Calcutta in A ncient Indian History and C ulture,
specialises in the social and econom ic history o f early India (up
to c. AD 1300), with a particular interest in the history o f trade
in early India. K now n for his research on the maritime trade o f
India in the Indian O cean befo re the com ing o f the
Portuguese, he is a regular contributor to several international
journ als, having published 75 research articles and the
following books: Warfare fo r Wealth: Early Indian Perspectives, Calcutta
(1986); A n Economic History o f Early India (in Bengali), Calcutta
(1991); edited (jointly with K um al Chakrabarti and A m it
Bandyopadhyay) a volume o f essays in Memory o f Professor Ashim Dasgupta
(in Bengali), Calcutta 2000; Associate Editor, Sourcebook o f
Indian Civilization, Calcutta 2000; Editor, Trade in Early India, New
D elhi 2001; Explorations in Trade and Society in Early India, New D elhi
2001. He is Professor in the Departm ent o f A ncient Indian
History and C ulture, University o f Calcutta, Calcutta, India.
Professor William Gervase Clarence-Smith teaches the

history o f the tropical world at the School o f O rien tal and
A frican Studies, University o f London. He is author o f The third
Portuguese empire, l 825~^975 (1985) ’ edited The economics o f the Indian
Ocean slave trade in the nineteenth century (1989), and co-edited
Hadhrami tranders, scholars and statesmen in the Indian O cean, l j $ 0 s to
i g 60 s (1997). He has also published widely on trading patterns
in the Atlantic, and on tropical agricultural com m odities.
xii C o n tr ibu to r s
Somasiri Devendra, Sri Lankan. B .A . (Ceylon) 1955 A fter

graduating taught at a m ajor secondary school (l9 5 5 —6 o),
com m issioned as Instructor O ffice r in the Royal C eylon Navy
(subsequently Sri Lanka Navy) retiring as Com m andant, Naval
& M aritim e Academy, Trincom alee in rank o f Lieutenant
C om m ander. D irector o f Som erville &. C o .L td . (Tea, Produce
and Stockbrokers) 1976—86 and associate companies and was
C hairm an C olom bo Brokers’ Association and a Founding
D irector o f the C olom bo Stock Exchange. Pioneered first
attempts at m aritime archaeology in the island and was invited
by the governm ent to conduct an (ongoing) rescue archaeology
project into Galle harbour, a historic port which is undergoing
developm ent. Now in retirem ent, acting as voluntary advisor to
the Dept, o f Archaeology and Ministry o f Cultural Affairs on
m aritime historical studies and new legislation on regulation o f
the practice o f m aritime archaeology; A uth or o f 'A History o f
the Navy in Sri Lanka: V o l.l - T h e Ceylon Royal Naval
V olunteer Reserve ( l 937 —*95 °)* and 'The Sri Lanka Navy: A
Pictorial Survey, I 937 - 1 998 * C o -ed ited 'M aritim e A rch aeol
ogy in Sri Lanka: T h e Galle H arbour Project 1992 * (with
G reen, J.) and the 'G alle H arbour Project 1996—97: Report
for the Sri Lanka D e p t.o f A rchaeology’ (with G reen, J. and
Parthesius, R.)
Series editor David Parkin is Professor o f Social A nthropology

and D irector o f the Institute o f Social and C ultural A n th ro
pology at the University o f O xford. He has carried out field
research in East A frica since 1962, much o f it while at the
School o f O rien tal and A frican Studies, L ondon University.
C u rren t interests include Islam, medical anthropology, socio
material prosthesis, and cross-cultural rhetorics. Recent p u b
lications include Sacred Void: spatial images o f work and ritual among the
Giriama o f K erya, Cam bridge, 1991, and A utoritf et Pouvoir C hez Les
Swahili (co-edited), Paris, 199^, the latter resulting from o n
going collaboration with French colleagues.
Himanshu Prabha Ray is Associate Professor in the C entre for

H istorical Studies at Jawaharlal N ehru University, New D elhi.
She has studied Sanskrit (Panjab University) and A rchaeology
C o n tr ibu to r s xiii
(University o f Cam bridge) and has a doctoral degree on Early

H istoric Trade in the Western Deccan from the Jawaharlal
N eh ru U niversity. H er publications include two books:
M onastery and Guild: Com m erce under the Satavahanas, O x fo rd
University Press, New Delhi, 1986; and The Winds o f Change:
Buddhism and the Maritime Links o f Early South A sia , O xford University
Press, New D elhi, 1994 (reissued in the O x fo rd India
Paperback series, 1998). In addition, she has edited a volum e
titled: Archaeology o f Seafaring: The Indian Ocean in the Ancient P eriod,
Indian C ou n cil o f H istorical Research M onograph Series I,
New D elhi, 1999 and another with J .-F . Salles, Tradition and
Archaeology: Early Maritime Contacts in the Indian Ocean, New Delhi:
M anohar, 1996-
She has participated in several workshops and seminars
including the C onferen ce on Seafaring Com m unities in the
Indian O cean (Lyon, 1996); the C onferen ce on New D irec
tions in M aritim e History (Fremantle, 1993 ); the U N E S C O
Sem inar on the Silk Roads (Madras, 1990); and the British
Museum C olloqu ium on the Indian O cean in Antiquity
(1988). She consulted for the National Institute for Science,
Technology and Developm ent Studies (N ISTAD S) on 'In d i
genous Traditions o f Indian Navigation’ from 1992—1994 and
was the jo in t-co o rd in a to r o f the international seminar on
Techno-archaeological Perspectives o f Seafaring in the Indian
O cean hosted by N ISTAD S in 1994 - M aritim e History and
A rchaeology o f the Indian O cean continue to be part o f her
research agenda.
Abdul Sheriff is Principal C urator o f Zanzibar Museums and

was Professor o f History at the University o f Dar es Salaam
where he taught since 1969 and has served as Head o f the
Departm ent o f History. He was born in Zanzibar but gained his
B A and M A at U C L A before going on to do his PhD in A frican
History at the School o f O rien tal and A frican Studies at the
University o f L ondon . He is author o f Slaves, Spices & Ivory in
Zanzibar. London: James Currey, (1987), editor o f The H isto y and
Conservation o f 4Jinzibar Stone Town (1995) and co-ed itor with Ed
Ferguson o f Zanzibar under Colonial Rule (1991).
This page intentionally left blank
P reface
Were early water-craft able to travel longer distances across the

Indian O cean than has often been thought? W hile some
answers to this question are inevitably speculative, scholarly
findings from archaeology, anthropology, history and p re
history o f peoples and objects are beginning to provide a m ore
acceptable picture, as suggested by some o f the chapters in this
volum e. These findings have implications for our u n d er
standing o f the kinds o f techniques used in ship construction
and their evolution, and o f the varying com m unities involved
in building watercraft and trading across oceans in them. As we
have indicated in the first volume (on Islamic prayer) in this
series on the transregional nature o f Indian O cean societies
and cultures, socio-geographical boundaries can be drawn, if at
all, only in relation to specific spheres o f activities. The
concerns o f religion, shipping technology and comm unities,
textile production and distribution, the provision o f herbs,
spices and medicines, architectural styles and textual scholar
ship and its dissemination constitute the separate themes but
are each associated with different if overlapping territories and
provenances in the region.
There is a paradox in the idea o f an ocean as being at the
basis o f a cultural complex o f resemblances and counterpoised
differences. Its distinctiveness as an entity may be claimed by
those with an interest in doing so, yet is rarely complete in any
one peop le’ s view: what some see as its socio-cultural horizons
are what others may see as its centre, and so on in the m anner
o f a latticework o f contacts and influences. B raudel’ s m on u
mental work, The Mediterranean and the Mediterranean World in the Age o f
Philip II (1972) has now been followed by another work o f
xv
xvi Preface
considerable scholarship and im portance for that region, The

Corrupting Sea: a Study o f Mediterranean History by H orden and Purcell
(^ O O O ), which addresses the question o f unity and continuity
incurred through, rather than in the face of, the socio-cultural
and m icroecological diversities making up the M editerranean
over 30 00 years, and the doubtful fate o f this unity as a result
o f radical changes brought about during the twentieth century.
T h e Indian O cean rim , by analogy with that o f the Pacific
O cean, might be thought o f as too widespread for there to be
sufficient o n -go in g co -o rd in atio n o f maritime activities to
constitute significant unity. In fact, what is remarkable are
indeed the lon g-term key relationships and seaways that, criss
crossing each other at certain points, do give a density to
Indian O cean interaction, so m uch so that it now seems
legitimate to refer to this ocean at least as a so cio-cultural area
o f com parable significance to lan d-based areas o f high
interactional density in the world. T h e popular term, liquid
continent, has come into vogue as an expression o f this general
view, although it must be admitted that it is the differences
between oceans as socio-cultural areas, rather than their
assumed similarities in relation to land-based regions, that
should be the object o f analysis.
For purposes o f convenience it may be appropriate to think
o f the Hadhram aut and Persian G ulf, the East A frican coast
and nearby islands, western/south-western India and Sri
Lanka, and Indonesia as the socio-cultural co-ordinates o f
the trade, population movements and transmission o f m aritim e
technological knowledge with which this present volum e is
concerned and which have historically inform ed the develop
ment o f the recurrent activities by which we identify the Indian
O cean as an area amenable to study as an entity. W ithin these
co-ordinates individual chapters provide particular cases o f
boat- or ship-building technology and o f the com m unities
whose settlement or trading activities are inextricably linked to
such knowledge.
T h e editors especially wish to thank H im anshu Ray, an
authority on the subject and indeed on a num ber o f Indian
O cean phenom ena, for writing the introduction in addition to
Preface XVI1
her own chapter in the volume itself. T h e other contributors

deserve the editors’ gratitude for their graceful acceptance o f
our editorial suggestions and the speed with which they
responded to them, special gratitude is due to D r. Dionysius
Agius for his help. T he chapters for this current volume
originate in the second conference in the series The Indian Ocean:
Transregional creation o f societies and cultures' organised by the Institute
o f Social and Cultural A nthropology at the University o f
O xford in May 199^- It was convened by the editors o f this
volum e, and was supported financially by the British C ouncil,
the British Academy, the Asian Studies C entre, St A n ton y’ s
College (director D r Steve Tsang) and the ESRC Transnational
C om m unities Programm e (based at ISC A) (director D r Steve
Vertovec) to whom we are very grateful. T he conference was
entitled 'Ships and the Developm ent o f M aritim e Technology
across the Indian O cean’ . Thanks are due to the staff o f IS C A
who assisted with the conference and the subsequent presenta
tion o f this volum e, especially to G ina Burrows, without whose
energy, enthusiasm and expertise, this publication would not
have materialised. We also want to thank Paul Agius, Isabella
Birkin, Valerie C oup lan d, Haas Ezzet, Vicky K night, D o m in
ique Lussier, Mike M orris, Russell Sharman and A nna W inton.
T he conference organisation and the principal oversight o f the
editing process have been the responsibility o f G ina Burrows
and thanks are due to colleagues and students, both those who
have contributed chapters and those, like D r Paul Dresch and
Zulfikar H irji, who have provided their own expertise towards
this publication. We thank our chairpersons, including Dr
Mark H orton, D r Janice Stagardt, and Jean-C laude Penrad
who provided a stimulating paper which was not, unfortunately,
available for this publication.
David Parkin
I
S h ip p in g in t h e In d ian
oc e a n : A n O ve rv ie w
Himanshu Prabha Ray
This introduction surveys the literature on shipping in the

Indian O cean and attempts to highlight the m ajor concerns
that have prevailed. A distinction may perhaps be made between
shipping and maritime trade, though the dem arcation is no
doubt fine and often the lines get b lu rre d , as sometimes there
is evidence for the nakhoda or master m ariner transporting his
own com m odities for trade. T he theme may be studied from
two perspectives, the two not being mutually exclusive: one, the
com m unities involved in the construction and sailing o f water
craft; and second, the technology o f the craft itself. W hile the
second approach has had a relatively larger following, the
literature on b oat-building and sailing com m unities continues
to be inadequate. A n aspect that has seldom been dealt with
earlier and one that this volume addresses is shipping in the
creation o f trans-regional comm unities and cultures. It is here
that the contributions in this volum e diverge from the trodden
path.
Literatu re & A r t
A pioneering work on Shipping in the Indian subcontinent

that is yet to be replaced is Radha K um ud M ookerji’ s study
(1912). A comprehensive survey o f textual and archaeological
I
2 H im a n sh u Prabha Ray
sources, the work encompasses a time span truly m onum ental —

extending from the protohistoric Harappan period to the early
twentieth century. T he author, a Professor o f History and a
M em ber o f Parliament, was, however, not unmoved by the
rhetoric o f pre-Independence India. This is evident from the
chapters dealing with the period o f 'H indu Im perialism ’ , as
also the objective o f the work which was to establish:
the central position o f India in the O rien t w orld, for

w ell-nigh two thousand years, not merely in a social, a
m oral, a spiritual, or an artistic reference, but also and
equally in respect o f colonizing and m aritime activity, and
o f com m ercial and m anufacturing interests (ibid.: xi).
In spite o f these obvious drawbacks, there is much in the work

that is com m endable. It is indeed deplorable that in the post
Independence discourse the work should have been m argin
alised under the rubric o f 'N ationalist’ historiography and little
attention paid to ship-building and m aritime com m unities in
historical writing.
Perhaps the first contribution o f M ookerji’ s study was to
provide a comprehensive com pilation o f textual references to
seafaring, including the use o f the Sanskrit work, Yuktikalpataru,
a treatise on ship-building attributed to king Bhoja and dated
to the eleventh century. T h e work refers to a variety o f woods
used in sh ip-building and classifies w ater-craft into two m ain
categories: ordinary and sea-going; and provides detailed
measurements for these. But perhaps the characteristic feature
o f water craft that it emphasises is the stitched tradition and the
absence o f iron in holding or jo in in g the planks together. This
feature o f Indian O cean vessels seems to have persisted almost
into the present as evident from notices by travellers to the
region.
M ookerji’ s com pilation has been reworked by subsequent
scholars. W hile Ghaudhuri (1976) based his paper on the
Yuktikalpataru, Schlingloff (1982) elaborated on the textual
sources and compared these with representations in art. It is
evident that none o f the textual sources on ship-building are
adequately detailed to allow for the reconstruction o f water
S h ip p in g in t h e In d ia n O cean: A n O v e r v iew 3
craft based on these. It is nevertheless significant that in the

eleventh century, these instructions should have been in co r
porated in a work on state-craft and attributed to the
authorship o f a king. In the wider context o f South and
Southeast Asian M aritim e History, it is from the ninth-tenth
centuries onwards that there are increasing references in
inscriptions to fishing rights; duties levied on com m odities
brought through the water-routes; and to revenue being
obtained from taxes on fishing. This is also a period o f
expanding maritime networks in the Indian O cean.
A second source used to advantage by M ookerji is the
evidence from sculptures, paintings and coins. In keeping with
his hypothesis o f H indu Imperialism, he argued that the vessels
represented on the B uddhist m onum ent at B o ro b u d u r
belonged to the Hindu-Javanese tradition — a contention that
has been disputed by M anguin (1980). But one aspect that
M ookerji missed was the active encouragem ent given to
seafaring activity by Buddhism at this time. M ore recent
analysis o f the iconographic data from the Indian subcontinent
has been undertaken by Deloche (1996) who also includes
m em orial stones within his purview. D eloche further expands
the scope o f this study by attempting an identification o f these
representations o f water craft on the basis o f ethnographic
parallels.
A noth er theme dealt with by M ookerji was the prom inent
place o f Bengal in ancient maritime trade. He based his
discussion on the sixteenth century collection o f poems in
Bengali known as Padma Purana or Manasamangala and composed by
m ore than fifty authors in celebration o f the glories o f the
goddess Gandi. Many o f these contained descriptions o f sea
voyages and are an addition to an earlier text on the same
theme, the graphic account o f the sea voyage o f Chand
Saudagar by the thirteenth century poet Narayandeva. The
depictions o f water-craft on the terracotta temples o f Bengal
escaped his notice, though these were later docum ented by
writers such as Dasgupta (19^3) and D eloche (1991).
A total o f nearly sixty-two panels o f the temples o f West
Bengal dated to the eighteenth and nineteenth centuries show
4 H im a n s h u Prabha Ray
realistic depictions o f river craft, while larger vessels form the

subject o f fifty-three panels (Deloche I 99 I: 2 ). These are in the
districts o f Bankura, Burdwan, Birbhum , H ooghly, Howrah,
M idnapur and 2 4 _Parganas- T h e river-craft represented are
rowing and racing skiffs, travelling houseboats or cerem onial
barges, while the m ore com m on types such as rafts, dug-outs,
cargo-carriers and fishing boats have been ignored. This
association o f ritual m aritime activity with the temple already
has a precedent in the eleventh century Jagannatha temple at
Puri where a cerem onial barge is prom inently depicted.
This emphasis on the ritual and cerem onial use o f the boat
in iconographic representations from Bengal and Orissa is in
marked contrast to contem porary evidence from the west coast,
especially Gujarat and the K onkan. H ere realistic representa
tions o f the ship occur on m em orial stones dated from the
eleventh to fifteenth centuries AD erected in h onour o f those
killed in sea battles. These depictions o f w ater-craft emphasise
structural details o f sea-going craft and focus on fighting scenes
(Deloche 1996). It is then evident that even near contem porary
societies in the Indian subcontinen t em phasised alm ost
antagonistic aspects o f seafaring activity. This is an aspect that
needs to be taken into consideration in the study o f ico n o
graphic depictions from other parts o f the Indian O cean as
well, e.g. those from east A frica (Garlake 1964) and parts o f the
Islamic world (N icolle 1989)*
In O m an maritime graffiti are to be found on the walls o f
fortresses, merchant houses and cliff faces and date from the
m id-seventeenth century to early in the present century
(Vosmer 1995 : I 3 )* A range o f vessels have been depicted
from purely Arab craft to hybrids o f Arab, Indian and
European design to A frican and pure European vessels. A ll o f
these representations depict vessels in profile and this seems to
be the preferred mode in most cultures. Perhaps an exception
to this are the rock art illustrations at Jabal Al-Jussasiyah in
north-east Qatar dating from as early as the tenth century
which show vessels on plan (Facey 1987)-
Th e association o f the site with pearl fisheries is significant
since the Bay around Jabal al-Jussasiyah form s the most copious
Sh ipp in g in th e In d ia n O cean: A n O ve r v iew 5
pearl banks in the region. In about I9 ° 0 , the Q atar pearl fleet

num bered over 800 boats m anned by 13 ,0 0 0 m en and there
are indications for the presence o f a large concentration o f the
pearling industry in the imm ediate vicinity o f Jabal al-
Jussasiyah during the seventeenth and eighteenth centuries.
T he evidence o f the boat drawings would suggest that the
exploitation o f pearl fisheries in the region dates to an earlier
period and Jabal al-Jussasiyah has been described as a look out
point for vessels com ing in from the pearl banks (ibid.: 205)-
T he ship-sym bol also occurs frequently on coins and
sealings. Iconographic representations, such as graffiti on
pot-shards, clay models, illustrations on seals and clay amulets
dated to the Harappan period indicate the nature and diversity
o f boat-types used in the western Indian O cean in the third
m illennium BC. A rectangular stone seal from M ohenjodaro
shows a double-ended vessel probably made o f reeds bundled
with lashings, with a crescent hull and a rectangular structure in
the middle o f the deck (Mackay 1938: 34 0 —i). A similar craft
with a small cabin in the centre o f the deck is depicted on a clay
tablet also found at M ohenjodaro. A graffito from the same
site, on the other hand, represents a vessel with a mast and
steering oars or side rudders at the stern o f the ship (ibid.:
183). A terracotta boat m odel from Lothal, in contrast,
represents a square-sterned vessel with a sharp bow. A
perforation through the bow has been interpreted as a fitting
for the rigging, while a second one at the upper edge o f the hull
is said to be an oar-rest (Rao 1985= 5 ° 5 > 5 11)-
It has been argued that these images o f 'craft, with bow and
stern rising sharply rem ind us o f the boats described by the
19th century observers and still used on the Indus river by the
M ohanna fisherm en’ (Deloche, 1996: 204). Rather than
discussing the identification o f these craft, Potts focuses on
the heterogeneous nature o f the representations; identifies at
least two types o f vessels in these illustrations; and suggests that
the square-sterned craft with a single mast may have been used
for coastal voyages (l9 9 5 : 5^1) •
As compared to this heterogeneity o f Harappan craft,
contem porary representations on seven stamp seals from
Failaka in the Persian G u lf show dou ble-ended vessels with

raised bow and angular stern, at times with a mast and
identified as plank-built. In contrast, the two Bahrain seals
show a very different kind o f watercraft with identically raised
and curved bow and stern (ibid.: 567)- Thus the heterogeneity
in the types o f watercraft in use was not only restricted to the
Harappan civilisation, but extended to the Persian G u lf as well,
with reed boats and plank-built vessels being the predom inant
types.
T h e literature on watercraft in ancient Mesopotamia

reveals that, on one occasion, a 60 gur ship at Lagas during
the U r III period could carry 150 talents o f bitum en,
while at Mari in the O ld Babylonian period a 6 ton
delivery o f wine, consisting o f 300 jars o f 10 qa (litres),
each weighing 20 kg., was considered the norm al load for
a 20 gur vessel (Potts et al., 1996: 302).
T h e seafaring capabilities o f these watercraft are evident from

the discovery o f bitum en mixed with straw from the sites o f Ras
al-H am ra and Ras al-Junayz in the Persian G ulf, bearing mat
and rope impressions and barnacle encrustation resulting from
long im m ersion in seawater. Bitum en was used for caulking
reed as also for plank-built vessels o f Magan and D ilm un as
suggested by a num ber o f U r III ship-building texts, as well as
several old Babylonian texts from Larsa (Potts 1995 : 5^2).
Ethno-historically, the caulking methods employed by the
indigenous boats o f the G u lf include the use o f fish -oil or a
tallow and lim e mixture and bitum en has traditionally been
used either by the Marsh Arabs o f southern Iraq to coat the
bundled reed boats or in parts o f India. This would then
suggest that the bitum en that was found at Ras al-Junayz came
from reed vessels o f either M esopotam ian and/or Indian
origin. N or are the seafaring capabilities o f reed vessels in
doubt after the successful voyage o f the reed boat Tigris by
T h o r Heyerdahl in 1977/78 from Iraq to Karachi and across
the Indian O cean to D jibouti at the m outh o f the Red Sea.
Enorm ous quantities o f palm -fibre and p alm -leaf ropes
referred to in the cuneiform texts suggest that at least some o f
Sh ipp in g in th e In d ia n O cean : A n O v e r v iew 7
the water transport was sewn or stitched. T he U r III texts list

around 8 tonnes o f palm -fibre rope and I tonne o f p alm -leaf
rope, together with 6 tonnes o f fish oil probably used as an
an ti-fo u lin g agent on the rope (Potts 1997 *' 126—7). A n oth er
text affirms the use o f 59*29° w ooden pegs for the boat yards
o f Umma during the U r III p e r io d , indicating thereby the use
o f w ooden pegs in com bination with stitching — a practice
traditionally attested to in the region (ibid.: 128).
A ro u n d the beginning o f the Christian era, the ship-sym bol
first occurs on punch-m arked copper coins from lower Bengal
and th irty-fou r such coins with representations o f a hull have
been reported in the collection o f the Asutosh Museum,
Calcutta. T h e symbol is m ore elaborately depicted on unbaked
clay sealings and single or double-m asted ships are shown often
with K harosthi legends. O n e o f these has been deciphered as
'trapyaga’ rem iniscent o f the trappaga m entioned in the Periplus
Maris Erythraei (M ukherjee 199 ° : *9 )- Further representations o f
two-masted vessels with the rudder shown as a steering oar are
to be found on Satavahana coins from the lower K rishna valley.
These have been identified as sea-going vessels similar to the
toni o f the C orom andel coast that plied the east coast at the
beginning o f the last century (Deloche 1994 : I 85 )- These
representations continue as evident from the depiction o f a
single-m asted vessel with high prows found in an inscription
from Duvegala in Sri Lanka (Paranavitana 197° : plate xxv).
In addition to these depictions are representations o f so-
called 'R om an’ vessels. A n incom plete graffito on a Rouletted
ware shard from Alagankulam on the Tam il coast is said to
represent a 'Rom an boat’ . Casson ( l 997 ) term ed it a 'Rom an
vessel’ on the basis o f parallels with the mosaic from Ostia, but
this analogy has been contested by Tchernia (1998: 447 ~ 463 )
who identifies the incom plete representation as that o f a three-
masted vessel which has been referred to in G raeco-R om an
texts, but an iconographic representation o f which has yet to be
found. A noth er pot-sherd with a ship graffito was recovered
during the recent excavations at Berenike in a securely dated
deposit o f the second h alf o f the first century A D . T h ou gh the
representation has been identified as a characteristic craft o f
the M editerranean, a survey o f the Red Sea coast in the vicinity

o f Berenike has provided no evidence for harbour installations
much less for any structures similar to those known from
Ptolem aic and Rom an harbours elsewhere in the M editerra
nean (Sidebotham and W endrich 1996). This raises a wider
question o f the reasons for the choice o f the ship-sym bol for
depiction on pottery and coins.
A noth er p ioneering work on the Indian O cean is H ou ran i’ s
study o f Arab Seafaring. H ourani associates the history o f Arab
seafaring in the Indian O cean with expanding com m erce that
reached its peak in the ninth-tenth centuries A D. A fter the
tenth century the references are few and far between and
present a continuation o f the earlier established traditions
(1975 : 83). T he sources used for the study continue to be
literary com bined with pictorial representations. H ourani,
however, accepts that shipping in antiquity cannot be strait-
jacketed into national boundaries and that his study is to be
located within the larger context o f traditional shipping in the
western Indian O cean. In spite o f this disclaimer, secondary
sources have continued to designate w ater-craft described in
A rabic sources as 'Arab ships’ and hence trade carried out in
these as 'Arab trade’ .
Similarly, A rabic accounts from the ninth to eleventh
centuries AD by writers such as Baladhuri, Dinewari, Tabari and
Buzurg containing references to C hina ships’ at U bulla and
other Persian G u lf ports have often been taken by historians as
indicators o f Chinese ships. T h e terms used include 'sufun min
a l-siri (ships com ing from C hina); 'marakib al-siri and 'markad sini
(ships o f China); and the m ore direct *al-sufun al-siniyeh’
(Chinese ships). T he available evidence, however, disputes
any correlation o f these terms with ships o f Chinese origin
(Tampoe 19 8 9 : 12o ). O n the contrary, it would seem that the
term refers to ships involved in the trade with China.
A nother topic where literary sources are often ambiguous is in
the size o f ships. M ore often the problem is one o f interpreta
tion and being able to correlate the dimensions in terms
comprehensible to a m odern reader. A n exercise o f this nature
was undertaken in the context o f Chinese vessels sailing to the
Sh ipp in g in th e In d ia n O cean: A n O v e r v iew 9
Indian O cean (Wake 1997)* Marco Polo, for example, provides

very detailed and precise descriptions o f the vessels and states:
they are so large that they carry quite 5 ° ° ° baskets o f

pepper, some 6000 . . . A n d these larger ships have such
large tenders that they carry quite IOOO baskets o f pepper
(as quoted in ibid.: 52).
It has been argued that the 'basket’ was a com m on term used in
the medieval period by the Venetians and others for a package
o f definite weight, i.e. 225 kilograms. T he figures then
translate to between 1520 and i860 tons burden for the large
junks and about 3 ° ° tons for the smaller vessels (ibid.: 57)*
These figures are supported by separate and independent
sources and are an indicator o f the developm ent o f C h in a’ s
maritime trade with South and Southeast Asia during the Song
dynasty (ibid.: 8o).
Ethnography
A third region in the Indian O cean often associated with a

distinctive sh ip-building tradition is the Indonesian archipe
lago. This construction system has been term ed the lashed-lug
planked boat and it has been suggested that this technique
required an iro n tool and replaced sewing which did not
(H orridge 1982: i). To further cite H orridge:
The Indonesian round dowel technique was either a

separate invention o f the Austronesian-speaking seafar
ers, or it came from India: certainly it contrasts with the
Chinese methods. Boats on the Indus in 25 ° ° to 1500 BC
could have had dowels between their plank edges, because
the technique was known, as shown by their construction
o f wheels (Johnstone, 198O: I 79 —80). T h e round dowel
technique is an adaptation to the tropical m arine
conditions (1982: 58)-
The link between the introduction o f metal technology and the

shift to the use o f dowels is, however, not universally accepted,
as sewn craft continued to coexist in Southeast Asia till the
twentieth century. Also the evidence from the Pontian boat

fou n d in Malaysia indicates that early attempts in this
transform ation date from arou nd the third century AD
(M anguin 1985= 34 °)* while the generally accepted date for
the beginnings o f the Metal Age in Southeast Asia is around
2000 BC.
T h e spread o f A ustronesian speakers throughout island
Southeast Asia has been attributed to their success in boat
technology. It has been suggested that around 2 0 0 0 BC they
already had a boat-b uilding technology based upon lashings,
protruding pierced lugs, and a hollowed base for the hull with
added planks. A t this stage, however, they must have adopted
their own unique triangular sail and the outrigger construc
tion (H orridge 1995 : I 35 )* It Is significant that unlike the
western Indian O cean, the discussion on boat technology in
Southeast Asia is based not on textual references, but on
archaeological finds and ethnographic studies.
Ethnographic studies based on present traditions o f boat
building and navigation indicate that a variety o f craft are used
along the coasts o f the Indian O cean ranging from the log rafts
and dug-outs to cargo carriers (G reenhill 195 ^, I 9715 Chittick
1 9 8 0 ; Prins 1986; K entley and Gunaratne 1987; K apitan 1987?
1988, 1989, 199I; Blue, et al. 1997)- Indeed the east and west
coasts o f the subcontinent are characterised by two distinct
boat-building traditions: while the form er is hom e to the log-
raft; it is the dug-out that prevails along the west coast. In
contrast, sewn craft are found scattered throughout the entire
coastline (Deloche 1994 : fig- xxxvii). It is true that at present
these boats are largely used for fishing, mechanised trawlers
having taken over m uch o f the coastal and long-distance
transportation by sea. It has been argued that the contribution
o f ethnographic studies is prim arily as indicators for resolving
questions on the technical capabilities o f water-transport. T h e
same problem s have arisen whenever and wherever m en have
built boats o f wood — hence the im portance o f the eth n o
graphic approach (G reenhill 1995: 9 )* But the scope o f
ethnographic studies should not be restricted to providing
technological analogies, as has been the case so far.
Sh ip p in g in t h e In d ia n O cean : A n O ve r v iew ii
A second m isconception often associated with ethnographic

data is the suggestion that the watercraft o f antiquity were
similar to the present traditional craft o f the Indian O cean.
Thus a link is sought to be established between plank-built
vessels depicted on the stamp seals from Failaka with the present
day Arab bum. Yet it would be naive to accept continuity,
especially since European influence in the region has co n
siderably m odified the build o f larger vessels engaged in trade
or warfare in the region. Some evidence o f this is provided by
the large num ber o f technical nautical terms in G u lf Arabic
which are loan words from Portuguese (Potts 1997 : I36)-
H orn ell (192O: 139—73 ) was perhaps a pioneer in this field.
Based on ethnological studies, he described distinct boat
building traditions that evolved along the m ajor regions o f the
Indian coastline. Thus the kotia is the ocean-going craft o f the
K utch and Kathiawar coast, while along the K onkan, pattamars
are used for coastal sailing. T he dugouts o f the Malabar coast
are best suited for the extensive inland network. A lon g the
entire east coast, the catamaran is the characteristic fishing
craft, the name being derived from the Tamil term kathu maram
or tied logs, together with the masula, also known as padagu
among C orom andel fisherm en. What is intriguing is the
distinction in the boat designs along the Palk Straits. O n the
Tamil side the catamaran and boat canoe alone are u sed , while
on the Sri Lankan side the outrigger canoe called oruwa is the
dom inant type (ibid.).
In 1946 (reissued in 197° ) ’ H orn ell published the distilla
tion o f a lifetim e’ s work in Water Transport in which his major
preoccupation was with tracing the evolution o f water-craft, as
also their com m on origins and diffusion. Many o f H o rn e ll’ s
conclusions continue to be repeated in academic writings and it
is only the boat typology form ulated by him that has been
m odified in recent years (M cG rail 1985)- A noth er argument
put forward by H orn ell relates to the com m on origins o f
certain boat types and their dispersion through migrations and
movements o f people, a typical example being the outrigger
canoe. T he single outrigger is characteristic o f the whole o f
Polynesia and M icronesia and is also found on the coasts o f
India, Sri Lanka, Madagascar and the C om oro Islands as a

result o f the voyaging and colonisation by the 'Island people o f
the Far East’ (H ornell 197° : 255 )-
A t the m om ent a seventh century AD date is generally
accepted for the arrival o f Austronesian-speaking people on
Madagascar. Two possible routes o f travel have been suggested:
one, along the coasts o f South Asia and Africa; and the second,
directly across the Indian O cean. As several islands o f the
Indian O cean such as those o f the Mascarene group and
Seychelles rem ained uninhabited and were colonised only in
the seventeenth century, it has, however, been argued that
seafaring across the central Indian O cean was not a regular
practice (H orridge 1995 : *38) • It should also be rem em bered
that the history and developm ent o f a boat-type is a complex
process.
T h e appearance o f similar types o f simple boats in widely

separated areas and totally different societies suggests that
the same solution to the problem o f providing water
transport, given a reasonable similarity o f resources,
developed independently at different periods in different
parts o f the world (G reenhill 1995 : ^x)-
In recent years, attempts have been made to docum ent b oat

b uilding traditions in several regions o f South Asia, but these
have seldom addressed larger issues, such as ownership or
relations with agrarian and other com m unities. Similarly there
is no in form ation on traditional b o a t-b u ild in g centres;
possible shifts and changes in these; and factors leading to
innovation and technical change. Most o f all the historical
dim ension is lacking, as also survey and exploration to identify
coastal structures and landing places. In the Indian subconti
nent recording o f boat-types was done under the aegis o f the
C o u n cil for Scientific and Industrial Research and carried out
at several centres inclu ding Tam il University, Thanjavur
(Rajamanickam and Subbarayalu 1988; Varadarajan 19 9 8 ).1
These reports, based on interviews with boat-builders, never
theless provide an overview o f boat-building activity along the
coasts o f the southern states (Ray 1996).
Sh ipp in g in th e In d i a n O cean: A n O v er v iew 13
O n e region where attempts have been made to docum ent

ethnographic data with a view to addressing questions o f
historical relevance is O m an (Vosmer 1999 )- I *1 places such as
Musandam in northern O m an, num erous comm unities co n
tinue to em ploy traditional maritime practices such as fishing,
coastal trading and boat-building. In addition to docum enting
the craft used by these and other comm unities, 'another facet
o f the research was the collection o f a lexicon o f b oat-building
and m aritime term inology, both to categorise the types o f vessel
and to examine the incidence o f loan words as a key to cultural
or technological exchange’ . Especially useful for the purpose
was the isolated village o f Kum zar in the northern province o f
Musandam, ninety kilometres across the Strait o f H orm uz
from Iran. T h e inhabitants have developed a unique nautical
term inology with borrowings from A rabic, Farsi, Portuguese
and their own language. Many o f the terms are corrupt forms
o f loan words from Farsi and Portuguese, but can provide
useful inform ation about cultural contact and perhaps the
source and tim ing o f the introduction o f certain technologies
into ship-building in the region (ibid.). Inform ation on the
possible size and provenance o f ships can also be obtained from
an analysis o f anchor finds (Vosmer et al. 1998), o f which three
basic types have been distinguished in the western Indian
O cean. A n oth er aspect o f the m ulti-disciplinary approach
adopted by the O m an project involved experim ental eth n o
graphy and attempts were made 'to learn by direct experience
about the sailing characteristics and handling o f one type o f
O m ani vessel, the battil o f M usandam’ (Vosmer 1999 )*
This approach has had an earlier proponent in A lan Villiers.
In 1938, he undertook a journey from M a’alla to Gizan in a
small Red Sea zarook, Sheikh Mansoor and published his im pres
sions o f the travel (Villiers 1954 )- T he prim ary interest was in
recording the 'way in which the ship was ru n ’ (ibid.: 173) and
the means o f navigation adopted. T he cargo that the 2 0 -ton
ship carried consisted o f trans-shipped goods from A den
including Australian flour, two cases o f Japanese matches, a
num ber o f bales o f Japanese cotton goods, some rice and coffee
and other items (ibid.: 178). In 1980, T im Severin recreated
the seven voyages o f Sindbad from O m an to China in a ship

made from Malabar tim ber and held together with coconut
rope (Severin 1982).
Archaeology
T h e potential o f archaeology, especially underwater archae

ology for a study o f boat-building techniques has yet to be
tapped in the Indian O cean region, though there have been
prelim inary forays into surveying and exploration o f some
areas. O n e o f the pioneers in the Indian subcontinent is Rao,
from the Institute o f O ceanography in Goa, who explored
coastal sites such as Dwarka and Lothal (Rao 199 ° ) • O th er
expeditions, such as those from the U nited States and Australia
have been involved in surveying o ff the shores o f O m an,
Yem en, Bahrain and the U nited Arab Emirates (Owen 1997 )*
while British expeditions have had some success with European
wreck sites in East A frica (Sassoon 1982). In Sri Lanka jo in t
projects sponsored by the Post-graduate Institute o f A rch ae
ological Research, the C entral C ultural Fund and the Western
Australian M aritim e Museum have made significant discover
ies, most o f them in the Galle H arbour area (Vosmer 1999 )-
Nevertheless, ship-wreck sites have yet to be discovered in the
Indian O cean on the same scale as in the M editerranean,
though it is no doubt true that the two regions present very
different environm ents for underwater archaeology.
Coastal features such as landing places or harbour installa
tions could be profitably used for a study o f seafaring activity in
the p re-m od ern p eriod , but the bulk o f the data on this theme
remains scattered and under-utilised. A noth er topic on which
little inform ation is available is coastal fortification. O n e o f the
characteristic features o f m aritime trade from the ninth/tenth
centuries onwards was the location o f markets in fortified
settlements along the Indian O cean littoral, as also in the
interior. Rules governing the payment o f taxes and regulating
the functioning o f the markets were often inscribed on copper
plates and provide useful insights into the organisation o f the
trade network. T h e Q u ilo n copper plates o f Sthanu Ravi, for
Sh ipp in g in th e In d i a n O cean : A n O v e r v iew 15
example, record trading rights granted to the N estorian

Christian church. A market was located within the precincts
o f the fortified settlement at the port o f Q u ilon , while the
church was situated outside the fortification wall (Abraham
1988: n o ).
A rchaeological excavations conducted at Siraf in the Persian
G u lf indicate its beginnings in the Sasanian period when a fort
was located at the site. Roughly square on plan, the south side
contained a m onum ental entrance flanked by hollow sem i
circular towers (Whitehouse 199^: 34 1)* O n the basis o f the
ceramic evidence, Tampoe (1989: 99 ) has argued that the fort
was largely military in character, while the settlement at Bushire
provided the base for com m ercial activities. This conclusion
has, however, been debated by W hitehouse (1996: 3 4 ^) who
contends that in addition to military functions, Sasanian Siraf
was also a com m ercial port as evident from im ported objects at
the site.
Similarly, archaeological exploration in the Indus delta has
revealed remains o f structures dating from the ninth/tenth
centuries onwards (Kervran 1999 )* A t the site o f Jaki Bandar in
the Indus delta, for example, the ruins o f a fortress were
discovered beneath the existing structure (ibid.: 117)- T he
extant constructions along the Indian coast, however, date only
from the thirteenth centuries onwards and forty-five o f these
have been identified so far (Naravane 1998: 127)-
Much o f the archaeological work along the Indian O cean
littoral has focused on trading centres and settlements, and a
range o f sites have been excavated dating from the protohistoric
period onwards. What is nevertheless required is an enquiry
into the mechanisms through which trading networks fu n c
tioned as well as into the w ater-craft used in the transportation
o f com m odities across the Indian O cean in the p re-m od ern
Navigation
In contrast to boat-building, nautical sciences have not
received the attention they deserve, especially in an historical
i6 H im a n s h u Prabha Ray
context. In the traditional system, navigation was based on

stellar knowledge, and nautical learning was founded on the
accum ulated experience o f navigators. These skills were
com m unicated orally and learnt during years o f apprentice
ship. The maritime literature in Greek, Sanskrit and A rabic
dated p rior to 1000 A D , was largely descriptive and integrated
accounts relating to coastal navigation, winds, ports, etc. into
descriptions o f countries. From the end o f the ninth century,
however, a change occurs and there are references to maps and
portulans, though the real expansion takes place in the
fifteenth and sixteenth centuries (G ro sset-G ran ge 199^;
M iquel 1996: 8 ll).
In the fifteenth century m aritime activity in the western
Indian O cean was focused between two main areas: the east
A frican coast with its num erous ports such as Mogadishu,
M alindi, Kilwa and Sofala on the one side and the Sultanate o f
D elhi, on the other. O n these voyages, the navigators crossed
the paths o f the Chinese m ariners who were pushing into the
area. In 1405, the m aritime expeditions o f the A dm iral Zheng
began, in the course o f which he reached Indonesia, India and
A frica. By the late mediaeval period indigenous nautical charts
developed which were comparable to the Portuguese portolanos.
These were based on systematic observation o f coastal con fig
uration, landmarks easily identifiable from the sea at a
distance, bearings in terms o f h orizon star azimuths and
underwater flo o r features. A large num ber o f references
indicate that such charts were already in use in the north Indian
O cean when the first European vessels entered the Indian
O cean.
Physical evidence o f the presence o f such charts has been
found among the K utchi seamen in the form o f a pothi or
manual written in K utch i and dated to 1664 A D .2 Th e manual
is not complete and only five maps rem ain. These are line
drawings o f lim ited coastal stretches, often com bining silhou
ette profiles o f coastal landscapes as observed from the sea
(Arunachalam 1996: 279 )-
Th e fifteenth century was also marked by interchanges
between the ship-building tradition o f the Europeans with that
Sh ippin g in th e In d ia n O cean : A n O v e r v iew 17
o f the Indian O cean. Q aisar’ s study (1982, I9 9 &) assesses the

nature, degree and pattern o f Indian response in accepting or
adapting European technology and remains the only work on
this im portant issue to date. Though the study is broad-based
and deals with technology as a whole, an entire section is
devoted to shipbuilding.
In addition to the Indian response to European technology,
it is also im portant to understand the extent to which Indian
O cean solutions to problems o f the sailing ship were adopted
by European shipwrights. This was perhaps one o f the factors
that led Adm iral Paris (1806—93) to undertake detailed
docum entation o f the extra-European’ craft o f the Indian
O cean — a source that continues to be under-utilised (Reith
1993). Similarly, the recovery o f the wreck o f the Santo Antonio de
Tanna o ff the coast o f Mombasa reveals a very shallow hull 'o f a
geom etric pattern unlike anything recorded in Portuguese
documentary sources’ (Barker 1989: 212)- The vessel has been
described as a Portuguese 'frigate’ built in India in 1681 and
lost in 1697. T he matter is further complicated by the fact that
many o f the vessels constructed in India and term ed 'frigates’
by the English and the Dutch may have been m odelled on
English ships visiting Goa (Boxer 1940).
F ish in g C o m m u n itie s
The next issue o f relevance here is the study o f fishing and

sailing com m unities. Secondary sources on this theme are by
and large ethnographic and anthropological studies o f localised
fishing com m unities in the Indian O cean (Barnes 1996;
Pokrant 1997)- T he theme has also been o f interest to
researchers involved in the process o f social and econom ic
developm ent o f the fisheries sector. In the context o f inland
water transport, a study was undertaken in Bangladesh to
docum ent the country boats including their construction,
operation, ownership and com petition with mechanised vessels
(Jansen et al. 1989). T he objective was to review the relation
ship between the country boat sector and processes o f rural
development and impoverishment, with a view to im proving the
i8 H im a n sh u Prabha Ra y
efficiency and effectiveness o f the country boat operations

(ibid.: 5- 6 ).
O th er papers that attempt to investigate the process o f
developm ent do so by analysing innovation and change in boat
technology (Patel 1989; K u rien 1996)- In addition, K u rien
explains seeming contradictions such as 'large boats in the poor
fishery’ . This phenom enon is explained by the need to resort
to alternate occupations such as coastal cargo trade to
compensate for the less productive m arine fishing grounds in
Gujarat and the lowest demand for fish in the hinterland
(K u rien 1996: 22 l). In contrast, fish resources are both
abundant in Malabar and also available close to the coast. As a
result the carrying capacity o f w ater-craft becomes a non-issue
and the traditional vessel used in the region is the small dug-
out or vallam.
Contributions that analyse the subject in an archaeological
context are few and far between. O n e o f these on the ethno-
archaeology o f fishing relates to the Harappan period (Belcher
1999). The m ethodology adopted consisted o f identifying fish
remains in the faunal assemblages o f Harappan sites. In order
to understand the procurem ent strategies, ethnographic studies
were undertaken in the region and fishing and fish processing
practices docum ented. These investigations have revealed that
the fishing industry contributed a significant portion o f food
for some populations in the urban and rural settlements o f the
Harappan civilisation. W hile the emphasis in B elcher’ s study
was on understanding the contribution o f fishing as a resource,
other studies have focused on developing m ethodologies for the
identification o f the fishing com m unities in the archaeological
record. This becomes particularly problem atic in the case o f
m obile com m unities, such as the O ran g Laut o f Southeast Asia.
These fishing com m unities, term ed 'true nomads o f the sea’ ,
represent a cultural adaptation that has survived almost to the
present-day. W riting in the nineteenth century, the Dutch
Consul G eneral in Bushire described them as a hybrid fishing
society com prising o f Arabs, Persians, Baluchis, Socotris,
Mahris and even natives o f the Hadramawt. They moved in
accordance with the season and spent almost three months o f
S h ip p in g in t h e In d ia n O cean : A n O v e r v iew 19
the year on the coast o f Baluchistan, where they salted and

dried their catch, before returning to O m an to sell a portion
o f it. A t the time o f the hajj, they sailed to Jeddah and the coast
o f the Hejaz to sell their fish (quoted in Potts 1990: 57).
A t present, maritime hunting-gathering often referred to as
boat nom adism as a subsistence strategy is lim ited to parts o f
island Southeast Asia. Traditionally the habitat o f the so-called
'sea-nom ads’ included travelling by boat over an area extending
from Burm a across the Indonesian archipelago to the Riau-
Lingga islands beyond Singapore to reach the Sulu Sea in the
southern Philippines (Sopher 1977 : 54 “ 5 )* O ne o f the groups
that has been studied is the 'Chaw Lay’ based on the Phuket
group o f islands — Phuket being the m eeting point and
cerem onial centre o f the widely dispersed network o f maritime
hunter-gatherers. O f the 45 ° ° Chaw Lay living along the
south-west coast o f Thailand in 1981, a third were resident on
Phuket island itself — the groups ranging in size from two to
m ore than eight hundred people. T he specialised subsistence
strategy o f these groups is based on an exploitation o f marine
resources, especially fish. As a result o f dependence on a
m obile resource such as fish, these com m unities inevitably
maintain a peripatetic way o f life,
with a compact and easily transportable material culture

and social attitudes which encourage the spirit o f
adventure, group co-operation and an outgoing h ospi
table attitude to other nom adic groups they may meet on
the foraging expeditions and upon whom they may have to
depend for food and shelter (Engelhardt and Rogers
1997= 179)-
O ne o f the problem s associated with this subsistence strategy is
that it is difficult to isolate it in the archaeological record,
especially in the sandy beach matrix o f the coast. These
habitation areas are generally occupied in short stretches and in
keeping with a complex spatial patterning. W hile a site or a
portion o f a site is being used as a base camp by one Chaw Lay
group, other smaller and transient groups could occupy
portions o f the site, sometimes tem porarily. A t the same time
the site could also be exploited by other Chaw Lay groups for
water and vegetable collection. T h e use pattern is further
com plicated by the fact that while some o f the sites indicate
seasonal occupation, others are left fallow to allow regeneration
(ibid: 182).
O n e o f the sites on the Indian O cean littoral where an
attempt was made to study the faunal remains, especially those
o f fish, was Shanga on the Larriu archipelago dated between the
eighth and fourteenth centuries AD. Relatively small numbers
o f fish remains were found in the earlier levels with a peak in
consum ption in the thirteenth century (H orton 1996: 3 9 1)-
T h e rarity o f fish and animal bones in the early levels was,
however, com plem ented by the presence o f shell-fish in
sufficient quantities.
Ind igenou s M aritim e Networks
O n this issue, the data c o n t i n u e s to be d isp e rse d a n d n e e d s to

b e c o n s o l id a t e d f o r an u n d e r s t a n d i n g o f the tr a d i n g system o f
the I n d i a n O c e a n in the p r e - m o d e r n p e r i o d . A s m u c h o f this
tr a d e c o n t i n u e d to b e tr a n s p o r t e d i n the t r a d i t i o n a l c ra ft o f the
r e g i o n , an analysis o f the n a tu r e a n d q u a n t u m o f c o m m o d i t i e s
t r a n s p o r t e d w o u l d h e lp u n d e r s t a n d i n n o v a t i o n a n d c h a n g e in
shipping.
Inform ation on n o n -V O C local ships is available in harbour
masters registrations from a num ber o f Javanese ports for the
period o f the late 1760s to the late I7 7°s and these provide
some inform ation on the quantum and nature o f shipping in
the region (Knaap 1996). Forty-seven different types o f water
craft have been m entioned in these records, some m ore
frequently than others. T h e most frequently m entioned types
were the jukong and paduwang, though the most com m on type was
the mayang (ibid.: 3 I—3 3 )* T he jukong was a large-sized dug-out
canoe, while the mayang was originally designed as a fishing boat,
but in due course came to be used for sea-borne trade.
Inform ation on the ethnicity o f the skippers indicates that o f
these, 40% were o f Javanese origin, while 3 ° % were Chinese
and 9% Malays. In addition were a num ber o f m igrant groups
Sh ippin g in th e In d ia n O cean: A n O v e r v iew 21
(ibid.: 64)- T h e average crew num ber per vessel was nine. Data
on ownership is not always available, but it would seem that J to
24% o f the skippers owned not only the ship, but also the
cargo. A m on g the other owners were Dutch colonial officials,
rich C hin ese and the elite o f Palem bang, though the
indigenous Javanese aristocracy does not figure (ibid.: 75 )-
Sim ilar data continues to be sparse for the indigenous
trading system o f the Indian subcontinent, though an early
attempt at com pilation o f the writings o f European travellers in
the Indian O cean in the sixteenth and seventeenth centuries
was made by H ill (1958). The best known o f these is Thom as
Bowrey (1650—1713) wô spent nineteen years in India and
Southeast Asia as a private trader mainly in pepper.
In his account o f the trade o f Surat in the first h alf o f the
eighteenth century, Das Gupta first drew attention to the
discussion on Indian shipping in Dutch sources, no Indian
evidence o f any significance being available for the period. He
showed that the strength o f Surat’ s com m ercial m arine at the
turn o f the eighteenth century was about one hundred, 'a very
impressive fleet for an Indian O cean p o rt’ ( l 979 : 289). Some
data is available in the sources on the size o f the vessels. This is
given in khandies, three khandies making one ton. T h e m arine
khandy, probably adapted from the measure used in weighing
raw cotton in Gujarat has continued almost to the present
(ibid.: 285). T h e problem , o f course, is that not all the vessels
calling at the port o f Surat have been listed, nor is it always
possible to identify the different types o f vessels m entioned in
the lists o f arrivals. W hile the larger craft have been generally
included, the same is not valid for the single - decked, single
masted craft involved in coastal trade:
But the entire trade o f the G u lf o f Cambay was carried by

such vessels and Surat’ s trade with Sind, Kathiawar and
much o f the west coast was dom inated by them (ibid.:
281).
O n the C orom andel coast, Nagore, located five kilometres

north o f Nagapattinam was a centre for ship-building from the
b egin n in g o f the eighteenth century onwards and also
22 H im an sh u Prabha Ray
prospered as a centre o f free trade under the ruler o f Tanjore.

Dutch records refer to a three-m asted ship, forty-two and a
quarter cobidos long over the keel (i cobido= l8 inches, hence
about 64 feet) built at Nagore for a Marakkayar merchant
(Bhattacharya 1995 : 35 &> footnote io ). In spite o f these
scattered bits o f inform ation available in the sources, a
comprehensive delineation is yet to emerge.
T h e two volumes by C haudhuri (1985; 199 ° ) deal with the
econom y and civilisation o f the Indian O cean from the rise o f
Islam to the m id-eighteenth century. He acknowledges his debt
to Fernand Braudel ( l 949 ) wô pioneered the idea that 'the
study o f civilisation might be named after a sea’ and attempts to
address the question:
Is it possible then to discover the unity and diversity o f the

Indian O cean civilisations through the study o f lo n g
distance trade, which o f necessity takes place across
geographical and cultural watersheds and which has a
com pelling tem poral dim ension? (1985: 2).
In his first book, Chaudhuri assigns two chapters to the study o f

ships and seafaring activity. He accepts the presence o f
com m unities that lived o ff the sea, but argues that on account
o f their lowly status, these could not develop into a 'single
national maritime tradition to be used as an arm o f the state’ .
Instead these com m unities took to piracy and sea-trade (ibid.:
122). T h e sources used for this study are either accounts o f
Arab geographers or archival material relating to the European
Com panies. T h e overall discussion, however, flounders at an
impressionistic level and provides very few details o f Asian
shipping. A similar deficiency o f analysis on shipping is evident
in Arasaratnam ’ s study o f the seventeenth century ( l 9 94 )>
though the author accepts that there are references to an
'abundance o f Asian sailing craft from the various regions
plying the ocean’ (ibid.: 247)- T h e discussion is largely based
on secondary sources such as H ill (1958), M oreland ( l 939 ) and
others.
In contrast is the paper by M anguin (1985b) which quotes
observations on nautical technology by sixteenth century
Sh ipp in g in th e In d i a n O cean: An O v e r v ie w 23
travellers such as Gaspar C orrea. C orrea describes craft on the

Malabar coast with strakes stitched with fibre o f the coconut
tree and with a few frames likewise sewn to the planks (ib id .: 4)-
A noth er technique for structurally strengthening the hull was
by inserting wooden dowels into the seams o f the planks. O ne
o f the vessels that stands out in Portuguese sources o f the late
fifteenth and early sixteenth centuries is the craft o f the
Gujarati merchants — the nao which was around 3 0 0 ~ 600 tons
on an average (ibid.: 9)-
O f relevance here is a discussion o f some o f the technical
problem s faced by Portuguese shipping in the sixteenth century
on entry in the Indian Ocean, as this is essential to dispel many
o f the myths associated with the entry o f European shipping in
the region. T h e vessel used by the Portuguese for travel down
the A frican coast from about 1436—41 was the caravela, widely
supposed to derive from Arab fishing craft. It was small and
required a relatively small investment, and significantly no
special skills or facilities were necessary for its construction.
Also it was small enough for repairs to be carried out by the
crew itself should the need arise (Barker 1992: 4 3 6 )-
Approaches to the problem o f repairing varied in the different
periods as also in diverse parts o f the world. Areas with higher
tides posed little problem as these could be utilised to berth
even laden coasters on the beach, though this also meant that
there was relatively less time available for repairs before the area
got flooded again. Nevertheless the process o f docking was a
comparatively late development and in England it was not until
the eighteenth century that dock structures and gates developed
'to anything recognisable in m odern docks’ (Barker I 99 I: *79 )-
A nother critical factor on long voyages was drinking water
and when the Portuguese arrived in the Indian O cean they
soon copied the local technique o f using integral tanks (Barker
1992 : 445 )-
The techniques o f carrying water were not mastered for a
full 300 years after Colum bus. Even great cities had
acutely unsatisfactory water supplies, yielding in many
cases no m ore than 5—10 litres (o f highly questionable
quality) per head per day, mostly at public fountains.

Simply to collect sufficient quantities o f clean water for a
ship must have been a m ajor problem almost anywhere
(ibid.: 444)*
From the response in early Portuguese sources it is evident that

the Europeans encountered large local trading vessels in the
Indian O cean.
O n e type o f ship encountered from the Bay o f Bengal to

C hina was the jonque (distinct from the Chinese junk), a
m erchant vessel o f some 2—300 tons. Little is now known
o f these craft: they disappeared quite rapidly, declining as
the Portuguese pressure increased through the 16th
century (Barker 1989*. 203).
If our knowledge o f the history o f ship-building in the Indian

O cean is inadequate, our inform ation on ownership o f these
traditional craft is practically non-existent. Two terms that
occur frequently from the fifteenth to nineteenth centuries
with regard to indigenous merchants in the Bay o f Bengal are
Chulias and K lings. W hile the form er was used to designate
M uslim merchants from the C orom andel coast, the latter was
often used interchangeably with Chetty and denoted H indu
merchants from the C orom andel coast drawn from three
linguistic groups: Tam il, Telugu and Kannada (M cPherson
1998: 187, 206, footnote 2).
Q uantification is difficult in the p re-m od ern p eriod , but
records available from Melaka on the Malay peninsula would
suggest that in the fifteenth century, in addition to the Chulia
and the K ling, nearly IOOO H indu and M uslim Gujarati
merchants were resident at the coastal settlement, while
10 0 0 —2000 were on the move across the Bay o f Bengal at
any one time during the sailing season’ (M cPherson 1993 :
155)* In the early seventeenth century, nearly 1500 Javanese
merchants travelled annually 'to Banda in search o f spices,
which they carried to ports such as Melaka for purchase by
Portuguese, South Asian and Chinese m erchants’ (ibid.). It
would seem that while the Chulias are referred to both as
merchants and ship-owners, the same is not the case with the
K lin g who seldom find m ention as ship-owners.
These two terms are o f much earlier circulation in the Bay o f
Bengal. Foreign and local merchants involved with tax farm ing
appear in the inscriptions from Java dated between 840 to
1 3 0 5 A D. T h e term kling refers both to people specifically from
India, and also as a general term for foreigners and in one
instance, it occurs as a part o f the personal name — si kling
(Barrett Jones 1984: 25 )- Similarly the expression Colika is
said to denote people from south India (Sarkar 1969: 169).
It is, however, problem atic to define 'aliens’ referred to in
inscriptions in ethnic terms and a good example o f the
changing meanings are terms such as 'damila or ja v a n a ’ . A
divergence o f ethnic association is evident in the case o f the
ja v a n a . The first m ention o f them is from the north-west o f the
Indian subcontinent and refers prim arily to the Ionian Greeks
and subsequently the In d o -G reek s. In contrast are the
inscriptions from central and western India which record
donations by ja v a n a to Buddhist monastic establishments; the
donors, in this case have Sanskrit names. In a different category
is the earliest corpus o f poems in Tam il dated from the
beginning o f the Christian era to the fifth/sixth centuries AD
which contain descriptions o f ja v a n a com ing with gold and wine
in their ships and returning with pepper. O n e o f the centres
frequently visited by them was M uciri. References in these texts
also indicate a m ore varied role for the ja v a n a . They are
described as sentries and bodyguards to kings and renowned for
their workmanship in making lamps. A separate quarter o f the
ja v a n a is referred to in the suburbs o f Kaveripum pattinam .
What adds complexity to the whole issue is the variation in
ritual status accorded to the ja v a n a , as evident from the
Dharmasastric literature (Ray 1988: 311—325)*
In the sixteenth and seventeenth centuries, even when
Indian ships were owned and controlled by one person, the
cargo was not owned by one corporate owner, nor were the
m en employed by a single individual. It was customary for
private merchants and travellers to hire space for their cargo as
well as for themselves (Qaisar 1987: 3 4 °)•
Details on seafaring and the organisation o f shipping are

seldom available in historical sources, an exception being the
C airo Geniza documents, dated between the eleventh and
thirteenth centuries AD. It would seem that a com bination o f
larger and smaller craft travelled together, especially in the
Indian O cean, since a smaller craft had greater chances o f
survival than a large vessel (G oitein 19 6 7: 3 ° 6 )- T he references
suggest that a ship was norm ally owned by a single p roprietor
and instances o f join t-ow nersh ip are rare. The largest single
group o f ship-owners were merchants, though there is m ention
o f high personages like members o f the royal family and
military commanders owning boats (ibid.: 310). References to
Jewish proprietors o f ships are relatively less frequent in the
M editerranean as com pared to those in the Indian O cean
(ib id .: 311). A noth er distinction between the two seas lies in the
fact that letters sent from boats operating in the Indian O cean
refer to cabins or compartments which were sometimes shared
by business associates (ibid.: 3I 5)*
A n exceptional find dated to the second century AD is that o f
a fragmentary papyrus that records a loan-agreem ent drawn up
at Muziris on the south-west coast o f India, between two
merchants whose names have been lost. T he three imports
from India that find m ention are Gangetic nard, ivory and
textiles. T he papyrus provides details o f payments to be made to
different people for transporting the com m odities, such as the
camel driver 'for use o f the road to K op tos’3 and o f the
different representatives to be inform ed 'for loading aboard
the river’ , as also for conveying 'downstream to the warehouse
that receives the duty o f o n e-fo u rth at A lexandria’ (Casson
1986 : 73 - 79 )-
It is evident that there are large gaps in our understanding o f
ship-building in the p re-m o d ern period. These lacunae need
to be attended to not only for a study o f the history o f
technology, but m ore so for an appreciation o f econom ic
interchanges in the region. It is crucial that ship-building
traditions be docum ented, not as exotic markers o f a bygone
era, but as a holistic study o f m aritime com m unities. This must
include an emphasis on rituals, social practices and beliefs.
Notes
1 See also u n pu b lish ed reports on the In digen ous T rad ition s o f In dian
N avigation, subm itted to the N ational Institute o f Science, Technology
and D evelopm ent Studies, New D elhi, R ajam anickam , et al., I 9 9 1)-
2 It is now in the N ation al M useum , New D elh i mss. no. 8 2 . 2 6 3 *
3 Presum ably the goods were to be tran spo rted fro m the Red Sea p o rt o f
disem bark ation to in lan d K o p to s, and then down the N ile.
2
S e a f a r i n g s , S hips and S h ip
O wners: In d ia and the Indian
O cean (a d ^ O O — I ^ O O )
Ranabir Chakravarti
This essay is an attempt to appreciate India’ s role in the Indian

O cean p rior to the arrival o f the Portuguese at the turn o f the
fifteenth century. T h e Indian maritime scenario is to be
presented here in a historical outline; in other words, what is
placed here relates to certain aspects o f the maritime history o f
the subcontinent before the arrival o f the Europeans in A frica
and Asia. M aritime studies, though having undoubted in te r
relations with oceanography, are, however, not merely an
appendage to this im portant discipline within the earth
sciences. A fundam ental issue in maritime studies is to situate
a given area in a well defined maritime space. Such a m aritime
space is not simply confined to a strict aquatic zone, but is
inseparably linked up with the adjacent coast and its hinterland.
Seen from this point o f view, it would be futile to lim it
maritime studies merely to naval encounters, naval diplom acy
and contracting naval treaties, though sometimes such specta
cular events did capture the imaginations o f the historian at
sea. Enquiries into any ocean/sea cannot but recognise the
im portance o f searching for the hum an face o f a given
m aritime space. But contrary to popular perceptions, they
should not be reduced to the study o f sea-borne trade alone,
although it is an integral part o f maritime studies. No maritime
region can hardly be examined in isolation; but its perusal can
28
Se a f a r in g s , Sh ips a n d Ship O w ners 29
be effectively carried on by highlighting the unity between the

sea and adjacent land, as Fernand Braudel’ s seminal ideas on
the history o f the M editerranean dem onstrate. N o less
im portant is his stress on the connections between one
m aritim e space and another. T he Braudelian perception
negates the image o f the ocean/sea as a geographical barrier,
but views it as a factor o f com m unication, linkages and unity
among diverse comm unities (Braudel 197^) *
M aritim e studies have in recent decades made considerable
strides as a result o f new discoveries in maritime archaeology.
Underwater explorations and excavations at or near possible
sites o f shipwrecks have yielded remains o f actual vessels. Such
underwater archaeological recoveries conducted are further
com pared, contrasted and supplem ented with ethnographic
and/or ethnoarchaeological data by the maritime historian to
offer a clearer image o f seafaring activities in the past,
particularly during the p re-m od ern times (Gasson 1991)-
M aritime archaeology, a relatively new branch o f archaeological
studies, have immensely helped generate technical data on
nautical history. T here may be seen an urge o f late to broaden
the scope and application o f maritime archeaology strictly from
the study o f nautical and ship-building technologies to the
appreciation o f what is termed as 'm aritim e culture ’1 (Wester-
dahl I 9 9 4 )- W esterdahl further feels that the seemingly
unchanging pattern o f ship-building and seafaring during
the p re-m o d ern period (i.e. before the advent o f steam
navigation and the industrial age) may be appreciated m ore
m eaningfully by the Braudelian concept o f long duration (la
longue duree). T h e booms and slumps in maritime trade on the
one hand and the ambitions and processes o f the states on the
other may be judged in terms o f B raudel’ s les conjonctures as these
certainly spanned over a shorter chronological range (Wester
dahl 1994: 269).
T he mastery o f Europe over greater parts o f the globe from
the late fifteenth century to the early twentieth century was to a
large extent the result o f the European dom ination over
various maritime spaces and related areas. This has encouraged
sustained academic outputs on the M editerranean Sea and the
30 Ra n a b ir C hakravarti
Atlantic O cean and the countries thereof. T h e scholarly

recognition o f the Indian O cean as a valid unit for maritime
studies is a relatively recent academic developm ent. A n ongoing
perusal o f the Indian ocean is only possible with orchestrated
efforts from diverse disciplines in the earth and hum an
sciences: geography, geology, oceanography, anthropology,
history, political science, nautical studies etc. As com m unities
o f the Indian O cean countries are generally considered to have
belonged to what is called traditional societies, the historical
studies o f the Indian O cean can hardly be divorced from
ethnographic and archaeological findings with regard to sea
faring activities o f both remote and recent past.
T h e emergence o f new nations in the 'T h ird W orld’ ,
follow ing decolonisation in Asia, A frica and Latin A m erica, has
immensely encouraged scholars to appreciate the role o f n o n -
Europeans in the historical developments o f these countries.
A n evident impact o f this perspective can be seen in the
growing im portance o f Indian O cean studies. T h e Indian
O cean dominates the sea-face o f Asia and provides vital
linkages o f Asians with num erous A frican com m unities along
the eastern sea-board o f A frica. T h e geographical area o f the
m aritim e space called the Indian O cean should be defined at
this juncture. T h e map o f the Indian O cean, published by the
National Atlas and Them atic M apping O rganization, places it
up to the Gape o f G ood H ope in the west, to Antarctica in the
south, and it includes the Red Sea, the Persian G ulf, the
A rabian Sea and the Bay o f Bengal, but leaves out the Java and
C hina Seas .2 It is one o f the three m ajor oceans in the w orld,
but actually extends over only a fifth o f the w orld’ s total
m aritim e surface, a fact that would on the other hand point to
the relative smallness o f this ocean. O n e cannot also lose sight
o f the fact that this relatively small m aritime space connects no
less than thirty seven countries inhabited by a third o f w orld’ s
population (G otthold 1988). T h e 'relative smallness, which
facilitates com m unications, explains, why the Indian O cean has
been, m ore than any other ocean, the vehicle o f most varied
hum an contacts, with very rich consequences’ (Verlinden 1987:
2 7 )-
S e a f a r in g s , Sh ips a n d Sh ip O w ners 31
The other relevant geographical fact is the more or less

central position o f the Indian subcontinent and Sri Lanka
among the countries o f the Indian Ocean. The two long coast
lines o f India, washed by the Arabian Sea and the Bay o f Bengal
respectively in the west and east, provide ample scope o f
maritime contacts along the two littorals and also with areas
overseas. A major factor o f unity in maritime Asia is certainly the
m onsoon wind system. Before the advent o f steam navigation
shipping and navigation in the Indian O cean zone were largely
shaped by the more or less predictable alterations o f the south
west (approximately from June to September) and the north-east
(approximately from October/ November to January) m onsoon
winds. This not only directly affected the patterns o f voyages —
and therefore, the fortunes o f ports and seafarers — but had
profound impacts on the primarily agrarian communities o f the
Indian O cean countries living in the interior. It has been
recently observed that a long-term perspective, say spanning over
the last two millennia, would project Asia and not Europe, as the
leading maritime continent o f the world (Broeze 1989: 8).3
Such a recognition has indeed been largely responsible for a
recent spurt in publications on the sea-faring activities in the
Indian O cean (Chaudhuri 1985, 199 ° ; Chandra 1987; Das
Gupta and Pearson 1987; McPherson 1994-)•
Such studies are marked by an emphasis on the period from
1500 to 1800 principally because o f the a) availability o f
enough data and specially statistical inform ation and b) vast
transform ation o f the Indian O cean situation during these
centuries as a result o f the spread o f European mastery over the
area. O ne cannot, however, deny that the European urge for
reaching out to the Indian O cean was prom pted by their
knowledge and perception — whether direct and received — o f
the Indian O cean as an intensely active and interacting
m aritime zone even during the p re-15 0 0 days (Landstorm
1964; Lach 1965; C haudhuri 1990a). The subcontinent’ s
nearly central position in the Indian O cean and its two long
coastlines must have facilitated regular movements o f men,
m erchandise, ships and ideas from and to India over centuries.
It has further been noted that around IOOO AD segmented
32 R a n a b ir C hakravarti
voyages between the western (e.g. Fustat in the Red Sea area and
Siraf in the Persian G u lf) and the eastern (south-east Asian
and C hinese harbours) term in i o f the A sian m aritim e
com m erce began to be increasingly preferred to the previous
practice o f a single voyage between the two points (Chaudhuri
1985). As seafaring in this area was largely regulated by the
alterations o f the m onsoons, it would have been impossible to
accomplish a round trip in a year. Indian littorals, dotted with
num erous ports and harbours, which were already known for
their brisk export and im port trade, were ideally located as
stop-overs o f ships sailing across the Indian O cean. Indian
harbours and ports must have gained econom ically under such
circumstances. Situating India in the affairs o f the Indian
O cean p rior to 1500 is, therefore, no less fascinating an
exercise than exam ining the centuries following it. T h e present
paper may be seen as a plea, already placed by Bouchon and
Lom bard (1987: 4 6 —70), for the study o f the Indian O cean
during the period p rio r to 1500. T h e m ajor problem for the
study o f India’ s role in the Indian O cean during the p re-150 0
days lies in the well-know n paucity o f relevant data. Statistical
and quantifiable data on the com m odities, types o f ships, their
carrying capacities and tonnage and their sailing patterns are
almost rare. Impressionistic data, however, can be culled from
both indigenous and non -indigenou s literary materials, the
latter category often in the form o f impressions o f n on -In d ian
travellers, geographers and religious personalities. Valuable
evidence o f what G oitein term ed as India trade’ and 'India
trader’ has been furnished by the famous Jewish business letters
from the Cairo geniiah (G oitein 1967, 1973 ’ J98o, 19 8 7).4
These different types o f sources may be introduced under the
category of, following L. Udovitch, qualitative data, as distinct
from quantitative (Udovitch 1970).
O f late attention is being increasingly paid to archaeological
evidence concerning ancient sea-borne contacts among Indian
O cean countries. Artefactual analyses and especially, the study
o f pottery have proved to be o f considerable help in tracing the
possible networks o f com m unication and movements o f men,
m erchandise and ideas. M arine archaeology as an academic
S e a f a r in g s , Sh ips a n d Sh ip O w ners 33
discipline is still in its infancy in India, though its progress is

encouraging. T h e marine archaeologist in India is still looking
for the first trace o f an ancient ship wreck, the in-d ep th
archaeological enquiries o f which have immensely enlightened
on shipping and navigation in the Atlantic and the M editerra
nean. Ethnographic accounts o f traditional vessels, shipping
and navigation in India going back to early m odern times are
now also being tapped to throw light on the Indian sea-faring
in the Indian O cean. T h e sources for the study o f India and the
Indian O cean p rior to 1500 are as scarce as scattered, both
spatially and tem porally. O n ly a careful com bination o f
different types o f evidence may offer some scopes to solve a
few o f the mysteries o f the jigsaw puzzle.
Th e early Indian perceptions about the seas on the three
sides o f the subcontinent were far from specific. Early Indian
literature, however, always highlighted the m aritime space to
the south o f the tip o f the peninsula as a feature o f the
configuration o f the country called Bharata or Jambudvipa.
T h e two m ajor segments o f the Indian O cean, viz. the Arabian
Sea and the Bay o f Bengal were simply designated by early
Indian writers respectively as the western (pascima) and the
eastern (purva/prak) seas. Pliny (Naturalis Historia, V I.X X L 56)
seems to have been the first to have designated the term mari
Indicum, literally m eaning the Indian sea.5 As Pliny died in AD
79 this expression must have gained currency in the Classical
world around the last quarter o f the first century A D . In the
medieval A rabic and Persian texts on travel and geography the
m aritime space was labelled as al bahr al Hindi (the Indian Sea)
and D a iy a -i-A kh za r (the G reen Sea). Two segments o f al bahr al
Hindi figured prom inently in such accounts: bahr Larvi (the Sea
o f Lar), i.e. the present A rabian Sea and bahr Harkal (the Sea o f
Harikela), i.e. the Bay o f B engal.6 As early as the m iddle o f the
second century A D , Ptolemy ( 1 9 3 2 ) had already labelled the
eastern sector o f the Indian O cean as the Gangetic G ulf. A
copper plate inscription o f 97 1* found near Dhaka, refers to
Vangasagara (the sea o f Vanga), providing thereby the earliest
indigenous base o f the name Bay o f Bengal (Chakravarti 1996c,
1998b). T he Chinese, in spite o f their distinct interests and
regular presence in the Indian O cean area, designated it

merely as the Western O cean, H si-hai (Ray 1987).
Academ ic attention to India’ s seafaring experiences in the
Indian O cean area does not have a long tradition and it is only
in recent times that scholars o f different disciplines have shown
an awareness o f the im portance o f this subject. T h e lack o f
adequate inform ation must have acted as a deterrent to
academic inquiries. But no less significant was the perception
that Indians were traditionally averse to seafaring activities.
This was because traditional Indian way o f life was largely
projected as rooted to inland, rural and agricultural activities;
trade and seafaring, which are often associated with an
adventurous, spirited and innovative society, were considered
marginal to traditional Indian experience. To this was added
the notion that crossing the sea was a forbidden act (Kalivarjya)
in the strict and seemingly invariant Indian sastric code
(Dharmasastras or legal treatises).7 So, if any historical in form a
tion on early Indian seafaring was available it was conveniently
explained as a result o f some external (preferably Western)
stimulus. A typical example o f such a scholarly stand will be
evident in the following quote from a celebrated historian o f
the trade between the Rom an em pire and India during the
early centuries o f the Christian era.
T h e moving force from the first to the last came from the
West, the little changing people o f the East allowed the
West to find them out. We have then, on one side, India
o f the O rien t, then as now, a disjointed aggregate o f
countries, but without the uniting force o f British rule
which she now has, and while open to com m erce, content
generally to rem ain within her borders to engage in
agriculture. (Warmington 19^8/1974:
Such a historiographical position was countered by culling

data from early Indian popular and creative literature (i.e.
beyond the sastric texts) offerin g narratives o f seafaring
activities o f Indians. Th ou gh many o f the textual gleanings
were actually myths and legends, they nevertheless drove home
the message that early Indians were not necessarily restricted by
S e a f a r in g s , Sh ips a n d Ship O w ners 35
the sastric dicta forbidding voyages across the seas. Scholarly

attention was further drawn to the Yuktikalpataru o f Bhoja, an
eleventh century treatise, for throwing lights on ancient Indian
ship-building techniques (M ookerji 1957 T h e m ajor
critique o f the dom inant European perception o f early Indian
seafaring or the lack o f it was provided by a num ber o f Indian
nationalist historians. Though Indian nationalist h istoriogra
phy sometimes exaggerated the extent and range o f early Indian
seafaring out o f proportion, it nevertheless established that
ancient textual accounts could enlighten the historian o f early
Indian maritime activities. Subsequent searches for the literary
evidence o f early Indian seafaring m ore or less followed similar
m ethodologies but with a m ore critical approach and less
nationalistic fervor (Basham 1967; Gopal 197° ; ^ ao I 97 °)-
These were marked by the utilisation o f Classical and early
A rabic and Persian notices o f early Indian seafaring. The
literary impressions o f early Indian seafaring are, however,
meagre and lack the specificity o f the descriptions o f actual
voyages. Such a problem has to some extent been offset by the
in-depth study o f early Indian maritime iconography. A
pioneering effort in this direction was launched by M ookerji
(1957), but a m ore critical, systematic and m ethodical approach
to the study o f visual representations o f early Indian riverine
crafts, coasters and sea-going vessels has considerably enriched
our understanding o f the subject (Deloche 1987, 1994 * ^996;
Schlingloff 1988). To this can be added the growing literature
on the ethnography o f traditional ship-building practices in
different regions o f South and Southeast Asia. It often
combines the ethnographic accounts o f indigenous vessels, as
noted by early m odern European observers, with data on the
traditional boat/ship building designs and technologies based
on recent field works (H ornell 1918—2 3 ; M cG rail and K entley
1985; M anguin 1986 and Arunachalam 1987, 1996).
In m aritim e archaeology some advancem ent o f our know l
edge is claim ed on the basis o f underwater explorations at
Dwaraka in the Saurashtra coast in western India and
Poom puhar, the latter identified with the fam ous harbour
and city o f K averipattinam in ancient Tam il texts and
Khaberos em porium in Ptolem y’ s Geography in the Kaveri delta

in Tam ilnadu (Rao 1988). O u r appreciation o f the brisk
m aritim e m ovem ents in the Indian littorals in the early
centuries o f the C hristian era has considerably im proved,
thanks to new translations o f the Periplus o f the Erythraean Sea, a
first century AD logbook type o f w ork by an anonym ous G reek
m ariner/sailor (H u n tin gfo rd 198O; Casson 1989). Th is has
paved the way for a reassessment o f Classical notices o f India
and the Indian O cean in the light o f m ore m ature study o f
th e-th en nautical technologies and sailing patterns (Casson
1980, 1988a, 1988b, 1992; Rom anis 1997 )- N o less sign ifi
cant is the attempt to com pare, supplem ent, juxtapose and if
possible, correlate field archaeological data on m aritim e
activities with ethnographic studies and textual im pressions in
the context o f India and the Indian O cean p rio r to AD 1500
(Ray and Salles, 1997 )-
That seafaring in the subcontinent was already well estab
lish e d d u rin g the days o f the H ara p p an c iv iliz a tio n
(c. 2 6 0 0 —1800 B e ) is well established in archaeological
researches (Ratnagar 1981, 1994 ; Tossi 1992). T he Mauryan
em pire (c. 324“ 187 BC) is known to have m aintained extensive
contacts with G reek kingdoms in West Asia; such contacts could
have been at least partly m aritime and facilitated by the Maurya
occupation o f the Gujarat, Kathiawad and K onkan coasts
(Thapar 1987). This is the period when the Persian G u lf
em erged to considerable prom inence as a m ajor sea-lane in the
western Indian O cean. This is clearly evident from recent
excavations in the G u lf region in particular at Failaka near
Kuwait as the key site (Salles 1987> 1996)- From late first
century BC to about the m iddle o f the third century AD as an
outcom e o f its brisk trade with the Rom an em pire South Asia
began to figure in a very prom inent way in the international
maritime network, with a shift from the G u lf to the Red Sea as
the m ajor sea-lane in the western Indian O cean. T he history o f
this com m erce is an aspect o f Indian O cean maritime affairs
that has been well researched into principally on the basis o f
the Periplus o f the Erythraean Sea, the N atural Histoiy o f Pliny and the
Geographike Huphegesis o f Ptolem y.8
Se a f a r in g s , S h ips a n d Sh ip O wners 37
In the Periplus’ s account o f Barygaza (m odern-day Broach

on the m outh o f the Narmada in Gujarat), the prem ier port in
coastal western India during late centuries BC and the early
centuries o f the Christian era, the author notes the extreme
difficulty faced by non -indigenous vessels to reach the port.
T h e ruler th ereof employed local crafts, trappaga and kottymba to
guide foreign vessels safely to the harbour. T he trappagas used to
frequent 'as far as Syrastrene (= Saurashtra, in the southern
part o f the Kathiawad peninsula) to meet vessels and guide
them upto Barygaza’ . (Casson 1989* section 44)
The trappaga and perhaps also the kottymba were therefore crafts
meant for coastal voyages and apparently capable o f undertaking
fairly long coastal cruises. The indigenous character o f these two
vessels have been established by their m ention as tappaka and
kotumba in the Jaina text Angavijja. These have been classified in the
said text as vessels o f the middle category, larger than small crafts
like kattha and velu, but smaller than the pota, ship with greater
space (mahavakasa) and suitable for high sea voyages.
A n inscribed seal impression (palaeographically assignable to
c. third century a d ) from Chandraketugarh (near Calcutta)
shows the figure o f a ship, labelled as trapyaga (i.e. trappaga).
Chandraketugarh, a well known excavated site bearing features o f
a riverine port in the Gangetic delta, is often sought to be
identified with the port o f Gange o f the Classical texts. This and
other seals/sealings which are discussed below are inscribed in
the mixed Brahm i-Kharosti script, according to B .N . Mukherjee
(Mukherjee 199 °)• These are mostly found from the south
western part o f present West Bengal, India. The sealing in
question gives us for the first time some clues to how the trappaga/
trapyaka could have looked like or at least how it was visualised by
the artist/designer o f the seal (fig. 2 .l ). The left half o f the hull
is longer, raised to a considerable height and also probably
pointed at the end. O ne can also notice at least three vertical and
parallel bands between the port and starboard sides o f the vessel.
This may be an attempt to show the beams o f the vessels. A t the
centre o f the vessel stands a single mast with a tripod base and the
stylised figure o f a flag atop. Near the top o f the mast may be
seen a rectangular object from which a pair o f parallel ropes
Ra n a b ir C hakravarti
Figure 2.1 Trapyaga vessel on an inscribed seal

(Chandraketugarh, West Bengal, India)
Figure An inscribed seal showing a ship with an oar

2, . 2,
S e a f a r in g s , S h ips a n d S h ip O wners 39
come down on each side o f the mast. These may be tentatively

suggested as the ship’ s riggings. But no sail is depicted. A close
perusal o f the sealing shows the figure o f a horse on the right
hand field and the right edge o f the ship. The horse is depicted
in profile with its head towards the mast; its mouth, an eye and
an ear are also clearly visible. The sealing provides the earliest
evidence o f the shipment o f the horse in the context o f an early
Indian harbour.
T h e same site has yielded representations o f several other
vessels. A vessel with a single mast in the foredeck is depicted
with the bow and the stern both curved upwards. A n oar is
fitted to the raised stern. T he legend on the seal suggests that it
was a royal ship belonging to Juja, the conquering king (fig.
2 . 2 ). T h e third specimen from Chandraketugarh has the figure
o f a ship with the bow and the stern both raised upwards; the
stern is fitted with an oar. A tripod mast stands on the foredeck
and atop the mast is flying a banner. A bout the three quarter
height o f the mast is a rectangular object with four holes, two
on each side o f the mast. It is difficult to identify this object. A
recent com m unication with J. Deloche suggests that it may
represent a furled lateen sail. T h e inscribed legend, palaeo-
graphically assignable to the third century AD, suggests that it
was fit to undertake journeys in 'three directions’ (tridesayatra) ,
i.e. to distant destinations (fig. 2 .3 )*
T he fourth specimen (fig. 2 .4 ) from the same site shows the
figure o f a ship with a single mast on the foredeck and clear
representations o f the beams. A t the centre is depicted a rather
disproportionately large basket containing stylized stalks o f
grain. T h e accom panying inscription labels the vessel as
jaladhisakra (literally Indra o f the ocean; Indra is the king o f
the Brahm anical gods). Such a name may imply that the vessel
was fit for making overseas journeys, and therefore, should be
differentiated from a coaster. T he depiction o f the vessel shows
that the stern is fitted with a steering mechanism which does
not resemble an oar, seen on other representation o f vessels on
the seals from Bengal. T h e steering mechanism is represented
by a vertical straight line which is crossed at the centre by
another long horizontal line (Chakravarti 199 6 b ).9
Figure 2-3 Ship on a terracotta seal

Figure 2-4 Ship on a terracotta seal showing a stylized stalk of

grain in a basket and a steering mechanism fitted to the stern
S e a f a r in g s , S h ips a n d Sh ip O wners 41
Figures o f ships also appear in the 'ship type’ coins o f the

Satavahana rulers o f the Deccan contem porary with early
centuries o f the Christian era. They show double masted ships
with a thick wad visible at the mast head. D. Schlingloff
(Sch lin gloff 1988; also D eloche 199^) suggests that the thick
wad could have represented the furled sail o f the vessel. These
ships, different from the ones found on the seals from Bengal,
appear to have been used along the eastern coast o f peninsular
India. A recent perusal o f ancient Tamil literature has brought
to light at least eighteen terms to suggest the rich varieties o f
indigenous vessels in south India. The largest o f these South
Indian vessels was called matalai (also kalam ), generally used for
long-distance voyages (Rajamanickam and A ru l Raj 1994 )- The
large sized vessel called colandiaphonta in the Periplus was, in the
opinion o f Lion el Casson, a south-east Asian ship frequenting
the eastern coast o f south India. It is significant to note that
such large ships are described in the Periplus in the context o f the
country called Masalia/Maisolia, corresponding to the area
around present Masulipattanam in Andhra Pradesh, India. In
this country was situated, according to Ptolemy (1932) an
aphaterion or a point o f departure for ships bound for Chryse
C hora and Chryse Chersonesis (i.e. referred to as Suvarnab-
hum i and Suvarnadipa in early Indian literary sources and
located in South-East A sia).10
It is apparent that in the discussions on the seafaring
traditions o f India there is a considerable gap for the period
from c 700 to I 5 0 0 - This is a phase which has not yielded
much archaeological data and nor any m ajor quantifiable data.
Yet this is the phase which witnessed the rise o f the Indian
O cean as a m ajor zone o f maritime contacts. T he European
docum entation since the early sixteenth century cannot but
leave a strong impression that there had already evolved a
distinct pattern and ethos o f trade in the Indian O cean before
the advent o f the E uropean s.11 T h e very arrival o f the
Portuguese and subsequently other European nations in the
Indian O cean trade did not immediately signal a collapse the
established seafaring traditions. Instead the European activ
ities, though bringing in significant changes, were accom m o
42 R a n a b ir C h akravarti
dated within the pattern o f the Indian O cean trade for another
century. That is why, we would try to examine here the
fragm ented, meagre and scattered evidence at our disposal to
offer some suggestions on the seafaring activities during those
crucial eight centuries from c. J O O to l5 0 0 .
T h e period stated above is well known in the cultural history
o f India fo r the com position o f lexical literature not only in
Sanskrit, but in several regional languages too. These lexicons
and their respective com m entaries were m odelled on the
famous Sanskrit lexicon, Amarakosa, datable to the sixth-seventh
centuries A D . These texts generally include a chapter on
different types o f water crafts and different parts o f such a
vessel. A m ore elaborate classification o f vessels is available in
an eleventh century Sanskrit text, the Yuktikalpataru o f Bhoja.
T h e follow ing discussions are made largely on the basis o f the
Yuktikalpataru and different lexical literatu re.12 T h e Yuktikalpataru
makes a broad classification o f vessels into two: ordinary
(samanya) and special (visesa). T h e latter term actually denotes
the sea-going vessel, while the form er should logically stand for
the riverine craft. T h ere are ten sub-types o f the riverine craft.
T h e special or the sea-going craft is further subdivided into two
groups: long (dirgha) and tall (unnata). T h e 'lo n g ’ variety has ten
further sub-types and the 'tali’ five sub-groups. How far such
classifications corresponded to the actual types o f water crafts
cannot be ascertained and hence is not elaborated here.
Literary creations o f the early medieval period rarely refer to
such classificatory terms for water crafts as laid down in the
Yuktikalpataru. A seagoing vessel, on the other hand, is frequently
term ed in literature o f the post 6 0 0 AD times, as ja n a p a tra
carrying samyatrikas or passengers (Gopal 1970). A scrutiny o f a
few Ilth century inscriptions from the K onkan coast in the
western littorals o f India demonstrates that the pravahana mainly
plied on coastal routes along the western sea-board. It was
clearly differentiated from the vahitra which is described as
being used for dvipantara voyages, or journeys across the high sea
(Chakravarti 1998a, 1998b). That the vahitra was em ployed for
voyages from and to distant areas is in fact recorded in an
inscription o f AD 592 from Kathiawad in Gujarat (Sircar 1953 ;
Se a f a r in g s , Sh ips a n d Sh ip O w ners 43
K osam bi 1958). T he textual image o f vahitras plying near the

famous port o f Tamralipta in the Bengal littorals is also offered
by the Dasakumaracarita o f D andin belonging to AD 7th century
(Gopal 1965)- It may therefore be reasonably inferred that the
vahitra was in use in both the Arabian Sea and the Bay o f Bengal
areas for high-sea voyages at least since the late sixth century
AD.
T he image o f the diversity o f sailing crafts in early medieval
texts and inscriptions is further supplem ented by the visual
representation o f vessels in m em orial/hero stones (viragals),
com m em orating the death o f persons by drowning and/or in
sea-battles. We owe our understanding o f these iconographic
delineations o f such water transports to the studies o f Jean
Deloche (Deloche 1987, 1996)- The well known panels o f
representations o f boats in the caves at Borivli, near Mumbai,
call for our attention. Several lines o f vessels are shown in a
battle scene. These are all crafts made o f planks which are
stitched together probably with coconut coir. T h e long
projected bow and the gunwale o f the sharp ended boat are
also clearly depicted. Also shown are oars which come out o f
the single level holes, cut below the gunwale, but only the shafts
o f the oars are visible. D eloche considers that there was possibly
a gap o f I meter between each hole and opines, on the basis o f
the num ber o f oars shown, that the hull was 13—16 metres in
length. The sculptures are thought to have portrayed a battle
between the Silaharas o f north K onkan and the Kadambas o f
Goa; it is generally ascribed to the twelfth century.
A num ber o f viragal, depicting sea-battles, are now in the Goa
Museum; these have also been studied by D eloche. A ll the
sculptures from Goa have been placed on stylistic grounds in
the tw elfth-thirteenth century. O ne m em orial stone shows a
double-ended and banana shaped craft. A second variety o f
craft is represented by a canoe-shaped one with an axial
rudder; seven rowers facing backwards are also depicted. No
less interesting is another viragal on which is carved the figure o f
a broad vessel with a sharp stern (fig. 2 . 5 ) - That this is a sewn
vessel is also clearly visible in the sculpture. N ine oar holes are
shown. But the most interesting com ponent o f this vessel is a
Figure 2-5 A stone sculpture showing abroad vessel with a sharp stern and
a stern-post rudder (Goa Museum, Goa, India: after Jean Deloche)
stern-post rudder which is prom inently portrayed. T ech n o lo

gically speaking, the stern-post rudder may be considered to
have marked an im provem ent on the axial rudder, depicted in
another specimen, and also on the steering paddle. M oreover,
this vessel is shown not afloat, but appears to have stood on
three log-like cylindrical objects. T h e possible intention o f the
sculptor is to show a vessel being lowered into the water. If this
suggestion is accepted, the figure can well be taken to depict a
ship in a dockyard where it could either have been built or
repaired.13 Literary accounts o f ships and voyages and visual
representations o f ships amply demonstrate the lively nature o f
Indian shipping traditions during the early middle ages.
From the seventh century AD onwards Indian coasts,
particularly the western sea-board, were regularly touched by
diverse types o f vessels from the Persian G ulf, the G u lf o f
O m an, the Red Sea and the east A frican coast. T h e rise and
spread o f Islam has a role in this traffic, as a m ajor impetus to
the increase in the sea-borne traffic certainly came from the
annual pilgrim age (hajj) to Mecca. W hile overland routes o f
com m unication and transportation definitely continued along
S e a f a r in g s , S h ips a n d Sh ip O w ners 45
with the m aritime voyages, the latter began to gain greater

prom inence. T h e research o f G oitein on the medieval Jewish
business letters strongly suggests the preference o f travellers for
sea to land routes not only during long journeys but also for
relatively short distances (G oitein 196*7: 275 )- T he Arab sea
going vessels, com m only grouped together as the dhow, became
the m ajor carrier o f goods, human beings and also ideas. The
nautical skills o f south-east Asia are perhaps best represented
by the prahu, im m ortalised in its sculptured depiction in
B orobudur, and the sampan. Farther east in China the Sung and
especially the M ing epochs left little room for doubt about the
marvels o f the huge Chinese ships, com m only labelled as the
junk. Ibn Battuta in the first half o f the fourteenth century
witnessed three types o f Chinese junks, the largest o f which
required a crew o f IOOO and were four stories high. However
Chinese attitudes to the sea were not consistent, and there was
the oscillation between closed and open door policies. L o n g
distance voyages became the order o f the day in the Indian
O cean zone from around the tenth century with its western
term ini at Siraf and al Fustat, and the eastern lim it marked by
the harbours o f China. As shipping in the Indian O cean prior
to the advent o f steam was largely guided by the m onsoon
winds, it was not possible for ships to complete a round journey
between the western and the eastern term ini in a single year.
Ports on both the sea-boards o f India, besides being known for
the availability o f coveted products, also fu n ction ed as
inevitable stopovers and gateways.
Seen from this point o f view, it is hardly surprising that
accounts o f voyages in large ships fit for overseas journeys
became quite regular from the tenth century (H ourani 195O.
T he Arab ships were generally branded as marakib; the term ja haj
also appears at least once in a genizah docum ent in the context
o f a journey from M anjrur (Mangalore) to al Dyyb (Diu) along
the western coast o f India (G oitein 1980). T he frequent
movements o f the jilab in the Red Sea with special reference to
the port o f Aydhab are prom inently referred to in the travels o f
Ibn Jubayr during the closing years o f the twelfth century
(Broadhurst, 1952). The jilab as a characteristic vessel o f the
Red Sea repeatedly and prom inently figures in the genizah

papers on the India trade and India traders (G oitein 1967).
A few types o f vessels plying in the western Indian O cean in
the early medieval times are m entioned in two Jewish business
letters (possibly dated in II 34 /II 36 ) which were sent from A den
to M anjrur, i.e. Mangalore in the south western coast o f India
(G oitein 1954 )- The principal content o f both the letters
concerns an invasion o f A den by the ruler o f Qais (i.e. Kish, a
small island in the Persian G u lf). T he said invasion also finds a
m ention in the accounts o f Ibn al Mujawir who died in I 29 1-
D uring this naval raid a fleet was employed that 'consisted o f two
big burmas, o f three shaffaras and ten jashujiyas’ (G oitein 1954 : 256 )-
The term burma literally means pots, but in this context actually
standing for large ships with roundish hulls. The three shaffaras
that also sailed along with the burma were smaller ships, generally
accompanying the bigger ones. G oitein’ s study o f numerous
other Jewish business letters may indicate that these ships,
though small, could still carry a weight o f 14 bahars or nearly
4200 pounds (one bahar= 300 pounds). The third type o f vessel
figuring in our letter were small boats (jashujiyat) which were also
put to use during the raid. This last category was small enough to
be carried on in the bigger ships and could be used as life boats
or as vessels for launching attacks. In this context G oitein draws
our attention to an observation o f al-Idrisi from second half o f
the twelfth century. While giving an account o f Kish, al-Idrisi
spoke o f the presence o f the meshiat at Kish. The meshiat is
generally identified with the mashwa or the machwa, figuring
prom inently in early m odern European accounts o f traditional
Indian vessels in the Guajart and K onkan coasts (Deloche 1994 )-
T here are occasional descriptions o f large sea-going vessels
in European accounts o f the early m odern times. Described by
Lewis as 'Indian-A rabic type’ (Lewis 1973 : 247 footnote 2),
these ships appear in the accounts o f Duarte Barbosa. Barbosa
inform s us about the voyages o f such ships from Mombasa to
Sofala and Cambay. That large ships o f the Malabar coast were
plying to and from the Red Sea was also noted by him .
A ccording to him , C alicut was noted for the construction o f
S e a f a r in g s , S h ips a n d S h ip O w ners 47
keeled ships o f a thousand and one thousand two hundred

baheres burden (about 35 ° —4 ° ° European tons). These
ships were built without any nails and the whole o f the
sheathing was sewn with thread and upper works differed
much from ours. They had no decks (Barbosa ig ^ 1 :76).
Irrespective o f the diversities in the types o f ships and vessels in

the Indian O cean zone they all were constructed without iron
nails. T he vessel in Sanskrit texts is invariably described as
bereft o f iro n for clam ping the planks (nilloham). T he wooden
planks were fastened together by coconut coir wound into the
shape o f rope. That is why traditional and indigenous vessels in
the Indian O cean maritime zone are described as 'sewn boats’ .
H ere lie two further factors o f comm onalty in the Indian
O cean shipping tradition. The two vital ingredients, viz. quality
teak wood for m anufacturing planks and coconut coir, were not
readily available in the thriving centres o f ship building in the
Persian G u lf and the Red Sea. Both the materials were
abundant in India which was regularly described in Arabic and
Persian texts as the principal source o f wood and coir rope for
ship construction. Arab and Persian texts on travel, geography
and maritime narratives (e.g. the Kitab Ajaib ul H ind o f Buzurg
ibn Shahriyar) are replete with the procurem ent o f these two
items from India to ship building centres in the western Indian
O cean. T he jila b , as Ibn Jubayr witnessed them at the Red Sea
port o f Aydhab, 'are sewn with a cord made from qinbar, which
is fibre o f the coconut’ . He also noted that the caulking o f the
jilabah was done with 'the shaving o f the palm trees’ . T h e vessel
was further smeared with grease or castor oil or the oil o f the
qirsh (shark), the last being considered the best material. 'The
purpose in greasing the boat is to soften and supple it against
the many reefs that are met with in the sea and because o f which
nailed ships do not sail through it.’ (Broadhurst 195 ^)
India witnessed a distinct developm ent o f local and regional
languages from c. the tenth century onwards. This resulted in
the growth o f lexical literature, m odelled on the famous
lexicon Namalinganusasana o f Amarasimha o f the seventh century.
The early medieval lexicons accom modated into them many
technical terms in local languages. A close perusal o f the

Desinamamala o f H emachandra or the Trikandasesa, for instance,
would indicate the coinage o f various terms in local/regional
languages concerning types o f vessels, ship-building, navigation
and m arine flora fauna. Such data, as and when systematically
collected and processed, could not only increase our existing
knowledge, but also bring in the flavour o f regional features in
the indigenous tradition o f shipping and navigation in the
different coastal segments o f the subcontinent. T h e creative
literature in Sanskrit and the lexical texts in Sanskrit and other
regional languages, however, appear to have been fam iliar with
terms denoting different parts o f a vessel. T he planks which had
to be sewn together were term ed as phalakas in Sanskrit and they
were to be clung to as the last resort by victims o f shipwreck.
T h e mast is variously called kupastambha/kupadanda/gunavrksa in
early Indian texts. Th ou gh textual notices o f the sail (sitapatta)
abound in early Indian literature, visual representations o f
vessels rarely show early Indian ships with sails (a m ajor
exception being the ship figure in the Ajanta painting). T h e
celebrated Arab navigator ibn Majid offers lengthy discussions
on the sail and interestingly refers to the 'Indian sail’ (al-q u lu al
hindiya; Tibbetts 1971 ^71) * T h e Sanskrit terms rajju and sutra stand
for ropes probably used for rigging; the word aritra stands for
oar and mandira corresponds to cabin in the vessel. That the ship
carried an anchor is evident from the use o f terms like nangara
(anchor), nangarasila (anchor stone) in Sanskrit and anjar in
A rabic. T h e use o f the grapnel is further illustrated by the
famous representation o f the al H ariri ship.
Indian O cean navigation was influenced by predictable
alteration o f the m onsoon. T h e southwest m onsoon wind
blowing from June to Septem ber was ideally suited for an India
bound voyage from the west, while the return west bound
jou rn ey could be facilitated by follow ing the north-east
m onsoon wind system. The early notice o f this wind system
from the western point o f view is seen in the Periplus o f the
Erythraean Sea and the N atural History o f Pliny. Variously called the
Hippalus and Etasian w ind, its intelligent utilisation trans
form ed the nature o f shipping and navigation between India
and the West.14 A ccording to Pliny, a ship starting from the Red
Sea port o f Berenice or Myos H orm os could reach the Malabar
harbour o f Muziris in forty days. A recent examination o f these
accounts has led Casson to suggest that the voyage could reach
Malabar from the Red Sea area in as few as twenty days (Casson
1980). T h e graphic details o f voyages following the north-east
m onsoon wind are available in the accounts o f Fahsien, the
Chinese pilgrim who sailed from the port o f Tamralipta in the
Bengal coast to Sri Lanka in a merchant vessel (Legge I 97 1)-
T h e word m onsoon is derived from A rabic mawsim m eaning a
fixed time in the year. It is used in plural as mawasim by Ibn
Majid who understood by the term the sailing seasons or dates.
He laid down detailed instructions o f ideal dates for sailing
from particular harbours to definite destinations. He also
strongly advised against sailing from and to a particular port
after the final dates for voyaging were over. Missing the
prescribed last date o f sailing or waiting for the m onsoon winds
to blow in the desired direction often resulted in prolonged
stop at a particular port in the Indian O cean (Tibbetts I 97 1)*
Thus Ibn Battuta waited for no less than three months at the
port o f Calicut in K erala to set sail for China. A close scrutiny
o f the Periplus and the text o f Ibn M ajid has led Casson to
conclude that most o f the Indian harbours were closed at the
height o f the southwest m onsoon (Casson, 1988b). These were
preferably reached from the western Indian O cean ports in the
m onths o f A ugust/Septem ber when the ferocity o f the
m onsoon winds and torrential rain would somewhat die down.
Th e key figure on board o f the traditional vessel was the
captain, distinct from the ship owner. He is known as the sasaka
(literally, the adm inistrator) and niyamaka (literally, one who
disciplines) in Sanskrit and Pali texts. A n expert in the art o f
navigation (naupracaravidya) and especially in bringing the ship
into (aharana) the port and taking it out o f the same (apharana),
often through difficult and treacherous passages, he should be
in total com m and o f the ship. O ne comes across a Indian
mahanavika or a master m ariner, named Buddhagupta in an
inscription, palaeographically assignable to AD sixth century,
from the Malay peninsula (Sircar 1965: 4 9 ^)-15 A ccording to
the inscription, he was a resident o f Raktamrttika, usually

placed in present West Bengal, India. T h e reasonable inference
is that the master m ariner undertook voyages between the
Bengal coast and the Malay peninsula. His epithet certainly
shows that he was no ordinary sailor (navika), but probably o f the
same rank and experience as that o f a niyamaka o f early Indian
texts or o f the muallim or navigator in the manual o f Ibn Majid.
T h e pivotal figure in the navigational manual o f Ibn M ajid is
the muallim or the pilot/navigator o f the vessel. T h e art o f the
pilot or navigator (m uallim ) is called siyasat. He is to m aintain the
ship in an impeccable m anner, must exercise his full control
and authority on his subordinates, but m oreover should be
firm with passengers (especially merchants) on board the ship
w ithout, however, h urtin g their prestige (Tibbetts I 97I:
3 8 7 -9 1). The muallim was in charge o f the correct fitting o f
the ship, inspection o f the gear, the stores and the supervision
o f the loading. He wielded immense authority over the crew
(a l-a sk a r). 'O n the muallim rests the governm ent o f the ship’ ,
according to Ibn M ajid. Between the muallim and the sailors
stood the tandil, the ch ief o f the sailors (Tibbetts I 971)*
Ibn Jubayr knew the pilots by the term rubban (Broadhurst
1952), who in the manual o f M ajid was viewed as a coasting or
localised pilot in the Red Sea and never in the same status as
the muallim (Tibbetts I 97 I: 60). T h e Jewish business letters,
relating to India trade, generally but not uniform ly distin
guished the ship-ow ner from the captain who is called therein
rayyis. O ccasionally the term nakhoda is also used to denote the
captain o f the ship, as was Buzurg ibn Shahriyar, the famous
author o f the Kitab A jaib ul H ind (Freem an-G renville 1982).
Textual impressions from early Indian sources suggest a
num ber o f subordinates o f the captain or the navigator o f the
ship. T h e niyamaka or the mahanavika should be assisted by the
karnadhara (helmsman), the rasmigrahaka (literally, holder o f the
string, i.e. one who m anouevred the riggings), datragrahaka (the
sickle-holder, ie, the crew who would cut the ropes with a sickle
during a storm) and utsecaka (one bailing out the excess water
from the hold o f the craft). T h e crew o f an ancient Indian vessel
is often term ed navikakarmakara in early medieval inscriptions.16
Se a f a r in g s , Sh ips a n d Sh ip O w ners 5i
A key figure in the movement o f ships in the Indian O cean

area was the ship-ow ning merchant. T he earliest known ship
owner in an Indian context is Hovajino who is said to have
owned the trappaga vessel, the figure o f which was depicted on an
inscribed seal from Chandraketugarh, datable to the third
century AD (M ukherjee 199 ° : 4 7 5 Chakravarti 1992 )- Since
IO O O AD available sources began to offer m ore inform ation on
a distinct type o f merchants, the ship-owners, whose active
presence is seen in western Indian littorals. Indian inscrip
tions, particularly from the K onkan and Gujarat sea-boards,
provide us with interesting impressions about nauvittakas. The
term nauvittaka denotes one who earned wealth by ships, i.e. by
owning ships. Epigraphic evidence o f the late eleventh century
from southern K onkan leaves little room for doubt that at the
early medieval port o f Balipattana (near m odern Kharepatan,
Distirict Ratnagiri, Maharashtra) there existed a flourishing
family o f nauvittakas or ship owning merchants. T hree genera
tions o f this family o f nauvittakas rem ained active at the
Valipattana and owned the pravahana type o f coasting vessels
which plied along the whole o f the K onkan coast. This family
o f nauvittakas appears to have employed sailors or navikakarmakaras.
(Chakravarti 1986; Chakravarti 1998a). Inscriptions from Goa
(from the late tenth to the thirteenth centuries) also throw
some light on local nauvittakas, active at the port o f Candrapura
(Sindabur o f the Arab authors), located to the south o f present
Goa (Chakravarti 1998). C ontem po rary A rab ic texts on
geography and travel and the geniza documents on 'India trade’
are replete with references to nakhuda, lord o f the nau or ship. In
other words, nakhuda also meant the ship owning m erchant and
corresponded to the nauvittaka o f the Indian sources.
The intense com m ercial activities in the western Indian
ocean since IO O O AD led to the regular presence and
interactions am ong Indian and A rabic ship owning merchants.
This explains why in Indian inscriptions from the coastal areas
o f Gujarat and K onkan the two terms were simultaneously used
to refer to ship-owners in these coastal segments. T h e two
terms nauvittaka and nakhuda were abbreviated respectively as nau
and nakhu in inscriptions and were also used therein as
interchangeable terms. This practice o f abbreviating the two

terms in official and/or royal records would strongly suggest
the fam iliarity o f the coastal society about the ship owners.
T h e medieval Jewish letters are m ore explicit regarding the
ship owners and their activities. Individual ship owners
p rom inen tly figure in these docum ents. T h ey were fabulously
rich and glo rified fo r their lavish donations in cash. In 1130
M adm un, a very p ro m in en t figure in the geniiah papers and the
trustees o f m erchants at the port o f A den, wrote about a ship
ordered by him for carrying com m odities and passengers
fro m A d en to Sri Lanka covering a distance o f 2 IOO miles
(G o itein 1973 : f ^ ) - Th at this vessel rem ained lon g in
operation and must have served its owner adm irably, w ould be
amply dem onstrated by references to its regular voyages in
1141 and I I 49- In ^ e letter o f 1141 Nahray b. A llan , a very
p ro m in en t 'India trad er’ m entions the 'ship o f the elder
M adm un’ in which he sent his goods (G o itein 1973 : 2 0 0 ).
A braham Y ish u , a Jewish m erchant from al Mahdiyya or
Tunisia who lived at the port o f al M anjrur (=M angalore on
the south western coast o f India) from II32 to I I 49> wrote a
letter in 1149 to his younger b ro th er from A d en , where he was
at that tim e on his way hom e. T h e letter inform s us that he
had earlier sent for his brother, who was in Egypt, some gifts
in the ship o f M adm un. It appears from the content and
context o f the letter that at the time o f sending the presents
fo r his b roth er A braham Y ish u had still been at M angalore. In
other words, the vessel owned by M adm un seems to have
continu ed to be in service in the M alabar-A den run fo r at
least two decades (1130—1149/50)-
Also o f relevance is a letter o f 1139 (G oitein 1973 : 185—9 ^) *
This particular docum ent contains considerable material on
ships and their respective owners. T he letter was written by
K h alaf b. Isaac b. Bundar, a cousin o f M admun, to Abraham
Yishu, in M anjrur or M angalore. K h alaf b. Isaac wrote:
I took notice, my master (i.e. Abraham Yishu), o f your

announcem ent o f the sending o f refurbished iron in the
boat o f the nakhoda (shipowner) Ibn A b i '1 K ata’ib. the
S e a f a r in g s , Sh ips a n d S h ip O w ners 53
shipm ent has arrived and I received from him two bahars
and o n e-th ird , as you noted.
We are then inform ed o f a voyage by nakhuda Joseph from

Dahbattan or Valarapattanam in Malabar to Aden, not in his
own ship, but in the ship o f Ibn al-Muqaddam. Thus one
encounters here two ship owners; it is also clear that the nakhuda
him self undertook sea-voyages, at least occasionally and was not
always a ship owner based ashore. The letter continues to tell us
that two ships o f one Fatanswami (i.e. Pattanasvami, literally meaning
the lord o f a port) were bound for the port o f Aden. The smaller
ship reached A den duly with all its cargoes intact. But the bigger
one foundered at Bab-el-M andeb. T he pepper therein was
completely lost, but portions o f iron carried by the ship could be
salvaged by employing professional divers from the port o f
Aden; there was also a charge paid for the service rendered by
these divers. The term Fatanswami is an arabicisation o f Sanskrit
Pattanasvami. The epithet Pattanasvami as the lord o f the port is
associated with the names o f a num ber o f merchants in several
epigraphic records from early medieval south India (Appadurai
1936). T he Jewish trade letter o f 1139 clearly demonstrates that
the Pattansvami/Fatanswami in question owned sea-going vessels. The
ship o f Fadiyar under the command o f Ahm ed is also m entioned
in the same letter as having duly arrived at A den from Malabar.
The ship carried some presents for Abraham Yishu from K halaf
b. Isaac b. Bundar. G oitein s study enables us to know that the
Fadiyar, which was a title like Fatanswami, was an Indian ship owner.
In fact several vessels o f his are m entioned in the geniza papers.
The inform ation regarding the Fatanswami and the Fadiyar is a clear
indicator o f the fact that at least some o f the Indian ship owners
sent their ships to distant overseas destinations in the western
Indian Ocean.
T he considerable wealth o f the nakhuda is well driven home
ship owner Ramisht o f Siraf whose ships touched not only
harbours o f coastal western India, but also A den at the m outh
o f the Red Sea (Stern 1967)- T he Jewish business letter
describing the invasion o f A den by the king o f K ish speaks o f
the arrival o f two ships o f Ramisht from the Persian G u lf port
o f Siraf to A den (G oitein 1954 )- Abraham Yishu who lived in

Mangalore from 1132 to 1149 was inform ed from A den that two
o f Ramisht’ s ships were com pletely lost, yet it did not prove
catastrophic for the Sirafi nakhuda (G oitein 1954 )* T h e wreck o f
the two ships o f Ramisht is also figuring in a letter, dated to late
1130s, written by Joseph b. Abraham to Abraham Yishu.
Abraham Y ishu probably sent some goods in the two ill fated
ships. Joseph b. Abraham consoled him: 'D o not ask me, my
master, how much I was affected by the loss o f the cargo
belonging to y o u .’ (G oitein 1973 : J93 )
O n A p ril 2 4 * II 4 I Nahray b. A llan wrote on the eve o f his
departure from the Red Sea port o f Aydhab to A den that he
had boarded the ship o f al-D ibaji (G oitein 1973 : I9^). T h e
term al-D ibaji, explains G oitein, stands for a merchant dealing
in brocade. T he owner o f the ship was therefore originally a
brocade textile m erchant. He must have becom e wealthy
enough to have owned a vessel meant for a voyage to India,
yet retained the epithet signifying his previous occupation.
Nahray b. A llan seemed satisfied that he managed to get a bilij
on this ship. T he term, as pointed out by G oitein, occurs
frequently in the genizah papers and stands for some kind o f a
cabin. It was actually a space screened o ff by mats and there the
passengers could keep their wares, sleep and also cook. This
was, o f course, far from com fortable. Arabic and Persian texts
would repeatedly narrate that extremely tough situation for the
passengers on board. They had to sleep over or beside their
merchandise which were often badly affected if water seeped
into them through the covers. If there were animals in the ship
they shared the same space with the hum an passengers. A bdur
Razzak in the fifteenth century left behind a pen picture o f the
conditions o f passengers on a ship which also carried horses;
these were regularly shipped from Arabia and Persia to India
(M ajor 1900). Ibn Battuta had sufficient reasons to be full o f
praise for the Chinese junks as these had private cabins for
im portant passengers like h im self (Lewis 1973 )- Indian
inscriptions and the genizah documents would further show that
influential and wealthy ship-owners could sometimes assume
administrative functions.
A n oth er ship-owner, named T inb u , figures in a Jewish

business letter o f 1145* written by Mahruz b. Jacob to his
brother-in -law , Judah b. Joseph h a-K o h en (G oitein 1973 :
62—65)- M ahruz was him self a nakhuda or a ship-owner, who
often plied between A den and India, and his b roth er-in -law
was a representative o f merchants at the Egyptian capital.
Mahruz in this letter was urging his brother-in -law , who had
the m isfortune o f being attacked by pirates somewhere along
the K onkan coast, to return at his earliest opportunity to
M anjrur or M angalore, where Mahruz had been waiting for
him, from Tana or Thana in the north K onkan coast. T in b u
seems to have been a ship owner based at M angalore and
'apparently a H in d u ’ (G oitein 1973 : 6 4 * footnote 2). A letter
from Mahruz was sent for Judah b. Jacob through nakhuda
T in b u who was on his way to Tana or Thana. T he em ergent
image from this letter is that T in b u ’ s shipping business
spanned from M anjrur to Tana along the K onkan littorals. The
whereabouts o f Judah were reported to Mahruz by a merchant,
Sheikh A b u ’l Qasim ibn Qattan (a cotton dealer). This clearly
highlights the inform ation network, though o f prim itive
nature, among merchants without which shipping and trade
in the Indian O cean zone would have collapsed.
Now briefly to the commodities shipped from and to India.
O n e o f the much cherished notions regarding maritime
transportation o f goods in the pre-m odern sea-borne trade is
that it was essentially based on trade in luxuries (van Leur 1955)*
It can hardly be denied that overseas trade in muslin, silk and
other textiles, exotic prestige goods and diverse spices, precious
metals and minerals, animals like horses and elephants used as
war machines created immense wealth and brought about
fabulous gains for a successful merchant. But ships in the Indian
O cean were also a carriers o f staples and bulk items, like iron
and copper, coarse cloth, dyeing materials like indigo, timber
and coconut coir for ship building17 and cowry shells. T he last
item is to be seen as small exchanges and ballast in the context o f
maritime commerce in the Indian O cean (Heiman 1990).
Maritime shipping among diverse com m unities, often widely
separated, brought about cultural exchanges and movement o f
ideas. T he Indian coasts are found to have been ideally suited

to the creation o f diaspora like settlements o f foreigners
(C u rtin 1984)- Thus a temple o f Augustus stood at the harbour
o f M uziris in Malabar, as m entioned in the Tabula Peutangariana.
A ncient epics in Tam il, the Silappadikaram and the M anim ekalai,
give us excellent images o f settlements o f ja v a n a (Roman)
merchants at Kaveripattinam , the leading port in Tam ilnadu
during the first four or five centuries o f the Christian era
(Cham pakalakshm i 1996). T h e archaeological rem ains at
Arikam edu near Pondicherry yielded the material proofs o f
such a Rom an trading station in the east coast o f India
(Wheeler and Deva 1946; Begley 1983; Begley and de Puma
1992). T he spread o f Buddhism and various Brahmanical
bhakti cults to South-East Asia is inseparably associated with
brisk and lasting maritime intercourse between India and that
region (Ray 1994-) • N o less significant was the contact between
India and the Arab Islamic world in the early medieval times.
Arab authors such as al Masudi spoke o f the settlements o f
O m ani, Baghdadi and Sirafi merchants at various ports in
Konkan. This is strikingly corroborated by contem porary
epigraphic records from the K onkan coast right up to Goa.
These records enlighten us on the presence o f Tajjika (Arabic)
merchants who appear to have been asked to be present at
im portant social and cultural events in coastal towns. No less
striking is the Sanskrit rendering o f the personal names o f such
Arab merchants, which implies their gradual integration into
the coastal society (Chakravarti 199 ° *» Chakravarti 1998a).
Th e genizah documents on the India trade are an eloquent
testimony on the spirit o f cooperation, tolerance and friendly
relations among Indian, Arab M uslim and Jewish merchants.
Th ere were obvious elements o f rivalry, com petition and
occasional quarrels among merchants, but there is no elem ent
o f animosity am ong different religious groups. We would like
to refer once again to the letter o f II45> telling us the
m isfortune o f Judah b. Joseph h a-C o h e n who fell a victim o f
piracy on the north K onkan coast. His b roth er-in -law sent
some m oney (21 mithqals or Egyptian dinars) for him through
the H indu nakhuda T in b u .
If my lord (i.e. Judah) you need any gold, wrote Mahruz,

please take it on my account from the nakhuda T in b u , for he is
staying in Tana, and between him and me there are bonds o f inseparable
friendship and brotherhood’ (G oitein 1 9 7 3 italics m ine).
T h e most illum inating illustration is the case o f the ship
owner (nakhuda) N uruddin Firuz who came from H orm uz to
Som nath, a famous H indu religious centre and a prom inent
port in Kathiawad in western India in 1264* A bilingual
inscription in Sanskrit and Arabic from Som nath tells us that
with the help o f a few local H indu elites (the most prom inent
o f them being Chada), the nakhuda procured some land,
constructed a mosque or mijigiti (i.e. Arabic masjid) there and
arranged for the provision o f various M uslim festivals and a
congregation (Sircar 1983)- T h e Sanskrit version o f this
inscription invokes A llah as Visvanatha (Lord o f the Universe),
Visvarupa (Universal), Sunyarupa (Aniconic) a n d Lakshyalakshya (O n e
who is at once Visible and Invisible). T h e remarkable attitude to
tolerance is further illustrated by the fact that N uruddin and
Chada, a local H indu land owner, were described in the said
record as dharmabandhava, i.e. righteous friends. 18
Two particular groups are known to us the least, the
fisherm an and the pirate. Piratical activities were not unknown
in the Indian O cean. Th ere were some notorious pockets in
the southern K o n k an and M alabar coasts where piracy
continu ed to be ram pant over centuries. T h e com m on
tendency was to rob a ship if it was driven by storm to an
unscheduled halt near a port, which was not its original
destination (Chakravarti 1989; Chakravarti, I 99 1)- Political
authorities occasionally curbed piracy and assured maritime
merchants and travellers adequate safety. Such an edict o f
security was issued by king Ganapati in coastal Andhrapradesh
in 1245 when he suppressed piracy at the port o f M otupalle
(Chakravarti 1995 )- But the piratical activities cannot be
considered as a foreru nner o f armed trade, initiated in the
Indian O cean zone with the advent o f Europeans in the
sixteenth century. The Indian O cean at the turn o f the
sixteenth century was widely traversed by and thoroughly
acquainted to merchants, preachers and sailors alike. There
was, therefore, no urge or thrust for exploration o f new lands

and for geographical discovery as one encounters in the history
o f contem porary Europe. No less im portant is the fact that
except for the C ola conquest o f Sri Lanka and Sri Vijaya in the
eleventh century (Nilakantha Sastri I 955)19 and the three
decades o f powerful official Chinese presence in the Indian
O cean under the leadership o f the great admiral C heng ho (or
Zeng he) from 1404 to 1433 (Mills 197°)* no political power
in Asia and A frica considered any part o f the Indian O cean as a
space to exercise and establish political mastery. T he literary
accounts and visual representations o f sea battles/raids are only
a handful in num ber and should be seen as exceptions to prove
the prevalent practice o f political powers being prim arily
interested in the revenue-bearing potentials o f the m aritime
trade. T here is little, if any, inclination on the part o f rulers to
exercise w ide-ranging dom inance over the Indian O cean as a
whole or over considerable parts o f it. T he traditional attitude
o f Indian and other Asian political powers to the sea was not
marked by any belligerent orientation to the m aritime zone.
The Indian O cean and segments th ereof aroused a sense o f
awe, marvel and fear among the rulers; but the maritime space
was rarely viewed as an arena for armed hostilities among
ambitious Asian powers. O ne cannot but agree with N eil
Steensgard that 'the principal export o f pre-industrial Europe
to the rest o f world was violence ’ .20
Notes
1 T h e term ’m aritim e cu ltu re’ was coin ed by the D utch social a n th ro
polo gist Prins (P rins 1 9 6 5 : 3) w^ ° defin ed culture as a pattern ed set o f
recu rren t events’ . H is concept was based on his study o f the L am u in
Islam ic East A frica. M aritim e culture, th erefore, may be looked fo r in a
given area/com m unity on the iden tification o f a recurren t set o f
m aritim e traits, which are not con fin ed to sh ip -b u ild in g and nautical
tech n ologies.
2 T h e chart p rod u ced by the Royal A u stralian Navy H ydrograph ic O ffice in
198 0 , en titled, 'A ustralia and the A djacen t waters: Lim its o f the O ceans
and the Se as’ , has created a So u th ern O cean, gird lin g the w orld aro u n d
the A n tarctica’ (D as G upta and Pearson 1987= 9 —i o ) . T h u s, this chart
S e a f a r in g s , S h ips a n d S h ip O wners 59
places the southern lim it o f the In dian O cean to the n orth o f the
Sou th ern Sea.
3 O n e may see in B ro eze’ s assertion an in direct and unstated critique o f
Pierre G h aunu who thought that the Atlantic in practice is the oldest
ocean on a h um an scale, the first that has been regularly crossed , the first
to be located at the heart o f an econom y, even o f civilization.’ (C h au n u
quoted in Das G upta and Pearson I 9 &7 : 8 ).
4 T h e term genizah is derived from the Persian word ganj, m ean in g a
storehouse or treasure. In medieval Hebrew, the term refers to a
repository o f discarded w ritings. T h e largest n um b er o f geniza papers were
discovered from a lu m b er ro om attached to a synagogue in al Fustat or
O ld C airo (G o itein 1967- I- 2)
5 T h e relevant passage reads:
'Indorumque gens incipit, non Eoo tantum mari adiacens verum et meridiano quod
Indiacaum aapellavimus’ H e re begins the In dian race, b o rd erin g not
only on the E astern Sea, but on the Sou th ern also, which we have
designated the In dian O cean ’ . (Rackham 1942 : 381)
6 T h e term L ar is derived from the Sanskrit w ord Lata, co rre sp o n d in g to

southern G ujarat in clu din g the coastal area; H arkal is the A rab icisation
o f the early m edieval nam e H arikela, d en otin g the sou th -eastern m ost
part o f presen t Bangladesh (the N oakhali, C o m illa and C h ittagon g
areas).
7 T h e m ost com prehensive list o f such fo rb id d e n acts, with a special thrust
on the textual pro h ib itio n s against seafarin g, is available in K an e (1962).
8 C asso n ’ s con trib u tion s deserve special m en tion in that he com bin es the
textual evidence with studies o f ancient seafarin g, navigational and b o a t
b u ild in g tech n ologies (1 9 8 0 , 1988a, 1988b, 1992). H e has also m ade a
masterly study o f a loan contract docum ent between two G rae c o -R o m an
m erchants at the fam ous M alabar port o f M uziris w herefrom six parcels
o f G angetic n ard , ivory, tusk and fine fabric (all b ein g In dian exports to
the R om an em pire) were tran spo rted to A lexandria, the p re m ie r p o rt o f
Egypt and the eastern M editerran ean, on b o ard the ship Harmopollon
( 1 9 9 °)• H is fresh exam in ation o f Pliny’ s account o f the developm ent o f
m aritim e routes between In dia and the West suggests that the m ost
im proved stage o f voyages brough t m arin ers and m erchants fro m a Red
Sea p ort to M alabar in twenty days, while it was previously thought to have
taken forty days. A lso o f singular im portan ce is the textual com p arison
between C lassical accounts and the Sangam literature on the m aritim e
com m erce between In dia and the R om an em pire by R om an is (1997)-
T h a p ar (1992) establishes the signal im portan ce o f p ep p er o f M alabar,
celebrated as 'black g o ld ’ , in the sea-b o rn e trade with the R om an
em pire; this set the tren d fo r the e n orm o u s dem and fo r M alabarese
p ep p er in E u ro p e right up to the early m od ern tim es. T h e im m ense
6o Ra n a b ir C hakravarti
dem and fo r eastern spices, in clu ding p ep p er, am on g R om an s has b een

exam ined — alon g with iden tification s o f plan t nam es — by M iller (1969).
T h e study o f the m aritim e com m erce between the R om an em pire and
South A sia un til recently was prim arily based on textual accounts, furth er
supp lem en ted by epigraph ic and num ism atic evidence. A rchaeological
field data fro m the R ed Sea ports like Myos H o rm o s and B ern ice have
now considerably en hanced o u r u n d erstan d in g o f the m ovem ents o f
go ods and peop le (Sideb oth am , 1 992). N o less interestin g is the
discovery o f the n am es o f In d ian m erch ants in an in scrip tio n fro m
Egypt; the nam es ap pear to have b een those typically o f early historical
D eccan (Salo m o n , I 99 1)- A first/secon d century A D B rah m i in scrip tion
was also fo u n d at Leukos L im en , a R ed Sea port in use both d u rin g the
R om an and m edieval Islam ic p e rio d (W hitcom b and Jo h n s o n 1982:
2 6 4 ). Such evidence clearly shows that the In dian m erchant did venture
to overseas coun tries and the com m ercial initiatives and m ovem ents were
n ot necessarily a m on op oly o f W estern traders.
9 For an exhaustive discussion on the im portan ce o f these seals and sealings
fo r the study o f ancient seafarin g in the Bengal coast see Chakravarti
(1992, 1996 [l9 9 2 ])- C h an draketugarh is one o f the m ost im portan t
archaeological sites in coastal B engal and appears to have b een a m ajo r
riverine p ort in the G angetic delta, n ot far away from the Bay o f B engal
(see G hosh 1 9 8 9 : 95—9 ^)* T h e epithets o f differen t types o f water crafts
m en tion ed above ap pear originally in Prakrit language; they figure in o u r
discussion in their Sankritised form s, follow ing the Sansk ritisation
suggested by M ukherjee (1 9 9 0 )
10 T h is in fo rm atio n figures in Ptolem y’ s (1932) Geographike Huphegesis,
V II.I.15.
11 A n overview o f this poin t is presen ted by K ulke ( l 993 )
12 R .K . M ookerji ( l 957) p io n e e re d the study o f the textual data in the
Yuktikalpataru, though the applicability o f the n orm s set by the text fo r boat
b u ild in g is nowadays not readily acceptable to scholars. T h e ir m ain
lim itation lies in the fact that these were not b o at-b u ild in g m anuals and
did not have even any rem ote association with seafarin g activities.
13 A sixth century AD in scrip tion from ancient Vanga (now com p risin g the
Dhaka, V ikram pur and Faridpu r areas o f Bangladesh) records the
existence o f a navatakseni, i.e. a boat b u ild in g or rep airin g station. T h e
notice o f such a h arb o u r-lik e place occurs only rarely in early In dian
d ocum en tation . But such a station in a deltaic area like lower B engal, not
far from the coast, is only to be expected (Sircar 1965-' 3 ^ 6 ; Chakravarti
1996)
14 A recent paper (M azzarino 199?) has suggested a new in terp retation o f
the account o f Pliny on the use o f the south-w est m o n so o n wind by
G rae c o -R o m an sailors. T h e Periplus o f the Eiythraean Sea and Pliny’ s Naturalis
Historia generally call this wind as H ipp alus, so nam ed after H ipp alus, a
Se a f a r in g s , S h ips a n d Sh ip O w n er s 61
sailor, who is often credited with the discovery o f the south-w est
m o n so o n w ind. O n the basis o f critical exam in ation o f Pliny’ s text,
M azzarino, however, argues that the com m only known H ipalus (H ip p a
lus) w ind was m isread in g o f the term hypalum, which actually occurs in the
text o f Pliny. T h e term hypalum, accordin g to M azzarino, sto od fo r the
seasonal south-w est wind and was not derived from any G reek m arin er,
H ipp alu s.
15 Sircar explains the term mahanavika as the captain a large ship or mahanau
(Sircar 19^ 5 : 49^> footn o te 4)
16 In an in scrip tion o f AD 1264 fro m Som n ath in the Kathiaw ad pen in sula,
western In dia the navikakrmakaras or ordinary sailors ap pear as M uslim s
(Sircar 1983)
17 T h e overseas tran sp o rtatio n o f copp er fro m western In dia to M e so p o
tam ia is known fro m the recent studies o f the H arap p an trade (R atnagar
1981). Letters o f 'In dia trad ers’ are replete with the accounts o f the
overseas tran sp o rtation o f the various types o f iro n fro m M alabarese
ports to A d en and other Red Sea h arb ou rs. T h ou gh C lassical and A rabic
texts are full o f description s o f the superb quality o f In d ian luxury
textiles, the Periplus o f the Erythraean Sea leaves little ro o m fo r doubt that
ordinary cotton cloth were also exported in bulk fro m the p ort o f
Barygaza to the West (C asson 1989: 8 3 ). In dian in digo as a regular item
o f export to al Fustat (O ld C airo ) and even to T u nisia figures in the
eleventh century G eniza busin ess docum en ts o f Ibn Awkal (Stillm an
1973; Chakravarti 1996a).
18 T h e sign ifican t p oin t is in AD 1264 G ujarat was yet to be captured by the
D elh i Sultan ate. T h e entire arran gem en t fo r the con struction o f the
m osque and the necessary provisions were approved o f by the local town
council o f Som n ath . T h e town council was h eaded by a p ro m in en t
H in d u religious leader (a Saiva P asupatacarya). Its approval was finally
ratified by the reign in g king Vaghela A rjunadeva.
19 Sch olars are engaged in lively debates on the question o f the nature and
motive o f the C ola naval expeditions in the Bay o f B engal. While
M ajum dar (1962) attributes it to the C ola ru le rs’ asp iratio n fo r gain in g
political glory, Spen cer ( l 977 anc^ *9 ^ 3 ) sees in these expeditions the
dom in an ce o f the p lu n d er motives o f the C ola kings. In -d e p th prob es
into C o la in scription s (H all 198O; C ham pakalakshm i 1995 an<^ *9 9 ^ )
w ould u n d erlin e that the C ola m aritim e expeditions were largely guided
by their lively interests in the se a-b o rn e trade between south In dia,
Sou th -east A sia and Su n g C h in a.
2 0 Steen sgard ’ s p o sitio n is approvingly quoted by O m Prakash (1998: 174)
3
S eafaring in P eninsular
India in the A ncient
P eriod: Of W atercraft and
M aritim e C om m unities
Himanshu Prabha Ray
T h e letter o f my lo rd , the most illustrious elder, arrived.

. . . Please give Sus Siti and K inbati [and Isha] and Ishaq
the Banyan my best regards, and tell them o f my longing
for them. Inform them in my name that as for pepper, in
this com ing year its value, [that is] the price per bahar, will
be thirty dinars, and m ore, and as for refurbished iron, a
bahar will be [worth] not less than twenty dinars, and that
the raqs [shining, glittering iron], which was in the city, is
completely exhausted. [Tell them also] to despatch a ship
from M angalore, if they can, and to send in it any
available pepper, iron, cubeb, and ginger; it should set
out at the first opportunity for a l-D jy b [Diu], taking some
coir [coconut fibre], fine aloes w ood, mango [?], and
coconuts, because all these are selling well.
If they are equipping a ship in A den, and they want me
to take part, I will share [in it] with them. If there were a
ship sailing from Mangalore this year, I would send them
gold, sugar, raisins and [other] goods. Be sure to inform
them o f all this, and do not be remiss, for you take the
place o f a letter o f mine to them.
These are excerpts from a letter o f II Septem ber H 49 in Arabic

written by M adm un bin Hasan bin Bundar, representative o f
62
Sea fa r in g in P e n in s u l a r In d ia in th e A n c ie n t P er io d 63
the merchants and superintendent o f the port o f A den. Like

his father, he was also a ship owner and the m iddlem an between
the Jews o f the M editerranean region and those active in India.
The addressee o f the letter was Abraham bin Perahya Ben Y iju ,
a learned Tunisian Jew who operated a brass factory on the
Malabar Coast. T he term 'banyan m entioned in this com m u
nication was the usual appellation for an Indian merchant at
that time.
The letter is an obvious example o f the close interaction
among the seafaring com m unities o f the Indian O cean in the
medieval period. N or is this an isolated instance. A Tam il guild
is known to have functioned at Barus, Sumatra, while another
settlement o f Tamil merchants was based at Takuapa on the
West Coast o f the Malay Peninsula. Merchants from various
regions are referred to as 'tax farm ers’ in the inscriptions from
Java dated between the ninth and fourteenth centuries A D .
These included groups from lower Burma, Cam bodia, the east
coast districts o f what are now A ndhra Pradesh and southern
Orissa, and from the western K onkan coastal region already
dom inated by the Ayyavole merchant association. Inscriptions
in A rabic from the Indian subcontinent dated from AD I 2 l 8 to
I 29 1 refer to traders and shippers from Arabia and Persia
resident in the towns o f Gujarat such as those o f Cambay,
Somnatha, Junagadh and Anhilavada. A n epitaph from Cambay
refers to a trader in betel nuts and another who was drowned,
while a third record from Junagadh com m em orates the
erection o f a mosque by the chief o f shipmasters.
N or was this interaction lim ited to the historical period.
There is evidence for maritime networks across the western
Indian O cean as early as the third m illennium B C . A triad o f
names that figures consistently in M esopotam ian texts includes
those o f D ilm un (the Bahrain archipelago), Magan (the O m an
peninsula) and Melukkha (identified with the H arappan
settlements). A n early reference appears in the boast by Sargon
o f Akkad ( 2 3 34-2279 BC) stating that:
'The ships from Meluhha

the ships from Magan
64 H im a n s h u Prabha Ra y
the ships from D ilm un

he made tie-up alongside
the quay o f Akkad’ .
These m aritime com m unities, i.e. com m unities involved in

sailing and navigation, form the focus o f this paper, though
many o f the other groups, such as pilgrim s, merchants,
adventurers, travellers and wandering scholars are by no means
insignificant and have been discussed elsewhere (Ray 1999 )- ^
is being argued here that these seafaring com m unities form ed a
distinct group and are to be distinguished from merchants,
traders and caravaneers involved in inland trade.1 Merchants
and traders in some cases certainly owned ships and water
craft, but they neither m anned nor sailed these. M ore often,
however, goods and cargoes were entrusted to the captain o f the
vessel who was then responsible for their sale and profit. W hile
the chronological fram ework adopted in this paper relates to
the ancient p eriod , i.e. up till the twelfth century AD,
reference would also be made to present practices adopted by
indigenous sailing com m unities o f the Indian coast.
Indigenous traditions o f boat-b uildin g and navigation in the
Indian O cean are both num erous and diverse. A t present a
m ajority o f water craft are used for fishing and ferrying
passengers across river-crossings, mechanised trawlers having
taken over much o f the coastal and long-distance transporta
tion by sea. Such ethnographic data are however, here regarded
neither as 'living traditions’ nor as 'survivors from India’ s past’
as has been suggested by some researchers (Blue et al. 1997 ;
19 1 ).2 W hile there have b een co n tin u ities such as the
persistence o f the stitched tradition, the hierarchies o f scale
have changed. From being used for cargo vessels, the stitched
tradition now survives in small boats employed for beach
seining along the east coast, (figs. 3 -1 & 3-2)- A distinction has
also to be made between the use o f water craft for fishing, as at
present, where investments are generally on a modest scale as
com pared to those for cargo vessels. This distinction between
boat and ship on the basis o f econom ic and social significance
is well worth stressing. As G reenhill argues:
S e a f a r in g in P e n in s u l a r In d ia in t h e A n c ie n t P e r io d 65
F ig u r e 3 .1 Stitched craft on the Andhra coast
Figu re 3.2 C lose-up o f the stitching

66 H im a n s h u Prabh a Ray
Ships are structures which constituted significant elements

in the econom ies o f the societies which built and operated
them. . . . Such structures we may call ships and, for
purposes o f convenience perhaps, describe all less
significant contem porary m obile floating structures as
boats, whatever their construction and lineage ( l9 9 5 : 9)*
T h e attempt would be to stress the points o f disjunction to

enable a judicious use o f the data.
Perhaps the single most significant point o f disjunction was
the introduction o f steamship navigation in the nineteenth
century. H istorical evidence indicates that as a result, the
indigenous system underwent radical changes. Seafaring activity
shifted from being 'fair weather’ to 'all weather’ . Instead o f the
traditional com m unities, lin er companies now dom inated
m aritime trade. T h e skilled manpower required for m anning
these mechanised vessels could no longer be provided by the
littoral sailing com m unities, and this led to the m arginalisation
o f coastal groups to traditional fishing activities. These
operations are in turn being threatened on account o f
m echanisation and the introduction o f com m ercial fishing.
Thus while docum entation o f any tradition o f boat-building is
o f prim ary im portance, by itself it can provide little for a study
o f the past and needs to be analysed in conjunction with data
regarding other aspects o f seafaring activity.
A t the outset this paper should be placed within the wider
context o f archaeological and historical studies relating to the
ancient period o f the subcontinent. These studies, dom inated
as they have been by the emphasis on material culture, have
focused on social evolution and the em ergence o f the state.
W ithin this framework, trade has been linked to agrarian
expansion and the em ergence o f urban centres. T h e aim o f
archaeological research has centred on a study o f the material
culture. As a result, m aritim e contacts have rem ained restricted
to studies on trade, especially luxury trade and fishing and
sailing com m unities have found no relevance, m uch less
maritime technologies. It is only recently that fish as a food
resource has been studied in the context o f the Harappan
civilisation (Belcher 1998)- Similarly, the emphasis in h istor

ical studies has been on analysing written texts, be they
inscriptions which prim arily record land-grants or literary
sources which have an obvious inherent bias in favour o f the
literate elite. These sources have little to contribute towards a
study o f ship-building technology. It is a sad comm entary that
after the pioneerin g work o f R .K . M ookerji on Indian Shipping
published in 1912, there has been little follow up on the
subject.
Inadequate sources are frequently quoted as the reason for
this disregard. What is missed is a comparative analysis o f these
sources that would make an integration possible for a study o f
seafaring com m unities and maritime technologies in a histor
ical context. In this paper the objectives are twofold: first, to
concentrate on continuity and change in sh ip -b u ild in g
technologies in peninsular India; and second, to draw
attention to the inter-relatedness o f the seafaring com m unities
o f the Indian O cean.
M aritim e T echnologies
T h e disjunction between the indigenous sailing pattern and

European intervention is nowhere m ore marked than in the
evolution o f ports and harbours along the Indian coast from
the nineteenth century onwards. P rior to these changes,
elaborate structures and harbour installations were n o n
existent in the region, nor were these necessary, (fig. 3 .3 ).
In the ancient p eriod , no specific term inology for 'p o rt’ is in
evidence or was for that matter required. A m on g others, the
term pattana was used both for a market centre as also for a
riverine settlement.
Ports were generally located at the mouths o f rivers, though
those on the coast were not unknown. This location provided
easy accessibility to the routes o f the interior, but one o f the
hazards was the problem o f silting resulting in constant shifts
and changes. Evidence for these dislocations o f port sites is
docum ented throughout the historical period in the Indian
subcontinent. T h e longevity o f the site o f Arikam edu from the
Iron Age M egalithic period to the medieval p eriod , vis-a-vis

other sites along the Tam il coast, is best explained by its
relatively larger h interland .3 A study o f remote sensing data on
palaeo channels and palaeo m ud flats for the Gujarat coast
indicates that the G u lf o f Kham bat was much wider in the past
than it is now, extending up to the south o f the present city o f
Ahm edabad. Both Vallabhi and M odhera which occur p ro m i
nently in the historical literature and are inland, landlocked
centres at present, were located closer to the coast and would
have been riverine ports (Thakker, Raval and Dasgupta 1993).
A nother pertinent example relates to shifts on the west coast
between the ports o f Bharuch, Sopara, Kalyan, Chaul and Surat
during the ancient and medieval periods. It was only with the
transform ation o f the motley collection o f seven islands with a
variety o f landing places into the port o f Bombay that an
identifiable port city was created on the west coast o f India. U ntil
this transformation, several landing places had been in existence
in the vicinity o f Bombay. It was between 1845 an<^ 1859 that
there was a need for the construction o f the first docks for the
steam-shipping companies as a result o f the increasing naval and
Figure 3.3 Boat repairs on Orissa coast

Se a fa r in g in P e n in s u l a r In d ia in th e A n c ie n t P er io d 69
commercial importance o f the site. It is significant that before

the construction o f these facilities, sailing ships anchored at
some distance from the coast and small boats were used to ferry
passengers and cargo to the shore.4
It would seem that similar problem s plagued shipping in an
earlier period as well and references to this practice have been
docum ented in early Buddhist literature. T h e Mahavastu, a
Buddhist text o f the early centuries AD, narrates the story o f
a brahm ana who returned from a sea voyage and stepped onto a
landing craft from the ship for transportation to the coast.5
Steam -shipping in the Indian O cean had been pioneered

since c. 1830, but long distances and low density o f traffic
meant that until the 1860s little else but passengers were
carried by the steamers o f the few successful companies
(Broeze, et al. 1987: 255).
It was only in the 1860s that regular cargo services entered the
ocean with the establishm ent o f the lin er com panies. T h e a ll
weather steam vessel was 'larger, faster, safer and m ore
reliable than indigenous craft on lo n g-h au l coastal and trans
oceanic voyages’ (M cPherson I 9 9 3 : 222). This resulted in the
indigenous sailing craft being driven to sh ort-h aul m arginal
and feeder routes, such as fo r example between the west coast
o f India and the Persian G ulf. Reduced dem and for cargo
carrying m eant reduction in the num ber o f vessels being
constructed and the closure o f traditional b o a t-b u ild in g
yards.
P rior to this tilt in power relations in favour o f steam
navigation, the trading system was governed by the regime o f
the m onsoon winds. D uring the summer months these winds
blow from the south-west and at this time they are violent and
stormy and frequently rise to gale force.
As a result sailing remains suspended from May until

Septem ber along the west coast o f India as well as along
the south Arabian coast. A reversal takes place around
O ctober and the north-east m onsoon dominates between
N ovem ber to A p ril. Thus the season for trade from
70 H im a n s h u Prabha R ay
Gujarat to A den was broadly Septem ber to May, for A den

to Malabar, O ctober to February (Das Gupta and Pearson
1987: 12 ; Tibbetts 1971: 3 6 4 -7 4 ).
This meant that sailing was seasonal and that for several months
there was no activity at the ports. T h e sailing season determ ined
the price and movement o f export com m odities. A letter
written in the eleventh century refers to the price o f flax falling
after the arrival o f a ship at Mazara on the south-western tip o f
Sicily, whereas at the end o f the sailing season the price o f lead
rose as no m ore supplies were expected (G oitein 1967)*
A noth er practice, indicated in early Buddhist literature, was for
local merchants to book a share o f the incom ing cargo by
paying m oney in advance (fataka, Book I: n o .4 0 ); a practice,
which could often be m anipulated to control prices.
B oat C la ssifica tio n
A n oth er area where European intervention has resulted in

radical changes is in the sphere o f classification o f water craft.
From the m id-eighteenth century onwards, both the English
and the French had expanded their trading enclaves on the
coasts o f India into fu ll-fled ged establishments. Th ere was, at
the same time, a shift in the European perception o f India’ s
contribution to the developm ent o f science and technology.
W hile the achievements o f the ancient 'H in dus’ in many fields
o f scientific enquiry were lauded, the same was not considered
applicable to the contem porary p eriod. W ith a view to
preserving these ancient achievements, attempts were made to
record in detail relics o f the past and a new discipline was born
— that o f maritime ethnography.
O n e o f the results o f this early developm ent o f the discipline
o f ethnography was the classification o f the water craft o f the
Indian O cean into neat categories, often based on European
term in ology. T h u s Portuguese and D utch shipping lists
described Asian vessels that called at different ports as fustas,
yachts, sloops, brigs, galliots, frigates and phats. W riting o f
coastal traffic o f Travancore in the early sixteenth century,
Sea fa r in g in P e n i n s u l a r In d i a in th e A n c ie n t P er io d 71
Duarte Barbosa term ed the small rowing vessels bargatim (H ill

1958: 202). In contrast, Tom e Pires had little acquaintance
with matters m aritime and the names o f boats that he gives are
all transliterations into Portuguese o f equivalent words (ibid.:
204.).
Thom as Bowrey (1650—1713) may be credited with the
introduction o f new classifications, such as the masula — a term
that continues to be used for the frameless stitched boats o f
India’ s east coast. This is not a term used by local boat builders
and users who adopt the generic term for boat such as padagu,
padava and padhua. N or does the term incorporate any typical
boat type, as there are significant variations in size, shape and
m ethod o f construction among the vessels said to form a part o f
the masula family (Kentley 199^: 25 °)- But perhaps Bowrey’ s
contribution should be recognised m ore in terms o f his boat
drawings, rather than boat typology.6
This system o f classifying boat types also presented a marked
contrast to the situation prevailing along the Indian coasts.
W hile European ship-types may be categorised on the basis o f
differences in their tackle, the same is not true o f the
indigenous craft o f the Indian O cean. In the case o f cargo
vessels, regional variations are often superficial and lim ited to
the nom enclature used rather than significant technological
differences between them. In contrast, the hull form s o f fishing
boats are m ore region specific, though here again m ore than
one nom enclature may be used to define a vessel. For example,
the terms patia and danga are often used interchangeably for
clinker built vessels on the Orissa coast. A noth er difference
between the two is that while fishing craft are often built either
by the com m unity themselves or by itinerant boat-builders, the
large cargo vessels are constructed at boat yards.
Sim ilar problem s o f classification are encountered vis-a-vis
cargo carriers. These travel to distant centres o f the Indian
O cean depending on the remunerative freights offered and
return to their hom e ports for overhauling during the south
west m onsoon. O ften boats o f the same type and build are
given different names on account o f the nationality o f the
owner. Divergences are insignificant and defined prim arily in
n H im a n sh u Prabha Ray
terms o f ornam entation. T h e sambuk, though o f Arab origin is

often constructed in boat-b uilding yards on the west coast o f
India, an im portant centre being at Beypore, south o f Calicut
on the m outh o f the river Chaliyar (Wiebeck 1987: 96). In
addition to the sambuk, a variety o f water craft, such as the
padava, kotia, pattemar, etc., were also built at this centre. Its
m ajor advantage was access to a wide variety o f tim ber from the
rain-forests which form ed the upper basins o f the river
(K unhalli 1993 : 56 )-
Thus the baghla and the gunjo are the Arab form s o f the
Indian kotia, the Arab bum the counterpart o f the Indian
dhangi and nauri, with the Indian bate 1representing the Arab
sambuk (H ornell 1970: 197) •
Any association o f boat-building traditions with ethnicity, e.g.

Arab tradition, Austronesian tradition, etc. raises the question
o f centres o f maritime culture. Boat building is not essentially
ethnic in character, though in certain areas, historically there has
been an association between certain groups and distinct boat
building traits. In antiquity, as at present, there were no national
barriers to seafaring activity and merchants and sailors travelled
on the routes most beneficial to them. For example, Buzurg
writing in the tenth century states that A l-R ubban A bharah had
learnt the art o f seamanship as a fisherman in the G ulf, having
first been a shepherd in the district o f K irm an. He then became
a sailor on a ship trading with India, and finally commanded the
markab sini or China ship itself (Tampoe 1989= 122).
In addition to the m obility o f m en, there is evidence for the
use o f raw materials like w ood, coir rope, etc. for boat building
from different parts o f the Indian O cean. The Periplus Maris
Erythraei, for example, refers to the im port o f teakwood and
beams and logs o f sisoo (Dalbergia sissoo) and ebony from the west
coast o f India to O m an (section 36). 'O m ana also . . . sends
out to Arabia its local sewn boats, the kind called madarate
(ibid.; Casson, 1989= 73 )*
A dm iral Paris (1806—93) form ulated the hypothesis that
indigenous watercraft were as much expressions o f a culture as
were palaces, religious buildings or fortresses. As a trained
Sea fa r in g in P e n in s u l a r In d ia in t h e A n c i e n t P e r io d 73
French navigator he circumnavigated aboard the Astrolabe, the

Favorite and the Artemis and catalogued the 'extra-European craft
(Reith 1993)* This trend continued in the writings o f H orn ell
who argued that a distinctive climate and coast form ation
dom inated or influenced by distinct ethnic stocks resulted in
the evolution o f characteristic boat-types (1946: 195)*
accordance with this correlation, boat-types were seen as 'c o
extensive in range with the limits o f race and language or the
influence o f foreign sea-trade’ (ibid.). T h e influence o f
Portuguese sea-trade, he stated, was particularly noticeable in
the larger vessels o f the far south .7
Based on ethnological studies, H orn ell (192O: 139—73 )
described distinct b oat-building traditions that evolved along
the Indian coastline. His m ajor preoccupation was with tracing
the evolution o f w ater-craft, as also their com m on origins and
diffusion through migrations and movements o f people, a
typical example being the outrigger canoe. Attem pts at
correlating boat-types with race and language do not explain
the presence o f different boat-types in the same cultural m ilieu
or similar boat construction methods independently evolved in
different environm ents. Several examples could be quoted, but
one would suffice: that o f the clinker-built patia in the estuary
o f the Pancapara in north Orissa together with flush-laid boats,
both types o f boats being built by the same group o f carpenters.
W hile today this tradition survives in a few pockets, it was
earlier com m on on the upper Ganga. Similarly, foreign
influences cannot be identified by isolating a few traits and
linking their incorporation to European contacts. Instead the
crucial issue should discuss the nature o f interaction between
different boat-building traditions and the reasons why one
dom inated and prevailed.
Ownership o f Vessels
A study o f the organisation o f shipping involves a series o f

issues to which no clear answers are available in the records.
For example, what does ownership im ply? Does it mean full
proprietorship o f the entire vessel by one person or does it
refer to form s o f shared property? Data on these aspects is

seldom available in a historical context. Early Buddhist texts
make a distinction between owners o f vessels, sea-traders and
the crew m anning the watercraft. Merchants who earned a
living by maritime trade or jalapathakamm ikena are referred to in
the Jatakas, while Panini makes a distinction between merchants
who owned two vessels (dvinavadhana) or who sailed with five ships
(pancanavapriya) (Agrawala, 1953 : I 55 )* In contrast, members o f
the crew responsible for m anning a big ship or mahanava
included a captain (sasaka), a pilot (niiyam aka), a m anipulator o f
the cutter and ropes and a bailer o f water (utsecaka, Arthasastra,
II.2 8.13). T h e Jatakas m ention a niyyamakajetthaka (Book XI: no.
463) and include amongst the crew oarsmen (Jatakam ala: XIV)
and sailors (kammakara, B ook I: no.31) who worked the sails
while the ships sailed in the open sea where the coast was no
longer visible (Jataka, B ook XVI: no. 518).
A noth er issue that has a bearing on this question is the
location o f fishing villages vis-a-vis agricultural settlements.
A lo n g the coast o f Orissa, there is a marked variation between
the northern and southern sectors. Th ou gh characterised by
stretches o f sand beaches, north Orissa provides pockets o f
agricultural land not far from the coast. T he fisherm en, as a
result, live in the villages populated by agriculturists, and
though their prim ary occupation continues to be m arine
fishing, they also work as agricultural labourers and at times
possess some land. Wealthy m en o f the village who determ ine
the share o f different crew members in the catch also employ
fisherm en on wages who often own boats.
In contrast, fishing settlements in south Orissa are situated
on the broad sandy beaches and there is a clear segregation
from agricultural villages located inland. Fishermen own boats
and nets and live exclusively by fishing. As a result they are able
to exercise far m ore control over the catch and its marketing
(Mishra 1997: 109).
A n oth er example may be quoted from Gujarat. T h e region
has the largest traditional fishing boats in the subcontinent,
though it has the least productive m arine fishing grounds and
the lowest demand for fish in the hinterland. This somewhat
anomalous situation may be explained by the fact that fisher

men have to resort to alternate means o f earning a liv elih o o d ,
and participation in coastal cargo trade presents one such
possibility (K u rien 1996: 22 l).
In n ovation and Change
In the sixteenth and seventeenth centuries there was a greater

use o f Asian boat-yards by Europeans for shipbuilding. A
majority o f these yards were located on the west coast o f India at
Goa, C ochin, Calicut, Bassein and Daman and at centres in
Gujarat and Kathiawar (Arasaratnam 1994 : 2 4 $). In I 5 l 5 >
Duarte Barbosa who came o f a Portuguese family in the service
o f the Duke o f Braganza was com m issioned to go to Calicut to
supervise the building o f two galleys. He returned to Portugal
and wrote his account o f the vessels o f the Indian O cean (Hill
1958: 2 0 2 ). T h e reason for this was that production and
labour costs were much lower in India as compared to Europe.
In 1619 it was estimated that a ship o f 500 tons built in India
could save the English East India Com pany £ l,0 0 0 , as against
one built in the shipyards o f D eptford and Blackwell in
England (Qaisar 1982: 22 ). In addition, European ship
designs were being copied and Indian built vessels were being
converted outwardly 'after the Christian m anner by fitting her
properly’ (ibid.: 37 )* Alternatively, European style craft were
being constructed fo r local clientele. In 1507 an Arab
merchant built what was described as a galleon in Portuguese
style in Gujarat (ibid.: 25 )-
W hile em erging European com m ercial and political dom
inance radically altered the Indian O cean system, this is not to
suggest that the system had continued unchanged in earlier
periods. Historically, iconographic evidence would indicate
m odifications in the design o f water craft, at least in the case o f
larger vessels operating on the west coast and the introduction
o f several features, such as for example the stern-post rudder
from the twelfth century onwards (Deloche 1996: 212 ).
It is difficult to docum ent these changes for an earlier
period , since representations o f boats in art and architecture
are an integral part o f the im portance attributed to seafaring

activity by society. Thus in the Indian subcontinent these
representations are found mainly in two historical periods: the
first from the second century BC to the seventh century A D ; and
the second from the eleventh to the fifteenth centuries AD
(ibid.). In the Early H istorical p erio d , the representations are
either on Buddhist m onum ents or on coins, seals and sealing
(fig. 3 .4 ), and are thus linked to the Buddhist support to
trading activities (Ray 1994-) • Sea-going vessels are nevertheless
depicted on Satavahana coins, in the Ajanta paintings, and
Aurangabad sculptures. These bear a close resemblance to the
craft plying the east coast in m ore recent times, as docum ented
by Paris (Reith 1993 :
In the second period, the depictions indicate two varied
perceptions. Elaborately carved memorial stones have been found
extensively in peninsular India dating from the fifth to the twelfth
centuries A D . These provide abundant inform ation on the cult o f
the hero and are also significant indicators o f sea battles,
especially those from the region o f Goa dating to the period o f the
Kadambas ( a d 9 g O — 1270) (fig. 3 *5 )* From the eleventh century
onwards, representations on memorial stones depict planked
vessels, sharp-ended with a long projecting bow strongly raked, in
addition to the canoe-shaped craft (Deloche 1996: 209).
In contrast to this emphasis on sea-battles and aggression
along the west coast are the iconographic illustrations from
Orissa and Bengal. A cerem onial barge is prom inently depicted
on the eleventh century Jagannatha temple at Puri on the
Orissa coast, stressing thereby the ritual and cerem onial aspects
o f the boat. This trend is continued in the terracotta temples o f
Bengal dated to the eighteenth-nineteenth centuries A D . A
total o f nearly sixty-two panels o f the temples o f West Bengal
(districts o f Bankura, Burdwan, Birbhum , H ooghly, Howrah,
M idnapur and 2 4 “ Parganas) show realistic depictions o f river
craft, while larger vessels form the subject o f fifty-three panels
(Deloche I 99 I: 2). T h e river-craft represented are rowing and
racing skiffs, travelling houseboats or cerem onial barges, while
the m ore com m on types such as rafts, dug-outs, cargo-carriers
and fishing boats have been ignored.
Seafaring in P e n i n s u l a r In d i a in the A n cien t P er iod 77
F igu re 3 .4 Clay sealings from Chandraketugarh
F ig u r e 3 .5 M e m o r ia l s to n e , G o a M u s e u m
A n argument often used against innovation is that since

instructions were verbal and no drawings used for boat
building, very little change was possible. In the case o f the
T u ticorin torn', it has been docum ented that no plans were used
in its construction. Instead outlines o f the ribs were drawn on
the floor and wooden templates chiselled over these markings.
It is from these templates that the trunks and branches o f the
babul tree are cut to make up the karavai or frame (Hawkins 1965:
151). W hile this may well be partially true, it is difficult to
extend the debate further, since there are references to the use
o f architectural drawings in the p re-m od ern period and the
same may have been the case with w atercraft .8
Similarly, though no charts or maps have survived, literary
sources do indicate their use, the Sanskrit term used for the
purpose being citra or alekhya, i.e. a painting or delineation. A
fifth century text, the Brhatkathaslokasangraha by Buddhasvami
refers to details o f a particular sea, direction and place being
inscribed on a wooden board with a cover (samputaka) and
being used to locate a ship that had gone astray (G ole, 1989:
16 - 23 ) -
Traditionally the navigators o f the Indian O cean learnt their
routes by identifying natural phenom ena and by checking their
positions by stars. Stellar navigation made sailing along the
latitude possible, though finding the longitude was relatively
m ore difficult. However, longitude tables were theoretically
prepared by mathematical computations and early versions o f
nautical charts were available for specific routes and sailing
schedules. L ong-h aul direct voyages were not unknown, but
these were not the norm .
The first century AD G reek text, the Periplus Maris Erythraei,
refers to several o f the navigational techniques adopted in the
western Indian O cean, such as identifying the approaching
coast by means o f changes in the colour o f the water; and the
presence o f sea-snakes.
Vessels com ing from the open sea in the vicinity o f these
places get an indication that they are approaching land
from the snakes that emerge to meet them; these are also
black in colour but shorter and with dragon-shaped heads

and b lo o d -red eyes (section 55 )-
This practice o f recognising the Indian coast, especially the

Gujarat and K onkan coasts, by the abundance o f sea-snakes is
also referred to in Tibbetts’ later work and continues to be
adopted by local sailors. In vernacular literature, there are
several references to the use o f disakaka, sea-crows and other
hom ing birds on board a vessel to judge the proxim ity o f land
or the approach o f an insular zone. For example, the approach
o f the Indian coast from the Arabian Sea is recognised by the
abundance o f sea-snakes about 20 zams offsh ore .9
Since sailing in the Indian O cean has been dom inated by the
regim en o f the m onsoon winds, the correct interpretation o f
the natural phenom ena becomes crucial for success at sea and
this trend has continued into the H istorical period.
In 1541, a Portuguese m arauding fleet in the Red Sea set

sail to return to India on the I O th o f July. T he headstrong
captain refused to listen to the advice o f his M uslim pilots,
who, basing themselves on centuries o f experience, told
him that by leaving at this time he would have no trouble
getting to the entrance to the Red Sea, but that once in
the Arabian Sea weather o f such vileness could be expected
that no ship could navigate. A nd this advice, o f course,
turned out to be correct (letter o f D. M anuel de Lima to
K ing, Goa, 18 Novem ber I 541 as quoted in Pearson 1994 :
148).
In addition to observations o f the natural phenom ena, maps

and traditional instruments were in use, for ascertaining
latitude when out o f soundings (Congreve 185O: 101—4)- T he
use o f fingers for altitude measurement was quite widespread in
antiquity. A m on g the Arabs it was indispensable not only for
m ariners but also for the desert dwellers who had to travel
along the trackless sand dunes (Fatimi 1996* 283).
Boat building has often been described as a craft marked by
extreme conservatism. A study o f recent attempts by the
G overnm ent o f India and other organisations to introduce
8o H im a n s h u Prabh a Ray
m ore efficient boat designs in place o f the traditional kattumaram

is instructive. It indicates that initial attempts at either
m otorization o f this traditional craft or the introduction o f
new designs o f small mechanised boats met with little success.
T h e kattumaram fishery is noted for its great diversity o f nets

and other fish catching devices, which is a function o f the
seasonality and variety o f fish species found in these
coastal waters. Since the nets were small and the quantity
o f fish caught was not unduly large to need m ore than
such a wash th rough’ platform to brin g it shore in small
flexible palmyra baskets. T h e additional fact that the fish,
though netted, continued to be awash in sea water and
thus stayed fresh for a longer period o f time was another
im portant factor in favour o f fisherm en retaining their
kattumarams (K u rien 1996: 221—222).
T h e strategy that finally worked was the result o f a com bination

o f factors, such as increased demand for shrimp; com petition
from large trawlers in fishing; rise in prices o f traditional craft
due to shortage o f appropriate tim ber resources; the develop
ment o f participatory organisations involving both fisherm en
and socially com m itted technicians; and the growing awareness
among the fishing com m unities themselves for technological
change in order to be able to compete (ibid.: 219—234)-
Ethnographic studies o f the boat thus cannot be separated
fro m developm ents in m aritim e activity. C hanges and
adaptations in boat types are linked to the purpose fo r which
the boat is used. T h e T u tico rin toni, for example, was
introduced by the local cotton-pressers around 1914* Initially
the toni was built on the English m odel, but this was rejected
by local boat-ow ners who p referred to copy in sim plified
form the Arab bums. As this m odel did not work too well,
fu rth er changes were in tro d u ced (H o rn e ll 194-5 : 2 I 7 )-10
O riginally these vessels were no m ore than 50 to J O tons
and carried a single lateen sail. Subsequent developments have
resulted in changes both in the size and hull form o f these
sailing craft (Hawkins 1965* I 4 9 )-n T h e recent evolution o f
the T u tico rin toni as a boat-type provides significant clues for
an understanding o f the processes involved and the m odel

adapted for local use.
O th er im portant variables resulting in changes in either the
design o f boats or the introduction o f new types are the
com m odities involved. Understandably the size o f the vessel will
depend on the nature and bulk o f goods that need to be
transported. For example, the size o f the Arab vessels known by
the generic term dhow increased with a shift to bulk cargo like
mangrove tim ber in the m id-twentieth century (Prins 1965:
4—5)- Demands o f maritime trade and the need to adapt to
changing circumstances are crucial if the competitive edge is to
be maintained.
Seafaring C o m m u n ities
This then brings us to the second issue: the seafaring

communities o f the Indian Ocean. Traditionally it was the
fishing communities that had provided sailors, while pilots and
navigators were drawn from the literate elite. This pattern was,
however, disrupted with the introduction o f the new technology,
i.e. the steamship. This required new skills such as engineering —
skills which were the exclusive domain o f the Europeans, thereby
relegating indigenous sailors to low-paid jobs such as those o f
deckhands, stokers, cooks, launderers and stewards. These sailors
were recruited through intermediaries and port-based agencies
who in turn tapped the depressed rural areas for unskilled
labour. Further change was brought about with the emergence o f
national navies, but here again the pattern o f recruitment
tended to be universal within the respective states. As a result,
sailors as litto ral comm unities’ have largely vanished from the
Indian O cean (McPherson 1993: 236).
Fis h in g C o m m u n it ie s
Littoral sailing and fishing comm unities were an integral part

o f m aritime activity in antiquity. It has been suggested that
when viewed from a cross-cultural perspective, fishing com
munities have much m ore in com m on than those in other
82 H i m a n s h u P r a b h a R a y
occupational positions. These com m onalties have been traced

to such cross-cutting factors as uncertainty o f the resource base
due to lack o f knowledge or control over the m arine bio-m ass;
territorial openness, since m obile m arine bio-m ass movements
are difficult if not impossible to either detect or control; and
m ultiple subsistence potential, such as both land and sea
exploitation (Smith 1977 : 7)- A noth er socio-cultural charac
teristic that fishing com m unities share is the explicit awareness
o f om nipresent danger associated with their subsistence activity
(ibid.: 8).
Evidence for coastal occupation in peninsular India dates
from the M esolithic period onwards, though fishing as a
resource strategy has yet to receive attention from Indian
archaeologists. Inform ation on these com m unities is also
available from another source, that o f rock art. In the southern
Deccan, nearly thirty sites with a little m ore than hundred
painted surfaces o f the M esolithic period have been found.
Narrative scenes are rarely represented, nor are hunting scenes
shown. It is significant that in southern Deccan fish are quite
com m on and are depicted at all rock art sites (e.g. at
Sanganunpalli, H am pi, Settavaray, see N eum ayer, 1993 :
123- 4 )-
O n e o f the few literary sources that provide graphic accounts
o f fishing and sailing com m unities is the early Tam il poetry.
For example, in Akananuru IO , 8—13, the anonymous heroine is
compared to the coastal settlement o f Tonti:
H er beauty is like the rich Tonti, where in the fragrant

seaside villages the fisherm en (paratavar) who still have new
nets, and whose old boat (tim il) was removed in the sand
heaps, broken by the glittering waves o f the ocean ruffled
by the east-w ind, having caught the shark in the high sand
shore, divide it into parts (De Romanis, 1997 : 93 )-
In another poem , fishing is described as one o f the m ajor

activities o f the settlement at M uciri or M uziris. T h e fisherm en
are described as harvesting the salt on the shore (Akananuru 280)
and collecting at the harbour 'where they take fat pearls from
the spreading waves and divide them on the broad shore’
Sea fa r in g in P e n i n s u l a r In d i a in th e A n cie n t P er io d 83
(Akananuru 280). The outer streets o f a town are said to be

inhabited by fisherm en (Pattinappalai j j ) .
O ne o f the sites that could have provided invaluable data for
the growth o f fishing and sailing in the region is Arikam edu
south o f Pondicherry. T he site grew along the river bank and
extended for almost 480 metres north-south. W heeler had
divided it into a northern and southern sector and Casal’ s
subsequent excavations had indicated that the southern sector
was first settled by fishing comm unities (Casal 1949 : 3 1)- This
had been m odified by W heeler, though he accepted that prior
to what he term ed Rom an contacts there was a fishing village at
the site (Wheeler 1954 : 47 )- M ore recent excavations have
confirm ed that the earliest settlement both at Arikam edu and
in the region around the site dates to the megalithic period. It
is therefore, incom prehensible that without any archaeo-
zoological analysis and solely on the basis o f pottery, the
excavator should conclude that:
In the archaeological re co rd , there is no convincing

evidence that A rikam edu was a fish erm en’ s village,
although today fishing, like toddy making, is a m ajor
occupation (Begley 1996: 14)*
Significantly, the site was a centre for the production o f bangles

and ear ornaments from marine shell in the subsequent Early
H istoric phase. The excavations made no attempt to u n d er
stand the mechanism by which the shell Xanthus (Turbinella) pyrum
may have been obtained and the possible role o f local fishing
com m unities in this.
Two inscriptions dated to the early centuries o f the Christian
era, inscribed on the back wall o f the Buddhist cave num ber 8
at Nasik record donations made by a fisherman Mugudasa
together with his family (saparivarasa) . Mugudasa is described as
belonging to the lay comm unity o f Chetikas and the object o f
the donation was the cave. For the benefit o f the ascetics
residing in the cave, a field was given by D ham anandin (Senart
1905 - 6 : 76 - 7 ) .
Fishing com m unities o f the Indian O cean also occur in
accounts by G reek geographers who refer to them as the
Fisheaters or Ichthyophagoi who inhabited the coasts and offshore

islands o f East A frica, the 'Erythraean Sea’ , southern Iran and
India. It seems that under this rubric, the Greeks included
various ethnically distinct coastal peoples whose prim ary source
o f food was fish and who spoke a variety o f languages that were
incom prehensible to the Greeks (Casson, 1989* 97 “ I00).
Nevertheless what makes the descriptions significant is that
these were often based on direct, if irregular contact with these
groups; for example, during the voyage o f Nearchus the Greeks
landed on the coast when required, either for supplies, or for
navigational purposes.
Seafarers and N a v ig a t o r s
In addition to the fishing communities there are references to

traders by sea who earned a living by jalapathakammikena meaning
'through the sea routes’ (Jataka, Book I: n o . 4 ; B o o kX : nos. 4 3 9 >
4 4 2 ; Book XI: no. 463; Book XVI: no. S1^)- A n inscription
from the site o f K anheri near Bombay records the gift o f a
sagarapaloganas interpreted as a community o f traders by sea
(Burgess 1883: n o .23). O n the testimony o f the Periplus, the
seafaring communities o f the Indian O cean presented a diverse
range and included the Nabateans, Sabaeans, Homerites, Arabs
and Indians. Muza at the mouth o f the Red Sea is described as a
port o f trade without a harbour (section 24) but with a good
road-stead for m ooring and teem ing with Arab ship-owners or
charterers and sailors (section 2 l). Leuko K om e on the Red Sea
coast was the harbour o f the Nabateans where craft, 'none large’
came to it loaded with freight from Arabia (section 19). T he
island o f Dioscurides or Socotra is referred to as being settled by
Arabs and Indians and even some Greeks who sail out o f there
to trade’ (section 30) and there are references to the export o f
sewn boats from Om ana to Q ana and Arabia (section 36).
M aritim e voyages in the early centuries o f the Christian era
were regarded as profitable ventures and Buddhist literature
describes a variety o f social groups who were involved (Jatakas,
B ook I, n o . 4 ; B o o k X , nos. 4 3 9 >4 4 2 ; B o o k X I, no. 4 ^3 ; B ook
XVI, no. 528). In addition to merchants there are references
Se a f a r in g in P e n in s u l a r In d ia in t h e A n c i e n t P e r io d 85
to princes who travelled across the seas to make m oney (Book

XXI, no. 539 )- T h e prosperity and status o f these m ariners is
evident from the donations made by them to the Buddhist
monastic establishments and recorded in the inscriptions o f the
Early H istoric period (Ray 1989: 437 —57 )*
This then raises the question o f the im portance o f using a
knowledgeable pilot in the successful com pletion o f voyages in
antiquity. W hile much o f this expertise was com m unicated
orally and learnt during years o f apprenticeship, there is
increasing evidence for the existence o f written 'charts’ which
have unfortunately not survived in the archaeological reco rd .12
By the late mediaeval p eriod indigenous nautical charts
developed which were comparable to the Portuguese portolanos.
These were based on a systematic observation o f coastal
configuration, landmarks easily identifiable from the sea at a
distance, bearings in terms o f horizon star azimuths and
underwater flo o r features.
A n eighteenth century map painted on cloth m easuring 4 0 6
— 4 4 0 cms. depicts the whole o f Gujarat and some parts o f
Rajasthan and is now in the possession o f the Baroda Museum
and Picture Gallery. It represents an earlier indigenous
tradition known from fifteenth century textual references and
one that was different from the European tradition that
originated in G reek cartography. T h e map when hung shows
east on top and all the place names are written bearing this
placement in m ind. T h e rivers are shown as zigzag lines, the sea
as a large area and the m ountains as blocks. T h e central
position in the map is occupied by Ahm edabad shown on the
banks o f the Sabarmati and the other m ajor towns o f Gujarat,
such as Surat, Bharuch, Kham bat, Baroda, etc. correctly
placed. T h e m ain function o f the map seems to be to indicate
the relative position o f the cities and to outline their road
system (Mehta, 1978—79 )*
B u d d h is m and M a r it im e A c t iv it y
It is not to be suggested that maritime activity was dom inated in

the ancient period by adherents o f Buddhism . Instead there are
86 H i m a n s h u P r a b h a Ra y
references to a range o f com m unities involved, owing affilia

tion to a variety o f religions, Brahmanism and Jainism. O ne
area in which Buddhism adopted an approach unlike its
contem porary religions was in evolving the notion o f a saviour
from the dangers encountered at sea (Ray 1994 : I 53 ~ 4 )-
Perhaps the earliest representation o f a seafarer in distress is to
be seen on a m edallion from Bharhut which shows a gigantic
sea-m onster threatening to swallow a boat (Cunningham 1879,
plate 3 4 > fig* 56 )- O n the basis o f the inscription the m erchant
has been identified as Vasugupta who was saved in his m om ent
o f peril by m editating on the Buddha.
C hronologically the next depiction is at Mathura (C u n n in g
ham 1871—7^: 2 0 —i) where the Bodhisattva in the form o f a
horse saves some ship-wrecked sailors from a group o f hungry
ja k s is or demonesses. Various versions o f the story occur in the
Jatakas and in Mahavastu and it is said to allude to the pseudo-
historical legend o f prince Vijaya’ s conquest o f Sri Lanka
(Sch lin gloff 1988: 256). The legend is represented in cave
XVII at Ajanta, whereas the painting in cave II depicts the
Buddhist m onk Purna flying down to save his ship-wrecked
brother, Darukarnin, the owner o f the vessel.
A related developm ent was the emergence o f the cult o f the
Boddhisattva Avalokitesvara as a saviour o f m ariners and
travellers in distress. This has generally been associated with
Mahayana and the Saddharm a-pundarika-sutra (chapter XXIV),
though an enum eration o f dangers o f travel is to be found in
earlier texts as well such as the Anguttara Nikaya (Kessivagga
119—20; vol. II: I2l) and the Divjavadana (p. 92: 25“ 8). T h e
cult was prevalent at the time that the Chinese pilgrim Fa H sien
came to India in the fourth century, though the first sculptural
representations o f the litany can be dated no earlier than the
fifth century AD (Ray 1994 : chapter V ). This theme is to be
found in the Buddhist caves o f Ajanta, Aurangabad, Ellora and
Badami, though nowhere is the com position so elaborate and
the treatment so elegant as in cave 90 at K anheri (Mitra 1980:
165)- For the first time around the sixth to eighth centuries A D ,
in the rock cut sanctuaries at Ellora, the function o f the
Avalokitesvara as the saviour from the eight perils is delegated
S e a f a r in g in P e n in s u l a r In d i a in t h e A n c i e n t P e r io d 87
to Tara (ibid.: 185). This theme occurs prom inently in the

monastic establishment at Ratnagiri in Orissa. A standing
image o f astamahabhaya Tara dated to the eighth century AD is
shown flanked by scenes o f the eight perils depicted in two
vertical rows o f four panels each (Mitra 1983: 4 4 4 —5)- A
second standing image is dated somewhat later to the eleventh
century on the basis o f a fragm entary in scrip tion and
graphically portrays the jalarnava-bhaya or fear o f drowning in
a sinking boat (ibid.: 4 ^8 ).
In the final analysis, certain conclusions may be drawn. First,
there is clear evidence for disjunction both in maritime
technologies in the nineteenth century leading to the m argin
alisation o f littoral comm unities involved in fishing, sailing and
boat-building (figs. 3-6 & 3 -7)* ^ a result, present boat
building and navigating traditions followed by fishing com m u
nities are not inevitably a reflection o f earlier practices.
Continuities are possible and do exist, but they need to be
confirm ed rather than taken for granted. Second, inform ation
on seafaring com m unities from historical and archaeological
sources indicates that their make-up cut across ethnic b o u n d
aries. H ence attempts at categorising them within present
national boundaries may be in keeping with current aspirations,
but will find little support historically. T h ird , historically as also
at present, boat-building was marked by diversity with the co
existence o f simpler forms such as dug-outs or rafts used for
fishing in local waters with more complex plank-built craft used
for carrying cargo and passengers. B oat-building should there
fore be studied in terms o f the functions it perform ed and the
needs it fu lfille d , rather than within an evolutionary framework.
This diversity o f water-craft required a range o f com m unities to
man and sail, including the use o f literate and skilled navigators
and pilots for long-distance voyaging. Fourth, while trade
provided an im portant motive for sea-travel, it was by no means
the only reason for travel by sea and needs to be studied within a
wider perspective o f seafaring activity and maritime contacts.
Finally, the role o f ideologies, such as those o f Buddhism or
Islam in motivating and supporting seafaring activity needs to be
recognised and accepted.
88 H im a n s h u P r a b h a R a y
Figu re 3 .6 Sh oe-dhoni along Andhra coast
Figure 3*7 Construction o f fishing craft

Se a fa r in g in P e n i n s u l a r In d i a in th e A n c ie n t P er io d 89
Notes
1 T h is is an argum en t that figures in several papers presen ted at the
C on feren ce on Seafarin g C om m u n ities in the In dian O cean h eld at the
M aison de 1’ O rien t M editerran een, Lyon, Ju ly , 199^ (R-ay an d Salles
fo rth com in g).
2 D ocu m en tation o f these craft has follow ed the tren d set by early
E u ro p ea n explorers to the region , with few exceptions. T h e seventeenth
an d eighteenth century observers often regarded in d igen ou s rafts and
boats as relics fro m In d ia’ s ancient past, not accom m odatin g in tern al
dynam ics o f water tran sp o rtation that considerably m o d ified the b u ild o f
larger vessels engaged in trade o r w arfare.
3 L ocated alo n g the b an k o f the A riyankuppam river, the earliest
settlem ent at the site dates to the Iro n Age M egalithic p e rio d and may
be term ed a fish in g village as suggested by one o f the excavators (C asal
1949: 31)-
4 A d escription o f Bom bay is illustrative o f this practice:
O rigin ally, sailing ships fo r the m ost part rem ain ed an ch ored on
the ro ad stead while go ods an d passen gers were lo ad e d and
discharged by m eans o f lighters which used an in creasin g n um b er
o f lan d in g places stretching fro m the C ustom s B an d er in the
south -east to M ahim to the north-w est. N on e o f these lan din g
places was satisfactory, as they could only be used interm ittently
d u rin g som e part o f the day while at low tide go ods and passengers
had to be carried through the m ud (Broeze 1 9 9 2 : 254 )-
5 T h is was true o f the east coast as well, as is evident fro m the follow ing
account:
From the south ern p oin t o f the coast o f C o ro m an d e l to the Bay o f

B alasore it is im possible to make a go o d lan d in g in E u ro p ean
boats. A n E u ro p ean b oat attem pting this passage w ould ru n the risk
o f touch in g on the bank and b ein g swallowed up by the waves. To
prevent this, flat-b o tto m ed boats called chelinques are con structed
without b eam s an d which have the planks sewed together instead o f
b ein g n ailed. T h is fo rm atio n gives them m ore elasticity; they are so
flat that they do not draw when loaded above six inches water, and
som e not even so m uch . . . they are in less danger o f b ein g filled.
They are generally m ann ed with nine Blacks, and when the sea runs
high with eleven. . . . W hen the sea ru ns so high that they are
appreh ensive o f an acciden t, they take extra p re ca u tio n by
p rovidin g a catimaron. . . . O n vessels o f this frail description the
natives o f In dia and particularly the islan ders o f the A n d am ans and
the Straits undertake lon g voyages by p uttin g up a sail (G ran d p re
1 8 0 3 :8 5 - 9 0 )
6 T h e significance o f these drawings has b een discussed by H ill:

T h e interest which attaches to Bowrey’ s boat drawings lies not in
any possible claim they may be su pp osed to have to ph otograph ic
objectivity. T h e perspective is often faulty, m aking the b oat look
m uch sm aller than he in ten ded. It is the structural details, the
ru d d er paddle, the outline o f the hull, the p ositio n o f the crew and
the superstructure, which repay a few m o m en ts’ study in the light
o f what other w riters have said about them (H ill 1958: 2 l 6).
7 H o rn e ll suggested that:
U n til the arrival in the In dian O cean o f the Portuguese at the end
o f the fifteenth century little im provem en t or change appears to
have taken place. It was the in tru sio n o f the E u rop ean s into the
trade o f the In dian O cean which b rough t about a revolution in the
design in g o f the larger craft o p eratin g there (197O: 2 3 6 ).
8 Early B uddhist texts provide details fo r the con struction o f m on um en ts

and refer to the p re p aratio n o f plans in verm ilion on lin en cloth b efore
u ndertakin g actual con struction ( Mahavamsa X X V II. 10.18). A few fra g
m ents o f pottery and sealing have been fo u n d in the archaeological
context at sites such as Kasraw ad and Salih u n dam which show geom etrical
designs id en tified as plans o f m on asteries (D iskalkar 1949 : I—18;
Subrahm anyam 1 964). Plans o f tem ples are also known to have been
drawn, the surviving exam ples on stone fro m the u n fin ish ed Siva tem ple
at B h o jp u r dated to the eleventh century b ein g strewn across the h ill-sid e
(Pande 1986: 3 3 ). In other con tem porary structures, these b lu ep rin ts are
often fou n d concealed on the u n derside o f paving stones. While
stonecutters probably used many o f these as guides, there are also
detailed designs fo r various p illars and an elevation o f the fro n t o f the
tem ple. A p olish ed red stone shows the gro u n d plan o f the in n er shrine
(ib id .).
9 N atu ral ph en om en a were also used fo r p redictin g unfavou rable weather
con dition s so essential to survival at sea. For exam ple, the bu n ch in g and
intertw ining o f sea snakes o ff the C o ro m an d e l C oast is co n sidered as
indicative o f an arriving d ep ression . Sim ilarly, the grou p flight n orth to
south o f the konga (a sea b ird ) and karigedu konge (black sea b ird ) on the
A n d hra coast is con sidered as a w arning o f a cyclone (A runachalam 1996:
270).
10 T h e presen t-day lighter appears as though it had evolved directly fro m
the p lan k-b u ilt canoe used in the coral trade, a course which it has not
follow ed as shown above; rather is it an instance o f gradual reversion to
the old b o at-can o e type (H o rn e ll 1945: 2 ^7)-
11 It has b een suggested that increase in size resulted fro m the growth and
expansion o f trade at T u ticorin .
In i9 6 0 there were som e 125 ^ oni type vessels registered at

T u tico rin an d ten at the n eigh bou rin g port o f K ulasek arip atin am .
A bou t 195° two or three vessels a year were b ein g built. However,
in 1955, there was a con siderable increase in con struction and
between the years 1958 and 1961 about 3 0 new vessels were b uilt. It
was d u rin g this p e rio d that hulls o f greater capacity came into
bein g, som e o f them stepping the extra mast (Hawkins 19^ 5 : I 49 )-
12 Physical evidence o f the presence o f such charts has now b een fou n d
am on g the K utch i seam en in the form o f a pothi o r m anual w ritten in
K utch i and dated to 1664 AD (now in the N ation al M useum , New D elhi
m ss. n o. 8 2 .2 6 3 ). T h e m anual is not com plete and only five m aps
rem ain. T h ese are line drawings o f lim ited coastal stretches, often
com b in in g silhouette profiles o f coastal landscapes as observed from the
sea (A runachalam 199^ : 279)*
4
A rch aeological E vidence
on S hipping C om m unities
o r S ri L anka
Osmund Bopearachchi
In his book: Taprobane: Ancient Sri Lanka as known to Greeks and Romans,
D .P .M . Weerakkody, correctly pointed out: 'S ri Lanka’ s
outstanding significance for international trade has been
recognized through the ages, and has been attributed to its
central position in relation to the countries o f the East and the
West. Cosmas Indicopleustes, who realized this fact as early as
the sixth century AD, called the island the 'm ediatrix’ between
the countries o f eastern and western Asia. His observations,
however, are only the climax o f a long series o f notices o f the
island by G reek and Rom an authors, most o f whom show some
awareness o f the com m ercial significance o f Sri Lanka’ s natural
sources’ .1 Apart from its central position in the Indian O cean,
in the middle o f sea routes between the east and west linking
the shores o f the Persian G ulf, Red Sea and Southwest Asia with
Southeast Asia and the Far East and its products o f high export
value such as precious stones, pearls, ivory, tortoise shell,
elephants, valuable woods, textiles and especially spices (cloves,
pepper and cinnam on), this island was gifted with num erous
bays, natural harbours, estuaries and navigable rivers facilitat
ing both sea-borne and inland trade.
T h e m ain aim o f this chapter is to show how the results o f my
research on the organisation o f maritime trade in the Indian
O cean connected with ancient Sri Lanka throw light on the
92
A r c h a e o l o g ica l Ev id en ce on Sh ipp in g C o m m u n itie s 93
role played by both local and foreign shipping com m unities,

during different periods o f its history. Obviously, it is beyond
the scope o f this chapter to deal with all the shipping
com m unities connected with ancient Sri Lanka. Besides,
various scholars have discussed the trading activities o f a
certain num ber o f them 2 and furtherm ore I too have already
dealt with this question in detail in the light o f numismatic and
literary evidence.3 So, I have made an attempt in this chapter to
highlight the activities o f some maritime com m unities whose
existence, in relation to Sri Lanka, is relatively little-know n. As
we shall see later, recent archaeological discoveries place them
in a different and im portant context. Although, it may appear
elementary, it should be emphasised that Sri Lanka being an
islan d, any item o f foreign origin has to be brought in through
the intervention o f mariners.
My study o f the shipping com m unities o f ancient Sri Lanka
is based on literary and epigraphical sources on the one hand
and archaeological evidence on the other. Foreign sources,
especially Indian, Greek, Roman, Persian, A rabic and Chinese
accounts, provide extremely valuable evidence on the flou rish
ing trade and econom ic prosperity o f ancient Sri Lanka.
A t least ten G reek and Latin authors give im portant evidence
on the m aritime trade activities and shipping comm unities
connected with Sri Lanka. A m on g them, the testimonies o f the
following authors are significant: O nesicritus o f Astypaleia
(fourth century B e),4 a com m ander o f A lexander’ s fleet and
the first G reek to write about Taprobane; Megasthenes (third
century B e),5 the ambassador o f Seleucus N icator to the
Mauryan em peror Chandragupta; Eratosthenes (third century
B e),6 the librarian o f the royal library o f Alexandria appointed
by Ptolemy Euergetes; Strabo (64 BC — AD 2 l ) , 7 geographer as
well as historian; Pliny the Elder, the Western writer who gives
the most comprehensive description o f Sri Lanka o f the first
century AD8; the anonymous author o f the Periplus Maris Erythraei,
most probably o f the first century AD; Claudius Ptolemy
(second century A d ),9 whose writings represent by far the most
comprehensive and scientific foreign account o f Sri Lanka that
has come down to us from classical antiquity; Dionysius
94 O smund Bopearachchi
(second century a d ) ,10 whose Periegesis is a G reek geographical

treatise written in hexameter; Palladius (fifth century a d ), 11 and
Gosmas Indicopleustes (sixth century a d ) ,12 probably an
Egyptian G reek who having made a long career in overseas
trade, became a m onk and devoted the rest o f his life to writing
theological works. T he most recent and complete study on all
the classical texts on Sri Lanka, by D .P .M . Weerakkody (1996),
is now available to the reader. C on cern in g the Chinese texts,
one may read the studies made by Sylvain Levi (1900), Paul
Pelliot (1903, 1904) and E .H . Schafer (1963). A good num ber
o f inscriptions, written either in Brahm i or in Sinhalese
scripts, dealing directly with the shipping com m unities, have
been brought to light by Senarath Paranavitana ( l 97 ° & ! 9 &3 )
and some other scholars. T h e Indian and Sinhalese texts
concerning political and econom ic aspects o f these m ariners
are analysed to a certain extent by N. Wijesekara (1952). For
Persian and A rabic connections with Sri Lanka, one may refer
to H. Hasan, 1928; G. Ferrand, 1914 ; S. K iribam une, 1990 *
Likewise the Sinhalese, Indian, Chinese and Arabic literary
sources have attracted a great deal o f attention from scholars
throughout the last two centuries, and I shall therefore confine
m yself to some observations which appear to me to be
particularly significant in the light o f archaeological evidence,
i.e. coins, ceramics, beads, seals, intaglios and all the other
relevant materials discovered either sporadically or in an
archaeological context.
T h e im portant archaeological discoveries made in Sri Lanka
especially during the last twenty years will be taken into account
in my analysis. New data have been obtained by Sri Lankan and
British archaeologists from the excavations conducted in the
citadel o f A nuradhapura.13 I shall also take the maximum
benefit out o f published data obtained from the excavations
conducted during the last fifteen years, at the Buddhist
monastery o f Jetavanarama14 at Anuradhapura, pleasure gar
dens at Sigiriya15 and Alahana Parivena at Polonnaruw a.16 We
have extremely lim ited inform ation on the discoveries made at
the three ancient seaports, Manthai, Kantarodai and K u ch -
chaveli, at the northwest tip o f the island.17
A r c h a e o lo g ic a l E vid en ce on S h ip p in g C o m m u n ities 95
In addition to these discoveries, there is new archaeological

and epigraphical evidence obtained from our own exploratory
program on the organisation o f maritime trade connected with
the ancient ports on the Western and southern coasts o f Sri
Lanka, launched by the French M ission o f A rchaeological C o
operation in Sri Lanka in collaboration with the A rchaeological
D epartm ent o f Sri Lanka ,18 throw much light on the various
maritime com m unities involved in international trade in the
Indian O cean.
I have pointed out elsewhere 19 that the most im portant
characteristic o f all the ancient ports o f the Western and
southern coast and many others around the island is their
geographical situation at the estuaries o f rivers. T h e location o f
em poria along rivers must have facilitated transactions with the
interior regions. T he starting point o f my investigations was the
spatial distribution o f ancient ports along the South Indian
coasts. It is significant that ancient ports like Ponnani, Muziris,
Porakad, K olkai, Karikal (Camara), Poduke (Putucceri) were
situated either on the banks or at the m outh o f the rivers,
especially o f the Ponnani, the A chenkoil, the Tam raparni, the
Kavaeri, the C enci. Arikam edu on the Ariyankuppam river,
Kaveripatinam at the Kaveri river and Alagankulam on the
Vaigai river are well known sites o f this nature .20 As K .
K arttunen ( l 9 9 5 : 85) has well observed: 'All the three Tamil
capitals were inland towns, but each had one or several marts
on the coast’ . For K arur, the capital city o f Chera kings,
situated at the Amaravathi river, a tributary o f the Kaveri river,
and for U raiyur, the capital city o f C hola kings, the main port
was Kaveripattinam , also called Pum puhar, on the Kaveri delta,
XaPrjpic; ’sjmroplov o f Ptolemy (7, I, 13)- F °r M adurai, the capital
o f Pandyas, a direct river connection along the Vaigai led to
Saliyur (m odern A lagankulam ), Sfjôup ’sjuTcopiov Baicov, o f
Ptolemy (7, I, II) near Rameswaram . . .’ .2I
As in India, the most im portant ancient capitals o f Sri Lanka
were inland, but each had one port on the coast. Manthai, the
most active port in ancient Sri Lanka, is located close to the
A ruvi A ri river which linked the port to the inland capital o f
Anuradhapura. Likewise, the geographical situation o f the
96 O sm und B opearach ch i
■Jaffna 1
Psilk ^
St rai t
INDIAN
OCEAN
Manthai ^
(Mannar)
4ruvi Aru
Trincomalee
«, Gokanna
Uruvelapatnam
Anuradhapura
Kala Oya
Giribawa
vSigiriyaY
‘Tputalam
Polonnaruvva
/
Efeduru Oya
Batticoloa
Salvattota I
Mahaveli Ganga
(Chilow) ^ Nariyagam^\
Panirendawa
Kurunagala
Maha Oya
Meegomuwa i -/Tedigamj
(Negombo) V
Wattala Pilapitiya
^Kelaniya^—v Kehmi Ganga
Kalu Ganga
ESyagama
Kalalittha\|
(Kalatura)\ Kirindi Oya
BhimatitthaN Bentora Ganga

(Bentota) Ridiyagama
Tissamaharama
NiJwaJa Ganga Walawe Ganga
Kirinda
Gimhatittha
(Gintota) Gothpabbata (Godavaya)
Mahavalukagama"
(Weligama) Nilwalatittha (Matara)
50 km
F igu re 4 .1 A n cien t ports o f Sri Lanka

A r c h a e o l o g ic a l Ev id e n c e o n Sh ip pin g C o m m u n it ie s 97
V asavasam udram
\« /
Ponnaiyar
Arikam edu
•/
'Karur
Karaikadu
Kodumanal
Kaveripatinam*
Ponnani Kaveri
Uraiyur
Muziris
SyKochi Vaigai
iA lagankhulam
Kollam
-s Korkai Manthai
V • /
Anuradhapura
lU ruvelapttna
Sri L a n k a
F ig u r e 4 *2 Trade Centres and ancient ports in South India
ancient capital o f Polonnaruwa on the banks o f the Mahavali

River, which flows to the sea at Gokanna, is not a co in
cidence.22 In the same way, Tissamaharama, Sri Lanka’ s ancient
city in the south became im portant in its glorious past because
it was established along the higher ground on the left bank o f
the K irin d i Oya connecting with the ancient port site o f
Kirin da.
The following sea ports subjected to our investigations are all
situated at the estuaries o f rivers: Salavattota (Ghilaw) at the
D eduru-oya, Wattala at the K elani Ganga, Kalalittha (K alu-
tara) at the K alu Ganga, Bhimatittha (Bentota) at the Bentota
Ganga, Gim hatittha (Gintota) at the G in -G anga, Mahavalu-
kagama (Weligama) at the Polwatta Ganga, Nilwalatittha
(Matara) at the Nilwala Ganga, Gothapabbata (Godavaya) at
Walawe Ganga and K irin da at the K irin d i Oya.
98 O sm und Bopearachchi
We obtained positive results from the excavations and

explorations conducted at Giribawa on the left bank o f the
Kala Oya, which flows to the sea at U ruvelapttna;23 at
Nariyagama, on the left bank o f Daduru O ya,24 which flows
to the sea at the ancient port o f Salavattota; at the village o f
Pilapitiya, on the right bank o f the K elan i Ganga, about seven
kilometres from the ancient sea port o f Wattala;25 at Diyagama,
on the right bank o f the K alu Ganga which join s the sea at the
ancient sea port o f K alalittha;26 and at Ridiyagma on the left
bank o f the Walawe Ganga which flows to the sea at the ancient
sea port o f Gothapabbata.27 It is beyond the scope o f this
chapter to describe each and every discovery o f our excavations,
however their im portance will be discussed in order to throw
light on the organisation o f some o f the shipping com m unities
o f ancient Sri Lanka.
Sri Lanka, because o f its geographical situation, established its
first trade relations with the Indian sub-continent. P ro to -
historic Sri Lanka was most closely linked with South India. In
the excavations conducted at Gedige (Anuradhapura), Pom par-
ippu, Kantarodai and Ibbankatuwa, substantial quantities o f
potsherds were found which parallel the Iron Age and early
historical wares o f South India, such as Megalithic Black and Red
Ware and the Rouletted Ware o f the Arikam edu types.28 As V .
Begley pointed out (197O: 94 ) > the chronological position o f
these sites parallel Arikam edu. A ccording S. Suresh ( l 993 )
around ninety Indian sites spread throughout most parts o f the
country have revealed the rouletted ware. T h e maximum num ber
o f finds is, not surprisingly, from Andhra and Tam ilnadu
region. He was correct to emphasise that the rouletted ware is
known to occur throughout the C orom andel coast and also in
Sri Lanka indicating a well-established com m unication network
linking the entire eastern coast o f India with northern Sri Lanka.
Th e results o f our own excavations and explorations on the
Western and southern coasts o f Sri Lanka, show that not only the
north, but also the west and south o f the island, should be
included in the com m unication network.
We were also able to collect, not only at Ridiyagama, but also
along the Walawe Ganga, large quantities o f early megalithic
A rch a e o lo g ic a l E vid en ce on S h ip p in g C o m m u n ities 99
Black and Red Ware, some o f which was engraved with early
historic symbols. Several discs made using broken rouletted
ware were unearthed from Tissam aharama.29 T h e excavations
conducted by us last year, at the village o f Pilapitiya close to
Kelaniya, about seven kilometres from the estuary o f the
K elan i Ganga, yielded not only early historic Black and Red
Ware, but also several fragments o f im ported Rouletted Ware,
similar to the ones found along the South Indian coasts. This
was the first time in the study o f Sri Lankan history that a site
situated in the western Wet Zone, away from the capitals o f the
Sinhalese kings in the northern Dry Zone, yielded archae
ological material which can be dated with certainty back to the
fourth and third centuries BC. It is also interesting to note that
as in Algankulam ,30 the Indian sea port closest to Sri Lanka, the
excavations at Pilapitiya (Kelaniya) revealed three out o f five
m ajor types o f rouletted ware,31 such as greyish pink ware (grey
slip inside, brown to sepia outside), greyish pink ware (black
slip inside, brown slip outside) and grey ware (black slip inside
and outside).
A part from ceramics, beads and coins found in southern
India and Sri Lanka highlight the close com m unication
networks linking both countries. H undreds o f beads made o f
crystal, glass, stone, ivory, bone, shell, clay and above all
semiprecious and precious stones, were found at Ridiyagama
and Kelaniya. A m on g the beads o f semiprecious and precious
stones, carnelian, lapis lazuli, rock crystals, agate, amethysts
Were found in hundreds. T h e discovery o f im perforated beads
together with fragments o f semiprecious stones confirm s the
existence o f a bead making industry at the site o f Ridiyagama. It
should also be emphasised that Ridiyagama is one o f the gem -
m ining sectors located away from the m ajor gem -bearing area
o f Sabaragamuwa Province.
In recent years there have been a series o f excavations at the
early historical sites o f A ndhra-Tam ilnadu. Beads constitute
an im portant class o f finds in most o f the southern Indian
sites. These sites include Amaravati, Dhulikatta, Kotalingala,
Peddabankur and Yeleswaram (in A ndra Pradesh), Arikam edu
(P on d ich erry), K an ch ip u ram , A p pu kallu , T iruvam ath ur,
Karaikadu, Mallapadi, Perur, K odum anal, K arur, Uraiyur and

Alagankulam (in Tam ilnadu). Some o f these, especially the
T am ilnadu sites, have yielded coins (both R om an and
indigenous), ceramics and especially beads similar to those
reported from Manthai, Anuradahapura, Kelaniya, Ridiyaga-
ma, Tissamaharama and other early sites o f Sri Lanka. T h e
beads from Ridiyagama are very similar, in colour and shape,
to the types recovered from four m ajor sites o f South India viz.
Arikam edu, Karaikadu, Uraiyur and Alagankulam .32
T h e beads that Peter Francis Jr. (1987- 8—9) classifies as
round beads o f India red glass, and small glass beads o f
different colours made at Arikam edu, were also attested to in
hundreds at Ridiyagma. T h e bead making industry at A rik a
m edu was large and productive.33 T h e m ajority o f the beads
found here are spheroid or pear-shaped, similar to those from
Ridiyagama. L ug-collared beads, identical to the ones from
Ridiyagama, have been reported in small quantities in the so-
called A rretine and post-A rretine periods o f Arikam edu i.e.
from the first century BG to the end o f the second century A D .
T h e small scale excavations conducted at Karaikadu, the
coastal site situated 4 ° kms south o f Arikam edu, have revealed
large quantities o f finished and sem i-finished stone glass beads
in association with a large brick structure dated to the first
century A D . A ccording to the excavators, this structure may have
been used for the production o f beads. M onochrom e drawn
glass tubes, like the ones from Ridiyagama and Giribawa, have
been unearthed in large num ber both at A rikam edu and
Karaikadu. These tubes were meant to be made into beads by
annealing them, smooth and round. Some o f these tubes are
believed to be the prototypes o f the collared beads. Likewise
Alagankulam, the extensive site on the m outh o f the River
Vaigai, is strategically located close to Sri Lanka. Beads made o f
precious and sem i-p reciou s stones and glass have been
recovered from stratified layers ranging in date from the
second century BG to the third century AD and even later. It is
now clear that most o f the beads collected in our excavations or
from our surface explorations at Ridiyagama are also co n
firm ed in all the im portant settlement sites o f South India. T he
A r c h a e o lo g ic a l Evid en c e on Sh ip pin g C o m m u n ities io i
conclusion to be drawn from this discovery is that they

belonged to the same trade network.
From the early period onwards, the South Indian m ercantile
and m ilitary com m unities, like Cholas, Pandyas, Cheras,
Virakkotiyar, Nanadesis and Velaikkarar ,34 in different periods
played an im portant role in the econom ic and political History
o f the island .35 T h e earliest references to Damilas in the
Mahavamsa, is in connection with Sena and Guttika, who
described them as assanavika, i.e. traders who came in ships
bringing horses for sale from South India .36 T he epigraphical
evidence for the active role played by the Tam il merchants in
the early phase o f Sri Lanka’ s history is num erous.
The inscription in early Brahmi script on a boulder in the
area to the north-West o f the ancient Abhayagiri Dagaba at
Anuradhapura, records that the terrace (Pasade) was o f the Tamil
householders (gahapatikana) and was made by Samana, the Tamil,
o f Ilubarata. The record states that the Tamil ship-captain
(Navika-Karavaha asane) was entitled to the seat o f honour and
probably was the leader o f the Dameda householders .37 Two
Brahmi inscriptions from Periya-Puliyankulama, are both o f the
same personage, a Tamil merchant named Visaka, referred to as
a householder (gapati).3S Perhaps the most evocative inscription
in this context is the one from Kuduvil in the Am pari District.
As Paranavitana (197O: xc) correctly observed: 'Dighavapi, given
as the place o f residence o f these brothers, was a seat o f royalty in
Rohana, second in importance to Mahagama only, and the place
might well have attracted merchants from foreign countries who
practised their own customs ’ .39
A recent study undertaken by I. Mahadevan (ig96)has
revealed the existence o f a num ber o f inscribed potsherds in
the Sinhala-Prakrit (old Sinhalese) language written in the
Brahm i script, found at or near ancient sea ports along the east
coast o f India. T h ey were rep o rted from K od u m an al,
Arikam edu and Algankulam .40 A ccording to palaeographic
and linguistic features, these inscriptions can be dated from
second century BC to first century A D .
A large number o f coins o f the South Indian dynasties
belonging to the historical period, were found in the island.
10 2 O smund Bopearachchi
Apart from the coins already published by Codrington, a

number o f new series o f Pandya coins have been attested in
recent years. The Sri Lanka-British Excavations at Salgaha Watta
in the ancient citadel o f Anuradhapura brought to light some
interesting coins o f the South Indian dynasties. The most
noteworthy specimen found in this period is the Pandya inspired
multi type coin.41 This coin is the only known specimen o f this
type depicting a caitya surmounted by a chatra, recalling the earliest
type o f stupa in India and Sri Lanka.42 The next local coinage o f
the island dates from the resumption o f independence in 28 BC
from the Pandya occupation. The new series o f multi-type
copper coins bears a prom inent elephant symbol. The coins
labelled as elephant and swastika, horse and swastika, lion and
swastika, tree and swastika and especially Lakshmi plaques
depicting the goddess Lakshmi seated or standing, belong to
the category o f local issues. Some o f these coins, certainly struck
in Sri Lanka, were found in the coastal regions o f South India.
Significantly a Lakshmi plaque o f Sri Lanka was recovered from
the river bed o f Amaravathi, near K arur, a city situated more
than two hundred kilometres from Kaveripatinam.
It is quite well known that after the decline o f the Gholas, the
Pandyans, under the leadership o f Aryachkravarty, a general o f
Kulasekhara, the Pandyan king, made several attempts to control
the trade along the western coast line o f Sri Lanka, during the
reign o f the Sinhalese king Bhuvaneka Bahu I (1284—I 29 1 a d ) ,
by conquering the ports o f Salavattota (Chilaw), Meegomuwa
(Negom bo), Wattala (Kelaniya), and C olo m b o.43
A m on g the other m ercantile com m unities o f South India,
the Nanadesis enjoyed great celebrity both in India and Sri
Lanka. As K .A . Nilakantha Sastri (i9 6 0 : 597 ) summarises their
activities: 'The Nanadesis then were a powerful autonom ous
corporation o f merchants whose activities apparently took little
or no account o f political boundaries. They visited all countries
in the course o f their trade, and everywhere they received
recognition alike from the central governm ent and from local
agencies like the village sabhas. They had their own m ercenary
army, doubtless for the protection o f the merchandise in their
warehouses and transit’ . As S. Pathamanathan ( l 9 9 0 ; *45 )
A r c h a e o l o g ica l Ev id en ce on Sh ipp in g C o m m u n itie s 103
observes: 'T h e largest num ber o f inscriptions and other

archaeological artefacts left behind by the Nanadesis and their
associates in any region outside the kingdoms o f South India
are to be found in Sri Lanka, where inform ation pertaining to
them has been steadily accumulating in recent years’ . These
records date from the first quarter o f the eleventh century to
the fifteenth century.
A cast bronze seal o f the Nanadesis discovered by a cultivator
in a field o f Ridiyagama, published by S. Pathmanathan has
great significance .44 T h e inscribed label on the surface o f its
pedestal contains the expression Nanadesin sakai, which Pathma
nathan (1984: 1^6) translates as 'the friends, allies or
companions o f the Nanadesis’ . So no doubt the seal had been
used for stamping receipts for the payment o f tolls, or cloth
and packs o f com m odities.
The second part o f the early historic period o f Sri Lanka

begins with the introduction o f Buddhism to the island by Thera
M ahinda, the envoy o f the Mauryan K in g Asoka during the reign
o f Devanampiyatissa (c. 2 5 ° ~ 2 I O B e ) . It was from this period
onwards that close political, cultural and com m ercial relation
ships were first established with N orth India.
However the island was known to the Indians o f the Indus
valley as early as the fourth century BC. Onesicritus o f
Astypaleia, a com m ander o f A lexander’ s fleet was the first
Greek to write about Taprobane (Sri Lanka) and he certainly
had access to inform ation from the Indians who were in contact
with the navigators sailing to Sri L an ka .45 As D .P .M .
Weerakkody ( l 997 : 29) points out: 'It is now generally accepted
that the Indians, who migrated to Sri Lanka in early times,
came from the western as well as the eastern part o f north
India. Those who came from the eastern parts are known to
have m aintained regular and intimate contacts with their
form er land, and there can be no doubt that, at least in the
earlier p eriod , their western counterparts did the same’ .
That m ariners from Sri Lanka continued to make voyages to
the western part o f India is evident from the early Brahmi
inscription from Andiyagama in the Anuradhapura district, in
io4 O sm und Bopearachchi
which a m ariner describes him self as having gone to Bhoja-

kataka: Bhojakatakasa navikasa padagadini, 'The steps o f the m ariner
(travelling to) Bhojakata’ .46 R .A .L .H . Gunawardana aptly
correlates this evidence with the rep ort o f O nesicritus,
pointing out that O nesicritus had apparently noticed certain
sailing vessels used on the route from Sind to Sri Lanka.47
It is most likely that Sri Lankans m aintained friendly
relationships with Northwest India even at later periods. T h e
M ahavam sa, the great ch ro n icle o f Sri Lanka, m entions
Buddhists from G reek territories o f the Paropamisadae, among
the foreign delegations who were invited by the Sinhalese king
D utugam unu, a close contem porary o f the In d o -G re e k
M enander (middle o f the second century B e),48 for the
inauguration o f the great stupa Ruwanvali D agoba.49 Frantz
G renet (1996: 67—9) argues in one o f his recent articles that
Amoghavjra, one o f the principal initiators o f Tantrism in
China, according to Chinese biographies, was b orn in Sri
Lanka, and later became a Sogdian m erchant. Sri Lanka’ s
m aritim e com m erce began to develop by leaps and bounds
once trade links were established with the Persian G ulf.
Cosmas Indicopleustes, the Egyptian G reek o f the sixth
century AD, bears witness to the presence o f Persian traders in
Sri Lan ka.50 A cco rd in g to a description in his Christian
Topography, Sri Lanka played an im portant role in transmitting
m erchandise between East and West, a role once perform ed by
western India. Cosmas (XI, 15), dem onstrating the central
position that the island held in international com m erce, said:
\ . . the island is a great resort o f ships from all parts o f India
and from Persia and Ethiopia, and in like m anner it dispatches
many o f its own to foreign ports. A n d from the inner
countries, I mean C hina and other marts in that direction, it
receives silk, aloes, clo ve-w o o d , sandalw ood, and th eir
products, and these it again passes on to the outer ports, I
m ean to Male, where pepper grows, and to Kalliana, where
copper is produced, and sesam e-wood, and material for dress;
for it is also a great mart o f trade; and to Sindu also, where
musk or coster is got, as well as Androstanchm e, and to Persia
and the H om erite country, and to A dole. Receiving in return
A r c h a e o l o g ic a l E v id e n c e o n Sh ipp in g C o m m u n itie s 105
the traffic o f these marts and transmitting to the inner ports,

the island exports to each o f these at the same time its own
products’ .51
As we shall see later, in addition to literary sources, the
epigraphical and archaeological evidence also shows that there
was regular intercourse between Sri Lanka and Central Asia.
It is interesting to note that in the citadel o f Anuradhapura,
as in India, fine Grey Ware and N orthern Black Polished Ware
were found in successive strata.52 Likewise, most o f the finest
im ported ceramics o f this period found in Sri Lanka were from
N orth India.
Punch-m arked coins known in ancient India as ' karshapana or
Kahapana foun d everywhere in the island, were certainly
im ported from N orth and Northwest India. Alm ost all the
coins reported from Sri Lanka belong to the Im perial series
that were m inted over a vast area o f India under the protection
o f a unifying authority.
T h e earliest epigraphical evidence for the circulation o f
karshapana on the island dates back to the end o f the third
century B C. Th e inscription o f M am pita-vihara53 (Kagalla
District in the wet zone), written in early Brahm i script and
referring to Kahapanas, indicates that trade, even in the early
days, was not barter alo n e.54 C odrin gton has assembled most o f
the epigraphical and literary references to payments in
thousands o f karshapana made by kings on different occasions,
for the construction o f religious m onum ents and donations to
monastic com m unities.55 It should be noted that the third
structural period I o f the Sri Lanka-British Excavations at
Salgaha Watta 2 , in the ancient citadel o f Anuradhapura (dated
between the third and second centuries B e ) brought to light
three Punch-m arked coins. These coins are found in th ou
sands, in hoards and as stray finds, not only at Anuradhapura,
the ancient capital o f Sri Lanka, but also in every sea port on
the north, West and southern coasts o f the island.56 T h e most
rem arkable fin d made in recent times was a terracotta
m oneybox containing five punch-m arked coins.57 Indian and
classical literary sources refer to Sri Lankan exports, especially
pearls, precious stones and textiles, for which there was a good
io6 O smund Bopearachchi
Indian market. So the Indian early issues cannot be isolated

from the other Indian imports attested in the same area.
Punch-m arked coins ceased to be issued in India after the
decline o f the Mauryan em pire, and India’ s earliest coins were
then replaced by the issues o f the Indo-G reeks followed by the
Indo-Scythians, the Indo-Parthians and the Kushans all o f
whom occupied the northwestern provinces o f the Mauryan
em pire. A certain num ber o f coins issued by all to these
dynasties o f diverse political and cultural origin, have been
found in Sri Lanka. N o doubt, when com pared to the
thousands o f karshapana, the Sri Lankan finds o f coins issued
by the successors o f the Mauryans are quite rare. Yet one
cannot deny their econom ic im plications. T h e next most
ancient coin after the karshapana to be found in the Sri Lankan
soil is an Indian-standard drachm o f the In d o -G reek king
M enander ( l 55 —13 ° BC) .58 Two posthum ous im itation coins in
the name o f H erm aios have been reported from two different
seaports o f the island, one from Puttalam and the other from
Tissam aharama .59
I have seen in private collections about ten coins o f Soter
Megas, which were hitherto unknown in the Sri Lankan
context .60 They were all supposed to be stray finds from
different places in the island. Soter Megas had been considered
an anonym om e ruler calling him self the K in g o f Kings, the
Great Saviour’ . Thanks to recent discoveries we know today that
Soter Megas’ dynastic name was Vim a Takto and that he was the
grand-father o f the Kushan king Kanishka I .61 T h e coin o f
Kanishka II, o f the K ushan dynasty, which was found in the
excavations conducted at Jetavanarama, under the U n esco-Sri
Lanka Project o f the C ultural Triangle, is significant in this
context .62 T he coin o f Kanishka II was found buried at the foot
o f one o f the frontispieces o f the Stupa along with two other
im portant coins: a silver coin o f Viradam an o f the Western
Ksatrapas and a bronze coin o f the Rom an em peror Trajan
struck in the city o f D o ra .63 I have seen at least ten m ore coins
o f Kanishka II in two private collections, and all o f them were
found on the southern coast o f the island. H .W . C odrin gton
(19 2 4 : 4 9 ) m entions four specimens o f the K ushan king
A r c h a e o l o g ic a l Ev id en ce o n Sh ipp in g C o m m u n itie s 107
Vasudeva bought in C olom bo. Several bronze coins o f Vasudeva

II, still unpublished, were found as a result o f clandestine
diggings alone with the posthum ous im itation coin o f
H erm aios m entioned earlier at Tissam aharama.64
A part from coins, the finding o f carnelian and lapis lazuli
beads and intaglios, not only at Mantai and Anuradhapura, but
also from my recent excavations and explorations at Ridiya
gama, is o f greatest significance, because both categories o f
stones were certainly im ported to the island from N orth India
and Afghanistan. As we know, carnelian belonging to the
chalcedony group is not found in Sri Lanka and was certainly
im ported from Gujarat, where, according to the archaeological
evidence, it was produced without interruption from the early
historic period. It is well known that the reddish colour o f
carnelian is artificially produced by heating dull brown stones
with a high iro n con ten t.65 The author o f the Periplus m entions
on three occasions that these stones were exported from
Barygaza.66 It is interesting to note that the late Brahm i
inscription from Bagavalena (Sri Lanka) refers to a cave
donated by a certain Mala, a m ariner sailing to Bharukaccha.
We know that Bharukaccha was a w ell-know n port in G ujarat.67
T h e num ber o f carnelian beads collected as surface finds at
Ridiyagama exceeds one hundred. In addition, the perforated
beads from Ridiyagama are similar to the ones found at the
Ibbankatuwa megalithic cemetery (c. 770 —395 B e ).68 T h e
presence o f early historic Black and Red Ware and N orth
Indian carn elian beads both at Ridiyagama and at the
megalithic cemetery o f Ibbankatuwa is not accidental. T he
most interesting discovery in this respect is a carnelian blank.
A n oth er carnelian intaglio o f the same dim ensions, depicting a
figure holding a flower found at Kataragam a,69 indicates that
seal carving was practised in the region, using carnelian
im ported from India. T h e same type o f carnelian blanks was
also found in the recent Jetavanarama excavations.70 T he
m ajority o f intaglios and carved jewels from the Jetavanarama
project were carnelian. O n e carved carnelian seal depicting a
figure holding a globe is certainly an im itation o f a Rom an
p rototype.71
io8 O smund Bopearachchi
The second category o f beads which deserves attention, is

those made from lapis lazuli, becamee the only known source for
this m aterial in antiquity was Badakhshan (in n orth ern
Afghanwestan). The author o f the Periplus mentions lapis lazuli
among the products exported from Barbaricum .72 This precious
material doubtless travelled along the sea route to reach the
southern coast o f Sri Lanka. Hema Ratnayake (1993: 84) has
also observed that on a painted slab belonging to one o f the
frontispieces o f the Jetavana stupa, there are traces o f lapis lazuli
underneath the line o f geese. He dates it to the third century AD,
to the reign o f king Mahasena, who built this feature o f the
stupa. The intaglio depicting a seated wild boar, unearthed along
with carnelian seals and beads from Akurugoda (Tissamaharama)
on the southern coast o f the island, is important in this context.
This type o f wild boar is known on Sasanian intaglios .73
T h e presence o f lapis lazuli on the southern coast o f Sri
Lanka cannot be an isolated event, because epigraphical
evidence bears witness to the fact that this area had close
relationships with the regions o f Afghanistan. 'K ab oja’ occurs
as a proper name in three inscriptions from Koravakgala
(Situlpavua) in the Hambantota District, on the southeastern
part o f the island, in ancient R ohana .74 S. Paranavitana (1970:
xc) believed that the K abojha, Kabojhiya and Kabojhika are to
be connected with the ethnic name Kam boja, which occurrs in
Sanskrit and Pali literature as well as in the V th and XHIth
inscriptions o f Asoka, Kabojhiya being equivalent to the
derivative term Kam bojiya and K abojika to K am bojika .75 T he
Brahm i inscription from Bovattegala on the southern border o f
the A m pari District, a few miles from the northeast lim it o f the
H am bantota D istrict, also in ancient Rohana, refers to
'Kabojhya-m ahapugiyana i.e. 'those who were members o f the
great corporation o f the 'K abojhivas ’ .76 T h e Brahm i inscrip
tion from Kaduruvava in the Kurunagala District, to the
southwest o f A nuradhapura, m entions a parumaka (C h ief) o f the
G ota-K abojikana, i.e. o f the corporation o f the K abojikas .77
These inscriptions indicate that the Kam bojas had organised
themselves into a corporation and were certainly engaged in
trade. T h e Sihalavatthu, a Pali text o f about the fourth century,
A r c h a e o lo g ic a l Ev id e n c e on Sh ip pin g C o m m u n ities 109
attests that a group o f people called the Kambojas were in

Rohana. In the third story o f this text, called Metteyya-vatthu, we
are inform ed that the Elder named Maleyya was residing in
Kam boja-gam a, in the province (janapada) o f Rohana on the
Island o f Tam bapanni .78 T h e Kam bojas are often m entioned
together with Yonas (Yavanas), Gandharas and Sakas. T h e
Kam bojas were a native population o f Arachosia in the extreme
west o f the Mauryan em pire, speaking a language o f Iranian
o rig in .79 T h e finds on the southern coast o f the island o f lapis
lazuli from northern Afghanistan and various coins o f Soter
Megas, Kanishka II, Vasudeva II and posthum ous H erm aios, all
from Bactria and Northwest India, and the references to the
Kam bojas o f Arachosia, com pel us to believe that there were
close relationships between Sri Lanka and the com m unities o f
Central Asia and North-W est India.
S. Paranavitana (197O: xci) did not exclude the possibility o f
the presence o f Sakas in the island. His starting point was the
inscription in Brahm i script, known as A nuradhapura Rock
Ridge West o f Lainkaram a ,80 which refers to 'The flight o f steps
o f Uttara, the M urundiya (Muridi—Utaraha seni). Since the epithet
'M u rid i’ is prefixed to the name ’ -U tara’ (Skt. Uttara), S.
Paranavitana believed that M uridi is a derivative o f M uruda,
which is the same as M urunda in the com pound Saka-M urunda
that occurs in the Allahabad inscription o f Samudragupta. S.
K onow (1929: XX), referring to the same inscription argued,
that murunda is almost certainly a Saka word m eaning 'm aster’ ,
'lo rd ’ , and he argued that the word murunda has becom e
synonymome with Saka, when applied to royalty.
Apart from the coins, beads and intaglios, the contacts
between Sri Lanka and the Gandhara region are revealed by
other pieces o f archaeological evidence from recent excavations
at various sites. A fragm ent o f a Gandhara Buddha statute in
schisst, still unpublished, was unearthed from the excavations
at Jetavanarama. Most o f the identified 'H ellenistic’ and
G reek-influ enced pottery from the citadel o f Anuradhapura,
and from our recent excavations at Kelaniya appears to be from
the G reek East, in other words, somewhere in Northwest India
or Bactria .81
iio O sm und Bopearachchi
The uninterrupted trade contacts o f Sri Lanka with Persia,

Central Asia and Northwest India are revealed by the recent
finds o f Sasanian ceramics, bullae and coins at Mantai,
A nuradhapura and Tissamaharama. S.U . Deraniyagala (l992:
713) reported that pale blue 'Sasanian’ glazed ware appeared for
the first time at the Anuradhapura citadel excavations as
foreru nner to the darker blue glazed varieties o f middle
historic times. T h e excavations at Jetavanarama yielded Partho-
Sasanian and Indo-Sasanian ware .82 J. Carswell (199O: 26)
published a baked-clay bulla, typical o f the sixth and seventh
centuries AD from excavations at Mantai with three seal
impressions on it: a tw o-hum ped Bactrian camel, a Persian
inscription and a Nestorian cross. O ne cannot avoid drawing a
parallel between the N estorian cross on the Sasanian bulla and
the stone cross now kept in the Anuradhapura Museum.
Persian Nestorian Christians were responsible for the increas
ing trade activities between the Sasanian em pire and the island.
Cosmas Indicopleustes, 'Indian Navigator’ , makes allusions to a
Nestorian com m unity, probably composed o f Persian m er
chants which was well established on the island .83
In publishing three coins o f the Sasanian king Yezdigerd I
( a d 397 - 417 ), C odrin gton put forward the hypothesis that
occasional finds o f small copper coins among the 'third brass’
provide evidence o f the dealings o f Persians on the island .84 I
have added three m ore Sasanian coins hitherto unknown in a
Sri Lankan context .85
Since then two m ore Sasanian coins have been reported
from the southern coast o f the isla n d , especially o f Shapur II
and another H orm izd IV. O ne Kushano-Sasanian coin o f
Piroz II (c. AD 3OO—325) ° f Gandhara type 86 and o f Shapur II
(fourth century a d ) o f the Taxila type were unearthed from the
excavations at Jetavanarama (Anuradhapura). A rock crystal
intaglio found at Akurugoda recalls a Sasanian prototype in its
depiction o f a scorpion seen from above .87
T he political and com m ercial relationship with C entral Asia
and Persia reached its peak during the reign o f king Kasyapa
( a d 4 7 7 —4 9 5 ) - I have also shown elsewhere, when defining the
function o f the interior city, the royal pleasure gardens and the
A r c h a e o lo g ic a l Evid en c e on Sh ip pin g C o m m u n ities h i
palace, that the layout o f the ancient site o f Sigiriya built by

king Kassapa ( a d 477 “ 495 ) was>to a certain extent, influenced
by Persian tradition .88 The lofty rock o f Sigiriya rises to a
height o f 360m above mean sea level and nearly 200m above
the surrounding plain. T he Sigiriya, (Sinhagiri, the 'm ountain
o f the lio n ’) can be seen as the abode o f a god-king built in a
m agnificently impressive situation. It is centred on the steep-
sided crag on which are located the citadel and palace which
form the focal point o f the whole complex. T h e natural
features are skilfully used to achieve impressive landscape
effects. T h is fortress city is a m asterpiece o f landscape
architecture o f ancient Sri Lanka.89
From the beginning o f the historical period o f the island,
which coincides with the introduction o f Buddhism, Sinhalese
architecture was dom inated by the requirem ents o f Buddhism
and by the desire o f successive kings to demonstrate their power
and acquire m erit by their activity in the erection o f buildings
for public purpose or for display. T h e most notable feature o f
the Sigiriya complex is the total absence o f any religious
building. As I know, even the small and short-lived capitals o f
Sri Lanka are almost covered with Buddhist architecture. A fter
he had killed his father, K in g Kassapa, knowing that he would
never be pardoned by the Buddhists for his crime, would have
taken refuge in this rock and have converted the whole site into a
non religious establishm ent.90 This complex shows the aston
ishing skill o f the architects in the organisation o f space and the
feeling for classical structure. Senerath Paranavitana correctly
interpreted the whole complex o f Sigiriya, as Alaka in the
Himalayas and Kassapa as a Devaraja or Lord o f A laka, that is
Kuvera or his representative on earth. Paranavitana, quoting W.
K irfel and J. Przylmeki, admitted that the concept o f Chakra-
vartin or 'universal m onarch’ in the Indian tradition derived
from the Ziggurat 'cosmic m ountain’ o f M esopotam ia.91
In spite o f the enorm ous chronological difference, the L -
shaped gardens o f Sigiriya rem ind us o f the Mughal 'chahar
bagh’ or four gardens o f water channels at right angles. Mughal
gardens Were introduced to India by the Mughal em peror Babur
after 1526, using an Iranian schema. T h e chahar bagh’ was the
ideal garden, the earthly paradise, providing coolness and shade

to offset the heat. T h e close contacts that Sri Lanka had during
that period with the Persians as it is revealed in the narration o f
Cosmas and the fin ding o f Sasanian ceramics and coins tempt
me to observe a possible Sasanian influence over Sigiriya.
T h e love o f flowers and interest in gardens have been
features o f life in Iran down the centuries. In Persia, the
concept o f Gardens as an integral part o f the royal palace
com plex was b o rn at the time o f Cyrus, around the m iddle o f
the sixth century B C . O n e o f the dom inant ideas o f this
conception was to create a landscape which included architec
tural elements, from which evolved the landscape garden later
popularly known as 'chahar bagh’ .
As David Stom ach ( l 9 9 4 : 2) correctly pointed out, until
only a few years ago it was still prevalent op in ion that the
celebrated garden carpets o f sixth century Safavid Iran provided
the oldest extant evidence for the form o f Persian garden.
From the appearance o f such carpets it was already clear that
Safavid gardens included num erous water channels, m ultiple
parterres, and a centrally placed garden pavilion which often
stood within a rectangular p ool on the long axis o f the plan.
Such carpets o f Safavid date indicate, m oreover, that the design
o f even the most com plex gardens o f the day depended on the
repeated use o f a single, basic m otif: namely that o f the chahar
bagh or fou rfo ld garden.
Recent archaeological work in Iran has done much, however,
to reveal the far older beginnings o f Persian garden construc
tion, including the characteristic presence o f a chahar bagh m o tif
like that o f Sigiriya. Thanks to the excavations conducted by
David Stronach, at Pasargadae, the capital o f Cyrus the Great
( 559 “ 53 ° the founder o f the Achaem enid em pire, it is
now possible to detect the origins o f Persian garden design not
just five hundred years ago, but two thousand, five hundred
years ago.92 David Stom ach has studied in detail the question o f
p re-Islam ic water gardens in the N ear East specially at
Pasargadae,93 and as he had correctly emphasized that it is
clear, fo r exam ple, that the Sasanians who deliberately
borrow ed many separate concepts from the Achaem enids, were
A r c h a e o l o g ic a l Ev id e n c e o n Sh ipp in g C o m m u n itie s 113
m ore than w illing to construct elongated, axial palace gardens

that featured long channels or pools. Stom ach ( l 9 9 4 : 8—9)
further argued that Sasanians also chose to underline, as did
the Seljuks a long time after them, the close association o f
palace, pavilion and garden, and it may indeed be suggested
that the huge, elongated gardens o f the Sasanians would have
called for the introduction o f fourfold designs o f a new
complexity. T h e quadripartite plan o f C yrus’ Royal G arden at
Passargadae could be viewed as an expression o f the new
geom etrical articulations. T he apparent borrowings — such as
were presumably directly derived from the tradition o f the
Near Eastern gardens o f the late Sasanian / early Arab world —
include the symmetrical m anipulation o f water, the frequent
presence o f stone water channels and not least, the fourfold
division or Chahar Bagh o f the available garden space .94
I may emphasize that one should not underestimate the
contribution made by the Sinhalese artists and architects to
make Sigiriya an unique example o f this nature. However, this
vast com plex is characterized by many exceptional features: its
ramparts which were meant to provide protection from
intrud ers ;95 its fountains and m iniature water gardens with
pebbled or m arbled floors, covered by shallow, slowly moving
water creating the impression o f a m irro r ;96 pools symme
trically arranged in 'chahar b agh ;97 its sum m er palaces
surrounded by water transform ing them into well ventilated
and shady place where it was pleasant to abide in a land o f the
sun ;98 the facade o f the rock depicting the 'celestial m aidens’
rising from the clouds evocative o f lovers’ rendez-vous; and
finally the top o f the rock where was located the palace or
Alakam anada, the abode o f a god-king. Does not all this rem ind
us o f the atmosphere o f a paradise attested in the M esopota
mian foundations, in the popular literature o f medieval Iran 99
and in the gardens o f the great M oghuls?
E .H . Schafer (1963= 12) also gives a vivid picture o f maritime

activities from the seventh to the ninth century in the Indian
O cean, which was a safe and rich ocean, thronged with ships o f
every nationality: 'The Persian merchants came to C eylon also
ii4 O smund Bopearachchi
called 'L io n C o u n try’ and 'Island o f Rubies’ where they

purchased gems. We then hear that the Buddhist m onk
Vajrabodhi sailed from Ceylon in AD 7^7 with thirty-five
Persian ships and arrived at Palem bang.100 H. Hasan (1928:
104) correctly summed up the situation o f Persian and A rabian
navigation in the Indian O cean during the eighth century AD:
Firstly the Persians were by nature bent on com m erce’ — a
conclusion independently established by the evidence o f
Cosmas and Procopius. Secondly, that the Persians sailed to
C eylon in search o f precious stones — inform ation also
supplied ten years earlier, i.e. in 7I7 AD by the voyage o f
Vajrabodhi. Thirdly, that the purchases o f 'silk piece goods and
the like ware’ made the Persians sail straight to C anton — a
conclusion already reached by G. Ferrand on the basis o f
linguistic material. A n d fourthly and finally, that the Persians
were in the habit o f sailing in big craft both on the western and
southern sea — a fact which shows that Persian navigation was at
its height in 7^7 AD » and therefore that it must have
com m enced long p rio r to this date. For these reasons I am
inclined to believe that Persian navigation o f early M uham m a
dan times was merely a continuation o f Sasanian navigation and
that as G. Ferrand has said , the Persians were the initiators o f
the Arabs in trade in the Far East.
It would seem that with the decline o f the Sasanian Em pire,
the Muhammadan Arabs began to reach the Malabar coasts and
Sri Lanka. Besides a great num ber o f literary sources and
inscriptions, archaeological evidence, such as ceramics and
especially coins, bears witness to the presence o f Muslims on
the island as early as the seventh century. G old and silver coins
o f almost every M uslim dynasty o f Baghdad, Alexandria, N orth
A frica and N orth ern India o f various periods between the
eighth century and the fifteenth century have been found on
the island. From about the seventh century, Chinese traders,
along with the Arabs and South Indians, developed com m ercial
activities with Sri Lanka. It seems that the first Sri Lankan
contacts with C hina (as early as the second century) were purely
religious. By the eighth century, the first trade links began to
form . It was with the sudden burst o f trade activities between
A r c h a e o l o g i c a l E v id e n c e o n S h ip p in g C o m m u n i t i e s 115
China and M iddle East from the seventh century onwards that
Sri Lanka began to play a decisive role in the maritime trade
between east and west. The main reason for these active trade
relations between east and west was that, on the one hand, the
unification o f the A rabian countries under Islam put an end to
Sasanian power in AD 650 while on the other the Tang dynasty
established power in China.
Im ported Chinese ceramics constitute the most characteristic
sign o f the trade contacts that Sri Lanka had established with
China from the eighth century. T h e first transactions are
revealed by the presence o f three heavy Chinese storage-jar
fragments found in an archaeological context.101 T h e earliest
wares fro m M anthai are products o f the Tang dynasty
(618—907). It was only from the eleventh century that China
developed extensively its commercial activities with Sri Lanka.
The South Indian conquests o f the island, which brought the
existence o f Anuradhapura as the capital o f the island to a
conclusion, diverted the trade centres. By this time Gokanna on
the eastern coast and all the ports in the western wet zone became
more important. T he excavations conducted at the Alahana
Parivena monastery complex over the last ten years have exposed
for comparison, the largest num ber o f Chinese ceramic sherds
o f this period. Four complete vessels, a G reen-yellow glazed grey
ware jar, a white ware dish, a white ware small jar and an O live-
green decorated bowl were also discovered from the site.102
E .H . Schafer (19 6 3: 13) ° f ^ e opinion that in the ninth
and tenth centuries the Chinese ships did not reach the Persian
G ulf, and the largest ships engaged in the rich trade came from
Ceylon. Sri Lanka seems to have furnished the Chinese markets
with ivory, gems and cotton. A ccording to Hasan (1928: g)> the
m onk Vajrabodhi found thirty-five Persian vessels in a port o f
Ceylon early in the eighth century, there for the purpose o f
trading in gems. E .H . Schafer (1963) shows how the Chinese o f
T ’ang obtained ivory from their own province o f Lingnam and
from m ore remote sources like the L ion C ountry = Ceylon, cf.
T F Y K , 971, 17B, (Ts’e fu yuan kuei (1642 ed.) and fine cotton
from Ceylon, cf. TS, 2 2 l b , 4 I 55 b = T ’ang shu (KM ): T F Y K ,
971, 17b = (Ts’e fu yuan kuei (1642 ed.).
ii6 O sm und Bopearachchi
H undreds o f coins belonging to the Song and Southern

Song dynasties found at Anuradhapura and Polonnaruwa, and
especially in the short-lived capitals o f the thirteenth and
fourteenth centuries, such as Yaphuwa, K urunagala and
Dambadeniya and many other places in the country, attest to
com m ercial exchanges o f C hina with ancient Sri Lanka.103
T h e hoard from Yaphuwa, according to Francois T h ierry
(1998: 191), covers a long p erio d , from 976/984 to 1 2 & 0 /
1^64* T h e group o f 381 coins, as it stands now, comprises 266
coins o f the N orthern Song dynasty, 114 coins o f the Southern
Song dynasty and one Vietnamese copy. As he concludes: 'It is
not unusual to find that this single group o f coins covers such a
long p erio d , nor to find that it contains a Vietnamese copy, for
these, as well as Javanese imitations, circulated in significant
numbers throughout Southeast A sia’ .
T h e absence o f Chinese coins for the period p rior to the
tenth century and the abundance o f coins dating from the
Zhenzong period in the island can be explained by two factors.
Firstly in 1075 the prohibition on making payments for
imports with cash was can celled, and secondly from the time o f
em peror Shenzong (1078—1086) the annual p roduction o f
coins was raised to six m illions o f strings, each containing a
thousand coins. T h e discovery and exploitation o f mines in the
centre and south o f C hina made it possible to produce such
enorm ous quantities o f coins. These Chinese coins were found
everywhere in Asia and A frica along the maritime route,
especially in Japan, Vietnam , Java, Sumatra, India, Sri Lanka,
Zanzibar, M ozambique and the M iddle East. T he exploitation
o f copper mines was so intensive that the Southern Song
dynasty began to feel a noticeable shortage. In 1 2 19 it was
ordered to pay for imports only with silk and p orcelain .104 As
Joe C ribb (1996 A: 256) clearly explains: 'In the period after
the fall o f the Song dynasty coin production continued to fall
and it is clear that for long periods o f the Y u an dynasty ( a d
1^7^—1368) there was no production at all’ . O nly very few coins
are attested in Sri Lanka belonging to the period that follows.
T h e first voyages o f Portuguese and Spaniards in the Indian
O cean and the dim inution o f trade activities o f C hina with Sri
A r c h a e o l o g i c a l E v id e n c e o n S h ip p in g C o m m u n i t i e s 117
Lanka take place almost at the same period. T h e last delegation

from Sri Lanka was sent to C hina in 1459 -105
In conclusion, it should be made clear that the results we
have so far obtained from recent excavations at A nurdhapura,
Sigiriya and Polonnaruwa and from our investigations on coins,
beads, intaglios, seals and ceramics and from the excavations at
Ridiyagama and Kelaniya oblige us to re-exam ine the question
o f the m aritime trade and comm unities o f the proto-h istoric
and early historic Sri Lanka. They make questionable most o f
the hypotheses, based solely on literary evidence, put forward by
some em inent historians on the organisation o f trade activities
in the island. Further investigations will enable us to test the
validity o f own hypotheses. Let this be a prologue to a larger
comprehensive study which will be based on all available
archaeological material and literary sources and on the results
that we hope to obtain from our own investigations o f all the
estuaries and o f river banks on the south and West coasts and, if
possible, also the east coast. T h e anticipated results should
certainly enlighten us further on both the international and
inland trade patterns o f ancient Sri Lanka.
Notes
1 D .P .M . Weerakkody refers to C osm as In dicopleustes, Topographia Christiana,
X I, 16: . . . auirj ouv r\ L ieê 5i(3a îsar) 7icoc; xuyxavoixra xf|c; ’Iv 5iKfj<;, e%ovoa
5s Kai xov 6&kiv0ov, e£, otaov xwv sîTiopicov 5ex£xai Kai oXok; (asxaPaM ei, Kai
fieya qiTiopiov xuyxavsi.
'So , this Sieledib a, situated as it were in the m iddle o f In dia, and with
the hyacinth p ro d u cin g country too, receives wares fro m all the m arts
and exports to them all, and it is a great centre o f trad e ’ . (D .P .M .
Weerakkody, 1997: 2 J 8 and 245)-
2 R egardin g the com m ercial and political relation sh ips b etien S ri Lanka
an d In dia, see K . In drapala, 199O; R A .L .H . G unaw ardena, 199O; S.
Pathm anathan, 197^, I 9 &4 > I 99 °> J 994 ; N . W ijesekara, 1952. T h e m ost
recent and com plete study on all the G reek and R om an relation s with S ri
Lanka, see D .P .M . Weerakkody (1996).
3 O . B opearachch i, 1 9 9 0 , 1992, 1993 A, 1995 A, 1995 B and 1996.
4 H is accounts are known through Strabo, Geography, XV , I, 15 an d Pliny,
Natural History, V I, 81.
5 K now n to us through Pliny, Natural History, V I, 81.
n8 O sm und Bopearachchi
6 Q u o ted by Strabo, Geography, XV, I, 14 an d Pliny, Natural Histoiy, V I, 81.

7 Geograph)!, I, 4, 2 , II, I, 14, II, I, 17, II, 5, 14, 32 & 35, XV , I, 1 4 - 5 .
8 Natural History, V I, 79“ 9 1-
9 Geographia, I, 14, 8 - 9 , V II, 4, I- 1 3 , V II, 5, I I , V III, I, 3, V III, 2 8 .
10 Periegesis, 59I—^ ° 5 -
11 Commonitorium Palladii, I —II.
12 Christian Topography, II,45“ 6; HI, 65, X I, 13—19, X I, 2 2 .
13 A total o f 149 coins Ire recovered fro m the trench A n u rad h ap u ra Salgha
Watta 2 (ASW 2), o f which 97 are iden tifiab le with certainty, a fu rth er
2 2 were iden tified with som e reservations an d the fin al 3 0 are
com pletely u n recogn isable. T h e oldest object which m ight rep resen t a
coin, a thin, extrem ely w orn cop p er-allo y square, was fo u n d in
structural phase 15 and dates to the m iddle o f the th ird century B C ,
whilst the youngest, a one cent coin o f the k in g -e m p ero r G eorge V I, was
m in ted in 1943 A D .
14 H . Ratnayake, 199° - T h e Jetavan aram a m onastery, with its Stupa known
u n d e r the same nam e — the largest o f all the B uddhist m on um en ts o f this
type in the w orld — was built by king M ahasen at the close o f the third
century o r the b egin n in g o f the fou rth century A D .
15 S. Bandaranayake, 1984* T h e fo rtifie d royal residence o f Sigiriya, as it is
well known, was b u ilt by king K asyapa ( a d 477“ 495 )> w^ 0 having killed
his father, took refuge on this natural rocky peak converting it into the
m ost sen sation al site in S ri Lanka.
16 P .L . Prem atilaka, 1 9 9 0 . Polonnaruw a, first m en tion ed in the reign o f
A ggabodhi III (624?— 6 4 0 ), becam e the capital o f the Sin halese kings in
the eleventh century.
17 T h e earliest excavations at M anthai date back to 1887. Since then, various
attem pts have been m ade fro m tim e to tim e; the m ost recent exploration s
were m ade by J . Carswell and many other S ri Lankan and foreign
archaeologists, fro m 1 9 8 0 . T h e excavations abruptly came to an end in
1 984 fo r political reason s; fo r b rie f notices o f these recent excavations,
see J . Carswell (1 9 9 0 , 1991), M. Prickett (1 9 8 0 ), and M. Prickett-
Fernan do ( l 99 ° B ). Very lim ited exploratory excavations were carried out
at K an taro d a i in the Ja ffn a pen in su la, first by P .E . P ieris ( 19 1 7 , I 9 J 9)
and later by V . Begley (1967). T h e m aterial fo u n d in both excavations is
still awaiting pu b licatio n .
18 I am m ost grateful to D r. W. W ijeyapala, D irector o f Excavations o f the
A rch aeological departm en t and the co -d ire c to r o f French M ission o f
A rchaeological C o -o p e ra tio n in S ri Lanka fo r his u n failin g help and
c ollab oratio n .
19 O . Bopearachch i, 1995 B & !996-
2 0 In his excellent study, J . D eloche (1985) has discussed in detail the role
o f ancient ports, situated beside rivers o r lagoon s on the coast o f
T am iln adu. See also R. Nagaswamy (1991).
A r c h a e o l o g ic a l E v id e n c e on Sh ipp in g C o m m u n itie s 119
21 T h anks to the generosity o f Tam il colleagues fo r whom I h old a great

esteem , I was enabled to visit al these very im portan t sites, and exam ine
archaeological m aterial den otin g m aritim e trade. I wish to thank very
particularly D r. Iravatham M ahadevan, Prof. K . R ajan, D r. S. Suresh ,
M r. R. K rish n am urth y, Prof. Y. Subbarayalu, D r. N atana K asin ath an
and Prof. S. G urum urth y fo r their un failin g help.
22 R .L . B ro h ie r (193 5 : assum es that the section o f the Mahawali G anga
betw een the islan d o f K alin g a in Polon n aruw a an d the sea o ff
T rin co m alee was o f sufficient depth at all tim es to be navigable fo r
sm all vessels. T h ou g h G okann a becam e a flo u rish in g h arb o u r fro m the
eleventh century, when the royal capital was tran sferred fro m A n u r
ad h apura to Polonnaruw a, there is sufficient evidence to show that it was
known as early as the fifth century AD to m erchants who reached the
island fro m the east. T h e Culavamsa (X L I, 7 0 —8 0 ) m en tion s G okan na at
the m outh o f the Mahawali G anga, at the Bay o f T rin co m alee, d u rin g the
reign o f K itsirim egh a ( a d 555“ 573) T ie (K ittisirim egh a) took him with
him and came in a m om ent to the G okan n a sea’ .
23 I carried out a series o f exploration s in the area arou n d G iribaw a. T h e
h istorical significance o f this site lies in the fact that it is situated at the
left bank o f the K ala Oya which flows to the sea at U ruvelapttna. T h e site
itself is not far away fro m the ancient b ridge built by Parakram bahu I.
T h e Culavamsa (L X X , 123—! 3 ° ) describes it as a solid b ridge passable by
files o f elephants, h orses and chariots, held together with iro n ban ds and
nails, m ade o f beam s o f tim ber and twenty cubits b ro ad . Even today the
ru in s o f this bridge can be seen. T h e archaeological site extends fro m the
bridge to a distance o f five kilom etres. I systematically ph otograph ed all
the ceram ics, iro n objects and particularly glass beads fo u n d by the
villages when p lou gh in g the fields o r digging holes in their gardens. T h e
m ost fascinatin g experience was the discovery o f glass furn aces, thousands
o f glazed tiles, glass beads and h un dreds o f glass fragm en ts and slag at the
site called Pabalugala.
24* At a distance o f seven kilom etres fro m the river m outh, at the village o f
N ariyagam a, we came across ancient fou n dation s on the left bank o f
D ad u ru Oya. T h e excavations conducted in sum m er 1998 by the
A rch aeological D ep artm en t an d the French M ission revealed the
existence o f settlem ents which may date back at least to the fourth
century AD as revealed by the late Black and R ed Ware fo u n d on the lowest
level.
25 T h e D ep artm en t o f A rchaeology an d the French M ission o f A rc h a e
o logical C o - o p e r a tio n in S ri Lanka jo in tly carried out three son dages
in su m m er 1997* the village o f Pilapitiya close to K elan iy a, ab out
seven k ilom etres fro m the estuary o f the river K e lan i, in o rd e r to clarify
the stratigraph y an d to d eterm in e the d iffe ren t ch ro n o lo g ica l phases o f
the early h isto ric settlem en ts o f the site. T h e m ost an cien t layers
b ro u g h t to light fragm en ts o f R ou letted Ware an d early Black an d R ed

Ware which can be dated with certainty back to the 4 an d 3 cen tu ries BC.
T h is ch ron ology is now co n firm e d by the calib rated d atin g o b tain ed by
C 14 analyses. T h u s these fin d s co n firm the w ritten H isto ry o f K elan iy a
which goes back at least to the th ird century BC. Even today rafts can go
up to Sitawaka, ab ou t forty km fro m the river m outh , where one o f the
sh o rt-liv ed capitals o f the sixteenth century was situated. H .W . Cave
(1 9 0 8 ) has d escrib ed how, even at the b e g in n in g o f the twentieth
century, sailin g craft c arrie d chests o f low -grow n teas 6 0 m iles down
this river to the sea and thence to the h arb o u r an d the w areh ouses o f
C o lo m b o .
26 D u rin g the exploration s carrie d -o u t in 1995 we came across a large
n u m b er o f fragm en ts o f ancient ceram ics in the vicinity o f D iyagam a on
the bank o f the K alu G anga, where the in scrip tion re fe rrin g to the
com m ercial centre known as K alyani was fo u n d . T h is in scrip tion refers
to the taxes paid by a villager to the m onastery o f K alaki in the
com m ercial centre known as K alyani. It is also noteworthy that the K alu
G anga, one o f the five m ajo r rivers o f Sri Lanka, flows through the
m iddle o f the ge m -b e arin g Sabaragam uw a Province, which has R atn apu ra
as its m ain centre. G em ston es, especially the pale blue sapphires and
rubies fo r which Sri Lanka had an ou tstand in g rep utation in antiquity,
were an im portan t part o f the eastern sea trade du rin g the R om an and
Byzantine perio d s.
27 T h e positive results o f the exploration s that we carried out in 1993 at
Godavaya, on the estuary o f the Walawe G anga, and along the river up to
Ridiyagm a, en couraged us to undertake a systematic study o f the
organ isation o f m aritim e and in lan d trade. O u r starting poin t was the
in scrip tio n o f G ajabah u (114—136 a d ) on the rock face beside a B ud dh ist
tem ple. T h is says that the king d on ated the custom s duties ob tain ed fro m
the port o f G odapavata to the V ihara at the site. From 1993 to 1996 I
carried out systematic exploration s an d excavations at Ridiyagam a, about
twelve kilom etres upstream fro m the m outh o f the River W alal. T h e
ancient settlem ents, clustered ro u n d the two ancient reservoirs, are today
com pletely u n d er water as a result o f an extension built by the B ritish at
the b egin n in g o f the century. In 1995’ the excavation brou gh t to light six
stratigraphic layers, subdivided into 45 d ifferen t contexts d eterm in ed by
sign ificant features. D ifferen t phases o f occupation go down to a depth o f
1.35m . Sh erds o f Black and Red Ware were n u m ero u s in the th ird layer.
T h e fourth layer yielded cop p er slags, m ica, b u rn t charcoal, early Black
an d R ed Ware, terracotta ob jects and b eads. T h is layer is thus
characterised by extensive d ifferen t h um an activities. T h e m ost ancient
layer b rough t to light a sm all n u m b er o f potsherds and copp er slag, thus
indicating the b egin n in g o f h um an settlem ent. T h e approxim ate dating
o f the differen t phases o f settlem ent from the fourth century BC to
A r c h a e o l o g ica l Ev id en ce o n Sh ipp in g C o m m u n itie s 121
seventh century AD is supp orted by R om an coins, Black and R ed Ware

and other ceram ics, copp er slag follow ed by iro n slags an d beads. In
1996, with the p articip atio n o f nine French , Italian an d D utch
archaeologists, m ore than twenty furn ace structures Ire excavated. Som e
o f these fu rn aces may have b een used fo r fo rg in g iro n or fo r
cem en tation .
28 See S. D araniyagala, 1972, 1986, 1992 fo r excavations at G edige,
A n u rd h ap u ra; V . Begley, 1970 fo r P om p arip p u ; V . Begley, 1967 fo r
K an taro d a i; an d H .- J. Iisshaar, 199^» f ° r Ibbankatuwa.
29 T h e excavations con ducted by Sri L ankan and G erm an archaeologists
u n d er H .- J. Iisshaar & W. W ijeyapala ( l 993 ) at A k urugoda (T issam a-
h aram a), yielded not only early h istoric Black and R ed Ware, but also
several fragm e n tso f R ouletted Ware. As a result o f clandestine diggings at
A kurugoda an d M in ihagodana in T issam ah aram a u n preceden ted n u m
b er o f engraved disks m ade o f R ouletted Ware were fou n d .
30 See R. Nagaswamy, 1991 and N . K asin ath an , 1992.
31 See S. G urum urth y, 1981: 3 ^ 0 .
32 R. Nagaswamy, 1991.
33 A ccordin g to P. Francis, (1987= 2 9 ) : 'W heeler thought that A rik au sd u ’ s
prim ary export was colou red cloth, fo r which In dia has lo n g b een
fam ou s. But ju d g in g fro m the scraps that rem ain, the b ead industry was
at least as im portan t as the cloth industry. M oreover, cloth was a prod u ct
o f many In d ian cities at that tim e, while bread m ak in g was very special.
S p e cial? Yes, in deed, the bead industries o f A rikam edu can even be
con sidered u n iqu e. N o other city in w orld H istory was both m an u fac
turer and m arketer o f both glass an d sem i-p recio u s stone beads. T h e
other great b ead m aking cities o f the w orld — A lexandria, Cam bay and
Venice fo r exam ple — have b een at som e tim e both m arketers and
m anu facturers o f their products, but they each specialized in one type o f
b ead m aterial. A rikam edu was d ifferen t. N early every sort o f bead m ade
anywhere in the w orld could be bough t there, either because it was m ade
locally fo r export o r because it had b een im p o rted fro m distant lands.
A rikam edu rightly deserves to be called the greatest o f Bead E m p o ria ’ .
34 A p art fro m the epigraph ic and literary evidence, the com m ercial
activities o f these South In dian com m un ities are known to us through
their coins fo u n d especially at A n u rad h apu ra, Polonnaruw a and other
ancient sites o f the islan d , see O . Bopearachch i, 1993 A.
35 For fu rth er details see K . In drapala, (1 9 9 0 ), S. Pathm anathan, (1976,
1 9 8 4 ,1 9 9 0 ,1 9 9 4 ) .
36 A ccordin g to S. Paranavitana (197O: lxxxix-xc): 'It is easy to recognise in
'D am e d a’ the prototype o f 'D em e la’ or D em ala’ , the designation in
Sinhalese literatu re as well as in the curren t speech o f the Tam il peop le
who inhabit the extrem e south o f the In dian Peninsula. Its affinity to
Pali, 'D am ila, Skt. D ram id a’ o r D ravida’ and Tam il, which are the
122 O sm und B opearach ch i
nam es, in the respective languages, o f the Tam il peop le, is also not
difficult to reco g n ise’ .
37 See S. Paranavitana, 1970, no. 9 4 :
38 See S. Paranavitana, 197° ’ n o - 35^ : 'T h e cave o f the h ouseh old er V isakha’
(Dameda-vanijha ga[pa]ti-Visakaha line); n o. 357 : 'T h e cave o f the h ou seh old er
Visakha, the Tam il m erch an t’ (Dameda-vanijha gapati-Visakaha line).
39 S. Paranavitana, 197° : n o - 4 ^ 0 : T h e cave o f the m erchants who are the
citizens o f D ighavapi, o f the sons o f . . . and o f the wife T issa, the T a m il’ .
4 0 Two other in scribed objects, a seal and a sealing were rep orted fro m
Bengal (see I. M ahadevan, 1996).
41 See R. K rish nam urth y, 1997: 34 ’ pi* 3 ’ n o - 29
42 T h ree types on the obverse: elephant, tem ple and and on the reverse,
fish sym bol, depicted on this coin are sim ilar to the ones on the coin
pub lish ed by M. M itchiner (1978: 6 2 9 ). He correctly dates this series c.
210—177 BC arguin g 'T h e earliest coinage o f Ceylon shows many parallels
with that o f the Pandyas, by which it was in sp ire d ’ . He furth er argued
'T h e in itial Pandyan issues have b een divided into two consecutive series
o f m ulti-type coins (c. 24 ° ~ ' 2 I 0 Be) that preceded the Pandyan
cam paign in C eylon d u rin g the second century BC. T h e earliest coin s in
Ceylon b ear designs derived fro m the secon d series o f Pandyan m u lti
type coins struck du rin g the p e rio d circa 210—175 BG and b ear a g ro u p o f
sym bols on the obverse am on g which an elephant norm ally fig u re s’ . It is
in teresting to note that this coin was fo u n d in structural p e rio d G that
dates to the secon d century BC. T h e Pandyan fish sym bol is also borrow ed
and appears on the reverse o f these earliest Sin halese issues. T h e Caitya o f
three cells su rm ou n ted by a chatra: & depicted on this coin is probably
an ad dition o f the Sinhalese coin engraver.
43 Nikaya Sangrahawa, p. 2 6 . T h e m ain pu rp ose o f this occupation, as B .J.
Perera (l9 5 2 : 198) points out, was 'to prevent Sinhalese traders from
dealin g directly with the A rabs and C h in e se ’ .
4 4 S. Pathm anathan (1984, 199° ) - A ccordin g to S. Pathm anathan (1984:
126) it appears to be a m in iatu re rep resen tation o f 'the holy h ill’ , the
figure o f which is said to have b een painted on the flag o f the N an adesis.
T h e p rin cipal figure depicted on this seal is a fo u r-a rm e d fem ale deity in
a standing (sthanka) p osture. S. Pathm anathan ( l 99 0 : I 4 I _ 2) correctly
iden tifies the figure as D urga, the deity whom the N anadesis held in deep
veneration.
45 H is accounts are known through Strabo, Geography, XV, I, 15 and Pliny,
Natural History, V I, 81. For the m ost recent and com plete study o f all the
classical texts on Sri Lanka, see D .P .M . Weerakkody (1997)-
4 6 S. Paranavitana, 197O: 8, no. 105.
47 R .A .L .H . G unaw ardana, 199O: 2 6 .
4 8 R egardin g the chronology o f M en and er see O . B opearachch i & W.
Pieper, 1998: 199—204-
49 and fro m A lasanda the city o f the Yonas (yonanagaraalasanda), came the
thera Y on am ahadam m arakkhita with thirty thousan d bhikkhus’ , Maha-
vamsa (X X IX , 39 )* T h e nam e o f the B uddhist m onk and the n u m b er o f
the delegation is o f course subject to the usual exaggeration, but one
cannot ign ore the fact that there was certain knowledge about the
im portan t B uddhist com m unity in the G reek territo ries at that tim e.
50 C osm as In dicopleustes: T o pograp h ie chretienne, trans. by Wolska-
C on u s, 3 vols., Paris, 1973*
51 J.W . M cC rin d le, 19O I: 1 6 0 - I .
52 S .U . D eraniyagala, 1992:
53 'Two kahapanas o f U ttara the officer in charge o f canals, . . . o f the lapidary
Phussa . . . a kahapana o f . . . (Adikaya-Utaraha duve kahapane manikara-pusa
. . . a .. . ciliya kasapane) (S. Paranavitana, 1970, n o. 791)-
54 T h e in scrip tio n o f Kaduruvava (S. Paranavitana, 197°> n o - 1205),
written in late B rah am i script, refers to a dealer in coined m oney
(R u a p a-v a p ara): 'H ail! T h e cave o f G op a, the G ran d so n o f the ch ief
V ahidi, the M inister, and eldest son o f H am ika, the cham berlain and
dealer in coin ed m oney, is dedicated to the San g h a’ . (i) Sidha [/* ]
Parumaka Vahidi = mataha marumakanake (2 ) dorakani-rupa-vapara-Hamikaha (3 )
jeta-puta G[o]paha lene saga-niyate.
55 See H.W . C o d rin g to n , 1924: 16, 191—93 *
56 As C o d rin g to n (1 9 2 4 : ^ I 9 I—93 ) correctly suggested, the absence on
these coins o f any symbol which can be attributed to S ri Lanka alone
leads us to assum e that all the genuine p u n ch -m arked coins fou n d in the
island were im p orted from In dia. He also correctly observed that they
were in circulation un til the end o f the third century A D . T h is is now
proved by the discovery o f many terracotta m oulds with Karshapana
im prin ts in the excavations at G edige (A n u rad h apu ra), T issam ah aram a
and many other places, See S .U . D eraniyagala, 1972: 5^-
57 T h is u n pu b lish ed m oneybox is now in the private collection M r. R.
W ickrem esinhe. H e also has two other terracotta m oneyboxes fu n d also at
T issam ah aram a, con tain in g R om an T h ird Brass.
58 See O . Bopearachch i, 1995 A: 128.
59 T h e coin fo u n d at Puttalam is in the private collection o f M r. F. M edis
and the one from T issam aharam a in the private collection M r. R.
W ickrem esinhe. Th ese two coins b elon g to the grou p 7 o f posth u m ous
im itations o f H erm aios that I have placed chronologically in the second
decade o f the first century AD, see O . Bopearachch i, 1997: I 9 I » 2 0 7 .
60 See O . Bopearachch i, 1995 A: 129-
61 See N . Sim s-W illiam s and J . C rib b , 1996; Fr. G ren et and O.
Bopearachch i, 199^-
62 Obv. K in g standing facing, h oldin g triden t in raised 1. h and and with his
r. han d low ered to an alter / Rev. E n th ron ed goddess facing. Tam ga on 1.
and nam e on r.: A rdokhsh o, see O . Bopearachch i, 1995 ! 29 *
63 See O . B opearachch i, 1993 A: 7 0 .

64 T h ese coins are in the private collection o f M r. R. W ickrem esinhe.
65 O n the techniques o f p ro d u c tio n o f carn elian, see M .L . Inizan, 1991.
66 'In land from this place and to the east, is the city called O zene, form erly a
royal capital; from this place are brough t down all things needed fo r the
welfare o f the country about Barygaza, and many things fo r ou trade: agate
and carnelian, Indian m uslin and mallow cloth, and m uch ordinary cloth’
( Periplus, 4 8 ); T h ere are exported from these places spikenard, costus,
ivory, agate and carnelian, lycium, cotton cloth o f all kinds, silk cloth,
mallow cloth, yarn, lon g pepp er and such other things as are brought here
from the various m arket-tow ns’ (Periplus, 4 9 ); 'There are brough t down to
Barygaza from these places by wagons and through great tracts without
roads, from Paethana carnelian in great quantity . . . ’ ( Periplus, 5l)-
67 'T h e cave o f the ch ief M ala, the m arin er sailin g to Bharukaccha [is given ]
to the San gh a’ ( Parumaka-Baruka [ca\ga-Malaha lene sagaha) . ’ S. Paranavitana,
1970, n o. 1183.
6 8 H .- J . W eisshaar, 1992.
69 T h is intaglio is now in the collection o f M r. D . Sam arasingh e and I am
thankful to him fo r au th orisin g me to publish it.
70 H . Ratnayake, 199O: 45, 4 9 .
71 H . Ratnayake, 199^ : 5^-
72 O n the other han d there are exported costus, b delliu m , Lycium , n ard ,
tu rqu oise, lapis lazuli, seric skins, cotton cloth, silk yarn, and in d ig o ’
( Periplus, 3 9 ). W .H . Sch off, (1912= 170) discussin g the w ord used in the
Periplus, argued: 'T h e w ord in the text is sappheiros, an d a natural
in clin ation w ould be to assum e this to be the same as sapphire, which is
also a prod u ct o f In dia; but accordin g to Pliny (X X XV II, 39) the stone
known to the R om an s as sapphire was an opaque blue stone with go lden
spots, which came fro m M edia, that is in a gen eral way, fro m the country
I call Persia. It was not suited fo r engraving because it was in tersected with
h ard crystalline particles. T h is can be n oth in g but lapis lazuli, which has
b een in dem and from a very early tim e fo r orn am en t and also as a
pigm ent, ultram arin e, which was so extensively used by the Egyptians in
their public b u ild in g s’ .
73 See R. Gyselen, 1993: pi. X X I, 3 0 .H .4 .
74 S. Paranavitana, 197O: no. 6 2 2 : ' Gamika-Kabojhaha lene’ The cave o f the
village-councillor K am b o ja; no. 623 ’- 'Gamika-Siva-putra gamika-Kabojhaha lene’
T h e cave o f the village-councillor K am bo ja, son o f the village-councillor
Siva’ (no. 625) t o Gamika-Siva-putra gamika-Kambojhaha jhitaya upasika-Sumanaya
lene. (2) gamika Kabojhaha ca sava-satasajesamage pat\ T h e cave o f the fem ale lay-
devotee Sum ana, daughter o f the village-councillor K am boja, son o f the
village-councillor Siva. May there be the attainm ent o f the Path o f Beatitude
for the village-councillor K am bo ja and for all beings.
75 J . Bloch, (195O: 103, 130),
76 S. Paranavitana, 1970, no. 553 : 'Kabojhiya-mahapugiyana Manapadasane agata-

anagat-catu-disa-sagasa [T he cave] M anapadassana o f the m em b ers o f the
G reat C o rp o ra tio n o f K am bojiyas, [is given] to the Saifigha o f the fo u r
quarters, presen t an d absent.
77 Paranavitana, 1970* no. 9 9 0 : ' Gota-Kabojhi[ya]na parumaka-Gopalaha bariya
upasika-Citaya lene sagasa Th e cave o f the fem ale lay-devotee Citta, wife o f
G opala, the ch ief o f the in corporated K am bojiyas, [is given] to the Saifigha.
78 Sihalavatthu, ed. by B ud dh adatta M ahath era, p. 10, q u o te d by S.
Paranavitane, 197° : xci-
79 P. B e rn a rd , 1985: 2 5 ~ 6 .
80 S. Paranavitana, 1970> n ° . 96.
81 J . Bouzek an d S .U . D eraniyagala, 1985*
82 H . Ratnayaka, 199O: 45 an ^ M. P rickett-Fernand o, 1990 A : 63.
83 (ii) In T ap rob an e, an island in in n er In dia, where the In d ian Sea is,
there is also a C h ristian church there, and clergy an d faithful, but I do
n ot know w hether there are any furth er on . Sim ilarly, in the place called
M ale, where p ep p e r grows, and in the place called K allian a, there is also
a b ish op, o rd ain ed in Persia. Christian Topography, III, 65, see D .P .M .
Weerakkody, 1997: 244 *
84 C o d rin g to n , I 924 : 3 ° -
85 O . B opearachch i, 1995 I 35 *
86 See J . C rib b , 1 9 9 0 , p. 171 and pi. IV, n o. 37.
87 See R. Gyselen, 1993: pi. X X X , 3 0 .T .I —3 0 .T .1 7 . T h is u n pub lish ed
intaglio is in the private collection o f M r. R. W ickrem esinhe.
88 See O . B opearachch i, 1993 B.
89 See S. Bandaranayake, 1984= 4
9 0 A ccordin g to the Cuhvamsa (X X X IX , 3—5): 'B ut as he was unable (through
these) to slay his b roth er, he b etook h im self through fear to Sih agiri
which is difficult o f ascent fo r h um an b eings. H e cleared (the lan d)
ro u n d about, su rro u n d e d it with a wall and built a staircase in the fo rm
o f a lion . T h en ce it took its nam e (S ih a g ir i).’ T h e chronicle fu rth er says:
'H e collected treasures and kept them there well protected and fo r the
(riches) kept by him he set guards at differen t places. T h e n he b uilt there
a fine palace, worthy to b e h o ld , like an oth er Alakamanada and dwelt there
like (the god) Kuvera ’ .
91 S. Paranavitana, 195O: 1 2 9 - 8 3 .
92 D . Stron ach , 1989, 1990 and 1994.
93 Ibid.
94 D . Stron ach , 1994: 8 - 9 .
95 It is very difficult to believe that the ram parts and m oats o f Sigiriya were
built only fo r p rotectio n . In many places the continuity o f the m oat is
distu rb ed by the natural rocks. N o effort was m ade to cut the rock to
convert it into a ditch. I f the m oat was built only fo r p ro tectio n then it
should have b een cut right through.
126 O sm und B o pearach ch i
96 T h e excavations carried out in m id 1 9 8 0 ’ s exposed a garden unit o f a

hitherto unknow n type that the Sri Lankan archaeologists nam ed as the
m in iatu re water gardens. T h e excavations o f this gardens has revealed an
elaborate network o f structures: pavilions, p o o ls, cisterns, courtyards,
con duits and w ater-courses. T h ere are at least five distinct units in the
garden , all com b in in g b u ild in gs and pavilions o f brick and lim eston e
with paved, water retainin g structures and w inding water courses. T h e
flo o r o f the p o o l is finely paved with large, quartz pebbles. A striking
feature o f this m in iatu re garden is the use o f these w ater-su rrou n ds with
p eb b led o r m arb led flo o rs, covered by shallow, slowly m oving water. See
S. Bandaranayake, 1984.
97 Passing the m in iatu re water garden s, one enters into the zone o f p leasure
gardens characterised by central islan d su rro u n d e d by L shaped fo u r
p o o ls, known accordin g to the ancient garden concept as quartered or
'chahar bagh. T h e island there was originally a pavilion, the rem ains o f
which consist o f flights o f steps on the fo u r sides. T h e causeways,
am bulatory and the sides o f the islan d are faced with brick, and flights o f
lim eston e steps go down fro m the causeways into the tank.
98 O n either side o f the foun tain garden , there are two sum m er palaces.
They are su rro u n d e d by high rubble walls and wide m oats. T h e access to
the sum m er palace was m ade possible by b ridges cut into the surface rock.
O n the flattened surface o f the islan d , on the partly built up b edrock,
were situated the water pavilions. T h e pavilions, as in dicated by the holes
still visible on the surface, were m ade out o f wood.
99 A ccord in g to the con cept evoked in the episode o f the S am a k-e-’A yyar’
(o f the sixth century m an u scrip t o f Iran , see M. G a illa rd , 1987) the
royal garden is in fact com pletely p rotected and totally isolated fro m
the out side w orld by high walls. T h e garden s inside the en closure
had a privacy strictly lim ited to the king and the n ob les. T h e
presen ce o f artificial fo u n tain garden s, su m m er palaces, an d their
close association with water evoke the Paradise like atm osph ere. T h e
king thinks that he is sh ielded fro m in tru d e rs if he is su rro u n d e d by
m aidservants enjoyin g his in n ocen t and lawful p leasures (M . G aillard
( 1991 : 21- 3 ).
IOO A ccordin g to H . H asan, (1928: Il8 ) 'Now accordin g to the Tchen Yuan sin
ting che kiao mou lou, com piled by Y iian -tch ao about the ninth century AD ,
V ajrabo di sailed fro m C eylon in 717 AD with thirty-five Persian ships
and arrived at Palem bang in a m o n th ’ . 'V ajrabodhi arrived at the islan d
o f Ceylon . . . Th irty-five Persian ships were fou n d there, com e to trade
in preciou s stones. As soon as the Persian, m erchants saw V ajrabodh i
they follow ed him with one accord. A fter a m o n th ’ s stay in Ceylon,
V ajrabodh i ob tain ed royal p e rm issio n to depart and sailed with the
faith ful Persian m erchants. A m on th sailin g b rough t them to Fo-ch e (=
C h e -li-fo -c h e ) o r Palem bang. T h e end o f the voyage was disastrous; all
A r c h a e o l o g ic a l Ev id e n c e o n Sh ipp in g C o m m u n itie s 127
the ships o f the m erchants were scattered by the tem pest, and only the
ship in which V ajrabodhi was sailing reached p o r t’ .
101 See M. P rickett-Fern ando, 1990 A: 7 0 .
102 P .L . Prem atilake, 199O: 2 3 5
103 See Fr. T h ierry, 1995 f ° r an exhaustive catalogue o f C h in ese coins
fo u n d at Yaphuwa.
104 A . Sch roed er, 1905= 373“ 5 - O n the d istrib u tion o f the C h in ese coins
in South A sian countries, see Fr. Th ierry, 1992 a n d j . C rib b , 1996 A.
105 P. Pelliot, 1 9 0 4 : 3 5 6 - 7 .
5
P re- M odern
S ri L ankan S hips
Somasiri Devendra
T here are many references to Sri Lankan ships in the historical

records o f Sri Lanka, as well as other countries. Yet, we have
little idea o f the appearance or structural characteristics o f the
early vessels. This paper, which tries to fin d an answer to these
questions, is presented in two parts. Part I states the hypothesis
and the path followed to test it. Part 2 describes the traditional
ships that survived into this century. T h e inland watercraft,
which are im portant for a fuller appreciation o f this subject,
are not dealt with here as I have dealt with them at length
elsewhere. (Devendra 1995 : 211—238)
Part I
H y p o t h e sis
A ll Sri Lankan ships and watercraft developed from two basic

form s that evolved out o f the interaction between the inshore
maritime environm ent and the biological resources o f the
island. Shared cultural links with, and technological form s
prevalent in, south India were the other parameters. W hen Sri
Lankan vessels eventually ventured farther out into the ocean,
these basic form s underwent further and greater m odification
to fit the new environm ent. Contacts with foreign ships calling
Pre- M o d er n S ri L a n k a n S h ips 129
at Sri Lanka and experiences gained by sailing in foreign

waters, exposed Sri Lankan mariners to types o f craft and
technologies that had originated in different parts o f the
Indian O cean (and beyond). Many useful and appropriate
features thus encountered were adopted and superimposed on
the original form s, which still retained their identities.
However, due to an insufficiency o f data, the sequence o f
evolution and the m orphology o f the most advanced ocean
going vessels have yet to be determ ined and a m ethodology
needs to be developed.
M eth odology
Th e first requirem ent was to gather all available references to

details o f construction. In many literary and historical works
Sri Lankan ships are described but, for the purposes o f this
study, these could be accepted only if they were specific and
credible. Some references (notably those o f O nesicritus and
Pliny) met these requirem ents. For example, Strabo quotes the
form er observing that the ships seen by him sailed badly 'since
their sailing gear is inefficient and they are built without belly
bolts on either side’ ; and the latter says 'The sea . . . is shallow,
not m ore than six yards deep but in certain channels so deep
that no anchors touch the bottom . For this reason ships have
prows at either end so they do not need to turn about in the
narrows o f the channel’ (Weerakkody 19 9 7: 225)- But even
these had to be measured against both the hard evidence
available (however meager), and an objective fram ework o f
reference, before being accepted. T he preferred hard evidence
had to be visual, graphic and three-dim ensional i.e. paintings
or carvings by com petent and knowledgeable artists; models
(preferably those made by seam en); technical drawings;
drawings, by artists with an eye for detail which followed
accepted conventions o f technical drawing, or perspective; and
early photographs. These were scarce in respect o f ancient and
medieval craft, but there was a surprising amount o f material
concerning the last two hundred years. Should a study o f this
material reveal some com m on and technologically feasible
130 SOMASIRI D E VE ND R A
characteristics, they could be tested against m ore recently

form ulated models o f the evolution o f watercraft and analysis
by structural characteristics. If the test proved positive, it could
be assumed that the characteristics isolated were the product o f
a specific line o f developm ent. Assum ing that the line o f
developm ent had been linear, the same characteristics could be
expected to have been present in earlier craft. A t this point, two
alternatives needed to considered.
i) If the craft studied were already technically advanced, their

predecessors would have been less so, and the characteristics
would appear in a less developed form .
2,) If there had been a lowering o f standards and sophistication
in technology and design, evidence would emerge o f
deterioration from a 'h igh ’ to a 'fo lk ’ technology.
Th e m ethodology adopted, therefore, was to (a) search for

com m on characteristics, (b) analyze them, and (c) project the
search backward in time from the known to the unknown. T h e
results expected were to:
1) form ulate a typology o f p re-m o d ern and surviving tradi

tional watercraft;
2) define their basic characteristics;
3) investigate possible links between seagoing and inland
watercraft (a much neglected field); and
4) trace a possible evolutionary path.
T he Search
T h e starting point was a study o f the models and drawings or

paintings in museums, published technical drawings, and early
photographs. T he technical drawings were good, but few,
notably those by Paris (1840?) and Edye (1834)- T h e models
were the best source as they could be studied in the round.
Seamen, down the ages, made models o f their own ships — a
practice not confined to any one region or culture. These were,
for the reason that they were built by sailors, very accurate. O ld
photographs were also objective representations, but most were
Pre- M od ern S ri L a n k a n Sh ips 131
in a poor state o f preservation. Engravings, sketches and

paintings were better preserved, but were likely to be less
objective as the artists’ unfam iliarity with ship construction led
to unwitting m isrepresentation o f detail. Finally, there was the
evidence from traditional ships that survived into this century
and from yet surviving craft, though the latter are o f the most
basic kind. W ith this material I started my study.
U nfortunately, I was deprived o f a very im portant and
definitive piece o f evidence. That was the 'technical m anuals’ .
It may, perhaps, appear strange to use such a term about a p re
m odern craft-technology, but they did exist in Sri Lanka. I am
not aware whether they were actually written down on palm -leaf
manuscripts, though that would have been a distinct possibility.
(In India, Raut and T ripati, in 1993, cite a personal
com m unication to say 'The Departm ent o f Archaeology, Govt,
o f Orissa, . . . collected a num ber o f p alm -leaf manuscripts
m entioning the details o f such as how to lay the keel, stern,
stem, ribs and cabins. T he manuscript also m entions the time
to start the building work, technical terms o f each part and
other details’). What did survive in Sri Lanka, till at least the
1930s, was the old oral tradition. A ll Sri Lankan learning and
technology was passed down the generations via the apprentice
ship system o f training; either through a father-son link or the
Guru-Shishya Paramparawa (teacher-pupil continuum ). In the late
1930s, a group o f C olom bo schoolboys had gone on a field trip
to Panadura, a coastal town 17 miles south o f C olom b o. T h ere,
they had visited a boat-yard where indigenous workm en were
building 3 -masted schooners using traditional technology. In
the 1960s one o f the group told me that he had questioned the
master craftsman about the rules he followed as regards size,
proportions, form , etc. In reply, the m aster-builder had
recited a series o f Sanskrit Slokas (Sanskrit was the language o f
science and technology in those days and Slokas are m nem onic
verses) which laid down the rules by which a ship should be
built. Even though some thirty years had lapsed, an attempt was
made to retrieve this lost knowledge. Th ere was a lead that one
person (no longer living by then) had published it privately
but, unfortunately, it had not found its way to the National
132 SOMASIRI D E V E N D R A
Archives and I could not find it even am ong his personal

papers. This lost oral tradition w ould, no doubt have answered
many questions. In a deja vu scenario, about ten years ago, I
asked the same question in a traditional Arab Dhow yard. T h e
answer was that there was nothing written down: the keel-log,
which was the most valuable item, set the proportions. A fter its
ends had been squared-off, its useable length determ ined the
rest o f the dim ensions, with the design and proportions borne
in the eye o f the builder.
Fa c t o r s I n f l u e n c i n g D e s ig n
1. Navigating Coastal Waters
T h e basic form o f p urpose-built indigenous watercraft (as

opposed to floats and rafts) emerges from an interaction o f
several factors, mainly the nature o f the coastline, problem s o f
navigating the inshore waters, and the materials available for
construction. These factors have to be appreciated to u n d er
stand the evolution o f the craft. T h e basic problem s faced in
the Sri Lankan context were: comparatively shallow waters,
shelving beaches, off-sh ore reefs, heavy surf close to land, and
currents. Sri Lanka is a m eeting, as well as a dividing point for
the current systems o f the Indian O cean, between the A rabian
Sea and the Bay o f Bengal. C urrents are strong, up to 5~ 6
knots, abrupt changes occur within short periods or distances,
and the m ajor current systems flow parallel to the shore while
inshore currents may flow in an opposite direction. (West Coast
o f India Pilot l g f o 2 0 - 2 l )
To navigate these waters, the craft had to be o f shallow draught
and hardy construction with a bottom sturdy enough to
withstand frequent abrasion in sandy shallows. T h e craft that
em erged was the dual elem ent form o f the oruwa (the logboat and
single outrigger configuration), which elim inated the need for a
keel and also reduced the draught. T h e keel’ s absence helped the
craft cross the reefs, and mitigated the effect o f currents and the
set o f the sea. B ein g a sh ell-b u ilt m onoxylon o f sewn
construction, the resultant flexible craft was able to cope with
Pre- M od ern Sri L a n k a n S h ips 133
the torque experienced in the heavy surf. Thus the dual elem ent
craft, whether logboat and outrigger, or the large tw in-hulled
beach seine fishing craft, ( madel paruwa) were capable o f dealing
with most o f the problem s posed by the coastal waters.
However, the most im portant question for the smaller
fishing craft, was that o f making land, given the shallow waters,
shelving beaches and a negligible tidal range. These craft did
not 'drop anchor’ in most places, as there was no anchorage in
the sense the word is understood in m ore northern latitudes.
Instead, they were beached. W ith the dual element construc
tion, with both elements being made o f tough w ood, the boats
could easily be hauled up the beach. They did not keel over but
stood upright on shore and the two elements (logboat and
outrigger) gave them a bi-pedal stability. The booms attaching
the two elements were convenient for the crew to manhandle
the boat up the beach. In larger craft, beams placed athwart-
ships and projecting outboard, also served this purpose. This
feature is also seen in traditional Indian and Arab craft even
today, though the deep-draught Tuticorin coasters do not have
them. This is probably because they now call only at m odern
ports where they secure alongside piers.
T here was another hazard the boats had to face: that o f the
sand spits that were constantly form ing across the entrance to
rivers and lagoons. These craft were equally at hom e in, and
used along, the lower reaches o f rivers and in the lagoons. In
fact, many o f the deep-water fishing fleets, depending on their
size, sought shelter either just within the river’ s m outh or inside
the lagoon. Due to reasons dealt with elsewhere (Devendra 1993 :
17—2 4 )’ sand spits would form across the entrances, making it
difficult for any but the shallowest draught vessels to cross over.
Here, again, the round-bottom ed logboats, with no keel to
ham per them, could enter and leave the haven .1
Constructional materials were easy to find in Sri Lanka. Unlike
in the clusters o f smaller islands in the Indian Ocean, Sri Lanka
has an overwhelming wealth o f timber resources. The south-west
quarter o f the island was, until a hundred and fifty years ago,
widely forested and builders had a wide spectrum o f timbers, with
differing characteristics, to choose from . Accordingly, different
timbers were selected for different parts o f the craft. Vitharana

(1992) in an exhaustive study o f the surviving oruwa fishing
craft, lists twenty-eight different types o f wood used for eleven
different parts o f ship. A lthough iro n was locally available and
steel adzes were used to h ollow -out the logboats, fastenings
were not made o f iro n but o f coconut fibre. These craft were
all sewn with coconut-fibre rope and, as the coconut palm was
found all around the coast o f the islan d , there was no shortage
o f this material. Gunawardana (199O: 25 “ 4 4 )> quotes al-Idrisi
recording that Arab ships from O m an and Yem en used to
come here to obtain rope, trunks o f coconut trees for masts
and tim ber for planking, as well as to place orders for ships
which were constructed here. C ocon ut plantation for com
m ercial purposes, possibly also for the production o f rope, has
been known from the early years o f the Christian era; the first
reference is an inscription from the reign o f K in g Mahadathika
Mahanaga ( 9—21 a d ) . That foreigners knew about this is borne
out by A elian, (170—235 a d ) who says that \ . . this island in
the Great Sea which they call Taprobane has palm trees
w onderfully planted in rows, just as in lush parks the park
keepers plant shady trees.’ (Weerakkody 1997 : 235 )
2. The Ocean
T h e Indian O cean is characterised by its alternating m onsoons,

the inter-m onsoonal cyclones, the current systems prevailing,
and the presence o f many island chains, reefs and shallows. In
the early historical p erio d , oceanic sailing took place only in
the m ore northerly waters .2 T h e Sri Lankan ships that first
ventured into the ocean were, presumably, larger versions o f
the fishing craft that evolved in inshore waters. T h e ir
developm ent into som ething m ore seaworthy would have
depended on the suitability o f the base form for adaptation
to the new environm ent and for the assimilation o f features
fro m other tech nologies. Sri Lankan seam en, ships or
travellers have been referred to in many parts o f the landmass
that was the then-know n world — Asia, Europe and A fric a .3
Vajrabodhi in the 8 th. century m entions 35 Persian ships in a
P re- M o d er n S ri L a n k a n S h ips 135
Sri Lankan port but en route farther east (Gunawardana (199O:

36, citing Silvan Levi). He also cites Ibn Battuta in the 14th
century m entioning a fleet o f "'one h un dred” vessels o f varying
size’ belonging to the recently-em erged northern Sri Lankan
kingdom o f Ariya Chakkravarti in an Indian port ( l 99 0: 3 &)-
Indian O cean Rim countries were jo in ed by econom ic, social,
religious and political ties. Sri Lanka was aware that it was one
country in a complex o f peoples and cultures. The country
itself, from very early times, was home to many peoples and
religions. In fact, this awareness extended to all the coastal
com m unities around the O cean, which led to the early
navigation o f this O cean.
3. Cross-pollination between technologies
Th e question arises whether the basic form o f the Sri Lankan

oruw a, and its sewn construction, proved viable for long oceanic
passages. In the absence o f any hard evidence o f a ship built
m ore than two centuries ago, it would seem necessary to accept
that the voyages described above, were undertaken by such
ships. This line o f reasoning is given credence by the survival,
into this century, o f large sewn-plank, shell-built, two-m asted,
outrigger equipped cargo vessels. These, the yathra dhonis, are
described below. They were used in this century between Sri
Lanka, both coasts o f South India, the Maldives and Malacca.
They have the m ajor characteristics referred to in relation to
early inshore craft: coconut-fibre rope sewing, outrigger,
double-ended configuration, and plank construction. The
seeming contradiction is the substitution o f planking for a
dugout. A n observation by H orn ell (1946) however, provides a
possible explanation:
T he final stage in the conversion o f the dug-out into a

fully plank-built boat is attained when the dugout u n d er
body is reduced to a keel-like axial beam, with sides raised
on its edges by num erous strakes o f sewn-on planking.
This was the m ethod o f construction em ployed by the
Persian and Arab shipwrights down to the sixteenth
136 So m a sir i D evendra
century . . . and the Sinhalese coaster o f the G u lf o f

M annar (H ornell 1943 >4 3 “ 6). In neither o f these vessels
were iro n fastenings present. T h e whole o f the planking
was sewn together. . . .’ (The 'Sinhalese coaster’ , accord
ing to H o rn e ll’ s own reference, is the yathra dhoni.)
Phillips-B irt ( l 979 : IJ) gives a drawing o f a 'section o f Indian

canoe from Dawki with vestigial dugout base and reverse clinker
planking above’ which shows an entirely acceptable evolutionary
stage, and reinforces H o rn ell’ s observation. His explanation is
given further credibility by the large, punt-shaped, sewn-plank
seining-boats, madel paruwa which are built upon a base o f two
vestigial dugouts. K entley & Gunaratne (19 8 7^ 35 —4^) chose
not to trace the evolution o f this feature in their study o f this
craft, preferrin g to call them chine strakes 'as they locate the
feature and im plies that it runs longitudinally’ . I have,
however, argued elsewhere (Devendra 1995 : 225 ) that this
craft was a dou b le-logb oat-h ulled craft 'that must be co n
sidered the end o f the line o f developm ent o f Sri Lanka’ s
logboat culture’. This craft operates close inshore: the yathra
dhoni which made the quantum leap that H orn ell describes is an
ocean-going one.
Finally, we have the bas-reliefs o f ninth century B orobudur
(Indonesia) which show detailed representations o f the same
class o f ship. T h e only difference o f note is the use o f hinged,
pivoted or detachable outrigger boom s, placed near the
waterline, in parallel with the surface o f the sea, unlike the
downward-curving flexible boom s o f the Sri Lankan ship,
attached to the hull just below the weather-deck. Both types o f
ship, it is im portant to note, were the products o f ship
building cultures characterised by logboats, sewn construction
and outriggers. W hether one was derived from the other is
open to debate but coincidence, as an explanation for the
similarities, is untenable.
Sailing the Indian O cean involved much m ore than building
ships. Navigation, knowledge o f the movement o f stars and
constellations, winds, currents, tides, the ability to know your
position to some degree o f accuracy were also needed.
P re- M o d er n S ri La n k a n S h ips 137
Presumably this knowledge was acquired through experience.

But it is difficult to overestimate the value o f borrow ing and
adaptation o f foreign technologies and practices.
A m on g the many borrowed and adapted features were the
fore-an d aft rig featuring A rab-In dian lateen sails and the
rudder. Previously, Sri Lankan craft had square sails and were
double-ended. They could not tack or effectively sail upwind.
W ith the need to keep the outrigger to windward, the best that
could be done was to 'go about’ , changing the set o f the sail.
W ith the adoption o f the fore-an d-aft rig, there was the ability
to tack, and 'goin g-abou t’ was unnecessary. W ith this new rig
came the rudder and the double-ended craft acquired a fixed
Bow and Stern. However, old practices die hard and even the
last yathra dhonis did not adopt a transom stern (which would have
been logical) but retained the double-ended hull configura
tion. They also retained the outrigger, which was no longer
essential. T he effect o f this mixture o f constructional features
on the sailing characteristics o f the ship is discussed below.
As in the case o f ship-building, navigation, reading o f the
stars, knowledge o f tides, pilotage etc. appear to have been
m aintained and passed on by oral tradition. T h e knowledge
resided in the master o f the vessel (navika) or the sailing master
(marakkalehe) o f fishing com m unities ashore. Even as late as
Dutch times, the Persian term for 'p ilo t’ (sambandar) was in use
and later became a personal name 'Sahabandu which is yet in
use: Indian texts, though, describe pilots and their knowledge
in the early centuries o f the Christian era. T here is no evidence
o f maps or charts. Stars were known and recognised and
Vitharana (1992) has docum ented fifteen names o f stars used
by the present-day fisherm en. However, the knowledge has all
but died out and his sources could not agree on the identity o f
the stars they were nam ing. As all deep-sea sailors have passed
on, this is not surprising. We do know that the Sri Lankan
ambassadors to the Rom an C ourt o f Claudius had com m ented
on the different starscape over Rom e, particularly rem arking
on the difference in the position o f the Great Bear and Pleides,
that Canopus was much brighter in the Sri Lankan sky, that the
sun was always to the south and therefore shadows in the
138 SOMASIRI DE V E N D R A
Rom an latitudes always pointed north. (Weerakkody 1997 :

226 —227 )- C om ing, as they did, from a country where
astrology was a serious study, this is not surprising: it is the
loss o f the application o f the knowledge o f the stars to
navigation that is regrettable. Fishermen however have retained
certain types o f knowledge. Vitharana has docum ented twenty-
four separate words to identify features o f the sea, twelve to
denote prevailing drifts, and twelve for different winds. Since
the craft are small, only the specialist names for the master and
the lo o k-o u t are retained as well as one collective noun
denoting the crew. Viewed against this background o f lost
knowledge it is com forting to note that the navigational
knowledge o f the Arabs, with whom we had much to do, has
been docum ented com petently. 'Sailing the latitudes’ by
m easuring the height above the horizon o f Polaris, Arab
sailors plotted the positions o f places they were interested in.
C om b in in g latitude, dead reckoning and days run as data, they
positioned themselves with sufficient accuracy to sail where they
wanted and find their way back home, on a regular basis. Not
maps but plotted positions constituted their charts. W hen these
plots are superim posed on a Transverse M ercator’ s Projection,
an acceptable degree o f correspondence is shown (Tibbets:
I 977 )- These maps also indicate the cross-oceanic routes — not
mere conjectural lines, but actual bearings and distances
between known and plotted points. Indian sources, too, are a
yet insufficiently exploited m ine o f in form ation , as no
satisfactory attempt has yet been made to collate the material
and chart courses. The attempt by Rao (1987) to plot Indian
sea-routes, particularly between the east coast o f India and
south-east Asia is, thus, a good beginning. Chinese maps, such
as those o f Ma Huan, also show navigation using stars. Bellec
( l 9 9 7 : 5 -1 0 ) presents an interesting explanation o f Ma H uan’ s
guides to celestial navigation giving, as his example, the passage
from Sumatra to Sri Lanka, using selected stars and constella
tions as guides. N ot long after that the Gujarati pilot, M u ’allim
Kanaka piloted Vasco Da Gama. This ocean, therefore, was a
highway, well traveled, with 'signposts’ and travel guides.
Sailors, being both gregarious and pragmatic by the nature o f
P r e - M o d e r n S ri L a n k a n S h ip s 139
their calling, are always on the look out for whatever is helpful
to them, whatever the source. T here was an ongoing cross
pollination o f ideas and technology among them. They picked
up from each other, by observation or other means, all the
knowledge necessary to navigate the Indian O cean. It is
reasonable to assume that some o f this com m on pool o f
knowledge was known to, and used by Sri Lankans, as they, too,
were long-distance sailors.
This, then, was the ocean Sri Lankan ships and seamen
sailed in. It cannot be claimed we were in control o f the seas,
or any part thereof, at any given time. But it is safe to assume
that we were participants in a m ulti-user environm ent, and that
our ships — in com m on with all ships that sailed this ocean —
incorporated constructional features, adopted from various
sources, that had been proved in blue-water sailing. These
features, superimposed on the basic form s evolved in coastal
waters, made these craft as suited as any other to undertake
oceanic passages.
Sh ip - B u il d e r s , S h ip y a r d s and Ports
The next step was to identify the builders. Today, only fishing
craft are built in Sri Lanka and there are no ship-builders left
— at least, in the Sinhala-dom inant parts o f the country.
Fishing craft are built by those who use them and most o f whom
belong to one caste. In India, there are yet many traditional
craft built by traditional builders: they belong to the H indu,
Muslim and Christian com m unities. A m on g the H indu (who
are culturally most akin to Sri Lanka), they have been variously
referred to as mestris (K unhali 1993 : 57 ) > mestas and acharis
(Sunderesh 1993 : 3 °)> and viswakaramas (Raut &, Tripati 1993 :
52). T h e two latter names are specific to the artisan caste. In
Sri Lanka, too, this caste is referred to by all the names above
and it is tem pting to believe that they were the builders o f old.
However, this was not so. A lthough the achari and visvakarma are
names denoting the Artisan class, the mestri or mesta name is used
by others to denote a master o f his trade: the word is derived
from Portuguese. In particular, the name is found am ong the
caste to which fisherm en o f Sri Lanka belong. Also, one

particular craft — carpentry — is carried on both by the artisan
and that caste (as well as others). Probably the ship-builders
then (as now) were from the people who used the craft. A clue
that fits in with this is the family name, specific to this caste, o f
Galappathi, which derives from the technical term 'to caulk’ .
Thus, in spite o f sharing a similar caste system with India, here
we come across a m ajor difference.
In the Tam il-dom inant north and the M uslim com m unities
o f the east, prevailing conditions have not favoured field
research since the beginnings o f my study. It was the Fishing
caste o f Jaffna who were the sailors o f the north and in the areas
dom inated by them ships larger than off-sh ore fishing craft
(the Jaffna thonis and the \ a tta i / vattal’ — see below) were built. As
in India here, too, ship-building activity was carried on by
different com m unities. Johnston (18 2 4 : 537 —54 ^), drawing
on his experience o f Adm iralty laws prevailing in or about 1802
comments:
The maritime laws and usages, which prevail amongst the

H indu and M oham m edan mariners and traders who
frequent Ceylon, o f which I made a complete collection
while presiding in the V ice-A dm iralty C ourt o f that
island, may be classed under four heads: First, those that
carry on trade in small vessels between the coasts o f
Malabar, Corom andel, and the island o f Ceylon; second
ly, those which prevail amongst the M ohammedan m ar
iners and traders o f Arab descent between the coasts o f
Malabar, C orom andel and the island o f Ceylon; thirdly,
those which prevail amongst the Arab m ariners and traders
who carry on trade in very large vessels between the eastern
coasts o f A frica, Arabia, the Persian G ulf, and the island o f
Ceylon; fourthly, those which prevail amongst the Malay
m ariners who carry on trade between the coasts o f
Malacca, the eastern islands, and Ceylon.
T h e first are in some degree m odified by the tenets o f
the H indu religion and by H indu law. T h e second, the
third, and the fourth are m odified in a great degree by
Pre- M od ern S ri L a n k a n Sh ips 14 1
the tenets o f the M oham m edan religion, and M oham

m edan law .’
This source indicates the variety o f shipping from Indian

O cean countries operating under indigenous legal systems even
after this country came under British rule in 179 ^- British rule
had been preceded by some 3OO years o f Portuguese and Dutch
control o f the maritime provinces. Even before this, the
waning power o f the central governm ent had weakened the hold
o f the Sri Lankan kings on the m ajor ports (and Jaffna in the
north) which were controlled by foreign traders, who adm i
nistered them according to their own laws. M ajor ports had, by
then, becom e foreign enclaves. T h e 'Galle T ri-lin gu al In
scription’ , indited in China with the date corresponding to 2
February, I 4 0 9 > and set up in the port-city o f Galle some two
years later by C heng H o, is inscribed in Chinese, Tam il and
Persian (in A rabic characters): a clear indication o f the
cosm opolitan nature o f the city (Devendra 199 ° C: 265)• O n
the eastern coast, in Trincom alee, the site o f an Arab
settlement in an em inently suitable inlet (N icholson’ s Cove)
has been attested by the discovery o f a graveyard containing
tombstones bearing 13th and 14th century dates (Devendra
1990 a: 209—2 I 7)- Johnston (1827) had already discovered one
in C olom bo, dated to m id IO th century. Ibn Batuta, visiting Sri
Lanka in the 13th century, speaks o f a 'Prince o f the sea’ ,
named Jalasti, who held sway over C olom bo com m anding a
force o f 500 Ethiopians. Jalasti has been variously described as
a powerful trader and a pirate. O th er pirates have also been
written about.
In earlier times, however, when a strong central governm ent
had prevailed, laws adm inistering ports and matters maritime
were prom ulgated on inscribed tablets set up in the vicinity o f
the relevant ports. A n inscription in Devundara (Dondra) in
the South states (Paranavitana 1953 : 53 —54 ) :
. . . apart from the levying o f such imposts as have been

approved by the M aha-P andithe, illegal imposts shall not be
levied. To those com ing from foreign countries, means
shall not be afforded to avoid the payment o f imposts and

duties that are due, which they do by establishing places o f
business, corrupting royal officers by means o f presents
and keeping with friends the merchandise smuggled from
their own countries . . .
A nother inscription, at Nayinativu in the extreme N orth, states

(Indrapala 1953 : 63 “ 7 ° )
. . . the foreigners should come and stay at Uratturai

(U ratota), that they should be protected and that
foreigners from many ports should come and gather at
our ports; as we like elephants and horses, if the vessels
bringing elephants and horses to us get wrecked, a fourth
(share o f the cargo) should be taken by the treasury and
the (other) three parts should be left to the owner; if
vessels with m erchandise get w recked, an exact half should
be left to the owner. . . .
T h e Culavam sa, the continuation o f the earlier Mahavamsa

(historical chronicles that record Sri Lankan history from
543 the traditional arrival date o f the Aryans to this
country) carries a most illum inating account o f the logistics o f
a punitive raid to Burma, stemm ing from a dispute involving
the trade in elephants. T he incident in question is supported
by a contem porary inscription recording the grant o f land and
privileges to the leaders o f the expedition. It makes reference to
Hatan N av, (lit. 'Warships’) but it is uncertain whether they were
a kind o f ship specially fitted out for war, or ships constructed
for the expedition. T he latter is m ore likely, following the
Culavamsa description o f their construction:
. . . gave the order without delay to make ready ships o f

various kinds, many hundreds in num ber. Now all the
country round about the coast was one great workshop
occupied with the building o f the ships taken in hand.
W hen within five months he had all the ships well built he
assembled them . . . (at) Pallava-vanka . . . he had provisions
supplied for a whole year . . . and abundant weapons o f
Pre- M od ern S ri L a n k a n Sh ips 143
war such as arm our . . . gokanna arrows . . . for defence

against elephants, also different kinds o f m edicines
preserved in cow -horns for dealing with venom ous
wounds caused by poisoned arrows . . . remedies for
curing the poison o f infected water . . . iron pincers for
extracting iro n arrow -heads, . . . lastly also skilful
physicians . . .
This description, which indicates knowledge o f the requ ire

ments for storing ship for an offensive action, is m ore
im portant for the reference to the ships o f various kinds being
built on the sea-shore. It is a pointer to ship-building without
the need for inland shipyards; to the ability to m obilize a large
w ork-force capable o f undertaking the building o f a large fleet;
and even to the possibility that the ships were o f the outrigger-
equipped and capable o f being built on the beach — that they
were, in fact, ja th r a dhonis. Even leaving room for exaggeration,
this account rings true in terms o f the building o f an
expeditionary force. T h e account continues, and even records
the loss at sea o f the majority o f the fleet. T here are other,
equally credible accounts o f attacks by sea between local warring
factions and other expeditions to intervene in south Indian
politics. A ll o f these indicate knowledge o f naval strategy in
offensive operations.
T ech nology T ran sfer
T he ships that sailed the Indian O cean were the products o f

technologies that evolved in their home waters. Still, all were
sea-worthy craft. W ithin the Indian O cean, several technology
zones existed, namely:
T he Dhow zone
The Sewn-boat zone
The Single outrigger zone
The D ouble outrigger zone
The H ybrid craft zone
Th e Shaped-log-raft m icro-zone.
W hile the first four zones above are well known (H ornell 194*6:
unnum bered last page, and M acPherson 199O: 261—264), I
have postulated the other two on the basis o f research and
interpretation. This differentiation into zones requires some
explanation (fig. 5 -1)-
Dhow technology
Characteristics: carvel built plank boats, sewn with coconut -

fibre rope, prom inent rudder, forward raking mast carrying
large Indo-A rab lateen sails on a long boom hoisted with a
pulley attachm ent. E volution and use: largely (but not
exclusively) in the Red Sea, A rabian G ulf, west coast o f India
and generally in the A rabian Sea.
Outrigger technology
Characteristics: the m onoxylon, or log-boat, as the m ain unit

o f construction (with or without plank extensions), very shallow
in draught, with attached outrigger (single or double) in place
o f a keel, coconut-fibre rope sewn and fastened throughout,
dou b le-ended , with square or lug sails (later fore-an d aft on
larger craft). Evolution and use: originated in the South-East
Asian seas and spread cross-ocean westwards to Sri Lanka,
Madagascar and East A frica and eastwards to all the Pacific
islands. N ot found in the Indian sub continent until late.
H orn ell (1946) concludes, from an examination o f the areas
where the two form s are u se d , that the single outrigger is m ore
seaworthy in open seas. He places the single outrigger to Sri
Lanka and the Bay o f Bengal islands (in the Indian O cean) and
to the Pacific O cean islands, and the double outrigger to east
A frica, Madagascar and South-East Asia. Ph illips-B irt ( l 9 79 :
135) has hazarded 'that the single outrigger preceded the
double outrigger, though for small craft and within the
lim itations o f prim itive technology the latter may not have been
an im provem ent.’ H ornell, however, had abandoned this
position, which he had earlier held, and said 'T here is now
Key
____ Dhow technology zone (after McPherson)
____ Single outrigger zone (after Hornell)
— Double outrigger zone (after Hornell)
Hybrid technology zone
.......... Shaped-log-raft micro-zone Sewn-boat zone (encompasses all others)
Figure 5*1 Ship building technology zones
little reason to doubt that the double outrigger antedated the

single type . . . ’ (1946: 268) I have, however, shown that in Sri
Lanka, the double outrigger did not appear. (1995) As the
m ajor country in the Indian O cean to have been by-passed by
the double outrigger, this country presents a strong argument
against the double outrigger being considered a necessary
predecessor to the single one.
Hybrid technology
This zone, which I proposed in 199 °> covers the old

Erythrean Sea and its limits are (subject to correction) the
lands and seas referred to in the fPeriplus o f the Eiythrean S e a . By
definition, this is the area where East and West met and
146 So m asiri D evendra
exchanged ideas and knowledge, and where the technological

peculiarities o f one type was adopted and adapted by another
to create the greatest num ber o f hybrid types. This cross
pollination process continued until the Indian O cean became
a 'British Lake’ . Example: In Jaffna in the north o f Sri Lanka,
until the I9 3 0s ships were built that — as regards hull form —
are almost clones o f nineteenth century British arm ed
m erchantm en, com plete with painted false gun -ports. C h a r
acteristics cannot be attributed as many types abound in the
indicated area.
Shaped-log-raft technology
This m icro-zone encompasses the south Indian and Sri Lankan

waters, where it was possible to sail the ocean with m inim al
com fort (Thivakaran and Rajamanickam: 1992: 2 3 - T he author
o f the Periplus and both H orn ell and Paris noted these craft.
They were the first to be depicted in a rock carving in p re -C .E .
Sri Lanka, and they are craft that are still in use on both sides
o f the Palk Strait. T he logs are either fastened with wooden
pegs, or tied (not sewn) with rope, and can be sailed or
(sometimes) used with a temporary, easily shipped outrigger.
U nlike the dhow and the outrigger, this craft is extremely
lim ited in its area o f use, being confined to India and Sri
Lanka only. A Sri Lankan chronicle, the Sam m ohavinodani,
describes rafts that were sailed from the port o f Jambukolapattana
(com m only identified with Kankesanturai in the north) to the
east coast o f India in the second century B C. These had three
decks: the lowest was awash. T h e passengers occupied the
second and the third carried goods. T h e passengers, it is said,
were unhappy to sail in them as they appeared unsafe. From the
description, it is a strong possibility that they were a precursor
o f the shaped-log-rafts o f today.
Sewn Boat technology
This was com m on to all the above zones, but retained until late
only in the outrigger and dhow zones.
Pre- M odern S ri L a n k a n S h ips 147
F ig.5 .1 shows the Indian O cean with all the zones marked on
it. Sri Lanka is the only location which falls into five o f the six
technologies m entioned: dhow, sewn-boat, hybrid, outrigger
and shaped-log-raft. Strangely, even dhow technology has not
affected Sri Lankan ships to any appreciable extent. O nly the
double-outrigger zone by-passes Sri Lanka completely. C o n
structional elements o f all these zones are seen in Sri Lankan
craft, either in pure form or in admixture. This fact is,
accordingly, reflected in the description o f the craft recorded
here in the typology.
M a jo r D if f e r e n c e s w it h in S ri L a n k a
N orthern Sri Lanka and South India are geographically and

culturally very close. T h e antiquity o f the cultural links
prevailing today are a matter for debate outside the purview o f
this paper. However, there is no gainsaying that, in terms o f
fishing, chank fishing and similar aspects relating to making a
living from the sea, the people on either side o f the Palk Strait
had many com m on activities. A m on g them was the type o f craft
used at sea. A case in point is the m icro-zone o f the shaped-
log raft technology referred to above. O n an even smaller scale
is the use o f two different types o f small fishing craft u sed , the
oruwa in the south o f Sri Lanka and the vallam and the kattumaram
in the north o f the island as well as the south o f India. H orn ell
(1943) was moved to com m ent on this phenom enon as
follows:
'N o greater contrast can be found in small craft designing

than that between the types used on opposite sides o f the
G u lf o f M annar, South o f latitude 9 degrees N. O n the
Indian, or Tam il, side the catamaran or boat canoe alone
are employed; on the Sinhalese side, the outrigger canoe
is the national and dom inant design, the catamaran being
used only in the northern, or non-Sinhalese part o f the
island and by migrant Tamil fisherman in C olom bo, with
the dug-out restricted to its proper sphere o f usefulness
on rivers and inland waters’
H ornell correctly uses the word 'catamaran (from Tamil kattu-

maram) to mean lo g raft’ , and not in the corrupted m odern
sense o f 'double-hulled craft’ . It is one o f the forms o f shaped-
log rafts which I have grouped together as form ing a distinct
m icro-zone. In Sri Lanka, H o rn ell’ s geographical differentia
tion is substantially correct: catamarans are largely used in the
N orth and N orth-East o f the island, with a dim inishing scatter
even as far South as the N orth o f Colom bo, and, on the East
coast, dim inishing south o f Trincom alee. This type o f craft is
also used in South India, particularly in Kerala, Tamilnadu and
A ndhra Pradesh in the Southern part o f peninsular India, and a
wider variety o f forms are in use. In Sri Lanka, there are but two:
the smaller four log pegged rafts and the larger five to seven
lashed rafts, which can be fitted with a sail and, occasionally, with
an outrigger. In India, in addition to these, the rafts are also
converted to virtual flat bottom ed boats by the addition o f
vertically attached logs to provide a degree o f freeboard, but
constructed entirely o f shaped logs, not planks. These craft,
ranging from seven to twenty one feet in length and two to five
feet, in breadth, are among the most primitive craft that sail even
the coastal seas but, if their antiquity is taken as an index o f a
successful design, they have to be considered as having proven
themselves. In Sri Lanka, a second century BC inscription at
Duvegala, at Polonnaruwa, shows an inscribed picture o f one
(Paranavitana 197O: Pl.XXIX) complete with mast topped by a
Nandipada (Taurine) sym bol .4 It is likely that this represents a craft
that was in use between India and Sri Lanka for more than two
m illennia. A n alternate possibility is that here, as elsewhere, the
full shape o f the hull o f a ship, rather than a raft, is shown in
profile. This, however, is a supposition.
T h e other, or 'dom inant design’ , H orn ell refers to is the
logboat and outrigger craft, the oruwa, whose origins are lost in
antiquity. H o rn e ll’ s remarks concerning the Sri Lankan —
Andam an Islands single outrigger have already been referred
to. These craft differ from the double-outrigger in other ways,
too. T h e larger, deep-sea craft particularly, but not solely are
fitted with vertical plank extensions sewn onto the gunwale o f
the m onoxylon. Thus there is a considerable difference in
relative heights above water between the freeboard o f the

logboat and the outrigger float. This imparts a downward curve
to the connecting booms, from the onboard ends to the
outrigger ends. They are therefore curved and made o f flexible
wood. U nlike in the outrigger ships seen on the B orobudur
bas-reliefs, pivots or hinges are not used at either end o f the
boom s, the flexibility o f the curved booms enabling them to
cope with the vagaries o f wave and swell. N either are they
removable. H orizontal booms are only used on inland craft, in
some o f which there are no vertical extensions.
R esults of th e Search
T h e first phase o f my study was to try to elicit com m on

characteristics (Devendra 199 ° : 265—27 1)- Using a m odel (fig.
5 .2 ) devised by P hillips-B irt (1979) as a control, I found the
following:
Most common features
T h e logboat as the unit o f construction

T h e use o f the outrigger balance log
H ull o f sewn plank
Sh ell-built hull of, carvel
Evolutionary paths
1) Log — dugout — shaped log rafts

2) Log — dugout — m ulti-hulled dugout — m ulti-hull, planked
3) Log — dugout — dugout and planks — shell construction,
carvel
Hypothesis arrived at
P re-m od ern ships were carvel built o f shell construction,

using sew n-plank technology; hulls were d ou b le-en d ed ,
carried squares sails, and were fitted with outriggers. Even
after the adoption o f fo re-an d -a ft rig and rudder after the
150 S o m a sir i D e v e n d r a
FLOATS
in f la t e d gourds ja r s reed bark lo g s rafts

s k in s b u n d le s
PRIMITIVE
BOATS
i
s k in o n s k in o n reed bark dugout shap ec shap ed
b a sk e t w ood r a ft r a ft
fr a m e lo g bam boo
dugout
d u g o u t p lu s m u lit-h u ll
p la n k s dugout
W OOD PLANKED
.^ BOATS
s h e ll s h e ll b u lk h e a d
c o n s tr u c tio n c o n s tr u c tio n c o n s tr u c tio n
carvel c lin k e r ca r v e l
& te n o n s s h e ll o n
r m o u l d s —m o u l d
s k e le to n
s k e le to n m u lti-h u ll
c o n s tr u c tio n p la n k e d
s h e ll
carvel _„
c o n s tr u c tio n BOATS OF
carvel METAL
PLASTICS
CEMENT
s k e le to n s h e ll m u lti-h u ll
c o n s tr u c tio n c o n s tr u c tio n m e ta l &
m e tal r e in fo r c e d p la s t ic
& fe r ro -c e m e n t p la s tic s
in f la t a b l e s
o f rubber &
tr e a t e d c lo t h
F ig u r e 5*2 Evolution o f constructional techniques
ninth century, the outrigger continued to be retained and,

in the smaller craft, even the dou b le-en ded form was
retained.
These points applied to both sea-going and inland craft.
However, after an in depth study o f inland craft, I came to the
following conclusions.
P r e - M o d e r n S ri L a n k a n S h ip s i5i
1) T h e evolution o f watercraft in Sri Lanka can be traced in

sequence from the oruwa to the paruwa and the ja t h r a dhoni.
T h e base form is a double-log craft. H orn ell has traced the
evolution from dugout to keel log.
2) T he madel paruwa (an inshore seining boat) and the paruwa (a
river/canal cargo carrier) evolved independently to suit
environm ental and functional needs.
3) Earlier paruwa prototypes which were superseded by the
river/canal paruwa which, in its last traditional form , was a
hybrid: being, in terms o f structure, influenced by Dutch/
European barges — thus being skeleton built unlike other
traditional craft, all o f which are shell built.
4) T he paruwa had its origin in the rivers o f the southwest,
consequent to the growing political and econom ic dom
inance, the dem ographic shift and the urbanisation o f this
region.
To analyse the craft according to structural characteristics use

was made o f another m odel (fig. 5 -3 )* This test confirm ed that
all known Sri Lankan craft were either rafts (the m inority) or
shell-built boats (the majority) in which the reduction and
construction processes were widely used, with the transform a
tion process used m ore sparingly. A ccordingly, they all fell into
the categories C l, C 2, C 3, C 4 and C6 o f that m odel.
T h e other product o f my search was a typology. It is
deliberately term ed 'tentative’ since new inform ation continues
to come in from time to time, particularly with regard to
certain variations o f craft types. First form ulated in 199 °» it has
undergone several revisions. There is, now, reason to believe
that no fundam ental changes need occur, and that the
inform ation concerning new craft could be accommodated
within the framework already drawn up. I am, however, still
chary about making it definitive.
This tentative typology (see Appendix) however, is not the
only, nor the first, to be attempted. Others have formulated
similar classifications and it is possible to accept each o f these as
equally valid in terms o f the parameters applied. The two that
come most readily to mind are those compiled by Gerhard
So m a sir i D e v e n d r a
W ater T r a n sp o r t
B u o y an c y
A p p lie d ?
D ire c t In d ire c t I
F lo a ts R afts & B o ats
B u o y an cy
v D e r iv e d ?
/ F ro m \ F ro m '
( In d iv id u al W h ole
V E le m e n ts J V essel .
R afts
F u n d a m e n t a l)
\ G o n c e P. /
KEY
water transport W atertigh t

W ater
p r o o fe d
S h e lf
F ra m e
attribute
attribute state Sh e ll - bu i It S k e le t o n - b u ilt
R reduction
C construction
P rm c l p a l \ P r m o p a lX
T transform ation T e c h n i q u e s ': / T e c h n iq u e s ‘r /
c:i C2 c;:5 c(> c:; C8 cq Gi o Cn C12 C13 C14
F igu re 5*3 T y p e s o f w a te r t r a n s p o r t
Kapitan, (1987= 137) and by the Sri Lankan Department o f

Fisheries in collaboration with U N D P and F A O (ig 9 4 )- There is
a very good reason why these typologies differ from each other
and mine: each has been form ulated with a different purpose in
P re- M odern S ri L a n k a n S h ip s 153
mind. T he Department o f Fisheries’ typology is a by-product o f
research into the gear, equipment and craft currently used by
fishermen and it deals with mechanised craft as well. Kapitan, on
the other h an d , only deals with the vanishing types o f traditional
craft, which are still in use, and which he is striving to save for
posterity, even as museum pieces. My own attempt is different in
that it is concerned with the past and surviving craft are only
studied to determine forgotten forms. I also place much
emphasis on inland watercraft, a class that has not been
specifically dealt with by earlier writers.
Part 2,
S ri L a n k a n S h ip s
A ll o f the foregoing provided only a framework o f reference

against which a particular description o f a vessel could be
judged credible, or not. Yet the exact appearance o f the ships,
their size, and types are questions that were left unanswered.
C onflictin g points o f view had been advanced. In 1966,
Toussaint (1966: 4 ” 5 ) wrote:
T he Sinhalese people never looked towards the sea and

the navigators whom history records were always fo reig n
ers. T h e outriggers themselves are o f foreign origin, and
it is not in C eylon that we shall really com prehend the
ocean’ s story.
Prof. R .A .L .H . Gunawardana and Y u m io Sakurai (199O:

2 J 0 — 2 8 0 ) , translating and com m enting on an extract from a
ninth century Chinese literary work, say the author o f the
original, referring to the ships from the Southern Sea which
arrive each year at A n -n a n and K uang C hou, states 'Am ong
these, the ships from the L ion K ingdom are the largest, with
stairways for loading and unloading which are several tens o f
feet in h eigh t.’ H ere we have two recent and very divergent
views. O n the basis o f the research conducted, I believe that
the first is too r ig id , and that an open m ind has to be kept on
the latter. Before com m enting further, it is necessary to
i 54 SOMASIRI D E V E N D R A
describe the next step in my search, which revealed the

m orphology o f the p re-m o d ern ships that survived into the
twentieth century.
The Jaffna thonis
In the course o f my quest for graphic representations I came

u pon H o rn e ll’ s photographs o f the Jaffna thoni (fig. 5 -4 ) which
is a ship belonging to the hybrid technology zone I have
referred to. This would m ore or less cover the old Erythrean
Sea, spanning the waters between the west coast o f India and the
eastern shores o f A frica, touching the northwestern coast o f Sri
Lanka and going southwards to the Maldives. In these waters
M editerranean, Arab, Indian and later, Portuguese, Dutch,
French and British ships met and m ingled. T h e Jaffna thoni is a
fine example o f this interm ingling and cross-pollination.
W hile basically a very South Asian craft in the way it is fitted out
o n b o a rd , its lines are very rem iniscent o f European ships o f a
century or m ore ago. H orn ell, com m ented in 1943 :
T h e larger type o f schooner is o f purely European design.

It diverges in no detail from the small w ooden schooners
F ig u r e 5-4 Jaffna T h o n i (H ornell)

employed in English coasting trade in the nineteenth

century except in one detail. No Jaffna schooner would be
considered shipshape unless a row o f im itation square
black ports were painted along each side, simulating the
appearance o f the gun ports o f A rm ed East Indiam en o f
the eighteenth century.
T h e European colonisers foun d , in this part o f the w orld, the

m aterial and the skills to build ships to their liking. Initially,
they m odified certain smaller European designs to incorporate
elements successfully proven in these seas. Later, large ships for
the European were constructed in India. G ill ( l 9 9 3 : 75 ) states:
T h e India built ships traded with U .K . and between June

and August, 20 such ships took rice from India and
returned with cargoes. T h en the English builders and
owners protested. Before a select committee (1814) o f the
House o f C om m ons (Chaired by Sir Robert Peel), John
H illm an a builder expressed that India built ships will kill
their industry. 'A n India built teak ship, after she had
perform ed 6 cargoes is equal to one o f ours after she has
perform ed three.
P hillips-B irt, speaking o f tenth century Arab ships, ( l 9 7 9 : 122)

comments:
. . . the teak often used for planking the better craft is

ideal for iro n fastening, containing as it does an oil that
preserves the metal, unlike oak, with its acid content,
which attacks it.
T h e technology transfer worked both ways, as the Jaffna thoni

demonstrates. Earlier in the colonial days, the Maldivians
picked up the lines o f the Portuguese Caravels and built them
late into this century. It is noteworthy that when Captain A llan
Villiers was preparing him self to sail the Mayflower II across the
Atlantic, it was on these Maldivian vessels and Arab dhows that
he refreshed his small ship sailing skills. T h e point is that these
designs continued to be used in the Indian O cean long after
they had been abandoned in the countries which gave them
156 Som asiri D evendra
birth. In about 1802, C eylon became a crown colony. T h e

C h ie f Justice, Sir Alexander Johnston noted (1824) that,
around 1802, he was in fo rm ed by M uslim priests and
merchants that, for some hundred years, they had been using
Arab translations o f Ptolemy. But, being in need, they had sold
them to merchants who were sailing the Sri Lanka-South East
Asian route. Shipowners o f the Jaffna thonis were either H indu
or M uslim. T h ere was only one point o f difference, as far as
design and decoration were concerned, which depended on
ownership. In the H indu-ow ned ships the sharply raked stem
ends in an inwardly coiled ornam ental head called a surul
marked with three horizontal bars to represent the three ash-
streaks that a H indu wears on his forehead. U nder this feature
is the ship’ s shrine. Also, on either bow a neat oculus is nailed,
to represent the eye o f the g o d , who will guide the ship through
the seas. The M uslim -ow ned ships do not have either o f these
features. To quote H orn ell ( l 943 ) again:
A t each end o f the vessel is a short decking, ending in each

case in a high transverse breakwater, 2V2 — 3 ft. in height,
sloped towards midships. T h e waist is undecked but is
covered by a penthouse roofin g o f palm-leaves overlaid by
closely set palm yra-palm reepers or battens tied down with
coir. T h e after decking is the longer; on it is a small
cooking galley or rather firebox and several water breakers
find accom m odation. A t the centre is a small hatchway.
Th ere is no poop.
This very same arrangem ent is found in the ja th r a dhoni (to

which I refer below) the river/canal paruwa (the largest o f the
inland watercraft) and the T u ticorin coasters. In the latter
tarpaulin sheets wrapped round the deck cargo have replaced
the roofed penthouse. It will thus be seen that, notwithstanding
the similarity to European craft, the interior was that o f a
typical Asian cargo ship (fig. 5 -5 )- The rig o f the Jaffna thonis,
however, is strictly nineteenth century British but with many
m ore spritsails. A ccording to records the last o f these sailed in
the 1930s and we cannot hope to see them again. H o rn ell’ s
photographs are therefore invaluable. He has also given typical
Pre- M odern S ri L a n k a n S h ip s 157
dim ensions: length between perpendiculars, IOO ft.; beam

amidships, 21 ft. 2 ins.; depth from gunwale to keel, 14 ft.,
carrying capacity, IOO tons. T here is also a bad newspaper
photograph o f what is claimed to be the last one built and
sailed. Nam ed the 'A nnapooran i’ , she is supposed to have
Figure 5*5 Jaffna Thoni, views of bow from midships and views of
cargo stowage arrangements (Hornell)
158 So m a siri D evendra
sailed for, and reached, England. H er whereabouts after that is

not known. T h e photograph shows her in the Suez canal.
T h ejath ra dhoni
Perhaps the most ancient type o f indigenous sailing craft that

we were able to research and reconstruct was the fo re-an d -aft
rigged, outrigger-equippedjat/ira dhoni o f Dodanduwa, a village
about ten miles from Galle towards C olom bo (fig. 5-6)- T ill
the 1930s, this was a port where these ships operated from ,
sailing largely to coastal ports o f India and Sri Lanka, the
Maldives and South East Asia. T h e find was, in every sense o f
the w ord, serendipitous. Its existence was known, there was a
published note by an eyewitness to the sailing o f the last o f its
kind, but no research published in English, other than that
and H o rn e ll’ s. Paris (1840?), has produced a much superior
description and drawings than H orn ell and added that this craft
was com m on to both Sri Lanka and the C orom andel coast o f
India: in fact, to judge from the rounded penthouse r o o f he
Figure 5«6 Yathra Dhoni Model from Dodanduwa

shows (quite unlike the typical Sri Lankan ridged ro o f) the

craft illustrated could have been an Indian one. A ll other
details, however, tally. H orn ell ( l 943 ) had noted them but
had, by some strange chance, not seen them. He says the last o f
them were seen in 1903 and 1908, adding:
T h e ir survival, or rather their presence on the Ceylon

coast until recent years, is o f great ethnological interest in
view o f the representation o f ships related to the outrigger
design among the sculptures on the great Buddhist shrine
o f Boro Budur in Java, dating back to between AD 75 ° and
900 .
There is a published photograph o f one, (fig. 5 -7) referred to
as a 'Calpentyn C oaster’ , where the author refers to it as a
surviving type. In the Mahinda College Magazine o f June, 1936, a
Senior Form er named N .A .W . A rthur Alwis, describes the final
voyage o f the last o f these ships. He says:
. . . it seems as if it was yesterday that I saw the last

schooner o f Dodanduwa lying in the harbour. However
the ship . . . left her native harbour for Hambantota never
to return. That was in the year O ne-thousand n ine-
hundred and thirty.
He goes on to relate the story o f the wrecking o f the craft o ff

the Maldives and the rescue and return o f the crew years later.
It is indeed puzzling why H ornell, who served in Ceylon,
missed seeing these ships, though he has somehow produced a
fine line drawing showing the craft in profile. He even says that
the very m emory o f these ships is passing away. My own
experience was completely the opposite. N ot only had I heard
o f them and knew o f much written and oral evidence o f details
in Sinhalese, but also I heard o f a very faithful large m odel o f
one in a Buddhist temple at Dodanduwa. T he senior m onk
came from a family which had owned these vessels and is a man
with considerable interest in ethnography, oral history, and
archival matters. He had preserved the m odel in a glass case, as
it had been built by his father, while yet a child, around 1892
and was awarded a gold-m edal at a Craft Exhibition. He
i6o SOMASIRI D E V E N D R A
F igu re 5-7 Yathra D hon i in port (Lewis)
allowed me to photograph the m odel and take it out o f its case

to make notes. However, I was later able to persuade him to
lend it to the National Museums for an Exhibition held in
connection with the U N E S C O Expedition o f the Silk Routes o f
the Sea in 1990.
In 1992 > while co-ordinating a jo in t Sri Lankan-Australian
M aritim e Archaeology Project in Galle H arbour I found the
m odel yet with the Galle M aritim e Museum. We had with us
Tom Vosm er o f the Western Australian Maritime Museum, who
used his skills to make the technical drawings o f the m odel. He
measured the m odel meticulously as essential raw data. T he
data was then tested on a M acintosh com puter, which was able
to generate an acceptable hull shape based on the data given in
the form o f a table o f offsets. T h e com puter-generated p re-fit
lines were worked on to manipulate the hull shape to the best
fit o f the original measurements. This was done by using a
'M acS u rf com puter program m e for boat design and analysis.
He was thus able to take the lines off, make detailed technical
drawings and arrive at the sailing characteristics o f the craft
Pre- M odern S ri L a n k a n S h ip s 161
(fig. 5 -8 )- My description o f this craft, therefore, is not

H o rn e ll’ s but Tom V osm er’ s ( l 9 9 3 : 109—Il8).
T h e m odel examined and recorded appears to exhibit a

hybrid o f influences including, Arabian, Indian, local Sri
Lankan traditions as well as Southeast and East Asian. As
the m odel had been built by a boat-builder, it exhibited
hallmarks o f his care. For example, it was noted that the
four hooked scarf joints in the keel-stem and stern-post
were made exactly as they would have been on the real
vessel. O th er details were also done with attention to
detail . . . In view o f this attention to detail, it was thought
the accuracy o f the m odel, both in scale and detail would
make a fairly reliable source for docum entation, (p. I ll)
T h e size o f the m odel studied is 150 cm long, 32 cm at the

point o f its greatest beam and 2,0 cm from keel to gunwale.
T h e ja th r a are large outrigger craft, ranging to IOO ft (30m)

in length but normally about 50—60 ft (15—18.3 m),
carrying 25—75 tonnes o f cargo usually averaging o f 50
tons (H ornell 1943)- Mukherjee mentions ja th r a dhonis as
being about 60 ft. (18.3 ni) in length with a beam o f 15 ft
(4.6 m ). They are sewn craft, planked from domba
(Callophyllum inophyllum), at least two inches thick (Vitharana,
199^: 69)- In recent times (the 50—IOO years to 19 3 °) the
ja th r a dhoni was used as a coastal trader and for voyages to
India and the Maldives. However the type appears to be o f
ancient lineage, with Pliny ( a d 23—7 9 ) reporting 'outrigger
craft o f large size west o f Taprobane [Sri Lanka]’ . . . (p. Ill)
T h e yathra can be described as double ended, with slack

bilges but full m id-sections. T h e forward sections are only
just slightly m ore fine than the aft sections, displaying a
subtle hollow entry at the bows. T h e forefoot is extended
forward by a gripe attached to the keel-stem and there is
also a skeg aft to which the rudder is fitted. Both these
devices would be aids to lateral stability, helping to reduce
leeway and balancing the helm while sailing. It should be
sr~-
construction m om
—jr ,
Sir
YATRA DHONI MODEL

QEMRAL •mUMOCMENT
m*
COWtTWUCnOW PLOW
OCWUUL OHMNOI W WT
F igu re 5*8 Yathra D honi Structural details (Vosmer)

Pre- M odern S ri L a n k a n S h i p s 163
noted that at least one drawing o f a ja th r a (H ornell, 1943 :

44) does not show these additions. T h e form o f the hull
shows affinities with ancient Arab and Indian craft
illustrated by Paris (1841) (p* 112).
A m on g a host o f features described by Vosm er the follow ing are

particularly relevant. T he use o f an outrigger, he comments, as
'curious’ as the hull rig configuration is stable, with hydrostatic
analysis showing the hull form to be seaworthy even without an
outrigger. He describes the rig as com m on to the Indian
subcontinent, it being ketch-rigged with square-headed lug
sails and a jib set on a short bowsprit. He considers the rudder
as 'enorm ous by most standards’ and rem iniscent o f some on
Chinese Junks and draws attention to a much narrower one
shown in H o rn e ll’ s sketch. C oncern ing the cabin he notes that
it is o f a style com m on to Indonesian water. In Sri Lankan,
cabins are known to be roofed with either 'cadjan’ or split
bam boo: Vosm er remarks the same practice prevailing in
Indonesian waters. He also comments on the hatches, with
slightly raised coamings along their sides, being located on the
starboard side. T he cargo handling gear, however, is clearly
located on the port side.
I have referred to the comments on the Cabin and Hatches
to show the similarity to the Jaffna thonis and river/canal paruwas:
all three craft are large cargo carriers and the stowage
arrangements are similar. The references to the rig, rudder
and outrigger need m ore com m ent. In my searches I came
upon two m o v e ja t h r a models. T he first one is not traceable, but
a photograph is available in the L ondon Science Museum. It
shows a w ell-constructed model, which conform s to the m odel
we studied. T h e other, which I examined at the Museum o f
M ankind ten years ago, is much m ore interesting and revealing.
It is described as a 'Cingalee m odel boat "Yathrawa” or
"D honey” , freight and passenger boat used between India and
C eylon’ . It had originally been presented to a Museum in 1854*
(The Museum staff was unaware to which museum it had been
originally been gifted, or where it is now, as artifacts have, over
the years, changed hands between museums.) It is a smaller
164 Som asiri D evendra
craft, d o u b le-en ded , showing sewing on the outside, with a

single square sail amidships and no rudder. It is undecked and
has two holes on the ship’ s side, aft, presumably for steering
oars. It had a single outrigger, slightly abaft o f midships on the
starboard side, (unlike the Dodanduwa m odel, where it is on
the port side) fitted in the same way as in the m odel we studied.
Th e m odel is damaged and has a loose bundle o f what looks like
split-bam boo matting: probably the remains o f the cabin roof.
T his m odel is interesting as its rig is different, there is no
rudder and the vessel it is based on probably did not tack but
went about by changing ends, as the outrigger fishing craft still
do, or by m anipulation o f the square sail. T h e outrigger is
firm ly fitted, allowing o f no change; it is, as m entioned, on the
starboard side. It is suggested that this m odel shows an earlier
configuration o f the ja th r a , which later developed further with
the introduction o f m ore sophisticated rigging and a rudder.
T h e older version probably survived into the nineteenth
century as a smaller craft.
I do not intend going into the details o f the analysis o f the
data that was carried out but will end this section with V osm er’ s
concluding paragraph:
Hydrostatic analysis
In view o f the foregoing, a vessel o f about 2 0 m in length
was the size chosen for detailed analysis. T he hull form o f
the ja th r a , with and without the outrigger, was analysed for
stability, displacement, wetted surface, drag and powering
requirem ents as well as cargo capacity for that size o f
vessel. The ja th r a hull alone was found to be reasonably
stable as indicated by the transverse GM and RM figures
(Garrett, 1987)- T h e addition o f the outrigger, however,
increased the righting m om ent (RM) by a factor o f
approximately IOO. It also, o f course, added to the drag
created, and therefore powering requirem ents o f the
vessel, (p. Il8)
Pre- M odern S r i La n k an S h ips 165
The hattal
Th e battal is included in the list o f Sri Lankan ocean-going

vessels, not because they were ocean going but to illustrate the
multitude o f influences that touched Sri Lankan shipping. T he
craft is essentially o f Indo-A rab form , about 50—60 ft. long,
com pletely undecked, with a single forward raking mast
carrying a huge sail hoisted by means o f a sheave atop the
mast. It was in use among the M uslim villagers o f M utur, in
K oddiyar Bay, just south o f Trincom alee harbour. T ill the
m id-1970s they used to regularly ferry the rice harvest from
M utur to Trincom alee and were seen daily (fig. 5 *9 ) * By 1988,
however, they had disappeared. T h e area o f M utur, has been
noted as 'M atura’ in A hm ed bin Majid al N ajd i’ s treatise
(trans. by Tibbets 1977)- In Trincom alee harbour is a sheltered
cove named N icholson’ s cove’ since British times, which is the
site o f an Arab settlement about eight centuries ago: I had the
good fortune to discover the site o f its burial ground and
rescue some gravestones (Devendra: 1970 and 1990 a :
Figure 5*9 Battal (Mutur, Trincomalee)

i66 So m a siri D evendra
2 I 2 —2 I 3 )- U nfortunately, I was too late on the field to study

the craft and that part o f the country is yet not safe for
researchers. L ionel Casson (1964= 160) has a picture o f an
Arab bum , which is remarkably similar to the battal. To make
matters m ore com plicated, that word itself derives from the
Portuguese batel, m eaning 'small boat’ and is used in various
form s all over the Indian O cean.
The vattai/vattal
T h e vattai or vattal is the Jaffna version o f the Battal, as the name

implies. These broad and beamy boats have the distinctive surul
decoration and were generally painted a bright blue or green.
A lthough the sail itself is smaller than the M utur battal, the mast
arrangem ent is similar. H orn ell ( l 943 )> w^ ° had seen many o f
this type in Jaffna, says:
T h e single mast is stepped vertically, a little forw ard o f

m idships. T h e rig is a short, broad lug, little rem oved
from the original square sail. A rectangular two-sheave
block is fitted horizon tally on the masthead. T h ro u g h
one sheave hole runs the m ain halyard, m an -han dled
without the assistance o f any tackle; through the other
peak is rove, run n in g direct to the other end o f the yard,
where the end o f a vang is also attached. . . . T h is two-
sheave pulley truck is a notable characteristic o f these
small dhonis.
W henever a pearl fishery is held in Ceylon waters, these
local craft furnish the m ajor part o f the diving fleet; the
divers like them best o f all the craft available; their broad
beam offers com fortable accom m odation, the low free
board facilitates diving operations and their light draft
makes it easy to row them when the divers change their
pitch in their search for a rich patch o f oysters.
A very well-made model is available at the London Science

Museum, where its catalogue description is Jaffna D honi (c, 1800).
Rigged m odel.’ (fig. 5 -IC>) Edye (1833) was not very impressed with
the craft although he gives a very good drawing o f one.
C o n c lu sio n
Sri Lanka had its own indigenous ships. These underwent

m orphological changes from time to time through adoption o f
elements considered m ore appropriate, but the basic elements
did not change. H ence, we can safely accept that the
developm ent was, by and large, linear, but that the process o f
developm ent was considerably influenced by selective regional
technologies. Chinese technology, for instance, had left no
im print here.
Seafaring was an ancient occupation and, although the
ability to sail long distances was there, Sri Lanka could lay no
Figure 5 -10 Vattai/Vattal (Jaffna)

i68 Som asiri D evendra
claim to have been a m ajor m aritime power. Undoubtedly, the

nautical tradition flowered and fa d e d , with the highest levels o f
sophistication being reached in the twelfth to fourteenth
centuries. Ironically, other Indian O cean maritime nations
were also experiencing the same phenom enon. In Sri Lanka,
however, the rapid collapse o f central power around the end o f
this period led to the maritime provinces com ing under the
control o f foreign merchants and m ariners. T h e lin k with
seafaring became tenuous, even though the knowledge was not
lost. T h e 'h igh ’ technology had been reduced to a 'fo lk ’
technology by the time the European powers established their
power in the Indian O cean. T h e point is succinctly set out by
Gunawardana: (1986—87: 4 ° )
Thus the traditions o f shipbuilding o f an era long past did

survive right into m odern times, but in a vitiated and
dim inished form . It is abundantly clear that an involution
had set in, pushing back the level o f nautical technology in
Sri Lanka to what it had been long before the eighth.
T h e country’ s greatest claim to navigational history is that it

still retains elements o f ship design from all over the Indian
O cean — and even beyond — which may have disappeared from
the countries o f their origin. Sri Lanka is thus a palimpsest:
layer upon layer o f developm ent and adoption can be traced
and it is, indeed here, that the search for the history o f the
ocean could be best begun. I personally believe that the ja th r a
dhoni o f Dodanduwa is the oldest o f the p re-m od ern craft that
survived into this century.
T h e question then arises whether this claim to significance
resulted from an exchange o f technologies in a scenario where
Sri Lanka played a pivotal role. It is suggested that its role was
pivotal in a purely geographical sense only. Gunawardana
(1986—87: 3 2 ) citing Cosmas, states about the sixth century:
'. . . the growth in im portance the island had achieved as a pivot
in a vast network o f trade routes . . .’ . But there is no evidence
that it played a m ore positive, or catalytic role. W hile its
position was significant in bringing together ships from both
the eastern and western extremities o f the ocean, similar scenes
would have been played out in any o f the major ports o f the
ocean.
T h e search for hard evidence o f the form and structure o f
the most advanced craft built in Sri Lanka is far from over.
How is it possible to reconstruct the ships described by a ninth
century Chinese as 'the ships from the L ion K ingdom were the
largest, with stairways for loading and unloading which are
several tens o f feet in height’ ? O r overlook the possibility that
the sewn-plank khun lun ships that visited China from the south
— so large that they could not be set in m otion by oarsmen
without the aid o f sails — were also from the L ion K in gdom ?
But, in spite o f the details given, one cannot accept these
accounts without the back up o f hard evidence. Ibn Batuta, in
the fourteenth century, found in Calicut, India, Chinese ships
with four decks, with massive oars m anned by ten to thirty each,
to row the ships when the wind failed. T he similarity o f Ibn
Batuta’ s observations to the two earlier quoted Chinese ones
may not be coincidental, but a com m on way o f describing
unfam iliar craft. In terms o f my own search techniques I have,
regretfully, to consider these descriptions unproven.
O n my own voyage o f discovery, my ship has not come
hom e. She has only cleared the bar and the voyage has just
begun.
A p p en d ix
P r e - m o d e r n S ri L a n k a n w atercraft
A tentative typology
Type Common local names

(S: Sinhala. T: Tamil)
1 In la n d watercraft
1.1 R iver and C anal C raft
1.1.1 Rafts o f log or bam boo. Pahura (S)

1.1.2 *L ogb oats without outrigger. Viliam oru (S& T )
1.1.3 Logboat with sewn gunwales and outriggers. Pila oru (S), Kula (T )
1.1.4 D ouble-h ulled logboats, connected by Angula (S)
platform deck, awnings and railings optional.

1.1 .5 *D o u ble-h u lled logboats with vertically 9
extended washstrakes, connected by platform
deck, with h alf-rou n d thatched canopy and
provision for sail.
1.1.6 *D o u b le-h u lled logboats (with/without ?Angula (S)
washstrakes) connected by platform deck, with
enclosed cabin.
1.1 .7 *D o u b le-h u lled logboat, broad beam ed, flat Parwa (S)
bottom ed, scow -ended, thatch ro o f at waist,
provision for single/double sail.
1.1.8 As 1.1.7, but without thatched roof. Vali Paruwa (S)
1.1 .9 *A s 1.1.7, but decked at waist and covered with ? Paruwa prototype
h alf-rou n d canopy and storage space under
deck.
1.1.10 *D o u b le-h u lled logboat, without washstrakes ?
with platform deck form ing floor o f a village-
type, full-sized thatch hut.
i.2 Lake and Lagoon C raft
1.2 .1 As at 1.1 .3 above.

1.2 . 2 *A s at I. I . 7 .
1. 2-3 Plank-hulled boat with ribs, skeleton built,
narrow and resem bling a longboat or canoe
with detachable outrigger and sail.
2 M arine C raft
2.1 I n s h o r e Fi s h i n g C raft
2.1.1 Logboat with vertically extended sewn plank Oruwa (S)

washstrakes and gunwales, connected by two
boom s to outrigger.
2.1.2 Large d o u b le -h u lle d , siene fishing boat, Madel Paruwa (S)
sim ilar to 1.1.8 but with sheer at ends, o f sewn Pattai (T )
plank, no superstructure or deck. No
outrigger.
2.1.3 As at 2.1.2 but sm aller, and fitted with
outrigger. O perated in pairs with outrigger to
port on one and to starboard on the other.
2. 2 O ffshore Fi s h i n g C raft
2.2.1 As at 2 .1.1 but larger, doub le-en ded with lee- Bala Oruwa/Varakan
boards and spritsail/lugsail. Oruwa (S)
2.2 .2 As at 2 .1.1, but sm aller and with A rab -In dian Oruwa (S)
lateen sail.
Pre- M odern S ri L a n k a n S h i p s

2.2*3 As at 2-1.1, double-ended with reversible sail. Oruwa (S)
2 . 2-4 L ash ed-log raft, with 3-5 shaped logs with Kattumaram (S& T )
outrigger (optional) and ability to sail with
bam boo mast.
2 . 2*5 Sim ilar to 2 -2-4 but sm aller, o f 4 logs pegged
Teppam (S& T)
together with stakes ru nn in g through the logs
with p rotru din g ends lashed athwartships.
2-2-6 As at 1.1.3 but much larger and beam ier, with Padavu (T )
m inim al washstrakes, no outrigger, used for
beach seining.
2. 3 Seag oin g / C o a s tin g G raft

2-3-1 *Carvel built hull o f shell construction o f sewn Yathra Dhoni/Yarthra
plank, with keel plank. One/two masts and bow Oruwa/Maha Oruwa (S)
sprit, rudder and tiller. Partially decked at bow and
stern. C adjan-roofed waist with cargo hatches, and
equipped with cargo-handling boom . Large,
single outrigger on two sturdy boom s Arab-Indian
fore and aft rigging with spritsail.
2.3 .2 *A s at 2 .3 .I but undecked but without rudder 2 -3-3
and tiller. Single mast with square sail.
*H u ll sim ilar to 2-3«l. Decked fore and aft with Varakan oruwa (S)
form sim ilar arrangem ents for cargo stowage
m idships. R udder and tiller. Sem i-circular
platform , extending outboard, for steersm an.
Single mast. N o outrigger.
2 -3-4 *L arg e schooner with two tall masts, well- Dhoni/Thoni/Padagu (T )
developed bowsprit carrying several spritsails.
Partially decked at bow and stern. Gadjan
roofed-over waist with cargo hatches. Sewn-
plank construction. Distinctive coiled term
post decoration ('S u ru l’), occuli, painted false
gu n -p orts.
2-3-5 *C arvel built hull o f shell construction, low Vattai/Vattal (T )
freeboard with convex stem and stern. Flush
decked with cargo hatches (optonally
undecked?). Single mast with lug-sail, rudder
and tiller. S u n il’ and occuli. Painted hull.
2-3-6 *C arvel built hull o f shell construction. Battal (S), Vattai (T )
U ndecked. R udder and tiller. Very large single
sail on forw ard-raking mast with block for
hoisting sail.
Note: * indicates obsolete types

172 So m a siri D evendra
Notes
1 In extrem e cases, the entrance w ould be com pletely closed up, leadin g to
the river waters backing up and fo rm in g lakes and lagoon s. T h e coastline
o f S ri Lanka is dotted with these.
2 It is not possible to be defin ite when, o r along what route, the south -east
A sian outrigger ships reached M adagascar, although this occu rred at
som e stage in the first m illen ium A D .
3 T h ese referen ces are w ritten ones but, fo r this p aper, these (whether
an cien t o r m o d e rn , prim ary o r secon dary) have served only as
c o rro b o ra tin g evidence, and what follow s, therefore, is only a sm all
sam p lin g culled fro m referen ces given by w riters who have dealt with, in
the m ain , trade and com m erce:
China (G unaw ardana & S a k u r a il9 9 0 : 277“ 2 8 l), (P ao -ch an g 1981:

6 8 - 7 0 , 89)
North India (N icholas 199O: 281)
South India (N icholas 1 9 9 0 : 2 8 4 —85)
45~ 48 )
Burma, (G eiger 1 9 2 I : 76, I I - 3 5 ,
Sri Vijaya (K irib am u n e 199° : 186)
Egypt, (G o iten 1973: * 8 5 ) ’ (C o d rin g to n 1919: 83)
Rome (M alalasekera 1935: ^ 3 ° ) » (M cC rind le 1 9 0 I : 103, 105)
Persia; (Im am 199O: 173)
Thailand (G u ru ge 199° : 2 4 6 )
Trading as an occupation: (G unaw ardana 199° : 33 ’ qu o tin g fro m Sahassa-
vattupakarana)
4 P rof. Su darsh an Seneviratna, com m en ted in a perso n al com m un ication ,

in 1987-.
T h e sym bol depictin g a vessel carrying a nandipada (taurin e) symbol

at the helm o f its mast is engraved alo n g with an early B rah m i
in scrip tio n at Duvegala in the Polonnaruw a D istrict. T h e in sc rip
tion records an endow m ent m ade by a B arata to the Sangha.
T h e nandipada sym bol is com m only fo u n d on the Early H isto ric
coins o f South central A sia. T h e vessel sym bol (without the
nandipada) occurs on the Proto H isto ric Black and R ed Ware o f
P en in sular In dia as a p o st-firin g graffti m ark. D u rin g the Early
H isto ric p e rio d a sim ilar sym bol is fo u n d on the coins o f the early
Pandyas and the A n d h ra Satavahanas.
T h is type o f vessel was used in the deltaic region o f Pen in sular
In dia an d it sailed alo n g coastal sea-ro u tes lin kin g Pen in sular In dia
with S ri Lanka. It is known as the N egapattanam seven log — type
b oat and un til recent tim es such vessels sailed between N egapatta
n am (in South India) and the north-w est coast o f Sri Lanka.
T h e Baratas (Paratavar) are known to the early in scription s and

the texts as m erch an t-m arin ers, who con ducted a lucrative luxury
trade lin king coastal areas o f P en in sular India and S ri Lanka. T h e ir
trade item s in clu ded h orses, gem s, salt, pearls, chank, spices and
paddy. T h e Baratas were ship-ow ners and they are known to have
con trolled port-tow ns d u rin g the Early H isto ric p e rio d .
R eferences
A lthough the Sri Lanka and the Silk Road o f the Sea is shown as the source o f many
papers below, it is an om n ib us volum e which b rough t together a large
n u m b er o f papers that had b een earlier pub lish ed elsewhere but the
origin al sources were not readily available. However, m uch o f the volum e
consists o f origin al con tribu tion s.
6
C la s s ify in g V e sse l-T y p e s
i n Ib n B a t t u t a ’ s R i h l a
Dionisius Albertus Agius
6j J-L-u <juuu (j-Aj 6j LuJuj I jic. ^ ajuluj L a L a ^juIj
I saw both [the sultan and the vizier] one day on the Tigris
in a harraqa (which they call shabbara, and it is a sallura) . . .
Ibn Battuta(II: 1 1 6 /3 3 6 —7 )
Th e R ihla1 o f Ibn Battuta (d. , 2 the most notable

travel book ever written in Arabic, rich in its descriptions o f the
great cities o f Islam, public m onum ents, pious institutions and
personalities, written in the conventional literary style o f the
time with an aim to inform and entertain readers, is the record
o f a dramatic personal adventure by land and sea which
provides us with a rich source o f geography, ethnography and
cultural history.
Ibn Battuta returned to M orocco (Tangier being his birth
place) in 1355 after some 3 ° years o f travel, going beyond the
M editerranean and crossing the territories o f the M iddle East
into the Indian O cean as far as China, covering an area o f
about 7 3 ,0 0 0 miles. We do not know the reason why he did not
set out to write his m emoirs on his return, what we do know,
however, is that in the same year he returned to M orocco, the
sultan A bu Inan Faris ( 745 —759 //i34'8 —1359 )> the M arlnid
ruler o f M orocco, who had heard about his extraordinary
experiences as well as his descriptions o f the W orld o f Islam
174
C la s s ify in g V e ssel-T y p e s in Ibn B a ttu ta ’s Rihla 175
(Dar al-Islam ), approached a young literary Andalusian scholar

M uhammad Ibn Juzayy al-K alb l (d.c. 75 &/I357 ) to down
in writing the memoirs o f Ibn Battuta. N othing is known about
the place where these two men met for work sessions, nor do we
know whether Ibn Battuta ever read the com pleted manuscript
and checked it for chronology and erro rs .3
In the writing o f the book, a few additions were inserted here
and there, some were from Ibn Juzayy’ s com position, others
came directly from the work o f Ibn Jubayr the twelfth-century
Andalusian traveller .4 In general, it needs to be said that Ibn
Juzayy relied on the personal interviews he had with Ibn Battuta
and the notes the latter left with him . But these notes must have
been the second set Ibn Battuta wrote from m emory as he
claimed to have lost the original diaries in the course o f his
travels. How much o f his m emory was reliable is an im portant
issue in the study o f Ibn Battuta’ s work, the chronology, for
example, o f his travels is flawed which does suggest a lapse o f
memory. In essence, the Rihla, as Ross D unn rightly remarks
(1986: x) is in itself, 'not a daily diary or a collection o f notes
but a work o f literature, part autobiographical and part
descriptive com pendium written at the end o f his career’ .
Here, 'work o f literature’ is taken to be what in Arabic is called
adab, giving im plicit attention to language and its delivery in an
artistic form at (Netton 199^: 103—4)-
T h e Rihla does not provide inform ation on fourteenth-
century shipbuilding and seamanship. This does not come to us
as a surprise, for often Muslim travellers were indifferent to
recording nautical term inology or ship-types as such. O n e o f
the main problem s com m on among classical A rabic writers is
their lack o f specific inform ation about material culture, let
alone descriptions o f ships or boats. Ibn Battuta’ s notes are
lively with comments on the characteristics o f port towns and
the people inhabiting them. His occasional observation on a
type o f ship is o f interest and useful but often puzzling.
T h e lack o f w ritten manuals on types o f ships and
shipbuilding techniques in both Arabic and Persian literature
is frustrating for the student o f maritime history. We can only
turn to M editerran ean manuals on m arine design and
176 D io n is iu s A lb e r tu s A g iu s
dimensions, inform ation from which could help to fill that

knowledge gap on contem porary designs (D otson 1994 :
16 0 -1). However, transferring that technical knowledge could
in my experience be m isleading because if we look at m odern
times rarely do shipwrights in the G u lf and the Arabian Sea
work from plans. Traditionally, dhow -builders work by the
accuracy o f the eye and the art is passed on from father to son.
(see Beckerleg, this volum e). T h e master builder squats and
draws his design on the sand and having erased the plan, works
purely from memory. O f course, the Arab or Iranian builders
over the centuries would not have deviated much from the
conventions but some changes in design did occur according to
the needs o f the users and progressive adaptations took place
through newer technology. M oreover, due to salinity and ship
worm, there are no wrecks in the G u lf and the A rabian Sea to
provide us with in form ation about design and b uildin g
techniques. Reconstructing, therefore, the type o f an Arab or
Indian vessel is difficult and we can only infer inform ation
about boat typology and the nom enclature o f such types from
geographical, travel and historical sources. O n e fact which
needs m entioning is the dearth o f sources on M uslim maritime
traffic in the Red Sea and the Indian O cean in general (let
alone during the time o f our present traveller). T h e occasional
references we have are very difficult to interpret in any
systematic way, Ibn Battuta’ s work being a typical case. Any
attempt, therefore, to reconstruct medieval vessels m entioned
by Ibn Battuta will to some extent rem ain hypothetical.
This chapter is mainly concerned with the classification and
names o f ship-types which Ibn Battuta travelled on or observed
during two o f his long journeys: the first down both shores o f
the Red Sea to Yem en, then from A den to the trading ports on
the coast o f East A frica, and along the southern coast o f Arabia
(Yem en and Hadhramaut) to D hofar in O m an and the G ulf;
the second adventurous jo u rn ey took him from central India
down the Malabar coast to Calicut, then to the Maldive islands
and C eylon (Sri Lanka) (see map). In the course o f these two
journeys he m entions 20 vessel-types (see table) which I list
here according to the function they perform ed: (a) generic
C l a s s i f y i n g V e s s e l- T y p e s in Ib n B a t t u t a ’ s R ih l a 177
SYRIA
,RAQ U b u lL
B«tr< F A R S
S I ND
vO alh at >
OMAN TSur C am bay-
Je d d a h
'A > d h i b i 7 U m i . ^ H c*J v X
I
S w ik in *
Arabian
YEMEN
H o n iv a r f ^
<i ’ Socotro IdanJ
M t n p l orr^fr
Zeila" I C alicu t V *
, ( J u ilo ifv
MaW.w
/M o p d u h u IJanJt
Kilw* I N D I A N
O C E A N
Ib n B aiiu U t voyage
F igu re 6.1 Ibn Battuta’ s voyage

£
o
C
Tear recorded Source Description Resembling type & region
vS
■§o
o
1
0
O
CD
ahawra [Indian O cean] river bo at tarida (M e d ite rra n e a n )
CO
J>-
CO
CO
£
m
O
O
CT oo
galley ukayri (Indian O cean)
e*
ghurab [M e d ite rra n e a n ]
CT
CO
CO
CM
\
£
O
O
CD
CD
in
harbi [M e d ite rra n e a n ] war vessel (g e n e ric )
OT
CO
CO
VO
OT
CD
CD
1
1
shabbara (Iraq); sallu ra
or
harraqa (pi. h arraqat, h a r a n q ) la u n c h
!>•
CO
CO
^CO
[M e d ite rra n e a n ] (M e d ite rra n e a n )
£
o
00
Of
w a rsh ip
to
jafn [M e d ite rra n e a n ]
CT)
CO
£
co
CD
OO
~co to
jafn [M e d ite rra n e a n ] w arsh ip
CO
or
or
1
jafn [M e d ite rra n e a n ]
oo
0 CD
CD
cv
jalba (pi. ja lb a t, jilab) [Red Sea] coastal sailing vessel
CO
£
£
U
a
u
Of
CD
C
>
or
kakam & zaw (In d ia n
OJ
junk (pi. junuk) [Indian O cean]
<N
<u
'co
'cr>
CO in CT
bo
CO
<D
<D
CD 00
O CO
O cean)
00~
CO
o
00 >
C
c
V
in CT <T> OO in
N
CD
0T
£
co
1
o#
rM
sm all-sized Chinese vessel

’■5
C3
kakam [Indian O cean]
'co
1 CT
CO
CO
d
CO CD CD 1
OS
00
00
O
in
CT CO
£
csr oo CD o
Of
Of
CT 0 oo" 00
00
in
CT
1 00 CD
kundura (pi. kanadir) [Indian O c e a r l] 744/1343 small b o at q a rib
oo CO
<T>
00
CD
00
m
CO
CO
>
1
00
in
in
CD
O
00
ma'diyya (pi. ma'ad) [Indian O cean] wooden raft
i"'.
CO
in
00
r^ or
ma'diyya (pi. ma'ad) [Indian O cean] reed raft
CT
CO
CO
CO
CO
£
\
O
in
00
ma'diyya (pi. ma'ad) [Indian O cean] b a m b oo -ca n e raft
CT
-
r^
Of
in
00
00
markab (pi. m ara k ib ) coastal or o c e a n - g o in g
cf)
CO
1
Of
00
in
in
CD
00
vessel (general te rm )
rC
00 CO CO
0
OJ
00
CD
CD
cr
00
or
markab (pi. m ara k ib ) coastal or o c e a n - g o in g
CO
CO
CO
vessel (general te rm )
o
e
o
oo
Tear record ed D escription Resembling type & region
3
C
o
o
-
T
O
markab (pi. m a ra l III: coastal or o c e a n -g o in g
\O
CO
CO
CO
<T>
vessel (general term )
o
00
cr
or
or
ô
qarib (pi. qawarib, small boat (generic term )
OS
CD
CO
cr
qarib (pi. qawarib, small b o at
a
or
1
or
lau n ch
r^
shabbara (Ira q )
Cl or CD
sallura (pi. salalir [M editerran e
CD
CO
CO
oo oo CO
CO
\
C
or
or
1
lau n ch sallura (M e d ite rra n e a n )
CD
shabbara [Iraq]
CD
r-~
CO
\ \O
00
O
O
cr
in
i'- ô
'ukayri (Indian O ce an )
<u
shillir [M e d ite rrar life b o at
nOS
CO CO
£
Sh
00
CD
or
03
coastal b o a t
’ o"
3 h
1
$ s
3*
1
in
_Q JO
W
coastal b o a t
'OS 103
OS
CO
a '3 a
C fi 0
on
00
or O 00 oo
'03
2 3 3
or or
or cr
coastal b o a t
S3
03
iq) [A rab ian
cr cr cr
""e? <U
oS CD
\CO
CD CD
CO CO
\
"E h 3
M
&
£03
C
'03
2
cr
nd
cr
or
00 CD in
<L)
'"e?
ocean-going vessel
GO
cr
<U
CO
CO
CO CO CO
CO
\
C
o
T
T3
O
O
ahawra (Indian O ce an )
icS
war vessel
r^.
\D
tarida (pi. tarayid , ta r a ’id, tara
CD
CO
CO
CO
[M e d ite rran ea n ]
\
o
&
oo
T3
Cl
or
O
war vessel
i«3
'co
tarida (pi. tarayid , , ta r a ’id, tars
!>-
CO
[M e d ite rran ea n ]
00
in
'ukayri [Indian Oc galley ghurab (M e d ite rra n e a n )
00 \
03 r"fl
O 00
CD O
or cr
O
O or
'ukayri [Indian 0( life b o at shillir (M e d ite rra n e a n )
£
X,
Sh
sh
'03
io3
or
or
os
in
1
life b o at
3
0Jh
<u
<T)
3 h
CO
CO
CO
g
I?
d
<U aj £ £
sh
tS
03 •OS '03
'ushari (pi. 'ushari sk iff
cu
03
<u
OS
6 0 - 1/800
<u <u
£
00
Ol or
20 zaw [Indian O ce ar m edium -sized C h in ese junk & kakam (In d ia n
CD
'co CO 'or CO CO
CO CO CO CO CO
CO
vessel O ce an )
-2
CD
£
oJ
-c -
fi
in
in
CQ
CS
C*
<A
V
$■
>-
Oh
<D
i8o D io n is iu s A l b e r tu s A g iu s
terms - qarib, markab, harbi, (b) ocean-going vessels - ju n k , zaw

and kakam, (c) war vessels - ukayri, shillir, ghurab, ahawra, ja fn , (d)
launches - harraqa, shabbara, sallura, (e) coastal sailing and rowing
vessels - ja lb a, sunbuq, (f) coastal rowing boats - kundura, ushari
and (g) raft - ma'diyya. O u r concern is the reliability o f Ibn
Battuta’ s inform ation and how many o f the terms could be
identified as local. As pointed out earlier, m em ory was crucial
in Ibn Battuta’ s case and when it came to recording vessel-types
it could have in some instances failed him . This could have
resulted in a confusion o f a local with a M editerranean name
or application o f a name he was not sure.
T h e vessel-types in Ibn Battuta’ s Rihla will be studied in the
context o f the Red Sea, the Arabian/Persian G u lf and the
Indian O cean. Giving the historical background o f the coastal
towns that the traveller visited is essential before discussing the
nom enclature o f ship-types with their historical and linguistic
origins. I will then look into the fourteenth-century Perso-
Arab and Indian vessel from fragmentary evidence o f co n
tem porary or earlier travellers and see how m uch o f this is
relevant to the understanding o f boat typology.
T h e Setting
In none o f the hum an and physical geography works o f the

tenth century, such as those o f al-Istakhrl (fl. 340/95I—2), Ibn
Hawqal (d. 367/ 977 —8)> al-M uqaddasi (d. 378/988—9) and the
encyclopaedist and also geographer al-M as'udi (d. 345 /
9 5 6 -7 ), can we find detail about port towns in the Arabian/
Persian G u lf and the A rabian and Red Sea coasts. Invaluable is
the geographical dictionary o f Yaqut (d. 626/1228—9) in the
thirteenth century, an extensive traveller who, in addition to
the com pilation o f a lot o f inform ation which he extracted
from earlier geographers and writers, has included his own
travel experiences and observations made throughout the area
under discussion. T h e period o f the M ongol invasion in 1258
and the fall o f the 'Abbasid caliphate mark the beginning o f a
time when Arabic works o f travel and geography almost becom e
non-existent. It is in this respect that Ibn Battuta’ s Rihla o f the
C la s s ify in g V e ssel-T y p e s in Ibn B a t t u t a ’s Rihla 181
fourteenth century fills an im portant knowledge gap on the

maritime history o f the area. His work and that o f M arco Polo
(d. 1323) are valuable sources on distances and port towns,
some inform ation on which, though true o f the time, echoes
similar descriptions o f the earlier travellers.
S iraf on the Persian littoral during the ninth century was a
true com m ercial em porium where sea-captains and ship
owners could find a thriving market with rich financial rewards,
as the geographer al-Istakhri reports (187O: 131—^, 170), and
al-M uqaddasi writes about the 'charm o f its markets, the
affluence o f its population, the extent o f its reputation (18 77:
378). It was at that time the gateway to China, after O m an. 'In
summary’ , he continues, 'I have not seen in the realm o f Islam
m ore remarkable buildings . . .; they are built o f teakwood and
baked brick . . T h e town had been by Ibn Battuta’ s time in
decline. It was struck badly by an earthquake in 977 > never to
recover again. What is peculiar, however, is the confusion Ibn
Battuta makes o f Siraf with the island o f Qays (II: 244/407). It
looks as though he may never have visited Siraf after all, but
travelled to Qays which was obviously a rich trading port
abundant in water 'with w onderful gardens o f scented herbs
and leafy trees amongst its houses’ (II: 2 4 4 /4 °8 ) and probably,
like the early 'Abbasid days, the island was still crowded with
ships com ing from India and Arabia.
Further north on the way to Basra (Basra) was another
island, Kharik, o ff the m outh o f the Shapur river. Famous for
the best pearl fisheries in the neighbouring sea, K h arik seemed
to have been as comm ercially im portant as Qays (Yaqut
1866—73 > I: 395 » 503). In Ibn Battuta’ s time the channel
between Siraf and al-Bahrayn was excellent for pearl fishing
which attracted 'large numbers o f boats to this place with divers
and merchants o f Fars, al-Bahra(y)n and a l-Q u ta (y )f . . .’ (II:
244 —5/4 °^ ). Water was in abundance at the city o f al-
Bahrayn 5: 'one digs with o n e’ s hands [in the sand] and there it
is’ , reports Ibn Battuta (II: 246/409)* Several medieval sources
m ention water as an im portant com m odity not only for locals
to drink but also, in particular comm ercially, to export it to
places where it was not accessible or for transporting on ships
for sailors and passengers to drink on long-distance voyages,

some lasting up to twelve weeks (Marco Polo 1982: 263)-
Basra was established as a com m ercial centre in the early
centuries o f Islam, after U bulla, south-east o f the city had
fallen into decay. As shipping increased in Basra, al-M as'udl
(1861—77> I: 229—3 ° ) inform s us, watchtowers were co n
structed as a lookout for pirates com ing from al-Bahrayn and
the Persian coast (al-Istakhri 187O: 32~3; al-M uqaddasi 18 77:
17). The pearl fishing industry attracted ships with merchants
from India and the province o f Fars as much as it did during
Marco P o lo ’ s and Ibn Battuta’ s visit to the Arabian/ Persian
G u lf ,6 though none o f these towers were m entioned by them.
By the eleventh and twelfth centuries the eastern com m od
ities reached the markets o f the Near East via H orm uz, 'Adan
(Aden) and A ydhab on the Red Sea. O th er places grew from
small towns to im portant trading cities such as Tabriz, which
was well watered by num erous streams and surrounded by
beautiful orchards .7 N ot only was it linked to the Tigris and the
G u lf but it was connected with the port o f H orm uz which
captured most o f the trade o f the Indian O cean. H orm uz rose
to im portance in the thirteenth century; it became prosperous
in the fourteenth century when M editerranean traders who
paid high prices for their goods at the Egyptian and Syrian
markets explored the new route o f H orm uz. It was visited by
Marco Polo in 1294 an<i had already by this time becom e the
trade and navigation centre o f the Arabian/Persian G ulf. 'It is a
fine large city’ , reports Ibn Battuta, 'with m agnificent bazaars,
as it is the port o f India and Sind, from which the wares o f
India are exported to the two 'Iraqs, Fars and K hurasan’ (II:
23 0 /4 0 0 ). C hinese historians m entioned H u -lu -m u -ssu
(Hurm uz) in their annals as a rich city which sent embassies
to China (Huzayyin 1982: 182). Qalhat, which lies south o f
Muscat in O m an, was another im portant port town, second to
H orm uz. It was the entrepot o f the m onsoon trade much
frequented by m erchant ships on which, Marco Polo notes,
'spices and other goods are carried to various lands and cities’
and, 'many fine horses are exported from here to India’ (1982:
272). Ibn Battuta, some 50 years later, speaks o f its fine bazaars
C la s s ify in g V e ssel-T y p e s in Ibn B a t t u t a ’ s Rihla 183
and that the Qalhatis made their living out o f 'what [came] to
them on the Indian Sea’ (II: 225 /396 )- The ruins o f big
buildings on the seafront o f Qalhat, which have been identified
as the foundations o f merchant houses, and the num ber o f fine
pottery, glass and ceramic scattered all over the area, are
evidence o f a once rich, wealthy and thriving city.8
The port o f A den was a great resort o f merchants. It was the
port o f Indian and Egyptian merchants, Ibn Battuta tells us, 'to
which come great vessels from Kinbayat [Cambay], Tanah,
Kawlam [Q u ilo n ], Q aliqut [Calicut], Fandaraina, al-Shaliyat,
M anjarur [Mangalore], Fakanur, Hinawr [Honavar], Sindabur
[Goa], and other places’ (II: 177/ 37 ^)• A lot o f the merchants
that lived here were wealthy; some owned large ships. 'In this
p o rt’ , Marco Polo remarks, '[merchants] transfer their goods
to other small ships’ (1982: 27o), implying that coastal boats
were better designed to cope with the reefs and currents o f the
Red Sea and other small port towns east o f A den, on the
Arabian Sea. T h e city o f Zafar, which was part o f Y em en in Ibn
Battuta’ s time (now part o f the D hofar region o f Om an) was an
im portant harbour famous for the export o f horses to India
(II: 196/382), and Marco Polo, giving some details, notes:
'The city stands on the sea and has a very good port, frequented
by many m erchant ships that im port and export great quantities
o f m erchandise. Many good Arab steeds, and horses from other
lands as well, are brought here . . .’ (1982: 27 ^)- They were
exported to India making a good profit. T he im portance o f
Zafar, as a trading centre from antiquity is corroborated by the
findings o f al-Shahrl ( l 994 )> in a survey o f rock carvings he
undertook on the mountains north o f Salalah in the D hofar
region. T here, he discovered several drawings o f horses, camels
and ships with an indecipherable Semitic script marking the
trade o f the Jabbala tribe with the neighbouring countries,
India and elsewhere. A m on g other exports, frankincense,9
perhaps the most valuable o f all, was (and until fairly recently)
exported from this region in large quantities.
T he M editerranean merchants, through their trade links
with Yem en, eventually discovered the coast o f Gujarat and the
Malabar, which became the m eeting place o f all traders from
the east and west. T h e main ports o f Q u ilo n , Calicut and

M angalore in the south-west o f the province o f Malabar were
sought for the num ber o f quality pearls gathered there in the
months o f A p ril and May and in some areas in Septem ber and
O ctober. T h e pearl fishing industry attracted merchants from
China, the Arabian/Persian G ulf, Southern Arabia and,
in d eed , the M editerranean. T h e horses im ported from
H orm uz, Zafar and A den came in large numbers to these
ports. O ne o f the kings o f the province used to buy '2 0 0 0 o f
them and m ore every year, and his brothers as many’ , but, the
Italian traveller comments in amazement, only a hundred o f
them survived due to ill-treatm ent and lack o f veterinary care
(Marco Polo 1982: 227 , 2 3 °)-
The increase in sea traffic signified a prosperous econom y,
an im portant part o f which was the im porting o f Chinese and
Indian goods such as iron, weapons, musk, aloes, porcelain,
pepper, cam phor and tin. In the thirteenth century, the
M ongols followed by the Il-K h an ids o f Persia ( 654 ~ 754 /
1256—1353) gained control o f the main sea routes from the
Arabian/Persian G u lf to India and China creating a network o f
caravan and sea routes in the Indian O cean, perhaps never
witnessed before. By good fortune, Ibn Battuta enters this part
o f the world in which the revival o f trade under the M ongols
found Q uilon, Mangalore, Hormuz, Qalhat, Aden and Aydhab,
the main harbours linking the Mediterranean with China.
G en eral Term s for Seacraft: q a r ib and m a r k a b
In the Rihla, Ibn Battuta is inclined to use qarib for any small
boat and markab when referring generally to a coastal or ocean
going vessel.
The name qarib (pi. qawarib) is a com m on term, a word also
used nowadays (practically all over the Arab world) to denote 'a
small boat; a s k iff. In the Classical Arabic lexicon it is defined
as a ship’ s boat, used by seamen to help them in carrying out
their duties (Lane 1984* 2 5 °8 ). Qarib in our context appears
to be a canal or coastal boat (II: 23 l/ 4 0 0 )- W hen Ibn Battuta
reached the island o f H orm uz some time between 132 5 and
C la s s ify in g V e ssel-T yp e s in Ibn B a t t u t a ’ s Rihla 185
I 3 5 4 > he went to Jarawn the chief city, described as the port o f

India and Sind. This place had water-springs from which the
locals collected the water in waterskins (qirab) and loaded them
on qaribs, cargo boats, which they ferried across to the main
land. I found no other m ention o f qarib in this context in the
rest o f the text dealing with the Red Sea and the Indian Ocean.
In another place the word occurs as a dug-out canoe. He
reports that when his ship arrived in Sumatra at a place called
Barahnakar, the natives came to them in little qaribs, 'each o f
which was a single log hollowed out’ (IV: 226/875)* It is clear
from the text that he was not able to obtain the native term
because, he says, the natives 'have a strange language which
nobody understands unless he has lived among them ’ (ibid). As
was often the case, he does not provide us with inform ation as
to how long these craft were and, in this particular type, from
which tree the log was hollowed out.
T h ere is no reason to believe that qarib was a medieval lingua
franca term among Muslims in the east, though we find
m ention o f it much earlier by the geographer al-M uqaddasi in
connection with Iraq and the Red Sea (Agius 1997 : S1^)- It was
definitely a term used in the M editerranean world (Picard
1997: 253, 2 9 3 - 3 I 4 > 37 l). hence it explains Ibn Battuta’ s
fam iliarity with the word as it was known to him in his home
town Tangier. That the word is related to the Arab two-banked
galley o f the early centuries o f Islam (c. second/eighth century)
appears to be accidental. For trade in the seventh century, qarib
was the typical cargo boat from the Maghrib to al-Andalus and
from al-Andalus to the Levant. The Geniza documents refer to
it as a barge or a lifeboat in the regular maritime links between
Seville and Alexandria via T rip o li (G oitein 1967, L 21, 42,
211—3, 3 0 5 -6 ). It is generally placed etymologically with the
G reek K & pa po q through K a p a p i o v , a word that became established
in the Romance languages (Fraenkel 1962: 2l8) such as
M edieval Latin carabus, Spanish caraba, Portuguese caravo
(K inderm ann 1934 : 76 )*
Muslim travellers and geographers o f the fourteenth century
and earlier used markab (pi. marakib, from Arabic rakiba 'to travel,
board’) as a generic term for a Muslim or non-M uslim type o f
i86 D io n is iu s A l b e r tu s A g iu s
ship (not o f small size). Some, not Ibn Battuta, were m ore
specific if they were writing about 'ocean -goin g vessels’ a l-
marakib al-bahriyya, 'transport ships’ al-m arakib al-ham m ala l-ja fy ya,
and 'merchant ships’ marakib al-tujjar (Picard 1997 : 298—302).
In Ibn Battuta’ s case we find the term markab being frequently
used to indicate any sailing vessel o f different hull design and
size .10 Also, safina is a general term for a ship though less
com m on both in the M editerranean and the Indian O cean.
Ibn Battuta hardly uses the term and not in the area o f our
concern. Interestingly enough, safina, in spite o f its classical
( Q u r ’anic) usage, cropped up regularly in my interviews with
the seamen on the Arabian Sea coast in N ovem ber 1996 and
A p ril 1998, though it needs to be said that both in the written
sources and oral tradition, markab and safina never refer to a
small boat.
T h e C h in ese O c e a n -G o in g Vessels
T he only time that Ibn Battuta took an interest in describing a

ship was when he saw the Chinese vessels in the port o f Calicut,
one o f the chief ports in Malabar. There, in 1342 , the traveller
saw all types o f ships and was enraptured by the bustling port
where the Chinese, Javanese, Sinhalese, Maldivians, Yemenites
and Persians met. We are told that its harbour was the largest in
the world (IV: 89/812). The Chinese vessels stuck out because o f
their size. T h eir sails were never lowered. O f these vessels was the
larger type, the ju n k (pi. ju n uk) which carried about a thousand
men, six hundred o f these were seamen (a num ber o f whom
were merchants) and four hundred m en-at-arm s (archers,
cross-bowmen and naphta-throwers); the latter were needed to
protect the ship and the passengers from looters and pirates so
predom inant on the Indian Ocean. Marco Polo (1982: 210)
inform s us that the Chinese vessels were built o f spruce and fir
and, according to Ibn Battuta (IV: 92/813), constructed in
Zaytun (C h ’iian-chou) or Sin Kalan (Canton); each was
accom panied by three smaller craft, reports Marco Polo
(1982: 2Il) and some were manned by sixty, eighty or one
hundred seamen. The Italian traveller has no name for the
vessel-type while Ibn Battuta, apart from ju n k, mentions two

other types, the zaw, a m edium -sized vessel and the small kakam.
To build a Chinese vessel, Ibn Battuta says, 'two walls o f tim ber’ ,
were constructed and the space between was filled with large
baulks o f wood 'fastened both in length and breadth by huge
nails each o f which is three cubits in length’ (IV: 92/813). Marco
Polo mentions thirteen bulk-heads (i.e. partitions) 'made o f
stout planks dovetailed into one another’ , the reason being that
if the ship’ s hull is damaged after it struck a rock it would be
repaired easily (1982: 210). There were twenty oars and each one
was as large as a mast and was operated by ten to fifteen rowers,
standing in two rows opposite each other. They pulled alternately
on two strong ropes fastened to the gunwale, singing as they did
so (Kinderm ann 1934 : 2Il). The Chinese vessel had one
steering-oar, a detail not given by Ibn Battuta. Each ju n k had
twelve sails o f bamboo cane mat; Marco Polo says that one type
carried four masts. There were cabins, suites and salons for
merchants on each o f the four decks. Ibn Battuta gives details
about the suites: each had a set o f rooms and a latrine. It was
usually reserved for wives and slave-girls and it could be locked.
He adds that sailors had children living on board ship and sowed
greens and ginger in wooden pots (IV: 92/814)-
O n e report by A bu 1-M ahasin in 1432 inform s us that some
o f the Chinese ju n ks sailed as far as A den selling porcelain, silk,
musk etc. (K inderm ann 1934 : 2 l). T h e large num ber o f
fragments o f Chinese porcelain and ceramic scattered in a
num ber o f port towns in the Arabian/Persian G u lf,11 on the
Kenyan coast in East A frica (Kam oika and Yajima 1979 : 66—9),
as well as other parts o f the Red Sea (Rougeulle 1996-* I 7 0—I)
does suggest an active shipping on the part o f the Chinese
vessels in the area, particularly during the Sung (960—1279 )
and Y u an ( l 2 79 —1386) dynasties (Jun-yan 198O: 92—3 ).12
Th e earliest record o f the name ju n k appears in 1300 and a
m ention o f how many passengers travelled on this vessel comes
in 1331 in the m emoirs o f Friar O d o ric ' . . . now on board that
ship were good 7 ° 0 souls, what with sailors and with merchants
. . .’ (Cathay 1866, I: 73 )- The seamen at Malabar must have
adapted the Javanese name djong for ship, a term which Arabs
i88 D io n is iu s A lb er tu s A g iu s
and Persians were well fam iliar with; while the English version
o f ju n k ’ could have come from the Portuguese ju n co . T h e other
ship-type zaw, m entioned by Ibn Battuta (1968, IV: 91) could
be the Chinese version o f sao, seou or ts’ao and, interestingly,
though not related semantically, the word zaw occurs in al-
Tabari’ s (d. 310/922—3) authoritative history as early as the
tenth century to denote a river boat on the Tigris (XI,iii: 682,
Il6 8 ). For kakam, Beckingham identified it as the Chinese h oa -
ch ’uan (l994> ^V 813, fn. 43) m eaning a trading vessel, the type
that Ibn Battuta sailed on (IV: g 5—6/815) beyond the Indian
coast to China.
T h e War Vessels
Some o f the anecdotes Ibn Battuta relates on sea incursions

include specific types o f war vessels. Attempts to find the real
native term for the vessel-type were rare. The generic term harbi
to indicate a war-vessel (IV: 206/865) only occurs once in the
context but its usage was quite prevalent in the medieval
p erio d , particularly in the M editerranean.
T h e first war vessel-type in our inquiry is a native term
ukayri. Travelling from Saghar to Cambay in 1342, Ibn Battuta
tells us that the Sultan o f Q andahar sent his son to accompany
him and his associates on a ship (markab) called ukayri which he
compares with a ghurab, only that it was broader with sixty oars
and during battle its ro o f was covered with some sort o f
material 'to protect the rowers from arrows and stones’ (IV:
59/800). I take it therefore, that the ghurab’ s deck was open,
hazardous to the rowers and the archers, an interesting feature
which Ibn Battuta thought worthy o f com m enting on, failing
however, to give us a complete picture o f what the rest o f the
'ukayri looked like. It is clear from the context that the war vessel
was an Indian O cean type, large enough to accommodate sixty
rowers and fifty Abyssinian m en-at-arm s (ibid), and it was used
to protect cargo vessels and passenger ships from pirates who
infested the Indian O cean and the Arabian Sea at the time.
No m ention o f ghurab13 occurs anywhere else in the text and it
is clear from the passage that his knowledge o f ships is based on
that o f his hom eland. T he ghurab is indeed a M editerranean

vessel. We know from the historian al-M aqrizI (d. 845 ///I 4 4 I)
that the M editerranean ghurab was a huge war galley14 and,
though Ibn Mammatl (d. 606/1209), a contem porary o f Ibn
Battuta, describes the ship as having 140 oars, larger than the
'ukayri, we are still left in the dark as to this ghurab’ s hull design.
To complicate matters, al-M aqrlzi refers to both the Christian
and Muslim galley as ghurab, which could be interpreted as
bearing similarities in hull design or in the num ber o f sails.
T h e Indian O cean ghurab which appears often in the Dutch
documents o f the latter seventeenth and throughout the
eighteenth century in connection with native Perso-Arab and
Indian cargo or war (pirate) vessels is o f interest,15 and maybe
one can draw some similarities with the M editerranean type
which could have been introduced in the G u lf and the Indian
O cean during the Portuguese presence as early as the sixteenth
century. Jean Deloche (1996: 208, 2l8), thinks that the
iconographic evidence o f a twelfth to fifteenth century galley
found on the K onkan coast o f west India (now in the Goa
museum) with a sharp stern and an unusual projection o f the
bow could be the forerunner o f the eighteenth century ghurab
used by the Marathas.
A noth er Indian O cean war vessel-type is the shillfr which Ibn
Battuta says has similarities with the fukayri. He was in Honavar,
south-west India in 1342 visiting the sultan, during which time
he was involved in a war expedition to conquer Goa. T h e sultan
Jamal a l-D in o f Honavar had fitted fifty-two vessels (markab) for
this expedition. Ibn Battuta embarked on one o f the vessels
with the sultan. The vessels moved in to attack and the
inhabitants fought back with mangonels, throwing stones at the
sultan’ s people, forcing them to jum p into the water, shields
and swords in hand. Ibn Battuta too jum ped into the water
while the sultan took refuge in an armed fukayri (IV: 107/820),
'which is like a shillir’ . I was unable to trace the latter name in
any medieval source I could lay my hands on. To assume that it
is a barque (Dozy 1967? I*' 7^ 2 ) with a shallow draft, as
Beckingham (Ibn Battuta 1994 ? IV: 820, fn. 55 ) understands
it, is only guess work. From the context it is certain that the
190 D io n is iu s A lb er tu s A g iu s
type is a war galley and, if Ibn Battuta is correct with the name,
it must be a local term.
W ith the fifty-two expeditionary vessels m entioned above,
Ibn Battuta reports that there were two other war ships o f the
tarida type 'open at the stern, carrying horses’ . T h e craft was
built in a way that 'the horsem an mounts his horse inside the
vessel, puts on his arm our and comes o u t’ ready to fight the
enemy (IV: 10 7/8 20 ).16 T h e traveller must have been im
pressed by this warfare technique to m ention it, though it was
not unique as the function o f the tarida was well known in the
M editerranean for the transport o f horses and heavy goods,
and having similar hybrid features o f oars and sails like the
ghurab.17 That the tarida transported horses is often reported by
the Muslim chroniclers o f the Crusades (Recueil des Historiens
1967, I: 6l2; III: 214 ; IV : 164). We have no knowledge o f its
hull design except that it was the largest among the vessels
known at the time. T h e Christian type was to some extent a
smaller design o f the round ship, with oars 'to supplem ent
pow er’ , in particular to get in and out o f the port. It was mostly
used by the Genoese and Venetians (U nger 198O: 125)* T h e
Arab tarida (pi. tarayid, tara id and taradat) had probably originated
from tarrad, 'a small but fast ship’ ; one still needs to establish if
the Byzantine xapixa18 and Low Latin tarida and tareta, originally
signifying a reed canoe in the Red Sea (Christides 1984: 4^)
has come from A rabic tarida or tarrada19 (Pryor 1992: 28).
T h e war vessel ahawra is associated with Sind, north-west
India. Ibn Battuta relates that in 1334 he and a num ber o f
passengers came to the river o f Sind on the border o f the
territories o f Sultan M uhammad Shah, king o f H ind and Sind
(III: 94 / 593 )-20 T h e ahawra stood out among the fifteen vessels
(markab) which the jurist 'A la’ al-M ulk o f Khurasan, governor
o f Laharl in Sind owned (III: 108—9 /6 0 0 ).21 It was likened
somewhat to a tarida in 'our country [i.e. M orocco]’ , reports
Ibn Battuta, 'but broader and shorter. In the centre o f it there
was a wooden cabin to which one clim bed up by steps, and on
top o f this there was a place prepared for the governor to sit in.
His suite sat in front o f him and the mamluks [i.e. the slaves]
stood to right and left . . . ’ (Ill: 109—II0/601). If the ahawra22
C l a s s i f y i n g V e s s e l- T y p e s in Ib n B a t t u t a ’ s R ih l a i 9i
was similar to a tarida there is nothing to suggest that it was open

at the stern nor, indeed, if it functioned as a vessel for
transport o f horses as the tarida did. In any case the fact that Ibn
Battuta draws our attention to two different terms for what
appears to be the same hull design, albeit to an inexperienced
observer o f ships, is in this context significant.
The frequent occurrence o f the term ja fn (pi. ajfan, jifa n or
ju fu n ) is used almost indiscrim inately o f m erchant vessel and war
galley.23 From the usage o f the word in the Rihla we notice that
ja fn is employed as what the traveller understood it to mean in
his own M editerranean culture i.e. war galley, the plural usage
being ajfan ghazawiyya (II: 311/446 and iV: 104/819) or ajfan
harbiyya (II: 3 50//4-66). Both ghazawiyya and harbiyya are used
adjectivally to mean fighting, otherwise ja fn when found on its
own refers generally to a 'cargo ship’ . Alternatively, i f ja fn was
accom panied by an adjective safari or safariyya (meaning 'seafar
in g’) as ajfan safariyya it would signify 'cargo ships’ (al-N akhili
1974: 26). Jafn, on its own, can also mean galley if it is found in
the context o f the ship taking part in an incursion between
Muslims and non-M uslim s. W hen Ibn Battuta was on the
Malabar coast in 1344 he saw eight vessels o f the ja fn -design
which belonged to the sultan o f Malabar, under the protection
o f the sultan o f Ceylon, a powerful man o f the sea and a tyrant
who owned pirate ships (IV: 166—7/847—8)- This ja/h-type was
not a fighting ship, it was a cargo vessel bound for Yem en, a
route to which ships from the east sailed and, in its harbours
unloaded and loaded cargo on local vessels, which then
continued the route through the Red Sea and finally into the
M editerranean. The sultan ordered troops to guard the jafns in
the port against the enemy who could loot and burn them, and
in fact this was the intention o f the seamen with whom Ibn
Battuta was travelling, to 'effect a surprise seizure o f them ’ . O f
course, having been alerted to the heavy guards placed with the
ships they called the operation o ff (IV: 103/818). T he vessel is
not described in the Rihla. Being such a com m on ship, no
medieval source has recorded any detail o f it, a fact which is
true o f many other vessel-types. The craft was definitely known
to him as he compared a vessel he was travelling on along the
192 D io n is iu s A l b e r tu s A g iu s
river Zaytun (C h ’iian -chou ), in southern China, to a ja fn in his

country (IV: 271/895)' j/^ most exclusively, the term appears in
a M editerranean context, in particular with references to the
Maghrib (N orth Africa) and al-Andalus (Spain), including the
islands o f the western and m id -M editerran ean .24
T h e Laun ch es
A m on g types o f launches Ibn Battuta m entions harraqa, shabbara

and sallura, im plying they were pleasure boats (II: Il6 —7/
336—7)- He arrives in Baghdad in the year 1327 and his arrival
coincides with the presence in Iraq o f the king, A bu Sa'id
Bahadur K han (716—736 / l 3 I 6 —35 ) ’ the last ° f the direct line
o f M onghol Khans o f Persia .25 'I saw both the sultan and his
vizier one day’ , he says, 'on the Tigris in a launch [harraqa]
(which they [i.e. the Iraqis] call shabbara, and is like a sallura) . . . ’
(II: 116—7/336—7)- From his inform ants Ibn Battuta reports
that harraqa (the type he is fam iliar with in the M editerranean) is
called in Baghdad a shabbara. We are left in suspense not
knowing in the first place what a harraqa is nor do we have a clue
what shabbara and sallura are. This is what I call the circular
approach o f describing unknown things: o f course the author
did not explain that with which readers were fam iliar and
therefore any description o f such ships was unnecessary. In
another reference to harraqa, when Ibn Battuta was in China, he
says that he sailed with the great am ir’ s son on the canal 'in a
ship [safina] which resembles a harraqa . . .’ (IV: 289/903)- From
the passage we can infer that it was a pleasure boat because o f
the reference to musicians and singers on board. G ildem eister
(1882: 438) describes a fifteenth-century harraqa as a gondola, a
small flat-bottom ed craft, with sharp pointed bow and stern
and a very low freeboard used by emirs and notables for
pleasure trips. Some harraqas were used to light fireworks for
festivals such as the sixteenth-century type on the Nile in C airo
(M oritz 1915: 439).
T h e M editerranean harraqa is known as a type equipped to
throw 'fire upon the enemy at sea (or on a large river)’ (Lane
1984, I: 552). Such was the harraqa m entioned by the historians
C la s s ify in g V essel-T y p e s in Ibn B a t t u t a ’ s Rihla 193
al-Tabari and al-M aqrlzI;26 the latter, writing about vessel-

types on the N ile, describes the harraqa as a fairly small warship
used to support the larger ships and galleys (1911—24> I-ii;
222). Also n on -A rab ic documents record the term as referring
to a type o f warship: the Genoese car(r)aca (in II 57)> the French
caraque referred to as the Saracen small boat, the Andalusi carraca
(variable o f the Venetian charachia 1418—145 ^) and the P ortu
guese carraca or caracca, a big vessel and warship that sailed to
India and South Am erica (Pellegrini 1978, II: 815), all o f
which seem to originate from the Arabic harraqa. T h e verb haraqa
means 'to burn something; to h urt’ and ahraqa 'to destroy by
fire ’ , the noun being harraq 'burning aflam e’ . The use o f harraqa
was definitely not restricted to wars, though it is difficult, on
the other hand, to determ ine its developm ent over the period
o f time.
T h e shabbara (from Farsi shah\-bar 'royal carrier’) is described,
according to the historian Ibn al-A th lr (d. 630/1232—3), as a
kind o f barge with an elevated cabin similar to that o f a gondola
(1965, HI: 9 6 ),27 but it appears in the chronicles o f al-Tabari
as a war canoe on the Tigris on num erous occasions in the
battle against the Turks in 865. Sallura (pi. salalfr or salariyya) is
known to be a boat that accom panied a warship on the
M editerranean,28 with 16 to 24 oars,29 as well as a river boat on
the N ile, the Euphrates and the T ig ris.30 T h e origin o f the
name has been traced through G reek aeAlapiov (K inderm ann
1934 : 4 1)- In the Rihla, shabbara was obviously a launch as much
as, apparently, sallura was.
Coastal S a ilin g and Rowing Vessels:

the j a l b a and s u n b u q
A com m on coastal sailing vessel in the Red Sea was the trading
jalba (pi. ja lb at, jila b ), probably a two-masted ship. Ibn Battuta
embarked from Jeddah in the Red Sea on a vessel the locals
called jalba (II: 15 8/361). There were two jalbas awaiting to sail to
Yem en. He boarded on one but refused to stay on because o f a
num ber o f camels; he recounts, 'I was frightened o f this, never
having travelled by sea b efo re’ (ibid). The interesting thing
194 D io n is iu s A lber tu s A giu s
about this anecdote is that it was the first time that he ever
journeyed by sea and perhaps the idea o f being in company
with camels was daunting. A fter this trip Ibn Battuta never
m entions any other ja lb a.
T h e text does not provide us with any inform ation other
than the name ja lb a . A description o f it was given by Ibn Jubayr,
the Andalusian traveller, some one hundred and fifty years
earlier, who wrote that the Red Sea jalba (jilaba as he called it),
was sewn together with cord made from qinbar (coconut fibre)
(195^: 65)- A n interesting detail that he adds is that the wood
and the coconut fibre for the vessel were im ported from India
and the Yem en. O n e feature about the jalba is that its sails, Ibn
Jubayr reports, are woven from the leaves o f M uql tree (a kind
o f gum tree) parts o f which were weak and unsound in structure
(ibid). O n this subject the historian al-Nuwayri 1-Iskandarani
(d. c. 779 /1377) notes an Indian O cean jalba carrying seven
square sails; these mats were woven with flax and coconut
thread .31 T he reference to the Indian O cean ship is unusual,
because historically the ja lb a - type is always connected with the
Red Sea. It is quite possible, however, that the name could be
linked to the Farsi galabat or galevat or the west Indian gallevat, the
latter though implies a galley with oars (down to the latter h alf
o f the eighteenth century) (Yule and B urnell 1994 : 3 ^1)-32
Th ere is no m ention o f this Indian gallevat ever being used as a
cargo ship though galleys often operated as cargo and transport
ships. In the Dutch documents (namely seventeenth century)
concerning the Arabian/ Persian G ulf, we come across the
name gallivat, gallouet, gallowet, mainly as warships or pirate ships
with the occasional reference to cargo vessels. In almost all the
references these vessels were sailed by Persian seam en .33
Sunbuq is another com m on medieval sailing ship as well as a
coastal rowing vessel. From the Basra quay Ibn Battuta
embarked in a sunbuq (pi. sanabiq) which was going to Ubulla,
the distance being ten miles. It passed 'through an u n in
terrupted succession o f fruit gardens and overshadowing palm
groves both to right and left, with traders sitting in the shade o f
the trees, selling b re a d , fish, dates, milk and fru it’ (II: 17/280)
from which I understand that sunbuq here means a ferry canal
boat or perhaps a cargo boat, though Ibn Battuta, aristocrat as

he was, would have refrained from travelling on a cargo boat
which smelled o f fish, vegetables or fruit. O ne characteristic
that Ibn Battuta thought worthy o f m ention was the fact that
'sailors in this country [Basra and Ubulla] row standing u p ’ (II:
17/281), which suggests that the norm al practice is rowing in a
sitting position. O n one other journey in East A frica, he sailed
from Zaila (Zeila) on the Somali coast due south o f A den to
Maqdashaw (Mogadishu), a town o f enorm ous size,84 the
inhabitants being merchants who have a large num ber o f
camels. He relates: 'It is the custom o f the people o f this town
that, when a vessel reaches the anchorage, the sunbuqs, which are
small boats come out to it’ (II: 181/374)* He reported further
that in each boat a num ber o f young men from the town rowed
to the vessel and boarding it presented food to the merchants.
O n disembarking the merchants went to the house o f the host
group o f young m en who took them on their sunbuq. T he sunbuq
was also used as the customs boat for the Sultanate o f
M ogadishu. Inspectors used to ferry themselves to the
oncom ing vessel and make inquires about the ship, its owner,
the cargo etc. (II: 183/375)- port town o f Zafar, it was
also a custom that when a vessel arrived from India or
elsewhere, the sultan’ s slaves went to the ship in a sunbuq (II:
198/383). Here, the custom was that the slaves carried a set o f
robes which they donated to the rubban (the sea-captain), the
kiram (the ship’ s accountant) and the owner o f the vessel or his
agent. N ot only that, but to each was brought a horse, on which
he m ounted in a procession 'with drums and trumpets playing
before them from the seashore to the sultan’ s residence’ (II:
19 8 /38 3-4 ).
In all the above passages the large vessel that anchored at the
port awaiting the sunbuq was simply referred to by the generic
term markab. Larger sailing types o f sunbuq seem to have been
around, as Ibn Battuta recounts how he sailed in 133 1—2 *n a
sunbuq from Jeddah to 'Aydhab but not coping with the wind was
driven back to a place called Ra’ s Dawa’i r .35 It was dou ble
ended as probably all dhows were at that time and it would have
carried a lateen sail and must have been popular for short
196 D io n is iu s A l b e r t u s A g iu s
journeys and perhaps m ore manoeuvrable in winds than larger

dhows.
I found one reference to sunbuq in relation to the Arabian/
Persian G u lf and the Arabian Sea as early as the tenth century by
the story teller Buzurg b. Shahriyar (al-Ram hurm uz! 1883—86:
190), him self being a sea captain. We do not know how the early
medieval sunbuq looked except that it had a curved bow (al-
Jawaliqi 1361/194-2: 177) • T h e historian al-Nuwayri 1-Iskandar-
anl, a contem porary o f Ibn Battuta locates it as belonging to
Yem en36 and India.37 It is m entioned later in several Portuguese
accounts as sambucho or zam buco.3& T h e name still prevails in
m odern times. In the course o f my interviews with seamen in
the G u lf and the Arabian Sea coast I could identify three types
o f sanbuq (as it is pronounced today), all o f which have a
scim itar-curve design at the lower end o f the bow. It looks as if
the medieval sunbuq could be the forerunner o f the Yem eni
double-ended type and it is quite possible that the G u lf square-
sterned design that you find in m odern times is an adaptation o f
the proto-Y em en i type giving it m ore space at the aft for fishing
nets and manoeuvrability o f lifting nets laden with fish. As for
the origin o f the w o rd , the G reek chx^Puks has come from Middle
Persian *sambuk giving Neo Persian sunbuk (al-Jawallql 1361/1942:
17 7 -9 ), the source o f which may be traced to Sanskrit gambuka
(De Landberg 1920—42, III: 1986—7, fn. i).
C oastal R owing Boats
Th e custom o f locals boarding a coastal boat from the shore to

a vessel that arrives in a port seems to have been com m on from
the Red Sea to the Indian O cean. T h e Maldivians met the crew
and the merchants on a coastal kundura (pi. kanadir) as the natives
called it, reports Ibn Battuta (IV: II9/826).39 T h e name,
however, remarked Gibb and Beckingham (Ibn Battuta 19 9 4 ,
IV: 826, fn. 17) was used in Sinhalese to describe a Maldivian
sailing boat while the locals used the word odi or doni, the latter
being com m on in South India.
A n oth er type Ibn Battuta m entions is the ushari. He relates a
dramatic experience on his intended departure from Jeddah
soon after the pilgrim age to Mekka had ended in 1331—2 . In it

he narrates how some crew members and some merchants after
severe distress escaped in an ushari as the ship they were on
foundered (II: 2 5 1/4jI3 ) * In a second anecdote he tells us how
an fushari (coastal boat) he boarded got stuck in the m ud at the
time o f low tide near the town o f Q uqa (Goga) in south-west
India: when we sank into the mud I was leaning upon two o f
my com panions, and the men o f the place gave me to fear the
turn o f the tide before I should reach it, since I was not a good
swimmer . . (IV: 60—1/800). There is no indication that the
ushari is a local Indian O cean name; most o f the medieval
sources point to the fact that it was a type used in the N ile and
the Red Sea. Several references by al-M aqrlzI illustrate the
usage o f fushari as a gondola on the N ile which was in part as
splendidly furnished as the harraqa and shabbara, referred to
earlier, and which was generally m entioned during the Fatimid
dynasty (s>97— ^ was a r o j a l type o f gondola
used by caliphs, princes, ministers and court officials. T he
name may have passed to the western uscerium, uxeria, usceri, uscerius,
uscerio and usseria.40 W e also know the different types o f ushari,
their size, weight and function; some were given different
names: rushari dimas, fushari dukasa, ushari fid d i etc.41 T he closest
description o f an 'ushari is found in the biographical dictionary
o f Ibn Khallikan (d. 681/1282—3): 'It has the form o f the one
called shabbara on the Tigris, but it is much longer and broader
. . .[ It is] decked with thick solid planks and has projections in
the form o f balconies . . . O n the deck [you find] a wooden
dom ed cabin with windows and curtains . . . ’ (1842—J l , I:
299)- From this sketchy description it clearly indicates that an
Iraqi fushari denotes a pleasure boat, which could be o f the
gondola-type. It was Ibn Battuta’ s understanding from his
M editerranean background that an ushari was a coastal rowing
boat, perhaps flat-bottom ed.
T h e Raft: m a diyya
In his travels Ibn Battuta recounts how he had to cross rivers or

lagoons on a raft. He does not seem to have ever bothered to
find out what the local word for raft was as he always uses the
term ma'diyya for any type. O n e anecdote relates his dramatic
experience o f a voyage to the land o f Malabar when a gale rose
and the ship (m arkab), with no skilled pilot on board, narrowly
escaped being wrecked. T h e sailors cut down the mast and
threw it overboard and then made a w ooden raft tying ropes to
it (IV: 185—7/857)* He recounts another story o f a certain
A m ir Bakht who fled with his associates to the land o f Sind
from the sultan who had a grievance against him . W hen they
came to the river o f Sind they were determ ined to cross it; so
they built a raft o f reeds (III: 359 / 7 ^8 ). In one other story, Ibn
Battuta tells us how the sultan o f Ceylon and his associates
reached the other side o f the river to go to M inneri-M andel
(Manar M andali) on a raft made o f bam boo canes. As far as the
name is concerned it is worth noting that ma'diyya is particularly
a medieval M aghribi (N orth A frican) term, including al-
Andalus (Spain) and Egypt. T h e word is missing from early
Classical Arabic dictionaries and al-K hafaji (1282/1865: 219)
allocates the m eaning to colloquial Arabic denoting 'ferry
boat’ ; the A rabic term can be traced in Portuguese almadia,
Spanish almadia or armadia, French almadie usually with the
m eaning 'ferry boat’ or 'raft’ . What is interesting to note is that
almadia is often m entioned in the works o f Portuguese travellers
in India with the understanding that it is a small native boat or
canoe (Yule and B urnell 1994 : 15—^6 ). Th ere is no medieval
or m odern Arabic source that records ma'diyya in the Indian
O cean or the A rabian coast.
T h e F o u rte e n th -C e n tu ry Perso-A rab

and In d ia n Vessel
T h e Rihla o f Ibn Battuta records hardly anything about the

construction o f Perso-Arab and Indian ships and the methods
used in navigation. Ibn Battuta had seen ships anchored in
Tangier, his home town, but never travelled on one until he
came to Jeddah, the beginning o f his first journ ey to the east by
sea (interrupted occasionally by land routes). Th ere is nothing
to suggest in his m em oirs that any o f the ships m entioned were
double-ended or square-sterned, a fundam ental distinction

made today by Arab shipwrights in the east whenever you
discuss boat typology in the Arabian G u lf and the south
A rabian coast. We can only surmise from a few illustrations (so
rare in the dom ain o f Islam) that such ships were double-ended
and, including various literary sources, that they were sewn.
T h e build o f boats all along the Arabian, east A frican and west
Indian coasts varied according to the localities for which they
were destined and each adapted to a particular need. W ith the
triangular trade around these coasts, there seems to have been a
cross-fertilisation o f hull designs that took place over many
centuries since times o f antiquity. It seems that Perso-Arab
designs were dom inant along the north-east coast o f India
where the Parsee comm unities mainly settled and influenced
ship construction, though the original design may have
disappeared, some m odifications were introduced due to
Portuguese and Dutch influence in the latter centuries.
W ood, as it still is today, was im ported to the Arabian/
Persian G u lf and the south Arabian coasts from India; also
from Yem en, as in the case o f ships built in 'Aydhab (Ibn
Jubayr 1952 : 65). Teak is often quoted by writers 42 as being very
hard but it is considered durable and pliable once it has been
seasoned. T he cord, from coconut husks ,43 was im ported from
the Maldives and Luccadives but also from Zafar as Ibn Battuta
inform s us (II: 204/387)* Both Ibn Jubayr and Ibn Battuta
were interested in the making o f the cord from qinbar or qanbar:
'The makers’ , in 'Aydhab, reports Ibn Jubayr, 'thrash [the
qmfear] until it takes the form o f thread, which then they twist
into a cord . . .’ ( l 95 2 : 65)- Ibn Battuta saw the qinbar in the
Maldives and described it as 'the hairy integum ent o f the
coconut, which they tan in pits on the shore, and afterwards
beat out with bars; the women then spin it and it is made into
cords’ (IV: 121/827 ).
Shipbuilding in the Perso-Arab and Indian tradition is the
shell-first construction, i.e. the keel first followed by h o r
izontal planks fastened to it on each side. N either Ibn Battuta
nor Marco Polo write about ribs supporting the sides o f the
hull. Obviously, this was com m on knowledge and it was
therefore not im portant to m ention it. Vessels would certainly

need ribs to give them strength and rigidity. W hen the ship was
constructed the carpenters caulked the planks even though the
medieval traveller, John o f M ontecorvino (d. 1328), reported
that it was not. Th ere is ample evidence that his observation was
wrong (Cathay 1866, I: 2 l8). T h e ships that the Andalusian Ibn
Jubayr saw on the Red Sea were caulked 'with shavings o f the
wood o f palm -trees’ (195^: 65) • Carpenters smeared the planks
and the joints with shark oil (qirsh), a m ethod still being used on
the G u lf and the south A rabian coasts (and see Beckerleg for
the east A frican coast, this volum e). Marco Polo notes that the
ships were not caulked with pitch 'but anointed with a sort o f
fish o il’ (1982: 53 )- Perhaps they used a mixture o f pitch and
shark or whale oil. I have observed in Saham, east o f Sohar
north o f O m an, during my field trip in Novem ber 199^* how
they cooked the shark’ s liver in a big cauldron until it melts.
This practice, I am told, was com m on throughout O m an and
the south Arabian coast. A l-IdrlsI (d. 561/1165) writes on this
subject that fisherm en caught the smallest whale which they
cooked until 'its flesh melts and changes into thick liq u id ’
(quoted by H ourani 1963= 97 )* Needless to say the smell is
hideous. T he greasing o f a boat is essential, reports Ibn Jubayr:
the reason is 'to soften and supple it against the many reefs that
are met with in the sea’ (1952: 65); the same reasons were given
by my inform ants. As a matter o f fact nailed ships could not sail
where there were many coral reefs. From his experience Ibn
Battuta tells us that whenever the sea was full o f reefs a nailed
ship was broken up on striking rocks. That could be the case
with smaller or m edium -sized ships but with larger ships such
as the Chinese junks, the bulkheads form ed o f thick planks
rabbeted together served against accidents in which water could
leak into the hull on striking a rock.
Marco Polo is in no doubt that the Perso-Arab and Indian
ships, 'were very bad, and many o f them founder, because they
are not fastened with iro n nails but stitched together with
thread made o f coconut husks’ (1982: 52, 2 7 °) and instead o f
nails they employed w ooden pegs. That ships were wrecked was
not u ncom m on in those days but there is no evidence to
C l a s s i f y i n g V e s s e l-T y p e s in Ibn B a t t u t a ’ s Rihla 201
support Marco P o lo ’ s theory that stitched vessels are frail and

therefore easy to break. Ibn Battuta explains that Indian and
Yem enite ships sewn together with cords escape this fate: they
are sturdy, their wood is pliable and with the cord are
sufficiently flexible not to 'fall to pieces’ (IV: I 2 l/ 8 s>7)-
M edium -sized sewn boats, such as the sanbuq I saw in ruins on
the D hofari coast in O m an in Novem ber 1996> existed until
fairly recently: some o f these vessels were ocean-going and
coped with strong winds and high seas as well as nailed boats
(many o f my older interviewees have testified to this). Th ere is,
therefore, no truth in Marco P o lo’ s and Friar John o f
M ontecorvino’ s statement that native vessels were wrecked by
strong winds because they were stitched (Cathay 1866, I:
2 17 -8 ); findings o f a wrecked ship can reveal the strength and
durability o f nailed or stitched planks but such wrecks in this
part o f the world (particularly the south A rabian coast and the
Arabian/Persian G u lf) are not to be found mainly due to the
w ood-eating teredo44 and to the salinity o f the ocean. T h e oil
substance described earlier is also good for blocking holes and
general cracks, thus strengthening the bottom o f the ship
against the seaworm and salinity. M oreover, a coat o f w hite
wash on the outside o f the sheathing board is renewed every
season.
We know nothing from Ibn Battuta’ s diaries as to whether
the Perso-Arab and Indian ships he saw carried a mast or a
rudder. Marco Polo, almost two centuries earlier, observed that
Arab ships in H orm uz had one mast, one sail and one rudder
and no perm anent deck (1982: 5^)- I do not think that the
ships in Ibn Battuta’ s time would be different. T h e widely
discussed illustration o f the M aqamat o f al-H arlrl (d. $ 1$ /
1 1 2 1—2)45 has am ong other features worthy o f note: the crow’ s
nest, a deck house located ahead o f the mast, a straight stem, a
curious curved stern and an iro n fo u r-p ro n ged grapnel anchor
hanging from a divided bowsprit. A dual bowsprit could be
found up to thirty to fifty years ago on the Kuwaiti bum , a
d ouble-ended ship with a straight bow set at an angle o f 45
degrees. What is o f particular interest in the illustration is the
stern rudder operated by the rope-steering system which was
typical o f the O m an i dou ble-ended badan, baggara and zaruka. In

another illustration taken from the Kitab al-furusiyya wa l-m anasib
al-harbiyya attributed to Hasan al-Ram m a (Egypt or Syria
thirteenth or fourteenth century)46 we see a drawing o f a vessel
with two masts, (one carrying a lateen sail), a straight keel and a
straight stem and stern post, with no evidence o f a rudder, but
again the hull design looks clearly similar to a Kuwaiti bum, even
though the latter’ s design is a twentieth-century invention.
Th e question o f Perso-Arab and Indian ships having no
decks, as the medieval travellers rep o rte d ,47 is intriguing.
Jordanus writes from Malabar that the undecked vessels he saw
took in so much water that sailors 'always, or almost always,
must stand in a pool to bail out the water’ (Cathay 1866, I: 53 )-
T h e Maqamat illustration showing seamen bailing out water does
seem to confirm this problem . Water seeping in through the
bottom o f the ship is not confined to undecked ships but
occurs with decked vessels too. T h e illustration shows cabin
holes, which can be interpreted to suggest that such ships were
decked. N one o f our medieval authors m ention any Chinese,
or for that matter, some Perso-Arab or Indian ships rem aining
dry. T h e fishing dhows I came across, such as the O m an i badan,
baggara, battil and the zaruka are undecked or partly decked on the
fore and aft. T h e quantity o f water that an undecked ship takes
because o f the crashing o f waves could contribute to the
problem , but my inform ants told me, that it was nothing
com pared to the quantity o f water that enters through the
bottom . O f course a lot o f the merchandise was covered with
reed mats or skins which absorb water, as Marco Polo observed
(1 9 8 2 :5 2 ).
T h e m onsoon winds were undoubtedly rough and boisterous
and ships were wrecked because their sailing gear became
entangled and broke when sails were raised or low ered’ (Ibn
Jubayr 195^ : 7 °)- Passengers often lost their lives. Talking
about him self and those travelling with him , Ibn Jubayr
remarks on the num ber o f times they 'died and lived again
(ibid). It would require a skilled rubban (sea captain) to steer the
ship in strait channels and on the open sea. Experienced pilots
knew the coasts and they could prevent their ship from a
disaster long before the storm hit the sea. T h e Yem enites and
the Persians were the most able sea captains and seamen (al-
Ram hurm uzi 1883—86), who sailed from Siraf, H orm uz and
O m an to India and China. Sailors o f Siraf, the geographer al-
Istakhrl states, 'passed their whole life in ships’ (187O: 138) and
al-MuqaddasI claims that most o f the shipbuilders and seamen
were Persians (1906:18). These comments come admittedly
three centuries earlier but there is no other inform ation to
suggest anything different in Ibn Battuta’ s time.
T he western shores o f India and the southern Arabian coast
were open to the strong south-west winds that blew from May to
early O ctober. Indian, Persian and Yem enite merchants would
trade during the winter months and seamen would do repair
work and careen their vessels during the summer m onths. They
used the north-east winds to sail to the Malabar coast when the
weather was fairer. In this context, Ibn Battuta reports how over
one lunar m onth he covered a distance from a port on the
south Arabian coast to a point in west India: 'I m yself crossed
(it) once from Q aliqut in the land o f India to Zafari in twenty-
eight days with a favouring w in d , sailing continuously by night
and by day’ (II: 196/382). Marco Polo comments that from
Malabar to Zanzibar the voyage lasted twenty days, whereas the
return trip took three full months, due to 'the continual
southward set o f the current. It flows in the same direction all
the time - southward, ever southward’ (1982: 263). The
m onsoon winds indeed played an essential role in the dynamics
o f trade, as Ross D unn (1986: I l 8 ) explains, 'the seasonal
rhythm o f the winds gave Indian O cean trade and travel an
element o f symmetry and calculability not possible in the
M editerranean’ .
T he fourteenth-century vessels were no bigger than m odern
traditional dhows, not exceeding 300 tons o f cargo weight and
with a length o f not m ore than 40 feet. T he large square-stern
A rab-In dian baghlas and ghanjas o f the early twentieth century
carried a weight o f 500 tons and their hull design was
influenced by the Portuguese. T h e Maqamat and the Kitab a l-
furusiyja illustrations clearly indicate that the ships are dou ble
ended like the Y em en i and D hofari sanbuq and the Kuwaiti
204 D io n is iu s A lb er tu s A giu s
bum. A ll o f this inform ation was obvious to the fourteenth-

century ship builder and, to the unfam iliar eye o f Ibn Battuta,
all ships looked tediously alike with no variations in hull
designs, as they all did to com m on people. Hardly any o f the
medieval authors give names o f vessel types, let alone describe
hull designs in general or speak about shipbuilding. O nly a few
did: Ibn Battuta’ s sporadic attempts to be specific do help us in
some way to locate and classify some Perso-Arab and Indian
types.
C o n c lu sio n
O u r Classical A rabic lexica hardly have entries concerning

names o f vessel-types or for that matter nautical term inology;
n or do they in general classify m aterial-cultural terms that are
known to the com m on people. Such terms have no place in a
dictionary as they are often foreign, and Classical A rabic lexica
deal mainly with religious, philological and literary term in ol
ogy and anything which is classified as pure Arabic. O nly ship-
types like markab, safina and qarib which are in any case generic
and Q u r ’anic terms would find their way into these lexica. Also
o f concern is that medieval Arabic dictionaries fail generally to
include names o f ship-types. A few such as ja fn , harraqa, jalba and
sunbuq have been listed however. T h e obscure terms (and these
are many) are often missed out. T h e question is: what possible
sources can help us find inform ation about vessel-types?
Geographical sources and travel works undoubtedly help, while
historical works are perhaps the most rewarding. However one
cannot look at historical works and ignore the rest. O ften one
complements the other, whatever the amount and quality o f
inform ation available. In some way Ibn Battuta’ s Rihla is
disappointing not as regards names o f vessel-types but in the
lack o f inform ation about them. Like many medieval authors
description o f ship-types is lim ited simply because there was no
need to provide it, it was com m on knowledge to all who read or
heard about them . T h is explains the circular approach
concerning the harraqa, shabbara and the sallura m entioned above,
none o f which tells us what they actually were. Nonetheless,
what Ibn Battuta should be credited for is his com parison o f

M editerranean vessels with Perso-Arab and Indian vessels. This
is unique in the Islamic history o f the nom enclature o f vessel-
types, and were it not for the interesting and often personal
anecdotes that Ibn Battuta excelled in, it would have been
impossible to determ ine the function o f some o f the vessels he
m entioned.
Notes
1 T h is is the short title o f Ibn Battuta’ s work, the full title b ein g Tuhfat al-
nuzzar f i ghara’ib al-amsar, w a-'aja’ib al-asfar (The Gem o f the Observers on the
Curiosities o f Cities and the Wonders o f Travels). In this chapter I will quote two
sources: (a) Voyage d lb n Battuta. A rabic text with tran slation by C.
D efrem ery and B. R. Sanguin etti, V olum es I—IV (Paris, A n th ropo s,
1968; first p ub lish ed in 1853—1858), (b) The Travels o f Ibn Battuta A.D.
1325 —1354 - T ran slation with revision and notes fro m the A rabic text o f
C . D efrem ery and B. R. Sanguin etti by H . A. R. G ib b, V olum es I—IV
(C am b rid ge: C am bridge U niversity Press, [Volum e I] 1958); (L o n d o n :
Hakluyt Society, [V olum es II—III] rep rin t 1995); [Volum e IV] translated
by H . A. R. G ibb (d.1971) tran slation com pleted with an n otation s by C.
F. Beckingham (d. I99&) (L o n d o n : Hakluyt Society, 1994)* C itation s will
be referred by the volum e n um b er follow ed by two page n u m b ers the first
fro m source (a) and the secon d fro m source (b).
2 For death, reign and dynastic dates, we will list two calendar years: the first
is the Islam ic hirjf year followed (with a slash) by the com m on era year.
3 T h ere are som e serious in consistencies 'if not im p o ssib ilities’ which H .
A. R. G ibb discussed in the A p pen dix o f V olum e II (1995): 5 ^ 8 —537 -
4 C opyin g fro m other literary sources was not con sidered im p ro p e r: som e
auth ors, however, did cite writers whether qu o tin g wholly or p a ra p h ras
ing the text, e.g. the geograph er al-M u qad d asi ( l9 0 6 ,I I I :3 ) scrupulously
credits any authority he quotes.
5 It is not clear whether he was referrin g to the m ain lan d known as al-H asa
or the island o f Bahrayn (B ah rain ) itself.
6 O n e in terestin g detail Ibn Battuta gives us is that the A rabs o f the Ban u
Saffaf, residin g in the island o f Qays were expert divers; G ibb notes that
these were A rab s fro m O m an 'lon g established on the coast o f Fars’ (Ibn
Battuta II: 4 ° 8 , f n . 139).
7 For detailed description s see: al-M aqaddasi l 877 : 375 »3 ^ ° ‘» al-M as'u d i
1861-77, I : 97 ; Y aqut 1 8 6 6 - 7 3 , 1:513
8 See note 9
9 I fin d that M arco P o lo ’ s description o f the tree and how its b ark is cut to
extract the sap o f the incense is quite accurate having seen it m yself on the
206 D io n is iu s A lb e r t u s A g iu s
hills n ear Salalah and M irbat in the D h ofar regio n o f O m an . H e writes:

'It is p ro d u ce d by trees o f no great size, like little fir trees. They are
gashed with knives in various places, and out o f these gashes oozes the
in cen se’ (M arco Polo 1 9 8 2 :2 7 2 ); see also the Periplus o f the Erythraen Sea :
'Th ese in cen se -b earin g trees are not o f great height or thickness; they
b ear the frankincense sticking in d ro p s on the bark, ju st as the trees
am on g us in Egypt weep their g u m ’ ( Periplus 1912:33)
10 See II: 1 6 0 /3 6 2 , 181/374, 3 5 4 /4 6 8 ; III: 1 0 9 - 1 0 /6 0 1 ; IV: I I9 —I ^ l/
8 2 6 - 7 ,1 8 5 - 7 / 8 5 7 ,2 7 9 / 8 9 8 , 4 3 2 /9 7 0 .
11 O n a recen t e xp ed itio n (M a rch -A p ril 1996) to Q alh at, n o rth o f S u r in
O m a n , sp o n so re d by the O m a n M aritim e H eritage p ro jec t, the W estern
A u stralian M aritim e M u seum in F rem an tle an d Earthw atch, we fo u n d
n u m e ro u s C h in ese p o rc e lain on the ru in s littered all over the p o rt
town; V osm er et a l (1 9 9 9 : 2 0 —3 0 ).
12 As early as the eighth and n in th cen turies we have rep orts fro m the
M uslim geograph ers and h istorian s about the C h in ese sea activity in the
A rab ian /P ersian G u lf (Ib n R ustah 1892 =9 4 an d al-T ab ari 1965, IV:
2023 ) som e 'ships fro m C h in a’ sailed up to U bu lla (ib id ., 1965, V :
2 3 8 4 ). T h e sea route to C h in a was then m uch safer than the lan d ro ute.
13 Several plurals are record ed by K in d e rm a n n (1 9 3 4 :6 8 ): ghurabat, aghrub,
ghirban, ghurb, and gharabin, the term probably gave M ediaeval Span ish
gorab, gorabo, corabo, 'vaisseau, galere, b rigan tin (P ellegrini 1978, II:
8 1 7 - 8 ).
14 A l-M aq rlzi Suluk 1 9 7 0 - 7 3 , III, i: 105,149*, IV, ii: 6 6 8 - 7 1 , 6 9 5 ; IV, iii: 2.
15 C D R A D /D u tch Archives, V O C 2 6 8 0 , O B (1747), 101 ( f f .4 3 - 4 ) , 34
(fol. 3 6 5 ); V O C 2 9 9 6 K h arg (N ovem b er 1759) fol. 2 0 ; V O C 9 1O I
M asqat (8 M arch 175^) Fol. 15; V O C 1913 B an dar A bbas (2 A p ril 1717)
fol. 2 9 ; V O C 2 6 8 0 (1741) fol 4 4 ; V O C 2 5 4 6 ,f f 1 8 0 4 - 6 .
16 A sim ilar d escription to that o f Ibn Battuta is given by the h istorian al-
Nuwayri 1-Iskan d aran i, see U A FA L (B erlin ) M S 667, fol. I 24 r: the tarida
is op en at the stern with d o o rs that open an d lock, designed to carry
h orses equ ip p ed fo r w ar’ . T h e re is, however, no in dication as to the
location where such craft was used.
17 Fatim id sources fro m T u nisia, see al-N ak h ill I 974 : 8 9 —9 1
18 'Batim ents courts destines au tran sp o rt de m archan dises, de trou p es et de
chevaux, doivent etre in trod u ites dans la langue byzantine par l ’in ter-
m ediaire du latin, c’est, du m oin s, l ’hypothese que suggere leur
ap partitio n tardive dans les textes grecs et leur m en tion en latin des les
9 c ’ (A n to n iad is-B ib ico u 1966: 168).
19 N o t to be con fused with a totally d ifferen t craft, the Iraqi m arsh canoe
tarrada, som e 36 feet lo n g (T h esiger 1 9 6 7 : 34 )-
20 Sin d was con qu ered by the A rabs as early as 9 4 / 7 ^
21 T h e town o f L ah ari is located som e 28 m iles south east o f K arach i (Ibn
Battuta I I I :6 0 2 ,f n .3 0 ) .
22 Phonetically sim ilar to the H in d i hurf, d u g -o u t canoe (still a com m on

craft in the G ulf, the R ed Sea and East A frica), but obviously the design
and fu n ction are differen t from the ahawra.
23 Roughly 15 tim es, I I : 234 - 5/ 4 0 2 , 3 n - 2/ 446 ,3 5 0 - 1 / 4 6 6 - 7 , 355/ 469 ,
4 3 3 / 5 0 9 ; ^ : 1 0 4 /8 1 9 ,1 6 7 /8 4 8 ,2 7 1 /8 9 5
24 Ibn al-K h atib 1956: 192; al-M arakush i I 949 : 3 X4 ; Ibn Ju b ay r 1852: 337 J
al-N akh lli 1974: 2 3 —6; Picard 1997: 162, 3O O -4, 3 9 8 ; see also
K in d e rm a n n 1934 : *8 .
25 See G ib b ’ s note (Ib n Battuta, 1995» H :335>fn -2 Il).
26 A l-T ab ari 1965, X I (3 ): 917; al-M aqrizi 1911-24* I, i i ; 3 ° 6 .
27 T h e re may be a con n ection with the w ord seebar, record ed by Edward Ives
( i 773 :I 96 ); he saw one b o u n d to M uscat in 1754 on a voyage from
E nglan d to India and it could be related to the M arathi shibar, a kind o f
coastal vessel, a patamar used in the M alabar trade (Yule and B u rn ell
1994: 827).
28 U AFAL (B e rlin ), M S 667, ff. I 23v I 24 r
29 UAFAL, M S 9 m ,ff. 2 0 - 1 .
30 UAFAL, M S 31m, fol. 140
31 U AFAL (B e rlin ), M S 667, fol. 127 r-v.
32 T h e M urcian jalfba (a river raft) o f the M editerran ean m en tion ed by the
geograph er al-Id risi ( d .561/1165—6), though m orph ologically sim ilar in
ro ot to the Red Sea jalaba, m ust be traced to a differen t o rigin.
33 C D R A D / D utch Archives, V O C 3184 (1766) 7 (fol. 8), II (fol. 17), 12
(fol. 18), 27 (f o l.51), 33 (fo l.61); V O C 2 8 6 4 L etter N o. I to Isaak Sweers
in A m sterdam 1756; V O C 2 9 9 6 K h arg (N ovem ber 1759) ff-8 —I I .
34 A trad in g city fo u n d e d by the A rabs from the Persian G u lf in the fo u rth /
tenth century (Ib n Battuta I 995 >H>fn -47)-
35 Probably re fe rrin g to D ukhana bay, som e 1 0 0 m iles south o f A ydh ab
(Ib n Battuta 1995, I I :4 l3 ,f n .6 )
36 M en tion ed by the h istorian al-M aqrizi in the context o f the R ed Sea as a
type o f vessel in the T u lu n id war fleet; although sunbuq may have existed in
the eigh th /fifteen th century, the events he was w riting about went back
som e 7 0 0 years earlier ( l 27° / l 853 > IL 1 8 0 ). O n e questions the reliability
o f the record o f such nam es.
37 U AFAL (B e rlin ), M S 667, fo l.I2 7 r.
38 C D R A D /P ortugu ese Archives, A N T T , M ucleo A n tigo, 592 f f.2 —5V
[1522].
39 O th er occurances o f kundura, see Ibn Battuta: IV: 135/833 an(I 2 0 8 /8 6 5 .
I fou n d a referen ce to a k.m.n.duri (the /m /p rob ab ly b ein g a copyist e rro r)
in a fifth/eleventh century source as recorded by the ph ilologist, A bu 1-
Q asim in the context o f Iraq (al-A zdi 1902 =107).
4 0 See A gius 1996: 291; C h ristides 1984= 4 6 ; Ahrw eiler 1966: 4 I 4 —5 ?
Fahmy 1966: 150.
41 A l-M aqrizi 1911-24, I: 4 6 9 , 471—4 , 475~ 9 , 4 8 2 - 3 ; II: 154—5.
42 Teak fo r ship b u ild in g was com m on : Periplus 1912: 152; T h eoph rastu s
1916, B oo k V, chapter 4 ; al-M as'u d l 1861—77» ^: 3 ^ 5 -
43 A rabic narjil o r nargil (originally fro m Prakrit nargil through Farsi), referred
to in south ern O m an in the D h ofar regio n where it grows as al-jawz al-
hindi 'the In d ian n u t’ .
44- Now w oods which decay in sea-w ater are eaten by the teredon, It is a
creature sm all in size, but has a large head and teeth; the h arm that these
do is easy to rem edy; fo r, if the w ood is sm eared with pitch, it does not let
in water when it is dragged down into the sea; but the h arm done by the
teredon cann ot be u n d o n e ’ (T h eoph rastu s 1916, B oo k V, chapters 4 ~ 6 ).
45 LA S (St Petersburg), M S S 2 3 .
4 6 B N (P aris), M S A rabe 1825* fo l.IO O r.
47 M arco Polo (1982: 5 2 ); D uarte B arb osa (1918—21, II: 49 ); L udovico di
V arthem a (1863: 152).
A rch ival Sources
ANTT A rquivo N acion al da T o rre do T o m bo (L isb on )

BN B ib lioth eque N ation ale (Paris)
CDRAD C en tre fo r D ocu m en tation and R esearch in A b u D habi
LA S L ibrary o f the A cadem y o f Sciences (St. Petersburg)
UAFA L U niversity o f A lexandria, Faculty o f A rts Library
VOC V ereenigde O o st-In d isch e C om p agn ie (A m sterdam )
7
N a v ig a tio n a l M eth o d s
in t h e In d ian o CEAN 1
Abdul Sheriff
T he long association o f the people o f the Indian O cean with

their sea naturally led to the growth o f an im portant branch o f
knowledge connected with navigation which had enabled them
to sail as far as China in the middle ages. U ntil recently there
has been little study o f this indigenous knowledge, and there
has been a tendency in western scholarship, for example, to
attribute the discovery o f the m onsoons to a G reek navigator,
Hippalus, and to see Vasco da Gam a’ s 'discovery’ o f India, or
the rise o f industrial capitalism from about I 75 °> as having
wiped out whatever there may have been o f the old tradition
and replaced it by new European knowledge. Tibbetts, who has
done an im portant study o f the works o f Ahm ed ibn M ajid, the
fifteenth century Arab navigator, concluded that 'the centuries
old methods o f finding o n e’ s way across the Indian O cean has
been completely lost [my emphasis] in favour o f those methods
learnt on the Atlantic and originally brought to the Indian
O cean by the Portuguese in Ibn M ajid’ s tim e.’ (Tibbetts 1971:
4 6 )2 Villiers, who sailed in a Kuwaiti dhow in I 93&~9 from the
Arabian coast to Zanzibar and the Persian G ulf, said Arab
navigation was at a very low ebb. Arab captains knew their winds
and their waters, but they merely coasted to A frica and India
within sight o f la n d , and made no use o f any navigational aids
to sail across the ocean. (Villiers I 9 4 0: 77 > T98, 2 ^ 0 — i)
209
210 A b d u l S h er iff
However, on his voyage V illiers came across, but did not

examine in any detail, an Arabic navigational directory by a
Kuwaiti nakhoda (captain) Isa al-Q u tam i, (Q utam i 1964) which
is still in use in the G u lf and continues to be reprinted.
Meanwhile the O m an Ministry o f National Heritage and
Culture has collected a num ber o f hand-written navigator’ s
guides from the m ajor dhow port o f Sur, and has published
one in Arabic (K hadhuri 1994 )- M oreover, there are a num ber
o f navigational guides in G ujarati/K utchi and Malayalam
languages in the Indian National Museum at D elhi, mainly
belonging to the seventeenth to the twentieth centuries, which
indicate a com m on source, and give in form ation about
location o f ports, routes, seasons o f sailing, and modes o f
measurement at sea (Arunchalam 1996: 262). A ll these texts
suggest that while the European navigational tradition u n
doubtedly had an impact, there are many elements o f the
indigenous tradition that have been preserved. T h e picture that
emerges is one o f a creative adaptation o f the new to evolve a
synthesis that served the needs o f the Indian O cean m ariners.
A n 'Ideal N avigator’
Consistent with the high degree o f regional intercom m unica

tion, similar titles have been given to navigators right across the
Indian O cean from the East A frican coast to Malaysia, although
there has been some local variation and change over time. Ibn
Majid seems to com bine the functions o f com m anding the ship
with that o f the navigator, but distinguishes the m u’allim as the
expert navigator on ocean -going ships, from the rubban who
piloted coastal shipping especially in the Red Sea. T h e Indian
pilot who took Vasco da Gama from M alindi to Calicut in I 4 9 &
was also called Malemo, and the sixteenth century Indian author
A b u ’l Fadhl describes the m uallim as the captain who must know
astronomy and guide the ship. A t the other end o f the Indian
O cean contem porary maritime laws o f Malacca compare the
position o f the malim with that o f an imam who leads prayers in a
mosque, and says he should be given the same authority
(Tibbetts 1971: 3, 9, 62).
N a v ig a tio n a l M eth o d s in th e In d ia n O cean 2ii
Ibn Majid also refers to the nakhoda, who was not, however,
equated with the m u’allim (Tibbetts I 97 I: 60). A b u ’l Fadhl
clearly distinguishes between the nakhoda as the owner o f the
ship who fixed its course, and the m u’allim who was the pilot.
T he word is o f Persian origin, m eaning lord o f the ship
(Ferrand I 9 2 4 : 238—9)- However, Buzurg’ s tenth century
collection o f sailors’ tales in A ja ’ib a l’Hind refers to the nakhhoda as
a pilot(D evic 1928: 17)- In the twentieth century, the Kuwaiti
dhow in which Villiers sailed did not carry a specialised pilot as
such, and the nakhoda com bined the functions o f captain and
navigator. Presumably when an Indian navigator was employed
on Kuwaiti and other dhows that had to cut across the ocean,
he occupied a tem porary honorary position. In the regular
crew the m u’allim was relegated to the position o f a mate who was
given com m and for shorter stretches, but as soon as the nakhoda
came on board, the m u’allim was merely a stand-by officer who
also kept accounts (Villiers 1940 : 2 3 > 87, 346 ). T h e general
term for captain and pilot throughout the Persian G u lf and
O m an is nakhoda (Hajji 1993 )- This is generally true o f Sur, the
other m ajor dhow port, where authority was often even m ore
centralised, since the nakhoda was often also the owner. This
usage extends to the Swahili coast where the nahodha is captain
and pilot (Prins 197° : 28), and Indian captains com ing to
Zanzibar were also generally known as nakhuda.3
Ibn M ajid set out in great detail the various qualities o f an
'ideal navigator’ which show a lot o f similarity with the
description o f a 'perfect p ilo t’ in an Indian work o f the first
century o f the Christian era. This highlights the continuity and
breadth o f the navigational tradition in the Indian O cean. A n
Indian pilot is described as one who knew 'the course o f the
celestial lum inaries, . . . was skilled in the establishment o f a
given time . . . by means o f m anifold marks, observing the
fishes, the colour o f the water, the species o f the grou nd , birds,
rocks, etc. . . .’ (Tibbetts 1971: I—2; G opal 196^: 324 )-
Fourteen centuries later Ibn M ajid deals m ore systematically
with the attributes o f an 'ideal navigator’ . They included 'the
knowledge o f lunar mansions . . . latitude measuring, . . . the
winds and their seasons, . . . [and] the instruments o f the sh ip .’
212 A b d u l Sh e r iff
Secondly they involved an understanding o f isharat, i.e. signs

and landmarks, including the Various guides such as m ud or
grass, animals or fish, sea-snakes and winds. . . .’ Finally, Ibn
M ajid goes beyond the technical expertise o f the navigator to
include siyasat, proper managem ent o f a ship at sea (Tibbetts
1971:77-8).
O f course not all Indian O cean navigators were as learned in
the highly specialised science o f navigation as Ibn M ajid who
was a renowned m u’allim. In his writings he described three types
o f routes and the corresponding expertise needed to sail them.
T h e first consisted o f m ere coasting which nevertheless
required keen eyesight as well as knowledge o f the coasts,
physical signs and landmarks. T h e second involved cutting
across a gulf, such as the G u lf o f A den, which called for an
accurate understanding o f the bearings o f the ports o f origin
and destination. As Severin points out, sailing from the Persian
G u lf to the western coast o f India or East A frica required only
this degree o f expertise. T h e third demanded an ability to sail
across the ocean by calculating latitudes o f the point o f origin
and destination. (Tibbetts I 97 I: 165—7 » 273—6; Severin
19 8 2 :5 1)-
T k e M aritim e C ale n d a r
T h e most basic requirem ent o f sailing in the Indian O cean that

every navigator was expected to master was o f course knowledge
o f the winds, in particular o f the m onsoons ( mawsim in Arabic,
mosum in H indi, msimu in Swahili, etc.). T h e m ajor seasons were
the northeast m onsoon facilitating exit from the South
Arabian, Persian G u lf and Indian ports between Novem ber
and February, depending on the exact location o f the port; the
southwest m onsoon (kaws in A r., kusi in Sw.) enabling return
from the A frican coasts between A p ril and May before the
m onsoon is fully developed, or at the tail-end o f the season in
August ( demani in A r. and Sw.); and the closed season in June
and July when sailing in the western Indian O cean was too
dangerous. W hile the m onsoons perm itted only a single round
trip between the Arabian and Indian coast to East A frica, the
N a v ig a tio n a l M eth o d s in th e In d ia n O cean 213
alignm ent o f the coast and the direction o f the m onsoons

blowing along it perm itted two-way traffic between India and
the A rabian coast several times during the year. (Tibbetts I 97 I:
3 6 0 -8 2 ; Prins 1970: 7, 15, 20).
The Islamic lunar calendar, with a 354~day year, is not
helpful in working out the sailing seasons, and the navigators o f
the Indian O cean have retained a solar calendar apparently
from the Persian maritime tradition. T he year starts with Nairuz,
which is obviously derived from the Persian new Year, Nauruz,
norm ally celebrated at the spring equinox on 21st March.
However, the 3 6 5 "day m aritim e year did not take into
consideration the shortfall o f a quarter day and the necessity
to adjust it with a leap year every four years. T h erefore the
navigators’ calendar has been slipping by 25 days every century.
In 1488 N auruz fell on 14th November. Ibn Majid and others in
his era were aware o f the attempts to reform the Persian
calendar, but the m ariners nevertheless stuck to the old
calendar, and not always uniform ly, with occasional ad hoc
adjustments to brin g it in line with the sailing seasons.4
The m aritime calendar year was not divided into months;
the num ber o f days was counted from N airuz. Even in the
Zanzibar nautical and agricultural calendar, which also starts
with Nairuzi or S ik u ja Mwaka (Day o f the Year), the year is divided
into decades (mwongo) with supplementary five days to complete
the 3 6 5"day year, although the total num ber o f days is also
m entioned (Gray 1955 : 2). Ibn M ajid gave the most favourable
days to start and complete voyages from different ports around
the Indian O cean counting from N airuz, and this gives an
indication o f the extent o f sailing in the Indian O cean in the
fifteenth century. Thus, for example, he recom m ended
departure from A den and the Y em en to the Zanj country
(East Africa) from the 320th to 330th day o f N airuz (8th to 18th
O ctober), from H orm uz on the 7 ° th (31st January), from
India up to the 80th (iO th February), and from Sumatra for
the A frican coast from the 60th day (21st January) etc.
(Tibbetts 1971: 2 3 4 > 360—2). The Indian manuscripts also
prescribe dates fo r the com m encem ent o f voyages from
different ports (Arunchalam 1996: 178).
214 A b d u l Sh er iff
Despite all the changes that have occurred during the past
five centuries, N airuz continues to guide sailing in the western
Indian O cean. A nakhoda in Sur, Said b. Ham ed al-Fanna,
emphasised that unless one knew the exact date o f N airui, one
could not go to sea and use the sailor’ s guide that Suri dhows
used to sail to India and the East A frican coast.
I s h a r a t/A la m a t
T h e second basic requirem ent o f navigation was p roper
identification o f signs and landmarks (isharat) . Ibn Majid says
it is not very scientific because it depended so m uch on the
experience and expertise o f individual navigators, but it was
nevertheless very useful to a navigator. He says that the
navigator 'should know all the coasts and their landfalls and
their various guides such as m ud, or grass, animals or fish, sea-
snakes and winds . . . consider the tides, and the sea currents
and the islands on every ro u te .’ (Tibbetts I 97 I: 77 > !92)
The most obvious is the configuration o f the coastline, with
mountains, capes and islands serving as the most com m on
landmarks, and in places, sailors set up cairns or lighthouses
where natural landmarks were missing. In Q utam i’ s directory
and those collected by the O m an Ministry o f National Heritage
and Culture, drawings o f these geographical features were
inserted in the text and tables, and nakhodas were able to identify
them instantly. In fact the drawings in Khadhuri and Q utam i are
almost exactly the same (Khadhuri 1994: 224> 233—4 - Qutam i
1964: 103, 114—5). In the Indian manuscripts these landmarks
sometimes carry latitudes and longitudes which were added later,
as well as their stellar bearings (Arunchalam 1996: 275 )*
Ibn M ajid also talks o f sand and m ud from the bottom o f the
sea as a guide, brought up by the bild (plumb line), and there
are reports o f nakhodas, even blind ones, feeling the texture o f
the m ud and sm elling it to identify the location. D iscoloration
o f water helped navigators recognise river mouths. A seven
teenth century Gujarati pothi (guide) explained in great detail
distance fro m shore by d iffe re n t colou rs o f the water
(Arunchalam 1996: 268). Seaweeds and sea grasses were also
helpful indicators. Ibn M ajid m entions a marine bank under

the water which produced seaweeds found on the north coast o f
Socotra. A m ber was collected near the coast o f Somalia, but
Ibn M ajid was not clear about its origin. (Tibbetts I 97 I: *97 )
Ibn Majid also describes different kinds o f birds seen at
different places to help the navigator identify his location, but
he says sometimes they appear, but sometimes they do not,
'there being no rule about it.’ (Tibbetts I 97 I: J97 )* Finally, sea
snakes are m entioned in specific areas, knowledge that
apparently continued into the later period. T he Indian guides
state that snakes are seen at 20 iams (see below) from the Indian
shore. In 1763 N iebuhr reported that 'when these serpents are
seen, they are an indication that the coast is exactly two degrees
distant’ (Bowen I 95 I: *79 - M ookerji 1957 : 5 2 ;- Arunchalam
1996: 268, 270).
T he use o f 'sh o re-fin d in g’ birds does not strictly fall under
this heading, but it was an old practice in the Indian O cean. A
Buddhist record o f the fifth century before Christ relates that
sailors used them, as does the Rigveda, the H indu scriptures, and
it is also m entioned in Pali literature. Pliny m entions this
practice among sailors o f Sri Lanka in the first century AD.
(Bowen 1951: 164* Arunchalam 1996: 27 ° . G opal 1962:
3 2 3 -4 )
The K a m a l
Central to Ibn M ajid’ s navigational theory is o f course the

ancient knowledge o f astronomy, which was developed to a high
degree during the classical period o f Islam, and the instru
ments used to measure the stars’ movements and their
direction. A sixteenth century O ttom an m iniature shows
M uslim scientists working in an observatory equipped with a
num ber o f astronom ical and navigational instruments (Aramco
43 /3 > back cover). Curiously, these instruments are never
clearly described in Ibn M ajid’ s texts o f practical navigation,
presumably because m u’allims were expected to know them.
T h e compass first appeared in Islamic literature in 1 2 32 A D ,
although its use at sea, either as a floating magnetised fish or a
2 i6 A b d u l Sh e r iff
needle suspended by a string, may have come somewhat later.

However, unlike the western compass, the various points on the
compass rose are associated with the rising and setting o f
prom inent stars (Tibbetts I 97 I: 3 > 29 °> 295—6). Stars were
used to describe directions even before the introduction o f the
compass, at least as early as the ninth and tenth centuries, and
many o f the A rabic names have a desert origin, while others are
Persian. T h e star compasses o f Arabia, Persia, coastal India and
the Laccadive islands are very similar, though those o f Gujarat
are different, and many o f the Swahili names for stars are
similar to A rabic names (Arunchalam 1996: 2 J I ; Tolmacheva
1995 : I^)- W hen the compass was introduced, the older star-
based system o f names was retained in the new system based on
magnetism to tell direction, an early example o f creative
adaptation o f a new discovery (D im m ock 1944 : I 56 , 16O;
Princep 1836: 144 ; Ferrand I 9 2 4 : 235 )-
Th e second vital instrum ent is one that could measure the
position o f the ship. Ibn Majid often m entions qiyas for star
measurement, and also khashabat (pieces o f wood) to measure
latitude. Fatimi shows that a similar instrum ent under different
names may have been in use since the ninth century. T h e first
to describe the kamal was a Turkish navigator Sidi A li b. Husain
C elebi, in his work on oceanography, the M ohit (The A ll-
Encompassing Sea), which was based on the works o f Ibn Majid
and Sulaiman al-M ahri and written in India in 1554 (England
1 9 5 5 :1 6 5 ).
T h e kamal consisted o f a num ber o f oblong pieces o f wood o f
different sizes with a string or different strings passing through
their centres. To measure the height o f Polaris in northern or
one o f the Bears in southern latitudes, the string was held
between the teeth with the wood at such distance from the eye
that while the lower edge was in line with the horizon, the
upper edge just touched the star. T h e latitude was calculated in
terms o f numbers o f finger widths (isbas) above the horizon
(Fatimi 1996: 290; A lbuquerque 1988: IO—2; O m an I 99 I:
9 7—IOO; P rin cep 1836: 13 9 —43)- T h e same system o f
measurement by finger widths occur in G ujarati/Kutchi and
in Malayalam guidebooks, although it is called by their
vernacular terms. It appears to have been im proved upon by

replacing the different pieces o f wood with a single tablet, and
the single string was calibrated by a num ber o f knots indicating
different latitudes. Indian O cean navigators, especially from
the Maldives, continued to use such an instrum ent into the
nineteenth century, and a specimen was obtained at Calcutta in
the 1890s for the H am burg Ethnographic Museum. It was
further im proved upon by replacing the string with a calibrated
rod (j a i or gaz, yardstick) to remove the slack o f the c o r d , with a
sliding block o f w ood, and this came to be known as ballisti
(Arunchalam 1996: 2 & 4 ; Tibbetts 197I: 2 7 2 , 313, 315—8).
Retracing the great medieval voyages to China in the O m ani
dhow Sohar in the 1980s T im Severin found m easurement from
such a simple instrum ent remarkably accurate (Severin 1982:
93) (see fig. 7.1).
A sim ilar board may have been made by a Chinese
instrum ent maker, Ma H uai-te, who called it a 'guiding-star
stretch b oard’ which consisted o f twelve boards. Fatimi suggests
that he may have been related to Ma Huan, the chronicler o f
the voyages o f C heng H o, the famous Chinese admiral. He may
have accom panied the expeditions in the early fifteenth century
as far as India and even East A frica, and may have learnt to
make the instrum ent in the Arabian Sea. If so, Fatimi says it is
an indication o f the 'waves o f . . . cross-cultural currents’ that
extended right across the Indian O cean as far as C hina (Fatimi
19 9 6 :2 8 9 ).
Indian O cean mariners, however, were not able to calculate
longitude before the introduction o f the sextant. T h erefore, a
voyage between two points often involved what was called
'parallel sailing’ , i.e. sailing first to the latitude o f the
destination, and then changing course in m id-ocean to sail
due east or west until landfall was made, or some com bination
o f these manoeuvres according to the wind. Speed o f the dhow
was calculated by dropping a floating object at the stem o f the
dhow, and calculating the time it took to reach the stern. East-
west m easurement o f distance was calculated in practice in
terms o f the zcim (watch), a standard three hours o f sailing. The
same term and measure o f three hours is found am ong many
2 i8 A b d u l S h e r if f
Foie star
H orizon
Pole star
H orizon
Figu re 7.1 The kamal and the cross-staff
sailors o f the Indian O cean from A rabia to Indonesia

(Arunchalam 1996: 263). T h e distance between Mombasa
and Sunda in Java, for example, was given by Ibn Majid as 306
Z a m s which translates to about 61 degrees. This is remarkably
close to the actual distance o f about 66 degree, spanning m ore

than 4 ,5 0 0 miles between them (Tibbetts I 97 I: 3 ^4: ~ 5 > 355 »
359). A t some stage a contraption, consisting o f a triangular
board tied to a rope was developed to measure the speed o f the
dhow. It was dropped from the stern o f the dhow. A specimen
o f this instrum ent is on exhibition at the Sur M aritim e
Museum and is illustrated in Q utam i (Q utam i 1964^ 75 )-
N a v ig a t io n a l M e t h o d s in t h e In d i a n O c e a n 219
The European Impact

T ibbetts’ rem ark that the pre-E uropean m ethod o f finding
o n e’ s way across the ocean was completely lost in favour o f the
Atlantic system introduced by the Portuguese (Tibbetts I 97 I:
46) seems to assume that the Europeans had already discovered
the m ethod and instruments for this type o f navigation. Vasco
da Gama had attempted to use an astrolabe to discover
longitude at sea, but it was very inaccurate (Encyclopaedia
Britannica, 191O: 19/284—90). To take him from M alindi to
India, he was forced to rely on an Indian pilot who used the
above-m entioned kam al Vasco da Gama took the kamal to
Portugal in 1499 > and Portuguese pilots were com m issioned to
experim ent with what they called 'Tavoleta da India’ ('little
boards o f India’) during C abral’ s voyage in 1500 which
'discovered’ Brazil. They eventually m odified it to read degrees
directly instead o f the standard finger widths o f the Arabs.
Vasco da Gama was also very pleased to see a chart o f 'the entire
coast o f India, set out in M oorish fash ion.’ A lbuquerque shows
that the so-called C antino map o f 1502 incorporated a lot o f
inform ation on parts o f the Indian O cean not yet visited by the
Portuguese from local charts, rutters and oral inform ation.
Thus, if any borrow ing occurred, it did so at least initially in
the opposite direction (England 1955 : I^5 ; A lbuquerque 1988:
2 9 -3 6 ).
T h e sextant was not invented until I 731 and the chronom eter
in 1735» and they mark the beginning o f m odern navigation.
W ith European dom ination o f the Indian O cean by that time,
and even construction or purchase o f E uropean-type naval
vessels by O m an and other Indian O cean powers during the
eighteenth century, it is not surprising that some aspects o f
European navigation were also adopted and adapted to the
existing tradition in the Indian O cean. T he chart o f the Indian
O cean that V illiers found in the hands o f a form er navigator,
dating to 174^, may have been based on western navigational
charts, but it was decorated with A rabic script and had
landmarks, distances and other inform ation o f im portance to
the Arab m ariner.
220 A b d u l S h e r if f
Villiers says that a copy o f a shipm aster’ s directory that he

was given in Kuwait was illustrated with drawings o f frigates and
not dhows. He says it had apparently been produced in Muscat
and had been brought to Kuwait, but he may have been
referring to the directory by Q utam i which was first published
in Kuwait only in 1915* A revised copy o f this directory, which
is still on sale in the Persian G ulf, shows dhows as well as
European-type sailing vessels that have long disappeared from
the western Indian O cean (Villiers 194O: 3 ° 6 ). T h e h an d
written O m ani directories are also illustrated with both dhows
and European sailing vessels. T h e use o f both types o f vessels to
explain the position o f the vessel in relation to the sun is a fine
example o f adaptation o f new knowledge to serve the needs o f
Indian O cean m ariners.
M ore to the point is perhaps the manuscript captured by the
British Navy on a slave dhow in 1844» now the British
Museum. Tibbetts says it shows the dhow captain used a sextant
to shoot the sun rather than depending on the stars (Tibbetts
1971: 46.)* Q utam i’ s and the O m an i directories, as well as the
Indian guides give latitudes and longitudes o f different places
around the western Indian O cean. Many o f the navigators also
used western printed charts. Many o f them have preserved and
are very proud o f their sextants as a sign o f their navigational
com petence. They call it kamal, transferring an old name to a
new instrum ent. Contrary to V illiers’ conclusion, they show
that Arab and Indian nakhodas from at least some o f the ports
were able to use these instrum ents and guides to sail across the
Indian O cean well into the twentieth century.
T h e Suri nakhodas stated, as did the Kuwaiti in V illiers’ dhow,
that they did not need the sextant when sailing to the East
A frican coast which ran parallel to the m onsoons, and that a
slightly straighter course across the sea saved very little time to
be worth the risk (Villiers 194 ° : Z j O — l ) . Several nakhodas even
in Sur and on the Batinah coast o f O m an close to the m outh o f
the Persian G u lf admitted that they did not know how to use the
sextant. However, the Suris were emphatic that an expert
navigator able to use the sextant was needed to cut across the
A rabian Sea from Sur to the west coast o f India, and even m ore
from the Indian coast directly to A frica. They often sailed from
the K erala port o f M angalore to the East A frican coast with
cargoes o f the famous red ro o f tiles.
These navigators used two types o f sextants, the apparently
simpler tw o-arm ed 'K u tch i’ (Indian) and the typical western
type, and the Sur M aritime Museum has specimens o f both. A
nakhoda at Ras al-K haim a within the Persian G u lf proudly
displayed his sextant and charts, and a photocopy o f Q utam i’ s
directory. Kuwaitis and others who did not know how to use the
sextant themselves hired Indian, Suri or Baluchi navigators to
cross the sea, and there were some Baluchi navigators at Saham
on the Batinah coast o f O m an, although not all were reliable.
A n Indian navigator was hired at K arachi by a nakhoda at Saham
to take him from Mangalore to the A frican coast, but in fact
they landed on the southern coast o f O m an despite all his
calculations, and naturally he was fire d .5
S iy a sa t
The final com ponent o f Ibn M ajid’ s attributes o f an ideal
navigator goes beyond the technical aspects o f navigation, and
touches on proper management o f the ship, or what he calls
siyasat (policy or nautical etiquette). A captain o f a ship in many
cultures is the ultimate authority at sea, traditionally com bining
many functions from commanding the ship to being a judge.
The Arabic term nakhoda comes from the Persian word naukhuda
which literally means lord o f the ship. The nakhoda was not only
captain in the usual sense, but also generally a trained navigator,
and in some cases, as in Sur, he was often also the owner. He was
ultimately responsible for the well-being o f the dhow and its
occupants. He was expected to be adept at running the ship. Ibn
Majid elaborates at length on his managerial and leadership
qualities. He emphasises patience in time o f fatigue and when
'arbitrating among m e n , and says he should be 'gentle in his
speech’ . He should be a good judge o f people, passengers as well
as crew, assessing them carefully and thus being prepared for
them in time o f trouble; listening to 'everything they said,
accepting anything good and rejecting the bad’ . He should be
firm and strong when speaking to them but always pleasant. He

should not fraternise with somebody who disobeys on matters
within his competence. He should not encourage backbiting.
The captain was also expected to be learned. Ibn Majid says he
should be literate, 'knowledgeable and learned in many things’ ,
and judging by the num ber o f books that he quotes, from
Ptolemy and the Persian poet Firdausi (whom he quotes in
Persian) to the famous Arab geographers and astronomers, Ibn
Majid lived up to that injunction. The captain should be just and
o f sound judgem ent. 'He should not anger the merchants over
regulations except in matters where they interrupt him in his
duties . . . If he has not these qualities’ , says Ibn Majid
emphatically, 'then he can never be the ideal navigator’
(Tibbetts: 7 7 -8 , 33, 3 9 -4 1, 58, 60, 73, 19 2 -3 , 270).
In contrast to Ibn M ajid’ s books, Q utam i’ s and K h ad h u ri’ s
twentieth century directories do not specifically deal with siyasat.
Judging from A lan V illiers’ account o f the voyage in a Kuwaiti
dhow in 1938/9, there was a well established tradition on this
subject that apparently did not need to be enunciated in a
book. Although the nakhoda in V illiers’ dhow may not have been
as learned as Ibn M ajid, he discussed diverse subjects from the
growth o f Islam in Japan and the great days o f Arab science, to
property laws in Hadhramaut. A form er nakhoda on the same
dhow was m ultilingual as were many other nakhodas, being able
to speak Swahili, Persian, Somali, H industani and several
Indian dialects, as well as Arabic, and some knew some English
as well. (Villiers 194 ° : 9 1* J4 9 )
Th e jo b description o f a nakhoda highlights his m anifold
duties that went beyond the mere technical com m and and
navigation o f the dhow. He was expected to be a master
shipwright, sail maker and rope maker, and an expert appraiser
o f dhows. He controlled spending on board, although the
m u’allim was required to keep accounts. To the rest o f the crew
he was a benevolent patriarch, or father, although sometimes
younger than many o f the sailors. In relation to the passengers
he was judge and ju ry in all disputes, and although the Sayjids
(descendants o f the Prophet) on board were sometimes allowed
to intervene in order to solve disputes, his word was supreme.
He ruled with a firm hand. O n the occasion o f a dispute

between a blind beggar and his two u n d er-fed boys, he got the
beggar to feed them properly. W hen three wayward Swahili
women from Lamu raised hell about conditions in the cabin,
unlike their Arab sisters who had endured the whole voyage, he
merely clamped down the hatch, locked it, and paid no
attention. A ccording to Villiers, nobody opened his m outh
except to repeat what the nakhoda had said. Passengers were
never regarded as correct, having been allowed on board being
privilege enough, a claim possibly exaggerated by Villiers
(Villiers 194O: 8 2 -3 , 85, 8 7 -8 , 16 8 -9 .).
T h e nakhoda was also the ch ief m erchant on board, trading
not only on his own account, but also on behalf o f the owner o f
the dhow. He was expected to be a judge o f quality and prices o f
all goods and their best markets, and to select the best agent in
the various ports visited by the dhow. It was his duty to find
passengers, even if it meant stealing a march on his rivals at the
next port by whatever means at his disposal. To further the
business o f the dhow he was expected to entertain at the
different ports, inviting local merchants, agents, passengers
and local notables (Villiers 194O: 23, 5 ^)-
C o n c lu sio n
T h e navigational methods used by the Indian O cean m ariners

show an interesting history o f syncretism by which they have
developed, adopting and adapting new methods to fit the
existing tradition and facilitate navigation. T h e tradition was
not confined to any single nationality, given that com m unica
tion between the different nationalities had been almost
continuous for several m illennia. This process may have begun
during the period o f Persian dom ination over the Indian
O cean in pre-Islam ic and early Islamic times when the Persian
calendar and m uch Persian nautical term inology may have been
adopted. C ontact with C hina was strengthened with the
expansion o f com m erce at the height o f the Islamic empires,
providing an avenue for transmission o f the compass to the
M iddle East. However, while the principle o f magnetism was
adopted to find the direction o f north, the older m ethod o f

identifying different directions by reference to the most
prom inent stars was retained in developing the Arab compass
rose. Finally Indian navigators influenced the writings o f such
m uallim s as Ibn M ajid (Tibbetts I 97 I: 3^)-
The entry o f the Portuguese in the sixteenth century certainly
inaugurated a new age in the history o f the Indian Ocean, but in
terms o f navigational methods, rather than an immediate
copying o f the Atlantic methods by the Indian Ocean mariners,
there may have been a case o f a reverse adoption by the
Europeans o f an indigenous instrum ent and m ethod o f
measuring latitudes, the kamal. The discovery o f the sextant in
the middle o f the eighteenth century, by which the longitude
could also be calculated by shooting the sun, revolutionised
navigation not only in the Atlantic but also in the Indian Ocean.
It is not surprising that the Indians and the Arabs eventually
adopted this instrument and named it after their old instru
ment, the kamal. But the name was not the only thing that was
retained; they continued to use it in com bination with the local
maritime calendar, the Chinese compass to find direction, and
the Arab compass rose based on the most important stars.
T h e Indian O cean navigational methods, however, went
beyond the technical and scientific aspects o f finding on e’ s way
across the ocean. It included an intimate understanding o f the
natural as well as hum an environm ent to guide a dhow to its
destination. A knowledge o f the m onsoons, o f the signs and
landmarks, the seaweeds, the snakes and the birds, was a vital
part o f the Indian O cean navigational tradition that had
developed in situ and could hardly have been swept aside by any
introduction from the Atlantic. That tradition was personified
in the nakhoda who typically com bined the functions not only o f
the captain and the navigator, but also often o f the ch ief
m erchant and sometimes even the owner. The tradition,
whether written down as in the books o f Ibn M ajid, or in the
lore o f the mariners o f Sur and Kuwait, was a living tradition
that has grown, not merely by borrow ing but also through the
long experience o f the Indian O cean m ariners in their
environm ent.
Notes
1 T his is part o f a research project on the history o f the dhow in the
Indian O cean which has been generously funded by the Centre for
Developm ent Studies, University o f Bergen. I would like to express my
sincere thanks to Professors Sean O ’Fahey and L e if Manger for
supporting the project.
2 Both Tibbetts 1 9 7 1 : 9 — I I and K houry 1983: 17 —22 have conclusively
shown that Vasco da Gama was taken to India not by Ibn M ajid but by an
Indian mu’allim or pilot, as the contemporary Portuguese sources have
always indicated. Tibbetts, however, adds that da Gama was piloted by
'some man from one o f the continually warring Indian states hoping to
boost his own country’ s fortunes at the expense o f a neighbour. T h e Arab
who had everything to lose by introducing more com petition would never
have done such a th in g .’ This is a strange, almost a 'colon ial’ argument.
A ccordin g to Tom e Pires, the Indian O cean was dominated by Gujaratis
by the beginning o f the sixteenth century, who stood to lose as much as
the Arabs (Lewis 1973 : 2 4 3 ” 6).
3 A n Indian comm unal benefactor in Zanzibar is described as a nakhuda,
and there used to be a restaurant called Nakhuda H otel in the same area.
4 Ferrand 1924: 2 3 0 - 4 , 24 9 ~ 50 . Tibbetts 197I: 44, 3 6 1 -3 . A t this rate, if
the slippage had been regular, the calendar may have been adopted some
950 years earlier, i.e. in the middle o f the sixth century when the
Sassanid Persians were expanding in the Indian O cean towards Ethiopia
and the Yem en in the west, and had an upper hand in the commerce o f
the Indian O cean ranging as far as China. G rosset-G range 1972:47 notes
that Nairuz corresponded to 4th August in Somalia, I2th in Qatar and
18th in Bahrain. He concluded that the dating o f Nairuz astronomically
had been abandoned.
5 Muhammad b. Hamed b. Najim al-G hailani, interviewed at Sur,
10 . 12 . 1997 - Kham is b . Salim al-Shidi, interviewed at Saham, 2 1 . I I . 1997 *
Bibliography
A lb u q u e rq u e , L . de (1988) Instruments o f Navigation, L isbo n : N a tio n a l B oard
fo r the C e le b ra tio n o f the Portuguese D iscoveries.
A ra m co W o rld , The Middle East and the Age o f Discovery, V o l.43/3*
A ru n ch ala m , B. (1996) 'T ra d itio n a l Sea and Sky W isdom o f In d ian Seam en
and th eir P ractical A p p lic a tio n s’ , in H . P. Ray and J .- F .
Salles, eds., Tradition and Archaeology: Early Maritime Contacts in the Indian Ocean,
N ew D elh i: M an o h ar, pp. 26 1—8 l.
B ow en, R. L e B a ro n (1951) 'T h e D how S a ilo r’ , American Neptune I I : 161—2 0 2 .
D evic, L . M . (1928 ) The Book o f the Marvels o f India, L o n d o n : G eo rg e R outled ge
& Sons.
D im m ock , H . L. F. (19 4 4 ) T h e poin ts o f the compass: A ra b ia n style’ ,

Mariner’s Mirror 3O: I 54—6 l.
Encyclopaedia Britannica, (19 10 ) E leven th E d ., N ew Y o rk : E n cyclo p aed ia
B rita n n ica C o .
E n g lan d , B. ( l 955) 'T h e K a m a l’ , Mariner's Mirror 41: 165—6.
Fatim i, S. Q . (1996) H istory o f the D evelop m en t o f the K a m a l’ , in Ray and
Salles, eds., p p . 2 8 3 —92.
F erra n d , G . (1924) L ’ elem en t Persan dans les textes n au tiqu es A rabes des
X V iem e et X V Iiem e siecles’ , Journal Asiatique, A v ril-J u in .
F re e m a n -G re n v ille , G . S. P. (1962) The East African Coast, O x fo rd : C la r e n d o n
Press.
G o p a l, L. (1962) A rt o f sh ip b u ild in g and navigation in a n cien t In d ia ’ ,
Journal o f Indian History 40 :313 —2 7 -
G ray, J. M . ( l 955) 'N a iru zi o r Siku ya M waka’ , Tanganyika Notes & Records 37 :
I—21.
G ro sse t-G ra n g e , H . (19 72), C o m m e n t naviguent a u jo u rd ’h u i les A rabes de
l ’ O c e a n I n d ie n ? ’ Arabica, 1 9 :4 6 —77-
H ajji, Y . Y . al- (1993) Nuwakhudha al-Safar al-Shira fi al-Kuwait, K uw ait: Shirkat
a l-R a b i’an.
K h a d h u ri, N asir b. A li b. N asir al- ( l 994 ) Madin al-A srarfi 'Ilm al-Bihar ('T h e
M in e o f Secrets in the Scien ce o f the Seas’), edited by H asan Saleh
Shahab, M uscat: M in istry o f N a tio n a l H eritage and C u ltu re .
K h o u ry , I., tr. (1983) As-Sufaliyya 'The Poem of Sofala’ by Ahmed Ibn Magid,
C o im b ra : Ju n ta de Investigacoes C ie n tifica s do U ltram ar.
Lewis, A . (1973) 'M aritim e skills in the In d ian O cea n , 136 8 —1 5 0 0 ’ , Journal of
the Economic and Social History of the Orient, 16: 2 3 8 —64.
M o o k erji, R. K . ( l 957) Indian Shipping, B om bay: O r ie n t Lon gm an s.
O m a n , Sultanate o f (1991) Oman: A Seafaring Nation, M uttrah: M in istry o f
N a tio n a l H eritage and C u ltu re .
P rin ce p , J. (1836) 'N o te o n the nau tical in strum en ts o f the A ra b s’ ,
re p rin te d in F. Sezgin ed. Islamic Geography, F rankfurt: G o e th e U n iversity,
v o l.15: I 3 9 -5 I -
Prins, A . H . J. (19 70 ) A Swahili Nautical Dictionary, D ar es Salaam : C h u o cha
U ch u n gu zi wa Lugha ya K isw ah ili.
Q u ta m i, Isa al- (19 6 4 ), Kitab Dalil al-Mukhtar f 'Ilm al-Bihar, th ird re p rin t,
Kuw ait: G o v ern m en t Press.
Ray, H . P. a n d J .- F . Salles, eds., (1996) Tradition and Archaeology: Early Maritime
Contacts in the Indian Ocean, N ew D elh i: M an ohar.
Severin , T im (1982) The Sindbad Voyage, H u tch in so n , L o n d o n .
T ib b etts, G . R. (I 97 1) Arab Navigation, L o n d o n : Luzac.
T olm acheva, M . ( l9 9 5 —6) Essays in Swahili G eo gra p h ical T h o u g h t’ , Swahili
Forum II & III.
V illie rs, A . (19 4 0 ), Sons of Sinbad, New Y o rk : C h a rles S c rib n e r’ s.
8
T he R ise and Fa l l of
H ad hrami S h ipp in g
in the I n d ian O cean ,
c.1750 - c.1940
William Gervase Clarence-Smith
T h e m iddle o f the nineteenth century was a golden age for

H adhram i Arab shippers, both those born in Hadhramaut
[eastern Yem en] and those o f mixed race born in the diaspora
(Freitag and C larence-Sm ith 1997 )- Indeed, shipping was
probably the most profitable o f the wide range o f Hadhram i
activities in the Indian O cean around 1850, even though close
links between shipping and trade make it difficult to disen
tangle the two. Despite the rise o f steamer navigation from the
1860s, H adhram i sail shipping did not decline markedly before
the l880s. M oreover, initial H adhram i attempts to break into
steam navigation met with moderate success.
A crisis point was reached around 19OO, although the tim ing
varied in different parts o f the Indian O cean. Despite a b rief
respite due to the First W orld War, a segmentation o f shipping
into three tiers became increasingly apparent (Broeze 1996).
Ever fewer and larger Western companies controlled long
distance scheduled lines, with ever bigger and costlier o il-fired
ships, ever m ore incestuous cartels, and ever closer governm ent
support. The second tier consisted o f tramp steamers, often
older vessels, which picked up cargoes wherever they could.
Asian shippers were fairly well represented at this second level,
but they tended to be Japanese, Chinese or South Asian. The
third tier consisted o f sailing ships, often o f local construction,
227
228 W il l ia m G e r v a s e C l a r e n c e - Sm it h
which survived repeated predictions as to their im m inent

disappearance from the high seas. Arab shippers were quite
num erous at this level, at least along the great arc from
Zanzibar to Calicut. However, O m an i and Kuwaiti Arabs were
m ore prom inent than their H adhram i cousins.
A lth ou gh H adhram aut had its own harbours, notably al-
M ukalla and al-S h ih r, the rise o f H adhram i shipping owed
little to these hom e ports. Nowhere was this clearer than in
the Malay w orld, the most profitable zone o f all by 1850.
Ind eed, many o f the most successful shippers and captains in
these waters came from the thickly populated in terio r o f
H adhram aut, and had no p rio r experience o f the sea (Berg
1886: 187—8). T h ey exploited evolving W estern technology to
develop the 'country trade’ , m iddle distance operations o f a
regio n al natu re. Links to H ad h ram i ports were m ore
p ro n o u n ced in the western Indian O cean , as was the use o f
m ore traditional types o f Indian O cean vessels, but H adhram i
shippers again concentrated on regional routes away from
their own ports.
Intense com petition, especially from neighbouring O m an,
shaped the excentred and regional character o f Hadhram i
shipping. O m an i entrepreneurs enjoyed the inestim able
benefit o f support from the rulers o f their growing maritime
em pire, based in Masqat and Zanzibar. T h e lords o f Masqat,
m ajor shippers in their own right, aggressively expanded the
O m an i m erchant and war fleet from the eighteenth century.
N eutrality in the wars o f I79 3 —I 8l5 made this strategy
particularly profitable, as the French withdrew and the British
were forced to make trade concessions. T h e O m ani purchased
large square rigged ships seized by the British as prizes, and had
others built for them in Bombay. In C ochin, Bombay and
A rabian ports, they also bought baghla, ships made o f Malabar
teak, two masted, lateen rigged, and o f 70 to IOO tons burden.
However, O m ani attempts to break into the trade o f eastern
Asia and the West proved unsuccessful. M oreover, peace from
1815 saw some shipping functions passing to Western firms,
while others slipped into the hands o f South Asian com m u
nities and Arabs from the Persian G u lf (N icholls 1971; Bhacker
T h e R is e a n d Fa l l o f H a d h r a m i S h ip p in g 229
1992; B enjam in 1976; Fattah 1997 ; Subram anian 1996;

Clarence-Sm ith 1989b).
C om petition from other Arabs and South Asians was weak in
the Malay w orld, but Hadhramis nonetheless had to overtake
European, Chinese, A rm enian and indigenous shippers to rise
to prom inence in the m id-nineteenth century. From their
original base in waters bounded by Malaya, Sumatra, B orneo
and Java, Hadhramis pushed further afield, at times settling
beyond the Dutch and British possessions. To the east, they
reached Portuguese T im o r and New Guinea. To the north,
they went to Th ailand [Siam], C am bodia and southern
Vietnam . The Philippines and East Asia were occasionally
visited, but these routes were dom inated by Chinese, Japanese,
and Western ships (C larence-Sm ith 1997 ; Berg 1886).
In the western Indian Ocean, H adhram i shippers did best in
the Red Sea, outside the O m ani em pire. Leaving long distance
routes to Western steamers, the Hadhramis overwhelmed
O m ani, G u lf Arab, Som ali and Gujarati com petition in the
southern reaches o f the Red Sea. They also extended their
activities northwards to Suez, at the expense o f shippers based
in Cairo (Ewald and C larence-Sm ith 1997)- T h e C om oro
islands, outside the O m ani empire and ruled by H adhram i
lineages, form ed a modest shipping hub in eastern Africa,
despite much com petition from O m ani, Gujarati and E u r
opean rivals (Shepherd 1985). Hadhramis m aintained precar
ious toeholds in western India, notably on the Malabar coast,
faced with a plethora o f South Asian, Arab and European
com petitors (Dale 1997 )* In Hadhramaut itself, shipping
rem ained largely in foreign hands (Ingrams 1937)*
Although long distance shipping links between the principal
zones o f Hadhrami activity were exceptional, people and ideas
flowed freely through the networks o f the far-flung Hadhrami
diaspora. Gill Shepherd has suggested that only sayyid lineages,
proud o f their descent from the Prophet, were able to maintain a
sense o f corporate identity across continents and generations
(Shepherd 1985)* Sa^yid supremacy in Hadhrami shipping circles
was indeed striking, and members o f these lineages were frequently
spread around the Indian Ocean. However, there were shippers o f
230 W il l ia m G e r v a s e C l a r e n c e -Sm i t h
other social origins, especially numerous in the southern Red Sea,

who may also have form ed widespread networks.
The sailin g boom in the Malay world

Hadhram i m igration to the East Indies grew as Dutch power
declined in the course o f the eighteenth century, and some
early shipping ventures were o f a frankly piratical nature.
Sayyid 'A ll b. 'Uthm an b. Shihab, o f a m ajor Hadhram i family,
seized the East Sumatran sultanate o f Siak in 1791 and was
described as the scourge o f the Straits o f Malacca (Trocki 1979 :
28—9; Berg 1886: 197 > n - 4 )- However, deep divisions over the
issue o f piracy were apparent within the Hadhram i comm unity.
T he pious 'alim Sayyid Hasan b. Ahm ad a l-Q a d ri [Algadri]
broke o ff all relations with his piratical son 'Abd al-Rahm an,
who in I 771 established a new sultanate in Pontianak, West
B orneo. Furtherm ore, Arab shippers engaged in legitimate
business provided valuable inform ation to European naval
forces, and were themselves targets o f piratical attack (M oor
1968: 21, 24> 30—I > an<i appendix: 101—4).
T he port records o f Java show that although Arabs accounted
for only 2% o f sea captains in 1774—77» they operated the
largest vessels after those o f the Dutch East India Com pany.
M oreover, Arab captains covered longer distances than any
com peting group, again with the exception o f the Com pany.
Arabs specialised in sailing between Java and the Straits o f
Malacca, especially Palem bang and Malacca [Melaka]. A
specifically Hadhram i involvem ent on this route dated back at
least to the m id -l7 5 0 s. T h e average size o f all ships owned by
Arabs was just over 5 ° tons, but some o f their vessels were
much larger. Arab captains resident in Java were mostly to be
found in the central and eastern ports o f the north coast,
notably Semarang, Pekalongan, Gresik and Surabaya. Crew
members were mainly Javanese, and were probably free rather
than slave (Knaap 1996: 65—73 » &3 > 212—3)-
T h e ships owned by Arabs were a mixture o f types, at a time
when older kinds o f Southeast Asian ships were giving ground
to ones o f European design, albeit mainly built in local yards.
T h e R is e a n d Fa l l o f H a d h r a m i S h ip p in g
A small num ber o f Arabs owned m edium -sized ships in the

European tradition, brigantines and barks o f up to 100 feet in
length, o f some 2 0 0 tons burden, and carrying a crew o f
around 3 ° m en. They were decked, steered with a central
sternpost rudder, and had two or m ore masts, square rigged for
the most part, but with fo re-an d -aft rigging on the rear mast.
T h e older lateen system o f fo re-an d -aft rigging was giving way
in these years to 'spanker’ or 'driver’ sails, which did not
overlap the mast and had a boom to spread the bottom o f the
sail. T he most popular ship with Arabs was the chialoup or
shallop, apparently the forerunner o f the schooner. It was up
to 75 feet in length, o f some 8o tons burden, with an average
crew o f 2 0 , single decked, single or double masted, and fore-
and-aft rigged. A central rudder was the norm on a chialoup, but
some were built in Java with two lateral rudders. A bout h alf the
ships operated by Arabs were smaller Malay craft, pencalang,
gonting, and mayang. They were up to 6o feet in length, between
10 and 25 tons burden, with no m ore than 10 crew, single
masted, lateen rigged with rectangular or triangular sails, and
steered with one or two lateral rudders. O n ly the pencalang was
decked, and none o f them had outriggers (Knaap 19 9 6 : 33 “ 7 >
6 6 , I 55 ~ 6 ; H orridge 1981; Ellacott 1958).
The British occupation o f the Dutch colonies from 1795 >
only effective in the case o f Java from l8 ll, appears to have
stimulated an inflow o f Hadhram i shippers, and they certainly
proved better able than the British to cope with the return o f
the Dutch in 1816. From 1818, all coastwise shipping was legally
restricted to Dutch flagged vessels owned by residents o f the
Dutch colonies. International shipping was confined to the
port o f Batavia, to which Riau, Semarang and Surabaya were
added in 1825* These regulations were tightened in 1834* to
prevent foreigners with only a few years’ residence from flying
the Dutch tricolour and engaging in coastal shipping, a
measure aimed mainly at British merchants. In 1850, ships
registered in the Netherlands were authorised to take part in
coastal shipping (Broeze 1979 : 2 5I—^).
This legislation helped to propel resident Hadhramis to the
forefront o f coastal shipping, as they were legally considered to
232 W il l ia m G e r v a s e C l a r e n c e - Sm i t h
be Dutch subjects. They moved into town quarters originally

reserved by the Dutch for South Asian Muslims in Java (Berg
1886: II^). T h e latter were mainly Bengalis and Tamils, who
played a significant part in recruiting crews for ships (Earl
197I: 60—2, 80—1). N ot only British but also Dutch concerns
declined, while ships owned by locally resident Chinese were
confined to short hops and were on average much smaller than
those belonging to Arabs. O ver m edium and long distances
within the Dutch East Indies, H adhram i ships gained supre
macy. From 22% o f the registered tonnage o f European rigged
ships in 1820, Arabs went to just over 5 ° % in x850. I*1 the
latter year, this compared to 29% for the Chinese, 9% for the
Dutch, 9% fo r the British, and 3% for the 'natives’ . Arabs had
no European rigged ships in Batavia in 182O, but they
possessed nearly 2 ,0 0 0 tons o f such shipping in the port by
1850. However, it was the twin harbours o f Gresik and
Surabaya in eastern Java which witnessed the fastest progres
sion, with some 11,500 tons owned by Arabs in 1850. In
contrast, Arab shipping stagnated in Semarang, which had been
their main centre in the I7 7 °s (Broeze 1979 : 257—60, 266—7).
Hadhramis were able to expand shipping links between
Dutch ports and those outside the Dutch possessions. Arabs
resident in Dutch ports captured the bulk o f business with the
British settlements in Malaya from the 1830s, as Dutch
regulations bore down m ore heavily on British firm s. Flying
the D utch flag, these 'fine vessels’ , built o f teak and ranging
from 150 to 500 tons, engaged in much contraband to avoid
Dutch tariffs (Buckley 1965= 3^4-)• Furtherm ore, independent
M uslim rulers allegedly accorded H adhram i sayyid shippers a
rem ission o f duties, on account o f their 'superior sanctity’
(Earl 1971: 68).
O n e effect o f this boom was to make H adhram i shipping
m ore elitist and specialised. Th ere was a concentration o f ships
into fewer hands, with the complete disappearance o f the
form erly dom inant A rab skip p er-o w n er in Batavia and
Surabaya by 1850. However, 7 Arab captains out o f 25 still
owned their ships in Gresik in 1850, and the practice rem ained
somewhat m ore com m on among Arabs than am ong Europeans.
T h e R is e a n d Fa l l o f H a d h r a m i S h ip p in g 233
The average size o f vessels rose, with the num ber o f Arab ships
o f over 2 0 0 tons in Gresik going from 9 in 1830 to 16 in 1850.
Shipping became a m ore specialised business, rather than a
mere appendage o f com m erce, although there are no trade
figures to match those on shipping (Broeze 1979 : 264—6). A
straw in the wind is that Haji Fattala, a m erchant in Bangkok
and possibly a G u lf Arab, placed a large cargo o f sugar on a
royal T h ai ship in the 1840s, to be transferred to a Bombay
bound vessel in Singapore (Neale 1852: 164, 170—2, 176).
Sayyid families involved in the great Javanese shipping boom
included the al-Saqqaf and the Ba Raqbah, but the most
notable among them were two members o f the al-H ibshi
family, Hasan and A law i. They seem to have been brothers,
and they both received the coveted Javanese title o f Pangeran in
the 1840s. By 1850, the two al-H ibshi possessed h alf the Arab
tonnage o f Surabaya, as well as owning ships registered in
Gresik and Batavia. A ll seven al-H ibshi ships in Surabaya were
barks, three masted square riggers, with some fore-an d -aft
rigging on the rear mast to allow sailing into the wind.
Together, these seven barks totalled over 2 ,0 0 0 tons (Broeze
I 979 ; 265—6). Sayyid Hasan was also the greatest Arab trader in
Java in the l820s and 1830s (Earl 19 71; 67—8). He was
entrusted by the Dutch with delicate diplom atic missions as far
away as Siam, where he probably had trading interests
(Lekkerkerker 1923: 2 0 2 —3; Broersma 1934 *- ! 33 “ 4 )*
N ot all m ajor Hadhram i ship-owners in Java were sayyid, for
the Ba Hashwan also possessed several European rigged ships in
Gresik and Batavia by 1850 (Broeze 1979 : 266). In the 1860s,
M uhammad b. 'Uthm an Ba Hashwan did a flourishing trade in
horses, transported from Sumba to Java in one o f his ships
(Parimartha 1995 : 214)- This family was drawn from the maskin
social stratum, the most lowly in H adhram aut’ s free p opula
tion. That said, the Ba Hashwan belonged to a 'bourgeois’
group, who claimed to have imm igrated from Iraq together
with the founder ancestor o f H adhram i sa p id families (Berg
1886: 4 9 -5 0 ) . Members o f the Ba Hashwan family were also to
be found on the coast o f Somalia (Ulrike Freitag, personal
com m unication).
234 W il l ia m G e r v a s e C l a r e n c e - Sm i t h
Palembang, in southern Sumatra, was the main rival to

Surabaya-Gresik as a centre o f H adhram i shipping (Berg 1886:
1 4 7 -8 ). O nly three Arab ship-owners were registered in
Palembang in the 1830s, but from 1840 there was a rapid
growth in the registration o f Arab barks, brigs and schooners,
generally between 200 and 600 tons. Foremost among the
Hadhram i ship-owners o f the port was Sayyid 'A ll b. A bu Bakr
b. al-Shaykh A bu Bakr b. Salim. He and his family owned
about half the total Palembang Arab fleet by 1850. T h e Dutch
appointed him Captain o f the Palembang Arabs from 1833 till
his death in 1878, and granted him the title o f Pangeran
(Peeters 1994 : 27 ~ 3 °)- T h e A bu Bakr b. Salim came originally
from 'Aynat in the eastern part o f the m ain wadi in
Hadhramaut, and they were exceptional in being allowed to
bear arms and dress like tribesm en (Berg 1886: 33 —4 > 5 I—3 ’
l8 l). T h e other h alf a dozen m ajor Arab ship-owners in
Palembang in 1850 were all o f H adhram i sayyid origins, many o f
them also active in Java and Singapore: al-H ibshi, a l-K af, al-
Munnawwar, Ba Raqbah, al-Saqqaf, and Bin Shihab. They had
initially engaged in shipping to and from Java, but an
increasing p ro portion o f their business was with Singapore
after 1819 (Peeters 1994: 2 7 -3 0 ).
A lthough Palembang and the ports o f north Java accounted
for a large p roportion o f ships belonging to Hadhramis, they
were also to be found in other parts o f the Dutch possessions
T h e Bin Shihab sultans o f Siak, East Sumatra, were m ajor
shippers, and in 1823 Arab ships controlled much o f the traffic
with Penang, across the straits in Malaya. Toop and pinsi were the
usual vessels for this traffic, locally built but borrow ing much
from European schooner design. T here were also small brigs
in the European tradition (D obbin 1983; 94 )* T h e a l-Q a d ri
sultans o f Pontianak, West B orneo, owned a brig and other
m erchant vessels in the l8lO s, but the port was declining as the
area’ s m ineral wealth became exhausted (M oor 1968, appen
dix: 101—6). In eastern Indonesia, Ternate, A m bon and Banda
were autonom ous and growing centres o f H adhram i-owned
sailing ships, with close links to Surabaya (C larence-Sm ith
1 9 9 8 : 3 9 )-
T h e R is e a n d Fa l l o f H a d h r a m i S h ip p in g
235
T he foundation o f British Singapore in 1819 provided m ajor

new opportunities for Hadhram i shippers. Raffles ordered that
a quarter suitable for up to two thousand Arabs be set aside in
the new free port, and Hadhramis came from far and wide to
take up his invitation (Buckley 19 6 5: 85; M orley 1949 : *55 5
Berg 1886: 122). A m ong them was the ship-ow ner Sayyid *Abd
al-R ahm an al-S aq q af [Alsagoff], who had earlier traded
between Malacca and Java (Buckley 19 6 5 : 5 ^4 )* T he al-Saqqaf
were a highly influential Hijazi family, with a well established
branch in Hadhramaut. T he other two great families o f
Singapore in the field o f shipping were both Hadhramis o f
sayyid origin, the a l-K a f [Alkaff] and the al-Junayd [Aljunied]
(Freitag forthcom ing).
Sayyid Ahm ad b. 'Abd al-Rahm an al-Saqqaf developed his
family shipping interests in Singapore in new directions. His
first move was a shrewd marriage to the daughter o f Haji
Fatima, the Makasarese Sultana o f Gowa in southern Sulawesi.
H aji Fatima drove much trade with Singapore, em ploying a
fleet o f ships rigged in either the European or the 'native’
manner, a 'contraband’ trade which the Dutch could not
control. Sayyid Ahm ad managed the business in Haji Fatima’ s
lifetim e, and took it over after her death (Buckley 19 6 5 : 5 ^4 ;
Wright and Cartwright 1908: J O 7; A lsagoff 1963: 9—II). Sayyid
A h m ad’ s next coup was to initiate a regular pilgrim service to
Jiddah in the 185Os, thereby linking up with H adhram i
shipping in the Red Sea and his own family roots in the Hijaz.
Singapore became the hub o f the rapidly growing pilgrim flow
from Southeast Asia, because the Dutch stubbornly attempted
to lim it pilgrim numbers for political and social reasons. The
sayyid com m unity exploited its religious prestige to take a large
share o f this lucrative business (Lee I 99 I: *65).
D uring the short British occupation o f Java from l 8 ll to
1816, the Arab shipping entrepreneurs o f Gresik made full use
o f the renowned teak forests o f east-central Java to build ships,
but this became m ore difficult when the returning Dutch
slapped restrictions on teak extraction (Broeze 1979 •' 256 “ 7)-
O ne unnam ed Arab nonetheless built a ship o f over 1,0 0 0 tons
burden on the river upstream from Surabaya, said to have been
236 W il lia m G ervase C l a r e n c e - S m ith
the largest ship ever built in Java at the time. T he cost o f the
enterprise was such that he was obliged to sell the ship at a loss,
as he lacked the necessary funds to fit her out (Earl I 97 I:
71—2). However, the ship appears to have rem ained in
Hadhram i hands, for this was probably the 'Fait A llam ’ , the
largest ship owned by any Arab in 1850, registered in Semarang
at 1,120 tons (Broeze 1979 : 264)- T h e Bin Shihab sultan o f
Siak also built his own vessels in the l820s (D obbin 1983: 94)-
H adhram i shippers sometimes employed senior European
officers. The al-H ibshi fleet in Surabaya in 1850 had Dutch
captains on four ships and Arab ones on three, the latter
relatives o f the owners (Broeze 1979 : 265—6). Similarly, in
182 7 > a vessel owned by Sayyid Husayn o f Pekalongan sailed
from Riau to Batavia under an English captain. Arabs were said
never to insure their ships, so that this was not the reason for
the em ploym ent o f a European (M oor 1968: 3 1)- By the 1880s,
however, a European mate was often employed to meet
insurance conditions, while the captain and factor, or super
cargo, were usually Arabs (Berg 1886: 147—5o).
O rdinary crew members were recruited from the seafarers o f
the archipelago (Berg 1886: I 47 ~ 5 o)- Thus, Bugis crewed an
Arab ship apprehended near Ternate in 1830 by the Dutch
authorities (Bastiaanse 18 4 5 : I 45 )- Many seamen on Arab ships
in the early 1830s were slaves from the O uter Islands o f
Indonesia (Earl 1971: 67). This practice declined, as the Dutch
gradually took m ore effective measures against the slave trade
and slavery.
Dutch regulations banning the entry o f European rigged
ships into small 'native’ ports may have led some Hadhramis to
acquire locally rigged ships (C larence-Sm ith 1998: 39 )* T h e
statistical data for such craft are unfortunately far m ore patchy
than for European rigged vessels. O n e needs to treat with
caution figures showing Malay rigged ships owned by Arabs
peaking between 1825 an<i 1840, at less than a tenth o f the
tonnage o f their European rigged ships (Mansvelt 1938: 98)-
Wealthy Hadhramis in Ternate in 1876 all owned perahu, rigged
and built in the m anner o f Makasar in South Sulawesi, for the
purpose o f coastal com m erce in the Moluccas [Maluku] (ANRI,
T h e R ise a n d Fa l l o f H a d h r a m i S h i p p i n g 237
30, 162a, Algem een Verslag Ternate 1876). W hether Ternate

Arabs also owned local kora-kora, fitted with outriggers, is not
clear. Makasar perahu in the 1880s rem inded A nna Forbes o f
small Chinese junks, probably because o f their sails. They were
around 80 tons burden, sloped down to the bows, and were
built using rattan, bam boo, and other local materials (Forbes
19 8 7: 45 —7)* Arabs throughout the Dutch possessions were
thought to own about a hundred perahu rigged in the local
m anner in the 1880s, a suspiciously round num ber. Few Arabs
engaged in fishing, but it was locally significant in Aceh, Nias
and Pekalongan (Berg 1886: 147“ 52 )-
T h e d eclin e o f sail in the M alay world
N o crisis in Hadhram i shipping was immediately apparent after

1850, but steamers provided ever m ore com petition. T he
num ber o f E uropean rigged sailing ships owned by Arabs in the
Dutch possessions continued to increase between 1850 and
1864, reaching a total o f 96 in i8 6 0 —64, but the Arab
percentage o f the total fell slightly, as Chinese shippers
expanded their operations even faster. In 1865—69, the last
quinquennium for which the Dutch published statistics, the
aggregate num ber o f European rigged sailing ships in Arab
hands fell for the first time (Mansvelt 1938: 98; Cam po 1992:
367, 683). Nevertheless, Arabs still owned 75 European rigged
ships in the m id -l8 8 0 s, am ounting to some l6 ,0 0 0 tons in all,
with Palembang and Surabaya as the main ports o f registration
(Berg 1886: 119, 1 4 7 -5 0 ).
A functional niche which such ships retained longer than
most was the transport o f live animals, pride o f place being
taken by horses brought from the Lesser Sundas to Java (Berg
1886: 150; C larence-Sm ith forthcom ing). However, a m ajor
problem for sailing vessels was that they could only leave the
Lesser Sundas when the m onsoon was blowing in the right
direction, as the condition o f horses deteriorated when ships
had to tack against the wind (Kuperus 193^: 28). Arab ship
owners only held their own against the new technology until
special facilities fo r the transport o f live animals were
238 W il lia m G ervase C l a r e n c e -S m ith
introduced on steamers around the turn o f the century. By the

interwar years, Arab square rigged ships had becom e obsolete
in this trade, although their perahu were still em ployed to
concentrate horses in the larger ports visited by steamers
(Kisterm ann 1991= 5 1* 62, 65).
A rab owned perahu continued to be used for other purposes
till the end o f the Dutch period. In the 1930s, they transported
bulky but low value chalk and lim e from Sulawesi and the
Lesser Sundas for use in the Javanese construction industry, a
business in which H adhram i firms were heavily involved. This
may have been a cost cutting response to the econom ic
depression o f that decade. Arab owned perahu were also o f some
im portance in fishing in East Java (K ro e f 1953 : 3*6).
A noth er niche was pearling in the far east o f the D utch East
Indies, where m o th e r-o f-p earl was especially plentiful. A
Western craze for m o th e r-o f-p earl buttons and similar articles
led to a m ajor boom in the last decade o f the nineteenth
century. W ith a schooner and a fleet o f small lateen rigged
luggers, it was possible to fish m ore efficiently and at greater
depths than indigenous peoples could manage. Shaykh Sa'id b.
A b dallah Ba A dilla [Ba A b d allah ?], Lieutenant o f the Arab
com m unity o f Banda, received perm ission to fish for pearl
oysters o ff the A ru islands in 1897, near New G uinea. He and
his brothers made a fortune in the process. A lthough the
m o th er-o f-p earl business never truly recovered from the First
W orld War, Arabs continued to fish for pearl oysters and
trochus shells in a small way (C larence-Sm ith 1998: 3 ®“ 9 )*
In terms o f general cargo and passenger work, Arab sailing
ships were relegated to less frequented routes from the 1880s,
especially in eastern Indonesia (Berg 1886: 150). T h e schooner
became the preferred Arab ship for such purposes, as it was
easier to manoeuvre than a square rigged brig or bark, and it
required a smaller crew per ton. Up to i g i 4 > Arab schooners
operated between the small islands o f the Moluccas and the
Lesser Sundas, serving ports and islands at which steamers did
not call. T h e Ba A dilla family was especially prom inent in this
business (Broersm a I 9 3 4 : I 4 4 —5 > 3^5“ 6; Broersm a 1935 —36:
425)-
T he R ise a n d Fa l l o f H a d h r a m i Sh ippin g 239
T h e false dawn o f steam navigation

in the M alay world
Steamers may have harm ed established interests, but they also

provided new opportunities in the Indian O cean (Broeze
1996). A few Hadhramis responded positively to the challenge,
even m aintaining the link with the Red Sea. However,
prom ising beginnings gave way to frustration. From around
the turn o f the century, Hadhramis were reduced to defending
a shrinking business, lim ited to pockets o f Southeast Asia.
Th e Dutch possessions apparently witnessed the beginnings
o f steam navigation by Hadhramis. In the early 1860s, a group
o f Arabs from A m bon were reported to be the real owners o f
the 'M enado’ , which operated between Java and eastern
Indonesia. T h e steamer was registered in the name o f an
Am bonese with European legal status, and had a European
captain. Despite somewhat scanty return freight, consisting
mainly o f spices and coffee, the 'M enado’ undercut the rates o f
subsidised mail boats in the 1870s (C larence-Sm ith 1998:
39—40). In 1878, two Arabs and a European founded the N .V .
Voorwaarts company in Batavia for a weekly steamer service to
Semarang in central Java. In 1879> the company purchased a
small 300 ton steamer in Singapore, able to carry J O
passengers, and extended its services to Surabaya. However,
the venture collapsed a year later, apparently for lack o f
sufficient capital (Cam po 1992: 4^; Berg 1886: 148—9). Two
Arab firms in the Dutch possessions owned a single steamer
apiece in 1886, one in A m bon and the other in Palembang
(Berg 1886: 149)-
T h e foundation o f the K oninklijke Paketvaart Maatschappij
[KPM] in 1888 sounded the death knell o f Hadhram i steam
navigation in Indonesia, for the Dutch were determ ined to
elim inate 'Foreign O rien ta l’ com petition in coastal shipping.
Financed by the m ajor D utch shipping com panies and
benefiting from generous official subsidies and assistance, the
K PM quickly became a quasi-m onopolistic force, closely linked
to the Dutch program m e o f military 'pacification’ (Cam po
1992). Arab steamer services within Indonesia were thus almost
entirely choked o ff by the end o f the 1890s. O ne exception was

northern Sulawesi, where Arabs still owned a few small tramp
steamers in 1920 (Jansen 199 ° : J9 )- M oreover, the K PM could
not control routes to Singapore as effectively as those lying
entirely within Indonesia. In 1911, an Arab steamer service ran
from southeastern Sumatra to Singapore, and Arabs also
participated in the considerable river traffic generated by the
rubber boom in this region (Purwanto 1992: 5 4 > 89—90).
Arabs at times acted as agents for E uropean steamer
companies, typically in remote and dangerous areas. A n Iraqi
Arab, Sayyid A bdallah b. 'Abd al-Rahim al-Jilani, was the
Lom bok agent o f the N ederlandsch-Indische Stoom vaart
Maatschappij in 1887. A lthough this company was a wholly
owned subsidiary o f the British India Steam Navigation
company, it was subsidised by the Dutch authorities p rior to
the creation o f the K PM . Sayyid A bdallah then played the same
role for the K PM until his assassination in 1891, when a
Hadhram i, Sayyid 'Abd al-Rahm an al-H ibshi, took his place.
However, the latter was in turn replaced by a D utchm an in
1894> following the conquest o f Lom bok by colonial troops
(G o o r 1982; Cam po 1992: 174—5, 179).
Singapore, somewhat m ore liberal in its adm inistration,
became the main bastion o f H adhram i steam navigation. Sayyid
Ahm ad b. A b d al-Rahm an al-Saqqaf acquired steamers to
expand his already flo u rish in g pilgrim age business. His
Singapore Steamship Com pany ferried pilgrim s to Jiddah as
early as 1871, in a steamer with a European captain and a
Chinese name (Sherry 1966: 47 )* In 1874> company had
four steamers and conveyed 3,476 pilgrim s to Jiddah. A fter his
death in 1875? his son Sayyid Muhammad continued the
business (Lee I 9 9 I: 165—7)* T h e follow ing year Sayyid
M uhammad became a partner in a Chinese ship chandlers
business in Singapore (Song I9 2 3 : 188).
In 1880, a m ajor scandal befell the Singapore Steamship
C om pany’ s Jed dah ’ . This was an eight year old steamer o f over
1,0 0 0 tons, built in D um barton, Scotland. T h e European
captain together with other European crew members and Sayyid
'U m ar al-Saqqaf, abandoned the stricken Jeddah ’ in the G u lf
o f A den, following a bad storm. They reached A den in a

lifeboat, and reported to the authorities that the ship had gone
down with nearly a thousand pilgrim s on board. To general
astonishment, the vessel was then towed into A den by another
steamer. T h e 'Jeddah’ was found to have had only six lifeboats
on board, but accusations that the incident was staged to gain
£ 30 ,0 0 0 in insurance m oney proved to be groundless (Sherry
19 6 6 :4 ,3 -6 4 ).
Despite this scandal, there were still two Arab steamer
companies active in Singapore in the m id -l8 8 o s, the al-Saqqaf
firm and Sayyid Masim [Muhsin?] b. Salih al-Jifri [Jeoffrie].
Arab capital from Surabaya and some Chinese funds were
invested in one o f these companies. They owned two steamers
apiece, employed European captains and engineers, and were
locked in fierce com petition with two Dutch lines for the
pilgrim traffic to Jiddah. They also ran services to the east, as
far as China. However, Sayyid al-Jifri’ s business collapsed some
time before his death in 1894> f ° r reasons which rem ain
unclear (Berg 1886: I 4 9 —5 ° ; Buckley 19 6 5: 564—5)•
T he long term problem for H adhram i steam navigation was
that British liberalism failed to prevent the em ergence o f
powerful cartels, the shipping 'conferences’ or 'rings’ which
enmeshed the Indian O cean in their coils from the end o f the
nineteenth century. O ne solution was to jo in these cartels.
Sayyid 'U m ar al-Saqqaf was a founding m em ber o f a syndicate
for the pilgrim traffic in 1896, in alliance with the British
company Holts, several Dutch shippers, and the Sharif o f
Mecca. The syndicate enjoyed a m onopoly over the transport o f
pilgrim s from the Straits Settlements to Jiddah (Lee I 99 I:
165—6). How long this arrangement lasted is not certain, but
the al-Saqqaf company was still transporting pilgrim s to Jiddah
in 1908 (Wright and Cartwright 1908: j 0 j ) .
T h e cartelisation o f steam navigation in Asian waters
increased greatly from the turn o f the century, leading
Singapore Arabs to jo in an unsuccessful public protest in
191O (H u ff 1994 : 135)* M oreover, rapid technological change
led to ever larger and m ore costly steamers, raising barriers to
entry in this sector (Headrick 1988, ch. 2,) . Nonetheless, some
Arab involvem ent in m odern shipping persisted into the

interwar years in Singapore. Salim b. Muhammad b. Talib, a
Hadhram i o f qabili [tribal] origins who was among the wealthiest
Arab entrepreneurs in Singapore after 1918, owned several
steamers until his death in 19 3 7’ when his family sold them
(Freitag forthcom ing). It is not clear whether these were o f the
new generation o f o il-b u rn in g ships, which accounted for 58%
o f steamers around the world by 1939 (Headrick 1988: 3 1)-
In Indonesia, Arabs rem ained defiant in north Sulawesi.
T h e K PM fought a long 'guerrilla war’ against Arab m otor
boats with w ooden hulls that brought copra to Manado for
shipment abroad in tramp vessels, often Japanese. In 1935 ’ the
K PM set up the Celebes Kustvaart Maatschappij [Cekum ij],
holding nearly 40% o f the capital and distributing the rest to
the owners o f m otor boats brought into the new company.
Sayyid 'Abd al-Rahm an b. Shaykh al-H asni, a prom inent
H adhram i trader in G o ro n ta lo , refused to accept this
com prom ise. He teamed up with Japanese interests to form
the rival M otor Kustvaart Maatschappij [M okum ij]. Sayyid 'Abd
al-Rahm an obtained a 45 % share in M okum ij, while his
Japanese partner, the Celebes D evelopm ent Com pany [C ede-
co], held the rest. New regulations in D ecem ber 1935 ’
reserving coastal shipping in N orth Sulawesi to Dutch vessels,
were evaded by registering ships in the name o f Sayyid 'Abd al-
Rahman, a Dutch subject. However, when the authorities
refused to renew operating licences, Sayyid 'Abd al-Rahm an
and his Japanese backers were forced to sell M okum ij’ s five
ships to the K P M in 1939 (Dick 1989: 257 “ 8, 266).
S h ip p in g in In d ia and the Persian G u lf:

a struggle for survival
Calicut [Kozhikode] on the northern Malabar coast had long

acted as a kind o f H adhram i cross-roads in the Indian O cean,
but it is difficult to detect how much Hadhram i shipping
survived the fierce com petition that developed from the m id
eighteenth century. T he O m an i role in the renaissance o f Arab
shipping in India was such that 'Muscat boats’ was the generic
term for Arab sailing ships in the 1780s, and the lords o f
Masqat were themselves m ajor shippers to Malabar (Das Gupta
1967: 19, 2 2 , 83, 90). A Dutch report on the trade o f the
Malabar coast in 1781 noted that Masqat was the furthest port
from which Arab vessels came to trade, bring mainly dates in
exchange for sugar (Galetti and G root I9 H : 2 l8 —2 l).
T he O m ani were not the only non-W estern shipping rivals
that Hadhramis faced. From the l820s, when O m an lost
control over access to the Persian G ulf, people from this region
were free to develop shipping links with India (Bhacker 19 9 2 :
88—94). 'U tbi Arabs, recently settled in Kuwait, Bahrein and
Qatar, proved particularly dynamic in shipping horses, pearls
and dates to Bombay. Kuwait was also a m ajor ship-building
centre. Q asim i Arabs from the Trucial Coast, Persians, and
South Asian m ercantile comm unities were all active on this
route, as were possibly Baghdadi Jews and Iraqi Eastern
O rthodox Christians (Fattah 1997 ; al-Rashid 1981: 22 —5 ).
Although many G u lf ships went only as far as Bombay, some
sailed annually to Malabar to buy Mysore coffee (Langley 1962:
37 )-
A further com plication concerns the relations between
Hadhramis and Mappila [Moplah] Muslims o f the Malabar
coast. T he latter were a dynamic trading and shipping group in
their own right, dispersed to Sri Lanka [Ceylon], Burm a and
Malaya by the 1900s (Innes 1908: 189—9 1)* Many Mappila were
o f Hadhram i descent, and they venerated H adhram i teachers o f
sayyid origin as holy men with m iraculous powers. However,
Mappila generally saw themselves as a Malayalam-speaking
South Asian comm unity, rather than as Hadhramis or Arabs
(Dale 1997 ; Lakshmi 1995 )- This may have reflected dem o
graphic factors, as figures for the Malabar district showed only
246 Arabs in 1881, compared to 495*248 Mappila (Logan
1887, vol. 2 ‘ iv).
Some Hadhram i shipping did survive in Malabar. Thus,
Sayyid Muhammad al-Zahir left his native Hadhramaut in the
m id -l8 30 s, and became a successful trader and ship-ow ner on
the Malabar Coast. He sent his son Sayyid A b d al-Rahm an as
supercargo on one o f his sailing ships to the Red Sea ports o f
244 W illia m G ervase C la re n ce-S m ith
the Y em en in 1848. Sayyid 'Abd al-Rahm an later developed

com plex m ercantile and political ties all round the Indian
Ocean, achieving lasting fame for his role in the opening stages
o f the long war between the Dutch and the Acehnese o f N orth
Sumatra from 1873 (Reid 197^)•
H adhram i shipping relations between Malabar and the Malay
w orld were closer at the beginning o f the nineteenth century
than at its end. Arab sailing vessels from the Malay world used
to come to Malabar to buy South Asian cloth, the staple o f all
com m ercial operations in the archipelago. In return, they
brought spices, benzoin and M uslim pilgrim s (M orley 1949 :
164). In the middle o f the nineteenth century, Sayyid 'A ll b.
A bu Bakr b. al-Shaykh A bu Bakr b. Salim, the greatest
Hadhram i shipper in Sumatra, traded regularly to Persia,
probably stopping o ff in Malabar (Peeters 1994 : 3 °)- the
1880s, however, European cloth had to a great extent replaced
South Asian cloth in the business dealings o f Hadhramis in the
Malay w orld, pilgrim s were travelling direct to Jiddah from
Singapore, and the Hadhram i square rigged fleet in Sumatra
was shrinking (Berg 1886: 144—5 ° ; L ee J99 I: I^5)-
Shipping contacts between Malabar and the western Indian
O cean are equally hard to pin down. A small trickle o f ' ulama o f
H adhram i sayyid provenance came to settle on the coast through
the eighteenth and nineteenth centuries, such as the related al-
J ifri and Bin Sahl families. They kept some com m ercial links
with their hom eland (Dale 198O: III—7). In the 1830s, wealthy
Arab 'ulama in southern India owned A frican slaves as personal
servants, and Arab ships based on the Malabar coast had
A frican slaves in their crews, possibly indicating shipping links
with East A frica or the Red Sea. However, the British cracked
down on the slave trade in the follow ing decade (Young 1987)-
O n e wealthy and influential Malabar Hadhram i family
played a significant role in building sailing ships in Beypore,
a port o f greater significance in coastal shipping than Calicut by
the 1860s (Blake 195^: 73)- T h e Ba Rami were established in
both C alicut and Beypore, but their shipyards were in the latter
port. They profited from local supplies o f teak to build dhows,
exporting a fair num ber to the Persian G u lf in the 1910s.
In d e e d , their yards were reputed to be one o f the single most

im portant sites for the construction o f dhows along the coast o f
southwestern India. T h e Ba Rami traced their origins to al-
Mukalla in the eighteenth century, claim ing sayyid origins and
relatives in Egypt and Indonesia. However, they did not appear
on Van den B erg’ s list o f Hadhram i sa p id families, and they did
not interm arry with other sayyid families in Malabar (Dale 1997 :
l 8 l - 2 ; Lakshmi 1995 : 5” 9 )*
Hadhram i shipping links with Malabar continued after the
ending o f the slave trade. Merchants o f al-Shihr, who operated
their own sailing vessels, were reported to be in regular
com m ercial contact with Malabar in the early 1890s, although
what they carried is unclear (Hirsch 18 9 7: 25 )* H adhram i
em igration to Malabar was said to be mainly from coastal
districts in the 1880s (Berg 1886: 124).
Equally murky was the Hadhram i role in shipping further
north, along the coasts o f K onkan and Gujarat. Most Arab
ships sailing to Bombay between 1800 and 1840 were O m ani,
but some came from points further west along the coast o f
southern Arabia (Benjam in 197^: 89)- T h e slave trade was one
link. In 1835* three vessels from al-M ukalla brought 79 slaves
to Porbandar, in western Gujarat, 24 ° f whom were marched
inland in charge o f two Arabs (Banaji 1933 : ! 5 I—^)- Although
it is not stated that these ships were H adhram i owned, South
Asians generally steered clear o f transporting slaves (C larence-
Smith 1989a). A small comm unity o f Hadhram i businessmen
was still active in Bombay and Gujarat around 19OO, but
shipping was not specifically m entioned among their activities
(Gazetteer 1899* v°l* 9 ~ x5—J7)- Navayat Arabs, often o f
mixed race, were famous as seafarers in Bombay and the coast
to the south, but their origins lay in the Persian G u lf (Imperial
Gazetteer 1909, vol. I: 226; vol. 2: 86, 98).
O ne particular Hadhram i shipping connection with Bombay
was m aintained by the Q u'ayti sultans and other m ercenary
leaders. A fter the British subjugation o f Gujarat in the early
nineteenth century, Hadhram i m ercenaries moved to the
inland M uslim state o f Hyderabad in southern India. However,
they kept in touch with their hom eland through Bombay. T he
246 W il l ia m G ervase C l a r e n c e - S m ith
Qu'ayti sultan, who doubled up as a military com m ander for

the Nizam o f Hyderabad, had considerable investments in the
thriving city o f Bombay (Khalidi 1997 ; Gavin 1975 : I 5 ^—6 o).
'Awadh b. 'Um ar al-Q u 'ayti, who reigned from l 866 and was
said to have been the single wealthiest man in Hadhramaut, had
most o f his fortune in India (Berg l 8 8 6 : 7 ^). From Blake’ s
confused account, it seems to have been this sultan who used
the O ld M ogul dry dock in Bombay to careen his dhows, but it
may have been his father (Blake 195 ^: 233 “ 4 )- It is n °t certain
whether these Q u'ayti dhows were based in Bombay or in
Hadhramaut.
Sultan 'Awadh al-Q u 'ayti purchased a steamer in Bombay in
1874, but it is unclear whether this steamer served for trading
purposes in India (Gavin 1975 : 160—73)- Even if it ever did so,
the rise o f the British India Steam Navigation Com pany
Lim ited [BISN] severely restricted the role o f Asian steam
navigation. Part o f the M ackinnon group, the BISN played a
similar role to that o f the K PM in the Malay world. Registered
in Scotland in 1862 with a capital o f £ 4 0 0 ,0 0 0 , the BISN
obtained subsidies and support from the British governm ent
from the outset. From British India, the company gradually
extended its operations all around the Indian O cean and the
South China Sea. T he BISN specialised in run ning small
steamers on 'feed er’ routes, either nipping local com petition
in the bud, or buying up rivals such as the Bom bay-H edjaz
Com pany [Shah Line] in 191O and the A rm enian Apcar Line
in 1912 (Blake 1956). Some 'native’ owners o f steamers
persisted right through the British colonial p eriod , but there
is no indication that any Hadhramis were in this position
(Headrick 1988: 366—71).
T h e hom e ports o f H adhram au t
Hadhram aut’ s own sea-borne trade offered few pickings. T he

territory lacked com m odities to rival with Y em en i coffee,
O m ani dates, or the dates, grain, pearls, and horses o f the
Persian G ulf, and H adhram aut’ s ports were poor natural
harbours. T h e small flow o f exports consisted mainly o f fish
T he R ise a n d Fa l l o f H a d h r a m i Sh ipp in g 247
products, salt, tobacco, and honey (Berg 1886: 73 “ 4 > 7 &;

Lekon 1997= 266—7). A l-M ukalla played a modest role as an
em porium p rior to the British seizure o f the ruined settlement
o f A den in 1839, but trade was depressed (Haines 1906: 228).
A l-M ukalla was even briefly used as a coaling station in the first
years o f steam navigation. However, as steamers became larger
and as the British developed A d en ’ s facilities, Hadhram i ports
were unable to compete. From the 1850s, al-M ukalla and the
older port o f al-Shih r found a niche in redistributing slaves to
a variety o f destinations in the M iddle East, but the British
stepped in to put an end to this traffic in the 1880s (Gavin
1975: 28, 15 6 -7 , 161).
A variety o f sailing ships were in use in the nineteenth
century in Hadhramaut. Markab were European rigged ships o f
200 to 400 tons. Dhows were called baghla, ghanja, and sambuk,
in descending order o f size (Berg 1886: 73 )- Dhows were built
in al-M ukalla in the first half o f the century, sometimes o f sewn
rather than nailed construction, presumably with teak im
ported from the Malabar coast o f India (Malecot 19 9 °: 3 ^)-
Long distance trade in Hadhramaut was largely in the hands
o f South Asians, and it seems that this was also true o f shipping.
In the m id - 1830s, Gujarati merchants already controlled most
o f the trade o f al-M ukalla, which received some 20 dhows a
day, o f IOO to 300 tons burden (Haines 1906: 228). Parsi
merchants o f Gujarati origins were still in this position in the
1890s, although the reigning al-Q u 'ayti family owned a couple
o f baghla based in al-Shihr, which had to take refuge in
neighbouring al-H am i in stormy weather. T he Bents took ship
in al-Shihr on a dhow from Kutch, which had a H indu captain
and crew, both o f low caste (Bent 190O: 75~ 7 > 2 ° 5 > 284).
Hadhram i shippers seem to have m aintained a better position
in al-Shih r than in al-M ukalla, in contact especially with
eastern A frica and Malabar, and on an occasional basis with
Basra in the Persian G ulf. Local captains were sometimes freed
slaves, as in the case o f the o n e-arm ed a l-N u b i, who
transported H irsch on a sambuk (Hirsch 18 9 7: 25 > 88—90).
T he advent o f steamers did not improve the H adhram i
position in their home ports. T he steamer purchased in
24 8 W il lia m G ervase C l a r e n c e -S m ith
Bombay by Sultan 'Awadh b. ’U m ar al-Q u 'ayti in 1874 played

an im portant part in his seizure o f al-M ukalla in 1881, but what
happened to it afterwards remains a mystery (Gavin 1975 :
160—73). W hen the sultan granted an export m onopoly in the
early 1890s for H adhram aut’ s tobacco, it went to an O ttom an
company, which chartered small steamers to handle the crop
(Hirsch 1897: 2 - 3 , 6 - 7 , 14).
In the interwar years, most o f H adhram aut’ s shipping was in
the hands o f two firm s based in A den. O n e was the old
G ujarati Parsi company o f Cowasjee Dinshaw. T h e newer
enterprise was the Halal Shipping Com pany, set up in 1924 by
A n to n in Besse, a French entrepreneur. In addition to two to
three steamers, Besse had a fleet o f ten to twenty dhows o f
around 150 tons, m ainly built with M alabar teak in his own
shipyards in A den. Large D utch and British liners called
occasionally at al-M ukalla, catering for the passenger traffic
spawned by the far-flu n g H adhram i diaspora. Som e 70 to 80
baghla left al-M ukalla every year, going to Bombay, the G ulf,
the Red Sea and Zanzibar, but it is unclear how many o f these
large dhows were owned by H adhram is (Ingrams 1937 : 69—7 °5
Footm an 1986: 56, 6 0 —2, 87, 94 —5 > I0 5 )- Som e dhows
calling at H adhram i ports certainly belonged to shippers o f
O m an i and Kuwaiti origin (Stark 1953 : 3 9 > 54 “ 5 ; M eulen
1947: 14, 235 )- In 1938, Freya Stark travelled from H adhra
maut to A d en in a dhow 120 feet long, crewed by 18 O m anis,
all related to one another. T h e ship was built o f Malabar
tim ber, decked at the stern, and with a wheel and compass.
T h e O m an i captain was related to the sultans o f Zanzibar and
worked for a 'rich m erchant’ in A den, possibly Besse. This
dhow plied the route lin king C alicut, Bombay, Basra, Jiddah
and Zanzibar, the crew taking time o ff to harvest dates in
O m an when the winds were too strong (Stark 1945 : 256—7 >
259 )- Dhows were still built in al-M ukalla, as were small boats
in al-S h ih r, the latter sown with coconut fibre rather than
nailed (Ingrams 1942 : 14 ^; Stark 1953 : 5 5 ~ 6).
T h e sh ip pin g bonanza in the R ed Sea
Sailing ships were pivotal to H adhram i capital accum ulation in

the Red Sea, reflecting an unusually highly developed transport
sector in a region with few natural resources. Dem and for the
services o f sailing vessels initially increased in the nineteenth
century, despite the use o f steamers as early as 1829 and a surge
in steamer traffic following com pletion o f the Suez Canal in
1869, so that the num ber o f ocean-going sailing ships calling at
Jiddah did not begin to decline until the 1880s (Ochsenwald
1984: 7 0 -3 ; Sha’afi 1985* c^. 2). European rigged ships found
sailing conditions difficult in the Red Sea, leaving the field
open to two masted baghla o f some 300 tons from India, as well
as smaller sambuk o f 150 to 200 tons from O m an and the G ulf.
T h e carriage o f goods across the narrow sea took place in even
smaller zarug, undecked boats o f shallow draft, measuring 45 to
60 feet, and with crews o f 6 to 12 m en. They were also used to
fish and dive for pearls and m other o f pearl (Malecot 199 ° :
28—32; M onfreid 1961).
H adhram i m igration to the southern Red Sea, o f consider
able antiquity, grew rapidly in the nineteenth century. As early
as 1814, they were num erous in Jiddah and the ports o f A s ir to
the south, and this became their main fie f (Burckhardt 1968:
391, n.; Ewald and C larence-Sm ith 1997 : 286—7). O n the
southwestern side o f the Red Sea, Hadhramis congregated in
Massawa [Mitsiwa] and Suakin [Sawakin] (Jedrej 19 8 5: !0 g ).
T here was a marked predom inance o f Hadhramis from Wadi
Daw'an, tenacious businessmen strongly attached to their
valley, and sayyid lineages did not dominate shipping (Berg
1886: 124 ; C arter 1984; Cornwallis 1976; H ogarth 1978).
N o n - s a y jid prom inence may have reflected greater geographical
proxim ity to Hadhramaut than elsewhere in the diaspora.
In the first h alf o f the nineteenth century, Hadhramis were
typically captains o f coasting vessels. They were often said to
come from the ports o f al-M ukalla and al-Shihr, although this
may have disguised an origin in Wadi Daw'an. They were
excellent seamen, able to dispense with m odern aids to
navigation. Crew members included some Hadhramis, but
250 W il lia m G ervase C l a r e n c e -S m ith
consisted mainly o f A frican slaves and Y em eni tribesm en, who

were both likely to be rewarded with a distribution o f the share
o f net profits at the end o f a jo u rn ey (Malecot 1990: 3I—2).
Owners o f ocean-going sailing vessels com ing from outside the
southern Red Sea appear to have been mainly O m ani, G u lf
Arab, and Gujarati (Ewald 1989: 82—3; Ochsenwald 1977 : 3 X4 ;
Fattah 1997; C arter 1984).
Hadhram i shippers, many o f them o f n o n -sayyid origins,
only slowly pushed their way up the ladder. Muhammad b.
'Ubayd b. Zagr came from the northern desert steppes o f
Hadhramaut, a m em ber o f the nom adic and raiding Sa'yar
B edouin people. He went on sailing expeditions to the G u lf
with his stepfather, and on the latter’ s death inherited the ship.
In 182 5 > he form ed the firm o f Shinkar and Binzagr & Go.
A lthough this was the ancestor o f a great m odern Saudi firm , it
took the Bin Zagr a fair time to becom e wealthy. Similarly, 'A ll
b. 'Abdallah Ba Eshan, from Wadi Daw'an, began modestly with
a quarter share in a dhow. Also a n o n -s a y y id , he prospered as a
Jiddah merchant at the end o f his remarkably long life, with
dhows sailing as far as Basra and Malabar from the 1840s to the
1860s (Carter 1984: 7 7 -8 , 9 0 - 5 ).
A m ajor source o f Hadhram i success was shipping slaves
across the Red Sea and the G u lf o f A den (Toledano 1982: 58)*
As the crossing was short, boats were small and needed no
special equipm ent to carry slaves, who were often transported
in small groups (Ewald 1989; M onfreid 1961: 38—4 0). In
1858, the H adhram i 'A ll Ba Junayd, resident in Massawa, sent
three Ethiopian slaves to his brother M uhammad in Jiddah, the
latter being the legal owner o f the slaves (Ewald and C larence-
Smith 1997= 290). Hadhramis featured prom inently in the
riots in Massawa and Jiddah that followed the 1857 O ttom an
prohibition on the sale o f slaves, resulting in several deaths in
the latter city. Two Jiddah Hadhramis were eventually co n
dem ned to death as ring-leaders, 'Abdallah al-M ohtessib [?]
and Shaykh Sa'id b. Husayn al-'A m ud i. Many other Hadhram i
merchants in Jiddah were im prisoned or exiled, including
Y u su f b. Ahm ad Ba Naji and Sayyid 'Abdallah Ba H arun
(Ochsenwald 1977 5 M arston 1961: 237 ” 8, 263—8; Ewald and
T h e R i s e a n d Fa l l o f H a d h r a m i S h i p p i n g 25i
C larence-Sm ith 1997 : 29 0—J)* Despite this crisis, H adhram i

involvement in the slave trade persisted. As late as 1880, the
slave trade o f al-Hudayda was dom inated by Sayyid 'A ll Ba
H arun (Toledano 1982: 5 1* 58 )-
In parallel with the slave trade, Hadhramis stepped up the
carrying o f Y em eni coffee from al-Hudayda to Egypt, taking
over sail shipping in the northern h alf o f the Red Sea in the
process. Absent from these tricky waters at the end o f the
eighteenth century, Hadhramis had come to dominate the local
trade between Jiddah and Suez by the 1850s. They pushed aside
other Arab merchants, mainly o f M aghrebi origin and based in
C airo, apparently taking advantage o f the disorganisation o f the
coffee trade provoked by Muhammad 'A li’ s attempts to turn it
into an Egyptian state m onopoly up to 1840 (G rohm ann 1933,
pp. 84—6). This went hand in hand with the growth o f a small
but wealthy H adhram i mercantile comm unity in C airo, among
them the great sayyid family o f a l-K a f from Tarim, whose
colossal fortune in Singapore derived in part from shipping
(Ingrams 1937 : 161). Several Hadhramis accumulated con
siderable wealth in shipping goods to Egypt from the southern
Red Sea (Ewald and C larence-Sm ith 1997 : 286—7).
It remains unclear when steamers took over this business,
but it was probably from the l880s, when transport by sail
entered into a phase o f decline. At that time, Jiddah merchants
owned ten large sailing ships o f 600 to 1,0 0 0 tons, which went
as far as Zanzibar, India and Singapore, and which were
presumably rigged in the European fashion. Ochsenwald does
not specify the ownership o f these vessels, but even if they were
all in Hadhram i hands, this was still a considerably smaller fleet
o f European rigged ships than that owned by Hadhramis in the
Malay world in the same decade. Between 200 and 400 dhows
also operated as far as the Persian G ulf. Jiddah was a centre o f
ship-building with teak im ported from Malabar, turning out
h alf a dozen two masted ships o f up to 80 tons in the 1850s.
The fall in the numbers and size o f ships built from the 1880s
reflected the decline o f sail (Ochsenwald 1982: 7 1)-
Nevertheless, Hadhramis rem ained prom inent in sail navi
gation in the southern h alf o f the Red Sea in the i g i O s .
252 W il l ia m G ervase C l a r e n c e - S m ith
A lthough their operations were increasingly confined to coastal

dhows, A b d al-Rahm an Ba N aji o f Jiddah, probably the
wealthiest H adhram i shipper at the time, retained contacts with
India as late as 1917 (Hogarth 1978: 62). T he Ba Jubayr firm
owned about twenty dhows operating out o f Jiddah, al-
Q unfudha and M idi, while 'Abd al-Rahm an Ba M uharram in
al-Q u n fu d h a owned four (Cornwallis 197^: 94-)•
This lingering prom inence o f H adhram i sailing ships owed a
great deal to a flourishing contraband trade in firearms, m uch
o f which originated in French Som aliland (M onfreid 1961:
7—15)- T h e O ttom an-Italian war o f 1911—12, followed by the
1914—18 struggle fo r con trol o f the A rabian peninsula,
provided particularly rich pickings fo r arms sm ugglers.
Hadhramis were high on the list o f merchants accused by the
British o f gun run ning through A s ir i ports, notably Ba Naji,
Ba Sahi, Ba Jubayr, Ba Junayd, and Ba M uharram , all
apparently o f n o n -sayyid origin. (Cornwallis 197^: 19 > 85, 9 4 >
98; H ogarth 1978: 58, 62). N ot m entioned in the context o f
arms smuggling was a m em ber o f the great al-'Attas sayyid
family, who was nonetheless prom inent in coastal shipping out
o f Jiddah (Hogarth 1978: 6l; C arter 1984: 69—71)*
M ore modest econom ic niches for dhows rem ained. O n the
eve o f the First W orld War, the Hadhramis o f Massawa used
their dhows to transport hides, m illet and camels across the
Red Sea, and to dive for pearls and m o th er-of-p earl in the
Dahlak archipelago (Istituto G eografico de A gostini i g i 3 :
207). However, the exhaustion o f oyster beds harm ed the latter
business towards the end o f the century (Ochsenwald 1982: 71)*
T h e H adhram i contribution to steamer navigation in the
Red Sea and the G u lf o f A den remains shrouded in mystery.
T h e al-Saqqaf operations have only been studied from the
Singapore e n d , and it remains unclear whether their steamers
just ferried pilgrim s to and from Jiddah, or whether they also
engaged in local shipping services. In any event, no trace o f
H adhram i steamers has been found in the inter-w ar years.
O n e lucrative sp in -o ff from the growth o f steam shipping
was the recruitm ent o f Y em eni m ountaineers to work on
steamers plying routes around the w orld. A p rom in en t
T h e R ise a n d Fa l l o f H a d h r a m i Sh ippin g 253
recruiting agent in A den in the interwar years was the

Hadhram i Aw adh b. Ahm ad Ba Sahi, with representatives in
D jibouti and several European countries (Lawless 1995 : 4 ° ,
45 “ 9)- Ba Sahi’ s business extended to N orth Am erica in the
interwar years, for his name appeared in connection with
Yem eni sailors jum pin g ship and taking em ploym ent in the
steel mills o f the northeastern U nited States (Richard Lawless,
personal com m unication).
N ich es in the O m a n i dom ains o f eastern A fric a
In O m an ’ s A frican em pire, Hadhramis drove a modest direct

trade with the C om oros, the Red Sea and Hadhramaut, unable
to offer serious com petition to dom inant O m ani, Gujarati,
and Western entrepreneurs. Nonetheless, ships from al-
Mukalla were reported in Zanzibar as early as l 8 ll (Nicholls
I 97 l: 79 )* the l820s, vessels from H adhram i ports went
yearly to Zanzibar to fetch cargoes o f slaves, mainly sold in the
southeastern ports o f the Red Sea (Kelly 1968: 4 I 3 )- Dhows
from al-M ukalla and al-Shihr brought emigrants, salt fish, salt
and onions to Zanzibar in the 1840s, calling at various ports on
the way down the coast. They returned with slaves, millet,
sesame, and sesame oil. In that decade, some 25 dhows came
annually to Zanzibar from al-M ukalla, al-Shihr, and the Mahri
country to the east o f Hadhramaut, and another 10 came from
the Red Sea. However, this compared to 50 dhows from O m an
and 35 from India (G uillain 1856—57 > vol. 3: 335 > 337 ’
357 - 6 l). Dhows, often built in the Persian G ulf, were usually
30 —50 tons, with a single forw ard-sloping mast, lateen rigged
with a single matting sail. T he larger baghla, built in Kutch,
western Gujarat, might exceptionally be two masted and decked
vessels o f some 150 tons. However, most o f them were in the
50—60 tons range. Crews consisted mainly o f A frican slaves
(Nicholls 1971: 7 5 -8 , 354).
The Lamu archipelago, outside O m ani control until the
l820s and a local centre o f boat building, was probably the
chief centre o f H adhram i sail shipping (Nicholls I 97 I: 92 >
I 2 o ) . The growth o f Hadhrami estates in the M am brui area o f
254 W illia m G ervase C l a r e n c e -S m ith
coastal Kenya from the l8 6 os to the l8 8 o s tem porarily boosted

this nexus o f H adhram i shipping. M illet and sesame from
M am brui were destined not only for Hadhramaut, but also for
a wider market. Ahm ad b. Salim [Salim?] Ba Sharahil, o f maskin
origins and b orn in Hadhramaut, was a trader in Lamu, while
his brother 'A ll owned an estate in M am brui. Together, they
owned two dhows, and dispatched sesame, sesame oil, m illet,
goats and skins to Arabia and India (C o o p er 1977 : 83 ~ 5 ’>
1886: 49 —5 °)* A lthough the fall in grain prices, the abolition
o f slavery, and the spread o f steamers soon brought this little
boom to an end, dhow traffic received a tem porary boost from
the First W orld War, when many steamers were sunk or
employed in war duties (Salim 1973 : 183) - However, the advent
o f m otor boats led many Hadhramis in Lamu to dispose o f
their sailing vessels (Le G u en n ec-C o p p en s 1979 : 95 )*
Hadhram i dhows nonetheless persisted on a small scale in
eastern A frica, up to the Second W orld War. In addition to the
older staples shipped to Zanzibar, Hadhramis brought gar
ments, honey, dates, henna, butter oil, and shark oil in the
1930s. T h eir return cargoes consisted mainly o f m illet.
M oreover, Hadhramis developed a lucrative business shipping
live animals from the East A frican m ainland to Zanzibar and
Pemba, where they also had a near m onopoly o f butchers’
shops. O n e such was Nasir b. 'Abdallah a l-K ath iri, who
supplied Zanzibar with meat on a large scale in his own dhows,
and accumulated enough wealth to have four wives. He was
expelled by the British in 1928, accused o f leading an anti-
O m an i faction o f Hadhramis blam ed for serious rioting on the
island (Ingrams 1937 : x5 8 ; Ingrams 1942: 4 4 —5 > 287—8).
T h ere was no sign o f Hadhramis taking to steamers in
eastern A frica. To be sure, M uhammad b. 'A ll Ba Kashmar,
b orn in Hadhramaut and holding high office under four
sultans o f Zanzibar, visited Britain several times as Sultan
Barghash’ s representative to purchase ships for the sultan’ s navy
(Pouwels 1987: 170—i). However, Hadhramis do not appear to
have made use o f this experience to branch out into steam
navigation.
T h e C om oros and southern A fric a
O m ani com petition encouraged Hadhramis to locate them

selves beyond the southern limits o f the O m ani em pire,
notably in the C om oros. T he great advantage o f these islands
was that Hadhram i lineages ruled them until French annexa
tion. In the veins o f the various petty sultans flowed the blood
o f several great H adhram i sa p id families, notably the Bin al-
Shaykh A bu Bakr b. Salim, already encountered as shippers in
Palembang and Persia. O th er prom inent families o f H adhram i
origins were the al-Masila, al-A ydarus and Jamal al-Layl.
A lthough these families had becom e quite local over the
centuries, a process o f're -A ra b isa tio n ’ began in the eighteenth
century, reinforced by a renewed current o f im m igration from
Hadhramaut (Shepherd 198O: 88 —9 3 ; M artin I 9 &3 > v°h I:
47 “ 9 ; el Z ein I 9 7 4 : 131). T h e Jamal al-Layl came to
prom inence on the other side o f the ocean, becom ing sultans
o f Perlis in Malaya in the m id-nineteenth century, a position
they have retained to our own day (Mohammad 1969).
Despite this favourable political situation, Hadhram i ship
pers never became as dom inant in and around the C om oros as
in the Malay world and the Red Sea. O m ani captains ranged
well beyond the southern limits o f the sultanate o f Zanzibar,
and were frequently encountered together with Hadhramis in
the C om oros, northwestern Madagascar, and northern M o
zambique (N icholls 1971: 3^0 ; Ferrand 1891—19°2, vol. 3: 78;
G randidier 1908, vol. 4 -1 & 4 ~ 2 : 2 2 3~ 4 , 321, 40 8-II, 658;
Renault 197^: 154 )- Gujarati merchants and shippers were
unusually weak in the Com oros themselves, but the same could
not be said o f north ern M ozam bique and northwestern
Madagascar. French and Portuguese traders, many o f them
locally born and o f mixed b lo o d , were also active competitors
(Frere 1873; H afkin 1973 ; Cam pbell 1995 )*
Th e C om oros experienced a b rie f burst o f prosperity,
roughly from the 1820s to the 1870s, as Malagasy raids were
contained by M erina expansion and the illegal slave trade
flourished. The islands were ideally situated to smuggle in
slaves from the m ainland, especially M ozambique, and send
them on legally as libres engages under the French flag to sugar

planters on the French islands, notably R eunion [Bourbon].
Th e Portuguese abolition o f the slave trade in 1836, given teeth
by the A nglo-Portuguese treaty o f 1842, placed much o f this
M ozambican traffic in Arab hands (C larence-Sm ith 1989a;
Cam pbell 1989; Shepherd 198O; Renault 1976; H afkin 1973)*
Inform ation about the illegal slave trade is somewhat patchy.
Strong personal links between sultans in the C om oros and
dom inant M uslim fam ilies along the coast o f n orth ern
M ozambique were revealed in correspondence seized by the
Portuguese authorities in 1879* T h e presence o f Hadhram i
sayyid lineages in both areas apparently facilitated trust between
the parties concerned. European surveillance made the use o f
the C om oros themselves rather dangerous by this date,
deflecting trade to ports in the lowlands o f northwestern
Madagascar (Hafkin 1973 '- 54 —7) * Shipping and trade became
m ore distinct over time, as ship-owners in the C om oros hired
out their dhows to others fo r increasingly risky slaving
operations. Crew members were often slaves, but both servile
and free men appear to have received part o f the profits o f
voyages (Shepherd 1985: 164—6; Shepherd 198O: 75—Ô, 83).
Ships in the C o m o r o s ’ trade with M ozam bique and
M adagascar were generally small and o f Indian O cean
construction and rigging. Arabs were caught shipping slaves
in such pangaios near M ozambique Island in 1873 (Hafkin 1 9 7 3 :
268). Dhows employed for the short hop across the M ozam
bique channel, were 35 to 4 5 feet long, partially decked at
either end, with a single mast raked forward and a triangular
lateen rigged cotton sail. They carried a crew o f around seven
m en. However, large dhows from further north, o f around 150
tons or even m ore, were occasionally to be found loading slaves
in these waters (Cam pbell 1989: 181—3 ^ 186).
T h e C om oros also m aintained long distance shipping
relations with parts o f the A frican coast to the north and
south. To the north, shipping connections were especially close
with the Lamu archipelago (Nicholls 19 71 : 92, 120, 354 )- To
the south, Arab traders from the C om oros sailed to Cape Town
as early as 182O. They were well received by the British
T he R ise a n d Fa l l o f H a d h r a m i S h i p p i n g 257
authorities, and they maintained trading and religious links at

least until the early l860s. Areca nuts were among the goods
supplied to South A frica, probably originating in southern
India (Mayson 1963: 12, 15, 27“ 8).
The Bin Sumayt family provide a famous example o f
successful H adhram i shippers in the C om oros. B orn in
Shibam in the centre o f the main valley in Hadhramaut, and
with a branch o f the family present in Java, Sayyid A bu Bakr b.
A b dallah b. Sumayt left his hom eland around the m iddle o f
the nineteenth century to settle in Itsandra, on the island o f
Ngazija [Grande C om ore]. A successful ship’ s captain and
trader, he came to own seven dhows, despite a tem porary crisis
in his business. His operations embraced southern Arabia,
India, Madagascar, M ozambique, and East A frica. His piety and
learning earned him an appointm ent as a Shafi'i qadi in
Zanzibar in the reign o f Sultan Majid. He retired to Ngazija
after Sultan M ajid’ s death in 1870, and died there four years
later. His son, Sayyid A hm ad, born in Ngazija in 1861, began
life plying the Indian O cean on his father’ s ships from
Madagascar to Lamu. However, on his father’ s instructions,
Sayyid Ahm ad exchanged the life o f the sea for that o f
scholarship, becom ing the most famous ' alim in Zanzibar until
his death in 1925 (Pouwels 198 7 : 13$—9 , *52—8; Salim 1973 :
1 4 2 -3 ; M artin 1971: 541—5).
T h e prosperity o f shipping based on the C om oros did not
last. Britain opened up labour recruitm ent in India for French
planters in the 1860s, in return for m ore vigorous French
repression o f the slave trade (Renault 197^, ch. 5)- However, it
was the collapse o f sugar prices and the spread o f colonial rule
in the 1880s that reduced slave exports to a trickle. To be sure,
some slaves continued to be exported from northern M ozam
bique in Arab dhows until i g i 4 > bound for M iddle Eastern
markets, but the trade was a mere shadow o f what it had once
been (C larence-Sm ith 1989a). A nother problem for local
shipping was the arrival o f BISN steamers in 18 74 > connecting
Zanzibar, the C om oros and northwestern Madagascar in a
regular service (Blake 195^: 149 )-
C o n c lu sio n
Most H adhram i shippers were 'innovative entrepreneurs’ ,

seizing new opportunities as they em erged, rather than
'ro u tin e e n trep ren eu rs’ , exp loitin g existing skills (Aung
1989). W hen the profits from shipping declined, they moved
their funds into m ore remunerative activities in the lands o f the
diaspora, notably trade, real estate, m oneylending, and,
eventually, m an u factu rin g (Freitag and C la re n c e -S m ith
1997). Even in the C om oros, where the golden decades o f
sail navigation gave way to prolonged econom ic stagnation,
sa p id lineages were able to exploit their status and educational
advantages to exert great influence on the religious life o f
eastern A frica (Shepherd 1985; el Zein 1974 ; M artin 1971)-
Furtherm ore, Hadhramis rem ained quite prepared to invest in
shipping if the opportunity arose. For instance, Pasha Ba
Khashab, an Eritrean Hadhram i, moved to Saudi Arabia after
the Second W orld War to found the first Saudi shipping line
(Q u'ayti 1996: 26).
Both the rise and fall o f Hadhram i shipping can thus be seen
as part o f a wider success story, the expansion o f a great and
versatile entrepreneurial diaspora around the Indian O cean.
T h e Hadhramis were akin to similar innovating com m unities
in the region, such as the Gujarati or the Hokkien, both o f
whom owed a great deal o f their success to early sail shipping
ventures. T he onward march o f the H adhram i diaspora was
tem porarily checked in most parts o f the Indian O cean by the
upheavals surrounding decolonisation, but there are encoura
ging signs that it is beginning to flourish once again (Freitag
1997)-
9
C o n tin u ity and A d a pta tio n
by C ontemporary S w ah ili
B o atbu ild ers in K enya
Susan Beckerleg
T he Swahili world extends from Somalia in the north to

M ozambique in the south, and occupies the space between the
land mass o f the A frican continent and the Indian Ocean.
W hile Swahili traders and their language and culture have
penetrated and enriched East Africa, much o f their orientation
continues to be towards the sea and the countries that border
the Indian O cean. In Kenya, one o f the m ain ways that the
Swahili people (WaSwahili) distinguish themselves from their
non-Sw ahili neighbours is to refer to them as the watu wa bara
(inland people).
Over the centuries trade and intellectual exchanges have
resulted in a recognisable Swahili culture which shares elements
o f diet, music, building styles, religion, medical practice and
bo at-building. The ocean going dhow trade has been in decline
throughout the twentieth century (Martin and Martin 197^)-
Nevertheless, traditional maritime crafts remain an important
part o f Swahili culture and economic life along the East African
Coast. A number o f different types o f boats are built by
craftsmen using tools and skills which appear to be little changed
for centuries. Observation o f the contemporary construction and
use o f boats was made largely on the Kenyan coast, particularly in
the tourist centre o f Watamu and on Lamu, the Swahili island
centre to the north. In addition, boat types were observed in
259
26 o Su sa n Beckerleg
large and small ports along the length o f the Tanzania coast in
1997 ’ and on Zanzibar in 1986. In 1996 I visited the boat yards
o f Sur in O m an where Swahili-speaking craftsmen work along
side other Omanis and migrant Indian specialists.
I will here review the types o f boats under construction in
Kenya and the uses to which they are put, outline the ethnic
identity, division o f labour and movements o f contem porary
seamen and boatbuilders, and describe in detail the building o f
a mashua, a small all-purpose sailing boat.
T h e Boats
W hile the vessels under construction and in use in Tanzania

and Kenya are similar, there is variation in the local names
used to refer to them. H ere the everyday terms used by
contem porary Swahili in Kenya are reported.
Some ocean going vessels are still built in a num ber o f
centres in Kenya, including Lamu and Watamu, and are usually
fitted with engines that enable them to ply routes between East
A frica and the Arabian G ulf. A ll such vessels are referred to as
bourn, although their design does not conform to the bum o f
O m an (Ministry o f N ational Heritage and Culture, Oman I99l)« O n e
bourn was under construction for several years in Watamu. W hen
the hull was finally com pleted, it was towed to Mombasa for
fitting out, before em barking on a career as a cargo vessel
voyaging up and down the East A frican coast and across the Red
Sea to South Arabia.
Most contem porary building projects are to construct boats,
known as ja h a ii or mashua, that rem ain in coastal waters.
A ccording to Prins, the ja h a ii is the com m on sea-going coaster
o f Lamu, while the boats o f the same design may be called
mashua if they do not sail the high seas (1965, 74)- He also
reports that:
The term mashua itself is H industani in origin, but is also

found in the Persian G ulf, O m an and Mahra. But there it
seems to refer to a small rowing boat mostly, especially a
ship’ s boat. (Prins, 1965* 7 ^)
C o n t in u it y a n d A d a p ta tio n by Sw a h i l i B o a t b u i l d e r s 261
These days the jahazi is used to transport goods, having been

adapted for use with an engine to ferry passengers from
m ainland to islands and between islands as well as for
excursions o f European tourists. Its main industrial use is in
the transporting o f mangrove poles which have been cut for the
domestic building industry and for export to Arabia. Some
transportation o f other goods still takes place, including, in
recent years, the smuggling o f h eroin from South Asia.
Ideally, the smaller mashua is crewed by three men and is used
for fishing and, increasingly, in the tourist industry. Mashua are
manned by crews carrying out deep sea fishing with lines to catch
sharks and other large fish or by dragging large nets (jarifa)
behind the boat. Some mashua have been fitted with outboard
engines for game fishing for tourists. It is largely the demand for
mashua for tourism that has stimulated the boatbuilding industry
in Watamu and attracted craftsmen to the town.
However, the dau, whence the generic term 'dhow’ for a
sailing boat is derived, remains in com m on use. It is a small
flat-bottom ed vessel used for laying traps and other fishing
activity along the shore line and within coral reefs, particularly
during the southern m onsoon (kusi) when the sea is rough. The
dau is popular amongst the northern Swahili, particularly the
Bajuni o f northern Kenya and southern Somalia.
T h e mashua and dau appear to have largely replaced the m tori, a
small keeled vessel, pointed at both ends. C urrently only one
mtori works out o f Watamu as a fishing craft with a Bajuni crew.
The mtori was reputedly a popular vessel amongst Bajuni, yet
surprisingly, Prins (1965) makes no m ention o f it.
O utrigger boats, ngalawa, are not constructed north o f
Mombasa in Kenya, but are popular in Tanzania. They are used
for line fishing, and although small and frail-lookin g they can
and do sail the high seas. A ccording to Prins, 'this type o f canoe
which is generally considered Indonesian in origin, must here
be derived from Madagascar and (or) the C o m o ro ’ s’ (Prins,
1 9 6 5 ,8 1 )
Finally, the huri, a simple dug-out canoe, is used for fishing
within creeks and inlets and for ferrying passengers from jahazi.
It is similar to the dug-out canoes, also called huri, used in
262 Susan Beckerleg
Om an, for fishing and ferrying passengers to and from larger

boats (Ministry o f N ational Heritage and Culture, Om an, ^99 ^)-
T h e Seam en
Watamu, a small but rapidly growing tourist town on the Kenya

coast, is not known for its m aritim e culture. Yet, the
com bination o f excellent fishing grounds, the existence o f
the M arine National Park and a buoyant tourist industry have
com bined to make Watamu a thriving boatbuildin g and
seafaring centre (Beckerleg, 1 9 9 1)-
T h e core com m unity in Watamu consists o f Bajuni fisher
people whose original hom e is to the north o f Lamu. They
moved perm anently to live next to their seasonal fishing
grounds in the 1960s after their m ainland villages on the north
Kenya coast were attacked by Som ali bandits (Beckerleg, 1994*’
Beckerleg, 1995)• From I 99 1* the migrants have been jo in ed by
Som ali refugees fleeing war in their country. Many arrived in
small fibre glass fishing boats, while others were passengers on
large dhows or m odern ships. H undreds drowned o ff the town
o f M alindi when their ship ran aground on the reef. T h e new
Bajuni/Som ali residents o f Watamu have continued their
traditional work as fisherm en. W orking together with Bajuni
families who have lived in Watamu for several generations, a
num ber o f mashua vessels return to fishing grounds within
Som ali territory. Fish are either sold directly to refrigerated
boats that are positioned in the area, or are dried and salted on
beaches before being transported back to Kenya. T h e catches
are good, but there is a danger o f piracy.
In Watamu, fish is traded in a purpose built, developm ent
agency-funded, market on one o f the three beaches in
Watamu. In 1997 there were about 10 mashua, owned mostly
by Bajuni entrepreneurs, and seven ngalawa working out o f
Watamu. The ngalawa all belonged to fisherm en from the island
o f Pemba, o ff the north coast o f Tanzania. O riginally, the
Pemba came as seasonal fisherm en and some return home to
spend the kusi period there. O thers have m arried Bajuni women
and are integrated into the life o f Watamu.
C o n t in u it y a n d A d a p ta tio n by Sw a h il i B o a t b u il d e r s 263
T h e C raftsm en
A t any one time there are between five and ten traditional
craftsmen resident in Watamu. Most o f the boatbuilders are
Bajuni, with origins in either Kenya or Somalia. Several work
exclusively in Watamu, while others may go and work in other
towns as necessary.
The boatbuilders are not organised into any form al union
or guild, but do follow a customary code o f practice. For
example, once one craftsman has started work an a new boat,
another will not take over the job , no matter what profit could
be obtained. Similarly, if a boatbuilder is to work on another
craftsman s boat, he must obtain his perm ission first. In spite
o f the tem ptation o f easy money, I never saw either o f these
rules broken in Watamu. Further, the rule that a boatbuilder
never owns and runs his own boat appears to be strictly upheld.
T he boatbuilders in Watamu are presided over by the chief
boatbuilder. This restrained and elderly Bajuni from Somalia
gained his position through dint o f seniority, skill and
experience. He is absent from the town much o f the time
working, with his sons, on ocean-going bourn in Mombasa. Back
in Watamu at weekends, the chief craftsman has no direct
powers, yet he commands the respect and is able to influence
the conduct o f the other boatbuilders. Most boatbuilders
follow his counsel, and he thus maintains the code o f honour
which is under threat from the influence o f a tourist econom y
that encourages the abandonm ent o f strict codes in favour o f
fast money. Yet, while disputes with clients over payment and
the execution o f work are com m on, as a professional group
with a code o f practice, they maintain a united front.
Boatbuilders acquire their skills through a long period o f
apprenticeship. T he convention was that they would work,
largely unw aged, under a master craftsman until he judged that
the apprentice had gained sufficient skill and experience to be
given his own bag o f tools, m arking his debut as an
indep en dent craftsm an. W ith the advent o f the tourist
economy, this system is being eroded. In 1998 young Swahili
m en could be observed learning boatbuilding skills from
264 Su sa n Beckerleg
craftsmen on the beach at Watamu. However, they are unlikely

to be w illing to labour long and hard for little reward, and
some o f the heavy work has been taken over by m en and boys o f
the nearby Giriam a ethnic group.
In M atondoni on Lamu island, where boatbuilding is
traditionally a major activity, one master boat builder told me
how since his apprentice had died, he had worked alone without
assistance. O f two other form er apprentices, one had gone to
work for the Forestry Department, and the other had become a
prison guard. He added that o f ten craftsmen in Lamu he did
not know any who had an apprentice. This breakdown o f the
exchange o f labour for the acquisition o f a skill is recent, as even
quite young men working in Kenya acquired their skills through
adherence to the traditional system.
T h e Giriam a people, who share with the Bajuni mythical
origins in Shungwaya somewhere to the north o f Lam u ( deVere
Allen, 1 9 9 3 ) now live side by side with the Bajuni around Watamu
and M alindi. T he Giriam a are farmers and have no m aritime
tradition. However, many Giriam a have been influenced by
Swahili culture, some have converted to Islam and others have
started fishing and even boatbuilding (Parkin, 19 8 $ ; Parkin l^ Q l) .
In Watamu, a M uslim Giriam a who was trained in Tanga,
Tanzania in assisting a boat construction now competes with
established craftsmen in mashua construction.
In Kenya, M atondoni is the traditional centre o f b oat
building and one Watamu craftsman was born and trained
there. As well as in Watamu, he has also worked in Tanga in
Tanzania. Like his brother who is also a craftsman, he learnt his
trade as an apprentice. However, it is rum oured that he left his
teacher before he had received his blessing along with the
customary basket o f tools which mark the independent status o f
a master craftsman, a fu n d i. Nevertheless, this man is a skilled
boatbuilder and is hereafter referred to as 'Fundi . In 1986, I
had the opportunity o f observing the building o f a mashua from
start to finish. Fundi was the com m issioned builder, although
at one time or another most o f the boatbuilders in Watamu
worked on the boat. Below is a detailed account o f the technical
and ritual process o f Swahili boatbuilding.
C o n stru ctin g a Vessel
Th e story o f the building o f a boat can be told simply and

accurately in just a few sentences. M toro bin Mwinyi Bakari,
writing on life in nineteenth century Bagamoyo in Tanzania,
describes the process o f boatbuilding that was used then and, as
I saw, continues:
First the builders go to the forest to split planks and to

find a strong keel and strakes and ribs. W hen they have
these, they bring them to town, and ask their master
craftsman and the teacher to cense them as they begin
their work, so that it goes well.
First the keel and the ribs, the stem and stern posts and
the guide plank. They fix the planks with nails wound
round with cotton and coconut oil to make them firm .
W hen the ship is finished, they deck it fore and aft. T h en
they caulk it by filling any open holes with cotton and oil.
T h en they p ro o f it with shark oil and fill the ship with
water for seven days or fourteen to see if it leaks or not.
T h en they fit the tiller, and when that is in place they sew
the sails and select a mast, a bowsprit, and a yard and an
anchor. T h en they spin ropes for hoisting and for
lowering the anchor, and the ship is launched’ (Bakari,
1981 , p 124)-
W ritten in the nineteenth century in Bagamoyo, Tanzania, this

description outlines the sequence and basic technique used by
Swahili boatbuilders, but tells little o f the ritual used to protect
the new vessel and to mark the im portance o f the relationship
o f the craftsman to his client. Here I tell the story o f the
building o f one small mashua, com m issioned in Watamu by Sizi,
a Bajuni man, and built by Fundi under the gaze o f Bajuni
sailors. Boatbuilding rituals also draw in the wider com m unity
who take considerable interest in the construction o f any new
boat. From the handshake o f agreement between Fundi and
Sizi, to the launching o f the vessel, took eight m onths. A
further m onth passed between the launch and the maiden
voyage. I was able to observe the painstaking work o f Fundi, the
266 Susan Beckerleg
F igu re 9.1 Fundi using the bow drill
social network o f the boatbuilders, and the ritual procedures

deem ed necessary to ensure the success o f the venture and to
protect all those connected to the new boat.
In his classic m onograph, Sailing From Lamu (1965), Prins
outlines two main types o f agreem ent between master boat
builders and the prospective owner (tajiri). T h e boatbuilder may
be paid a lum p sum, out o f which he must recruit and pay
workers labouring under his supervision. In the second form o f
contract the tajiri pays the master boatbuilder a lesser sum, but
is then also responsible for paying the other boatbuilders on a
daily basis. In either case, according to Prins, the tajiri should
send food to the boatbuilders (Prins, 1965.* 93). T h e b oat
builders bring their own bag o f tools, including adze, saw and
bow -drill.
A fter much discussion, Fundi settled on a fee o f J 0 0 0
Shillings (approximately £400 at 1986 rates o f exchange), with
10% to be handed over in advance. Sizi would have to pay
various workm en chosen by Fundi. For his part, Sizi specified
that he wanted a mashua with a slanting prow like an Arab bourn,
rather than the conventional near-vertical prow. O n e such

craft had already been constructed in Watamu, and Fundi was
happy to try out this new style. T h e verbal contract also
encompassed the traditional proviso that if the planned mashua
went wrong, Sizi would have no comeback nor receive any
com pensation.
W ork finally started by cutting wood in the governm ent
managed Arabuko-Sakoke Forest behind Watamu. Fundi was
able to obtain perm ission to enter the forest to cut tree
branches for the ribs (mataruma). A t this point the disputes and
battle o f wills between craftsman and client began. Seventy ribs
were ordered — a large num ber are needed because many prove
the wrong shape and are discarded — but only forty three were
delivered. Fundi argued that there were sufficient ribs for his
purpose, but Sizi felt it necessary to have the wood checked by
another boatbuilder. T he 'simple cerem ony consisting o f the
burning o f incense over them (matarum a), invoking the name o f
A llah ’ (Prins, 19^ 5 : 9^) was om itted, and not even suggested by
Fundi. Sizi, w ondering if this bad start augured ill, considered
com m issioning another builder. He approached one, but as he
wanted far too much money, and as Fundi had already received
his first instalment, Sizi decided to stay with the original
agreement.
Fundi now introduced his assistant, A li, who was to do much
o f the early work before falling into dispute with Fundi. A li was
a M uslim Giriam a who had started his working life as a charcoal
cutter. He was then employed by Bajuni to cut boat ribs in the
forest. Gradually, as they saw that he was good with wood they
started teaching him how to work with an adze, the main tool in
boat construction. A li received his m ain training in Tanga,
Tanzania, where he spent some three months, for which he
never received payment. Back in Watamu, Fundi su b -co n
tracted many o f the easier, but still skilled, work to A li.
A ccordingly he was responsible for selecting, cutting and
transporting tim ber for the keel, stern and aft posts from the
forest. These he 'cleaned’ with the adze, ready for the approval
o f Fundi and Sizi. He was paid for the delivery o f the actual
wood by Sizi, and had a private arrangem ent with Fundi over
268 Susan Beckerleg
F igu re 9.2 Sizi poses while Fundi works
the cost o f his labour. However, A li started dem anding m ore

money and was clearly attem pting to style him self as a master
boatbuilder. A fter a few days work, he quarrelled with Fundi
and went to work on other B ajuni-ow ned boats in Watamu.
O nly at this point, once the keel is ready, is the construction
o f a new boat said to have started. T h e 'cleaned’ keel also
marked the beginning o f a sustained criticism o f the work,
materials and craftsmanship. T h e keel was too short, everybody
said, especially Sizi’ s cousins! T h e boatyard rapidly became a
m eeting place for Bajuni m en and boys, some passing from the
beach, but others making a special journ ey to check Fun di’ s
work. Bajuni m en and boys, most o f whom are excellent and
knowledgeable sailors, all seem to be experts on the construc
tion o f boats, and even when this is not the case, are happy to
discuss the work o f others in m inute detail. Fundi took the
criticism in his stride, apparently enjoying the company, and
advising me that it was all in the norm al course o f events. T he
great attention the new boat was receiving did, however, worry
Sizi, He quickly came to the view that the boat would need
protection against envious ill-wishers. Spurning resort to

techniques forbidden by Islam, he felt protection would best be
obtained from the Q uran and called local Muslim teachers to
read at the boatyard.
Planks for the hull could be obtained only from saw mills
but were in short supply. A fter great effort Sizi purchased
planks in Mombasa and arranged for them to be transported by
bus to Watamu. T h e quality o f the wood was excellent and this
made Sizi nervous o f both theft and the power o f envy. The
delivery o f the planks meant that it was time to em ploy an
assistant for Fundi. A young Giriam a boy, the brother o f the
man who erected the shelter for the boat, and a neighbour o f
A li, was engaged at 2 0 / - per day. This m oney was to be paid by
Sizi, and was separate from Fundi’ s fee.
The first person approached by Sizi to read prayers for the
protection o f the boat was a local Sharif, a descendent o f the
Prophet M oham m ed. Sizi was not reassured by this first reading
and the following day arranged for a bigger event, including an
offering or Sadaqa. A mixture o f p opcorn and sugar was
prepared and taken to the boatyard in the m orning. A teacher
from the mosque school arrived, shortly after, with seven small
boys and for about fifteen minutes they recited a variety o f
prayers, with the boys and a few bystanders intoning the
responses in a matter o f fact fashion. A t the conclusion o f
prayers the p opcorn was handed around to all present. Fundi,
who had worked throughout the reading, stopped to eat some
popcorn. Afterwards I spoke to the teacher who had led the
prayers. He said that a num ber o f different prayers and
Q uranic verses had been used to keep those involved with the
new boat safe and to counter envy.
The new assistant was away on an errand whilst the reading
was being co n d u cted , so he missed the distribution o f popcorn
and nobody rem em bered to save him any. A ccording to Fundi
this was an oversight, but not a serious one, as ill wishers were
assumed prim arily to rank amongst those friends and relatives
who would especially covet their own boat. But the assistant
proved lazy and uninterested in the work, as he considered that
the 2 0 /-, constituting the legal m inim um wage, was insufficient
270 Susan B e ck e rle g
F igu re 9.3 Fundi’ s assistant awaits instructions
reward for his labour. A fter a week or two he left the jo b ,

thereby missing an opportunity to learn a craft that was in
dem and, and potentially to com m and a high financial reward.
W ith all wood and nails supplied, work proceeded. Fundi
worked almost completely by eye, disdaining the spirit measure
favoured by other Swahili craftsmen. He took measurements
based on the cubit, the distance from his elbow to finger tips,
and the boat under construction was referred to as having a 12
cubit keel. First the stern and aft posts, which had been
prepared by A li using an adze to carve the wood into the
appropriate shape, were attached to the keel.
C o n tin u ity and A d a p tatio n by S w a h ili B o a tb u ild e r s 2J1
T h e next task was to fit the first planks into the two sides o f
the narrow keel. In order to shape the planks, a small fire was
lit and a plank placed in the flames. Using crude wooden tongs,
each plank was bent into the correct shape for insertion into a
small groove along the two edges o f the keel. T h e work was slow
and painstaking, as the angle and fit o f the planks largely
determ ined the shape o f the finished vessel. A fter the first
planks were inserted and pushed in by gentle ham m ering,
Fundi started fitting the tem porary ribs. He carefully studied
the branches brought from the forest, searching for the ones
that were already close to the required shape. He then set to
work with his adze working on each branch, fashioning it into a
rib. T h e ribs were fitted into the boat so that they stuck up in
the air hinting at the shape to come. As noted by Prins (1965),
m ore planks were fitted around the ribs until five planks on
either side were in place.
With the keel, stern and aft posts, tem porary ribs and lower
planks in place the boat started to take shape. T he next stage was
to fit the transom, or durusi. Fundi warned Sizi that this was a
vital com ponent requiring great skill. Fundi now spent hours
staring at a single piece o f wood and marking it with black lines
made from the contents o f discarded household batteries. Next,
he started cutting the wood with his adze and tacked the finished
piece to the boat so that he could study its shape further. A t this
point the m urm ur o f criticism from the observers increased.
Most people seemed convinced that the durusi was the wrong
shape and would slow the boat or make it unstable. Fundi,
pondering his work, came to the same conclusion. A new durusi
would have to be made and Sizi would have to find the wood. A
piece o f cam phor was procured from Mombasa and delivered to
Fundi. This time no mistake was made and the new durusi met
with the approval o f the ever present critics.
T h e fisherm en com ing from the beach and seeing a fitted
durusi asked when the goat was to be slaughtered. A ccording to
custom, a goat should be purchased by the owner o f the new
boat and the meat distributed to the fishing and maritime
com m unity on the beach. Sizi resisted this expense and instead
gave Fundi a tip. From this point, coinciding with the start o f
272 Su sa n Beckerleg
F igu re 9*4 Fundi prepares to fit the final plank
the high tourist season, work slowed and Fundi started doing
jobs for other contractors. T h e relationship between Fundi and
Sizi deteriorated. Yet, Sizi was advised that other craftsmen
would refuse to break their professional code and take on the
jo b . Sizi’ s family also advised that any such break in tradition
would result in disaster, such as the loss o f life when the boat
inevitably sank.
Fundi continued working on the boat in fits and starts: the
perm anent ribs were selected and each carved to fit its place in
the boat exactly. T h e planks making up the hull were put in
position and smaller pieces o f wood fitted to fill the gaps
C o n tin u ity and A d a p tatio n by S w a h ili B o a tb u ild e r s 273
between the planks. These were attached to the hull with pieces
o f wood nailed onto the planks. T he inserted wood was eased to
fit the spaces through delicate work with the adze. W hen these
nails were removed each hole was plugged with a piece o f wood
fashioned to fit it. A ll this took months o f painstaking work,
interspersed with Fundi’ s absences. Unlike the boat yards I
observed in Sur in O m an, no labour saving devices nor
electrical equipm ent were used. Before a plank could be fitted
to the hull, the desired shape had to be marked and the wood
hand sawn, a procedure that often took all afternoon.
W ork next started on the structural features inside the hull,
including a deck. T he fitting o f one o f these, the fu n do, or beam
to support the mast, marked another customary gift. Sizi
should now have given Fundi a chicken; he gave him a cash tip
instead.
O th er craftsmen, with Fundi’ s permission, started working
on the boat. Fundi, however, retained overall control and
started talking about the kambe, or ritual prohibitions sur
rounding new boats. It emerged that already one im portant
kambe had been violated: months earlier, people had sat on the
keel. Previously Fundi would have fined them for this offence,
now he does not bother and the rules are falling into abeyance.
Fines would have been paid in the form o f food , and 'not just
beans’ . O ne kambe that Fundi tried to maintain was that nobody
should sit in the boat just for fun, and they should never climb
in wearing shoes. A noth er kambe was that nobody should
approach the boat carrying a walking stick. Rather it should be
laid on the ground so that it could not touch the boat.
W ith com pletion at last in sight came the time to caulk the
vessel with cotton soaked in coconut oil. Fundi supervised this,
but a num ber o f the regular group o f critics jo in ed in caulking
in a sudden outburst o f comm unity spirit. O nce again, Fundi
inform ed Sizi that a special gift was required by custom.
Sweetened water, maji tamu, should be distributed to everybody
who passed the new boat. T he measure was to diffuse envy. Sizi
agreed to provide the refreshments and the older m en seemed
to appreciate this belated respect for custom. W ith the vessel
almost ready for launching, Fundi wanted to test its seaworthi
274 Susan Beckerleg
ness. T h erefore, the boat was filled with water and the degree o f
leakage observed over several days (Bakari, i g 8 l ) . Any gaps in the
hull were filled in by Fundi. Finally, using buckets, the water
was emptied.
Sizi wanted to launch the boat as quickly as possible. He
required Fun di’ s perm ission, which he requested and was
granted. In Swahili culture a boat launching is often a big event
marked by ritual and celebrations that draw in the entire village
or town. Sizi, wary o f m ounting costs, opted for a quiet
launching that relied on the customary co-op eration o f the
m aritime com m unity. Indeed, the usual observers and critics
turned out on the appointed m orn in g to assist in the
launching. Several ropes were attached to the vessel and at
low tide the m en hauled the boat over the beach to the shore
line. A fter a wait o f a few hours the tide came in and the boat
floated. Fundi watched anxiously, but liked the way the boat sat
in the water, and was delighted to receive the enthusiastic praise
o f his erstwhile critics. Those who had assisted in the launching
were rewarded with goat meat, but following custom, expected
no cash paym ent. Fundi, however, did not receive the
customary gift that both Prins (1965) and Bakari (1981)
m ention as due to a craftsman at this point.
T h e boat still required a rudder, mast, sails and paint. W ith
F undi’ s perm ission, other craftsmen worked on the boat, and
were paid at a daily rate by Sizi. A considerable delay occurred
when it was time to make a metal brace (ramada), which would
fix the rudder to the transom (durusi). Two ramada, known as
'm ale’ and 'fem ale’ , were required, but no Swahili blacksmiths
with experience o f boatbuilding were to be found in or around
Watamu. Eventually in M alindi, Giriam a blacksmiths working
with an open fire and hand-held goat skin bellows, fashioned
the braces according to Fundi’ s specification.
Preparation o f the sails, however, was m ore easily achieved.
Four Bajuni and Pemba fisherm en volunteered to cut and sew
the cloth. O ccupying a wide path leading to the beach they laid
out the sail cloth and set to work cutting and sewing with large
needles they provided themselves. They appeared to be happy
to carry out this work for free and expected no payment.
C o n tin u ity and A d a p ta tio n by S w a h ili B o a tb u ild e r s 275
Indeed Prins observed that sailmaking is the only type o f labour

which requires no wage. He noted that:
. . . Sailmaking is an 'h o n o u r’ , sailors and passersby

considering it a favour to be allowed to assist the master
sailmaker — or so I have been told — and the craft o f
sailmaking is considered a venerable one. (Prins: 1965:
112)
Finally, only the making o f the hole into which the mast was to
be inserted rem ained. Such work requires skill, and a mistake
could result in the mast toppling over. The boatbuilders were
united in their assertion that this task must be perform ed by
the man who built the boat, for the hole is the 'heart’ o f the
mashua. Eventually, after much delay, Fundi made the hole and
thereby com pleted the terms o f his contract. H enceforth the
boat was considered ready for its maiden voyage.
D uring the last ten years the boat has perform ed well as a
traditional fishing craft. W hen working o ff Watamu the boat
sails into waters far beyond the reef and drops anchor. Baited
lines are then used to catch large fish, including shark and
kingfish. But the mashua has also sailed to fishing grounds o ff
Somalia and to Pemba island with the varied crews that have
worked her. In recent years she has been managed by a
fishm onger in Watamu who, after selling the catch, apportions
monetary shares to the crew and to Sizi who is still the owner.
This mode o f management follows customary lines and is an
effective means o f spreading the profits between all those with a
stake in a particular fishing vessels. Yet, possession o f or
working on a mashua can never have yielded great returns. Fish
are becom ing m ore scarce as East A frican waters are harvested
by Japanese and other foreign trawlers. In 1998 Sizi’ s mashua
was still fishing the waters o ff Watamu, despite being in need o f
repairs and further care from Fundi and his colleagues. By
early 1999 the vessel was beached while her manager decided
whether m eeting the costs o f repairs would be a worthwhile
investment.
276 Susan Beckerleg
C o n c lu sio n
Fundi and his colleagues do not come from prom inent Swahili
families, yet are accorded respect by the maritime com m unities
o f which they are a part. W hile fisherm en and other sailors
continue to rely on locally built vessels, they depend on the skill
o f the boatbuilders for their livelihood and safety at sea.
H ence, although hierarchy in Swahili society is usually based on
ethnicity (Beckerleg, i 994 >
* de êre Allen, 1993), boatbuilders may
achieve considerable status, despite hum ble origins, and even
the taint o f slave ancestry. T he senior boatbuilder in Watamu
has been num bered among the elders who are called upon by
local governm ent to manage village affairs. Craftsm en who
com bine their expertise with business acumen may also becom e
prosperous. However, Fundi has been unable to capitalise on
his skills and has rem ained a poor man often perform in g
itinerant labour.
Some observers have ascribed Fundi’ s m isfortune to his
defiance o f his master when he was an apprentice, so that he
works without the blessing o f his teacher. However, Fundi is
not unusual in working against or simply ignoring customary
rules, which are falling into abeyance. Yet, while many ritual
procedures have been dropped, traditional methods o f b oat
building have not only endured in Kenya but have been
adapted and developed to fit the changing maritime econom y.
T h e customary code o f practice is still adhered to by b oat
builders in Watamu, so that, while not organised into a form al
guild or society, they m aintain the following rules: they do not
own vessels; they will not work on a boat their colleague is
responsible for without perm ission; they accept no liability for
the finished craft.
Fundi’ s early rebellion paved the way, and boatbuilders have
recently been forced to adapt the harsh terms o f apprentice
ship. Nevertheless, master craftsmen continue to train young
m en in Watamu. Indeed, as the emphasis o f m aritime activities
has switched to tourism , m ore young m en have been attracted
into the boatbuilding industry. M ore passengers on mashua are
European tourists being taken to observe sea birds or dolphins,
C o n tin u ity and A d a p ta tio n by S w a h ili B o a tb u ild e rs 277
or to go on game fishing trips. But older links with Arabia and

the Indian O cean also endure. M otorised bourn, including one
constructed in Watamu, ply the waters from South Arabia to
Mombasa and on to Zanzibar. However, the tradition o f
maritime trade between Arabia and East A frica continues to
decline (Martin and M artin, 1 9 7 8 ) . O n the other hand, the use o f
smaller craft used for fishing and tourism is thriving in
Watamu.
Prins, in his study o f 'm aritim e culture in East Africa,
enthusiastically noted the names o f vessels he saw in Lamu and
other Swahili ports. He wrote:
Ships are neither mere items belonging to so-and-so nor

are they simply a dau, a mashua or a jahazi: they have an
identity o f their own. Always when people are talking
about them, they refer to any under their proper name.
T he names give curious inform ation about the ship
owners’ minds (Prins, 1965= 84).
In Kenya, 4 ° years later boat names still indicate local and

regional preoccupations o f their owners. The names o f the
mashua, dau and one mtori m oored at Watamu reveal a com m unity
aware o f both the potential o f tourism ('R am bo’ the H ollywood
character; 'C lub soda’ the soft drink and 'm ixer’ with whisky)
and current events in Arabia ('Arafat’ , the Palestinian leader
and 'Scu d’ an Iraqi missile) but also with a local orientation
('Subira’ m eaning patience in Swahili and 'U p ep o’ m eaning
wind) and a continental perspective ('A frica’).
N ote
E arnest H em m ingw ay, fam ous fo r his h u n tin g and fish in g as w ell as w ritin g,
is said to have visited nearby M alin d i, and on e o f the recen tly b u ilt hotels
in W atam u is called H em m in gw ay’ s.
IO
The H isto rica l C o n te x t o f

th e C o n stru ctio n o f th e
Va t t a i F i s h i n g B o a t a n d
R e la t e d F r a m e -F ir s t V e sse ls
o f T a m il N a d u and B eyo n d
Lucy Blue
This paper summaries fieldwork undertaken in Tam il Nadu,

Southeastern India in January 1997 (Blue et al. 1998). T h e
project is part o f a five-year program m e o f research funded by
the Society o f South Asian Studies, one aspect being an
investigation into fram e-first vessels o f Tam il Nadu. T h e term
'fram e-first’ (also known as skeleton) describes boats and ships
that are built by erecting the frames or ribs o f the boat before
adding the shell o f planks. F ram e-first con struction is
com m only observed, certainly within the context o f European
boat building, but is in contrast to the other traditional types o f
boats that are currently built along the shores o f eastern India
(Blue et al. 1997 )* T h e m ajority o f traditionally built boats in
India are constructed 'p lank-first’ , which involves most o f the
shell o f planking o f the hull being fastened together and to the
plank-keel and the posts, before the fram ing o f floors and
crossbeams are positioned in the hull.
Background to the T a m il N ad u Fieldwork
T h e unusual adoption o f fram e-first construction techniques,

and in particular the methods used by the Tam il shipwrights to
design the fram ework o f their ships, prom pted further study o f
these craft. Five types o f fram e-first vessels were encountered
278
C o n stru c tio n o f th e Vattai F i s h i n g Boat 279
on the Tamil coast. These included the T u ticorin thoni, the kotia
o f Cuddalore, the vattai fishing boat o f Palk Bay, the vattai
(a dumb lighter), and a second fishing boat known as the vallam.
It would appear that these seemingly disparate craft all belong
within a single boatbuilding tradition: 'tradition’ being defined
as 'the perceived style o f building generally used in a certain
region during a given time range’ . T he aim o f the 1997
fieldwork, therefore, was to record the Tamil Nadu fram e-first
methods in detail.
The earliest known vessels in the world were constructed by
plank-first methods (M cGrail 1997 )- It was not until the 15th
to 16th centuries that a marked shift from plank-first to fram e-
first m ethods was observed in the central and western
M editerranean. 16th and 17th century texts have survived which
describe, albeit not unambiguously, the design and building o f
ships by fram e-first techniques, methods which have becom e
known as 'M editerranean M oulding’ , and subsequently studied
by specialists such as Sarsfield (1984), Bellabarba ( l 993 )> and
Rieth (1996). W ithin the scope o f fram e-first construction a
varying degree o f dependence upon the form al design o f a
vessel has been noted: from building mostly by eye to total
dependence on drawings. This means that, at one end o f the
spectrum, the builder has in m ind an ideal framework that will
produce the desired boat shape. T he builder may m odify the
design depending upon the availability o f tim ber or to
incorporate some perform ance-enhancing feature. A t the
other end o f the spectrum, the builder will design a boat
based strictly on naval architects drawings, with no scope for
innovation and adaptation, relying solely on pre-designed
plans.
In 14th century Venice, however, it would appear that this
evolving innovation o f designing fram e-first ships was som e
where between the two spectrums o f designing fram e-first
vessels: between the builders’ inform al approach (rem iniscent
o f the plank-first building methods), and the naval architects’
form al fram e-first drawings. A num ber o f rules and p ro ce
dures were adopted whilst still allowing for individual scope.
Yet, this Venetian design m ethod still ensured a shape o f hull
280 L u c y B lue
I K a rik a i
Bay
of
A rliram pat in am
\fu u u p e« Bengal
'\M a llip a n n a m # J V ed aran yam
I N D I A F .r i p u r a k a r a T
s*r
Nambu Ih a la i-
Tondi
D cvipattinam m
Pulk
K il.ik .ii.u _ Ban
Tunconn G u l f
of
M a n n a r
S R 1
1 A N K A
C a lx Comorin
/4 -
5 /,
V
Oc ' A A-
Figure IO.I Coastal sites in southern Tamil Nadu (Drawing: L. Blue)
that p r o v e d to have d e sir a b le q u alities ( f o r e x a m p le capacity,

speed, stability etc.) a n d that c o u l d b e r e p e a t e d aga in a n d
again. In the c o u r s e o f u n d e r t a k i n g f i e l d w o r k i n T a m il N a d u i n
1 9 9 7 we o b s e r v e d f r a m e - f i r s t vessels b e i n g b u i lt b y a very
sim ila r d e sig n to that p e r c e i v e d as h a v in g b e e n a d o p t e d by the
V e n e t i a n sh ip w righ ts o f the 1 4 th c e n tu r y . T h e fo c u s o f this
p a p e r is thus, the s p e c ific d e sig n o f the f r a m e - f i r s t vessels o f
T a m il N ad u .
C o n stru c tio n o f th e Vattai F i s h i n g B oat 281
T h e 1997 Fieldwork in T a m il N ad u
A num ber o f sites were visited during the course o f our three
week field season in Tamil Nadu, during which all five vessel
types outlined above were observed and recorded to a greater or
lesser extent. This paper will focus on the vattai fishing boat o f
Palk Bay, observed both beached at Eripurakarai, and under
construction at a boatyard at Adiram pattinam ; it will also refer
to the larger thoni cargo ship o f T u ticorin (fig. 10 .i ).
The T u ticorin thoni is a three-m asted sailing cargo ship (now
m otorised) that is used for trading with Sri Lanka (fig. 10 .2 ).
These ships have a full form , with a near-rectangular cross
section, and they range in capacity from 225 to 650 tonnes,
with keel lengths from 29 to 38m. T h eir length/ beam/ depth
ration is 8 : 2 :1. 25 * The Tuticorin thoni was examined and
photographed whilst it was being built on the foreshore
building site at T u ticorin but not recorded in detail. T he mestri,
the master shipwright in charge o f building this vessel, was
asked about the design methods and building techniques.
T he vattai, open fishing boats o f much smaller dimensions,
are found all along the shores o f Palk Bay (fig. 10 .3 ). They were
docum ented in detail, whilst being built at Adiram pattinam ,
whilst beached at Eripurakarai and other nearby beach landing
sites, and whilst underway in the northern coastal waters o f Palk
Bay. A beached vessel at Eripurakarai was measured and drawn
at 1:10 scale, photographed from all viewpoints, and notes on
her structure were com piled (fig. 10 .4 )* T he inform ation
about the beached boat was backed up by interviews with boat
users.
Vattai, are flat-bottom ed, have a box-like transverse section
and are near wall-sided over much o f their length. They range
in size from around 13.72m long, with a beam o f 2.13m and a
depth o f 1.37m , to the smallest vessels o f c. 5 -l 8 m X 1.07m X
0.76m . However, irrespective o f their size, they are all similar
in shape with very high bows, and two or three masts each with a
settee-lateen sail, a balance board, and, uniquely on this coast,
leeboards.
282 Lu cy Blue
Figure 10.2 A thoni at a foreshore bu ildin g site at T uticorin
Figure IO .3 A vattai on the foreshore at Adiram pattinam

<u
"<Su
W
bio
f3
U? 03
u
,-d
’T(U3
^5
'u
<U
i"-
CD
03
3d
03
'— 5
<U
in
o
X
u<v
£
au
as
^h\_
03
Sh
2
V "C
W
E <u
in™
£
O
’-o
<u
-0
S-i
O
u
<L)
S-i
5
a
»
co 03
«-H
o
s
03
N -d
CD
5
a
<u
6
&>
OU
E
284 L u c y B lue
D e sig n in g tk e V attai
The vattai is a fram e-first boat and thus, before building can
begin, the fram ework has to be designed. This includes
designing a num ber o f frames o f the same size and shape
(known as 'equal fram es’), which form the central section o f
the boat; and a num ber o f pairs o f frames, which are erected
before and aft o f the central section o f 'equal’ frames. These
are known as the 'unequal fram es’ and produce the narrowing
and rising o f the hull.
Thus, in the vattai that we recorded at Eripurakarai, the
'equal fram es’ are the central fifteen frames. These fifteen
frames equate to seven sets o f equal frames that are positioned
either side o f a central or master frame, from the seventh from
the bow to the twenty-first. A ll fifteen frames are 'equal’ in size
and dim ension, their shape having been determ ined by the
shape o f the central or master frame.
Next the pairs o f 'unequal fram es’ are erected; one o f each
pair being placed forward o f the central section o f frames, and
the other aft. These are frames six and twenty-two; five and
twenty-three; four and tw enty-four. Each pair o f frames is
identical, but each pair is different from the equal frames and
from the other pairs. T he six other frames, or rather h alf
frames — the forem ost three in the bow and the three in the
stern — are not part o f the design process but have their shape
determ ined by the planking (they are essentially shaped by
'eye’). Regardless o f their size, all vattai have an odd num ber o f
frames, for example the one we measured had twenty-seven
frames. The master frame is erected at the centre o f the boat
and then each set o f frames is placed either side, thus form ing
the skeleton o f the boat.
T h e Tuticorin thoni also employs the same system o f equal
and unequal frames. O ne 225 ton vessel noted, was some 29 m
in length and had seventeen equal frames and ten pairs o f
unequal frames. T here is no fixed ratio o f unequal to equal
frames used in either the construction o f the vattai or the toni,
but by changing the ratio o f equal to unequal frames, the
shipwright changes the shape o f the hull.
To fix the shape o f the main hull o f the vattai, therefore, the
boat builder must produce at least four frame shapes — one for
the equal frames and three for the unequal frames. The num ber
o f unequal frames used in the design and construction o f the
vattai ensures the simple hull form o f the boat, with a long, box
like central length and only a short length where the hull is rising
and narrowing. Conversely, the relatively greater num ber o f
unequal to equal frames employed in the construction o f the
larger Tuticorin thoni, dictates a more shapely ship.
Thus, the essential requirem ent for the construction and
particular design o f hull o f both the vattai and the thoni, is based
around the shape o f the frames, in particular the central frame
or master fram e. T he shape o f the master frame determines the
shape o f both the equal and unequal frames. This master frame
takes the form o f a wooden m ould or template. A builder
would have a num ber o f such moulds appropriate to different
form s and sizes o f hulls. In the case o f the vattai that were
recorded at Eripurakarai, one equal frame was shaped direct
from the master m ould, and three unequal frame shapes were
derived from the m ould.
The process o f determ ining the shape o f these frames begins
by preparing a 'scrieve board’ for the boats, or a 'm ould loft
flo o r ’ for the ships: thus providing a horizontal surface on
which the shapes can be drawn. A rectangle is scribed or drawn
in ink — its breadth and height being equal to the beam and
depth o f the intended vattai (fig. 10.5 A ) . These proportions
are determ ined by form ulae (as are those o f the thoni), that
relates to the overall length o f the ship.
T he dim ensions o f the m ould are determ ined by and
determ ine the shape o f the master frame. Thus the m ould is
then positioned on the scrieve board (fig. 10.5 B ) and the
curve o f the m ould is marked o ff to give the shape o f one h alf
o f the master frame (fig. 10.5 C ) . The m ould is then flipped
over to the symmetrical position on the other side o f the board.
T he m ould is then marked off, thus creating the full shape o f
the master frame (fig. 10.5 D) (this is equivalent to the cross
section o f the boat amidships). This line equates to the shape o f
the master frame and hence the equal frames.
286 L u c y B lue
(A) (B)
(G) (D)
(E) (F)
(G)
F i g u r e I O . 5 Diagram to illustrate the design o f vattai frames using a scrieve board

and a single m ould (Drawing: E. Kentley)
T h e m ould is then adjusted on the scrieve board to mark o ff

the shape o f the first unequal fram e. A particular form ula is
used to calculate the changing angle: for every inch along the
diagonal (from the last drawn frame) the half-breadth narrows
by 3/4 inch. Thus the rising o f the bow and sterns, is allowed for
on the diagonal, and their narrowing, along the top edge o f the
scrieve board. Each o f these points is marked o ff on the scrieve
board and using the m ould, a line is drawn between them,
form ing part o f the curve (fig. 10.5 E ) . T h e m ould is then
rotated and the rest o f the curve drawn (this is sometimes
drawn by 'eye’) (fig. 10.5 F ), giving h alf the shape o f one pair
o f unequal frames. Again the m ould is flipped over and the
symmetrical shape is marked o ff on to the scrieve board
alongside that o f the equal frame (fig. 10.5 G ) . This line
equates to the shape o f the first pair o f unequal frames.
This procedure is repeated to design the shape o f the second
and third (in the case o f the vattai) unequal frames, rising by an
additional inch each time on the diagonal and narrowing by an
additional 3/4 inch along the top. T h e m ould in each position is
marked o ff on the scrieve board. These four curves are then
used to mark out the tim ber (they function as a sort o f
pattern), and fashion fourteen plus one equal frames, and
three pairs o f unequal frames. For each frame the builder
marks three timbers — one form ing the floor tim ber and the
other two the futtocks. The floor and futtocks overlap in a
complex dovetail jo in t. The overlap extends right through the
turn o f the bilge. Identical dovetail joints were observed in the
fastening o f the Tuticorin thoni, also overlapping at the turn o f
the bilge. O nce the timbers have been marked they are
transported to the sawmill where they are cut into shape.
Thus, the shape o f the equal and unequal frames is derived
from one master frame with the use o f a m ould (in the case o f
the vattai) and a template (in the case o f the T u ticorin thoni),
using the rule o f thum b. In both the building o f the vattai and
the T u ticorin thoni (and for that matter all the other fram e-first
vessels that we observed), the size o f the boat is pre-determ in ed
(usually by the boat owner) and the design o f the vessel is based
on the following parameters:
1) Basic length o f the vessel = L

2) G eneral shape specified by ratios relative to the length, L
(the maximum breadth and depth o f the hull is determ ined
in terms o f ratios relative to L) e.g. thoni length/ beam/
depth ratio is 8:2:1.25*
3) Shape o f the master frame has to be known (this is
determ ined by breadth and depth dim ensions o f the vessel)
4) N um ber o f spacing o f frames has to be known (this is
standard within each vessel)
288 L u c y B lue
5) Total narrowing and rising o f the designed hull has to be

known
T h e specific dim ensions o f the hull shape are then determ ined
by a com bination o f the shape o f the equal frames (established
by the shape o f the master fram e); the shape o f unequal frames
by a simple m odification o f the shape o f the master frame; and
the shape o f the extreme end frames that is determ ined by a
single tem porary shaping batten (known as a ribband) that runs
the length o f the boat.
C o m p a riso n in the D esign o f T a m il N ad u F ra m e-

F irst Vessels and those o f 14 th C en tu ry V enice
Many similar conclusions can be drawn with respect to the

design methods used in the construction o f fram e-first ships in
14th century Venice:
1) the basic use o f the length (L) to determ ine by relative

ratios the maximum breadth and depth o f the hull
2) the use o f a master frame, the shape o f which was conceived
by a wooden template or m ould
3) the adoption o f a total num ber o f designed frames
4) the total rising and narrowing o f the designed hull in
relation to the depth and breadth o f the master frame
5) the use a w ooden tablet and a stick to derive the shape o f the
designed frames from the shape o f the master frame,
instead o f a scrieve board (see Sarsfield 1984; Bellabarba
1993; Barker 1991; Rieth 1996).
6) the designed part o f the hull was usually slightly less than the
keel length. Beyond the last o f the designed frames, towards
the posts, the hull shape changed too rapidly for the frames
to be calculated by the set o f design rules or partison method.
T h e bow and stern frames were therefore fitted later, after
the hull had been partly planked, again in the same way as
they are on the Tamil Nadu fram e-first boats and ships.
Thus, we can see many features o f the Tam il Nadu design

m ethod that correspond with the Venetian method. This
possible transmission o f design across regions has also been

noted in Brazil. Sarsfield (1988) docum ented the use o f a very
similar design m ethod still in use in Brazil. He furtherm ore
noted the use o f the dovetail jo in t between the floor and
futtock, also identified on several Iberian (that is Portuguese,
Spanish or Basque) wrecks o f the 16th century, excavated from
A m erican and British waters.1
It seems possible therefore, that the foreru nner o f the design
system used in Tam il Nadu was taken there and to Brazil by the
Portuguese in the 16th century. The Portuguese first settled in
Tuticorin in 1543 - T he Portuguese Jesuit missionary Francis
Xavier noted in 1542 —3 that the Paravas (the indigenous
fisherm en and pearl fishers o f Tamil Nadu) had open fishing
boats know n as thoni and vallam (Sch u rh am m er, 1977 :
30 6—328). We cannot, o f course, be certain that there is any
continuity between these boats o f the 16th century and those o f
the same name used and built in Tamil Nadu today. However,
the uniqueness o f this design (particularly when compared to
the majority o f traditionally built boats in this region) and the
strong connection between the design m ethod and the use o f
the remarkably intricate dovetail joints, could imply that the
Tamil design o f the 20th century is perhaps a sim plified version
o f the M editerranean m oulding or Atlantic coast design,
docum ented by Bellabarba ( l 993 )> Barker (1988) and Rieth
(1996). Thus, the evidence as it now stands, suggests that the
16th century Portuguese may well have brought a version o f
these European practices to Tamil Nadu. The sim plified hull
forms o f the Tamil thoni, kotia and especially the vattai, no longer
have the need for an intricate design. Based on m ore elaborate
fram e-first design principals, amalgamated with an elem ent o f
'building by eye’ , and with the adoption o f personal experience
and requirem ents, Tamil shipwrights have evolved a vessel
design that suits their purposes. O f course, it is impossible at
this stage to prove that Tamil building tradition does in fact go
back as far as the 16th century: at best we have accounts by
James H orn ell ( l 945 ) to support the use and construction o f
these boats in Tamil waters for the last fifty years or so. U ntil we
have archaeological evidence to allow us to look m ore closely in
290 L u c y B lue
to the history and evolution o f this design, our observations on

fram e-first vessels o f Tam il Nadu must simply rem ain a
hypothesis.
N otes
I Stu dlan d Bay (H u tch in son 1991); Cattewater (R edknap, 1 9 8 4 ); R ed Bay

(G re n ie r, 1988) an d two in the C arib b ean (K ieth , 1 988).
II
Y ami B o a ts and B o a t
B u i l d i n g i n a W lD ER
P ersp ective
R.H. Barnes
D uring a tour o f Taiwan lasting some three weeks in August and

Septem ber 1997 which I made in the company o f D r. David
Faure, we visited Lanyu or O rch id Island (also known as Botel
Tobago) from Septem ber 10 through 12. The two evenings we
were there we had supper in the home o f Mr. Shiyman Feaien
(K uo C h ie n -p ’ing) and his wife in the village o f Iraralay (Lang
Tao). O n the Ilth, he accom panied us on a tour o f the island,
including among items o f interest the nuclear waste disposal
plant against which he had helped lead a dem onstration in
1995. O n that evening, before being shown an hour and a h alf
long video o f the dem onstration and upon request, I showed
slides o f the eastern Indonesian whaling and fishing village o f
Lamalera, Lembata about which I had recently written a book
(Barnes 1996)- In return he showed us slides o f Yam i fishing
ritual and house building. Fortunately, given my lack o f any o f
the relevant languages, Shiyman Feaien has good English.
W haling is not part o f the Yam i fishery, and my hosts could not
understand why Lamalera bothers to take whales. My attempts
to explain did not satisfy them. W hen I m entioned that like the
Yam i, people in Lamalera catch flying fish and said that at
certain seasons a single man in a boat might catch as many as
200 in a day, their response was that in Lanyu fishing with nets
they can catch as many as 2000 in a night. What Lanyu and
291
292 R .H . B a r n es
Lamalera have in com m on, other than speaking languages o f

the M alayo-Polynesian branch o f the A ustronesian family and a
com m on interest in taking protein from the sea, is a shared
tradition o f boat construction.
Boat C o n stru ctio n M ethods
As is well known, H orn ell drew attention to the fact that a

similar method o f boat construction was found in ancient
Scandinavia, the Solom ons and Botel Tobago, as well as in
Indonesian islands to the west o f New G uinea (H ein e-G eld ern
1932: 602, H orn ell 1936, 1946: 188). H orn ell was using
'Indonesian’ as a cultural, rather than political, term, by the
way. His examples were principally in the Moluccas, but there is
no reason why it should not be taken to include the
Philippines, as well as other parts o f what is now the Republic
o f Indonesia. In doing so, he confirm ed the observation by J.
W. Davidson that the boats o f Botel Tobago are 'almost an exact
counterpart o f craft constructed by the Papuans in the
Solom on Islands’ (Davidson 1903)- In fact, as will be seen
below, there is one significant respect in which they differ.
Lanyu boats are well known throughout Taiwan for their
beautiful shape and decorations, and several examples are held
in museums in Taiwan and elsewhere, including Europe.
Davidson’ s description is worth quoting.
It is a built up boat; and, considering the crude tools used

in its construction, is a remarkably creditable affair. The
tribe possess no saws, and consequently each plank is
adzed down, apparently from single trees. These planks,
form ing the sides o f the boat, are so carefully shaped that
they fit closely. Holes are bored near the seams, through
which rattan lashings are passed and drawn tightly,
literally tying the parts together. T he bottom planks are
fastened to the strong V shaped keel in the same m anner
(Davidson 1903: 5^6).
H o rn e ll’ s description o f the com m on construction m ethod is

as follows, 'perforated lug-shaped cleat projections were left in
Yam i Boats an d B oat Bu ild in g in a W id e r P e r spe c tiv e 293
vertical and transverse series at definite intervals on the inner

side o f the hull planking, and to these cleats, shaped frames
were fitted and lashed’ . This m ethod has also been called the
'lashed lug planked’ technique (H orridge 1982; M anguin 1985:
333). By coincidence Edm und Leach’ s earliest articles were
about Botel Tobago, then called by the Japanese K oto-sh o.
A lthough Leach had not yet been trained in anthropology, his
writing showed the same verve which was characteristic o f his
m ore mature work, for example his description o f the island as
a tropical Eden somewhat m odified by the existence o f a
perm anent gale (Leach I 937 a: 4^ 7) • Freshly back from a visit o f
two and a half months to Botel Tobago as a m em ber o f the
Stewart expedition, Edm und Leach challenged the idea o f
cultural diffusion o f construction details, noting the smaller
two-m an boats (called tatara) is unseaworthy in even a m ild
swell, and that it becomes unmanageable in any sort o f wind
(fig. I I . i ). He did acknowledge though that the larger boats
(called chinedkulan1) were m ore stable and had form erly been
used to make the 5 ° mile voyage to the Batans in northern
Philippines, where a mutually intelligible language is spoken
(Leach 1937b: 18 6 -18 7, 1950 : 258 n. 41) (fig. I I . 2). 'It is
F igu re II.I Small tatara (source Leach I 937a: 4 ^9)

294 R .H . B arnes
F i g u r e I I . 2, A ten -m an boat (source K ano and Segawa 1956: frontispiece)
necessary to attack the b elief that the historical diffusion o f

culture can be traced by com paring similar tricks o f technical
design which may arise under different cultural conditions.
Technique may be so violently m odified by the effects o f
cultural and physical environm ent that the search for diffu -
sionist origins becom es not so much invalid as irrelevant’
(Leach 1938a: 9). His attitude was characteristic o f an anti-
diffusionist current in Britain at the time and associated with
the name o f his future teacher Bronislaw Malinowski. That
particular fashion need not concern us here, but it should be
noted that the problem s about independent invention and
diffusion are perennial and thus still with us.
In fact the sea worthiness o f tatara is irrelevant to the
question o f ethnological connection. Leach made the necessary
connection him self when he referred to contact with the
Batans. He observed that boats similar to those o f Lanyu had
been constructed in the Batans (Leach 1937b; 186); he also
noted that the A m i o f Taiwan used the same methods. In fact
the same basic methods were used in a continuous area which
included Lanyu and the whole o f M aritim e South-East Asia.
Yam i B o a ts a n d B o a t B u ild in g in a W id e r P e r s p e c t iv e 295
For a m ore recent consideration o f ethnological contacts with

reference to boat design and confirm ation o f the affinity o f
Solom on Islands boats to those o f eastern Indonesia see
M anguin (1985: 336—340). From archaeological finds and
written sources M anguin is able to conclude that:
T h e 'lashed-lug and stitched-plank’ technique . . . was

used for vessels o f all sizes during the first m illennium
A .D . and the first few centuries o f the second m illen
nium . Progressively, during the first half o f the latter
m illennium it gave way in the larger South-East Asian
trading vessels to structures where frames and planks were
assembled with w ooden dowels only . . . (M anguin 1996:
183).
T he South-East Asian jo n g built using these techniques, as
described in texts, had an average burden o f 400 to 500 metric
tons. Jong is a Malay word for a seagoing ship. T h e Chinese
ju n k is a very different kind o f vessel to which this Malay word
has been applied (Manguin 198O: 266—267). The smallest
figure given is 85 tons and the highest 700* A jo n g built in
Japara, Sumatra in around 1513 might have reached 1,0 0 0 tons
and carried 1,0 0 0 persons. A jo n g o f Pasai, Sumatra, which the
Portuguese under Alfonso de Albuquerque attacked in the early
sixteenth century dwarfed the Portuguese nau. Its four layers o f
planks defeated the Portuguese cannon which could only
penetrate two layers. T h e jo n g surrendered only when the
Portuguese tore its two rudders away after two days and nights o f
fighting. A n account o f a Javanese jo n g by Father Nicolau
Perreira in 1582, comments that these very large ships contained
no iron nails, but were built with wooden dowels inserted into
the seams o f the planks. There is no inform ation as to how the
ribs were fastened to the hull, either with lashing, as in Lanyu
and Lamalera, or with dowels, as also occurs in South-East Asia.
M anguin thinks that lashing would be insufficiently strong for
such large ships. Today ships in Sulawesi and Madura have their
planks pinned to rigid frames with dowels. M odern Bugis ships
have a burden o f 50 to 100 tons with the largest about 200 tons
(Manguin 1980: 253 —^69)-
296 R .H . Barnes
F igu re II.3 Tapping the barok between the seam pegs

(Kano and Segawa 1956: 318)
In order to consider the points o f com parison, it is useful to

summarize the principal features o f the Yam i boat. A ccording
to H o rn ell’ s description o f the Yam i boat in the collection o f
the H am burg Museum fur Volkerkunde,
Th e planks are held together initially by means o f w ooden

dowel-pins inserted into opposed edges . . .; no sign o f
their presence is shown outwardly. As each strake is made
up o f three lengths or sections butted end to e n d , care is
taken to break the joints.
As in the orem ba [o f the M oluccas] the dowel
attachment o f the planking is reinforced by a system o f
Yam i B oats an d B oat Bu ild in g in a W id e r P er spe c t iv e 297
A large boat h alf assembled

F igure I I . 4
(source K ano and Segawa 195^= 3*5)
indirect lashings. W hen dubbing out the planks, lu g

shaped perforated cleats (com b-cleats) are shaped out and
left upstanding at predeterm ined places. In the present
boat they consists o f three sets. T h e first o f these give
attachment to a U -shaped transverse frame, located
amidships. . . . T h e second serve to hold in place a
triangular bulkhead board near one end . . .; the third are
used to tie together the conversing sides at each end
(H ornell 1936: 146) (figs. I I . 3 & I I . 4 ).
A ccording to Leach, there are two types o f boat, the small tatara
for one or two m en and the large chinedkulan for as many as
298 R .H . Ba r n e s
Figu re I I .5 Interior view (source H ornell I 9 3 ^ : Plate L)
Figu re II.6 View from above o f the method o f attaching the rib
(source H ornell 1936: Plate L)
Y am i B oats an d B oat Bu ild in g in a W id er P er spe ctiv e 299
ten oarsm en plus a helmsman. In the latter type, the single

centrally located U -shaped transverse frame (or rib) are
replaced by two such frames placed symmetrically about one-
third o f the way from each end. Leach’ s next statement is to my
m ind somewhat confusing. He says that in this case, 'the rattan
lashings are threaded through two holes in the rib and are then
lashed back onto the comb cleats; there is not a double row o f
comb cleats on each side o f the fram e’ as seen in H o rn ell’ s
photograph o f the rib in the H am burg tatara, 'but only one row
on the inside o f the rib (i.e., on the side nearest the centre o f
the b oat)’ (figs. I I .5 & I I . 6 ). If I understand him , he means
that the ribs o f the chinedkulan are adjacent to the row o f lugs
(as can by seen in the photographs in K ano and Segawa 195^:
320—321) (figs. I I . 7 & I I . 8 ). As for the tatara, to my eye the rib
shown in H o rn ell’ s photograph is not placed between two
parallel rows o f lugs, which seems to be Leach’ s interpretation,
but actually rests on a single row o f lugs which project on either
side o f it, as is the case in Lamalera boats (the photograph on
the left in K ano and Segawa 195^: 357 would seem to confirm
this understanding o f the matter) (fig. I I . 9). Shadows in the
photograph make it impossible to be completely certain o f this
interpretation though. Leach further comments that there are
four strakes on each side o f the keel in the chinedkulan, rather
than the three in the tatara. 'The top strake is not attached in
any way to the U -shaped transverse frames and is held in place
only by dowel pins and the com b-cleat ties at each en d’ (Leach
1937^ : 186). His photograph and fig. 5 >show a series o f thwarts
in the chinedkulan (five in the one in the photograph) (figs.
II. 10 & I I . I I ) . These appear to be dowelled to lugs, but not
lashed to ribs. O nly two would coincide with a rib in any case
(see fig. I I . 7). If this interpretation is correct, it is an
im portant point in the engineering, as will be explained below.
Leach’ s account needs to be m odified in view o f the
inform ation provided by Hsii Y in g -c h o u .2 First o f all, there is
m ore variety to the two kinds o f boats than Leach described.
His inform ation appears to have come from the southern-m ost
village o f Im urud (Im orod) (the village whose fishery is now
threatened by radio active spillage from the atomic waste
300 R .H . Barnes
F i g u r e I I . 7 T h e h u ll is p r o p p e d u p w ith rib s
(s o u r c e K a n o a n d S e g a w a 195 ^= 3ô)
F ig u r e I I . 8 G iv in g th e f in a l to u c h e s to th e h u ll
(s o u r c e K a n o a n d S eg a w a 195 ^: 3 2 l)
Y am i B oats and Bo at Bu ild in g in a W ider P er spectiv e 301
F i g u r e I I .9 P r o je c t in g lu g s in a p a r t ly fin is h e d ta ta ra
(s o u r c e K a n o a n d S ega w a 1 9 5 6 : 357 )
F ig u r e I I .I O I n t e r io r , s h o w in g a r r a n g e m e n t o f o a r s , la te r a l c r o s s - m e m b e r s ,
a n d r ib s ( s o u r c e L e a c h 1937 b : P la te M )
302 R .H . Ba r n e s
JB,
F igu re II.II H alf section (transverse) o f the boat from Botel Tobago
(source Leach 1937b: Fig. 5)
disposal plant), which may or may not have something to do

with his rather simpler picture (Leach 195 ^ : n * 4 1).3 T he
small Likey A Tatala, as they are called there, are one man
Pikatangiyan, two m en Pikavangan and three m en P inonog-
nogan. A ll three may be used for nighttime fishing, but only
the first two for daytime fishing. T h e large Chinedkeran are
built in sizes large enough to accommodate crews o f six, eight,
or ten m en. Two further issues appear from H sii’ s account.
O n e is that both the Likey A Tatala and the Chinedkeran may
be built up with planking either from a broad dugout keel or
from a narrow keel. T h ere is nothing surprising from a
comparative perspective in these alternatives. As it happens they
correspond to a stage in H o rn e ll’ s evolutionary scheme o f
developm ent o f planked boats from dugouts. ’T h e final stage in
the conversion o f the dugout into a fully plank-built boat is
attained when the dugout underbody is reduced to a keel-like
axial beam, with sides raised upon its edges by num erous strakes
o f sewn on planking’ (H ornell 1946: 19^)* Boats in which the
planks are placed edge-to-edge and literally sewn together are
Y am i B oats and B o at Bu ild in g in a W ider P er spective 303
characteristic o f the methods o f Persian and Arab shipwrights

down to the sixteenth century and o f the Lamu Archipelago o f
East A frica. They were described by various earlier writers,
including Marco Polo.
W hen all is complete, rib frames are inserted athwart the

hull and held down in place by lashings which pass
through holes in the planking. Thereafter, pegs are driven
from the inner side into the stitch holes and broken o ff
short. W herever the twine shows on the outside this is cut
away flush w ith the surface. . . . (H o rn e ll 1946:
234 - 235 )-
O f course the matter o f sewn planking marks a difference from
the techniques o f Lanyu, but the point still obtains. T h e planks
o f Yam i boats, like those o f Lamalera, are jo in ed edge to edge
by dowels, but the edges are not sewn.
Th e second point is that the boats may be built o f a single or
m ultiple layers o f planking. T he statement is not to be
confused with the m ultiple layers o f sheathing form erly
frequently found in South-East Asian ships (M anguin 198O:
27o ). Instead, as can be seen in the diagram by K ano and
Segawa (see fig. II . 1 2 ), what is intended is that in some boats
there are as many as four strakes lying above the keel, each
consisting o f a central plank and another plank on either end.
In Lamalera, where the planks have a somewhat similar pattern,
the plank names apply essentially to the central planks, while
the end planks are spoken o f as extensions. O f the dugout
based boats on Lanyu, the Likey a Tatala may have a single layer
o f planking, that is to say a single strake made o f a middle plank
and two end planks, thus three planks on each side for a total o f
six, or a double layer o f planking (two strakes o f three planks
each) for a total o f twelve planks. The Chinedkeran built in this
way have three layers o f planks with three planks each on each
side making for a total o f eighteen planks. Small boats built on
a true keel have three layers o f planking, with three planks each
on each side, for a total o f 18 planks. Large boats on a keel have
four layers o f planking, with three planks on each side, for a
total o f tw enty-four planks. The boats range in length from
304 R.H. Barnes
Pak.
Pak.
Pam.
Kab. pata.
Kab. pata.
Pab. pata.
Kab. pab.
Kab. pab.
Pab.
\Ip. y Kab. pato.
Kab. pato.
Pab. pato.
Rap.
F i g u r e I I . 1 2 A r r a n g e m e n t s o f th e p la n k s ( s o u r c e K a n o a n d S e g a w a 1 9 5 6 : 3 0 6 )
2-3 m for a single passenger boat to 7-6 m for boats holding a

crew o f ten (Hsii 1982: J — 8 ). K ano and Segawa (1956: 282)
indicate that the large boats measure about J metres, while the
small boats are about 3 metres long.
A ll the boats share a similar outline, with a characteristic

vertical extension form ed by the prow and stern strakes
MOW AS, which sweep upwards above the head-hight o f
m en seated in the boat. The planks are fastened together
with w ooden pegs PA SEK A K A Y O (Morus australis, Poir.
[mulberry]), and the interstices are bunged with V A R O K
fibre (Zanthoxylum integrifolium M errill [prickly ash])
(Hsii 1982: 8).
Pasek a kayo has no relation to the Lamalera word for wooden

dowels, ketilo, but compares directly with the Indonesian
equivalent pasak kayu.
T he im pression to be drawn from these accounts is o f a fairly
flexible approach to boat construction, which uses a basic set o f
techniques and styles to produce a variety o f boats suited for
specific purposes. These features include edge to edge plank
ing, internal dowels jo in in g planks, protruding lugs to which
are lashed U -shaped ribs. The thwarts provide, according to
Leach (1937b: 186), lateral thrust to the sides o f the boat. A ll
Y am i B oats and B o at Bu il d in g in a W id er P er spective 305
o f these features are found in Lamalera boats, but what is

lacking is a further feature providing compression to the
structure.
Lamalera builds, or built, three types o f boat. T h e first, no
longer built, are dugouts (bero) with double outriggers and
with the sides built up by p alm -leaf splash screens. T h e second,
a newly popular type (sapa), are two man planked boats without
outriggers built up from a keel using the same techniques as the
third type. This third type is the traditional boat (tena) used for
hunting large game, such as sperm whale, porpoise, manta ray,
whale sharks, sea turtles, etc. This boat is ten metres long and
accommodates fourteen men. There are six strakes on each
side, although most consist in two or three pieces o f wood
jo in ed at the ends by a lock scarf. Each plank overlaps those
below, thus providing further structural strength (see fig.
II. 13). T he internal dowels restrict the movement o f the
planks, which are further secured by ten ribs lashed to lugs left
in the hull in the same m anner as on Lanyu, although o f course
there are many m ore o f them. Unlike the Lanyu boats, the two
ends are not the same. The bow has a raised bowsprit and a
harpooning platform extending a metre beyond the bow. The
stern is finished o ff with a decorated stern piece.
A n additional feature o f the structure apparently lacking in
the Yam i boats is transverse thwarts lashed to the lugs and ribs,
which squeeze the hull together, further fixing the planks in
place (see fig. I I . 14). A ccording to H orridge (198^ : 59), the
1. ara belikang
2. ara kenati
3. ara tuka
4. nulu futu
5. ara bela
6. nefi
F i g u r e I I . 1 3 T h e p la n k p a tt e r n o f L a m a le r a b o a ts (s o u r c e B a r n e s 1 9 9 6 b : 2 I l)
306 R.H. Ba rn es
Figure I I . 14 I n t e r n a l c o m p r e s s io n s tr u c tu r e o f th e h u ll o f a L a m a le r a b o a t
(s o u r c e R . H . B a r n e s 1 9 9 6 b : 2 0 5 )
compression o f the edges o f the planks against each other

increases the friction, so that the whole hull becomes a rigid
pre-stressed structure. H orn ell ( 1 9 2 O : 5 9 ) writes that this
m ethod, as used in Ternate, produces a hull possessing great
elasticity and which 'stands bum ping in the surf in a way that no
m etal-fastened boat would long survive’ . N ot only do Lamalera
boats have to be brought ashore daily through a surf, which can
be quite rough, they have to withstand hard treatment by sperm
whales. H o rridge’ s point about the pre-stressed structure was
graphically dem onstrated one day when the Sili Tena returned
to shore after having been struck in the bow by the tail o f a
whale. T he force carried through the hull and broke a plank
diagonally opposite in the stern. Except for repairs required to
a plank in the bow and one in the stern, the boat was not
otherwise damaged (Barnes 199^: I 7I—173 > 201 —219 )-
To return to the question o f cultural influences, G ib son -H ill
(1952) describes boat planks found at Pontian, Pahang, Malaysia
and subsequently radiocarbon dated to 293 ± 6 0 AD A ccording
to M anguin (1985= 333 ), this shipwreck is actually the co
location o f three techniques seldom found together on the same
Y am i B oats and B o at Bu ild in g in a W ider P er spective 307
craft: sewing o f the planks, edge-dowelling and frames lashed to

carved-out lugs’ . G ibson-H ill (195^: 113—114) speculates that
this boat would have had thwarts, thus opening up the possibility
that it also had the compression feature o f the Lamalera tena
and other eastern Indonesia boats. M anguin comments that the
Pontian boat is the earliest and best example o f an intermediate
construction design representing the transition between purely
sewn-plank boats comparable to some Cam/Vietnamese craft
and the fully edge-dowelled, lashed-lug boats described above.
A nother early wreck in Butuan, M indanao, Philippines, which
has been dated to 3 2 O ± IIO AD is described by M anguin as the
earliest example o f a fully fledged lashed-lug design (Manguin
1985= 333 - 334 )- It is worth remarking in passing that the
Solom on Islands boats (mon) lack edge to edge dowelling, but
instead the edges are sewn together (Woodford 1909: 5 ° 9 >Ivens
19^7: 150, H ornell 1946: 208, Manguin 19^5 * 337 )- O n an
imaginary linear evolution line, the m on would stand som e
where before the edge-dowelled and sewn Pontian boat, which
would be followed in turn by the fully edge-dowelled Butuan
wreck’ (see below; Manguin 1985: 337 )-
It would seem appropriate therefore to regard Yam i boat
b u ild in g skills as part o f a widespread tradition where
borrow ing and innovation were always part o f the picture,
producing a great range o f boats o f varying sizes and specialized
uses. This point is reinforced by discoveries o f seven shipwrecks
ranging from South-East Asia to K orea, including two on the
southern Chinese coast, dated between the thirteenth and the
seventeenth century. These ships are all hybrids between what
had previously been considered quite different Chinese and
South-East Asian traditions (M anguin 1984).
K ano and Segawa (1956: 299 ) provide photographs o f
tree felling and shaping o f logs to become planks (see fig. I I . 15).
I have participated in parties which have travelled from
Lamalera to fell and shape trees in the forest for use in
constructing boats. K ano and Segawa say that this tree cutting is
done years in advance o f building the boat, as it should be, but
is not always, done in Lamalera. Yam i choose an auspicious day
for cutting the trees, because the future prosperity o f a boat is
308 R .H . B a r n e s
F i g u r e I I . 1 5 R o u g h s h a p in g a tr e e (s o u r c e K a n o a n d S e ga w a 1 9 5 6 : 2 9 8 )
influenced from the m om ent o f felling the tim ber. As in

Lamalera, the Yam i use axes to cut the trees, rather than saws. I
am afraid that I have produced large blood blisters on my hands
doing this sort o f work. As in Lamalera, the Yam i then roughly
shape the timbers into planks with these axes, although
Lamalera would also use adzes and field knives at this stage.
Th e planks must be cut from a live tree among the Yam i, as
there is a prohibition on using dead wood. 'The process is in
itself one o f sculpturing, a single large piece com ing from a
single standing tree’ (Kano and Segawa 1956: 298—299). In
Lamalera, each section o f the keel is made from a separate log.
However, generally two planks are taken from a tree. T h e trees
are felled, branches lopped off, and then a groove is chopped
lengthwise through the trunk with axes or large adzes. Wedges
are worked into the groove, eventually driving the halves apart.
Planks are shaped roughly, but leaving them nearly twice as thick
as the eventual plank. T h e center o f the trunk becomes the
outer hull, and they adze down the curved outer trunk, leaving
in place the necessary blanks for the eventual lugs (Barnes 1996:
Y am i B oats and B o at Bu ild in g in a W id e r P er spectiv e 309
F i g u r e I I . 16 S h a p in g th e p r o w ( s o u r c e K a n o a n d S e ga w a 1 9 5 6 : 3 0 1 )
213—Î^)- D uring such a session, the participants on Lanyu are

provided with periodic meals o f boiled m illet, fruit and other
foods (Kano and Segawa 195^: 302). In Lamalera, the meal is
o f rice, maize and pig and fish; and in Lamalera too there are
several food breaks during the day. H ornell (1936: 146) says
that each h alf end-piece is cut from the solid. Leach ( l 937 b :
186) confirm s this statement and adds that 'it is always cut from
the base o f a tree so that the wood grain may follow the curve o f
the prow’ . What is meant by these statements is best understood
by looking at the photographs and reading the explanations
given by K ano and Segawa (1956: 30 I—3 ° 3 > see fig- H .16).
These photographs show large buttress roots at the bases o f
tropical trees and Yam i shaping the fallen trees, taking
advantage o f these roots to produce the great up-sweeping stem
and stern pieces which rise more than a metre above the sea.
O n Lanyu tim ber should be felled, worked into rough shape
and carried back to the village on the same day. T h e bow and
3io R.H. Barnes
stern pieces require some twenty m en taking turns to carry

them across the island. Lamalera similarly brings the wood back
in a single trip, although they do so using their boats. T h e
Yam i do not now use even their large boats for transportation,
but the people o f Lamalera do. W hen the Yam i arrive with their
wood at the village, they shout exultantly, and the wom en
welcome them with cooked taro, fish and fruit. In Lamalera the
return to the village is less marked and there is no ceremony.
Th e Yam i finish the wood in a special w ork-shed using an axe,
chisel and gouge (see the photographs in K ano and Segawa
1956: 30 6—311, see fig. 11.17)* In Lamalera, this work is done
in the boat shed o f the boat being worked on. They use a variety
o f adzes, chisels, gouges, borers and m odern tools, such as a
brace and bit, to prepare the pieces for assembly. A description
o f this work with appropriate photographs can be found in
Barnes (1985, 1996: 2 0 l- 2 ig ) .
In both com m unities the keel is laid first, and then the
planking on either side gradually built up through a process o f
fitting, correctin g and refittin g u n til the planks match
perfectly. In Lamalera, this process involves placing short logs
across the top o f the plank being fitted and wrapping lashing
F i g u r e I I . 1 7 W o r k - s h e d u s e d in b u i l d i n g a la r g e b o a t
( s o u r c e K a n o a n d S eg a w a 1 9 5 6 : 3 0 6 )
Y am i B oats and Bo at Bu ild in g in a W ider P er spectiv e 311
made from palm branch skin around one end o f the logs,
passing them under the boat and wrapping them around the
other end o f the logs. W hen this lashing is twisted, the planks
are squeezed together with such force that the soft, tem porary
dowels holding them together may be crushed (Barnes 1996:
215, fig. 28). This technique is not described for the Yam i, but
m uch else in the photographs in K ano and Segawa (1956:
313—3 J5) would be fam iliar to the boat builders o f Lamalera
(compare the photographs in Barnes 1985: 352, 354, see fig.
I I . 18). Yam i boats are caulked with the yellowish-white cotton
like fibre from the root o f the barok (Zanthoxylum integri-

folium , M errill), a plant which is restricted to Lanyu and the
Philippines. Like most South-East Asian boats, those o f
Lamalera are caulked with the dusty black fibre from the
trunks and branches o f the sugar or areng palm (Arenga
pinnata, M errill), called in Lamalera rapo and in Indonesian
ijuk (Kano and Segawa 1956: 316—319, Barnes 1996: 216). In
Lamalera and on Lanyu the ribs are placed in the hull only
after the hull has been constructed, and o f course decorating
the hull comes last (Kano and Segawa 1956: 321—323, Barnes
F i g u r e I I . 1 8 T h e p a ta k a r a n b e in g j o i n e d ( s o u r c e K a n o a n d S e ga w a 1 9 5 6 : 3 1 3 )
312 R .H . Bar n e s
1996: 173, 2 l7 ~ 2 l8 ). Large boats require over two months o f

work to construct in Lanyu, a few weeks in Lamalera.
It has been com m on to disparage the tools available to the
Yam i, as in the quote from Davidson given above. H ein e-
G eldern (1932: 6 o i—603) had argued that the plank boats o f
Botel Tobago, the Moluccas and the Solom on Islands belonged
to a tradition derived from a culture associated with a smoothly-
ground celt with oval or cylindrical cross section which has been
found in Indonesia, O ceania and Japan. To this Leach
responded that the clumsy tools described by H ein e-G eld ern
would be insufficient to make boats like those o f the Yam i. 'The
use o f iron tools is definitely essential for the constructions o f
the boat in its present form and it seems significant that,
according to Yam i legend, their ancestors acquired the art o f
forging iron before they learnt to build a boat’ (Leach 1938b:
162—163). H orridge (1982: i) also argued that, 'the dowelling
technique required an iron tool and replaced sewing which did
n o t’ . Manguin, however, offers a variety o f counter arguments
to the effect that holes suitable for dowels can be produced with
stone tipped drills, and that such were still used in the
Solom ons in the twentieth century (1985: 339 —3 4 °)• Ivens
(19 2 7: 150) comments that in the Solom ons holes were
punched using pieces o f flint or a sharp pointed shell.
It is difficult to understand how the natives were able,

before they became acquainted with iron tools, to adze
down the canoe planks to the requisite degree o f thinness
and shape them with the aid only o f stone implements,
but it appears from the descriptions [by various early
explorers] above quoted that they undoubtedly did so, and
that the canoes have changed but little in type since the
days o f Mendana (W oodford 1909: 5 ° 8 )-
C on clu sion s
Lanyu is evidently much changed from the days when Leach and
Hsii first visited to judge by what we were shown on our visit
and by the excellent photographic record o f earlier times
Y am i B oats and B oat Bu ild in g in a W id er P er spective 313
available in the works which I have cited. A ny new account

would have to take these factors into consideration, as I have in
my book on Lamalera. Lanyu and Lamalera differ in the scale
and types o f their fishery, and their boats differ in types, size
and purpose. In both comm unities boats and fishing tech ni
ques are carefully adapted to the specific conditions in which
the fisherm en must carry out their work. In this respect we can
speak uncontroversially o f rational econom ic adaptation to the
environm ent. From conversations with friends in Lamalera, I
can confirm that they discuss and debate the relative merits o f
techniques in boat building and fishing, that they m odify these
techniques as opportunities arise, and that they are quite
conscious o f the effects o f any changes introduced. I cannot
confirm ed that the same is true on Lanyu from personal
experience, but I am sure that it is.
To a great extent, the nature o f the occupation will mean
that there will be similarities in the organization o f the work in
fishing com m unities around the w orld, even those which have
no, even distant in time, ethnological con n ection .4 N ever
theless, Leach’ s scepticism as to any ethnological connection
between the methods in which boats are built in Lanyu and
those used in M aritim e South-East Asia and western O ceania is
misplaced. T he basic technique, including internal dowelling,
lashed lugs, and compression was described for the Moluccas in
a Portuguese docum ent (probably authored by A nton io Galvao)
around 1544 (Jacobs I 97 I: *57 ) and in greater detail in Spanish
for the Philippines in 1668 by Alcisco Alcina, SJ (Scott 1981;
H orridge 1982: 6—18). We would probably be justified in
speaking o f a com m on maritime culture, including boat
building and fishing methods, calendar and ritual cycle
throughout the island world where Austronesian languages
are spoken, in which the details vary, but where the basic
themes are usually the same, except in O ceania where certain
themes have been lost and innovations have been introdu ced .5
At any rate, this survey has shown, I think, a considerable
similarity o f approach between Lanyu and Lamalera not at first
evident to the eye, despite the great and obvious differences
between them.
3i 4 R .H . B a r n e s
Notes
1 Such is L each ’ s spelling. H sii spells it chinedkeran, while K an o and
Segawa have chin urik uran. I am not in the p ositio n to decide which is
correct o r w hether the differences reflect local dialectical differen ces or
the language b ack groun d o f the various auth ors. T h e re fo re , I spell the
word in every case as does the au th or I am currently re fe rrin g to.
2 I note the b ib liograp h ical references to works on Y am i b oat con struction
by H sii, K o k u b u N aoich i and K an o T adao, which are inaccessible to me
both physically and because they are in Ja p a n ese , a language which I do
not read (H sii 1982: 155)*
3 T h e fron tisp iece in H sii (1982) shows Im o ro d as it w ould have looked
about the tim e Leach visited it.
4 See fo r exam ple W eibrot (1958) fo r a com p arison o f Portuguese and
N orw egian fish eries.
5 Bellw ood ( l 995 : 105) com m en ts: In o rd e r to settle the W allacean islands
(P hilippin es, Sulawesi, L esser Sundas, M oluccas), as also A ustralia and
New G uin ea, the o rigin al Pleistocene colonists m ust have had som e
degree o f seafarin g capacity, even if rudim entary, by at least 4 ° > 0 0 0
years ago. D id the A u stro n esian s learn a n u m b er o f seafarin g skills from
them . . . R ather little seafarin g term in ology is recon structible fo r P roto-
A u stro n esian . While it w ould be unwise to deny P roto-A u stron esian s the
knowledge at least o f canoes, it is w orth rem em b erin g that m uch o f the
early expan sion o f the A u stro n esian -sp eak in g peop les was through
Wallacea, especially the Ph ilippin es and Sulawesi with their m an ifold
satellite islands. It is am ongst the m ore watery topography o f Late
Pleistocene Wallacea, rather than lan d -b rid g e d S u n d alan d , that one
m ight expect pre-A u stro n esian m aritim e tradition s to have flou rish ed
and to have b een transm itted to later arrivals.
B ibliography
A brah am , M . (1988) Two Medieval Merchant Guilds of South India, New D elhi:
M anohar Publishers.
A gius, D .A . (1996) Siculo Arabic, L o n d o n : K egan Paul Interan tional.
---- ( l 997) H istorical-lin gu istic reliability o f M uqaddasi’ s in fo rm a tio n on
types o f ships, in Across the Mediterranean Frontiers: Trade, Politics and Religion,
6 5 0 —1 4 5 ° ’ eds. D .A . A gius & I.R . N etton (T u rn h o ut: B repols): 3 0 3 —29.
---- (1999) M edieval Q alhat: Travellers, Dhows and Stone A n ch ors, in
Archaeology of Seafaring: The Indian Ocean in the Ancient Period, ed. H . Prabha Ray
(D elhi: Indian C o u n c il o f H istorical Research): 1 73—220 .
Agrawala, V .S . ( l 974 ) Brhatkathaslokasangraha, Varanasi.
A h rw eiler, H . (1966) Byzance et la men La Marine de guerre, la politique et les institutions
maritimes de byzance, Paris: Presses U niversitaires.
A l-A z d i (190 2) Hikayat Abi l-Qasim al-Baghdadi, ed. A . M ez, H eidelberg: C arl
W inter.
A lb u q u erq u e, L. de (1988) Instruments of Navigation, Lisbon: N ational B oard fo r
the C e leb ra tio n o f the Portuguese D iscoveries.
A l-Istak h ri (1967) Kitab masalik al-mamalik, ed. M. J. de G oeje (Biblioth eca
G eo gra p h o ru m A ra b ico ru m , I), Leiden: E. J. B rill (first editio n 18 70 ).
A l-Jaw aliqi (1361/1942) Kitab al-mu'arrab min al-kalam al-ajam i ra/d huruf al-mujam,
ed. A . M . Shakir, T eh ran: D ar a l-K u tu b .
A l-K h a fa ji (1282/1865) Shifa al-ghalil fi ma fi kalam al-'arab min al-dakhil, C airo .
A l-M a q riz i ( l 270//l853) Al-Mawa'iz wa l-i'tibar fi dhikr al-khitat wa l-athar, V olum es
I —II, C airo : B ulaq.
---- (1911—2 4) Al-Mawaiz wa l-i'tibar f i dhikr al-khitat wa l-athar, ed. M. G. Wiet,
V olum es I-IV , C airo : Institut Franâis d A rch eologie O rien ta le .
---- (x957—64) Kitab al-suluk li marifat duwal al-muluk, ed. M. M . Ziyada, V olum es
I —II , C airo .
315
3 i6 B ib l io g r a p h y
---- ( 1 9 7 0 -7 3 ) Kitab al-suluk li ma'rifat duwal al-muluk, ed. S. 'A. 'Ashur, V olum es
III-IV , C airo: D ar al-K u tayyib .
A l-M araku sh i (1949) AI-Mujib fi talkhis akhbar al-maghrib, eds. S. a l-'A rya n & M.
'A a l-'A la m l, C airo .
Al-M as'udi (1861—77) Muruj al-dhahab wa maadin al-jawhar (Les prairies d’or), ed. &
trans. C . Barbier de Meynard and P. de Courteille, Volumes I-IX , Paris:
Imprimerie Imperiale.
A l-M u q a d d a si (190 6) Kitab ahsan al-taqasim fi ma'rifat al-aqalim, ed. M . J. de G oeje
(Biblioth eca G eo gra p h o ru m A ra b ico ru m , III), Leiden : E. J. B rill (first
e d itio n l 8 j j ) .
A l-N akhili, D . ( l 974 ) Al-Sufun al-islamiyya 'aid huruf al-mujam, Cairo: A l-Ahram
al-Tijariyya.
a l-Q u 'a y ti, G halib b. 'Awadh (c. 1996) Hadhrami migration to the Hejaz through the
ages: a general survey, T ypescript.
A l-R am h u rm u zi, B uzurg b. Shahriyar (188 3—86) Kitab Ajaib al-Hind (Le Livre des
merveilles de I’lnde), ed. P.A . van der L ith and trans. L .M . Devic, L eid en : E.J.
B rill.
a l-R ash id , Zam il M. (1981) Su'udi relations with eastern Arabia and 'Uman, 1 8 0 0 —1871,
L o n d o n : Luzac.
A lsagoff, Syed M ohsen (c.1963 ) The Alsagojffamily in Malaysia, A.H. 124 ° (a d 1824)
to A.H. 1382 (AD 1 962) with biographical and contemporary sketches of some members of the
Alsagojf family, Singapore: privately p rin ted .
a l-S h a ’afi, M uham m ad S. (1985) The foreign trade ofjuddah during the Ottoman period,
1 8 4 0 —1916, Jiddah: S ha’afi.
A l-S h a h ri, 'A .A .'A .M . (1994) Ktyfi2 ibtadayna wa kayfa irtaqayna bil-hadara 1-insaniyya
min shibh al-jazjra l-'arabiyya \ufar: Kitabatuha wa nuqushuha 1-qadima, D ubai: A l-
G hurayr.
A l-T a b a ri (1965^ Ta'rikh al-rusul wa 1-muluk, eds. M,J. de G oeje et al., V o lum es I-
X V , B eirut: Khayats.
Alwis, N .A .W . A rth u r. (1936) T h e Last o f the Sailing Ships in The Mahinda
College Magazine. V I . M ahinda C o lle ge , G alle. pp. 138—139.
A n to n ia d is-B ib ico u , H . (1966) Etudes d ’histoire maritime de byzance apropos du 'Theme
des Caravisiens’, Paris: Ecole Pratique des Hautes Etudes.
A p p ad urai, A . (1936) Economic Conditions in Southern India lOOO—lÔO, M adras:
U niversity o f Madras.
A ram co W orld , The Middle East and the Age of Discovery, V o l. 433 -
A rasaratnam , S. (1994) S hipp in g and Seafaring, Maritime India in the Seventeenth
Centuiy, New D elhi: O x fo rd U niversity Press, pp. 2 4 6 —264 .
---- (1994) Maritime India in the Seventeenth Centuiy, O x fo rd U niversity Press.
A ru n achalam , B. (1987) 'T h e Haven F in din g A rt in Indian Navigational
T ra d itio n and C arto grap h y’ , in Satish C h an dra ed., The Indian Ocean:
Explorations in Histoy, Commerce and Politics, D elhi: Sage: 191—221.
A ru n achalam , B. (1996) T rad itio n al Sea and Sky W isdom o f Ind ian Seam en,
Ray, H .P . and Salles, J .- F . eds., Tradition and Archaeology: Early Maritime Contacts in
the Indian Ocean, New D elhi: M an ohar Publishers, p p .261—82.
B i b l io g r a p h y
317
A u n g T u n T h e t (1989) Burmese entrepreneurship: creative response in the colonial economy,
Stuttgart: Franz Steiner.
D avidson, Jam es W . 190 3. The Island of Formosa: Past and Present. L o n d o n :
M acm illan.
Bakari, M toro b in M winyi (1981) The customs of the Swahili people ed. by J .W .T .
A lle n . L o n d o n : U niversity o f C a lifo rn ia Press.
Banaji, D .R . (1933) Slavery in British India, Bombay: D .B . Taraporevala Sons &. C o .
Bandaranayake, S. (ed) (1984) Sigiriya Project, First Archaeological Excavations and
Research Report (Jan.-Sep. 1982), C o lo m b o .
B anerjee, R uchira (1991) M appilla m ercantile network o f M alabar in the
eighteenth century, in A n iru d h a G upta, ed., Minorities on India’s West coast:
history and society, pp. 151—201, D elhi: K alin ga Publications.
Barbosa, D uarte (1918. 1921) The Book of Duarte Barbosa, an Account of the Countries
Bordering on the Indian Ocean and Their Inhabitants, written by Duarte Barbosa and
Completed about the year 1518, trans. and ed. M ansel Lon gw orth Dam es,
V olum es I—II, L o n d o n : Hakluyt Society.
Barker, R. (1988) M any may peruse us: ribbands, m oulds and m odels in
dockyards. Revista da Universidade de Coimbra 34 : 539~~59 -
---- (1989) Perspectives on the 15th century ship, Congress Internacional Bartolomeu
Dias E.A. Sua Epoca, 2, Porto, pp. 20 1—222.
---- (1991) D esign in the dockyard c. 160 0 . Pp. 66—69, in R. R einders & K .
Paul (eds.) Carvel Construction Techniques. O xbow M on o. 12- O x fo rd .
---- (1992) Shipshape fo r D iscoveries, and R eturn , The Mariner's Mirror, 78, 4>
PP- 433 - 447
---- (1991) C areen in g: A rt and A n ecd o te, Mare Liberum, 2, pp. 177—207 -
Barnes, R .H . 1974- *985 ' W haling vessels o f Indonesia. In Sean M cG rail and
E ric K en tley, eds., Sewn Plank Boats: Archaeological and Ethnographic Papers based on
those Presented to a Conference at Greenwich in November, 1984. (N ational M aritim e
M useum , G reenw ich, A rchaeological Series N o. io ) O x fo rd : B .A .R .
---- (1996) Sea Hunters of Indonesia: Fishers and Weavers of Lamalera, O x fo rd :
C la re n d o n Press.
Barrett-Jones, A .M . (1984) Early Tenth Century Java from the Inscriptions, D ordrecht:
Foris Publications.
Basham , A .L . (1967) Studies in Indian History and Culture, Calcutta: Sam bodhi
Publications.
Bastiaanse, J .H . de Boudyck (1845) Vcyage fait dans les Moluques, a la Nouvelle Guinee
et a Celebes, Paris: A . B ertrand.
Beckerleg, Susan (19 9 1) Le reseau d un C on stru cteu r de Bateaux’ in Eds D.
Parkin and F. Le G u e n n e c-C o p p e n s Les Swahili entre Afrique et Arabie Paris:
Kathala.
---- (1994) Watamu: lost la n d , but a new Swahili town’ in J. Maw and D.
Parkin Eds Swahili Language and Society Beitrage zur A frikanistik 23, Vienna:
Institut fo r A frikan stik u nd A egyptologie der Universtat W ein.
---- ( l 995) Brow n Sugar or Friday Prayers: Y o u th C hoices and C om m u n ity
building in Coastal K en ya’ African Affairs 94, 2 3 —38.
318 B i b l io g r a p h y
Begley, V . (1967) A rch aeolo gical explo ratio n in n o rth e rn C eylon , Bulletin of the
University Museum of the University of Pennsylvania, IX, 4 , Sum m er 1967, pp. 21—2 9
---- (1970 ) Excavations at P om p arip pu (1970 ): Part I, Ancient Ceylon, May
1970 , pp. 53- 95
---- (1983) A rik am ed u recon sid ered , American Journal of Archaeology LX X X V II,
1983, pp. 4 6 1 - 4 8 1 .
---- (I 99 1) C eram ic evidence fo r p re-P erip lu s trade on the Ind ian coasts, Rome
and India: the ancient sea trade.Madison-. U nivesity o f W iscon sin Press,
P P - 157- 196.
Begley, V im ala and de Pum a, R ichard D an iel eds. (1992) Rome and India, the
Ancient Sea Trade, D elhi: O x fo rd U niversity Press (Indian R eprin t).
B elcher, W .R . (1998) Fish Exploitation of the Indus Valley Tradition: an Ethno-archaeological
Approach to the Study of Fish Remains, D octo ral D issertation, U niversity o f
W isconsin.
---- ( l 999) T h e E th n o-arch aeology o f Fishing in a Baluch V illage, Ray, H .P .
ed. Archaeology of Seafaring: The Indian Ocean in the Ancient Period, New D elhi: Indian
C o u n c il o f H istorical Research, pp. 22—51-
B ellabarba, S. ( l 993) A n cie n t m eth od o f design ing hulls. Mariners Mirror 79 :
274 - 2 9 2 .
B ellec, Francois. (1998) Early Pilots in the M uslim Waters. Proceedings of the
Indian Ocean Week, 1997 G raem e H en d erso n , ed. W estern A ustralian M aritim e
M useum , Frem antle, W estern A ustralia, pp. 5—1°-
B ellw ood , Peter 1995- A ustron esian Prehistory in Southeast Asia: H o m e lan d ,
Expansion and T ra n sfo rm a tio n . In Peter B ellw ood , James J. Fox & D arrell
T ryon , eds., The Austronesian: Historical & Comparative Perspectives. C an b erra,
A ustralia: T h e A ustralian N ational U niversity.
B en jam in, N ., (1976) A rab m erchants o f Bom bay and Surat, c. 18 0 0 —18 40 ,
The Indian Economic and Social History Review, 13, I, pp. 85—95 -
B ent, T h e o d o re and M rs. (190 0 ) Southern Arabia, L o n d o n : Sm ith, E lder & C o .
Berg, L .W .C . van den (1886) Le Hadramout et les colonies arabes dans Varchipel indien.
Batavia: Im p rim erie du G ou vern em en t.
B ern a rd , P. (1973) Fouilles d ’A'i Khanoum IV. Les monnaies hors tresors. Questions d ’histoire
greco-bactrienne (MDAFA 27)- Paris: K lin cksieck.
Bhacker, M . Reda (1992) Trade and empire in Muscat and Zanzibar, roots of British
domination, L o n d o n : R outledge.
Bhattacharya, B. (1995) T h e D utch East India C om p an y and the Trade o f the
C hu lias in the Bay o f Bengal in the late eighteenth century, M athew K .S .
ed., Mariners, Merchants and Oceans: Studies in Maritime History, New D elhi:
M anohar, pp. 347_ 361.
Blake, G eo rge (1956) B. I. centenary, 18^6—ig ^ 6 , L o n d o n : C o llin s.
B loch, J. (1950) Les inscriptions d'Asoka, Paris: Les Belles Lettres.
B lue, L. ( l 997) Boats o f South Asia, The International Journal of Nautical Archaeology,
2 6 .4 , pp. 339- 343 -
B lue, L ., K en tley, E. & M cG rail, S. (1998) T h e vattai fish ing boat and related
fram e-first vessels o f T am il N adu. South Asian Studies 14: 4 I_ 74 -
B i b l io g r a p h y
319
Blue, L .K ., K en tley, E ., M cG rail, S. and M ishra, U . ( l 997) T h e patia fish ing
boat o f O rissa: A Case Study in E thnoarchaeology, South Asian Studies, 13,
p p . 2 I 2- 23 I* PP - 1 8 9 -2 0 7 .
---- (1997) T h e patia fishing boat o f O rissa: a case study in ethnoarchaeology.
South Asian Studies 13: 212—231.
Bopearachchi O . (1990) Som e O bservations on R om an G oin s fo u n d in
recent Excavations at Sigiriya, Ancient Cey/on. no 8, pp. 20 —37
---- (1991) Monnaies greco-bactriennes et indo-grecques. Catalogue raisonne, B ibliothequ e
N ationale, Paris.
---- (1992) Le com m erce m aritim e entre R om e et Sri Lanka d ’apres les
donnees num ism atiques, Revue des etudes anciennes, pp. 10 7—121.
---- ( l 993 A ) La circulation des m onnaies d ’origin e etrang'ere dans l ’antique
Sri Lanka, Res Orientales.V, p p .63—87.
---- ( l 993 B) Les jard in s de Sigiriya au Sri Lanka, Asies II, Amenager I’Espace, Sous
la d riectio n de F. B lan ch ard, I 993 > PP- 239 “ 261.
---- ( l 995 A ) R ecent discoveries o f ancient foreign coins, hitherto u nkn ow n in
Sri Lankan context, Sesquicentennial Commemorative Volume, Rcyal Asiatic Society of Sri
Lanka 1 8 4 5 -1 9 9 5 , C o lo m b o , 1995, pp. 1 2 7 -1 3 9 , pi. I & II.
---- (1995 B) S ea -b o rn and in lan d trade o f ancient Sri Lanka: first results o f
the Sri Lanka French exploratory program , Association of South Asian Archaeologists
in Western Europe, 3—7 July 1995 > C am bridge, South Asian Archaeology, I 995 >
pp. 285- 299
---- (1996) Seafaring in the Indian O cean: A rchaeological evidence from Sri
Lanka, in Ray, H im anshu Prabha and Salles, Jean-Francois (eds) Tradition and
Archaeology; early maritime contacts in the Indian Ocean New D elhi: M anoharpp. 59—77*
---- ( l 997) T h e posthum ous coinage o f H erm aios and the conquest o f
G andhara by the Kushans, in F. R aym ond et al (eds) Gandharan Art in Context.
East-West Exchanges at the Crossroads of Asia. New D elhi: Regency Pu blications
p p . 18 9 -2 13 -
B opearachch, O . and Pieper, W . (1998) Ancient Indian Coins. B rjp ols, B elgium :
T u rn h o u t.
B ou ch on , Genevieve and L o m b ard , Denys (1987) T h e Indian O cea n in the
F ourteen th C e n tu ry ’ , in A sh in Das G upta and M .N . Pearson eds., India and
the Indian Ocean 1 5 0 0 —1 8 0 0 , Calcutta: O x fo rd U niversity Press.
Bouzek, J. and D eraniyagala, S .U . (1985) Tessons de vases hellenistiques
trouves en S ri Lanka, Bulletin de correspondance hellenique C IX , pp. 58 9—59^-
Bowen, R. L eB aro n (1951) T h e D how S a ilo r’ , American Neptune I I : 161—2 0 2.
Boxer, C .R . (1940 ) A d m ira l Joao Periera C o rte -R e a l and the C o n stru ctio n o f
Portuguese E ast-Ind iam en in the 17th century, The Mariner's Mirror, 26,
PP- 396 - 4 0 4 -
Braudel, Fernand (1972) The Mediterranean and the Mediterranean World in the Age of
Philip II, 2vols., L o n d o n : Fontana (Translated in to English by S. Reynolds).
Broadhurst, R .J .C . trans. (1952) The Travels of Ibn Jubayr, L o n d o n : J. C ape.
B roersm a, R. ( l 934 ) K oopvaardij in de M olukken, Koloniaal Tijdschrift, 23,
pp. 129-47 and 3 2 0 - 5 0 .
320 B ib l io g r a p h y
---- ( 1935- 36 ) Lan d en volk in M o lu k k en -Z u id , Koloniaal Tijdschrift, 24 *

pp. 4 1 6 - 3 4 , and 25, pp. 42 - 71-
Broeze, Frank ( l 979) T h e m erchant fleet o f Java, 18 20 —185O: a prelim inary
survey’ , Archipel, 18, pp. 251—69.
Broeze, Frank ed. (1989) The Brides of the Sea, Ports Cities of Asia from the l 6 —20 th
Centuries, K en sin gton : New South Wales University Press.
---- (1992) The External Dynam ics o f Port C ity M orphology: Bom bay
1815—19! 4 ’ Banga, I., ed., Ports and their Hinterlands in India 1J 0 0 —1$ $ 0 , New
Delhi: Manohar Publishers, pp. 245 —272.
---- (1996) T h e ports and port systems o f the Asian seas: an overview with
historical perspective from c. 175° , in Agustin G uim era and D olores
R om ero, eds., Puertosj sistemas portuanos, siglos XVI-XX, pp. 9 9 —121, M adrid:
M inisterio de Fom ento.
Broeze, F .J.A ., M cPherson, K .I. and Reeves, P.D . (1987) E ngin eerin g and
Em pire: The Making o f the M odern Indian O cean Ports, C h andra, S. ed.,
The Indian Ocean: Explorations in History, Commerce and Politics, New D elhi: Sage,
pp. 256- 3 0 1 .
B roh ier, R .L . (19 35^ Ancient Irrigation Works in Ceylon, T h e C eylo n G overn m en t
Press, C o lo m b o .
Buckley, C .B . (1965) An anecdotal history of Singapore in old times, K uala L u m pu r:
University o f Malaya Press, 2n d ed.
B urckhardt, Jo h n L. (1968) Travels in Arabia. L o n d o n : Frank Cass, 2n d ed.
C am p bell, Gwyn (1989) M adagascar and M ozam bique in the slave trade o f the
western Indian O cean , 18 0 0 —1861, in W illiam G . C lare n ce-S m ith , ed.,
T h e economics of the Indian Ocean slave trade in the nineteenth century, pp. 166—93,
L o n d o n : Frank Cass.
C am p o , J .N .F .M . a (1992) Koninklijke Paketvaart Maatschappij, stoomvaart en
staatsvorming in de Indonesische archipel, l 888 —i g i 4 , H ilversum : V erlo ren .
Carsw ell, J. (19 9 ° ) T h e excavation o f M antai, Ancient Ceylon n o .7, 1990,
pp. 17 -2 8 .
---- (1991) T h e p ort o f M antai, Rome and India: the ancient sea trade. Madison.-
University o f Wisconsin Press.
C arter, J .R .L . (1984) Merchant families of Saudi Arabia, L on d on : Scorpion Books.
Casson, L io n el. ( 1 9 6 4 ) Illustrated History of Ships and Boats. D oubleday 8l C o , Inc.
New Y o rk .
---- (1986) P. V in d o b G 4 0 8 2 2 and the S h ip p in g o f G ood s fro m India, Bulletin
of the American Society of Papyrologists, vol. 23 * 3—4 * PP- 73~ 79-
---- (1997) The Indian Express, Madras, IO February.
---- (1992) A n c ie n t Naval T ech n o logy and the Route to In d ia ’ , in Vim ala
Begley and R ichard D aniel de Pum a eds., Rome and India: the Ancient Sea Trade,
D elhi: O x fo rd U niversity Press (Indian reprin t): 8—II.
---- (1980) 'R o m e ’ s Trade with the East: T h e Sea Voyage to A frica and In d ia ’ ,
Transactions of the American Philological Association, IIO: 21—36.
---- (1988a) 'R o m e ’ s M aritim e Trade with the Far East’ , The American Neptune,
XLVIII: 149- 53 -
B i b l io g r a p h y 321
---- (1988b) R o m e ’ s Trade with the Eastern Coast o f In d ia’ , Cahiers d ’histoire,
Lyon: 3 0 3 —08.
---- (1990 ) 'N ew Light o n M aritim e Loans: P. V in d o b G 4 ° & 2 2 \ Zgitschrift fur
Papyrologie und Epigraphik, 8 4 : 195— 2 0 6 .
---- (1991) T h e A ncient M ariners, Seafarers and Sea Fighters o f the
M editerranean in A ncient Tim es, Princeton.- Princeton University Press.
Casson, L io n el edited and translated (1989) Periplus Maris Erythraei, Princeton:
Prin ceto n U niversity.
Cathay, see Y u le .
Cave, H .W . (190 8) The Book of Ceylon, L o n d o n : Cassell &. C om p an y L td ., .
C hakravarti, R anabir (1986) 'M erchants o f K o n k a n ’ , Indian Economic and Social
History Review, XX III: 207 —15
---- (1989) 'O verseas Trade in H orses in Early M edieval India: S hip pin g and
P iracy’ in D .C . Bhattacharyya and D evendra H anda eds., Praci-prabha
(Professor. B.N. Mukherjee Felicitation Volume), Delhi: Harman Publishers: 243 ~~^°-
---- (1990) M onarchs, M erchants and a Matha in N o rth e rn K onkan
( 9 0 0 —1 0 5 3 )’ , Indian Economic and Social History Review, X X V II: 189—2 0 8 .
---- (1991) 'H orse T rade and Piracy at Tana (= T hana, M aharashtra, India):
G leanings fro m M arco P o lo ’ , Journal of the Social and Economic Hisotry o f the Orient,
XX XIV : 1 5 9 -8 2 .
---- (1992) 'M aritim e Trade in H orses in Early H istorical Bengal: A Seal from
C han d raketu garh ’ , Pratnasamiksha, I: 15 5—60.
---- (1995) 'Rulers and Ports: Visakhapattanam and M otupalli in Early
M edieval A n d h ra ’ , in K .S . M athew ed., Mariners, Merchants and Oceans, Studies in
Maritime History, D elhi: M anohar: 57~ 7&-
---- (1996a) 'T h e E xport o f Sindan i Indigo fro m India to the West in the
Eleventh C en tury’ , Indian Historical Review, X V III: 18—30 . (volume originally
b elon ging to 1992, but published in 1996).
---- (1996b) 'M aritim e Trade and Voyages in A n cie n t B en gal’ , Journal of Ancient
Indian History, XIX: I 45—71 (volum e origin ally b elo n g in g to 1992, but
publish ed in 1996).
---- (1996c) Vangasagarasam bhandariyaka: A R iverine Trade C en tre o f Early
M edieval B en gal’ , in D ebala M itra ed., Explorations in Art and Archaeology of South
Asia, Essays dedicated to N.G. Majumdar, Calcutta: G ovt, o f West Bengal: 557—72
---- (1998a) Coastal Trade and Voyages in K o n kan : T h e Early M edieval
S ce n ario ’ , Indian Economic and Social History Review, XX XV: 9 7—124
---- (1998b) Vangasagara and O th e r Related Term s: an E xam in ation ’ , in A sok
Datta ed., History and Archaeology of Eastern India, New D elhi: B ooks and Books:
254- 64 -
Cham pakalakshm i, R. (1996) Trade, Ideology and Urbanization: South India 300 BC to
AD 1^ 0 0 , D elhi: O x fo rd U niversity Press.
C han dra, Satish ed. (1987) The Indian Ocean: Explorations in History, Commerce and
Politics, D elhi: Sage.
Ch audh uri, K .N . (1985) Ships and Sh ipbuildin g in the Indian O cean, Trade
and Civilisation in the Indian Ocean, C am bridge University Press, pp. 138—159.
---- (1985) Trade and Civilisation in the Indian Ocean from the Rise of Islam to ljO O ,
C am bridge: C am bridge U niversity Press.
---- (1990a) Asia Before Europe, Economic Civilisation in the Indian Ocean from the Rise of
Islam to 1JOO, C am bridge: C am bridge U niversity Press.
---- (1990 b) 'R eflections o n the O rgan isin g Principles o f P re -m o d e rn T ra d e ’ ,
in Jam es D . Tracy ed., The Rise of Merchant Empires, C am bridge: C am bridge
U niversity Press: 4 2 1— 42-
C h a u d h u ri, M . (1976) S h ip -b u ild in g in the Yuktikalpataru and Samarangana
Sutradhara, Indian Journal of History o f Science, II , pp. I 37- 147-
C hittick, N . (1980) Sewn Boats in the western Ind ian O cean , and a survival in
Som alia, The International Journal of Nautical Archaeology and Underwater Exploration,
9 .4 , pp. 297- 3 0 9 .
Christides, V. (1984) The Conquest of Crete by the Arabs (ca. 824)'- A Turning Point in the
Struggle between Byzantium and Crete, A thens.
C lare n ce-S m ith , W illiam G . (1989a) The economics of the Indian Ocean slave trade in
the nineteenth century, L o n d o n : Frank Cass..
---- (1989b) Indian business com m un ities in the western Ind ian O cea n in the
n i n e t e e n t h c e n t u r y , The Indian Ocean Review, 2 , 4 > PP - 1 8 — 2 1 .
---- ( l 997) H a d h ram i e n tre p re n e u rs in the M alay w o r ld , c. 175° to c. 19 4 0 ,
in U lrik e F reitag and W illiam G . C la r e n c e -S m ith , eds., Hadhrami traders,
scholars and statesmen in the Indian Ocean, i y ^ 0 s - ig 6 0 s , pp . 297“ 3 ! 4 > L e id en :
B rill.
---- (1998) T h e e co n o m ic role o f the A rab com m un ity in M aluku, 1816 to
19 4 0 , Indonesia and the Malay World, 26, 74 > PP - 32—49
---- (forth com in g) H orse trading: the eco n o m ic ro le o f A rabs in southeastern
Indonesia, c. 18 00 to c. 1940 , in N ico K a p tein and H u ub de Jon ge, eds.,
Arabs in Southeast Asia, Leid en .
C o d rin g to n , H .W . (1919) A Sinhalese Embassy to Egypt. Journal of the Ceylon
Branch of the Royal Asiatic Society. V o l. X X V III, N o . 72. C o lo m b o .
---- (1924) Ceylon Coins and Currency, Memoirs of the Colombo Museum, Series A, n o. 3,
C o lo m b o .
---- (1925) N otes on C e ylo n topography in the twelfth century, Journal of the
Ceylon Branch of the Royal Asiatic Society, I 925 > PP- 7° —9 1-
C on greve, H . (1850) Madras Journal of Literature and Science, 16.
C oop er, Frederick (1977) Plantation slavery on the east coast of Africa, New Haven:
Yale University Press.
C ornw allis, K in ah a n (1976) Asir before World War I: a handbook, C am bridge and
New Y o rk : O le a n d er Press and Falcon Press, 2nd ed. u n d er new title.
C rib b , J. (1990) N um ism atic evidence fo r K u sh an o-Sasan ian chron ology,
Studia Iranica, I 99 °> PP- I 51—J93 *
---- (1996 A ) C hin ese coin finds fro m South India and Sri Lanka, in Numismatic
Panorama. Essays in the Memory of Late Shri S.M. Shukula, ed. by K .K . M aheshwari &.
B. Rath, New D elhi, 1996, pp. 2 5 3 —269 .
---- (1996 B) T h e early K u sh an kings: new evidence fo r chron ology, in the
P roceedings o f the Weihrauch und Seide Miinzen, Kunst und Chronologie. Ein Symposium
B ib l io g r a p h y 3^3
des Kunsthistorischen Museums, der osterreichischen Akademie der Wissenschaften und der
Universitat, Wien, A p ril 1996 (in prin t).
C unningham , A . (1871) Annual Report of the Archaeological Survey of India, III, pp. 13—46.
---- (1879) The Stupa of Bharhut, L o n d o n .
C urtin , Philip D . (1984) Cross Cultural Trade in World History, C am bridge:
C am bridge University Press.
D ale, Stephen F. (1980) Islamic society on the South Asian frontier: the Mappilas of
Malabar, 1 4 9 8 —1Q22, O x fo rd : C lare n d o n Press.
---- ( l 997) T h e H adhram i diaspora in south-w estern India: the ro le o f the
Sayyids o f the M alabar coast, in U lrike Freitag and W illiam G . C la re n ce-
Sm ith, eds., Hadhrami traders, scholars and statesmen in the Indian Ocean, 1 ^ 0 s - i g 6 0 s ,
pp. 1 7 5 - 8 4 , Leid en : B rill.
Das Gupta, Ashin (1967) Malabar in Asian trade, -*$00, C am bridge:
C am bridge University Press.
---- (1979) Som e Data on Indian S hippin g fro m D utch Sources, Indian
Merchants and the Decline of Surat, c. l j o o —l j ^ o , W iesbaden: Franz Steiner
Verlag, pp. 2 8 0 —292.
Das G upta, A sh in and Pearson, M .N . eds. (1987) India and the Indian Ocean
1 5 0 0 —1 8 0 0 , Calcutta: O x fo rd U niversity Press.
Dasgupta, T . (1923) Sh ipbuildin g and Com m erce in Ancient Bengal, Journal of
the Department of Letters, University o f Calcutta, 10, pp. 111—28.
Davidson, James W. 1903. The Island of Formosa: Past and Present. L o n d o n :
M acm illan.
De Landberg, L .C . (1920 —23) Glossaire Datmois, Volum es I—II, Leiden: E.J. Brill.
De Lan db erg (1942) Glossaire Datinois, ed. K .V . Zettersteen, V o lu m e III,
L eid en : E.J. B rill.
De Rom anis, F. and T ch ernia, A . eds. ( l 997) Crossings: Early Mediterranean Contacts
with India, New D elhi: M an ohar Publishers.
De Vere A lle n , Jam es (1993) Swahili Origins: Swahili Culture and the Shungwaya
Phenomenon L o n d o n : Jam es C urrey.
D eloche, J. (1996) Iconograp hic evidence o n the developm ent o f boat and
ship structures in India (2nd cent. BC — 15th cent. A .D .) : A New A p p ro ach ,
in Tradition and Archaeology: Early Maritime Contacts in the Indian Ocean, eds. H .
Prabha Ray & J.F . Salles (New D elhi: M anohar): 199—224*
---- (1980 ) La circulation en Inde avant la revolution des transports, tom e II, La voie d’eau,
Paris: Ecole francaise d ’E xtrem e-O rien t; depositeur, A. M aisonneuve.
---- (1987) K on k an Warships o f the X lth -X V th centuries as represented on
M em orial Stones, Bulletin de FEcole Francaise dExtreme-Orient, j 6 , p p .165—84.
---- (I 99 1) Boats and Ships in Bengal T erracotta A rt, Bulletin de VEcole Francaise
d'extreme-Orient, j 8 , pp. I—23.
---- ( l 994 ) Transport and Communication in India prior to Steam Locomotion, vol. II: Water
Tran sport, D elhi: O xford University Press.
D epartm ent o f Fisheries and A quatic Resources. ( l 994 ) Fishing craft and Gear of Sri
Lanka. U N D P F A O S R L 9 1 0 0 2 . M arine Fisheries M anagem ent Project.
C o lo m b o .
3^4 B ib l io g r a p h y
D eraniyagala, S .U . (1972) T h e citadel o f A n u rad h ap u ra 1965= Excavations in

G edige A n u rad h ap u ra, Ancient Ceylon n o. 2, 1972, pp- 4 8 —169.
---- (1986) Excavations in the citadel o f A n uradh ap u ra G edige 1984= a
Prelim in ary R eport, Ancient Ceylon n o. 6, 1986.
---- (1992) The Prehistory of Sri Lanka: an Ecological Perspective, 2 parts, C o lo m b o :
D epartm ent o f A rch aeolo gical Survey, G overn m en t o f Sri Lanka.
Devendra, S. (19 7 °) A rabic Gravestone from T rin com alee Dockyard. Journal of
the Ceylon Branch of the Royal Asiatic Society, New Series. X IV. C o lo m b o , pp 28—35.
---- ( l 99 0 a) N ew Light o n some A ra b ic L ithic R ecords in S ri Lanka, in Sri
Lanka and the Silk Road of the Sea, Bandaranayake S., Dewaraja L ., Silva R .,
W im alaratne K .D .G . ed. T h e S ri Lanka N ation al C o m m issio n fo r
U N E S C O and the C en tra l C u ltu ral Fund. C o lo m b o .
---- ( l 99 0 b) P re-m o d e rn Ships and W atercraft o f Sri Lanka, in Sri Lanka and the
Silk Road of the Sea, Bandaranayake S., Dewaraja L ., Silva R, W im alaratne
K .D .G . ed. T h e Sri Lanka N ational C om m ission fo r U N E S C O and T h e
C en tra l C u ltu ral Fund. C o lo m b o .
---- ( l 99 0 c) T h e G alle T ri-L in g u a l Slab In scriptio n (E ditorial n ote) in Sri
Lanka and the Silk Roadof the Sea, Bandaranayake S, Dewaraja L, Silva R,
W im alaratne K .D .G . ed. T h e S ri Lanka N atio n al C o m m issio n fo r
U N E S C O and T h e C en tra l C u ltu ral Fund. C o lo m b o .
---- ( l 99 0 d) Ships and S h ip b u ild in g T raditio n s o f the Indian O cea n as
preserved in Sri Lanka, in The Indian Ocean Review 3 -3 * C en tre fo r Indian
O cea n R egional Studies & C en tre fo r Peace Studies. Perth, pp. 2—5
---- (1993) Rem ains o f riverine C raft: source m aterial fo r ecological and
com m un ity studies, in Bulletin of the Australian Institute of Maritime Archaeology. 17-2.
A ustralian Institute o f M aritim e A rchaeology. A delaide, pp. 17—25 *
---- (1995) Pre -m o d e rn Sri Lankan W atercraft: the tw in -h u lled logboats
SesquicentennialCommemoration Volume of the Royal Asiatic Society of Sri Lanka.
C o lo m b o .
Devic, L .M . (1928) The Book of the Marvels of India, L o n d o n : G eo rge R outledge &
Sons.
D i V arthem a L. (1863) The Travels of Ludovico di Varthema in Egypt, Syria, Arabia Deserta
and Arabia Felix in Persia, India and Ethiopia, AD 1503 to 1508, ed. G .P . B adger and
trans. J.W . Jones, L o n d o n : Hakluyt Society.
D ick, H oward (1989) Japan’ s e co n o m ic expansion in the N etherlands Indies
betw een the first and second w orld wars, Journal of Southeast Asian Studies, 20, 2,
p p . 244 - 72 .
D im m ock, H .L .F . (1944) 'T h e points o f the compass: A rabian style’ , Mariner’s
Mirror 3O: I 54- 6 l.
Diskalkar, D .B . (1949) Excavations at K asrawad, Indian Historical Quarterly,
M arch, 25 * PP* I — 18.
D o b b in , C h ristin e (1983) Islamic revivalism in a changing peasant economy: central
Sumatra 1784—1847, L o n d o n : C u rzo n Press.
D otson , J-E. (1994) Cogs, Caravels and Galleons: The Sailing Ship, 1 0 0 0 —16$0,
L o n d o n : Conway M aritim e Press.
B i b l io g r a p h y 325
Dozy, R .P .A . (1967) Supplement awe dictionnaires arabes, V olum es I —II, L eiden : E J .
B rill.
D u n n , R .E . (1986) The Adventures of Ibn Battuta: A Muslim Traveller of the 14th Century,
B erkeley-L os A ngeles: U niversity o f C a lifo rn ia Press.
Earl, G eorge W. (1971) The eastern seas, Sin gapore: O xford University Press,
reprin t o f 1837 ed.
Edye, J . (1834) D escription o f the V arious Classes o f Vessels constructed and
em ployed by the natives on the coasts o f C orom an del, M alabar and the
Island o f Ceylon for their coasting navigation, in the Journal of the Royal Asiatic
Society of Great Britain and Ireland. I- pp. I—14-
el Zein, Abdul H am id M. ( l 974) The sacred meadows, a structural analysis of religious
symbolism in an East African town, Evanston: Northwestern University Press.
Ellacott, S .E . (1958) The story of ships, L o n d o n : M ethuen, 2n d ed.
Entyclopaedia Britannica, (1910) Eleventh E d ., New Y o rk: Encyclopaedia B rit-
annica C o .
Engelhardt, R.A . and R ogers, P.R. ( l 997) M aritim e adaptive strategies in
post-Pleistocene Southeast Asia, Bellw ood, P. and T illotson , D. eds. Indo-
Pacific Prehistory: The Chiang Mai Papers, C an b erra: A u stralian N ation al
University, pp. 177—92*
E nglan d, B. (1955) 'The K am al’ , Mariner's Mirror 41: 165—6.
Ewald, Ja n e t J . (1989) The N ile valley system and Red Sea slave trade, in
W illiam G . C laren ce-Sm ith , ed., The economics of the Indian Ocean slave trade in
the nineteenth centuiy, pp. J I —Q2 , L on don : Frank Cass.
Ewald, Janet, and C lare n ce-S m ith , W illiam G . ( l 997) T h e eco n o m ic role o f
the H adhram i diaspora in the Red Sea and G u lf o f A d en , l8s>Os to 1930s,
in U lrike Freitag and W illiam G . C lare n ce-S m ith , eds., Hadhrami traders,
scholars and statesmen in the Indian Ocean, l7 5 0 s-ig 6 0 s, pp. 281—96, Leiden: B rill.
Facey, W . (1987) T h e boat carvings at Jabal Al-Jussasiyah, n orth -east Q atar,
Proceedings for the Seminar for Arabian Studies, vol. 17, pp. 199—221.
Fahmy, A .M . (!966) Muslim Naval Organization in the Eastern Mediterraneanfrom the
Seventh to the Tenth c. A.D., C airo: N ational Pu blication .
Fatimi, S .Q . (1996) The History o f the K am al, Ray, H .P. and Salles, J . - F . ,
eds., Tradition and Archaeology: Early Maritime Contacts in the Indian Ocean, New Delhi:
M anohar Publishers, pp. 2 8 3 —92*
Fattah, Hala M. ( l 997) The politics of regional trade in Iraq, Arabia and the Gulf
1J 4‘5 ~ 19 c>0 > Albany: State University o f New York.
Ferran d, G abriel (1891—1902) Les musulmans a Madagascar et aux Comores, Paris:
Ernest L eroux.
---- (1 3 1 4 )Relations de voyages et textes geographiques, relatifs a VExtreme Orient, vol. II,
Paris.
---- (1924) 'L ’JIJm en t Persan dans les textes nautiques Arabes des XViem e et
X V Iiem e siecles’ , Journal Asiatique, A vril-J u in .
Filliozat, J . (1956) Les relations exterieures de I'lnde (I), Pondichery.
Footm an, David (1986) Antonin Besse of Aden: the founder of St Antonys College Oxford,
Basingstoke: M acm illan.
Forbes, Anna (1987) Unbeaten tracks in the islands of the Far East: experiences of a
naturalist’s wife in the 1880s, Singapore: O xford University Press, reprin t o f
1887 b ook entitled Insulinde.
Foucher, A. ( l 947) La vieille route de I’lnde, de Bactres a Taxila, MDAFA, I, Paris: Les
Jd itio n s d ’art et d ’histoire.
Fraenkel, S. (1962) Die Aramaischen Fremdwdrter im Arabischen, H ildesheim : G eorg
O lm s.
Francis, P. J r . (1987) Bead Emporium: A guide to the Beads from Arikamedu in the
Pondicherry Museum, G overnm ent o f Pondicherry, Pondicherry.
Freem an-G renville, G .S .P . (1962) The East African Coast, O xford: C laren don
Press.
---- (1982) ed. and trans. The Book of the Wonders of India by Captain Buzurg ibn
Shahriyar of Ramhormuz, L o n d o n and H ague: East West Publications.
Freitag, Ulrike (1997) C on clusion : the diaspora since the age o f in d ep en
dence, in Ulrike Freitag and W illiam G. C larence-Sm ith, eds., Hadhrami
3^5—^ 9 » Leiden:
traders, scholars and statesmen in the Indian Ocean, lJÔ s to ig 6 o s,
Brill.
---- (forthcom ing) A rab m erchants in Sin gapore: attempt at a collective
biograph y in N ico K a p te in and H u u b de Jon ge, eds., Arabs in Southeast Asia,
L eid en .
Freitag, U lrike, and C laren ce-Sm ith , W illiam G ., eds., Hadhrami traders, scholars
and statesmen in the Indian Ocean, lyÔ s-igÔ s, L eiden: Brill.
Frere, Sir Bartle (1873) M em oran dum by Sir B. Frere regarding Banians or
natives o f India in East A frica, Irish University Press Series of British Parliamentary
Papers, Slave Trade g i, Sh ann on : Irish University Press.
G aletti, A ., and G ro o t P., com ps. (1911) The Dutch in Malabar, M adras:
G overn m en t P rinter.
Garlake, P .S. and M. (1964) Early Ship Engravings o f the East A frican Coast,
Tanganyika Notes and Records, vol. 63*
G arrett, R. (1987) The Symmetry of Sailing. A d lard C oles, L o n d o n . (C ited by
V osm er).
G avin, R.J. (1975) Aden under British rule, 1839—1967, L o n d o n : C . H urst.
Gazetteer of the Bombay Presidency (1899) Bom bay: G overn m en t C e n tra l Press.
G eiger, W. tr. (1921) The Culavamsa: being the more recent part of the Mahavamsa.
H um phrey M ilford , O xford University Press.
G hosh, A ed. (1989) An Encyclopaedia of Indian Archaeology, in two volum es, D elhi:
M unshiram M anoharlal.
G ib so n -H ill, C .A . 1952- Further N otes on the O ld Boat fou n d at Pontian in
South ern Pahang. Journal of the Malayan Branch of the Royal Asiatic Society 25 : I:
m -13 3 .
G ildem eister, J. (1882) U eber arabisches Schiffswesen, in Nachrichten von der
Koniglischen Gesellschaft der Wissenschajten und der Georg-Augusts Universitat zu
G ottingen, 1—23= 43 I_ 4 &-
G ill, J .S . ( l 993) O u r H eritage o f T rad itio n al Boat B u ild in g. Journal of Marine
Archaeology. 4 * G oa. pp. 74—75-
B ib lio g r aph y 327
G lover, I .C . (1996) 'R ecen t A rch aeolo gical Evidence fo r Early M aritim e
C ontacts betw een India and Southeast A sia ’ , in H im anshu Prabha Ray and
J.F . Salles eds. Tradition and Archaeology, Early Maritime Contacts in the Indian Ocean,
D elhi: M anohar: 2 9 3—310.
G ob i, R. (1971) Sasanian Numismatics, Braunschweig: K lin kahardt & B ierm an n .
G o ite in , S .D . (1954) 'Two Eye W itness R eports o n the Invasion o f A d e n by
K ais in the T w elfth C e n tu ry ’ , Bulletin o f the School o f Asian and African Studies, XV I:
247- 57
---- (1963) 'Letters and D ocum en ts on the India T ra d e ’ , Islamic Culture,
X X X V II: 1 8 8 -2 0 5 .
---- (1967) A Mediterranean Society: The Jewish Communities o f the Arab World as Portrayed in
the Documents o f the Cairo Geniza, V olum es I-IV . Berkeley: U niversity o f
C a lifo rn ia Press.
---- (1973) Letters o f Medieval Jewish Traders, P rinceton: P rin ceto n U niversity Press.
---- (1980) 'F rom A d e n to India: Specim ens o f C o rresp o n d en ce o f India
T raders o f the T w elfth C e n tu ry ’ , Journal o f the Economic and Social History o f the
Orient, XXI: 4 3 —66.
---- (1987) 'Portrait o f a Medieval India T rader: T h ree Letters from the C airo
G en iza ’ , Bulletin o f the School o f Asian and African Studies, XLVIII: 4 4 8 —64*
G o le, S. (1989) Indian Maps and Plans, New D elhi: M anohar Publishers.
G o o r, J. van (1982) 'T h e death o f a m iddlem an: schem ing in the m argins o f
the D utch East In d ies’ , in G . Schutte and H eather S u th erlan d , eds., Papers o f
the Dutch-Indonesian Historical Conference, 2 3 —2 7 J une PP* 223~ 49 > L eid en
and Jakarta: B ureau o f Indonesian Studies.
G op al, Lallanji. (1962) A rt o f sh ip bu ildin g and navigation in ancient In d ia’ ,
Journal of Indian History 4 0 :3 1 3 —27*
---- (1965) T h e Economic Life o f Northern India (c. AD 700 —1200), D elhi: M otilal
B a n a r a s id a s .
---- (1970 ) 'In d ian S h ip p in g in the M edieval P e rio d ’ in Lokesh C h an dra et al
ed., India’s Contribution to World Thought and Culture (Swami Vivekananda Commemora
tion Volume) M adras: 10 8—22.
G otth old, Ju lia (1988) A Bibliography o f the Indian Ocean, O xford: C lio.
G ran d id ier, A lfre d , and G uillaum e (1908) Histoire physique, naturelle et politique de
Madagascar, Tome IV, Paris: H achette.
G ran d p re, L. D e (1803) A Voyage in the Indian Ocean and to Bengal, I, L o n d o n .
Gray, J.M . ( l 955) N airu zi or Siku ya Mwaka’ , Tanganyika Notes & Records 37:
I—21.
G reen , J .N . et al. (1998) Joint Sri Lanka-Australia-Netherlands Galle Harbour Project
19 9 & ~ 1997 > Archaeology, History, Conservation and Training, C o lo m b o .
G reen h ill, B. (1956) T h e K arach i Fishing Boats, The M ariner’s Mirror, 42-1,
pp. 54- 66 .
---- (1971) Boats and Boatmen o f Pakistan, N ew ton A b b ott.
---- (1976 and 1995) The Archaeology o f Boats and Ships, L o n d o n : Conw ay M aritim e
Press.
---- (1995) The Archaeology o f Boats and Ships, Conway M aritim e Press.
328 B iblio g r aph y
G renet, Fr. (1996) Les m archands sogdiens dans les m ers du Sud a l ’epoque
preislam ique, Cahiers d ’Asie Certtrale, 1996, pp. 65—84 .
G renet, Fr. and Bopearachchi, O . (199G) U ne m onnaie en or du souverain
in do-parth e Abdagases II Studia Iranica, 25 * x9 96, pp. 219—231.
G ren ier, R. (1988) Basque whalers in the New World. Pp. 6 9 —84* in G .F.
Bass ed. Ships and Shipwrecks o f the Americas. T h am es & H udson . L on d on .
G rohm an n , A d o lf (1933) Siidarabien als Wirtschaftsgebiet, zweiter Teil, B riin n : R u d o lf
M. R ohrer.
G rosset-G ran ge, H . (1972) 'C om m en t navignent au jo u rd ’hui les Arabes de
l ’O cean In d ie n ? ’ Arabica, 1 9 :4 6 —77
---- (1996) A ran N autical Science, R ashed, R ., ed., Engjclopedia o f the History o f
Arabic Science, L o n d o n and New Y ork: Routledge, vol. 3, pp. 2 0 2 —242-
G uillain, C. (1856—57) Documents sur Vhistoire, la geographie et le commerce de VAfrique
Orientale, Paris: A . B ertran d , 3 vols.
G unaw ardana, R .A .L .H and Y u m io Sakurai. (1990) Sri Lankan Ships in
C h in a . Sri Lanka and the Silk Road o f the Sea, Bandaranayake, S ., Dewaraja L,
Silva R . , W im alaratne K .D .G . (eds.) Sri Lanka N ational C om m ission fo r
U N E S C O and T h e C en tra l C u ltu ral Fund. C o lo m b o .
---- (1990) Seaways to Sielediba: C h anging patterns o f navigation in the
Indian O cean and their im pact on pre-colo n ial Sri Lanka. Sri Lanka and the
Silk Road o f the Sea, Bandaranayake S., Dewaraja L, Silva R ., W imalaratne
K .D .G . (eds). (Revised from Kalyani Vols. V & V I o f 1986—87). Sri Lanka
N ational C om m ission fo r U N E S C O and T h e C en tral C ultural Fund.
C olom b o.
G uruge, A.W .P. (1 9 9 °) T h e Sri Lankan Factor in the D evelopm ent o f the A rt
o f Sukothai and Lann a Tai. Sri Lanka and the Silk Road o f the Sea, Sri Lanka and the
Silk Road o f the Sea, Bandaranayake, S., Dewaraja L, Silva R ., W imalaratne
K .D .G . ed. Sri Lanka N ational C om m ission for U N E S C O and T h e
C en tral C ultural Fund. C olom b o.
G urum urthy, S. (1981) Ceramic Traditions in South India, M adras.
Gyselen, R. ( l 993) Catalogue des sceaux, camees et bulles sassanides, Bibliotheque
N ationale, Paris.
H afkin , N ancy J. ( l 973) Trade, society and politics in northern Mozambique, c .l// 53 ~ 1913 ’
PhD Thesis, B oston U niversity.
H aines, S .B . (1906) M em oir o f the south coast o f Arabia, fro m the entrance o f
the Red Sea to M isena’t, in G eorge W . Forrest, ed., Selections from the travels and
journals preserved in the Bombay secretariat, Bom bay: G overnm en t C en tral Press.
H ajji, Y .Y . al- (1993) Nuwakhudha al-Safar al-Shira’ f i al-Kuwait, Kuwait: Shirkat
a l-R a b i’an.
H all, K .R . (1 980) Trade and Statecraft in the Age o f the Colas, D elhi: Abhinav.
H asan, H . (1928) A History o f Persian Navigation, L on d on .
Hawkins, C.W . (1965) T h e T u tico rin T h o n i, The M ariner’s M irror, 51*
p p . 147-154.
H eadrick, D aniel R. (1988) The tentacles o f progress, technology transfer in the age of
imperialism, 1 8 5 0 —1 9 4 ° ’ O xford: O xford University Press.
H eim an, J. (1980 ) Sm all Exchange and Ballast: Cow ry Trade as an Exam ple
o f Indian O cea n E co n o m ic H istory’ , South Asia, II: 4 8 —69.
H e in e -G e ld e rn , R obert 1932. U rh eim at u n d friiheste W anderungen der
A u stron esier. Anthropos 27 : 5 4 3 —619.
H ill, A .H . (1958) Som e Early A ccoun ts o f the O rien ta l Boat, Mariner’s Mirror,
44 , pp. 2 0 1 - 2 1 7 .
H irsch, Leo (1897) Reisen in Siid-Arabien, Mahra-Land, und Hadramut, Leid en : E .J.
Brill.
H ogarth, David G . (1905) The penetration of Arabia, L o n d o n : Alston Rivers.
— (1978) [revision o f 1917] Hejaz before World War I: a handbook, C am bridge and
New Y ork: O le a n d er Press and Falcon Press, 2nd ed. u n d er new title.
H o rd en , Peregrine and N icholas Purcell (2 0 0 0 ) The corrupting sea: a study of
Mediterranean history, O x fo rd : Blackwell.
H o rn ell, J. (1918—23) 'T h e O rigin s and E thn ological Significance o f Indian
Boat D esign s’ , Memoirs o f the Asiatic Society o f Bengal, V II: 139—256.
---- (1920 ) T h e origin s and ethnological significance o f Indian Boat designs.
Memoirs o f the Asiatic Society o f Bengal. 7- Calcutta, pp. I 24 —I 4 I -
---- 192 0 . T h e O u trig ge r C an oes o f Indonesia. Rept. No. 2 , Madras Fisheries Bulletin
1 2 : 4 3 - 114 *
---- 1936. Boat C o n stru ctio n in Scandinavia and O cean ia: A n o th e r Parallel in
B otel T obago. Man 36 (Art. 20 0 ): 144—! 47 *
---- (1943) Fishing and coastal craft o f Ceylon. The Mariner’s Mirror. 29 -1-
PP 40 - 53 *
---- (1945) Pearling fleets o f S. India and C eylon . Mariners M irror 3I: 2 I 4 ~ 23 °*
---- (1946) The Sailing Craft o f Western India, The Mariner’s Mirror, 32, pp. 195—217*
---- (1946) Water Transport. Origins & Early Evolution. C am bridge University Press.
---- (1970) Water Transport, Newton Abbott.
H o rrid ge, G . A d rian (1981) The prahu , traditional sailing boat o f Indonesia, K uala
L u m p u r: O x fo rd U niversity Press.
---- 1982. T h e L ash ed -L u g Boat o f the Eastern A rchipelagoes. Maritime
Monographs and Reports, National Maritime Museum, No. $ 2 . G reenw ich: N ational
M aritim e M useum .
---- (1995) T h e A ustronesian C onquest o f the Sea — U pw ind, Bellw ood, P.,
Fox, Jam es, J. and T ryon , D arrell, The Austronesians: Historical and Comparative
Perspectives, C an b erra: A ustralian N ational U niversity.
H o rto n , M. (1996) Shanga: the archaeology o f a Muslim Trading Community on the Coast of
East Africa, L o n d o n : T h e British Institute in Eastern A frica.
H o u ra n i, G .F . (I 951) The Arab Seafaring in the Indian Ocean during the Ancient and Early
Medieval Times, B eirut: Khayats.
H ouran i, G .F. (1963, reprinted 1975) Arab Seafaring in the Indian Ocean in Ancient
and Early Medieval Times, New York: O ctagon Books.
Hsii Y in g - c h o u 19^2. Yami Fishing Practices: Migratory Fish. (Taiwan A b o rig in e
M on ograp h Series i) Taipei, Taiwan: S ou th ern M aterials C e n ter.
H u ff, W .G . ( l 994 ) The economic growth o f Singapore; trade and development in the twentieth
century, C am bridge: C am bridge U niversity Press.
330 B ib lio g r a p h y
H u n tin g fo rd , G .W .B . translated (1980) The Periplus of the Erythraean Sea, L o n d o n :

Haklyut Society.
H u tch in so n , G . (1991) Early 16th century wreck at Studland Bay, D orset.
Pp. 171—175» in R- R ein ders & K . Paul (eds.) Carvel Construction Techniques.
O xbo w M o n o . 12. O x fo rd .
H uzayyin, S .A . (1982) Arabia and the Far East, C a iro : Institut Franâis
d A r c h e o lo g ie O rien ta le au C aire.
Ibn a l-A th ir (1965) Al-Kamil ft l-Tar\kh, V o lum es I-X II, B eirut: D ar Bayrut.
Ibn a l-K h a tib (1956) A'mal al-alam ft man buyi'a qabl al-isldm min muluk al-isldm
(H istoire de l ’ Espagne m usulm ane), ed. L. Provencal, B eirut.
Ibn Battuta (1958, 199 4, 1995) The Travels of Ibn Battuta A D 1 3 2 5 -1 3 5 4 .
T ran slation with revision and notes fro m the A ra b ic text ( o f C . D efrem ery
and B .R . Sanguinetti) by H .A .R . G ib b , V o lu m e I, C am bridge: C am bridge
U niversity Press -V olu m es II—III, L o n d o n : Hakluyt Society (a rep rin t) -
V o lu m e IV, translation com pleted with annotations by C .F . B eckingham ,
L o n d o n : H akluyt Society.
---- (1968) Voyages dlbn Battuta. A ra b ic text with translation by C . D efrem ery and
B .R . Sanguinetti, V o lu m es I-IV , Paris: A n th ro p o s.
Ibn Hawqal (1873) Kitab surat al-ard , ed. M.J de G o eje, Leid en : E.J. B rill, 1873.
Ibn Jubayr (1852) Rihlat Ibn Jubayr, ed. W . W right, L eid en : E.J. B rill.
---- (1952) The Travels of Ibn Jubayr, trans. R .J .C . B roadhurst, L o n d o n : Jo n athan
C ap e.
Ib n K h a llik a n ( 1 8 4 2 —7^ Wafayat a l- a ja n wa anba’ al-iam an (Ib n K h a llik a n ’ s
B io g r a p h ic a l D ic tio n a ry ), tran s. M c G u c k in de S lan e, V o lu m e s I - I V ,
Paris.
Ibn Rustah (1892) Kitab al-alaq al-naftsa, ed. M .J. de G oeje (Biblioth eca
G eo gra p h o ru m A ra b ico ru m , V I), Leid en : E.J. B rill.
Im am , S .A . (1990 ) C u ltu ral R elations betw een Sri Lanka and Iran. Sri Lanka
and the Silk Route of the Sea, Bandaranayake S ., Dewaraja L, Silva R .,
W im alaratne K .D .G . ed. S ri Lanka N atio n al C om m ission fo r U N E S C O
and T h e C en tra l C u ltu ral Fund. C o lo m b o .
Imperial Gazetteer o f India (190 9) Calcutta: S u p erin ten d en t o f G overn m en t
Prin tin g.
Indrapala, K . (i9 6 0 ) T h e N ayinativu T am il Inscrip tio n o f Parakkram abahu I,
in The University o f Ceylon Review. X X I :I: C o lo m b o , pp. 63—7° •
---- (1990 ) S outh Ind ian m ercan tile com m un ities in C eylon , circa 9 5 0 —1250 ,
Sri Lanka and the Silk Road of the Sea, 1990, p p. 153—160.
Ingram s, W .H . (1937) A report on the social, economic and political conditions of the
Hadhramaut. L o n d o n : H M S O .
---- (1942) Arabia and the isles, L o n d o n : J o h n M urray.
Inizan, M . - L ., (1991) La p ro du ction de la cornaline en O rien t. Les collections
de T ello au M usee du Louvre, La pierre prehistorique, Paris, pp. 133—146.
Innes, C .A . (190 8) Malabar and Anjengo, M adras: G overn m en t Press (Madras
D istrict G azetteers).
Istituto G eo gra fico de A go stin i (1913) VEritrea economica, N ovara and R om e.
B iblio g r aph y 331
Ivens, W .G . 1927- Melanesians o f the Southeast Solomon Islands. L on d on : K egan Paul,
Trench, T ru b n er & C o.
Ives, E. ( l 773) A Voyage from England to India in the Year 1754> L on don .
Jacob s, H ubert, ed., I 971- A Treatise on the Moluccas ( c. 1544 ) Probably the Preliminary
Version o f Antonio Galvao’s Lost Historia das Molucas. R om e: Je su s H istorial Institute.
Jansen, E.J., D olm an , A .J ., Jerve, A .M . and Rahm an, N . (1989) The County
Boats o f Bangladesh, Dhaka: T h e U niversity Press.
Jansen, W .H .M . (1990 ) De economische ontwikkeling van de Residentie Menado,
19OO—1 9 4 0 > D octoraal Scriptie, Rijksuniversiteit te L eid en .
Je d re j, M .C . (1985) T h e Red Sea and the eastern Sudan in the early
nineteenth century, in Jeffrey C. Stone, ed., Africa and the sea, proceedings o f a
colloquium at the University o f Aberdeen, pp. I04~~l6, A berdeen: A berdeen
University A frican Studies G roup.
Jo h n sto n , Sir A lexander. (1824) A C ufic Inscription found in Ceylon., in
Transactions o f the Royal Asiatic Society o f Great Britain and Ireland. I. I. L on d on .
PP- 537- 548 .
John ston e, P. (1980) The Seacraft of Prehistory, L on don: Routledge and K egan Paul.
J u n -yan , Z. (1980) R elations between C h in a and the Arabs in early tim es,
Journal of Oman Studies, 6, i: 91—109.
Kam ioka K . and Yajima H. (1979) The Inter-regional trade in the Western Part of the Indian
Ocean (Studia Culturae Islamicae, No 9), Tokyo: Institute for the Study o f
Languages and Cultures o f Asia and Africa (original work in Japanese).
K an e, P .V (1962) History o f the Dharmasastra, V , Poona: B handarkar O rien ta l
Research Institute.
K an o , Tadao and K okichi Segawa 1956- An Illustrated Ethnography o f Formosan
Aborigines, Vol. 1: The Yami, revised edition. Tokyo: M aruzen Com pany.
K ap itan, G . 1987* Records o f native craft in Sri Lanka, in The International Journal
o f Nautical Archaeology and Underwater Exploration. 1 6 . 2,. T h e N autical A rch aeolo gy
Society, pp. 1 3 5 - 1 4 7 .
---- (1987, 1988, 1989* ! 99 J) Records o f Native C ra ft in Sri Lanka, The
International Journal o f Nautical Archaeology and Underwater Exploration, l6 .2 ,
pp. 135- 147; 17-3 * PP- 2 2 3 - 2 3 5 ; 1 8 .2 , p p . 1 3 7 — 149 ; 2 0 . 1 , p p . 2 3 - 3 2 .
K a rttu n en , K . (1995) Early R om an trade with South India, Arctos: Acta Philologica
Fennica, XX IX , I 995 > PP - 81—91.
K asinathan, N . (1992) Alagankulam. A preliminary Report, M adras.
Kelly, Jo h n B. (1968) Britain and the Persian Gulf 1 79 5 —18 8 0 , O xford: C larendon
Press.
K en tley, E. (1996) T h e Sewn Boats o f In d ia ’ s East Coast, Ray, H .P . and
Salles, J .- F . ed ., Tradition and Archaeology, New D elhi: M anohar Publishers,
pp . 2 4 7 - 2 6 0 .
Kentley, E. and G unaratne, R. 1987* T h e M adel Paruwa sewn boat with chine
strakes. The International Journal o f Nautical Archaeology and Exploration. 1 6 .1. The
N autical Archaeological Society, pp. 35” 48-
K ervran, M. ( l 999) M ultiple Ports at the m outh o f the river Indus: Barbarike,
D eb, Daybul, L ah ori Bandar, D iul Sinde, Ray, H .P. ed. Archaeology of
Seafaring: The Indian Ocean in the Ancient Period, Indian C o u n cil o f H istorical
Research, pp. J l —145-
K h adh uri, N asir b. Ali b. N asir al- (1994) M a ’din al-Asrar f i 'Ilm al-Bihar ('T h e
M ine o f Secrets in the Science o f the Seas’) edited by H asan Saleh Shahab,
Muscat: Ministry o f N ational H eritage and Culture.
K halidi, O m ar (1997) T h e H adhram i role in the politics and society o f
colonial India, I 75° s to I 95 ° s * in Ulrike Freitag and W illiam G. C laren ce-
Sm ith, eds., Hadhrami traders, scholars and statesmen in the Indian Ocean, 1 ^ 0 s - ig 6 0 s ,
pp. 67—81, L eiden: Brill.
K h o u ry , I., tr. (1983) As-Sufaliyya, The Poem o f Sofala by Ahmed Ibn Magid, C o im b ra :
Junta de Investigacoes C ientificas do U ltram ar.
K ieth , D .H . (1988) Shipwrecks o f the explorers. Pp. 45—68, in G .F . Bass ed.
Ships and Shipwrecks o f the Americas. T ham es & H udson. L o n d o n .
K in d e rm a n n , H . ( l 934 ) ^ c^ j f 1171 Arabischen. Untersuchung iiber Vorkommen und
Bedeutung der Termini, B on n: Zwickau.
K irib a m u n e , S. (1990) M uslim s and the T rade o f the A rabian Sea with special
referen ce to Sri Lanka fro m the birth o f Islam. Sri Lanka and the Silk Route o f the
Sea, Bandaranayake S ., Dewaraja L, Silva R ., W im alaratne K .D .G . ed. Sri
Lanka N ational C om m ission fo r U N E S C O and T h e C en tra l C u ltu ral
Fund. C o lo m b o .
K isterm an n , H . ( l 99 x) De paardenhandel van Nusa Tenggara, l 8 l^ —i g 4 1‘ Student
graduation paper, H istory D epartm ent, V rije U niversiteit, A m sterdam .
K n aap , G e r ritJ . (1996) Shallow waters, rising tide, shipping and trade in Java around 1*775,
Leiden : K I T L V Press.
K o n o w , S. (1929) Corpus Inscriptionum Indicarum, vol. II, Part I, Kharoshthi Inscriptions
with the Exception o f those o f Asoka, C alcutta.
K o sam b i, D .D . (1958) Indian Feudal T rade C h a rters’ , Journal of the Economic and
Social History o f the Orient, IV: 281— 93.
K rish n am u rth y, R. (1991) O b lo n g cop p er coin with M other G oddess sym bol
fro m K a ru r, T am ilnadu, Journal o f the Numismatic Society of India, LIII,
PP- 59- 61.
---- (1994) Late Roman Copper Coins from South India: Karur and Madurai, M adras:
G arnet Publications.
---- (1997) Sangam Age Tamil Coins, Madras.
K r o e f, Justus van der (1953) T h e Arabs in Indonesia, The Middle East Journal, J ,
3, pp. 300 - 23 -
K u lk e, H e rm a n n (1 9 9 3 ) 'A Passage to In d ia: T e m p les, M erch an ts an d the
O c e a n ’ , Journal o f the Economic and Social History o f the Orient, X X X V :
15 4 -8 0 .
K u n h alli, V . ( l 993) Ship B u ild in g in Beypore — a Study in M aterials, W orkers
and T echn ology, Marine Archaeology, 4 * PP* 56—59 -
K u p eru s, G e rrit (1936) Het cultuurlandschap van West-Soembawa, G ro n in g e n and
Batavia: J .B . W olters.
K u rie n , J. (1996) D iffu sio n as a process: the case o f plywood boats in m arine
fishing, Review o f Development and Change, July D ecem b er, pp. 219—234 *
Lach, D on ald F. (1965) Asia in the Making of Europe in three volum es, C hicago:
Universtiy o f Chicago.
Lakshm i, L .R .S . (1995) From Y em en to the shores o f M alabar: H adram i
m igrants to C alicu t, a case study, Paper presented at co n feren ce Hadhramaut
and the Hadhrami Diaspora, late 18th Century to c 1967, S O A S , L o n d o n , A p ril.
Landstorm , B jo rn ( 1 9 6 4 ) The Quest for India, Stockholm : Doubleday.
Lane, E.W. (1984) Arabic-English Lexicon, V olum es I —II, C am bridge: Islam ic
Texts Society Trust.
Langley, W .K .M . (1962) Century in Malabar: the history of Peirce Leslie & Co. Ltd.,
M adras: T h e M adras A dvertising C o .
Lawless, Richard I. (1995) T h e role o f seam en’ s agents in the m igration fo r
em ploym en t o f A ra b seafarers in the early twentieth century, in D iane
Frost, ed., Ethnic labour and British imperial trade: a history of ethnic seafarers in the UK,
pp. 3 4 - 5 8 , L o n d o n : Frank Cass.
Le G u e n n e c-C o p p e n s , Franôise ( l 979) ^ es M asharifu Jam alilil a Lam u,
K enya, Annuaire des Pays de VOcean Indien, 6, pp. 91—102-
Leach, E.W . 1937a. T h e Y am i o f K o to -sh o : A Jap an ese C o lo n ia l E xperim en t.
Geographical Magazine 5 = 6 (O cto b e r): 4 17—434*
---- , 1938a. Stone Im plem ents fro m B otel T obago Island. Man 38 (Art. 189):
16 1 -16 3 .
---- 1938b. E co n o m ic Life and T ech n o logy o f the Yam i o f B otel T obago. Man
38 (Art. 4): 9.
---- 1950 . Prim itive C alendars. Oceania 20: 4 : 245“ 262.
Lee, Edwin (1991) The British as rulers: governing multiracial Singapore, 1867—1914-
Singapore: S ingapore U niversity Press.
Legge, J .trn . (1971) A Record of the Buddhistic Kingdoms, D elhi: O rien ta l Publishers
(Indian rpt).
Lekkerkerker, C . (1923) H et voorspel der vestiging van de N ederlandsche
macht op Bali en L om bok, Bijdragen tot de Taal-, Land-, en Volkenkunde van
Nederlandsch-Indie, 79 * pp* J 9 8 —322-
Lekon, Christian (1997) T h e impact o f remittances on the econom y o f H adhra
maut, in U lrike Freitag and W illiam G. C larence-Sm ith, eds., Hadhrami traders,
scholars and statesmen in the Indian Ocean, I7 5 0 s-ig6 0 s, pp. 264—80, Leiden: Brill.
L eu r, J .C . van ( l 955) Indonesian Trade and Society, T h e H ague and B andung: van
H oeve.
Levi, S. (190 0 ) Les m issions de W ang H iu e n -ts’e dans l ’ ln d e, Journal Asiatique,
19 0 0 , pp. 297- 341 . 401-468.
Lewis, A . (1973) 'M aritim e Skills in the Indian O cea n 1368 —1 5 0 0 ’ , Journal of
the Economic and Social History of the Orient, XVI: 23 8 —64*
Lewis, J.P. (1914) Boats and canoes o f Ceylon. Times of Ceylon Christmas Number.
Colom bo.
Logan, W illiam (1887) Malabar, Madras: G overn m en t Press.
Mackay, E.J. (1938) Further Excavations at Mohenjodaro, New D e lh i:.
M acPherson, K . (1 9 9 °) T raditional Indian O cean Sh ippin g Technologies. Sri
Lanka and the Silk Road of the Sea, Bandaranayake, S ., Dewaraja, L ., Silva R,
W im alaratne K .D .A . ed. Sri Lanka N ational C om m ission fo r U N E S C O

and T h e C e n tra l C u ltu ral Fund. C o lo m b o .
M ahadevan, I. (1996) O ld Sinhalese in scrip tion s from Indian ports: New
evidence fo r ancient India — Sri Lanka contacts, Journal of the Institute of Asian
Studies, X IV pp. 5 5 - 6 5 .
M ajor, R .S . (190 0 ) India in the Fifteenth Century, L o n d o n : Haklyut Society.
M ajum dar, R .C . (1962) 'T h e Invasion o f C h o la R ajendra to S ou th-E ast A sia ’ ,
Artibus Asiae, V : 155—62.
M alalasekera, G .P . ed. (1938—39) Vamsattappakasini Vol. II. Pali Text Society.
London.
M alecot, G eorges (1990) Q u elques aspects de la vie m aritim e en M er Rouge dans
la prem iere m oitie du XIXe siecle, L ’Afrique et I’Asie Modernes, 164, pp- 22—43 -
M angnin, P ierre-Y ves 198 0 . The Southeast A sian Ship: An H istorical
A p p ro ach . Journal of Southeast Asian Studies II: 2 5 3 —269.
---- (1980) T h e Southeast A sian Ship: an H istorical A p p ro ach , Journal of
Southeast Asian Studies, II, pp. 2 5 3 - 2 6 9 .
---- (1984) R elationship and C ro ss-In flu e n ce betw een Southeast A sian and
C hin ese S h ip b u ild in g T radtions. In Final Report: SPAFA Workshop on Shipping and
Trade Networks in Southeast Asia, 1984- Bangkok: SPAFA.
---- (1985) Sew n -P lank C ra ft o f S ou th-E ast Asia. In Sean M cG ra il and Eric
K en tly, eds., Sewn Plank Boats: Archaeological and Ethnographic Papers Based on Those
Presented to a Conference at Greenwich in November, 1 9 8 4 • (N ation al M aritim e
M useum , G reen w ich, A rch a e o lo g ica l Series N o. io ) O x fo rd : BAR.
International Series 276, ppp 319—343
---- (1985) Late M ediaeval A sian S h ip b u ild in g in the Indian O cea n , Moyen
Orient & Ocean Indien, Paris: Societe d ’H istoire de l ’ O rie n t, pp. I—30 .
---- (1986) 'Late M edieval S h ip b u ild in g T echn iqu es in the Ind ian Ocean.- A
R eappraisal’ , Moyen Orient et Ocean Indien, III: 31—60.
---- 1996. Southeast Asian Sh ip p in g in the Indian O cea n D u rin g the First
M ille n n iu m AD In H im anshu Prabha Ray and Jean-Franôis Salles, eds.,
Tradition and Archaeology: Early Maritime Contacts in the Indian Ocean. New D elh i,
India: M anohar.
M ansvelt, W .M .F . (1938) De prauwvaart in de 19c eeuw, Koloniale Studien, 22, I,
p p . 8 9 —10 2 .
M arston, T ho m as E. (1961) Britain’s imperial role in the Red Sea area, l 8 0 0 —l 8 j 8 ,
H am den (C o n n .): Shoe S trin g Press.
M artin, B rad fo rd G . (1971) N otes o n som e m em bers o f the learned classes o f
Zanzibar and East A frica in the n in eteen th century, African Historical Studies, 4>
3 > PP- 525- 45 -
M artin E .B . and M artin C .P . (1978) Cargoes of the East: the ports, trade and culture of
the Arabian Seas and Western Indian Ocean L o n d o n : Elm T ree Books.
M artin, Jean (1983) Comores, quatres iles entre pirates et planteurs, Paris: L ’H arm attan,
2 vols.
M ayson, Jo h n S. (1963) The Malays of Cape Town, C ape Tow n: A frican a
C on n oisseu rs Press, rep rin t o f 1861 ed.
M azzarino, S. (1997) O n the N am e o f the H ipalus (H ippalus) W in d in P lin y ’

in Federico R om anis and A n d re T ch ernia eds., Crossings, Early Mediterranean
Contacts with India, D elhi: M anohar: 72—79 *
M cC rin d le, J W . (1901) Ancient India as Described in Classical Literature. A rch ib ald
C on stable. W estm inster. R eprin t I 971*
M cG rail, Sean. (1985) Towards a C lassification o f W ater T ransp ort, World
Archaeology, 16, pp. 28 9—30 3.
---- (1987) Ancient boats in N.W. Europe The archaeology of water transport to AD 1{)00.
Lon gm an , L o n d o n & New Y o rk.
---- (1997) Early fram e-first m ethods o f b u ild in g w ood en ships and boats.
Mariners Mirror 83= 76—80.
M cG rail, S. and K en tley, E. eds. (1985) The Sewn Plank Boats, O x fo rd : B A R
Publications.
M cP herson , K . (1993) The Indian Ocean: A History o f People and the Sea, N ew D elhi:
O x fo rd U niversity Press.
---- (1998) Trade and Traders in the Bay o f Bengal: F ifteenth to N in eteen th
C en tu ries M ukherjee, R. and Subram aniam , L. ed., Politics and Trade in the
Indian Ocean World, New D elhi: O x fo rd U niversity Press, pp. 183—2 0 9 .
M ehta, R .N . ( l 97^—79) A n O ld Map o f G ujarat, Bhowm ik, S .K . ed., Reflections
on Indian Art and Culture, Vadodara, pp. 165—9*
M eulen, D . van der, and W issman, H . von (1932) Hadramaut; some of its mysteries
unveiled, L eid en : E.J. B rill.
M iller, J .I. (1969) The Spice Trade of the Roman Empire, O x fo rd .
M ills, J .V .G . (1970) The Overall Survey of the Ocean’s Shores, translation o f Ma
H u an ’ s Ying-Yai She-lan, C am bridge: Haklyut Society.
M inistry o f N ational H eritage and C u ltu re, O m an (1991) Oman: a seafaring nation
Sultanate o f O m an : M inistry o f N ational H eritage and C u ltu re.
M iqu el, A . (1996) G eography, Rashed, R ., ed., Encyclopedia of the History of Arabic
Science, L o n d o n and New Y ork: R outledge, vol. 3. PP* 796—8 l2 .
M ishra, U . ( l 997) Seafaring in the Context of the Material Culture of Early Historical Orissa
(3rd century B C -^th-^th centuries ad ): An Ethnohistorical Approach, M. Phil
dissertation, Jawaharlal N eh ru U niversity, New D elh i.
M itchin er, M. ( l 977—79) Oriental Coins and their Values. The Ancient & Classical World,
6 0 0 BC — AD 6 5 0 , L o n d o n : Hawkins Publications.
M oh am m ad, H ussain Baba b in (1969) Sejarah n egeri dan raja-raja Perlis,
Journal of the Malaysian Branch of the Royal Asiatic Society, 42, 2, pp. I 73—94 *
M o n fre id , H enry de (1961) Les secrets de la Mer Rouge. Paris: Le Livre de Poche,
2n d (?) ed.
M ookerji, R .K . (1912) Indian Shipping: A History of the Sea-borne Trade and Maritime
Activity o f the Indians from the Earliest Times, L o n d o n : Longm ans.
---- ( l 957) Indian Shipping, a History of the Sea-borne Trade and Maritime Activity of the
Indians from the Earliest Times, second edition , B om bay (originally published
1912): O r ie n t Longm ans.
M oo r, J .H ., com p. (1968) Notices of the Indian Archipelago and adjacent countries,
L o n d o n : Frank Cass, rep rin t o f 1837 ed.
M o rela n d , W .H . (1939) T h e Ships o f the A ra b ia n Sea about 150 0 AD, Journal

of the Royal Asiatic Society, Ja n u a ry-A p ril, pp. 6 3—74; I 73—^
192-
M o ritz, B. (1915) E in F irm an des S ultan s S elim fu r die V e n e z ia n e r vom
Jah re 15^7» Festschrift Eduard Sachau, ed. G . W eil (B e rlin : G e o r g R e in e r):
422 - 43 -
M orley, J .A .E . ( l 949 ) T h e A rabs and the Eastern Trade, Journal of the Malayan
Branch of the Royal Asiatic Society, 22, I , pp. I 43 ~ 76 -
M ukh erjee, B .N . (1990 ) K h arosth i and K h aro sth i-B ra h m i Inscriptions in
West Bengal (India) Indian Museum Bulletin, 25 -
Nagaswamy, R. (1991) A lagan akulam : A n In d o -R o m a n trad ing p ort, in
M argabandhu C . et al (eds). Indian Archaeological Heritage, K.V. Soundara Rayan
Felicitation Volume 1991, New D elh i, A gam K ala Prakashan, p p. 247“ 254 -
N aravane, M .S . (1998) R estoration, preservation & utilisation o f the coastal
forts o f India, La Hausse de Lalouviere ed., Coastal Fortifications, M auritius:
H eritage, pp. 126—131.
N eale, Fred. A. (1852) Narrative of a residence at the capital of the kingdom of Siam,
L o n d o n : O ffice o f the N ation al Illustrated Library.
N etton , I.R . (1996) Seek Knowledge: Thought and Travel in the House of Islam, L o n d o n :
C u rzo n Press.
N eum ayer, E. (1993) Lines on Stone, New D elhi: M anohar Publishers.
N icholas, C .W . (1990 ) Sinhalese Naval Power. Sri Lanka and the Silk Road of the
Sea, Bandaranayake, S ., Dewaraja, L ., Silva R, W im alaratne K .D .A . ed. S ri
Lanka N ation al C o m m issio n fo r U N E S C O and T h e C e n tra l C u ltu ral
Fund. C o lo m b o .
N icholls, C .S . (1971) The Swahili coast, politics, diplomacy and trade on the East African
littoral, 1*798—18 56 , N ew Y o rk: A frica n a P u blishin g C o rp o ra tio n .
N ico lle, D . (1989) S h ip p in g in Islam ic A rt: seventh th ro ugh sixteenth century
AD, The American Neptune, vol. 49 » n o .3.
Nilakantha Sastri, K .A . (1955) The Colas, second editio n , Madras: U niversity o f
M adras.
O chsenw ald, W illiam ( l 977) T h e Jid d a Massacre o f 1858, Middle Eastern Studies,
I 3 > 3 . PP- 314- 26 .
---- (1982) T h e com m ercial history o f the H ijaz Vilayet, 18 20 —190 8 , Arabian
Studies, 6, pp. 57—76-
---- (1984) Religion, society and the state in Arabia: the Hijaz under Ottoman control,
18 4 0 —1908, C olu m b u s: O h io U niversity Press.
O m an , Sultanate o f (1991) Oman: A Seafaring Nation, M uttrah: M inistry o f
N ational H eritage and C u ltu re.
O w en , J. (1997) D o A n ch o rs M ean S hips? — U nderw ater Evidence fo r
M aritim e T rade A lo n g the D h o fa r C oast, S o u th ern In d ian O cea n ,
A lessandra A van zin i (ed) Profumi DArabia, A tti del C on vegn o , Saggi di
Storia A n tica I I , L ’erm a di B retschn eider, p p. 351 - 363 -
Pande, B .M . (1986) A S hrin e to Siva: an u n fin ish ed house o f prayer in
B h ojp u r, The India Magazine, 6, p. 33 -
P ao-ch an g tr. (1981) Pi-chiu-ni-chuan. Lijung-hasi, Osaka: T okohai.
Paranavitana, S. ( l 953) T h e Shrine o f U pulvan at D evundara. Memoirs of the
Archaeological Survey of Ceylon, V o l. V I. D epartm ent o f A rchaeology. C o lo m b o .
---- (1970 ) Inscriptions of Cey/on. V o l. I. D epartm ent o f A rchaeolgy. C o lo m b o .
---- (1983), Unscriptions of Cey/on, vol. II, part I, Department o f Archaeolgy, C olom bo.
Parim artha, I. G de ( l 995) Perdagangan dan politik di Nusa Tenggara, 1815—1915,
D octo ral dissertation, V rije U niversiteit, A m sterdam .
Paris, E. (18 4 0 ? N o date o n first edition) Essai sur la construction navales des peuples
extra-europeens. Paris: A rthu s B ertrand.
Paris, F.E . (1841—3) Essai sur la construction navale des peuples extra-Europeens, Paris.
Parkin, D avid (1985) 'B ein g and selfhood am ongst in term ediary Sw ahili’ in J.
maw and D . Parkin Eds Swahili Language and Society Beitrage zur A frikan istik
23 >V ien n a: Institut fo r A frikan stik u n d A egyptologie der U niverstat W ein.
---- (I 99 1) Sacred Void: spatial images of work and ritual among the Giriama of Kenya
C am bridge: C am bridge U niversity Press.
Parry, J .H . (1974) The Discoveiy of the Sea, C alifo rn ia.
Patel, H .G . (1989) A lternative T echn ologies and S o cioeco n om ic contexts o f
adaptation: a study o f the coastal fish erm en o f Saurashtra in W estern India,
van der Leeuw, S .E . and T o rren ce, R. eds. What’s New? A Closer Look at the
Process o f Innovation, L o n d o n : U nw in H ym an, pp. 54—^2.
Pathm anathan, S. (1976) T h e Velaikkarar in M edieval South India and Sri
Lanka The Sri Lanka Journal of the Humanities, 197^, pp. 12 0 —137
---- (1984) T h e Nagaram o f the Nanadesis in Sri Lanka, circa AD I O O O — 130 0 ,
The Sri Lanka Journal o f the Humanities, 1984, pp. 122—163.
---- (1990) T h e bron ze seal o f the Nanadesis fro m H am bantota (Sri Lanka) in
Asian Panorama .- Essays in Asian History, Past and Present, edited by K .M .d e Silva, S.
K irib a m u n e and C .R .d e Silva, pp. 139—150.
---- (x994 ) T h e T am il slab in scriptio n o f the Virakkoti at Budum uttava,
Nikaweratiya u rban ization at Magala, The Sri Lanka Journal of the Humanities,
1994, pp. 15- 3 0 .
Pearson, M .N . ( l 994 ) Pious Passengers: the Hajj in Earlier Times, New D elhi: S terlin g
Publishers.
Peeters, J e ro e n ( l 994 ) Kaum Tuo Kaum Mudo; social-religieuze verandering in Palembang,
1821—1942, D octo ral dissertation, U niversity o f Leid en .
Pellegrini, G .B . (1978) Term in ologia m arinara di origine araba in italiano e
nelle lingue europee, in La Navigazione mediterranea nellalto medioevo (Settimane di
Studio nel Centro Italiano di Studi sullAlto Medioevo), Volum es I-II (Spoleto) II:
7 9 7 -8 4 1.
Pelliot, P. (190 3) Le F o u -n an , BEFEO, 1903, pp. 2 4 8 - 3 0 3 .
---- (190 4) D eux itineraires de C h in e en Inde a la fin du V ile siecle, BEFEO,
19 0 4 , pp. I 3I - 4 I 4 -
Perera, B.J. (1952) T h e Foreign Trade and C om m erce o f A n cie n t C eylon . I
T h e Ports o f A n cie n t C eylon , in The Cylon Historical Journal 1.2 C o lo m b o .
---- (1952) T h e fo reign Trade and com m erce o f ancient C eylon 4 parts, The
Ceylon Historical Journal, I, Jan. 1952, pp. 10 9—119, 192—2 0 4 , A p ril. 1952,
3 0 1 - 3 2 0 , July - August 1952, pp. 1 4 - 2 2 .
Periplus, see Sch off.

P h illip s-B irt, D ouglas ( l 979) The Building of Boats. W .W . N o rto n &. C o . Inc. New
Y o rk .
P icard, C . ( l 997) L ’Ocean atlantique musulman: De la conquete arabe a Vepoque almohade,
Paris: M aisonneuve & Larose.
Pieris, P.E . (1917 & J9! 9) N agadipa and B uddhist rem ains in Jaffna, Journal of
the Royal Asiatic Society (Ceylon Branch)., X X V I, 1917* pp. n —30; X X V III, 1919,
pp. 4 0 -6 6 .
Pliny (1942) The Naturalis Historia of Pliny, trans. H . Rackham, C am bridge, Mass:
Loeb Classical Series.
Pokrant, B. ( i 997) T h e Fishers and Fisheries Sector o f Bangladesh: an
an th ro p olo gical overview, Journal of Social Studies, 76, A p ril, pp. 32 —48.
Pokrant, B ., Reeves, P.J. and M cG u ire, J. (1998) T h e N ovelist’ s Image o f
South A sian Fishers: exp lo rin g the w ork o f M anik Bandopadhyaya, Advaita
Malla B arm an and T hakazhi Sivasankara Pillai, South Asia, 21, I, pp. I2 3 ~ 138.
Polo, M arco (1982) The Travels of Marco Polo, trans. R. Latham , New York: A baris.
Potts, D .T . (1990) The Arabian Gulf in Antiquity, O x fo rd : O x fo rd U niversity Press.
---- ( l 995 ) W atercraft o f the Low er Sea, F in kbeiner, U ., D ittm an n , R. and
H auptm ann, H. eds., Beitrage zur Kulturgeschichte Vorderasiens, M ainz: von
Zabern , p p - 559- 571-
---- ( l 997) Mesopotamian Civilization: The Material Foundations, London-. Athlone
Press.
Potts, D . T . , Parpola, A ., Parpola, S. and T idm arsh, J ., 1996, G u h lu and
G uggulu , Wiener Zgitschrift fuer die Kunde des Morgenlandes, 86, pp. 291—3 0 5 .
Pouwels, Randall (1987) Horn and crescent, cultural change and traditional Islam on the East
African coast, 8 0 0 —1 9 0 0 , C am bridge: C am bridge University Press.
Prakash, O m (1998) 'E u ro p ean C orp o rate Enterprises and the Politics o f
T rade in India, 16 0 0 —1 8 0 0 ’ , in Rudrangshu M ukherjee and Lakshm i
S ubram anian eds., Politics and Trade in the Indian Ocean World, Essays in Honour of
Ashin Das Gupta, D elhi: O x fo rd U niversity Press: 165—82.
Prem atilleke, P.L. (1990) C h in ese C eram ics discovered in Sri Lanka, Sri Lanka
and the Silk Road of the Sea, pp. 233—237 -
Prickett, M. (1980) Excavations at Mantai i g 80 : Preliminary Report of the Field Director,
C am bridge, Mass., ( i O , 17, 1980) m anuscript.
---- (1990 ) Sri Lan ka’ s foreign trade before AD 6 0 0 : archaeological evidence
in Asian Panorama: Essays in Asian History, Past and Present, edited by K .M . de Silva,
S. K irib a m u n e and C .R . de Silva, pp. 151—180.
P rickett-F ern an do, M . (1990 A ) D urable goods: T h e archaeological evidence
o f Sri Lanka’ s role in the Indian O cea n trade, Sri Lanka and the Silk Road of the
Sea, pp. 61—85.
---- (1990 B) M antai-M ahatittha: the great port and entrepot in Indian trade
in Sri Lanka and the Silk Road of the Sea, pp. 11 5—122.
Princep, J . (1836) 'N ote on the nautical instrum ents o f the A rab s’ ,reprinted
in F. Sezgin ed. Islamic Geography, Frankfurt: G oethe University, vol.15:
139- 51-
Prins A .H .J. (1965) Sailing from Lamu: a study of maritime culture in Islamic East Africa
Assen: van G orcum and Com pany..
---- (1970 ) A Swahili Nautical Dictionary, D ar es Salaam: C h u o cha U ch u n gu zi wa
Lugha ya K isw ahili.
---- (1986) A Handbook of Sewn Boats, L o n d o n : M aritim e M on ographs and
Reports n o. 59 -
Pryor, J .H . (1992) Geography, Technology and War: Studies in the Maritime History of the
Mediterranean, 6 4 9 —1571’ C am bridge: C am bridge U niversity Press.
Ptolem y, C lau diu s (1932) Geographike Huphegesis, trans. in to English by E .L .
Stevenson, N ew Y ork: N .Y . Public Library.
Purwanto, Bam bang (1992) From dusun to market: native rubber cultivation in southern
Sumatra, 1 8 9 0 —1 9 4 0 , PhD Thesis, U niversity o f L o n d o n .
Q aisar, A .J. (1982, paperback 199^) The Indian Response to European Technology and
Culture AD 1 4 9 8 —1707, New D elhi: O x fo rd U niversity Press.
---- (1987) F ro m Port to Port: L ife on In d ian Sh ips in the sixteenth
and seven teen th cen tu rie s, Das G u p ta , A . and P earson , M .N . e d s., India
and the Indian Ocean 1 5 0 0 —1 8 0 0 , C alcu tta: O x fo r d U n iversity Press,
PP- 331- 349 -
Q u tam i, Isa al- (1964) Kitab Dalil al-Mukhtarfi Tim al-Bihar, third reprin t, Kuwait:
G overn m en t Press.
Rajam anickam , G .V . and A ru lra j, V .S . ( l 994 ) Maritime History of South India,
T hanjavur: T am il U niversity.
Rajam anickam , G .V . and Subbarayalu, Y . (1988) History of Traditional Navigation,
Thanjavur: T am il U niversity.
Rajam anickam , G .V . et al., (1991) Maritime History of South India, vols. 5>
unp u blished reports, subm itted to the N ational Institute o f Science,
T ech n o logy and D evelop m en t Studies, New D elhi.
Rao, S .R . (1970 ) Sh ip p in g in A n cie n t India, in Lokesh C han dra et al eds.,
India’s Contributions to World thought and Culture (Swami Vivekananda Commemoration
Volume) Madras: 83—107*
Rao, S.R . (1985) Lothal: A Harappan Port Town 1955—62, M em oirs o f the
Archaeological Survey o f India, j 8 (i).
---- (1987) Progress and Prospects of Marine Archaeology in India. N ational Institute o f
O cean ograp h y. G oa.
---- (ed) (1988) Marine Archaeology of Indian Ocean Countries, G oa: N ational Institute
o f O cean ograp h y.
---- (1990 ) Excavation o f the legendary city o f Dwarka in the A rab ian Sea, in
Marine Archaeology, G oa, pp. 59“ 9&*
Ratnagar, Shereen (1981) Encounters: The Westerly Trade of the Harappans, D elhi:
O x fo rd U niverstiy Press.
---- ( l 994 ) 'H arap p an T rade in Its W orld C o n te x t’ , Man and Environment, XIX:
117 - 2 7 .
Ratnayake, H . (1990) T h e Jetavana Treasure, Sri Lanka and the Silk Road of the Sea.
Raut, L .N . &, T rip ati, S. (1993) T rad itio n al Boat B u ild in g C entres arou n d
C hilika Lake o f O rissa, in Journal of Marine Archaeolgy. 4- G oa. pp. 5I—55-
Ray, H .P . (1986) Monastery and Guild: Commerce under the Satavahanas, New D elhi:
Oxford. U niversity Press.
---- (1989) Early H istorical Trade an Overview, Indian Economic and Social History
Review, v o l u m e X X VI, n o 4 > PP - 437“ 57-
---- ( l 994 ) The Winds o f Change: Buddhism and the Maritime Links o f Early South Asia, New
D elhi: O x fo rd U niversity Press.
---- ( l 995 , first pu blish ed 1988) T h e Yavana Presence in A n cie n t India,
Boussac, M .-F . and Salles, J .-F . ed., Athens, Aden, Arikamedu, N ew D elhi:
M anohar Publishers, pp. 75—95 -
---- (1996) M aritim e A rchaeology: T h e Ethn ographic Evidence, Man and
Environment, vol. 21 (i) pp. 74~ 85.
---- (1998, first published 1994) The Winds o f Change: Buddhism and the Maritime Links
o f Early South Asia, New D elhi: O x fo rd India Paperbacks.
---- (ed) (1999) Archaeology o f Seafaring: The Indian Ocean in the Ancient Period, New
D elhi: Ind ian C o u n c il o f H istorical Research.
Ray, H .P . and J .- F . Salles, eds., (1996) Tradition and Archaeology: Early Maritime
Contacts in the Indian Ocean, New D elhi: M an ohar.
Ray, H .P . and Salles, J .- F . (1999) Communautes maritimes de YOcean indien,
T u rn h ou t: B repols Publishers.
Ray, H araprasad (1987) C h in a and the W estern O cea n in the Fifteenth
C en tury, in Satish C h an d ra ed., The Indian Ocean, Explorations in History, Commerce
and Politics, D elhi: Sage: 109 24 -
Recueil des historiens des croisades. Historiens orientaux (1967), V olum es I-IV , Paris:
Im p rim erie N ationale.
Redknap, M . (1984) Cattewater Wreck. B A R 131. O x fo rd .
Reeves, P.J., M cG u ire, J ., Pokrant, B. and Pope, A . (1996) T h e K o li and the
B ritish at Bom bay: the structure o f th eir relations to the m id -n in etee n th
century, Pearson, M .N . ed., South Asia, 19, Special Issue: A sian and E u rope:
C o m m erce C o lo n ialism and C ultures, pp. 97—119-
R e id , A n th o n y (1972) H abib A b d u rra h m a n az-Z ah ir, 1833—^ 9 6 , Indonesia,
13* PP- 37- 6 0 .
R eith, E. (1993) Voiliers et Pirogues du Monde au debut du XIXe siecle, Paris: D u May.
R enault, Francois (1976) Liberation d ’esclaves et nouvelle servitude, les rachats des captifs
africains pour le compte des colonies franâises apres Vabolition de 1’esclavage, A b id jan : Les
N ouvelles Ed itions A fricain es.
Rieth, E. (1996) Le Maitre-Gabarit, La Tablette et Le Trebuchet. Paris C .T .H .S .
R om anis, Federico de (1997) 'R om e and the N otia o f India: R elations betw een
R om e and S ou th ern India fro m 30 BC to the Flavian P e rio d ’ , in R om anis,
Federico de and T ch ernia. A eds. Crossings, Early Mediterranean Contacts with India,
D elhi: M alles, M anohar: 8 0 —160.
R om anis, Federico de and T ch ernia. A eds. ( l 997) Crossings, Early Mediterranean
Contacts with India, D elhi: M anohar.
R o uggeulle, A . (1996) M edieval trade networks in the eastern In d ian O cea n
(8—14 centuries): Som e reflectio n s fro m the d istribu tion pattern of
C h in ese im ports in the Islamic W o r ld , in Tradition and Archaeology: Early Maritime
Contacts in the Indian Ocean, eds. H . Prabha Ray & J.F . Salles (New D elhi:
M an ohar): 159“ 180.
Salim , A .I. (1973) Swahili-speaking peoples o f Kenya’s coast, 189 5—196$, N airo bi: East
A frica n P u blishin g H ouse.
Salles, J.F . (1987) 'T h e A rabian Persian G u lf u n d er the Seleucids’ , in A .
K u h rt and S. Sherw in W hite eds. Hellenism in the East: 75—184*
Salles J.F . (1996) H ellenstic Seafaring in the Indian O cean , a Perspective fro m
A rabia, in H im anshu Prabha Ray and J.F . Salles ed., Tradition and Archaeology,
Early Maritime Contacts in the Indian Ocean, D elhi: M anohar: 293“ 31 0 -
Salom on , R. (1991) Epigraphic Rem ains o f Ind ian T raders in Egypt, Journal o f
the American Oriental Society: 73I—36-
Sarkar, H .B . (1969) South India in O ld Javanese and Sanskrit Inscriptions,
Bijdragen tot de Taal-, Land- en Volkenkunde, 125* PP- x93—20 6.
Sarsfield, J. (1984) M editerranean whole m oulding. Mariners’ M irror J O : 8 6 —88.
---- (1988) Survival o f p r e -l6 th century M ed iterran ean lo ftin g techniques in
Bahia, B razil. Pp. 63—86, in O .L . Filgueras ed. Local Boats. B A R S. 438 -
O x fo rd .
Sassoon, H . (1982) T h e sinking o f the Santo Antonio de Tanna in M om basa
H arb o u r, Paideuma, 28, pp. I O I — 108.
Schafer, E .H . (1963) The Golden Peaches o f Samarkand. A Study o f Tang Exotics,
Berkeley: U niversity o f C a lifo rn ia Press.
S ch lin gloff, D . (1982) Indische Seefahrt in rom isch er Zeit, Q ir geschichtlichen
Bedeutung der friihen Seefahrt, M iinchen: V erlag C .H . Beck, 5 1 -8 5 .
---- (1988) Ships and Seafaring, Studies in Ajanta Paintings, D elhi: A janta
Publications, pp. I9 5 —2 l8 .
---- (1988) Ajanta Paintings, D elhi: Books and Books.
Sch off, W .H . (1912) The Periplus of the Erythraean Sea: Travel and Trade in the Indian
Ocean by a Merchant o f the First Century, translated fro m the G reek and annotated
by W .H . S ch off, second editio n , New Y o rk, 1912.
Sch roeder, A . (1905) Dai Nam Hoa Te Do Luc. Annam. Etudes Numismatiques, Paris.
Sch urh am m er, G. (1977) Francis Xavier, his life and times. V o lu m e 2 (India
I 54 I~ I 545)- Jesuit H istorical Institute. R om e.
Scott, W .H . 1981. Boatbuilding and Seamanship in Classic Philippine Society. (A n th ro
p ological Papers N o . 9 -) M anila, Philippines: N ational M useum .
Senart, E. (19 0 5 —6) T h e Inscriptions in the Caves at Nasik, Epigraphia Indica, 8,
PP- 59- 96 .
Severin, T im (1982) The Sindbad Voyage, H u tchin son, L o n d o n .
S h e p h e r d , G ill (19 8 0 ) T h e C o m o r ia n s an d the East A fr ic a n slave trad e,
in Jam es L . W atson , e d ., Asian and African systems o f slavery, O x fo r d : B asil
B lack w ell.
S h ep herd , G ill (1985) T ra d in g lineages in historical perspective, in Jeffrey C .
Stone, ed., Africa and the sea, proceedings of a colloquium at the University o f Aberdeen,
pp. 152—77’ A b erd een : A b erd een U niversity A frica n Studies G ro u p .
S herry, N o rm a n (1966) Conrad’s eastern world, C a m b rid g e : C a m b rid g e
U niversity Press.
Sidebotham S .E . and W endrich, W .Z . ed., (1996) Berenike 1995, Leiden :

Research S ch oo l C N W S.
Sidebotham , Steven G. (1992) Ports o f the Red Sea and the A rab ia-In dia
T rad e ’ , in Vim ala Begley and Richard D aniel de Pum a eds., Rome and India:
The Ancient Sea Trade, D elhi: O xford University Press (Indian edition ): 12—38.
Sim s-W illiam s, N . and C rib b , J. (1996) A new Bactrian in scrip tio n o f
K anishka the G reat, Silk Road Art and Archaeology 4 > PP- 75—!42-
Sircar, D .C . ( l 953) 'T h e C h a rter o f V ishnusena — AD 592 ’ , Epigraphia Indica,
X X X; 16 3 -8 1 .
---- (1965) Select Inscriptions Bearing on Indian Histoiy and Civilization, I, Calcutta:
University o f Calcutta.
---- (1983) Select Inscriptions Bearing on Indian History and Civilization, II, New D elhi:
M unshiram M anoharlal.
Sm ith, M. Estellie ( l 977) Those Who Live from the Sea: A Study in Maritime Anthropology,
New York: West Publishing Com pany.
S on g O n g Siang (1923) One hundredyears' history o f the Chinese in Singapore, L o n d o n :
Jo h n Murray.
Sopher, D .E . ( l 977) The Sea Nomads, Sin gapore: N ational M useum .
Spen cer, G eo rge (1983) The Politics o f Expansion—The Cola Conquests o f Sri Lanka and
Srivijaya, Madras: G eetha Publications.
Spencer, G eorge W. ( l 977) T h e Politics o f Plunder: T h e Colas in Eleventh
Century Ceylon’ , Journal o f Asian Studies, XXVI: 4 ° 5 —2 0 .
Stern, S.M . (1967) 'Ram isht o f Siraf: A M erchant M illionnaire o f S ir a f,
Journal o f the Royal Asiatic Society o f Great Britain and Ireland : IO—15-
Stillm an, N orm an A. ( l 973) T h e Eleventh Century M erchant H ouse o f Ibn
Awkal (A Geniza Study)’ , Journal o f the Economic and Social History o f the Orient,
XV I: 1 5 -8 8 .
Stronach, D. (1989) T h e Royal G arden at Pasargadae: Evolution and Legacy in
Archaeolia Iranica et Orientalis: Miscellanea in Honorem Louis Vanden Berghe, vol. I, ed. by
L. De Meyer et E. H aerinck, G hent: Peeters.
---- (1990) T h e G arden as a political statement: Som e case studies from the
N ear East in the first m illennium BC, Bulletin of the Asia Institute, 4* PP- 1 7 1 — 1 8 0 .
---- (1994) Parterres and stone w atercourses at Pasargadae: notes on the
A ch aem en id c o n trib u tio n to garden design, Journal o f Garden History, pp. 3—I 2-
Subrahm anyam , R. (1964) Salihundam, a Buddhist Site in Andhra Pradesh, H yderabad.
Subram anian, Lakshm i (1996) Indigenous capital and imperial expansion; Bombay, Surat
and the West coast, D elhi: O x fo rd U niversity Press.
Sundaresh. (1993) T radition al Boat Building Centres o f K arnataka C oast A
Special Reference to H onavar, Bhatkal and Gangolly, in Journal o f Marine
Archaeology. 4 * G oa. pp. 29—35 -
Suresh, S. (1993) A Study o f the Roman Coins and Other Antiquities in India, with Special
reference to South India, (unpublished thesis) Jaw aharlal N ehru University, New
Delhi.
Tam poe, M. (1989) Maritime Trade between China and the West, BA R International
Series 555.
T ch ernia, A . ( l 999 ) A rikam edu et le G raffito Naval d ’Alagankulam . Topoi 8.

T hakker, P .S ., Raval, M .H . and Dasgnpta, A .R . ( l 993^ Study o f Ancient Ports o f
Gujarat using Remote Sensing Data, Paper presented at the International W orkshop
on 'A p p lic a tio n o f R em o te S en sin g fo r A rc h a e o lo g y and R elated
D iscip lin es’ , D ecem b er 17—19*
T hap ar, R om ila (1987) E pigraphic Evidence and Som e In d o -H elle n istic
Contacts d u rin g the M auryan P erio d , in S .K . Maity and U p en d ra T h a k u r
eds., Indological Studies, Professor D.C. Sircar Commemoration Volume, D elhi: A gam
K ala Prakashan: 15—19
---- (1992) Black G old : South Asia and the R om an M aritim e T ra d e ’ , South
Asia, XV : 1—28.
S ch off, W .H . (trans) (1912) The Periplus o f the Erythraean Sea, trans. New York:
Longm ans, G re en and C o .
T heoph rastus (1916) Enquiry into Plants, trans. A . H o rt, V olum es I —II, L o n d o n -
New Y ork: W illiam H e in e m a n n -G .P . Putnam .
T hesiger, W . (1967) The Marsh Arabs, L o n d o n : Penguin.
T h ie rry , Fr. (1986) Les collections monetaires. Monnaies d’Extreme Orient, Paris:
A d m in istratio n des m onnaies et m jdailles.
T h ie rry , Fr. (1992) D u b o n usage des trouvailles m onetaires chinoises, South
Asian Archaeology, Proceedings o f the Third Conference o f the European Association of
Southeast Asian Archaeologists, I 99 °> ed- by I. G lover, H u ll, C en tre fo r S o u th
East A sian Studies, pp. 29—43
---- ( 1995) Les m onnaies chinoises du M usee d ’A nuradh ap u ra provenant du
site de Yaphuwa, 52, 1995* 43 I—437 -
Thivakaran, G .A . and Rajam anickam G .V . (1992) Catam aran o f South India:
A marvel o f T rad itio n al T echn ology, in Journal o f Marine Archaeology. 3. G oa.
pp. 17- 3 0 .
T ibbetts, G .R . (1961) A rab Navigation in the Red Sea, Geographical Journal, 127,
pp. 3 2 2 - 3 4
---- (1971) Arab Navigation, L o n d o n : Luzac.
---- (1971) Arab Navigation in the Indian Ocean before the Portuguese, L o n d o n : T h e Royal
Asiatic Society o f G reat B ritain and Ireland.
Tibbets, G .R . ed. (1981) Trans, o f A h m ed b. M ajid a l-N a jd i’ s treatise o n Arab
Navigation in the Indian Ocean before the coming o f the Portuguese. Royal Asiatic Society
o f G reat B ritain and Irelan d, L o n d o n .
T oled an o , E h ud (1982) The Ottoman slave trade and its suppression, Princeton:
Prin ceto n U niversity Press.
Tolm acheva, M . ( l 995—6) 'Essays in Swahili G eograp hical T h o u g h t’ , Swahili
Forum I I & III.
Tossi, M . (1992) 'T h e H arappan C ivilization beyon d the Ind ian S u b co n ti
n e n t’ , in G regory L. Possehl ed., Harappan Culture, a Contemporary Perspective,
second edition , New D elhi, O x fo rd and IBH : 36 5—78-
Toussaint, A . (1966) History o f the Indian Ocean R outledge & K egan Paul. L o n d o n .
T rocki, C arl A . ( l 979) Prince o f pirates: the Temenggongs and the development o f Johor and
Singapore, 1 7 8 4 —1885. Singapore: Singapore U niversity Press.
U dovitch, L. (1970 ) 'C o m m e rcia l T echn iqu es in Early M edieval Islamic

T ra d e ’ , in D .S . Richards ed., Islam and the Trade o f Asia, Pennsylvania:
U niversity o f Pennsylvania: 37—62.
U n ger, R. W . (1980) The Ship in the Medieval Economy 6 0 0 —1 6 0 0 , L o n d o n -
M ontreal: C ro o m H elm & M cG ill - Q u e e n ’ s U niversity.
Varadarajan, L. (1998) Sewn Boats of Lakshadweep, G oa: N ational Institute o f
O cean ograp h y.
V erlin d en , C harles (1987) 'T h e Indian O cean : A n cie n t Period and the M iddle
A g e s,’ in Satish C h an d ra ed., Indian Ocean -.Explorations in History, Commerce and
Politics, D elhi: Sage: 27—53 -
V illiers, A .J (1940 ) Sons of Sinbad, New Y o rk: C harles S crib n e r’ s.
---- (1940 ) Sons o f Sindbad: an account o f sailing with the Arabs in their dhows, Lo n d o n :
H o d d e r & Stoughton.
---- ( l 954 ) Passage in a Red Sea D how, The M ariner’s M irror, vol. 4 °> no- 3>
pp. 17 1 -18 2 .
V itharana, V . (1992) T h e O R U and the Y A T H R A : T rad itio n al O u trig ge r
C raft o f Sri Lanka. C o lo m b o . Privately published.
V osm er, T om . ( l 993 ) Boat ethnography: the yathra d h on i o f Sri Lanka, in
Maritime Archaeology in Sri Lanka: The Galle Harbour Project, 1992. G reen , J. and
D evendra, S. ed. A rch aeolo gical D epartm ent o f Sri Lanka, T h e C en tra l
C u ltu ral F un d , W estern A ustralian M aritim e M useum and Post G raduate
Institute o f A rchaeology. C o lo m b o .
---- ( l 995 ) M aritim e G raffiti in O m an , The Indian Ocean Review, T h e Indian
O cea n C en tre fo r Peace Studies, N edlands, vol. 7, n o. 4 > PP- I 3—23 -
---- ( l 999 ) M aritim e A rchaeology, E thn ography and H istory in the Indian
O cean : an E m ergin g Partnership, Ray, H .P . ed. Archaeology o f Seafaring: The
Indian Ocean in the Ancient Period, Indian C o u n cil o f H istorical Research,
pp. 265- 279 -
V osm er, T . , A gius, D ., Baker P. and C arp en ter, J. (1998) Field Report, Oman
1999 ’ W estern A ustralian M aritim e M useum R eport N o 144’ Perth.
Wake, C . ( l 997) T h e G reat O c e a n -g o in g ships o f Sou th ern C h in a in the A ge
o f C hin ese M aritim e V oyaging to India, Tw elfth to F ifteenth C en tu ries,
International Journal of Maritime History, 9, 2, pp. 5I—81.
W arm ington, E .H . (1928 —1974) The Commerce between the Roman Empire and India,
L o n d o n : Luzac. (first published 1928).
W eerakkody, D .P .M . ( l 997) Taprobane. Ancient Sri Lanka as known to Greeks and Romans,
Indicopleustoi. Archaeologies o f the Indian Ocean, vol. I, B repols, B elgium : T u rn h o u t.
W eibrot, K u rt 1958- The Crew as a Social System. O slo: N orsk Sjofartsm useum .
Weisshaar, H .-J . & W ijeyapala, W ., (1993) A n cie n t Ruhuna (Sri Lanka). T h e
Tissam aharam a Project: Excavations at A kurugoda 1992—1993’ Betrage zur
Allgemeinen und Vergleichenden Archaologie, 13, 1993, pp. I 27~ 166.
Weisshaar, H .-J . (1992) M egalithgraber u nter K o kosp alm en , Archaologie in
Deutschland, Jul. Sept. 1992, pp. IO—15-
West Coast o f India Pilot, ( i 975) H ydrograph er o f the Navy. Eleventh ed itio n .
B lue B ibliography.
W esterdahl, C h rister ( l 994 ) 'M aritim e C ultures and Ship Types: B rie f
C om m en ts on the S ignificance o f M aritim e A rch a eolo gy’ , The International
Journal o f Nautical Archaeology, 23 -4 : 2 6 5 —7° •
W heeler, R .E .M . and Deva, K rish n a (1946) 'A rikam edu: A n In d o -R o m a n
T ra d in g Station on the East Coast o f In d ia ’ , Ancient India, II: 17—124-
W hitcom b, D on ald and Janet Jo h n son (1982) Quseir al-Qadim 1980. Preliminary
Report. American Research Centre in Egypt Report 7- M alibu: U n den a.
W hitehouse, D . (1996) Sasanian M aritim e Activity, Reade, J ., ed., The Indian
Ocean in Antiquity, L o n d o n : K egan Paul International.
W iebeck, E. (1987) Indische Boote und Schiffe, Rostock: V .E .B . H in sto rff Verlag.
W ijesekara, N . (1952) H istorical backgroun d o f S in h alese-fo reign relations
fro m earliest tim es to the twelfth century, The Ceylon Historical Journal, 1952*
pp. I 72- I 9 I-
W ilson, N .F .J. (1909) The Native Crajt, Bom bay: Bom bay Port Trust.
W o o d fo rd , C .M . 190 9 . T h e C an oes o f the B ritish S o lo m o n Islands. Journal of
the Royal Anthropological Institute 39 : 2: 5° 6—
W right, A r n o ld , and Cartw right, H .A . (1908) Twentieth century impressions of British
Malaya, L o n d o n : Lloyds G reater B ritain, u nabridged ed.
Yaqut (1866—73) Kitab mu jam al-buldan, ed. F. W ustenfeld, V olum es I-V I,
Leipzig: F. A . Brockhaus.
Y u le , H . trans. and ed. (1866) Cathay and the Way Thither, V olum es I —I I , L o n d o n :
Hakluyt Society.
Y u le , H and A . C . B u rn ell ( l 994 )> Hobson-Jobson: A Glossay of Colloquial Anglo-
Indian Words and Phrases, ed. W. C ro oke, Sittingbourne.- Linguasia.
Index
24—Parganas 4 history o f 8 , 45
ships 4 5 , 155
Abbasid caliphate 180 Arabic and Persian texts 33 , 47 * 51* 54
A bu Inan Faris 174 see also Rihla, the
A b u ’l Fadhl 2 10 , 211 Arabian Sea 31* 33
A bu 1-M ahasin 18 7 Arasaratnam, S. 22
adab 175 archaeological
A den 2 4 1, 248 evidence ch. 4; excavations and
invasion o f 4 6 , 5 3 ; port o f 183 sites 7- 8 , 15 , 3 6 , 3 7 , 8 3 , ch. 4;
adze 2 6 6 , 2 6 7 , 2 7 0 , 271, 2 7 3 , 3 0 8 , remains 56 , see also faunal
310 see also boat-b uild ing remains
tools archaeology
Afghanistan 108—109 maritime/m arine 14 , 2 9 , 32—3 3 *
agriculturalists 74 see Qlso farmers 35- 36 , ch. 4.
ahawra 18 0 , 190 ethno-archaeology o f fishing 18 , 20
Akananuru 82 —83 archaris 139
Alahana Parivena 94 architecture, Sinhalese III
A lc is c o A lc in a , SJ 3 1 3 Arikam edu 5 6 , 6 7 , 8 3 , 98
almadia 198 Arjanta 86
Alwis, N A .W . A rthu r 159 artisans 139—140
Amarakosa 42 astrolabe 219
Angavijja 37 astronom ical and navigational
Anguttara Nikctya 86 instrum ents 215—21 see also
A nuradhapura 94—95 * I0 2 , 10 5 , 115 latitudes, longitudes
aphaterion 4 1 Austronesian speakers 10 , 12
apprentice/assistant 2 6 4 , 267—8 , Aydhab, port o f 45 * 47 * 54
2 6 9 - 7 0 , 276 axes 3 ° 8 , 310 see also boat-build ing
apprenticeship 141, 2 6 3 , 276 tools
Arab seafaring azimuths 16, 85
347
348 In d e x
babul tree j S boat nom adism 19

Babylonian texts 6 boat technology see also sh ip-build ing
baghla 7 2 , 2 2 8 , 2 4 7 * 2 4 8 , 2 4 9 - 253 technology
B ajuni people 2 6 2 , 268 changes in 18 , 80—81 see also steam
Bakari 2 6 5 , 274 navigation
Baladhuri 8 boats see also ships, vessels
Baluchistan coast 19 as distinct from ships 6 4 , 6 6 ;
Bangladesh 17 European classification o f
Barbosa, D. 4 6 —4 7 * 71* 75 70—7 3 ; evolution o f 7 3 ,
Barker, R. 289 I 35~ I 39 * I 49“ I 5I ; ownership o f
Barygaza, port o f 37 1 2 , 2 1 , 25 - 2 6 , 51- 5 7 , 73- 7 5 .
Basra, com m ercial centre o f 182 230 —2 3 2 ; types o f bero 3 0 5 ;
bas-reliefs 136 dugout IO, II see also vallam;
Batans, the 294 fishing 71, 80—81, 8 8 ; Kenyan
Batavia 231—32 see K en ya; Lamalera boats 2 9 9 ,
bazaars 182 3 0 3—13 ; launches or pleasure
bead making industry 9 9 —IOO 18 0 , 192—9 3 ; Likey A Tatala
beads 9 4 , 9 9 - I O O , I O 7 - I O 9 30 2—3 0 3 ; log raft IO ; m otor
Beckingham, C .F . 1 8 8 , 1 8 9 2 4 2 , 254 see Qlso vessels,
Begley, V . 9 8 m echanised; Muscat 2 4 2 ; ngalawa
Bellabarba, S. 2 7 9 , 289 (outrigger) 2 0 1; Pontian IO ;
Bellec, F. 1 3 8 reed 6 , J ; sapa 3 ° 5 ; sewn craft
Bengal 3 > 4 * IO ; Solom on Island (mon) 307 ;
Bay o f 24 . 31. 33> 50 Swahili 260 —6 2 ; tena 3 0 5 ; Yam i
Bengali poems 3 boats see Yam i boats and boat
Besse, A n ton in 248 building
Beypore 72 Bodhisattva Avalokitesvara, cult o f 86
Bharuch, port o f 68 Bombay, port o f 68
b in A llan, Nahray see Nahray borers 310 see also b oat-b uild in g tools
bin Hasan bin Bundar, M admun see B orobu du r 45
M admun B orivli caves 43
bin Isaac bin Bundar, K h alaf see Botel Tobago 29I —93 see Qlso Lanyu
K h alaf B ouchon, G . and Lom bard, D. 32
bin Jacob, Mahruz see M ahruz bourn 2 6 0 , 2 6 3 , 2 6 6 , 277
bin Mwinyi Bakari, M toro see Bakari bow drill 266 see also b oat-b uild ing
bin Perahya Ben Y iju , Abraham 63 tools
birds, in navigation 7 9 » 215 Bowrey, T . 2 1, J l
Blake, G . 246 brace and bit 310 see also b oat-b uild in g
boat-builders and craftsmen ch. 9 see tools
also ship-builders Brahm anical bhakti cults 56
b oat-b uild in g yards/centres 12 , Brahmanism 86
21- 2 2, 25, 71- 72, 75* 259- 60, B rah m i-K harosti script 37
2 6 2 , 2 6 4 , 277 Braudel, F. xiv, 2 2 , 29
b o at-b uild in g tools 9 , 2 6 6 , 3 0 8 , 3 10 , Brazil 289
312 Brhatkathaslokasangraha j 8
b o a t - b u i l d i n g t r a d i t i o n s I O —I I , British India Steam Navigation
80- 81 company (BISN) 24 ° > 2 4 6 , 257
association with ethnicity 72—7 3 > Buddhagupta 49—5°
2 7 6 ; diversity o f 87 Buddhasvami j 8
In d e x 349
Buddhism 3 , 85—87 charts see nautical charts
introduction to Sri Lanka 10 3 ; Chaudhuri, M. 2 , 22
notion o f saviour in 86—87 ; Chaul, port o f 68
spread o f to Southeast Asia 5 6 ; 'Chaw Lay’ 19—20
support to trade j 6 ;jaksis C heng Ho 5 8 , 14 1, 217 see also Zheng
(demonesses) in 86 He
Buddhist C hina 8- 9, 104, 114- 17, 192, 223 see
caves 8 3 , 8 6 ; literature 69—7 0 , 7 4 , also C hina ships, C h in ese, Marco
7 8 , 8 4 , 8 6 ; monasteries 2 5 , Polo
85- 8 7 , 9 4 ; monks 8 6 , 114- 15, C hina ships (markab sini) 8- 9 , 72
13 4 ; m onum ents j 6 ; stupa 10 2 , chine strakes 136
10 4 , 108 chinedkulan/Chinedkeran 2 9 3 , 2 9 7 , 2 9 9 ,
bullae, Sasanian II O see also seals/ 302- 304
sealings Chinese
bum I I , 7 2 , 8 0 , 1 6 6 , 2 0 1 - 2 0 2 , 2 6 0 attitudes to the sea 4 5 ; ceramics
Buzurg see ibn Shahriyar, Buzurg 115; dynasties 9 , 4 5 , 115, I l 6 ,
18 7; mariners 16 ; mines 116 ;
C ab ral’ s voyage 219 pilgrim Fahsien 49 , 8 6 ; silk 10 4 ,
calendars, maritime 212—14 114, I l 6 , 18 7; texts 9 4 , 15 3 ;
Calicut 4 6 , 4 9 , 7 5 , 18 4 , 18 6, 242 vessels 186—8 8 , 200
Cambay 46 chisels 310 see also boat-b uild ing tools
Cantino map 219 Christianity see N estorian Christians
captains 210—I I , 221—23 , 230 , 249 see chronom eter 219
also navigators and pilots Chryse C hora and Chryse Chersonesis
cargo 13 , 5 3 , 254 see com m odities, 41
trade goods 6 , 9 , 2 6 , 5 5 ’ ^2; Chulias 24—25
ownership o f 2 5 ; shift to bulk C od rin gton , H.W. 102, IO5- IO 7, IIO
8 l ; vessels see vessels coffee 2 3 9 , 2 4 3 , 2 4 6 , 2 5 1 see also trade
Carswell, J. IIO coir (coconut fibre) 6 2 , 7 2 , 13 4 , 248
cartels, shipping 2 4 1 see also sh ip-building materials
Casal 83 coins 3 , 5 , 7 , 7 6 , 9 4 , 99- 10 2 , 10 6 ,
Casson, L. 7 , 4 1, 49 , ^ 6 10 9 , IIO , 114, Il6 see also ships,
caste I39—4 0 , 247 see Qlso ritual status, representations
status punch-m arked copper 7 , 105—10 6 ;
caves 4 3 , 8 3 , 86 multi type 10 2 ; silver and
cerem ony 2 6 7 , 3^0 see also custom s, bronze 106
prayers, p ro h ib itio n s, C ola conquest 58
p ro tection , ritual/rituals, colandiaphonta 41
superstition C om oros, the 255- 57
Celebes Developm ent Com pany com m odities 5 5 , J82 see also cargo
(Cedeco) 242 price and movement o f 70
Celebes Kustvaart Maatschappij com m unities, maritime 17—2 0 , ch. 3
(Cekum ij) 242 see also fishing com m unities
ceramics 9 4 , 9 9 - I O O , 1 0 5 , I I O , 1 1 5 , com m onalities between 8 l—8 2 ;
1 8 3 , 1 8 7 see also pottery, cultural adaptation o f 18 ; impact
potsherds o f steam navigation on 66 see also
chahar bagh’ gardens I I I —13 steam navigation;
Chaliyar river 72 marginalisation o f 8 7 ; vis-'a-vis
Ghandraketugarh 37_ 4 ° , 51, 77 agricultural settlements 74
350 In d e x
compass 2I5~ 16 , 224 dictionaries or lexical texts

contraband 2 3 2 , 2 3 5 * 252 see also Sanskrit 4 2 , 4 7"~4 8 ; Classical
h e ro in , firearm s Arabic 198, 2 0 4 ; twentieth
C orrea, Gaspar 23 century 222
cotton 13 , 2 9 * 115 see °k° ship-building diffusion see cultural diffusion and
materials com m on origins
country trade’ 228 D ilm un (the Bahrain archipelago) 63
Cowasjee Dinshaw company 248 Dinewari 8
crew, recruitm ent o f Dionysius (second century A D ) 93—94
ethnicity 2 3 0 , 2 3 2 , 2 3 6 , 2 4 9 - 5 0 , Dioscurides or Socotra, island o f 84
25 2- 5 3 , 2 3 2 , 236 literacy 8 l , disakaka (sea-crows) 79 see Qlso birds
222 Divyavadana 86
C rib b, J. Il6 docks 2 3 , 68
cross-staff 2 l8 dovetail join ts 2 8 7, 289 see also ship
Culavamsa 142 building technology
cults 5 6 , 86 drachm, Indian-standard 106
cultural adaptation 18 D unn, R. 175 , 203
cultural contact 13 Dutch colonies 231
cultural diffusion and com m on dvipantara voyages 42
origins I I , 7 3 , 293- 94 Dwarka 14 , 35
cultural exchanges 55—5^ al Dyyb (Diu) 45 * ^2
customs 2 71* 273—7 4 * 275
customary code o f practice, boat Early Historical period 76
buildin g 2 6 3 , 276 sites 8 3 , 10 3 ; representations o f
customary gift 2 7 3 * 274 boats in art and architecture 76
Edye, J. 13 0 , 166
damila 25 elephants 5 5 * 9 2 , 142 see also trade, in
Das Gupta, A shin 21 live animals
Dasakumaracarita o f D andin 43 engineering skills 81
dates 2 4 3 * 2 4 6 , 2 4 8 , 254 see also engines 260 —61
trade, in staples/bulk items and epigraphical evidence I O I , 1 0 5 , 1 0 8 see
foodstuffs also inscriptions/inscribed
Davidson, J.W . 2 9 2 , 312 legends
Deccan, the 82 Eratosthenes (third century B C ) 93
decolonialistion, role o f n on - ethnic identity 260
Europeans 30 ethnicity 2 0 , 2 5 , 7 2 , 8 4 , 8 7, 276 see
D e lo c h e , J . 3 , 3 9 , 4 3 , 18 9 also crew, recruitm ent o f
Desinamamala o f Hemachandra 48 ethnographic studies
Dharmasastric literature 25* 34 see Qlso contributions o f I O — I I , 35 ; o f
prohibitions maritime com m unities 17—2 0 ;
dhow 4 5 , 8 1, 195 , 20 2- 2 0 3 , 2 14 , 2 2 0 , o f traditional ship-building
223, 244- 45 , 2 4 6 , 247- 48 , practices 35
251- 54 . 2 5 6 , 261 European dom ination 2 9 , 75* 219
technology 1 4 3 —4 4 ? measure speed exports see cargo, com m odities
o f 218
diasporas 5 6 , 2 2 7 * 2 4 8 , 258 see also Fahsien (Fa Hsien) 49 * 86
trading diasporas, Hadhram i Fait A llam ’ 236
diaspora farm ers 263 see also agriculturalists
al-D ibaji 54 Fatima, Haji 235
In dex 35i
Fatimi 2l 6 , 217 G ujarat 4 , 6 3 , 6 8 , 74- 7 5 , 7 9 , 8 5 ,
Fatimid dynasty, the 197 10 7, 245
Fattala, Haji 233 Gujarati merchants 2 3 , 24 > 247 ’ 255
faunal remains 20 see also Gunawardana, R .A .L .H . 10 4 , 134’
archaeological remains 168
Ferrand, G abriel 94 > IJ4 and Y u m io Sakurai 153
field knives 308 see also boat-build ing G u ru -S h ish y a Param paraw a (teacher-pupil
tools continuum ) 131
Firdausi, Persian poet 222
firearm s 252 see also contraband Hadhramaut
First W orld War 227* 2 5 2 , 254 harbours 228; ports 246—48
fish 2 6 2 , 275 Hadhram i
products 246—4 7 ’ 253 diaspora 227>229, 248, 249 see
fisherm en 5 7 ’ 74“ 7 5 ’ I3^ , 2 6 2 , diasporas; mercenaries 245“ 4 6 ;
276 m igration 2 3 0 , 2 4 9 ’ 2 5 5 ?
fishing 3 , 19, 237- 3 8 , 2 6 1 - 6 2 , 2 6 4 , sailing ships 249~53’ 255;
2 7 7 ’ 291 shipping ch. 8,
ethno-archaeology o f 18 com petition 228, 229,
fishing boats 88 see also boats, types o f 237 41 247 48
- ’ - ; in
fishing com m unities I j —2 0 , 7 4 ’ 3J3 Dutch colonies 231—32;
Forbes, A . 237 em ploym ent o f Europeans
fore-an d -aft rig 13 7, 2 3 1, 233 2 3 6 ; m ajor S ayyid families
fu n d i (master craftsman) 2 6 4 , 265 o f 230,233 ’ 234 ’ 235 ’
al Fustat 45 238, 240- 42, 244- 45’
251, 255, 257; slave and
Galle, port-city o f 141 contraband trade 249_52
Ganapati 57 see alsocon traban d, slaves
Ganga 73 and slave trade; trade in
Geniza(h) documents, Cairo 2 6 , 3 2 , specific regions 242—4 6 ,
45- 46 , 50- 5 2 , 54 ’ 56, 185 see 253- 55 ’ 255- 57’ 258
a /50 Jewish documents hajj 19, 44 pilgrims/pilgrimage
see also
G ibb, H A .R . and Beckingham, C .F. Halal Shipping Com pany 248
196 Harappan period/civilisation 5 , 18,
G ib so n -H ill, C A . 3 ° 6 —3°7 36, 66—67
G ildem eister, J. 192 harbour installations 8, 14
G ill, J.S . 155 a l-H ariri 201
G iriam a people, the 264 harraqa 174, 180, 192—93, 197
glass beads 99_ 100 see Q/50 beads Hasan, H. 94, 114 —*5
Goa 17, 5 6 , 76 H atan N a v (warships) 142 see also vessels
G oitein, S.D . 3 2 , 45- 46 , 53- 54 H ein e-G eld ern , R. 312
gold 2 5 » 62 see also trade Hejaz 19
gouges 310 see also boat-b uild ing tools heroin, sm uggling o f 261 see also
G reek and Latin authors 93_ 94 contraband
Greek texts see Periplus H ill, A .M . 21
Greeks 83—84 H indu Imperialism 3
G reenhill, B. 64 Hippalus 209
G renet, Fr. 104 historiography 2 , 34~ 35
Grey Ware 105 see also ceram ics, H orden and Purcell xv
potsherds, pottery H orm uz, port o f 182
35^ In d e x
H orn ell, J. i i , 7 3 , 135- 3 6 . x44 > Indicopleustes, Cosmas (sixth century

146- 4 8 , 151, 15 4 , 156- 5 9 . 16 6 , AD ) 9 2 , 9 4 , 10 4- 10 5 , n o , 112,
2 8 9 , 2 9 2 , 296- 9 7 . 2 9 9 . 3 0 2 , 114, 168
3 0 6 , 309 Indonesia 9 , 2 4 2 , 291
H orridge, G .A . 3 0 5 , 312 Indus delta 15
horses 3 9 , 54~ 55, I O I , 14 2 , 182- 8 4 , Indus river 5
19 0 , 233 . 237- 38 , 243 . 246 inland com m unities 31
see also trade, in live animals inscriptions/inscribed legends
H ourani, G .F . 8 A rabic 6 3 ; from Bharhut 8 6 ;
Hsii Y in g -ch o u 299 . 3 ° 2 , 312 Brahm i or Sinhalese 94 . IO I,
hybrid technology 4 . I45- 4 6 , 15 4 . IO3- IO 4, 10 5 , 107, 10 8 , 10 9 ;
30 6—307 see also sh ip-build ing from Chandraketugarh 3 9 , 5 1;
technology, syncretism in Devundara 14 I; near Dharka,
hydrostatic analysis 164 3 3 ; from Duvegala, Sri Lanka J ,
14 8 ; from G oa 5 1; from Java 25,
Ibn a l-A th ir 193 6 3 ; from K anh eri site 8 4 ; from
Ibn Battuta 4 5 , 4 9 , 5 4 , 13 5 , 14 1, 16 9 , K onkan and G ujarat 42 —43 > 51 ;
ch. 6 from the Malay Peninsula
voyage o f 177; classification o f 49_ 5° ; at Nasik 8 3 ; at
vessel-types 178—IJ 9 Nayinativu 14 2 ; Q u ilo n copper
Ibn Hawqal 180 plates o f Sthanu Ravi 1 4 ;
Ibn Jubayr 4 5 , 4 7 , 5 0 , 175 , 19 4 , Sanskrit and A rabic from
1 9 9 - 2 0 0 , 202 Som nath 5 0 , 57
Ibn Juzayy a l-K albi, Muhammad 175 insurance 2 3 6 , 2 4 1
Ibn Khallikan 197 intaglios 9 4 , 10 7—108
Ibn M ajid, A hm ed 48—5 ° , 20 9—2 l8 , iron
221- 24 cargo 52—5 3 > 5 5 , 6 2 , 184 see also
Ibn Mammati 189 trade, in metals and m inerals; in
Ibn al Mujawir 46 ship-building 155 see also boat
ibn Shahriyar, Buzurg 8 , 5° , 72, 19 6 , buildin g tools, sh ip-build ing
211 technology
iconographic illustrations 76 see also Islam 44 see also hajj, Jeddah, pilgrims/
ships, representations pilgrimage
A l-Id risi 4 6 , 134 al-Istakhri 180—l 8 l , 203
imam 210 Ivens, W .G . 312
imports 2 6 , 184 see also cargo, ivory 2 6 , 9 2 , 9 9 , 115 see also trade, in
com m odities, trade prestige goods
Indian O cean ch. 2
maritime studies 3 0 —3 2 ; marine/ Jabal Al-Jussasiyah 4 —5
maritime archaeology 3 2 — 3 3 , Jaffna thoni 154- 158 , 159—16 4 , 282
35—36 ; early Indian seafaring Jagannatha 4 , 76
34—3 5 ; representations o f ships Jaina text 37
3 7 — 4 1. 4 3 —4 4 ; m ajor zone o f Jainism 86
maritim e contacts 4 I—4 2 ; Jaki Bandar 15
classification o f vessels 4 2 , 4 6 ; jaladhisakra 39
sh ip-build ing technology jalapathakammikena (maritime trade) 7 4 .
47—4 8 ; navigation 48—4 9 ; 84
captain and pilot/navigator Jatakas, the 74 * 86
49 —5 ° ; ship owners 5I_ 57 Jeddah/Jiddah 19 , 235, 24 ° —4 1. 244 .
In d e x 353
252 see also hajj, pilgrims/ kotiaAottia I I , 7 2 , 279, 289
pilgrimage K u rien , J. 18
Jewish docum ents 2 6 , 3 2 , 45“ 46 , 5° - Kushan dynasty, the 106
52—55 see Geniza(h) K utchi 16
documents, C airo
Jews 63 labour recruitm ent 8 1, 257 see Qlso
John o f M ontecorvino 20 0 —201 crew
Johnston, A . Sir 140 —141, 156 Lakshmi, goddess 102
junks 4 5 , 5 4 , 18 0 , 186- 8 , 2 0 0 , 295 Lamalera boats and b oat-b uild ing
299- 303- 13
Kadambas, period o f 76 Lamu 259- 2 6 0 , 2 6 2 , 2 6 4 , 277
Kalyan, port o f 68 Lamu archipelago 2 0 , 253- 256
Kambojas, the 108—109 Lanyu see Yam i boats and boat
K anh eri 8 4 , 86 building
K ano, T . and Segewa, K . 30 3—3 0 4 , lapis lazuli 9 9 , 107—109
30 7- 3 0 9 , 311 latitudes 7 8 , 138 , 2 14 - 2 l6 , 220 see also
Kapitan, G. 151- 153 longitudes
Karttunen, K . 95 launches or pleasure boats 18 0 , 192—3
Kassapa, king I I I Leach, E. 2 9 3 - 9 4 , 297, 299 , 3 0 2 ,
Kathiawad 4 2 , 57 3 0 4 , 3 0 9 , 3 1 2 - 13
Kaveripattinam 3 5 , 56 legal systems, indigenous 141
Kaveripum pattinam 25 Lewis, A . 46
Kentley, E. and G unaratne, R. 136 Levi, S. 94
Kenya ch. 9 lineages 2 2 9 , 2 4 9 - 2 5 5 - 258
boats and boat-b uild ing centres liquid continent xv
259- 6 0 , 2 6 2 , 2 6 4 , 2 7 7 ; long duration, concept o f 29 see also
boat-b uild ing process and rituals Braudel, F.
265- 76 see also longitudes 7 8 , 2 14 - 2 l7 —2 0 , 224 see
cerem ony, custom , prayers, also latitudes
prohibitions, p ro tection , ritual/ Lothal 5 , 14
rituals, superstition; craftsmen
263—6 4 ; seamen 262—63 Ma Huan 138 , 217
Khaberos 36 Madagascar 12
K hadhuri 214, 222 Madmun 5 2 , 62—63
al-K h afaji 198 Mahadevan, I. I O I
K h alaf b. Isaac b. Bundar 52—53 Mahavamsa 10 4 , 142
Kham bat, G u lf o f 68 Mahavastu 6 9 , 86
khandy (pi. khandies, measure o f weight) Mahayana 86
21 al-M ahri, Sulaiman 216
Kharik, island o f 181 Mahruz b. Jacob 55—57
Kilwa, port o f 16 maji tamu (sweetened water) 273 see Qlso
K iribam une, S. 94 ritual/rituals
Kitab Ajaib ul Hind 5 0 , 211 Malabar coast I I , 18, 2 3 , 4 6 , 243- 44
Kitab al-furusiyya wa l-manasib al-harbiyya Malay Peninsula 24 - 49—5° - 63
20 2- 203 Malay w orld, the 2 2 8 , 244 - 255
Klings 24—25 Malay craft 231 see also ships, types o f
K oninklijke Paketvaart Maatschappij Malaysia IO
(KPM ) 2 3 9 - 4 0 , 2 4 2 , 2 4 6 M alindi, port o f 16
K onkan 4 , I I , 4 2 , 5 1 , 5 5 - 5 6 , 7 9 - 2 45 Malinowski, B. 294
354 In d e x
M anguin, P .-Y . 3 , 2 2 , 295> M esolithic period 82

30 6- 3 0 7 , 312 M esopotamiam texts 63
Manimekalai, the 56 mestri/mestris/mesta (master shipwright)
M anjrur (Mangalore) 45—4 6 , 5 2 , 55’ 13 9 , 281
6 2 , 184 metal technology 9 » 3*2 see also boat
Manthai, port o f 95 building tools
Maqamat o f al-H ariri, the 20 1—203 m iddlem an 63
al-M aqrizi 18 9, 19 3 , 197 M idnapur 4
M arco Polo 9 , 181- 8 3 , 186- 8 7 , m illet 2 5 2 , 2 5 3 , 254 see also trade, in
199- 2 0 2 , 303 staples/bulk items and foodstuffs
m aritime M ing dynasty 45
archaeology see archaeology; M odhera, port o f 68
com m unities see com m unities, Mogadishu, port o f 16
m aritim e, fishing com m unities; M ohenjodaro 5
culture 2 9 , 7 2 , 3 J3 ; Mohit, the 2l6
ethnography 7 0 ; expeditions 16 ; M ongol invasion 180
history 3 , 8 , ch. 2; hunting- m onsoon 31—3 2 , 4 5 ’ 48—4 9 ’ 69—7 0 ,
gathering 19 ; laws and usages 79, 13 4 , 20 2- 2 0 3 , 2 0 9 ,
140 —4 1 , 2 10 ; literature (Greek, 212- 13 , 2 2 0 , 2 3 7 , 26l
Sanskrit, Arabic) 16 ; networks see m onsoon trade 69—7 0 , 18 2, 203
trading diasporas M ookerji, Radha K um ud I—3 , 35’ 67
markab (pi. marakib) 4 5 ’ 18 0 , 184—8 6 , M orocco 174
195’ 247 mosque or mijigiti (Arabic masjid) 57’ 63
markets 14—15> I&2 M otor Kustvaart Maatschappij
mashua 260 —6 2 , 2 6 4 , 276—77 (M okum ij) 242
construction o f 265—75 M uciri 25> 82
al Masudi 5 6 , 18 0 , 182 A l-M ukalla 2 2 8 , 247
maps 7 9 , 138 M ukherjee, B. N. 37’ J6l
and portulans 16 ; and charts 7 8 , al-M uqaddasi 181, 18 5, 203
8 5 ’ 219 museums
Mauryan em pire 36 Asutosh Museum 7 ; Baroda
mechanisation Museum and Picture G allery 8 5 ;
attempts at 8 0 ; and com m ercial British Museum 2 2 0 ; Galle
fishing 66 Maritime Museum 1 6 O ; Goa
m echanised vessels see vessels Museum 43 ’ 77; Ham burg
M editerranean 8 , 18 5, 186 Ethnographic Museum 217;
manuals on marine design 175 ^ Ham burg Museum fur
merchants 18 3; vessels Volkerkunde 2 9 6 ; Indian
see vessels, harraqa National Museum, D ehli 2 10 ;
'M editerranean M ou ldin g’ 279 L ondon Science Museum 16 3 ,
M egalithic Black and Red Ware 16 6 ; Museum o f M ankind 16 3 ;
98- 9 9 , 107 Sur M aritime Museum 2 l8 , 221
M egalithic period 83 musk/coster 10 4 , 18 4 , 187 see also
Megasthenes (third century B C) 93 trade, in spices and aromatics
merchants 24—2 5 > 56 M uziris 2 6 , 49 ’ 56, 82
G ujarati 2 3 , 24 ’ 247- 2 5 5 ; myths and legends 34
M editerranean 18 3; ship-
owning 51- 5 4 , 74 see nakhoda/ Nahray b. Allan 5 2 , 54
nakhuda/nauvittakas-, Tamil 6 3 , IOI Nairuz or Nauruz 2 I 3—214
In d e x 355
Namalinganusasana o f Amarasimha 47 Nestorian Christians 1 5 , H O
Nanadesis, the 102 Nilakantra Sastri, K .A . 102
Narayandeva 3 niyyamakajetthaka 74
Nasik 83 N orthern Black Polished Ware 1 0 5
nautical charts 16 , 78 , 8 5 , 219 al-Nuwayri 1-Iskandarani 1 9 4 , 1 9 6
nautical sciences 15—16 N .V . Voorwaarts company 2 3 9
nautical technology 22—23 see als°
ship-building technology occupations/trades 18 , 74“ 7 5 > 81
nautical term inology I I , 13 O chsenwald, W. 251
nautical terms O d oric, Friar 187
anjar 4 8 ; aritra 4 8 ; ballisti 2 1 7 ; bild O m an 4 , x3 > *9
2 14 ; bilij 5 4 ; citra or alekfya 7 8 ; M inistry o f National Heritage and
datragrahaka 5 0 ; durusi 2 71 ; C ulture 2 10 , 214
Fatanswami/Pattanasvami 53; fundo O nesicritus o f Astypaleia (fourth
2 7 3 ; isharat/alamat 2 1 2 , 2 1 4 - 2 1 5 ; century B C) 9 3 , I O 3 - I O 4 , 129
kamal 216- 17, 2 l8 , 2 1 9 - 2 0 , 2 2 4 ; oral tradition 1 3 1 —3 2 , 1 3 7
karavai 7 8 ; kamadhara 5 0 ; khashabat O ran g Laut 18
2 l 6 ; kupastambhaAupadanda/ Orissa 4 , 6 8 , 7 1 , 7 3 - 7 4 , 8 7
gunavrksa 4 8 ; mahanavika 4 9 — 5 ° ; oruwa I I , 1 3 2 , 134 - 3 5 . 147 - 4 8 , 151
mandira 4 8 ; mataruma 2 6 7 ; mu’allim O ttom an-Italian war 252
5 0 , 210- 12, 215, 2 2 2 , 2 2 4 ;
nakhoda/nakhuda/nauvittakas I , 5 ° , Padma Purana/Manasamangala 3
51- 5 7 , 2 1 1 , 2 1 4 , 2 2 0 , 221 - 2 4 ; Palembang 234
nangara 4 8 ; nangarasila 4 8 ; Pali texts 4 9 , 1 0 8 - 1 0 9
navikakarmakara 50—5 1; niyamaka Palk Bay 2 8 1
49 - 5 0 ; pasek a kayo/pasak kayu Palk Straits I I
3 0 4 ; phalakas 4 8 ; pothi (guide) Palladius (fifth century A D ) 9 4
2 14 ; qiyas 2 16 ; rajju and sutra 4 8 ; Pandyans, the 102
ramada 2 7 4 ; rasmigrahaka 5 0 ; rayyis Pangeran, title o f 233 ’ 234
5 0 ; rubban 5 0 , 2 10 ; sasaka 4 9 ; Panini 74
sitapatta 4 8 ; siyasat 5 0 , 2 12 , parallel sailing’ 217
221—2 3 ; tajiri 2 6 6 , tandil 5 0 ; Paranavitana, S. 9 4 , I O I , 1 0 8 —1 0 9 , I I I
utsecaka 5 0 ; zflms 2 15 * 2 17” 18 Paravas, the 2 8 9
naval strategy 143 Paris, A dm iral 72 —73, 7^, 13 0 , 14 6 ,
navigation, indigenous knowledge o f 158
7 8 - 7 9 , 1 3 7 - 3 8 , ch. 7 see also partison method 2 8 8
navigators and pilots maritime paruwa 151, 156, 1 6 3
calendar 2 1 2 —1 4 ; identifying Pathamanathan, S. 1 0 2 — 1 0 3
signs and landmarks 2I4~ x5 ; pattana 67
instrum ents 215“ 18 ; European pearl fisheries 4—5 ’ ! 6 6 , l 8 l —8 2 , 1 8 4 ,
impact 2I9—2 1; proper 238
m anagement o f ships 2 2 1 —2 3 pearls 8 2 , 9 2 , 10 5 , 18 4 , 2 4 3 , 2 4 6 ,
navigational directories and guides 2 4 9 > 252 see also trade, in
210, 220 prestige goods
navigators and pilots 8l m oth er-of-p earl 2 3 8 , 2 4 9 , 252
'ideal’ 210—12 ; authority o f 2 10 , Pelliot, P. 94
211, 221- 23 pepper 9 , 2 5 , 5 3 , 6 2 , 9 2 , 10 4 , 184 see
Nederlandsch-Indische Stoomvaart also trade, in spices and
Maatschappij 24 ° aromatics
356 In d ex
Periegesis 94 Qalhat, port town o f 1 82

Periplus Maris Erythraei J , 36—3 7 , 4 1 Q atar 4- 5
4 8 - 4 9 , 7 2 , 78- 79- 8 4 , 9 3 , Qays, island o f 181
107- 10 8 , 1 4 5 - 46 qirtbar/qanbar 1 9 9 see also ship-building
Perreira, Father Nicolau 295 materials
Persian dom ination, period o f 223 Q u ilo n , port o f 1 5, 1 8 4
Persian G u lf 6 , 3 6 , 4 6 , 6 9 , 196 Qûran, the 2 6 9
Perso-Arab and Indian vessels see al-Q u tam i, Isa 2 1 0 , 2 1 4 , 2 l 8 , 2 2 0 ,
vessels 2 2 1 , 222
Phillips-B irt, D. 13 6 , 14 4 , 14 9 , 155
Phuket 19 Ramisht o f Siraf 53~ 54
pilgrims/pilgrimage 19 7- 2 3 5 al-Ram m a, Hasan 202
2 4 0 - 4 1, 2 4 4 , 252 Rao, S.R . 1 4 , 1 3 8
pilots 8 5 , 137 see navigators and Ratnayake, H. 1 0 8
pilots 're-A rab isa tio n ,’ process o f 255
piracy 2 2 , 5^—5 7 - 2 3 0 , 262 Red Sea
pirates 5 5 , 5 7 , 141, 18 2, 188 ports 4 5 - 4 7 - 4 9 - 5 4 - 8 4 ; vessels see
Pires, Tom e 71 vessels
Pliny (the Elder) 3 6 , 48 —49 - 93 - ! 29 - religions, variety o f 85—87
161, 215 religious tolerance 57
Polonnaruwa 94 - 97 Ridiyagama 107
Polynesia and Melanesia I I Rieth, E. 2 7 9 , 2 8 9
ports, ancient Rigveda, the 2 15
Indian 97- 101 > Sri Lankan 94—98 Rihla, the ch. 6 see also Ibn Battuta
Portuguese portolanos 16 , 85 vessel-types in 1 7 6 , 1 7 8 — 1 8 0
Portuguese shipping 23—2 4 - 4 1- 7° - 75
2 2 4 - 289 see also ships, types o f ritual/rituals see also cerem ony,
pothi 1 6 , 2 1 4 custom s, protection,
potsherds 9 8 , I O I see also ceramics superstition boatbuilding
pottery 10 9 , 183 see also ceramics 2 6 5 — 7 5 - 3 0 9 - I O ; prohibition
Potts, D .T . 5 {kambe) 2 7 3 see Qk° pro h ib itio n s;
prayers see protection status 2 5 see also caste, status
Prins, A .H .J. 26 0 - 6 1, 2 6 6 , 2 7 4 , 2 7 5 , rock
277 art 8 2 ; carvings 1 8 3 ; crystals 9 9 ,
prohibitions 3 4 , 2 7 3 * 3 ° 8 see also IIO
cerem ony, custom s, protection, rock-cut caves 86
ritual/rituals, superstition Roman em pire, trade with 3 6
p rotection Roman merchants 5 6
from dangers o f sea 86 see also 'R om an’ vessels, representations o f 7
B uddhism ; prayers and Rouletted Ware 7 - 98—99
offering (Sadaqa) for 269 A l-R ub ban Abharah 7 2
see also cerem ony, custom s, ruins
p ro h ib itio n s, ritual/ Jaki Bandar fortress 15 ; Q alhat 1 8 3
rituals, superstition
Ptolemy, Claudius (second century Saddharma-pundarika-sutra 86
A D ) 3 3 , 3 6 , 4 1 , 9 3 , 15 6 , 222 sagarapaloganas 8 4
salt 8 2 , 2 4 7 - 2 5 3 see also trade in
Qais (Kish) 46 staples/bulk items and foodstuffs
Qaisar 17 sambuk 7 2 , 2 4 7 - 2 4 9
In d e x 357
Sammohavinodani 146 as distinct from boats 6 4 , 6 6 ;
samyatrikas (passengers) 42 diversity/variety o f 5 , 4 I_ 4 6 , 7 2 ,
Sanskrit texts 2 , 4 2 , 48 —49 see 2 30 —3 1, 2 4 7 ; representations
Yuktikalpataru on bas-reliefs 13 6 ; on
Sargon o f Akkad 63 coins 3 , 5 , 7 , 4 1, 7 6 ;
Sarsfield 279. 289 heterogeneity o f 5—6 ; on
Sasanian period 15 m em orial stones 3~ 4 >
ceramics, bullae and coins o f I I O ; 4 3 - 4 4 , 76 - 77 ; o n
decline o f 114—H5 m onum ents 7 6 ; in
Satavahana rulers o f the Deccan 4 I paintings 3 , 7 6 ; in ritual
saw 266 see also b oat-b uild in g tools and cerem onial use 4 , 7 6 ;
Sayyids, the 222 in rock carvings 14 6 , 18 3 ;
sayyid origins or lineages 2 2 9 , 2 3 5 > on sculptures 3 , 43_ 4 5 >
243- 245- 249 . 258 7 6 , 15 9 ; on seals/sealings
Severin, T . 13- 1 4 , 2 12, 217 5- 7. 37- 40 , 76- 77; o n
Schafer, E .H . 9 4 , I I 3 - I I 5 temple panels 3—4 , 76
Schlingoff, D. 2 , 4 1 seafaring capabilities o f 6 , 2 6 , 3 7 ,
sculptures 3 * 4 3_ 4 5 > 7 6 , 159 see 3 9 , 4 1 ; types o f see also boats,
stones, m em orial vessels ahawra 18 0 , 19O ; almadia
sea-snakes 7&—7 9 > 2 14 . 215 19 8 ; baghla see baghla; bargatim J l ;
sea/trade voyages, accounts o f 3 , 21, batel 7 2 ; battal 165—6 6 ; battil 13 ;
3 4 - 3 5 . 4 5 . 5 2 , 6 9 . 84 boum see boum; bum see bum; burma
s eal s /s eal i ngs 5 - 7 , 3 7 ~ 4 0 , 5 1 , 7 7 , 9 4 , 4 6 ; caravels/caravela, Portuguese
10 3 , 10 7 , IIO 2 3 , 15 5 ; catamaran (kathu maram)
seaports see ports I I , 14 8 ; chialoup (shallop) 23 1;
Second W orld War 254 C hina ships (markab sini) 8- 9 . 72;
sewn-boat see also ship-building colandiaphonta 4 1 ; dau 261, 277 see
technology also dhow; dhangi and nauri 7 2 ;
construction 2 4 6 ; technology dhow see dhow; gallevat 19 4 ; ghanja
146- 47 2 4 7 ; ghurab 18 0 , 1 8 8 - 8 9 ;
sextant 2 l7 ~ 18, 219—2 1, 224 see also gondola 19 7; gonting 2 3 1 ? harbi
astronom ical and navigational 18 0 , 18 8; harraqa see harraqa; huri
instruments (dug-out canoe) 2 6 1; jafn (pi.
shabbara 18 0 , 192- 9 3 , 197 ajfan, jifan or jufun) 18 0 , 191—9 2 ;
al-Shahri 183 jahaj 4 5 ; jahazi 2 6 0 , 2 7 7 ; plba
Shanga 20 18 0 , 193—9 4 ; j Qshujiyas 4 6 ; jilab
shaped-log-raft technology 146 see also 45- 47 ; jong 2 9 5 ; jonque 24 ; jukong
sh ip-build ing technology 2 0 ; junks see junks; kakam 18 0 , 18 7;
shark oil 254 see also ship-building katha and velu 3 7 ; kattumaram 8 0 ,
materials 1 4 7 ; kora-kora 2 3 7 ; kotia/kottia see
Shepherd, G . 229 kotiaAottia; kottymba 37; kundura
Sihalavatthu 108 (pi. kanadir) 18 0 , 19 6 ; madel
Shinkar and Binzagr & C o. 25° paruwa 13 6 , 15 I; ma'diyya 18 0 ,
shipping 197—9 8 ; markab see markab; mashwa/
organisation o f 2 6 ; segmentation machwa 4 6 ; masula I I , 7 1; matalai
o f 227- 28 (also kalam) 4 1 ; mayang 2 0 , 2 3 1;
ships meshiat 4 6 ; mtori 2 6 1, 2 7 7 ; nailed
carrying capacity/size o f 8—9 , 18, ships 2 0 0 —2 0 1; nao 2 3 ; nau 2 9 5 ;
4 6 - 4 7 , 8 1, 2 3 3 , 28 1, 2 8 7, 2 9 5 ; oruwa see oruwa; padagu/padava/
358 In d e x
padhua I I , 7I_ 72 ; paduwang 20 ; 278—7 9 ; sewn-plank or stitched-

paruwa see paruwa; patakaran $11; plank 2 , 7 ’ 9~ n , 2 3 > 4 3 ’ 4 7 ’ 6 5 ,
patia and danga J l , 73 ; pattamars/ 135- 36, 149’ 1 6 1 , 295> 307 ;
pattemars I I , 72; pencalang 231; shell-first construction
perahu 236—3 8 ; pota 3 7 ; prahu 45 > x99—2 0 0 ; Swahili mashua
pravahana 4 2 , 51 ; qarib 18 0 , 265- 275; wooden dowelling 7 , 9 ,
184—8 5 ; safina 18 6 ; sallura 2 3 . 2 0 0 , 295- 9 6 , 30 3- 3 0 4 ,
192- 9 3 ; samanya and visesa 4 2 ; 3 1 1 ; zones o f 1 4 3 —4 7
sambuk see sambuk; sampan 4 5 5 sh ip -b uild ing texts
shabbara 18 0 , 1 9 2 - 9 3 , 19 7 ; Babylonian 6 ; G raeco-R om an 7; U r
shaffaras 4 6 ; shiller 18 0 , 18 9 ; shoe- III 6—7 ; Yuktikalpataru 2 , 35> 42
dhoni 8 8 ; sunbuq 18 0 , 194—9 6 ; ship-owners see boats, ownership o f
tarida 190—9 1; thoni see Jaffna ship-wrecks 14 , 17 » 2 9 * 2 8 9 , 30 6—307
thoni; toni, Tuticorin (coasters) see al-S h ih r 2 2 8 , 2 4 5 , 247
toni; toop and pinsi 2 3 4 ; trappaga/ silk see Chinese silk
trapyaga 37—3 8 , 5 1; ukayri 18 0 , Singapore 2 3 3 , 2 3 4 , 2 3 5 , 239- 4 1 ,
1 8 8 - 8 9 ; ushari 18 0 , 1 9 6 - 9 7 ; 251
vahitra 4 2 —4 3 ; vallam (dugouts) see Singapore Steamship Com pany 24 °
vallam; vattai/vattal see vattai/vattal; S iraf 15, 4 5 , 53- 54 , 181
janapatra 4 2 ;jathra dhonis see jathra Silappadikaram, the 56
dhonis; zarook 13 ; zaw 18 0 , 187—88 slaves and slave trade 2 3 6 , 24 4 “ 4 5 ’
ship-builders I39- 4 ° see boat 247* 250 - 51, 253’ 255- 57
builders and craftsmen Sofala, port o f 16, 46
sh ip-build ing materials Som nath 57
babul tree 7 8 ; bam boo cane mat Song (or Sung) dynasty 9 , 45 * IT6 , 187
18 7; barok root 311'* cam phor Sopara, port o f 68
271 ; coconut fibres/coir 2 3 , 4 7 > spices see trade, in spices and
5 5 , 7 2 , 134., 19 4 . 19 9 , 2 4 8 ; aromatics
coconut oil 2 6 5 , 2 7 3 ; cotton Sri Lanka 49 ’ 5&» cb- 4 - c^- 5
2 6 5 ; ebony J 2 ; grease or fish/ ancient ports o f 94—9 8 ; Buddhism
shark oil 7 * 4 7 >2 0 0 , 2 6 5 ; palm - in I O 3 - I O 4 , 1 0 9 , I I I , 114 ; early
branch skin 3 H ; palm -fibre and historical sites 99—I° 3 ;
p alm -leaf ropes 6—7 , 4 7 ; 51500 archaeological evidence on
7 2 ; spruce and fir 18 6 ; tallow shipping com m unities in
and lim e, bitum en 6 ; teak wood 94—9 5 ’ 98—117 see also ships,
4 7 , 72, 155. 199- 232 , 235 representations
2 4 4 , 2 4 7 - 2 4 8 , 251 Sri Lankan Departm ent o f Fisheries
sh ip-build ing technology 13 1 see also 1 5 2 -53
boat-b uild ing technology, Sri Lankan ships, p re-m od ern ch. 5
hybrid technology battal 165—6 6 ; evolution and cross
caulking 6 , 47 ’ 311; Chinese 18 7; pollination o f techniques
dovetail join ts 2 8 7, 2 8 9 ; fram e- 135- 3 9 ’ I49- 51 ; factors
first (skeleton) construction influencing design 132—3 9 ;
278- 79; lashed-lug 9- IO, 293, involution o f 16 8 ; Jaffna thonis
295’ 3° 7; 'M editerranean 154—5 8 ; major differences
M ou ldin g’ 279’ 2 8 9 ; metal within Sri Lanka 147—49 ; ship
tools, use o f 9 , 3^2 see also boat builders, shipyards and ports
building to o ls; oral tradition 139—4 3 ; technology transfer and
13 1; plank-first construction zones 143—4 7 ; typology o f
In d e x 359
I5 I~ 5 3 > 169- 7 1 ; vattai/vattal 16 6 , see also vattai/vattal
16 7, 282—8 3 , 2 8 6 ; jathra dhoni Tamil shipwrights 278
158- 64 Tamralipta, port o f 43 , 49
Sri Vijaya 5 8 , 86 Tampoe, M. 15
Stark, F. 248 Tang dynasty I I 5
stars 7 8 , 13 7, 2l6 tatara 293, 294 , 297 , 299 , 3QI
status 276 see also caste 'tax farm ers’ 2 5 , 63
steam navigation 3 1, 6 6 , 6 9 , 2 2 7 Tchernia 7
2 3 7 , 239—42 see also temples
vessels, mechanised in West Bengal 3—4 * 76 ; terracota
Steensgard, N. 58 3 - 7 6 ; o f Augustus 5 6 ; in Orissa
stones, m em orial (viragals) 3“ 4 > 76 , 8 7 ; at Dodanduwa 159
4 3 - 4 4 , 7 6 , 77 textiles see trade, in textiles
stones, precious and semiprecious 9 2 , Thera Mahinda 103
9 9 - I O O , 1 0 5 , 114 T hierry, F. I l 6
Stom ach, D. 112 Tibbetts 7 9 , 2 0 9 , 2 1 9 - 20
Strabo ( 6 4 B C - A D 2 l ) 9 3 , 1 2 9 T in b u 55- 57
subsistence strategy 1 9 —2 0 Tissamaharama 97
sugar 6 2 , 2 3 3 , 2 4 3 » 257 see Qlso trade, toni, T uticorin (coasters) J , J 8 , 8 0 ,
in staples/bulk items and 133, 156, 2 7 9 , 2 8 l, 2 8 2 ,
foodstuffs 284- 8 5 , 2 8 7 , 289
Sultanate o f D ehli 16 T on ti 82
Surabaya-Gresik 234 tourists/tourism 261, 2 6 2 , 2 6 3 , 2 7 2 ,
Surat, port o f 2 1, 68 276- 77
Suresh, S. 9 8 Toussaint, A . 153
superstition 30 7—308 see also trade
cerem ony, custom s, in coffee 13 , 2 3 9 , 2 4 3 , 2 4 6 , 2 5 I;
p ro h ib itio n s, in contraband 2 3 2 , 2 3 5 , 252 see
ritual/rituals also h e ro in , firearm s; in gold 2 5 ,
Swahili, the ch. 9 6 2 ; in live animals 3 9 , 54“ 55 >
boats 260 —6 2 ; seamen 2 6 2 ; 9 2 , 10 1, 14 2 , 182- 8 4 , 19 0 , 2 3 3 ,
craftsmen 263—6 4 ; technical 237- 3 8 , 2 4 3 , 2 4 6 , 2 5 2 , 2 5 4 ; in
and ritual process o f metals and minerals 52—5 3 ’ 5 5 >
boatbuilding 264—76 see also 6 2 , 7 ° > 10 4 , 18 4 ; in prestige
custom s, prayers, p ro tection , goods 2 6 , 5 5 , 9 2 , 10 5 , 115, 18 4 ,
p ro h ib itio n , ritual/rituals, 18 7, 2 2 1, 2 3 8 , 2 4 3 , 2 4 6 , 2 4 9 ,
superstition 2 5 2 ; in spices and aromatics 9 ,
syncretism 223 see Qk° hybrid 24- 2 6 , 53, 55 , 6 2 , 9 2 , 10 4, 183,
technology 184, 187, 239, 2 4 4 , 25 3; in staples/
bulk items and foodstuffs 13, 55’
A l-T abari 18 8, 193 6 2 , 6 3 , 82, 115, 184, 233’ 243’
Tabula Peutangariana 56 246, 247’ 248, 25 2, 253’ 254’
Tajjika (Arabic) merchants 56 2 5 7; in textiles 2 6 , 54—55, 92, 10 4,
Takuapa 63 105, 2 4 4 ; in timber and woods 55,
Taiwan 2 9 1, 2 9 2 , 294 62, 8 l, 92, 10 4 , 184, 199, 251; in
Tamil ancient epics 56 weapons 184; in wine 6 , 25
Tamil merchants 6 3 , I O I trading diasporas 56 see also diasporas,
Tamil poetry 2 5 , 82—83 Hadhram i diaspora
Tamil Nadu fram e-first vessels ch. IO networks
360 In d e x
com m unication 9 8 ; inform ation Vosm er, T . 160—164

5 5 ; maritim e 20 —2 6 , 6 3 ;
trade 14- 1 5 , I O I ; trade Watamu 262- 6 4 , 2 6 7 , 2 7 5 , 276- 77
routes 184 watu wa bara 259
'Travoleta da India’ 219 water
Trikandasesa 48 for drinking 2 3 —2 4 * l 8 l —8 2 ; to
export l 8 l
'ulama 244 water-craft see boats, ships, vessels
war vessels see vessels
Vallabhi, port o f 68 warfare 4 * 11 > 76
vallam (dugouts) I I , 18, 14 7 * 2 6 1, 279 warring factions 143
289 weapons 1 8 4 see also contraband
Van den Berg 245 Weerakkody, D .P.M . 9 2 , 9 4 * 1 0 3
Vasco de Gama 2 0 9 , 2 10 , 219 W esterdahl, C . 29
vattai/vattal 166, 167, ch. I O , 282—8 3 , W heeler, R .E .M . 8 3
286 W hitehouse, D. 15
fram e-first design o f 284—8 8 ; Wijesekara, N. 94
com pared with fourteenth wind systems see m onsoon
century Venetian design 288—90 woods 9 2 , 1 0 4 , 1 8 7 , 1 9 9 see also tim ber
Venice 279* 288
vessels, classifying ch. 6, 178—179 see Xavier, Francis 289
also boats, ships
cargo 6 4 , 71, 2 6 0 ; Chinese Yam i boats and boat-b uild ing ch. II,
186—8 8 , 2 0 0 ; coastal sailing and chinedkulan/Chinedkeran 2 9 3 * 2 9 7 *
rowing 18 0 , 193—97 ; Indian 299* 3 0 0 , 30 2- 3 0 4 ;
O cean 4 2 , 4 6 ; mechanised I j , Lamalera boats and b oat-b uild ing
6 4 , 6 6 , 80—8 1; M editerranean 2 9 9 ,com pared with 30 3—13 ;
18 9, 19 0 , 191- 9 2 , 205 see also Lanyu 2 9 2 , 294 * 305 * 3° 9- IO,
harraqa; ocean -goin g 18 0 , 312—13 ; lashed-lug technique
186—8 8 ; Perso-Arab and Indian 293 * 295* 307 ; tatara 293 * 294 *
198- 2 0 5 297* 299 * 301
boat typology 19 9 ; hull designs 19 9 ; Yaqut 180
shell-first construction jathra dhonis 135- 3 7 , 14 3 , J5 6 ,
199—200 see also ship 1 5 8 -6 4 , 168
building technology; javana 2 5 * 56
Red Sea 4 5 - 4 6 , 19 3 , 1 9 6 - 9 7 ; Yishu, Abraham 52—54
Tam il Nadu fram e-first design Y u an dynasty 187
ch. I O see also vattai/vattal, ship Yuktikalpataru 2 , 35 * 42
buildin g technology; war 14 2 ,
18 0 , 188- 9 4 Zafar, city o f 183
Vitharana, V . 134* I 37—3^ Zanzibar 253
Villiers, A . 13 , 15 5 , 2 0 9 - II , 2 1 9 - 2 0 , Zheng He 16 see also C hen g Ho
222- 23

(Routledge Indian Ocean Series) Ruth Barnes (Editor), David Parkin (Editor) - Ships and The Development of Maritime Technology On The Indian Ocean-Routledge (2002)

Uploaded by

Copyright:

Available Formats

(Routledge Indian Ocean Series) Ruth Barnes (Editor), David Parkin (Editor) - Ships and The Development of Maritime Technology On The Indian Ocean-Routledge (2002)

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Routledge Indian Ocean Series) Ruth Barnes (Editor), David Parkin (Editor) - Ships and The Development of Maritime Technology On The Indian Ocean-Routledge (2002)

Uploaded by

Copyright:

Available Formats

S h ips a n d the D evelopm ent

Editors: David Parkin & Ruth Barnes, University o f O xford

T h ere is a need to understand the Indian O cean area as a

I slam ic P rayer A cross the I n d ia n O cean

E dited by David Parkin and Stephen C . Headley

D avid Parkin and Ruth Barnes

Published 2016 by Routledge

Simultaneously published in the USA and C anada by Routledge

Routledge is an imprint o f the Taylor & Francis Group, an informa business

(' 2002 David Parkin and Ruth Barnes

All rights reserved. N o part o f this book may be reprinted or reproduced or

British Library Cataloguing in Publication Data

Library o f Congress Cataloging in Publication Data

ISBN 13: 978-0-700-71235-9 (hbk)

List o f Figures Vll

8 T h e Rise a n d Fall o f H a d h r a m i S h i p p i n g in the

10 T h e H i s t o r i c a l C o n t e x t o f the C o n s t r u c t i o n o f the Vattai

2.1 T r a p y a g a vessel o n an i n s c r i b e d seal 38

6.2 V e sse l-ty p e s in Ib n B a ttu ta ’ s Rihla 178

Dionisius Albertus Agius is a reader in Arabic and the

Robert Barnes is a Faculty Fellow o f St. A n ton y’ s C ollege and

Ruth Barnes has carried out extensive field research in eastern

form s. Together with Joanne B. Eicher, she edited Dress and

Susan Beckerleg is a Social A nthropologist who has been

Osmund Bopeararchchi specialises in Central Asian, Indian

(G .N .R .S .), Paris. He is also, since I9 9 4 > the director o f the

Ranabir Chakravarti, with M .A. and Ph.D . degrees from the

Professor William Gervase Clarence-Smith teaches the

Somasiri Devendra, Sri Lankan. B .A . (Ceylon) 1955 A fter

Series editor David Parkin is Professor o f Social A nthropology

Himanshu Prabha Ray is Associate Professor in the C entre for

(University o f Cam bridge) and has a doctoral degree on Early

Abdul Sheriff is Principal C urator o f Zanzibar Museums and

Were early water-craft able to travel longer distances across the

considerable scholarship and im portance for that region, The

her own chapter in the volume itself. T h e other contributors

This introduction surveys the literature on shipping in the

A pioneering work on Shipping in the Indian subcontinent

sources, the work encompasses a time span truly m onum ental —

the central position o f India in the O rien t w orld, for

In spite o f these obvious drawbacks, there is much in the work

craft based on these. It is nevertheless significant that in the

realistic depictions o f river craft, while larger vessels form the

pearl banks in the region. In about I9 ° 0 , the Q atar pearl fleet

Failaka in the Persian G u lf show dou ble-ended vessels with

T h e literature on watercraft in ancient Mesopotamia

T h e seafaring capabilities o f these watercraft are evident from

the water transport was sewn or stitched. T he U r III texts list

the M editerranean, a survey o f the Red Sea coast in the vicinity

Indian O cean (Wake 1997)* Marco Polo, for example, provides

they are so large that they carry quite 5 ° ° ° baskets o f

A third region in the Indian O cean often associated with a

The Indonesian round dowel technique was either a

The link between the introduction o f metal technology and the

twentieth century. Also the evidence from the Pontian boat

A second m isconception often associated with ethnographic

India, Sri Lanka, Madagascar and the C om oro Islands as a

T h e appearance o f similar types o f simple boats in widely

In recent years, attempts have been made to docum ent b oat

T h e potential o f archaeology, especially underwater archae

stern-post rudder which is prom inently portrayed. T ech n o lo