ARCHAEOLOGICAL REPORT

Archaeological Recovery

of the

JAVA SEA WRECK

PACIFIC SEA RESOURCES

ARCHAEOLOGICAL REPORT

Java Sea Wreck

Pacific
Sea
Resources

MATHERS
FLECKER

1997
ARCHAEOLOGICAL REPORT

Archaeological Recovery

of the

Java Sea Wreck

Edited by

WILLIAM M. MATHERS
MICHAEL FLECKER

Pacific Sea Resources

1997
 Editors: William M. Mathers
Michael Flecker

Authors: History - John Miksic, Ph.D.

Iron - Bennet Bronson, Ph.D.
Ceramics - Roxanna Brown
Archaeology - Michael Flecker

Layout & Design: Catherine Noon

Pacific Sea Resources, Inc.

726 Second Street
Annapolis, Maryland 21403
U.S.A.

Telephone: 410-268-5661
Facsimile: 410-268-9371

1997 Pacific Sea Resources, Inc.

All rights reserved. No part of this book may be
reproduced in any form or by any electronic
means, including information storage and
retrieval devices or systems, without prior written
permission from Pacific Sea Resources, Inc., except
for brief passages which may be quoted for
reviews.
 THE JAVA SEA WRECK
ARCHAEOLOGICAL REPORT

TABLE OF CONTENTS

1.0 Introduction 1

1.1 Objectives 1
1.2 Locating the Wreck 1
1.3 Political Considerations 4

2.0 Historical Background 5

2.1 Chinas Position in the Nan Hai Trade from 500 B.C. Through the
Tang Dynasty (A.D. 906) 5
2.2 Trade Interaction During the Song Dynasty (A.D. 960-1260) 18
2.3 Trade Interaction During the Yuan Dynasty (A.D. 1260-1367) 23
2.4 Trade Commodities 27
2.5 Monetary System 30

3.0 Archaeological Program 34

3.1 Archaeological Objectives 34

3.2 Recovery Vessel and Diving System 36
3.3 Excavation Techniques 43
3.4 Photographic Procedures 46
3.5 Registration System 50
3.6 Ceramics and Artifact Handling Procedure 53
3.7 Conservation 56

4.0 Site Conditions 58

4.1 Site Description 58

4.2 Oceanographic and Meteorological Conditions 59
4.3 Environmental Monitoring and Marine Life 60

5.0 The Wrecking Process 64

6.0 Interpreting the Ship 67

6.1 Hull Structure 67

6.2 Stowage Pattern and Cargo Capacity 70
6.3 Wood Identification 71
6.4 Identifying the Ship 72
6.5 Dating the Ship 76
7.0 Artifacts 77

7.1 Iron 77
7.2 Ivory 80
7.3 Resin 81
7.4 Balance Weights and Bars 81
7.5 Figurines and Finials 83
7.6 Trays 85
7.7 Gongs 86
7.8 Copper Ingots 87
7.9 Miscellaneous Copper Alloy Artifacts 88
7.10 Glassware 89
7.11 Rocks 91
7.12 Sharpening Stones 92
7.13 Organic Finds 93
7.14 Tin Ingots 94

8.0 The Iron Industry and Trade 95

8.1 The History of Iron Production in Asia 95

8.2 Iron-Making Technology 97
8.3 Historical and Archaeological Parallels 99
8.4 Analysis of Iron from the Wreck 101

9.0 High-Fired Ceramics of Song Dynasty China: The Java Sea Cargo
in Context 103

9.1 Introduction 103

9.2 Chinas Supremacy 103
9.3 Sources and Properties of Clay 104
9.4 Firing Temperatures 105
9.5 Glazes 105
9.6 High-Temperature Glazes 106
9.7 Glaze Color 106
9.8 The Java Sea Ceramics Cargo 107

10.0 Thirteenth-Century Potting Techniques: as Evidenced by the

Java Sea Cargo 110

10.1 Collecting and Preparing the Materials 110

10.2 Preparing the Glaze 110
10.3 Furnishing the Kiln 111
10.4 Making the Molds 111
10.5 The Potters Wheel 111
10.6 Finishing the Form 112
10.7 Decoration 112
10.8 Firing 112
10.9 Illustrating the Potting Process 113

iv
11.0 Ceramics Inventory 116

11.1 Painted Ware 116

11.2 Black-Glazed Ware 130
11.3 Molded White/Qingbai Ware 131
11.4 Qingbai Ware 137
11.5 Celadon and Olive-Glazed Ware 149
11.6 Decorated Green-Glazed Ware 153
11.7 Undecorated Green-Glazed Ware 160
11.8 Brown-Glazed Ware 165
11.9 Indian-Style Earthenware 172
11.10 Miscellaneous Ceramics 178

12.0 Conclusion 182

Bibliography 185

Appendices 191

A. Artifact Database 191

B. Selected Pages from Artifact Registration Report 198
C. Artifact Distribution Plan 203
D. Ceramics Distribution Plots 205
E. List of Marine Species 218
F. Soil Analysis Data 223
G. Wood Identification 231
H. C14 Dating of Resin 241
I. Tin Ingot Analysis 244
J. Iron Analysis 257
K. Project Personnel 261
L. Acknowledgments 265

v
1.0 INTRODUCTION
1.1 Objectives and Results
In March 1996 Pacific Sea Resources (PSR) learned of a shipwreck in the Java Sea. The
wreck had a cargo of iron and thirteenth-century Song dynasty Chinese ceramics. It became
known as the Java Sea Wreck.

Through Indonesian partner PT Sulung Segarajaya, PSR was awarded a license to salvage
the Java Sea Wreck in July 1996. During August a barge, tug, equipment and personnel were
mobilized in Singapore. Departure was on 28 August, and after clearing in at Belitung the
excavation vessel was anchored over the wreck site on 30 August 1996.

Excavation took two months. On 29 October the anchors were retrieved and the barge
and finds returned to Singapore, arriving there on 4 November. In a conservation laboratory in
Singapore, the ceramics cargo was cleaned, desalinated, and then systematically shelved for
cataloguing and study.

This Archaeological Report details the historical background at the time of the loss,
describes the archaeological methods that were used to excavate and document the wreck,
and discusses the site, the ship, and the likely wrecking process. It presents the artifacts and
ceramics in inventory form and goes on to study the context of the finds in time and place. The
iron cargo, the most extensive ever discovered on a ship of this era, is discussed in terms of the
history and technology of iron production in Asia.

The ceramics present a wonderful cross section of export wares from southern China.
Over half of the ceramics cargo consists of simple, utilitarian bowls and dishes, provincial ware
for day-to-day use. At the other end of the spectrum are exquisitely molded qingbai (pale blue)
ewers with dragon handles and rare Cizhou-type ewers and bowls, obviously meant for the elite.
In between are dozens of varieties of fine bowls, dishes, vases, bottles, boxes, and jars, in all
shades from white through green to blue.

None of the ships structure has survived. However, from evidence gleaned from the iron
concretions and fragments of wood, it seems that the ship was of Thai origin. The primary cargo
of iron cauldrons and bars was loaded in holds the full length of the ship. Tens of thousands of
ceramics from the kilns of southern China were stowed on top of the iron. The ship then made its
way south along the coastal route, possibly stopping in Vietnam, Thailand, and Sumatra, where it
traded for ivory, earthenware kendis, and aromatic resins. Having passed through the
treacherous Banka Strait, it headed on towards eastern Java where the Majapahit empire was
taking shape. But for reasons never to be discovered, it sank in the Java Sea, well short of its final
destination.

1.2 Locating the Wreck

The wreck lies in the northwest region of the Java Sea. There is no land in sight. There is
no bottom trawling. And yet fishermen found the shipwreck and can return to it again and
again. It is thought to have been first discovered in the late 1980s. It is most likely that birds gave
away the position that had remained a secret for over seven hundred years. Birds feed on
schools of fish that are attracted to the wreck. On an otherwise featureless sandy seabed, a
wreck provides food and shelter for huge schools of fish and large individuals - a place for corals,

1
sponges, and oyster shells to thrive. Birds are always feeding on schools of fish, but most of the
schools are pelagic and therefore move around. Fish over a wreck stay in more or less the same
location. When fishermen observe birds feeding for a long time in one location, it is a very good
sign that a wreck lies below.

Java Sea Wreck location map.

2
 The fishermen are usually more interested in fish than the wreck itself. They keep the
position to themselves, as it provides a continuous supply of red snapper and grouper. Their
ability to relocate the site is remarkable, considering the birds are only there on the rare
occasions that the schools of fish come to the surface. In the case of the Java Sea Wreck the
closest navigation marker is an offshore oil-production platform more than thirty miles away. By
steaming for, say, four hours at 10 from the platform, they come to the area using only a watch
and a rudimentary compass as instruments. Sometimes they go straight to it and confirm the
location by catching reef fish where normally there would only be pelagic species. Otherwise,
they drop a marker buoy and do a slow search using a lead bar on a rope. When the bar hits
an iron hull, a ballast stone, or a piece of porcelain, they can feel it and confirm it by fresh marks
in the lead. In no time they are anchored over the site, pulling snapper in one after the other.

It is normally only a matter of time before a hook snags on the bottom and, when pulled
free, comes to the surface with a lump of iron concretion or a ceramic shard. Many of the
fishermen in the southern portion of the Java Sea live on a few islands in the Thousand Island
Group just north of Jakarta. Several of them are also sea cucumber and aquarium fish divers
and have small compressors rigged on their boats. These divers are also on the lookout for
shipwrecks, which abound in the Indonesian archipelago. World War Two tin cargoes, Ming
blue-and-white, and Song celadons are often recovered by these divers and disposed of very
quietly. As soon as they heard that a ceramics shard had been fished up in the middle of the
Java Sea, they headed for the site to salvage what they could.

As there is no bottom trawling in the area, there must have been large numbers of intact
ceramics on the wreck mound that could be easily recovered without any excavation
equipment, such as airlifts. It is now known that much of the surface material was removed by
fishermen. As an aside, some fishermen have admitted to using explosives on the site to catch
fish before and after ceramics were recovered. The shock waves from the explosion kill the fish
and just about every other living organism in the near vicinity. About a third of the dead fish
float to the surface, where they are quickly snatched up. The other two-thirds sink to the
bottom, where they are left or collected by fishermen with primitive diving equipment. Entire
coral reef ecosystems are destroyed by this very short-sighted technique.

The authorities are aware of the loss of national heritage caused by looting and
dynamiting. Eventually, the navy got word of these activities and arrested the fishermen
concerned. Aware that it was too risky to return to the site, the fishermen sold the position of the
wreck to John Anderson, an American, who had been carrying out survey and salvage work in
Indonesia for several years. Under contract to the Government, Anderson visited the location in
a fishing boat in 1993 and recovered approximately 1,000 pieces during a preliminary
investigation. The salvage company returned with a barge and excavation equipment in 1994
and in less than a month recovered an additional 7,000 ceramic items. The barge almost sank
on the wreck and was towed back to Jakarta for major repairs. Shortly thereafter the company
went into receivership for reasons unrelated to the shipwreck.

In March 1996 Pacific Sea Resources carried out a brief survey of the wreck site, which
suggested that a significant amount of hull structure and undisturbed cargo remained.
Anderson then sold the rights to Pacific Sea Resources and a joint venture was established.
Unfortunately, it was only after excavation was under way that Anderson informed Pacific Sea
Resources that the wreck had been looted previously by enterprising local interests who had
recovered at least 4,000 intact ceramic pieces, and that no hull structure remained.

3
 1.3 Political Considerations

In order to carry out shipwreck excavation work in Indonesia, a salvage license must be
awarded to a local company, by the national shipwreck committee, Panitia Nasional. Pacific
Sea Resources formed the Indonesian company, PT Sulung Segarajaya, through Indonesian
partners to apply for a license.

With Pacific Sea Resources record of the successful archaeological excavation of the
Manila galleon Nuestra Senora de la Concepcion in Saipan, Panitia Nasional readily accepted
the application. Before the license could be awarded, however, approval had to be obtained
from twenty-two Government departments. As soon as these approvals were in hand, the
license was issued and a deposit of Rps. 100,000,000 was paid to Panitia Nasional.

Before work could commence, security clearances had to be obtained for all personnel
and vessels. Even Indonesian nationals, who made up the vast majority of the crew, had to be
cleared before they could work on a wreck site in Indonesian waters.

A security officer representing Panitia Nasional was permanently posted on board the
recovery vessel while excavation was under way. His role was to ensure that excavation was
done in accordance with the license terms and to guard against outside interference, such as
piracy. Every evening the ceramics finds for the day were inventoried and jointly signed off by
the security officer and Pacific Sea Resources. When the excavation was completed and all
export documentation was in order, the security officer accompanied the barge and cargo
back to Singapore.

4
2.0 HISTORICAL BACKGROUND by Dr. John
Miksic

2.1 China's Position in the Nan Hai Trade from 500 B.C. through the
Tang Dynasty (A.D. 906)
2.1.1 Introduction

In the year A.D. 100, a distant observer of the Earth would have seen humankind on the
verge of an explosion of commercial activity. Between the Mediterranean, where the Roman
Empire under the Caesars was at its height, and China, stable and prosperous under the Han
Dynasty, the kingdoms of South Asia were encouraging foreign traders to come to their shores.
Ivory and incense from India were used in Roman palaces. Traders from the Graeco-Roman
regions established outposts along the southern coasts of India, where Roman coins and pottery
have since been found. The British archaeologist Sir Mortimer Wheeler, who excavated several
of these sites, concluded that "it is fair to envisage Indo-European commerce of the first century
A.D. pretty closely in terms of that of the seventeenth century" (Wheeler, 1954: 125). The Romans
established trading companies with powers and procedures similar to those of the Dutch and
British factory systems used in Asia over a thousand years later.

The Indians formed similar groups. The members of the Indian trading corporations were
"related by a common interest in trade that had to pass through a particular center ...[T]hese
traders were given royal charters" (Kosambi, 1959: 282). Indian trading guilds, engaged in the
wholesale commerce of luxury goods, and maintained diplomatic relations with rulers of regions
where those goods originated.

The Chinese, however, were not yet party to this communication. Indeed, the Chinese
maintained an attitude of aloofness and official disinterest until much later, that is, until events
during the early Song dynasty (960-1279) conspired to bring about a change in the policy
toward trade. It was then that China finally integrated into the trading system of the Nanhai
(the Chinese phrase designating the Indian Ocean and South China Sea). But the nature of
China's participation in this ancient and sophisticated maritime trade network was unique. In
order to understand the factors which shaped the trade of the South China Sea in the thirteenth
century, it is necessary to understand the traditional Chinese perspective on trade in general,
and on maritime commerce in particular.

2.1.2 Trade in the Indian Ocean, A.D. 100

Two major Mediterranean sources enable us to construct an outline of the early maritime
trade system of Asia. One is a Geography written by the Alexandrian Greek astronomer, Claudius
Ptolemaeus (Ptolemy), around the year A.D. 100. The second is an anonymous work entitled
Periplus Marae Erythraensis, "Sailor's Guide to the Indian Ocean," written around the same time.
These texts describe the commercial institutions of the Indian Ocean and show that commercial
practices along the trade routes there had already been standardized.

The main commodities known to be shipped over these routes were high in value and low
in bulk. They included such items as spices, various kinds of incense, gold and silver, special textiles
including silk, and items that are now mainly considered oddities but commanded high prices in
those days such items as rhinoceros horn and kingfisher feathers.

5
 There were two basic sectors in the trade network: those in the Indian Ocean, lands "above
the wind," and those in the South China Sea and Straits of Malacca, lands "below the wind." The
traffic along these routes was governed by the regimen of the seasonal winds or monsoons.

It was not possible to sail from one end of the route, such as the Persian Gulf/Red Sea area,
to the other end, in China, in one voyage. The normal practice was for Indian ships to sail within the
Indian Ocean as far east as the Straits of Malacca. From there, the voyage to China was
undertaken in Southeast Asian ships.

According to the normal rhythm of the winds, ships would sail from China to the Straits of
Malacca area in January and February. The voyage from Southeast Asia to China could be
undertaken between June and August. Sailors left India for the Straits between April and August
with the southwest monsoon. It was technically possible to sail to India on the same wind, but most
ships only departed from the Straits for India in December or in January. So just as the ships from
China would be returning to Sumatra or Java with their cargoes of silks and metal, the other ships
would be leaving those islands on their way to India.

Merchants from the two legs of the network would not meet unless they remained in the
Straits for nearly a year. Ships did not usually stay over, but traders could. They changed from ship
to ship in relays.

The third leg of the triangle involved what is now eastern Indonesia, the sources of some of
the rarest spices, specifically cloves and nutmeg. The ships from the Moluccas and Sulawesi usually
left between May or June and October, arriving in the Straits or north Java about two months later.

Until the Song dynasty, shipping in the South China Sea was a monopoly held by Southeast
Asians, of whom the people of what is now Indonesia formed the majority. The people of Java
and Sumatra were the ancient sailors of the world par excellence. Their relatives settled the lands
from Hawaii to Madagascar, off the west coast of Africa. Huge ships known as kolandiaphonta
came from the eastern seas; these seem to have been non-Indian, and were almost certainly from
Java or Sumatra. Indian ships by contrast seem to have been largely coastal vessels.

2.1.3 Emporia: The Treaty Ports

The Graeco-Roman texts cited here use a precise word to designate an international
trading port in the Indian Ocean: emporion, which of course has come down to us in modern
English as emporium. The Periplus mentions 27 emporia in the Indian Ocean, divided into three
classes: "designated," "lawful," and "authorized" (Miller, 1969: 19, Schoff, 1912: 51, n. 1; Warmington,
1928:53). The distinctions between these categories are unclear, but the general term means "a
legal mart where foreign trade is allowed and taxed." Ptolemy, describing the India seas and non-
Roman but well-known territory, intends emporion to mean an authorized coastal mart in the
Orient where non-Roman dues were levied by non-Roman authorities (Warmington, 1928: 107).
Ships calling at other ports might be escorted away under armed guard.

Emporia existed from the Red Sea to Southeast Asia. Graeco-Roman traders resided as far
east as south India. Some Tamil kings even possessed bodyguards from "Yavana" (from "Ionia," an
old name for part of Greece). Rome exported glass and wine to India, and grain was shipped
from the Roman colony of Egypt, but most of the spices and silks had to be purchased with gold
and other metallic money.

6
 As previously mentioned, Roman coins are found in many sites of this first period of large-
scale Asian trade in India. It is interesting to note that Roman emperors as early as Vespasian, A.D.
78, complained that the trade in luxury goods was draining currency out of the empire. The same
problem arose many centuries later, when the British, in order to finance their imports of tea and
porcelain, forced the Chinese to accept opium in lieu of metals.

A few Roman artifacts have even turned up in sites in Southeast Asia, especially at Oc-eo in
southern Vietnam, apparently the major trading center in the South China Sea from the second
through the fifth centuries.

The Roman trading stations in south India survived at the most for about 200 years. Then
they fade out of history, as the empire began its long, slow decline. Even after Rome fell, however,
the name "Rum" continued to shine with a bright luster in Asia. The Byzantine Empire was also
known by this name, and coins from fifth-century Byzantium are common discoveries at sites of this
period in Sri Lanka (Wolters, 1967: 80).

Europe forgot completely about its earlier involvement with Asia for a thousand years and
sank back into the darkness of the Middle Ages, when all that remained of the early long-distance
trade were a few legends of fabulous wealth in the distant East.

The story has an interesting postscript. In 1410, a Byzantine-era copy of Ptolemy's

Geography was discovered in a library in Florence and was translated into Latin. This translation
was studied at the Academy at Sagres established by Prince Henry the Navigator. The information
it contained, although more than a thousand years old, provided a vital stimulus to the planning of
the explorations of the south coast of Africa which culminated in the Portuguese discovery of the
Cape of Good Hope and the route to India. This episode is one of the most intriguing examples of
the use to which ancient classical literature was put upon its rediscovery during the Renaissance,
and the clear connection between the beginnings of European commercial expansion into Asia
(and accidentally into America) and the birth of European humanistic studies.

2.1.4 While Europe Slept

The period that concerns us here is that between the fall of Rome and the arrival in the Far
East of their distant descendants, the Portuguese. During this thousand-year phase of
development, only now attracting historians' attention, the Asian maritime trade fared extremely
well without European involvement or, indeed, awareness. Very few Asian records of this long
interim survive, and those that do tend to be written in archaic or extinct languages and deal
mainly with ritual and politics. The principal exception is Marco Polo, who traveled the entire
length of the Asian maritime trade route during his return from Yuan-dynasty China to Italy in 1292-
93. Polo's descriptions of this world seemed so unbelievable to his contemporaries that for centuries
his tales were thought to be imaginary. Only in the early twentieth century when archaeological
discoveries begin to corroborate his reports did historians begin to take his account seriously.
Although one author has recently challenged the current view that Polo was recounting a true
story, the vast weight of evidence suggests that he was in fact attempting to do no more than
record what he had actually seen and done.

The very earliest reports from the Graeco-Romans depict a system whose framework
continued to exist right up to the twentieth century. This system was highly efficient, given the
technology and communications of the time, in distributing goods over an enormous stretch of
coastline. The main elements of the system included the designation of certain ports as open to
foreign traders and others as closed, the Roman emporia. This characteristic was obviously

7
designed to ensure that the benefits of the trade would accrue to certain people, namely the
rulers of the regions concerned.

In traditional kingdoms the more distant parts of the realm could not be trusted to send the
customs duties collected from foreign merchants to the central treasury. Furthermore, the wealth
accumulated by merchants could well be used to mount a challenge to the established ruling
clique. For these reasons, the main ports where foreign trade was allowed were those most closely
supervised by the principal ruler, who might reside inland, in the center of the more fertile
agricultural zones of the kingdom.

In the port zones, the foreign merchants usually had their own quarters. These would
correspond to the foreign quarters of the treaty ports of nineteenth-century China. Within these
quarters the foreign communities enjoyed more or less complete autonomy, as long as they did not
interfere with the rest of the citizens of the kingdom. This arrangement suited both parties: the
foreigners were allowed to follow their own customs and religion, and the local administration did
not have to worry about trying to control motley groups of people with different languages and
legal systems.

The trading system itself is difficult to define, but in essence there seem to have been two
separate sets of exchange practices for different categories of merchandise. The "rare and
precious" items (in the stock Chinese phrase) were coveted by the wealthy as status symbols, and
the rulers often limited access to them. Some were claimed as monopolies of the ruler himself; in
other cases, he claimed a duty of up to 30%, and the rest of the luxury goods could only be sold to
a designated group of buyers.

The possession of such luxury goods obviously had political symbolism, and the rulers did all
they could to control their distribution. This category of merchandise was usually not bought and
sold in the usual fashion, by a process of haggling until a bargain was struck between buyer and
seller. Instead, a customary or diplomatic procedure was used to negotiate a set of
"equivalencies" between the items brought by the foreign merchants and those returned by the
local inhabitants. The ideology of exchange was more of a generalized than a specific reciprocity,
in anthropological terms. It was not seemly to perceive this kind of exchange as trade. In most
instances the foreign items were presented to the local ruler in the guise of tribute or gifts, and the
local ruler reciprocated by giving presents. The nature of the gifts returned by the rulers was,
however, fixed by custom; a certain amount of foreign produce would elicit a specific quantity of
"gifts." Thus the appearance of trade (and the use of money) were avoided.

The second category of exchange involved more utilitarian items. These are even less well
documented than the luxury items, but stray references to the transport of bulk necessities such as
rice and salt are found in the earliest sources. It seems that these found their way into markets
where money was used. Many forms of currency were used in Asian markets, and coins in some
areas were treated as just another form of metal which might be melted down and recast into
other objects. In general, however, marketplaces in Asia were monetized at a very early period.

Indonesian inscriptions show that weights and measures had become standardized
between A.D. 850 and 900. These weights were heavily influenced by Indian units and linguistic
terminology, but there were some indigenous elements in both units and names.

By the year 900, silver and gold coins were minted in multiple denominations in Java. It is
worth noting that the Javanese coins look nothing like those made in India. The Javanese word wli
(modern Indonesia beli, "buy") first appears in an inscription dated 878. The word pirak, analogous
to modern perak, "silver," was a synonym for money. Indian influence in the form of the Sanskrit

8
word wyaya, modern Indonesian biaya, "expenses," also was current in 878. It seems that in
addition to precious metal, other currency in the form of iron bars was used until the early tenth
century (Wicks, 254-259).

An integral figure in the organization of ancient trade was the official who acted as the
intermediary between the foreign merchant communities and the local ruler. This official had
different designations in each area, but his duties and privileges were remarkably similar in all times
and places. In the Southeast Asian realm he was called the shahbandar (from Persian "Lord of the
Harbor"). In modern parlance this term is used to translate the English "harbormaster," but this gives
only a pale reflection of his activities. In many cases he was chosen from the foreign community,
and to ensure his loyalty he was frequently betrothed to a member of the local royalty. He
supervised the collection of customs duties and the warehousing of the imported goods. He also
acted as the manager for the local ruler's own commercial activities; many Asian rulers themselves
invested in ships and cargoes. They settled disputes between the foreign merchants and rulers,
and could act on their own initiative to make treaties with foreigners. Such people obviously had
numerous opportunities to enrich themselves.

A good description of the system is given by Tome Pires, one of the first Europeans with an
intimate knowledge of its working. He resided in Malacca from 1512 to 1515 and describes the
shahbandars there, of which there were four, each with jurisdiction over merchants from different
areas: "They are the men who receive the captains of the junks.... These men present them to the
Bemdara (the royal treasurer), allot them warehouses, dispatch their merchandise, provide them
with lodging if they have documents, and give orders for the elephants..."(Cortesao, 1944: II, 265).

These characteristics of the early ports of trade in Asia were already in existence when the
veil of history lifts during a short period around two thousand years ago. One can even trace their
origins further back in time; the oldest documents on trade, from Mesopotamia, already indicate
the existence over 4,000 years ago of officials in charge of long-distance trade, officially
designated trading ports, tributary trade in luxury items, and foreign quarters. This is not to suggest
that these traits owe their existence to Mesopotamian influence; rather they simply reflect a logical
approach to the similarity of conditions prevailing in these different times and places.

2.1.5 Early Chinese Trade

The Chinese were not among the original members of the Indian Ocean trading network.
When the Graeco-Romans first entered the Asian world, the Chinese were still largely confined to
the Yellow River area of northern China. The main route to China at this time was still overland, via
the Silk Route. For many centuries China's northern frontier had been a zone of constant
contention between the settled agrarian way of life and the nomadic adaptation of the steppes
beyond. The nomads were always a military threat to the Chinese.

One of the tactics that the Chinese used to counteract the nomads' military superiority was
to encourage exchanges of prestigious goods across the frontier. The Chinese produced
numerous manufactured items that the nomads desired; by dangling exotic and rare commodities
in front of the northern barbarians, it was sometimes possible to bribe them into maintaining
peaceful relations. Domestically, the imperial dignity could be maintained by casting the gifts to
the nomads in the form of a ruler-vassal relationship.

The nomad elite received imperial gifts in return for their tribute, recognition of political
autonomy in their own territory, and permission to conduct limited private trade with the Chinese.

9
This gave the people of the steppes access to Chinese silk, tea, and porcelain. The Chinese
obtained jade and, importantly for their military position, good horses.

The earliest evidence for foreign trade between Chinese and non-Chinese groups dates
from the fifth century B.C. It involved exchanges between the Han people living in Shandung and
the people occupying what is now southern China, known as the Yeh. The carriers of the trade
were Yeh people living around mouth of Yangzi. China south of the Yangzi became Sinicized
only after the fall of the Han, when many people from the north migrated south.

Between 500 and 221 B.C., the late Zhou Dynasty, demand increased among the
(northern) Chinese for southern luxuries like ivory, pearls, tortoise shell, kingfisher feathers, rhinoceros
horns, scented woods, and spices; "this demand may have been the strongest single motive for the
southward expansion of Chinese political power" (Wang, 1958: 5). Qin Sihuang, the "first emperor,"
was the first ruler to extend northern Chinese control to parts of the south coast, bordering on the
South China Sea. At this time the entire south coast was still populated by Yeh people. "The few
prehistoric sites along the coast which have so far been examined merely suggest that the people
were quite different from the Chinese and had a different cultural pattern" (Wang, 1958: 7).

The Yeh, unlike the northern Chinese, were expert sailors and already frequently traded
with other regions. One kingdom, Nan Yeh, with its capital near modern Guangzhou, was
described in the first century B.C. by the contemporary author Su-ma Ch'ien as "the collecting-
center of rhinoceros horns, elephant tusks, tortoise-shells, pearls, fruits and cloth (of hemp and other
fibrous plants)" (Wang, 1958: 8). Lo Yeh, another kingdom located in the Red River basin near
modern Hanoi, was captured in 214 B.C., but not until Chinese troops had been defeated several
times and reinforcements had been sent, consisting of "criminals, banished men, social parasites
and merchants" (Wang, 1958: 10). This interesting set of associations for merchants shows the low
esteem in which they were held. Who were these early merchants exiled to the southern colonies?
No information exists.

According to Confucianism, merchants were the lowest rung on the social totem pole. The
documentation of Chinese economic history is therefore extremely poor. Because Confucian
scholars monopolized the bureaucracy, owing to an education system predicating access to high
office on memorization of the Confucian classics, Chinese history was written by officials who
considered trade a degrading subject, not worthy of notice. However, indirect references, usually
complaints from ministers, show that by 178 B.C. Chinese people were becoming interested in
making profit from the luxury goods trade. A rising class of merchants encouraged powerful nobles
to invest in trade in salt, metals, and luxury goods. Meanwhile ministers praised rulers who "despised
gold and jade" and wanted the emperor to make everyone return to farming.

The relationship between the governing elite of China and the mercantile classes has been
marked by suspicion and hostility for over two thousand years. One of the main characteristics of
Chinese history has been the tension between the culture of the north, the original center of
Chinese culture, marked by militarism, austerity, isolationism, and centralized rule, and the culture
of the originally non-Han Chinese south, marked by a spirit of free enterprise, commercialism, social
mobility, and outward-lookingness.

Four military commanderies were set up to administer the newly conquered territory,
corresponding to the regions of Fuzhou, Nanhai (Guangzhou), Guilin, and Xiang (Hanoi). "Very little
is known of these commanderies" (Wang, 1958: 10) other than that they had a governor and
military staff.

10
 The commandery of Nan-hai soon split off and became a "miniature empire," Nan Yeh,
which ruled the coast from southern Fujian to Vietnam until 111 B.C., when the Han retook the
area. In 196 B.C. the chief was made a feudal lord by the Han empire, and a regular trade with the
Han developed, mainly based on the acquisition of iron in exchange for pearls, tortoise shell, ivory,
and rhinoceros horncommodities that were imported from the South Seas.

Han Wudi, an energetic emperor who died in 87 B.C., allowed trade, but after his death the
ministers had their way and trade was restricted. The court debate on the morality of trade is
preserved in a collection of essays entitled "Discourses on Salt and Iron." In 77 B.C. a Yeh native
who had become head of a commandery was executed for having earned more than one million
cash in illegal trade.

The "rare and precious objects" from beyond China played an important part in the
legitimization of the ruling dynasty. As Chinese civilization developed, so did the search for more
unusual and exotic items from abroad, until something resembling a cult of imported rarities
evolved. The imperial court formulated a policy of sending missions abroad in search of rare and
precious goods. This duty often was assigned to eunuchs. An early text refers to "chief interpreters
attached to the Yellow Gate [eunuchs serving in the palace] who go to sea with the men who
answer their appeal [for a crew] to buy bright pearls, pi-liu-li (opaque glass), rare stones and
strange things, taking with them gold and various fine silks to offer in exchange.... The merchant
ships of the barbarians are used to transfer them [the Chinese] to their destination" (Wang, 1958: 19-
20). These emissaries embarked from ports in western Guangdong, where few Chinese lived; they
had already been active as centers for the pearl trade. The eunuchs may have been barbarians
themselves.

In the later Han Dynasty, in the second century, barbarians began sending their own
seaborne missions to China. One supposedly from Daqin (Rome) came in the year 166, but was
rejected by the Chinese as spurious. The envoys brought elephant tusks, rhinoceros horn, and
tortoise shell as tribute; these were South Seas produce, not Roman. This record is, however, useful
as the only account of what any missions brought to China at this time.

The Han collapsed in 220 and was succeeded by a period known in general as the Three
Kingdoms, celebrated in Chinese literature. One of these three kingdoms, Wu, held the southern
coast; the other two, Wei and Shu, held the north and west. The Wei therefore had to deal with
the Wu in order to obtain southern luxuries. In 236 the Wei sent envoys to Wu to obtain pearls,
kingfisher feathers, and tortoise shell in exchange for horses.

The Wu kept sending missions into the South Seas in search of these items. Two of these
envoys, Kang Tai and Qu Ying, made a special mission to the South Seas in 245250. They both
wrote books, long lost, about their mission; some quotations from Kang Tai have been preserved in
other texts. He mentioned Java/Sumatra, which he apparently visited, as well as Bangka, Belitung,
and Borneo in all, ten places on the Malay Peninsula and in the Southeast Asian archipelago.
During the Three Kingdoms period Chinese ships were apparently still incapable of sailing in the
open sea; the emissaries traveled in foreign ships.

2.1.6 The Tribute Missions

The kingdom of Wu was conquered by Qin, A.D. 280. More men were sent to sea to
acquire the standard South Seas luxuries, and foreign tribute missions to China accelerated. In the
short period of 284287 Funan, Linyi (Champa), and twenty other countries came to present tribute.
An image of the wealth which this commerce generated found its way into the literature of the

11
period. The Governor of Qing Zhou, on the route between south China and Loyang, the northern
capital, was famous for flaunting his wealth, including coral trees, ivory, pearls, and scented woods.
One account says that he had "powdered gharu-woods as fine as dust sprinkled over an ivory bed,
and asked those that he specially loved to step on it" (Wang, 1958: 35).

Trade routes and tributary kingdoms of Southeast Asia in the fifth to seventh centuries A.D.

During the Wu and Qin dynasties, 226-405, China received missions from the following
countries: Funan (7 missions), Champa (6), Daqin (the Eastern Roman Empire) (1), Sri Lanka (1), and
Tangming(?) (1) (Wang, 1958: 120). Shortly thereafter, before the beginning of the next major
historical phase, the famous Chinese Buddhist monk Faxien returned to China from India by sea
rather than by the more common overland route. In 414 he called at Yeh-po-ti, a large trading
port somewhere near the south end of the Straits of Malacca (quite possibly in northwest Java).
From there he sailed directly to Guangzhou with over two hundred other people. During all his

12
years in India he saw no Chinese, nor were there any among the merchants on the ship taking him
to China. This is the first record of a sailing connection between India and China; all indications are
that Chinese merchants were not participants in this trade.

The next period of important tributary activity occurred during the Southern Dynasties, 420
589, when a large mission visited China "to ensure safe and profitable trading for the foreign
merchants and trading envoys" (Wang, 1958: 38). In all, 99 foreign tribute missions visited China
between 420 and 589 (Wang, 1958: 51), a great increase over the previous period. The flow of
tribute to China was not consistent, however; rather it came in spurts. The pattern of tribute exhibits
two peaks, one during the years 420460, the other in the period 502540.

The Southeast Asian countries active in this tributary trade include the following:

Holodan, in Java, sent 6 between 430 and 440.

Pohuang, Tulangbawang, southeast Sumatra: 7 between 445 and 464.
Gantoli, probably south Sumatra, 455564, 5.
Poli (Bali?): 3, 470524. Poli's products were dominated by those obtained from the sea:
tortoise shells, shell-fish, purple cowries, and corals.
Panpan, somewhere on Malay Peninsula, 455589, 12.
Langkasuka, Malay Peninsula, 51556, 3;
Champa, 420589, 25. There is a record of the types of commodities brought by this most
active tributary. These include metals: gold, silver, and copper (there were said to be lots of
gold and silver articles produced there); also tortoise shell, cowries, gharu wood, grass mats,
cotton cloth, rhinoceros horn, and ivory. On the whole, this list represents the common items
sought by the early Chinese nobility.

Funan, 430589, 17. Trade goods from Funan included some items for religious rituals like
gharu wood (incense), ivory and sandalwood stupas and statues, and glass vessels used for
temple rituals. The items meant for use by the nobility (as opposed to the temples)
consisted of gold and silver articles, cowrie ornaments, scented woods, ivory, peacock
feathers, tortoise shells, re-exports of items like gems, coral, opaque glass, cotton, and
storax. Chinese "gifts" to these areas list only silk and brocade.

Although the Southern Dynasties were an active period of communication with the South
Seas, the records of this time still exhibit no references to Chinese ships going overseas. There are,
however, rare pieces of evidence which enable us to read between the lines and to conclude
that certain Chinese were exhibiting a "modern" attitude toward the benefits of engaging in
foreign trade. In the 479502 period, there is a reference to a certain Zhang Qing-zhen who
"calculated carefully the silks and brocades which he used to trade with the 'K'un-lun p'o'" (Wang,
1958: 60). The phrase "K'un-lun p'o" refers to the ships of the people who lived in the South Seas.
Apparently there was a mercantile class in south China, but other than their bare existence, no
other information regarding them survives.

Two other references are even more indirect, but betray something of the true nature of
the trading activity and its importance for at least some sectors of Chinese society. In one
reference to the Liu Sung Dynasty, one of the southern dynasties, the following note appears:
"When the two Han dynasties had sent expeditions these [overland] routes had been found to be
particularly difficult and merchandise, on which [China] depended, had come from Tongking; it
had sailed on the waves of the sea.... Precious things come from the mountain and the sea by this
way. There are articles such as rhinoceros' horn and kingfisher feathers and rarities such as serpent
pearls and asbestos; there are thousands of varieties, all of which the rulers eagerly coveted.

13
Therefore ships came in a continuous stream, and merchants and envoys jostled with each other"
(in Wolters, 1967: 77).

The history of the dynasty that succeeded the Liu Sung and ruled from 479 to 502 also refers
to the active maritime trade of that time: "Of all the precious things in the world none are better
than those of the southern barbarians. They are hidden in the mountains and in the seas. They are
innumerable. Merchant ships arrive from afar and bring these things to the southern provinces.
Thus it is that Tongking and Kuangtung are rich and well stocked. The goods are stored in the
imperial treasury."

Very little information regarding the administration of the trade of this period survives. A
mission from the Javanese kingdom of Holotan in 430 was partly intended to acquaint the Chinese
emperor with the difficulties that the Javanese merchants sometimes experienced with corrupt
port officials (Wolters, 1967: 165). This indicates that the collection of duties was not always properly
supervised by the central government.

The Southern Dynasties came to an end with the reunification of China by the Sui Dynasty
in 581. The emperor in 605 moved his capital and ordered the provinces to deliver such southern
luxuries as rhinoceros horns, elephant tusks, furs and feathers, to adorn his new court (Wang, 1958:
63). The short-lived Sui recorded the following missions: Champa (2); Dandan (2); Panpan (1), Red
Earth Land (3), Cambodia (1), Jialoshi(?) (1) (Wang, 1958: 122).

2.1.7 Trade During the Tang Dynasty

After the establishment of the Tang Dynasty in 618, the southern ocean trade expanded. In
addition to luxuries for the court and goods of a religious nature, items used for medicinal purposes
and condiments for food formed an increasing portion of the imports from the South Seas (if the
sources are a true reflection of the situation). Yangzhou, at the junction of the Yangsi and the
Grand Canal, became the main center of trade. A road built in 728 connected Guangzhou with
Yangzhou. A source says that "the various countries from across the sea may now daily transport
their merchandise, so that the wealth of tusks, hides, feathers and hairs, and that of fish, salt, clams
and oysters can...meet the needs of the treasury and...satisfy the demands of the Qiang-Huai
region" (Wang, 1958: 79).

Through the eighth and into the early ninth century, trade continued to expand under the
energetic Tang rulers. Guangzhou in particular benefited from this expansion. As a source from
841 says, "Guangzhou enjoyed the profits of the barbarian ships where all the valuable goods were
gathered.... Of all those who served at Guangzhou, not one returned without being fully laden
[with the wealth they acquired]" (Wang, 1958: 83).

The missions which visited China more than once during the 278 years of the Tang Dynasty
are as follows: Champa (26), Holing, Java (8), Sri Vijaya (6), Shih-tse (4), Cambodia (3), Qulomi (2),
Jambi (2), Doholo (2), Kanqifo (India) (2), Mola (India) (2) (Wang, 1958: 122123).

Even during the Tang, there are still no references to private Chinese traders voyaging
overseas. The only descriptions we possess are the incidental references in court records, which
tend to overlook commercial matters completely, and the record of another Buddhist monk-
adventurer, Yijing. He voyaged from China to Srivijaya, south Sumatra, in 671, in a ship belonging
to the ruler of Srivijaya. He remained in south Sumatra for 6 months, studying Sanskrit at a large
monastery there; his experience was positive enough that he recommended it to other future

14
Buddhist travelers. He then took ship in another Srivijayan vessel and sailed to India via Malayu
(Jambi) and Kedah.

After residing in India for 18 years, Yijing returned to Srivijaya, noting that in the interim it had
absorbed both Malayu and Kedah. He intended to remain in Srivijaya for some time, but one day,
while he was aboard a ship in the harbor in order to send a request to China for more paper and
ink, the ship unexpectedly weighed anchor and sailed straight to China, leaving him no option but
to go along and buy his own supplies. Nevertheless he so desired to spend more time in Sumatra
that he returned there for some more years before finally returning for good to China in 695.

Yijing's account is useful because it mentions other Buddhists who made the pilgrimage to
the Buddhist holy land, including a monk who sailed in a merchant vessel "heavily loaded with
goods" from Guangzhou or Hanoi to Holing. The ship then went to Malayu, but after passing it sank
in a storm because it was overloaded; the monk drowned. No doubt this incidental report
represents the tip of the iceberg; voyaging between South Sumatra and China by Indonesian ships
carrying a wide variety of goods from both Southeast Asia and the shores of the Indian Ocean was
probably a common activity during the Tang period.

The report of Yijing is of utmost importance to the study of early Tang maritime trade,
because it is the only description of the sea routes used before 750. He describes direct voyages in
Indonesian ships from Guangzhou to Palembang or (less often) Java. From there the standard
ports of call were Malaya (Jambi) and Kedah. Return voyages followed the same route in reverse.

In the late seventh or early eighth century, Srivijaya "sent several missions to the court to
submit complaints about border officials seizing [their goods] and an edict was issued ordering [the
officials at] Guangzhou to appease them [by making inquiries]" (Wang, 1958: 97, quoting the Tang
shu). "This is the only mention in Tang records of any mission from the Nanhai successfully inducing
the central government to act on behalf of the merchants at Canton. This is evidence that
Srivijaya was the dominant trading power and had already earned the respect of the Chinese"
(Wang, 1958: 98-99). Srivijayans were leaders of the foreign merchant community at Guangzhou
until 742.

In 684 the governor of Guangzhou was killed in manner which suggests that the murderer
was an Indonesian distressed by officials' misdeeds (Miller, 1969: 186). Some such incident may
have been responsible for the remark of Tome Pires over 800 years later explaining why foreigners
were not allowed to go to Guangzhou: "They say that the Chinese made this law about not being
able to go to Canton for fear of the Javanese and Malays" (Cortesao, 1944: I, 122).

In 714 we first learn of the existence of a Chinese official called the Superintendent of
Shipping Trade (alternatively called Superintendent of Barbarian Shipping). This is the first sign that
an official like the shahbandar of the Indian Ocean was appointed in China. The office was rather
independent of the provincial authorities, being administered directly from the central
government, and was dominated by eunuchs. A record from the early ninth century gives some
indication of this official's duties:

When [the laden Nanhai ships] arrive, a report is sent to the Court and
announcements are made in all the cities. The captains who command them [or
chief merchants] are made to register with the Superintendent of the Shipping
Trade their names and their cargo. [The Superintendent] collects the duties on the
goods and sees that there are no [prohibited] precious and rare goods [of which
the government had a monopoly]. There were some foreign merchants who were
imprisoned for trying to deceive [him]" (Wang, 1958: 101). Another source says that

15
 head of the office "dealt with the translation of languages, the offering of valuable
gifts, and every year conducted the sending of tributes (either sending those due
from the province itself or arranging for foreign tribute missions to go to the capital).

After the last ship of the season arrived, 30% of the non-monopolized goods would be taken
as duty, and the rest would be given back to the envoys to dispose of themselves (Hirth and
Rockhill, 1911: 15).

Provincial authorities continued to have some duties too in regard to the supervision of
maritime trade. A governor of Lingnan in 820 was cited for halting smuggling: "When the foreign
ships arrive and are docking, they are charged a lowering-anchor-tax. There is an examination of
the merchandise. Rhinoceros (horns) and pearls were so numerous that bribes were offered to the
servants and retainers; the Governor stopped this" (Wang, 1958: 101).

The same record says, "Far across the sea in the South, there were those who died in the
countries there. The officials held their goods. And if their wives or their sons did not come within
three months to claim them, these would be confiscated. The governor [stopping this practice]
said 'The sea journey back and forth is calculated in years; why fix the time in months? If anyone
has proof, no matter whether he comes early or late, let him have all" (Wang, 1958: 101-102).

In the ninth century another port of trade was opened to foreigners: Quanzhou (also called
Zaitun, an indication of the large number of Muslim traders who frequented it).

In the time of the succeeding Song Dynasty (9601279), similar practices apparently
existed:

Concerning foreign ships and merchants, the Superintendent examines the boats
entering the harbour for 'prohibited goods,' takes into the godowns all the legal
imports, collects the taxes due on these goods, buys on behalf of the government
those goods of which it has the monopoly, and examines the boats leaving the
harbour for 'prohibited goods'. He further protects the foreign merchants while they
are at the port.

Concerning Chinese ships and merchants, the Superintendent examines the cargo
of the ships when they leave for the Nanhai and when they return, and collects the
taxes due on their goods.

In 878 this period of prosperity came to an end. Guangzhou was pillaged during the
Huang Zhao rebellion, foreign merchants were murdered, and all trade there ended for a century.
Foreign merchants then established a rendezvous at a place known to the Arabs as Kalah (which
appears in the Arabian Nights as a place visited by Sinbad). Archaeological sites at both Takuapa,
on the west coast of South Thailand, and Laem Pho, on the east coast, contain abundant
evidence of trade with the Near East and China at this time and may correspond to this toponym.

Wang, (1958: 113) divides early maritime trade with China into three phases:

Phase One: 200 B.C.A.D. 300. Precious things desired by courts.

Phase two: A.D. 300600. Holy things.
Phase three: A.D. 600900. Drugs and spices.

The brief Five Dynasties period (907959) saw little commercial activity. Only four missions to
China are recorded: three from Janzheng and one from Srivijaya. The foundation of the Song

16
Dynasty in 960 ushered in a new and even more prosperous period of trade, when many more
parties became involved, including for the first time Chinese merchants voyaging into the Nanhai.

2.1.8 Early Southeast Asian Ships

The oldest known boat yet discovered in Southeast Asia has been carbon dated to A.D.
260430 (Manguin, 1993: 236). It was discovered in Pontian, on the east coast of the Malay
Peninsula. This date agrees with that estimated in 1926 by I.H.N. Evans based on ceramics found on
the site with the boat, which are similar to those discovered at the contemporary port of Oc-eo in
south Vietnam. The origin of the boat cannot be determined, but its lashed-lug design marks it as
Southeast Asian rather than Chinese. It seems to have been a trading vessel, as its cargo suggests.
The ship is estimated to have been about 12 meters long.

Few other early boats have been discovered in Southeast Asia. One other, which has not
been scientifically dated but which resembles the boat from Pontian, was discovered near Khuan
Luk Pad, in south Thailand, a site of A.D. 1500 that has been systematically looted because of the
huge quantity of early trade beads discovered there.

Two important finds, although found in disturbed condition, were both made near
Palembang, Sambirejo, carbon dated to A.D. 610775; and Kolam Pinisi, Palembang, A.D. 434631.
The Sambirejo ship is estimated to have been 26 meters long. From a later period, probably the
late Song, is a fragmentary vessel from Paya Pasir, near Kota Cina, northeast Sumatra. This too was
a relatively large vessel like the Sambirejo ship. Finally, the only other important vessel of the pre-
Ming period to be discovered and studied was found at Butuan, Mindanao, south Philippines, and
dated to the period A.D. 12701410; its original length is estimated to have been 20 meters
(Manguin, 1993).

These ships can only be partially reconstructed, so no detailed comparisons are possible.
But some basic construction techniques can be discerned. One of the most important traits of
these early Southeast Asian ships is that no iron was ever used in their construction. Instead their
planks were lashed together with vegetable fibers. The use of outriggers would seem to have
been another important early characteristic; large seagoing ships of this type are depicted on the
Javanese monument of Borobudur, carved around 800. Dr. Manguin (1993: 264) does not think the
largest ships would have used them. The use of quarter rudders such as these reliefs indicate is
highly probable, but some early Chinese ships may have used them too.

A fourth-century Chinese text on medicinal plants refers to foreign ships called bo,
analogous to the K'un-lun po of other writers. The word po would seem to be a transliteration of
the Malay perahu, "boat." These ships were over 50 meters long, carried four masts, and had a
capacity of 600700 people and about 600 tons of cargo (Li, 1979: 90).

Chinese influence, such as the use of iron nails, seems to have become common in
Southeast Asia in the fourteenth through the sixteenth centuries. The shipwreck at Bukit Jakas,
Bintan Island, Riau, Indonesia, carbon dated to between 1400 and 1460, is an early example of the
hybrid Chinese-Southeast Asian variants that became standard in the period after the Chinese
began to participate as shippers in Southeast Asian maritime commerce. It is significant that this
influence only appears after the end of the Song Dynasty; it is another strong piece of evidence
supporting the inference that Chinese shippers did not participate in the trade until after the year
1000.

17
 2.2 Trade Interaction During the Song Dynasty (A.D. 960 - 1260)
2.2.1 Introduction

In 960 the Song Dynasty reunified China after a period of civil war, and commerce in the
ports where foreign trade was allowed quickly recovered. A decade later the Department of
Foreign Trade at Guangzhou was reorganized as commerce was expanding at a rapid pace.
The Song helped to stimulate this recovery by sending four missions abroad in 987. These missions
"consisted of eight court officials, who carried with them imperial edicts, gold, and cloth, to
various barbarian countries in the South Seas to induce the import trade of aromatics, rhinoceros
horns, pearls, and Baroos camphor (so-called after Barus, north Sumatra, source of the highest
quality camphor)" (Ma, 1971: 33).

The Northern Song opened new ports, equivalent to the treaty ports persisting into the
nineteenth century. The first alternative port to Guangzhou was Quanzhou, in 1087. Other
Offices of the Maritime Trade Superintendency (Shih-po Si) were located at Hangzhou and
Ningbo. Later, Offices of the Maritime Trade Bureau (Shih-po Wu) were located at Suzhou,
Wenzhou, and Jiang-yin Chun. Although little quantitative data exists to make a firm judgment,
enough anecdotal information has been preserved to support the conclusion that "by the Song
period, the scale of maritime trade had become so large that it may be deemed the first period
of great oceanic trade in the history of the world" (Ma, 1971: 23).

Port cities of Song dynasty China.

18
 It was still a criminal offense for individual Chinese to engage in direct trade with
foreigners, however, and the system of hoarding all foreign incense and other luxury goods in
government warehouses was still maintained. Thus for the period between 960 and 1126, known
as the Northern Song, foreign commerce was continuing to expand, but still within the mold cast
by the Confucianists a thousand years earlier.

Merchants who induced foreigners to bring cargoes that yielded duties of over 50,000
strings of cash were offered official rank. In 1115 the Chinese government established a hotel at
Quanzhou for foreigners, which cost the government 300 strings a year. In 1132 another hotel
was founded at Guangzhou. A welfare service was set up for shipwrecked seamen, with an
allowance of 50 cash and 2 pecks of rice a day.

As early as the late Tang Dynasty, the government had begun to invest in maritime trade
infrastructure by developing a harbor near Fuzhou. The Guangzhou harbor was dredged at
government expense in the early eleventh century (Wheatley, 1959: 26-27).

2.2.2 The Song Government Trade Monopoly

Imports via maritime trade in the Northern Song still consisted of two main categories: the
less expensive, consisting of textiles (mostly cotton), spices, and drugs, and the much more
valuable: jewels, ivory, rhinoceros horn, ebony, amber, coral, aromatic products and perfumes.
The sale of the luxury goods was still a government monopoly; only licensed dealers could buy
them at government warehouses in Quanzhou, and to a lesser extent at Guangzhou, in fixed
quantities and at fixed prices.

Licensed goods were acquired by the government as import duty (all duties were paid in
kind, not in cash) or through purchase by the superintendent of merchant shipping. For this
purpose he was allocated about 100,000 strings of cash a year in the late eleventh century, and
300,000 strings in the early twelfth century. Interestingly, the capital for this activity was obtained
by taxing priests' diplomas. The returns on this investment were substantial; the Quanzhou office
between 1128 and 1134 made a profit of 980,000 strings (Rockhill,1914: 421, n. 1.)

The government attempted to enforce an imperial monopoly over the import and
possession of eight items: tortoise shell, elephant tusk, rhinoceros horn, a special kind of steel
used for weapons, skin of a lizard used for making drums, coral, agate, and frankincense (Ma,
1971: 37-38). Nevertheless, indirect references suggest that the reality did not correspond very
closely to the ideal which the laws sought to impose. Again, the records that were preserved
obviously show a very inaccurate picture of what was actually going on.

2.2.3 The Currency Drain

Already in 1074 an official named Chang Fang-p'ing mentioned seagoing junks that
were leaving China on their return voyage with full cargoes of cash, so that "the currency was
drained off like the waters of the sea into the wei-lu" [literally, rear gate] (Rockhill, 1914: 422). By
the mid-twelfth century, illicit trade in expensive merchandise was so great that Chinese
smugglers were paying for all goods with gold, silver, iron, and especially copper cash, and this
drain was causing the Chinese government serious concern. By 1159 only one-tenth of the
money intended to be coined was actually cast. This discrepancy was generally attributed by
official accounts of the time to illegal foreign sea trade.

19
 In 1194 the governor of a military district in Fujian refused to allow people under his
jurisdiction to go abroad to trade with foreign people "whose many ships coming from abroad
laden with aromatics, rhinoceros horns, ivory and king-fishers' feathers were already draining all
the copper cash out of the land." Sumptuary laws were passed repeatedly (in 1107, 1157, 1201,
and 1214) against use of kingfisher feathers and gold for ornaments, but to no avail. In 1248, a
"Censor, Ch'en Ch'iu-lu, attributed the drain of cash out of China to the extravagance of its
people in purchasing such luxuries as perfumes, ivory, and rhinoceros horns, and to the sea-trade
generally" (Rockhill, 1914: 423).

Thus sea trade was a double-edged sword; rather than providing additional resources for
the state coffers, it seems to have had the opposite effect. This did not, however, prevent the
rulers from dreaming that maritime trade might provide a solution to their serious financial
problems.

2.2.4 The Southern Song Dynasty: Trade for Revenue

The ancient strictures on foreign trade maintained by the Chinese government began to
weaken when, in 1126, the Song rulers were defeated by the Khitan nomads. When the northern
heartland of China was lost to the invaders, the Chinese court escaped and moved south, to
Hangzhou. The port of Quanzhou, the nearest to Hangzhou, eventually surpassed Guangzhou as
the largest entrepot in China. Not only was the court now cut off from the overland route to the
west, and conversely nearer to the centers of maritime trade with the South Seas; now the
potential capacity of maritime commerce to augment the coffers of the kingdomin order to
defend the remnants of the Song lands against the northern invadersbecame a matter of
significant moment to the empire.

Thus official trade expanded, partly by design, in order to secure additional revenue. The
Emperor Gao-zong in 1137 issued an edict on the subject, stating, "The profits from maritime
commerce are very great. If properly managed, they can bring a million (strings of cash). Is this
not better than taxing the people?" (Ma, 1971: 34).

In fact Gao-zong's projections were based on available statistics; in 1128 maritime trade
had already yielded customs duties of 2 million strings of cash, which was 20% of the
government's entire revenue.

With the establishment of the Southern Song Dynasty, there came a revolutionary
development: for the first time in history, private Chinese were allowed, even encouraged, to go
overseas to trade. The entire structure of the maritime trade system of the South Seas underwent
fundamental changes as a result of this new policy. Status as an official tribute-bearing country
was no longer a prerequisite for conducting commercial relations with China. Trade was freed
from its link to diplomacy. The tribute system which had regulated contacts between Southeast
Asians and Chinese declined significantly in importance.

Even the ostensibly aloof attitude of the nobility toward trade, behind which often lay a
very different code of conduct, eroded significantly. "In the late Southern Song period, the
relatives of the Song royal family were vying with private enterprises in Quanzhou.... They used
their political status to toy with the maritime trade rules to earn much profit from maritime trade.
Hence they were very unpopular with the merchant community" (Kwee, 1997: n 35).

The decline of the tribute system probably contributed significantly to political changes
in the South Seas. In particular, the advent of Chinese merchants and shippers in Southeast

20
Asian waters broke down barriers that had fostered the prosperity of a few centralized trading
ports. The empire of Srivijaya, based in south Sumatra, had received a severe blow from the
Cola invasion of 1025; the appearance of Chinese ships in Southeast Asian waters probably
sounded the old thallasocracy's death knell.

The upsurge in Chinese shipping activity did not, of course, mean that Southeast Asian
society suffered economically as a whole; in fact the converse is almost certainly true.
Numerous new port sites, marked by abundant shards of Song ceramics, date from this time,
suggesting that prosperity probably increased for the Southeast Asians as commerce grew.
Moreover, the new wealth was probably more widely distributed. New ports gave traders direct
access to Chinese merchants, bypassing the Srivijayan rulers who would have raked off most of
the goods in dues and fees.

The complexity of maritime trade in the early Southern Song is well indicated by an
enormously varied inventory, dated 1141, which included 339 types of imports. The most
important in terms of value as well as volume were still aromatics and drugs: frankincense,
ambergris, liquid storax (a kind of resin), gardenia flowers, pucuk , myrrh, cloves, nutmeg, and
sandalwood. These were not trivial commodities; they were commonly used for a wide range of
purposes. In addition to their religious uses, various kinds of aromatics were needed in the
household to perfume clothes and bathwater and serve as wall decorations and in the
preparation of food. One of four imperial warehouses was used solely to store incense and
aromatics.

By late Song times, traders had to file official forms specifying destinations and needed
guarantors "who assumed full responsibility in case the traders violated the trade laws." Crew
members of trading ships usually engaged in small-scale maritime trading. They were organized
into 5-member units, and had some sort of "papers" issued for them (Kwee, 1997: 16).

Some Chinese had probably been residing in Southeast Asia before 1126, but it is unlikely
that any accurate accounting of them will ever be made. Since it was potentially a capital
offense to disobey the laws against private overseas trade, those who flouted the law would try
to ensure that their acts would not be recorded. One of the earliest pieces of evidence that
Chinese were living overseas for long periods appears in 1150, when a Chinese and some "dark
natives" were shipwrecked, apparently while trying to sneak back into China. The Chinese had
lived in Indonesia for a long time and had an Indonesian wife (Hirth, 1917: 76).

As might be expected from the clandestine nature of the early trade and the intensity of
the smuggling practiced even after private trade was no longer an offense, little documentary
information exists to illuminate the conditions under which early Chinese shipping was
conducted.

2.2.5 Chinese Ships of the Song Dynasty

Of the few details that can be adduced, one is that ship captains were given a qu-qi ,or
"vermillion pass", on which was written his name and that of his first mate, the number of
passengers, and the size and type of his ship. Ships had about ten oars, each worked by four
men, as well as sails. The sailors were armed against pirates, with bows and arrows. Ships had
two anchors at the bows. There were no cabins; each passenger was allotted a certain amount
of deck space. Zhu Yu of the Song period described merchant ships at Guangzhou:

21
 The ships were several hundred feet long, and wide. Merchants divided space in
the ships for stowing goods, each getting several square feet of floor space, while
they slept above. Most of the goods were ceramic vessels, one placed within
another according to size with little space between...." (Quoted in Li Zhiyan and
Cheng Wen, 1989: 102).

The main vessels towed behind them a small boat used when landing. In the early
twelfth century the magnetic compass began to be used, but texts describe another method of
navigation which made use of a hook on a long rope to bring up mud from the sea bottom,
which captains used to smell and inspect to determine their position (Wang, 1958).

Maritime archaeology is just beginning to contribute new insights into the physical
conditions of the early Chinese maritime trade activity. In MayJuly 1995, an expedition based
on preliminary work in 1990 investigated a shipwreck just north of Fuzhou, in the Dinghai area.
The majority of artifacts recovered (69%) were porcelains. There were also two concretions of
iron. Remains of ship timbers still existed; more excavations are planned, which may shed much
more light on the subject.

A shipwreck at Ningbo, contemporaneous with Bai Jiao, has also been discovered.
Another ship dating from the Song dynasty was found at Quanzhou and excavated in 1974. It
had a deep v-shaped bottom, a true keel, and a stern rudder. The Quanzhou ship was 34.6
meters long, 9.82 meters wide, and displaced 374.4 tons, making it as large as any merchant
vessel then known in the West.

2.2.6 Southeast Asian Markets

As has been noted earlier, some Southeast Asian societies had already become
accustomed to the use of currency before the end of the first millennium. Of these societies, the
most comprehensive data indicating a highly monetized economy comes from Java. Taxes in
Java were expressed as money, not as a proportion of the harvest as was still current practice in
India, for example. Irrigated rice land, orchards, and houses were privately owned and could
be sold, although rights to land not under continuous cultivation were still vested in the village as
a corporate body.

Most villages had periodic markets. Market officials are mentioned in the oldest
Javanese inscriptions from the eighth century (Christie, 1992). The inscriptions also describe two
levels of economic activity: one level was part-time (probably); activities in this category
included dye-making, dyeing, weaving, some pottery making, sugar making, and bamboo mat
making. These occupations were probably conducted as sidelines when the agricultural cycle
or local resources made them feasible.

The second level of activities concerned specialized traders and craftsmen, who could
choose where to live for economic reasons; they were mobile. These people were termed
masamwyawahara, "those who carry on commerce." Some were middlemen somehow
connected to an international network, dealing in imported produce, including tin and other
metals, but they also sold rice. This suggests that there was a relatively large number of people
who bought food on the market.

How did the long-distance maritime trade network interact with the local distribution
system? Sadly, the Javanese sources are no more helpful in answering such questions than the
Chinese. Once more we must turn to archaeology, and once more we find that very little data

22
has been collected. The only important class of Chinese artifact for which distributional data
has been collected in Java is porcelain, and only one regional study has so far been
conducted: in north-central Java. In a study conducted by a Dutch scholar in the 1940s, Song
pottery was found at over 20 sites in Rembang, and in more than a hundred sites in the
neighboring regencies of Semarang, Grobogan, Demak, Jepara, Pati, Kudus, and Blora. The
distribution pattern seems to be correlated with settlement areas and transport routes.

The data indicates that, by the eleventh century, pottery in the area of north-central
Java surveyed by Orsoy de Flines (2,500 square kilometers) was distributed by some sort of
integrated marketing system. The north-central pattern contrasts with the pattern so far
detected in a haphazard fashion for the Tang dynasty ceramics found in south-central Java;
most of the finds of Chinese porcelain there are associated with temple sites. There are,
however, several different variables that might account for this, including different functions for
the pottery, lesser importance attached to ceramics in the Tang, different geographical areas,
and different settlement patterns, and so on. So much more information on the distribution of
Chinese imports in Southeast Asian sites must be collected before the possible effects of
changing Chinese involvement in maritime trade on the Southeast Asian society and economy
can be observed and disentangled from other subsystems evolving at the same time.

2.3 Trade Interaction During the Yuan Dynasty (A.D. 1260-1367)

2.3.1 Introduction

Although a fugitive Song court managed to survive for another 19 years, until 1279, for
all practical purposes the Yuan Dynasty began in 1260, when the Mongols largely completed
the conquest of China. If Chinese economic history can be said to be neglected, the economic
history of the Yuan has been positively ignored. Only one English-language study has been
devoted to the foreign trade of the Yuan, and that work, Schurmann's Economic Structure of
the Yuan Dynasty (1956), contains only one chapter on maritime trade. Yuan trade has also
been largely neglected in Chinese-language sources. A 1955 study by Fang Hao is still the main
work, though its conclusions are now outdated (Kwee, 1997).

Chinese sources tend to conflate the Song and Yuan, so that details about the Yuan are
often obscure or contradictory. Fang Hao, studying Sino-Western interaction in general, argued
that foreign trade during the Yuan was stunted by monopolizing acts of the Yuan government,
that big merchants were suppressed, but that nevertheless trade increased compared to the
Song!

Chen Gao-hua has suggested that Yuan foreign trade was conducted mainly by sea
rather than overland (Kwee, 1997: 2, n. 3), a view that contradicts Reid's generalization (1993:10).
He makes the assertion, briefly treated, that the maritime trade reached a peak in Yuan; so does
Schurmann, also without substantiation. Li Donghua (1984) argued that maritime trade in
Quanzhou during the Yuan was more prosperous than in the Song, but, once again, offered no
evidence (Kwee, 1997: 2, n. 4).

However, anecdotes indicating the wealth of Quanzhou are plentiful. We know, for
instance, that the son-in-law of the superintendent of trade at Quanzhou, who died in 1293, had
80 seagoing ships and 130 pikuls of pearls (Wheatley, 1959: 29).

Such indirect evidence does support the idea that the trade expansion characteristic of
Song times continued to accelerate during the Yuan Dynasty. Moreover, Yuan sources show

23
clearer distinctions among various maritime territories. Such terms as East and West Seas (Dong-
xi-yang), Bigger East Sea (Da-dong-yang), Little East Sea (Xiao-dong-yang), and Little West Sea
(Xiao-xi-yang) first appeared. These are obvious signs of increasing Chinese familiarity with the
maritime world.

Other sources of indirect evidence must suffice to estimate the nature of the expansion
of Yuan trade, no statistics having survived. Schurmann (1956: viii; cited in Kwee, 1997: 5-6, n. 13)
observes that "not all the important economic institutions of the Yuan are covered [in the shi-
huo-zhi section of the Yuan shih]...nothing on tenancy, private commerce, stores, pawnshops,
and manufacturing. None of these institutions directly concerned the government, although
they are important in the economic history of China; only those of direct interest to the
government such as taxes, land survey, maritime grain transport, and monopoly taxes are
treated."

One text, the Da-de-nan-hai-zhi, states that there were "many treasures" in the Yuan
Dynasty, many times more than previous dynasties (Kwee, 1997: 3, n. 7). According to Chen
Gao-hua (1991), there were more than 160 types of maritime goods in the Southern Song, but
more than 220 types in the Yuan (Kwee, 1997: 4, n. 9). The number of foreign polities trading with
China, as well as the variety of produce in China, increased during the Yuan.

By way of interest, a rare instance of the use of maritime trade as a poetic theme occurs in
the work of Sung Pen in the early fourteenth century. "The foreign ships have sailed away from their
anchorage. Year after year they come as if the seas were always tranquil...In the sixth month
when the south wind blows they come, and we greet them with wine and music. Is it not a joyful
occasion?" (Wheatley, 1959:3).

2.3.2 Yuan Government Interests in Maritime Trade

The Yuan established their first maritime trade office in 1277 at Quanzhou, the most
prosperous port of the late Song. "Every year, an invitation was to be extended to foreign
traders to trade in China" (Yuan shih, juan 94; Kwee, 1997). Three more were established shortly
thereafter: at Ningpo (King-yuan), Shanghai, and Kanfu (near Hangchou, Polo's Ganfu). By 1293
there were seven. "So much emphasis was placed on revenue from maritime trade that even
monks, who were often granted privileges, were not exempted from taxes."

The Yuan, being Mongols, lacked the ancient Confucian prejudice against trade. Thus it
is not surprising that they liberalized the system even further than the Song emperors had done.
Official veneration of Ma-zu, Goddess of the Sea, increased. In 1278 "officials memorialized the
throne's calling attention to the importance of encouraging trade relations with the peoples of
the south-eastern [or southern and eastern] islands, all of whom, the writers declared, were filled
with the most loyal devotion to China" (Rockhill, 1914: 429). In 1279 an envoy was sent to Java.
In 1282 an envoy from Java came to China with a shrine of gold as tribute.

The Yuan in 1284 actually instituted a policy of government investment in maritime trade.
The prefects of Hangzhou and Quanzhou chose certain persons to go abroad to conduct trade,
providing them with ships and capital. Net profit was distributed according to a formula of 70%
for the government, 30% for the trader. Seagoers and families were exempted from corve. This
policy was resented by big merchants; its main objective seems not to have been to stimulate
trade but to prevent a few large mercantile families from becoming dangerously powerful by
making it possible for more small operators to compete with them. This new form of government
involvement in maritime trade may reflect the traditional symbiotic relationship between the

24
ortogh and the Mongolians in their original homeland in the steppes. Ortogh ("partners" in Turkic)
were commercial groups, consisting largely of central Asiatic Muslim merchants, who became a
leading class under the Yuan; extensive government funds were loaned to them for commerce
and usury (Kwee, 1997: n. 41). Under the Yuan government, more Muslim tombstones were
erected in Quangzhou than during the Song.

The government even went so far as to forbid the use of private capital in foreign trade
(Rockhill, 1914: 425). However, "The efforts of the government to prevent private trading must
have failed signally, for in 1303 we learn that the prohibition against private sea-trading was
repealed..." (Rockhill, 1914). By then the Yuan government had instituted a maritime trade tax, in
1292 (Kwee, 1997: 15, n. 46), and by 1295 smuggling to avoid the tax and to export forbidden
goods such as bronze coinage was so rampant that officials were sent out to sea to examine
ships.

In the late Yuan period the pendulum of official attitudes toward commerce swung back
and forth. The offices of maritime trade were closed in 1294, but reopened later the same year;
closed in 1303, reopened in 1308; closed in 1311, reopened in 1314; closed in 1320, reopened in
1322. The reasons for these closures are never given in the sources. Scholars have suggested
that political reasons, such as the fear that rich merchants might turn against the state, were
responsible, but it is also possible that these actions were meant to penalize private traders who
violated trading laws. There are references to "22 rules of Yuan trade." In an interesting theory,
Chen Gao-hua has argued that the prohibitions may not have been all that significant, for there
is evidence to suggest that foreign traders still came to Shanghai when the office of maritime
trade there was closed.

2.3.3 The End of Tributary Trade

In 1326 the Tai-ding emperor renounced tributary trade. "In 1329 the presenting for
transmission to Court of expensive and useless objects, all of which had to be paid for at
regulated prices and which were now held to be but a canker devouring the riches of the state
was strictly forbidden."

The Yuan were more expansionist than the traditional Chinese empires had been. They
fought numerous battles on the Southeast Asian mainland, with Vietnam, Champa, Cambodia,
and Burma, often to attempt to force the Southeast Asians to accept a more formal degree of
submission to their overlordship than the Chinese had ever required in the context of the
"tributary trade."

In 1292 the emperor, Khublai Khan, decided to send a large expedition to Java to
avenge a mutilation of his envoy. The naval expedition was placed under the governor of
Fujian, using ships requisitioned from private traders, there being no formal Chinese navy. In 1293
the fleet sailed from Quanzhou, but the expedition became embroiled in a confused political
situation in Java, was ultimately betrayed by an erstwhile Javanese ally, and sailed back the
same year. Khublai Khan died in 1293, and relations with Java soon returned to normal;
Javanese missions arrived at court in 1298 and again in 1300.

2.3.4 Chinese Trade in the Moluccas

One of the more interesting questions about Southeast Asian commerce in the Yuan
period concerns the breadth of the area over which Chinese sailors actually ranged. One of the

25
prizes of the maritime tradeone which attracted such early explorers as Columbuswas
access to the mace and cloves of the Spice Islands, the Moluccas. Some authors have argued
that the Chinese actually reached the Moluccas as early as the Yuan Dynasty, while others have
rejected this idea.

The Yuan dynasty author Wang Dayuan (Rockhill, 1914: 25960) and early Portuguese
sources say that Chinese traders once visited Ternate and Tidore to buy cloves.

Barros held that the Moluccans lived like savages until Chinese junks began to arrive to
buy their cloves, providing in exchange the Chinese cash that became their major currency.
Eventually "the Javanese also responded to their commerce, and the Chinese stopped coming"
[Joao de Barros, Da Asia, 1563, Dec. 3, livro 1, 57679]. Other Portuguese, Spanish, and Dutch
sources report similar stories circulating among the Ternatans. Galvao, one of the earliest and
most careful of these, concedes that the Ternatans differed as to whether the first junks arriving
for cloves were Chinese, Malay, or Javanese. "Most of them incline towards the view that it was
the Chinese, and that seems to be the truth" (Galvao, A Treatise on the Moluccas [c. 1544] in
Reid, 1996, n. 20).

But by the early fifteenth-century Ming period, there is no indication of direct Chinese
sailing to the Moluccas. What happened?

Reid (1996) suggests that Majapahit incorporated the Chinese sailors, by then largely
resident in Southeast Asia and in the process of becoming absorbed into the local populations,
into the Javanese shipping industry. "'Chinese' merchants may no longer have been reported as
making the voyage between Java and Maluku because they ceased to be identified as such.
The confusion of Ternatans as to whether the early traders were Chinese, Malays, or Javanese
was probably justified." This is plausible, but not completely convincing. The early fifteenth-
century Ming voyages under Zheng He were Muslim-led, so it seems likely that some knowledge
of this ancestry would have been communicated to them. It also creates the implication that
the Javanese were less aware than Chinese of the routes to eastern Indonesia. Another
possibility is that the early Ming prohibition against foreign trade severed the China-Moluccas
connection.

2.3.5 Chinese Communities in Southeast Asia

Our first written confirmation that some permanent communities of Chinese in Southeast
Asia had formed comes from the author Wang Dayuan, a trader who spent a number of years in
Southeast Asia in the late Yuan period. He mentions two such communities but does so in such
an offhand manner as to suggest that they were so common that they did not deserve special
attention. In one case, he reports that some Chinese from the Yuan dynasty fleet, on its way to
attack Java, were shipwrecked, fell ill, and had been left behind on Goulan Shan (possibly the
island of Gelam, off southwest Borneo). In his day, 40 years later, some men (or their
descendants; "over 100") "live mixed up with the native families" (Rockhill, 1914: 261). A second
reference to overseas Chinese appears in the context of his description of the Longya men, at
"Dragon's Tooth Strait," the western entrance to Keppel Harbor, Singapore. That location was
known as a particularly dangerous pirate lair. From his account some Chinese lived there,
although it would seem more likely that they were resident at Pancur, his name for the
settlement on the Singapore River, about 8 kilometers away. By the early fifteenth century,
several other communities existed in Sumatra and Java; it is likely that their roots go back at least
to the end of the Yuan, for the installation of the Ming Dynasty in 1368 was accompanied by a
total ban on Chinese emigration.

26
2.4 Trade Commodities
2.4.1 Historical Sources

There are few sources which contain usable information on China/Southeast Asia trade
of this period, largely due to the official prejudice against commerce even during the relatively
free atmosphere of the Song and Yuan Dynasty. The Ling Wai Dai Da, "Information on What is
Beyond the Passes, by Zhou Qufei (1178), has been lost but was quoted by later sources. The
oldest well-preserved text was written by the harbormaster of Guangzhou, Zhao Rugua, Zhu Fan
Zhi, "Records of Foreign Peoples, in 1225. He did not leave China, but accumulated his
information by interviewing sailors. The most interesting text is the Dao Yi Zhi Lue [DYZL], the
"Description of the Barbarians of the Isles" written by Wang Dayuan, cognomen Huan-chang, a
native of Nanchang in Jiangsi, in 1349. This is the first in-depth account of Southeast Asian trade
written by an eyewitness. Another important Chinese reference work on Southeast Asian
commerce was written by Ma Guan, entitled Ying-Yai Sheng-Lan [YYSL], "A Comprehensive
Survey of the Shores of the Ocean," probably written between 1425-1432, by an otherwise
unknown Chinese Muslim who knew foreign languages and went as interpreter and recorder
with the 1413 expedition of Zheng He. The Xing Cha Sheng Lan [ [XCSL] or "Description of the
Starry Raft," was written by Fei Xin in 1436; he made several voyages with Zheng He, in an
unknown capacity.

Before the Southern Song, aromatic woods and resins were the most sought-after Nanhai
products. During the Southern Song and Yuan, demand expanded to include bulk commodities
such as pepper and a wide range of other commodities listed in the texts referenced above.
The lists cannot be taken as comprehensive, however. In Wang, for example, the equivalent of
et cetera appears at the end of the lists of trade goods. Some ports, e.g., Hua-mien,
associated with the Batak area of north Sumatra, are identified as stops for provisioning only, not
as trading ports. Here the ships would purchase cattle, sheep, fowls, ducks, betel nuts,
sugarcane, sirih leaves, and cotton. The shippers would barter iron bars, blue cotton cloth,
coarse bowls, and Quzhoufu porcelain.

The Chinese authors sometimes were mistaken about which products were local and
which were re-exports. Zhao, for example, knew that Srivijaya was a great source of Arab
products such as pearls, frankincense, rosewater, gardenia flowers, myrrh, aloes, asa-oetida, etc.
(Hirth and Rockhill, 1911: 61). In other cases errors due to the inability to distinguish between
exports and re-exports have crept in.

2.4.2 Commodities Traded Within Southeast Asia

Following is a list of commodities traded between various ports in Southeast Asia as noted
in the above-mentioned sources. Some are Chinese products, while others are goods carried by
traders from outside the region.

Glass beads Coral beads Cotton

Taffetas Damask Chintz
Silk Satin Patola cloth
Brocades Muslin Gold
Silver Mercury Borax
Tin Betel nuts Cowrie shells
Salt Spirits Lacquerware
Musical instruments Wooden combs Pepper

27
 Rosewater Camphor Gypsum
Alum Brimstone Sulphur
Rhinoceros horns Hides Ivory
Tortoise shell Iron Laka wood
Sapan wood Gharu wood Rosewood
Sandal wood Calambac wood Cardamon
Kingfishers feathers Dragons blood Gems
Rattan Beeswax Ambergris
Pearls Coconuts Nutmeg
Cloves Mace Pandanus
Mats

2.4.3 Commodities on the Java Sea Wreck

The two main commodities, iron and ceramics, are treated in Sections 8.0 and 9.0, respectively.

2.4.3.1 Ivory

Sources of ivory and routes for the ivory trade during the Song dynasty.

Wild elephants were still found in south Fujian until A.D. 1050, and in Yunnan until A.D.
1388 (Arasaratnam, 1991). However, ivory was imported to China from at least as early as the
tenth century as a major item of tributary trade. Northern Vietnam sent large quantities on
nearly every mission. In the year 980, 100 tusks of ivory were sent; in 1164, 30 pieces of ivory
tusks; and in 1177, 70 ivory tusks. In 1173, 11 domestic elephants were sent to China as tribute.
Champa (central Vietnam) sent 168 ivory tusks in 1155. The Srivijaya empire of southern Sumatra
sent 87 tusks weighing 4,065 katis in 1156, and a further 60 tusks in 1178. Ivory was sent from as far
as Africa via Arab traders. The Abbasid dynasty of Arabia sent 209 pieces of big ivory tusks in
1131 (Wong, 1979). Ivory, according to Zhao, came from Annam, Red River, Cambodia, the east
coast of the Malay Peninsula, Sumatra, and Java (the latter lies outside the elephant's range
and in reality must have re-exported ivory).

28
 Whilst the primary cargo on the Java Sea Wreck was iron and ceramics from China, the
ivory cargo strongly suggests that the ship stopped at a port, or ports, along the route from
China to Java for secondary trade. The ivory may have come from Vietnam, Thailand or
Sumatra.

2.4.3.2 Resin

Damar is the Malay name for a wide range of resins. The resin recovered from the Java
Sea Wreck has not been identified. There are, in fact, over a thousand species of resin-
producing plants in Southeast Asia, so identification is problematic. The most commonly traded
resin was Styrax benzoin, which is tapped from a species of tree that occurs in Sumatra, Malaysia
and western Java, but it was only cultivated in Sumatra. Ibn Battuta mentions it as a product of
Sumatra in the fourteenth century. Wheatly (1959: 55-59) mentions that benzoin was being
traded north to China by the eighth or ninth centuries A.D. Zhao says it was imported from
Cambodia, and shipped to China in porcelain [stoneware?] vessels. Wang surprisingly does not
mention it at all. The YYSL lists it only as a product of Malacca.

Throughout Asia, resin was burnt as an offering, or as a way of calling certain spirits to the
aid of a healer. In Malay culture it was important for rice planting and reaping rituals in which
the supplicant is trying to coax the rice spirit to stay and allow an abundant harvest.

The presence of resin as an item of cargo on the Java Sea Wreck is very clear evidence
of secondary trade during the voyage from China. And a port in Sumatra is the most probable
candidate, possibly the same port at which the ivory was loaded.

2.4.3.3 Glass

This material was mentioned as an import to China in A.D. 80 in the Han Shu; collected by
"chief interpreters attached to the Yellow Gate [eunuchs serving in the palace] who go to sea
with the men who answer their appeal [for a crew] to buy bright pearls, pi-liu-li [opaque glass],
rare stones and strange things, taking with them gold and various fine silks to offer in exchange."
(Wang, p. 19).

Zhao found Arab glassware better than Chinese glass because it was heat resistant, less
brittle, and sometimes decorated with engravings. He says (p. 227) "Liu-li [glass] comes from
several of the countries of the Ta-sh. The method followed in melting it is the same as that of
China, that is to say, it is made by burning oxide of lead, nitrate of potash, and gypsum. To these
materials the Ta-sh add southern borax, which causes the glass to be elastic without being
brittle, and indifferent to temperature, so that one may put it in water for a long time without
spoiling it. It is, therefore, more valuable than the Chinese product." A seventh-century source
says that Funan ships brought glass mirrors to China (Hirth and Rockhill, 1911: 228). Idrisi the Arab
mentions the manufacture of glass in China, at Khanfu, Hangzhou.

A tributary mission to China in 1156 brought West Asian ceramics, 4 glass vessels of
Arabian sugar, and 16 glass vessels of Arabian dates. In 1178, 4 glass vessels of foreign sugar, 3
glass vessels of foreign dates, and 4 glass vessels of gardenia flowers were brought (Ho Chuimei,
1994).

X-ray florescence analysis of the few glass objects found on the Java Sea shipwreck
indicate that they came from two origins. A dimpled glass base of a vessel, similar to a

29
European onion bottle, is clearly of Chinese manufacture (see Section 7.10). There is only one
reference in Zhao to Chinese glass exports: beads and bottles were sent to Borneo. In the DYZL,
only glass beads are listed as having been exported. A large quantity of Chinese glass
fragments of bottles and other containers has been found in a fourteenth-century site in
Singapore. A very few intact examples of the same period have been recovered from burial
sites in Riau. It seems that other than beads, however, Chinese glass was not a common export
during the Song-Yuan era. In a recent study by a Chinese scholar (An, 1996), the author only
discusses the trade of glass into China, and does not mention any exports.

Analysis of a fragment of crumbly green and yellow glass from the wreck indicates that it
is almost certainly not of Chinese origin (see Section 7.10). While there is no firm evidence of
glass manufacturing from primary materials anywhere in Southeast Asia during the pre-modern
period, it is strongly suspected that the Javanese in the late first millennium A.D. engaged in this
activity. If this fragment was not from India or the Middle East, it may have been made in Java.

2.5 The Monetary System

2.5.1 The Monetization of China

The economy of China began to be monetized after the An Lu-shan Rebellion in A.D.
755. For making bronze, China had to import tin, but domestic deposits of copper were
relatively numerous. Of China's copper producing districts, five or six were located in the Khitan
(northern) area, 26 in the southern Song area. The Khitan and Xixia kingdoms used Song coins
rather than minting their own. During the northern Song period there were over 100 Chinese
copper mines in production. A gradual shift of copper mining centers from Fujian and
Guangdong to Hunan can be observed; this is probably due to the exhaustion of the earlier
mines (Ch'en, 1965: 615).

In China, copper was used for two competing purposes: to cast ceremonial and
household goods; and as money. The government tried to restrict the private use of copper; an
edict of A.D. 960 by the last emperor of Later Zhou stated that "apart from ceremonial objects
and weapons and such articles as bells, cymbals, and handbells used in temples, other copper
ware and statuettes must be surrendered to local officials within 50 days at an estimated price.
Those who fail to obey will be punished by death if the amount of copper secretly kept at home
exceeds five catties." The same proclamation was repeated in 969 and again in 977.

"Because of the acute shortage of the metal, in the 97683 period people were digging
up copper articles from ancient tombs or destroying Buddhist statues in order to obtain copper
for illicit minting" (Ch'en, 1965: 617), and an edict issued in 991 "forbidding people from melting
down coins for manufacturing purposes" acknowledges the problem.

After a period of relative balance between supply and demand for copper, the problem
arose again in the mid-twelfth century. To set an example, the emperor in 1154 gave the mint
1,500 bronze articles from his palace collection to melt down. In addition, 2,000,000 catties of
bronze items were procured from the people. Brass and bronze items used in temples had to be
registered and were subject to tax (Ch'en, 1965: 618).

The issue of coins in China can be plotted for some years. The following table (from
Ch'en, 1965: 619, table 3) gives some data for the numbers of coins minted in particular periods.

30
 Tang Dynasty: Northern Song: Southern Song:

74256: 327,000 995-8: 800,000 1131: 80,000

80405: 135,000 1000-1: 1,350,000 1132-3: 100,000
82021: 150,000 1007-08: 1,830,000 1155: 140,000
8345: 100,000 1016-17: 1,250,000 1156: 220,000
1021-2: 1,050,000 1157: 230,000
1023-32: 1,000,000
1041-9: 3,000,000
1049-54: 1,400,000
1064-8: 1,700,000
1073-4: 6,000,000
1080-1: 5,949,234
1106-7: 2,890,000
1124-5: 3,000,000

"The Southern Song suffered almost continuous inflation. The smallness of the figures is
due to the exclusion of 'big coins', iron coins, and paper notes. In a period of serious copper
shortage the standard cash could have been issued only in small quantities" (Ch'en, 1965: 619-
620). A process of debasement can also be detected in Song coins (Ch'en, 1965):

Copper content of Song cash: 976: 65.5%

1019: 64.5%
1035: 60%
1135: 60%
post-1135: 54%

An alloy known as dan-tung, essentially a copper sulphate solution, was also used to
copper-plate iron coins.

For Southeast Asians before the twelfth century, Chinese coins represented no more than
a convenient source of bronze. As early as 1074 an official complained that cartloads of coins
were crossing China's frontiers. The Song tried to push the export of silk as a way of offsetting the
unfavorable balance of trade. Between 1160 and 1265 the export of cash was forbidden ten
times, Java being specifically mentioned once. In addition to repeated injunctions against the
export of coins, in 1293 private trade in gold, silver, ironware, military equipment, and so on was
forbidden. In 1296 the export of gold and silver was again forbidden.

Another instructive set of quantitative data concerns the edicts that sought to stem the
hemorrhage of coinage from China. The list shows that during the Song Dynasty the main
sources of the leak shifted from the north to the south border.

Locations cited in injunctions against coin exports issued during the Song Dynasty (Ch'en, 1965:
621):

Year/Prohibited area
960 - north of Great Wall, southern barbarian regions
970 - all barbarian regions
1041 - outside the empire
1085 - outside the empire
1091 - outside the empire

31
 1165 - outside the empire
1178 - outside the empire
1182 - Guangzhou, Quanzhou, Mingzhou, Suzhou
1198 - all seaports
1212 - all Yangzi ports
1216 - only iron coins allowed in frontier area
1234 - all seaports
1244 - all seaports
1250 - all seaports
1252 - all seaports
1253 - all seaports

The Southern Song began issuing paper money in 1121, but because it was not secured
by metallic reserves it became greatly depreciated. Khublai Khan also began to issue paper
money in 1260. This paper money made its way to such distant places as Wutieh in India,
according to Wang Dayuan.

2.5.2 The Monetization of Indonesia

The economy of Java began to be monetized around the same time as that of China.
Silver coins with a design based on the sandalwood flower motif originated in central Java,
perhaps around A.D. 800, and spread from there to Sumatra and the isthmian area of the Malay
Peninsula. At least one example has been found as far away as Fostat, Egypt, in a twelfth-
century context. It is not known when Srivijaya in south Sumatra began to issue coinage; no
specimens of native coinage have been reported from the vicinity of Palembang, but
Palembang has been very cursorily excavated. Three gold sandalwood-flower coins were found
at a temple site at Muara Jambi, as part of a foundation deposit that probably dates from the
eleventh or twelfth century.

According to the DYZL, in Java they cast coins of an alloy of silver, tin, lead, and copper,
"the size of a section of a small conch shell. They are called 'silver coins,' and are used in business
transactions in exchange for [Chinese] copper cash" (Rockhill, 1914: 237).

Silver coins are rare in Sumatra, in contrast to Java. The only significant discovery of silver
coins is a cache of 79 found at Krui, Bengkulu, in 1904 (Wicks, 1985: 219). Other major finds of
sandalwood flower coins, at Barus and Bengkulu, have been of gold.

Tin coins were used in Sumendala, Aceh (YYSL). The tin coins were the main currency
used in trading, but they also cast gold coins called dinar. The Hsi Yang Chao Kung Tien Lu says
in a note that "They cast a coin of pale gold an inch and 5 tenths in diameter and on both sides
of which there are designs. It weighs 3 candareens 5 li. One authority says that 48 of them is
equal to one tael four candareens in gold." (Rockhill, 1914: 154, n. 1.)

In 1079, a Srivijayan embassy was given 64,000 strings of cash. These do not seem to
have found their way into circulation, however; they were probably melted down and used for
some other purpose.

Sometime within a few years of 1300, the Javanese adopted Chinese copper coinage as
an official medium of exchange, a landmark in the interaction between the internal economy
of Indonesia and its links with China. Chinese coins had already been in use as money in a few
ports in Sumatra and elsewhere in Southeast Asia as much as a century earlier, but the

32
conversion of the Javanese economy can be viewed as a significant watershed in the evolution
of Southeast Asian society.

Monetization had perhaps reached a stage beyond which it could not evolve further
without an efficient base-metal coinage, which Chinese money was able to provide. The
amount of Chinese coinage in Southeast Asia must have been considerable. The Javanese
inscriptions shift almost overnight from quoting monetary amounts in Javanese currency to
Chinese picis. Nevertheless, Wang Dayuan in the late Yuan mentioned that, in his day, Chinese
coins were routinely exchanged with native coins (Rockhill, 1914: 236), so native coins must still
have been in circulation.

The coming of Islam to Sumatra in the early fourteenth century introduced a new form of
coinage. "By the early fifteenth century this coinage would become the commercial and
accounting standard throughout island Southeast Asia, a position it maintained for more than
two hundred years" (Wicks, 1993). Chinese coins no longer formed the basis of trade.

33
3. ARCHAEOLOGICAL PROGRAM by Michael Flecker

3.1 Archaeological Objectives

The primary archaeological objectives of this project were to document as much of the
wreck site as possible, commensurate with actual site conditions, and then to use that
documentation and follow-up research to assign the shipwreck a place in history.

A preliminary investigation of the wreck site indicated a primary cargo of iron and Song
Dynasty ceramics. Before the excavation commenced, research was undertaken to discover
more about thirteenth-century trade throughout Asia, as well as Chinese and Southeast Asian
shipbuilding techniques at the time. By knowing what to expect in terms of artifacts and hull
structure, excavation, documentation, conservation, and storage could be planned with
specifics in mind. Actual finds always differ to some extent from expectations, but it is a
worthwhile exercise to find out as much as possible about a site before it is disturbed.

Once a wreck site is excavated, there is no going back. Any information not recorded is
lost forever. The two most important details to record before an artifact or structural element is
moved are position and context. With these, and a good understanding of site conditions, the
site can be reconstructed on paper, with a computer, or by modeling to determine cargo
stowage patterns, vessel layout, construction techniques, crew and passenger dispositions, and
the wrecking process.

On a well-defined and undisturbed site it is desirable and often feasible to plot precise
positions for all significant artifacts. Unfortunately, the Java Sea Wreck site was highly disturbed,
largely by the natural wrecking process, as described in Section 5.0. It would have been an
impossible task to record the individual positions of tens of thousands of ceramics spread over an
area of 70 x 55 m. Instead, the area was divided into 5 x 5 m grid-squares (134 in total) and
ceramics positions were recorded by grid-square.

Apart from a few fist-sized chunks

of eroded wood, nothing whatsoever
remained of the ship itself. In its place
were massive iron concretions that
concreted together before the hull
disintegrated. All that could be learned
of the ships layout had to be gleaned
from the configuration of the iron
concretions. As a result, they were
photographed, videotaped, measured,
and (in very small areas) carefully
dissected. A photomosaic was also
constructed, but due to changes in
vertical relief and mediocre visibilty, it Scatter of ceramics on the seabed prior to excavation.
cannot be interpreted quantitatively.

The artifacts themselves must be stabilized and conserved. It is the artifacts that provide
so much information on trade patterns, ancient technology, cultural interactions, life on board,
and life in general at the time of the loss. The artifacts and their constituent materials were
studied in detail. Follow-up research was undertaken. Much was learned, as detailed in the
following sections.

34
 TABLE 3.1 ARCHAEOLOGICAL PROCEDURE FLOW CHART

Research Surveys Excavation Recording Conservation Results

Trade/ Pre- Artifact Unique Artifact Conservation Major

Economic Disturbance Location Numbers of Archaeological
Background Survey Recording Assigned Artifacts Report

Artifact In-Situ Computer Archaeological Special Studies

Vessel
Videos/Stills Imaging Illustrations Technical
Construction
Reports
and Fittings

Systematic Measurements Computer Professional

Present Legal Excavation Imaging/ and Popular
and Video Articles
Contractual

Artifact Artifact Artifacts Presentations

Regional Recovery Topological Stored (Video/Slides)
Environmental Classification Topologically
Conditions

Survey/ Environmental Input

Recovery/ Documentation Archaeological
Documentation Data
Methodology

Pacific Sea Resources

3.2 Recovery Vessel and Diving Systems

3.2.1 Recovery Vessel
35
 The 53 m long accommodation/dive support barge, ABEX TS, was hired for the
excavation. This barge with its spacious covered accommodation area, ample deck space,
and high-capacity water and fuel tanks was an excellent work platform for a two- to three-
month project with minimal shore support.

General arrangement of DSB, ABEX TS.

The accommodation area consisted of sleeping quarters for up to thirty personnel, two
ablution cabins, a galley and mess, a dry-store container, and a large office. The office housed
the computer and video systems used for data recording, together with the charts and drawings
needed for the day-to-day monitoring of excavation work. An SSB radio installed in the upper-
level accommodation provided a worldwide communications link via Jakarta and Singapore
Radio. Forward of the accommodation were the permanently installed bosuns store container,
compressor/workshop container, dive chamber/control panel container, and bow-mounted A-
frame.

36
ABEX TS moored over the site. ABEX TS accommodation.

All onboard facilities were powered by a 195 KVA and a 125 KVA generator, which were
rotated on a twelve-hour basis, the smaller unit being used for the night supply. Two 390 cfm
low-pressure industrial compressors were installed near the bow to run the air lifts and three air-
tugger winches, 1 x 1 ton, 1 x 2 ton, and 1 x 5 ton. One compressor was sufficient to run all
systems, while the second provided full backup. They were alternated on a three-day cycle.

A four-point mooring system, run off two hydraulic double-drum winches, enabled the
barge to be maneuvered precisely over the grid being worked. The ground tackle for each
mooring point consisted of a three-ton stockless anchor connected to a pennant buoy by 27 m
of stud-link chain and 40 m of 25 mm pennant wire. The barges 25 mm mooring wire ran
through fairleads to the pennant buoy. The anchor pattern required a very wide spread, so that
the barge could be moved anywhere over the 70 m long wreck site without having to redeploy
anchors.

Three-ton anchor ready for deployment.

Taking up slack on a mooring winch.

Anchor pennant buoy.

37
 Five containers were loaded onto the forward deck, two of
them double stacked. Three contained packing material for
the ceramics from the wreck; one was for packing and
storing the ceramics; and one was an artifact store and
conservation area. Before departing from Singapore, 300
tons of potable water, 100 tons of washing water, and 80
tons of diesel fuel were loaded into the four designated
tanks, sufficient for an extended operation without
resupply from shore.

Loading stores onto the barge.

The ABEX TS was towed between Singapore and the wreck site in the Java Sea by the
660 BHP tug ATSA I. A steady five knots was achieved both ways in moderate weather
conditions. The ATSA I assisted with anchor handling and remained on standby throughout the
project.

Engine room of ATSA I.

Tug and anchor handler, ATSA I.

A 10 m long Indonesian fishing boat was hired

to ferry temporary personnel to and from Jakarta, 115
nautical miles to the south, and to supply fresh fruit,
vegetables and spare parts. The trip typically took 19
hours and occurred on average once every ten
days.

An Indonesian fishing boat used for resupply.

38
 3.2.2 Diving Systems

Diving System On Board DSB ABEX TS

HP HP
LP HP Compressor Compressor
Compressor Quad

Oxygen
Quad

Communications Communications
Single Double
Scuba Tanks Tanks

2 Aquarium

Dive Panel
Recompression Chamber
Air
Communications
Kluge

Scuba
O2 Divers
Regulator

Down
Line

Stage

KMB Divers

The wreck lay in 26 m of water, an ideal depth for surface-supplied-air diving with surface
decompression on oxygen. With a core of twelve surface supply divers, the first pair leaving the
surface the moment the sun crept above the horizon and the last pair leaving the
decompression chamber just after dark, twenty diver hours of bottom time were achieved on a
daily basis.

Decompression chamber operations.

The compression room.

39
The diving spread was mounted at the bow
of the barge, so the barge was moored with
the stern facing the prevailing weather in
order to put the bow in the lee. The main
equipment was installed in two 20 ft
containers. The forward container, with a
clear view of the working deck, housed a
two-diver air-control panel and a 48 in
double lock Deck Decompression Chamber
(DDC). The second container housed an 86
cfm low-pressure compressor, a 9 cfm high-
pressure compressor, an 8-cylinder
emergency air bank, and workshop and Oxygen racks for surface decompression.
storage space. In a row alongside both
containers were nine 16-bottle racks of medical oxygen needed for surface decompression.

Kirby Morgan Band Masks (KMB) and

occasionally a Superlite helmet were used
for the surface-supplied diving. They were
connected to the air control panel by
umbilicals that incorporated an air hose,
pneumofathometer hose for determining the
divers depth, communications cable, and a
rope-strengthening member. The Superlite
could be used with a helmet-mounted
videocamera and a light; however, the
Indonesian divers found the KMB more
comfortable.

Adjusting a KMB mask.

Two-diver control panel.

Checking the emergency air supply prior to a

dive.

40
 Diving operations were controlled from the forward container at the air control panel.
Air was supplied to the two-diver panel from both the low-pressure compressor and the
emergency air bank. Pneumofathometers built into the panel allowed accurate monitoring of
the depth of each diver, and a radio mounted just below the panel provided two-way
communications.

Divers were deployed in a two-man staging operated from an A-frame permanently

mounted off the bow of the barge. A 2-ton air-tugger winch was used to lower the staging from
the deck to the seabed and to lift it back to the deck at the end of the dive. The advantages of
using a staging are that ascent and decent rates can be carefully controlled from the surface,
and umbilicals can be tied off at the bottom to drastically reduce current drag on the diver. A
clump weight, consisting of a solid 7 cm bar of steel, was rigged such that the staging could be
lowered down to it on two guide wires, thus preventing the staging from twisting and fouling the
umbilicals. The clump weight could be raised and lowered by an 1-ton tugger in order to keep
the staging a meter or two above the seabed, where it could not foul the grid ropes. In heavy
seas this clearance was increased.

Divers about to leave surface. The staging on the way to the seabed.

40
 Surface decompression on oxygen at the depth of this site has three main advantages.
First, the divers can be removed from the water and complete their decompression in a carefully
controlled environment. Second, decompression time is reduced by breathing oxygen rather
than air. Lastly, by transferring the divers into the DDC, it frees up the diving equipment, allowing
the next pair of divers to immediately enter the water.
Decompression schedules were taken from the US Navy Surface-Decompression-on-
Oxygen tables and modified according to standard offshore diving practice. Bottom times
ranged between 100 and 105 minutes at a depth of 85 ft (26 m). Instead of using a 110/90
table, the 130/90 table was chosen (an increase of two tables) to take into account the
nitrogen build-up due to many consecutive days of diving. The US Navy procedure is to travel
directly to the surface without any in-water stops, enter the chamber, and be pressured down to
40 ft for 48 minutes, breathing oxygen. On completion the diver travels to the surface at 25 ft per
minute. Modifications to this procedure included a 3 minute water stop at a depth of 30 ft,
pressuring the chamber down to 50 ft while breathing oxygen for 10 minutes before ascending
to 40 ft, and then slowing down the ascent to the surface such that it took 10 minutes rather
than 1 minute, 40 seconds. An air break of 5 minutes was also taken for every 20 minutes on
oxygen to minimize the chances of suffering from oxygen toxicity. These modifications greatly
reduced the risk of decompression sickness.

While surface-supply diving was used for excavation work, scuba was used for setting up
the rope grid, video and photography, measuring and recording, and supervision. Up to four
divers used scuba, the numbers fluctuating to ensure a full team of twelve surface supply divers.
A special scuba platform was rigged on the side of the barge, with a ladder and a heavily
weighted decompression line. A guide line ran from the bottom of the decompression line out
to a buoyed down line. The down line led to one of the grid stakes, well away from the working
surface supply divers, so that there was no interference between the two. Normally, twin tanks
were used and dives were kept to 30 to 40 minutes for a first dive and 20 to 30 minutes for a
repetitive dive. All decompression was done in water, using standard US Navy tables. However,
as an added safety precaution, oxygen was breathed at the 10 ft stop via low-pressure hoses
fitted with second stage regulators.
In all, 880 dives were made during the project, totaling 1,190 diver hours bottom time, without
any serious mishap.

Company at the scuba decompression

staging. A scuba diver leaves the barge.

41
 3.3 Excavation Techniques
Before excavation could commence, a grid system had to be set up over the wreck site.
A longitudinal gap ran between the rows of iron concretions that protruded from the center of
the site. This gap was the only remaining evidence of a longitudinal bulkhead that must have
run the full length of the wrecked ship. It provided a very convenient center-line orientation.

A 12 mm polypropylene rope was run over the concretions parallel to the gap rather
than through it, as there were obstructions at both ends. It was extended in both directions to
several meters beyond the surface scatter of ceramics, where 1.5 m-long pipe stakes were
hammered through the soft sediment into the stiff clay below. The rope was stretched taut and
tied off to the stakes. The center-line rope was 70 m long.

Using a fiberglass tape and the 3,4,5-pythagorus-right-angle technique, more stakes were
driven in perpendicular to the center line and 5 m on either side. Having installed the
longitudinal ropes between these new stakes, the procedure was repeated until six lines were
established. By measuring 5 m along the lines from the outer stakes, perpendicular transverse
grid lines could be installed. Initially three transverse grid lines were put in, and more followed as
the excavation progressed.

Longitudinal lines were initially

numbered 1 to 7. As the site widened,
additional longitudinal lines had to be
installed until numbers ranged from -1 to 9.
Transverse lines started at A and eventually
reached N, being 55 m long at the widest
point. Large plastic grid tags were tied in to
the grid lines to the front and to the right of
each grid in the direction of progression for
easy reference by divers. The tags were
painted gray/blue to prevent flaring from
the camera strobe and video light.

Grid rope and tag in an unexcavated area.

Divers worked in pairs, with each diver being allocated one grid square. If it were
possible, adjacent grids were not worked simultaneously to prevent sediments stirred up by one
diver from reducing the visibility of the other. Each diver used a 15 cm diameter airlift fitted with
a 45- degree elbow at the base to prevent it from digging itself into the seabed. One airlift was
painted red and the other white to minimize confusion on the bottom in times of low visibility. By
placing the airlift away from the work face, divers could fan sediments into it to expose ceramics
and artifacts. In areas with a heavy concentration of broken ceramics and poor visibility, it was
often more efficient to dig up through the work face with both hands, so that artifacts and
intact ceramics could be felt as well as seen.

Each grid had a plastic basket, color-coded in the same manner as the airlifts and
numbered with a small tag that matched the grid being worked. Artifacts and intact ceramics
were placed directly into these baskets, so that their point of origin could be recorded on the
surface. Large broken pieces were placed in another unmarked plastic basket that was
emptied 5 m behind the work face whenever full.

42
 KMB diver excavating within a grid square. Surface signs of airlifts at work.

Artifact basket close at hand. Filled basket awaiting recovery.

At the end of a dive each diver would place his working basket in a steel lift basket,
regardless of how much was in it. There were always spare labeled plastic baskets in the steel
basket in case the first one was filled. The steel basket was winched to the surface at the end of
each dive. Working baskets were removed, and new ones put in and lowered to the bottom for
the next pair of divers.

43
 Ceramics were recorded by
grid-square, due to the huge
quantities involved. Artifacts such as
the stone anchors and the iron
concretions were plotted with
reference to the grid lines. The
concretions were more thoroughly
recorded by means of a
photomosaic, which will be discussed
in more detail in Section 3.4.4.

Artifact basket being placed in the steel lift basket.

Site cross sections and

concretion heights were determined
by taking soundings with a
decompression meter that was
accurate to 0.1 m. A reference point
on a grid stake ensured that
soundings could be corrected for tidal
variations.

Excavation progressed across

the full width of the site before moving
forward to the next grid, up to the iron
concretions. At this point, work was
restricted to the south of the
concretions to prevent equipment
being dragged over them during
barge moves, and because the site
had become so wide that it took too
long to shift the barge transversely as
well as longitudinally as work
progressed. Having passed the
concretions, work continued on the
south side of the site until reaching the
end of the grids. Then excavation
moved up and back along the grids
on the north side of the concretions.

The lift basket breaks the surface.

45
3.4 Photographic Procedures
Every step of the project was documented on still film and on videotape.

3.4.1 Still Photography

All still photographs, including those shot

underwater, were taken with Nikon cameras, mainly
the F90, N90S and F801 models, and occasionally with
an F90X. A selection of Nikon lenses was used for
surface photography. Zoom lenses included a 28-70
mm f3.5-4.5 D, and a 35-105 mm f3.5-4.5, both D and
non-D series. A 105 mm f2.8 D macro lens was used for
photographs of small objects, and a 20 mm f2.8 D was
used for wide-angle shots in confined spaces.
Occasionally, a Sigma 75-300 mm f4-5.6 zoom lens was
used for distant subjects.

Nikon SB-25 and SB-26 flashes were used

extensively, being particularly effective for fill-in light
on faces where there was a bright background.

Sixty-five rolls of Fuji Provia 100 ASA slide film

were exposed during the project, along with several
rolls of Fuji Super HGV 100 ASA print film and a few rolls
of Fuji Sensia 100 ASA slide film. One roll of Provia 400 Fully-equipped underwater photographer,
ASA slide film was also used, rated at 800 ASA, with Anthony Holley.
500 watt lights underwater. All rolls were 36
exposures giving a total of over 2,600 photographs taken during mobilization and on site.
Additional film was taken of the conservation of ceramics and other follow-up work in Singapore.

3.4.2 Underwater Still Photography

Both archaeological and aesthetic shots were taken underwater. A black and yellow 30
cm scale rule was included in most of the archaeological photographs. A black and white rule
was not used, as the high contrast causes halation, where the white is overexposed and distorts
the apparent size of the scale divisions.

An F90 camera with a 20 mm lens and an SB-25 flash were sealed in a Subal Miniflex
aluminum underwater housing for the documentation of large fixed objects, such as the
concretions. Concretions of particular note were photographed from many different angles and
exposures. As the concretions were too heavy to recover, and were too degraded to respond
to normal conservation treatments, the photographic documentation was that much more
important. Special attention was paid to certain blocks for interpretation purposes, such as
Concretion 19, which was thought to be molded to the shape of the bow. Others clearly
showed the packing of the cauldrons and bundles of iron bars.

Airlifting work on the seabed was also photographed many times with this setup from
varying viewpoints and with different lighting angles. The very wide angle of the lens behind the
dome port allowed these shots to be taken at close range in poor visibility. The flash was usually

46
held out at full arms length on its long jointed arms to reduce the reflection of the light off the
suspended matter in the water (back scatter). Normally the main flash was above and to the
left of the subject, but this was varied according to the subjects orientation. An Isotechnics
Isotender slave flash, fired by a sensor detecting the main flash and aimed from the right-hand
side, provided additional lighting many times. For the concretions it was placed on the seabed
to light up overhanging portions.

A pair of Nova 500 watt tungsten halogen lights was used for different lighting effects.
These were mounted on an aluminum bar with a power supply cable from the surface. The
arrangement was made less cumbersome with a small lift bag, but still had to be maneuvered
by an assistant diver. For the standard 100 ASA film the exposure required was 1/8th of a second,
wide open at f2.8. With the damping effects of the water and by lying on the bottom, slow
shutter speeds can be used underwater, but the movements of the airlifting would be blurred.
Therefore, 400 ASA film was used, upgraded to 800 ASA, to give reasonable speeds of 1/60th of a
second.

The photographer was always on standby to record one-off events, such as the recovery
of a large stone anchor and the excavation of unique ceramics. The wide-angle lens was used
when a diver was required to fill the frame, and the zoom lens was invaluable for shooting
smaller items, such as the ewers, in detail.

Marine life photographs were taken with the wide-angle lens for schools of fish, and the
zoom for individuals, corals and sponges. Many of the groupers that had taken up residence
between the concretions were very cooperative, enabling close-ups of an eye or a gaping
mouth. The 105 mm macro lens was ideal for small creatures, as it could focus down to life size,
equal to the 24 x 36 mm film. Waiting for the right moment and the right angle for a shoveler
shrimp and attendant goby required great patience, but could be well rewarded with the
results produced with this lens. Details of sea slugs and sea snakes were also recorded.

3.4.3 Video Photography

The new Sony DCR-VX 1000

Handycam Digital 3CCD videocamera was
used for surface and underwater video
work. The Mini DV digital videocassettes
give no loss in quality on repeated playing
or on digital copying, thus giving excellent
results in the final edited version. Sony
DVM60 cassettes were used exclusively. In
most instances, the videocamera was
tripod mounted for surface use. An
Amphibico VH1000 aluminum housing
provided the necessary protection for the
camera for underwater filming.

A Sony Marine Pack HVL-80DA 80- Digital videocamera in action.

watt light provided illumination for close
range filming of fish and corals and sponges, but was not powerful enough for longer range
work. The pair of Nova 500 watt tungsten halogen lights described above improved the range,
but required great coordination between the cameraman and the lighting diver. These lights

47
provided the necessary illumination for filming airlifting operations and for colorful shots of the
undisturbed ceramics field. They could also be used for artistic background illumination.

Like still photography, video was used for archaeological recording as well as aesthetic
work. The ability to freeze-frame the digital video, with no loss of picture quality or flickering,
made the video a powerful tool for quick analysis of various aspects of the site. Details could be
studied immediately after a dive. There was no need to wait for film developing.

Before excavation work began, pre-disturbance runs were made over the ceramics
fields and along the central axis of the wreck, following several of the longitudinal grid lines.
Many shots were taken of the concretions to study their interrelationship and their individual
composition.

3.4.4 Photomosaic

The mediocre visibility over the site made it impossible to get an overall view of the
concretions while diving. As no hull structure survived at all, it became imperative to fully record
the concretion layout. The individual blocks depicted the configuration of the no longer existent
hull. Unfortunately, they presented the worst possible photomosaic situation with sheer vertical
relief changes of up to 1.6 m. Accepting that it would be impossible to get a meaningful scale
out of the mosaic, it was decided to compile it by means of digital video and computer
manipulation.

Three parallel rope lines were installed at the same level across the concretions, the
center one being clearly marked with one meter divisions. The camera was held vertically with
the assistance of a small bubble level glued onto the housing. Planar alignment was achieved
by keeping one of the rope lines parallel with the edge of the viewfinder. And a constant level
was achieved by maintaining a constant depth reading. Of course, it was impossible to watch
all three indicators at the same time, but results were reasonably consistent nonetheless. A long
lead-in swim gave time to adjust and stabilize. Lines were swum, or rather drifted, with the
current in an east to west direction. In all, eight lines were filmed, and six of those were used for
the mosaic.

No artificial lighting was used for the photomosaic runs over the concretions, due to their
rapidly varying heights from the seabed. This would cause the camera to constantly alter
exposure levels and result in harsh shadows and glare. The camera is more light sensitive than
the human eye, so reasonable results were achieved without artificial lighting.

The video runs were played back at very slow speed, and individual frames were
selected to give more than 50% overlap. Having selected a frame, it was stored as a bitmap file
on the computer via a digital image capture system called Snappy (see Section 3.5, Registration
Systems). The images for each run were then recalled one by one into a image manipulation
program called Fauve Matisse. First the edges of each image were cropped, where distortion is
greatest. Then the scale and alignment were slightly adjusted to match as closely as possible
the previous image. When the best fit had been achieved, the new combined image was
saved and the steps repeated until the complete line (typically 20 images) was laid out.

In theory, successive lines could have been combined with the first to provide a
complete mosaic of the concretions. In actuality, the PC (see Section 3.5, Registration Systems)
did not have the power to do this and suffered memory overload. Instead, each run was
processed and printed out as a mosaic strip. These strips were then cut and pasted to form the

48
complete mosaic. This was used in conjunction with actual measurements to sketch the
concretion field.

Two digitally produced mosaic strips.

49
3.5 Registration System

Three databases were required

for the project: ceramics registration
with grid location as the pivotal data;
artifact registration with the same
emphasis; and a packing inventory for
Government monitoring. Instead of a
conventional card registration system,
all information including images was
recorded on computer.

Jeff Nicklason at his artifact imaging and registration

station.

Microsoft Access was chosen as the most appropriate database due to its widespread
acceptance, compatibility with other software programs, ease of operation, and ability to
incorporate digital images. Imaging was done with an 8-mm videocamera linked to the
computer by an off-the-shelf digital imaging package called Snappy, a cigarette packet-
sized device with accompanying software. The main computer was a Compaq Presario 7222
(Pentium) PC, and the deck registration computer was a Toshiba Satellite 115CS (110 Pentium)
notebook.

3.5.1 Ceramics Registration Database

The ceramics registration database is the primary tool for determining ceramics
distribution over the wreck site, and hence the likely stowage pattern for the different wares. It
contains 21 fields, 4 of them being for images of the piece taken from different angles and
scales. Many of the fields had a pull-down list of the data most frequently input. Of course, new
data could always be typed in. The input fields with associated pull-down lists are included as
Table 3.2.

3.5.2 Artifact Registration Database

The artifact registration database contains all information on the artifacts and can be
updated as the artifacts are cleaned and conserved, which often reveals details not apparent
when they are first recovered. The database contains 12 fields, including 2 for images. The fields
and associated pull down-lists are included as Table 3.3.

3.5.3 Ceramics Packing Inventory

50
 The ceramics packing inventory is a simplified version of the ceramics registration
database. It was required to keep track of the contents of each box that was packed and to
monitor the quantities being recovered on a daily basis. Only six fields were necessary, and the
pull-down lists are identical to those in the ceramics registration database.

Table 3.2: Ceramics Registration Database

Grid

Recovery Information

Material

Earthenware Stoneware Porcelain

Form

Box with cover Jar Basin

Box cover Jarlet Bottle
Box base Kendi Vase
Ewer Lid Dish

Size

Very Small Small Medium Large Very Large

Trait

Fine Provincial No Footring Unglazed Ring Octagonal Round

Glaze

Green Qingbai Brown

Pale Green Gray Burnished
Olive White Cizhou

Decoration Method

Combed Appliqu Molded/Painted

Molded Painted None
Incised Incised/Painted

Internal & External Decoration

Combed Plain Wavy Zigzag

Floral Swept Striated Scrolls
Lotus Ribbed Ringed

Description

Body Condition

Poor Marginal Good Excellent

Glaze Condition
51
Poor Marginal Good Excellent
 Table 3.3: Artifact Registration Database

Artifact Number

Location/Grid

Recovery Information

Material

Cu Alloy Ferrous Wood Glass Rock Organic Unidentified

Form

Figurine Ivory Resin Gong

Balance Weight Ballast Stone Sharpening Stone Tray
Ingot Scales Bar Finial Unidentified

Features

Condition Dimensions Description

Classification

Cargo Personal Possessions Ships Stores Ships Fitting

Ship Structure Intrusive Unidentified

Stored Images

52
 3.5.4 Data Analysis and Display

Access is a powerful database program that allows filtering, sorting, and summing on
many levels. Thus the ceramics database can be analyzed in great detail and in many ways.
The main use, however, is to determine the distribution of ceramics over the site. This was first
done for ceramics as a whole by summing the ceramics found in each grid. In order to
graphically display the results, the new data was transferred to Microsoft Excel, a spreadsheet
program with more powerful graphics than Access. The results were printed out both on 2-D and
3-D color-coded plots that depict the grid layout of the wreck site. Just a glance at the plots is
enough to picture the unusual distribution that has taken place.

Similarly, the sum of the different types of ceramics was determined for each grid. By
plotting this in the same manner, it is very easy to see which types were stowed in which part of
the ship.

By using the form function in Access, excellent-quality catalogs with color images could
be produced for both the ceramics and the artifact databases.

3.6 Ceramics and Artifact Handling Procedure

3.6.1 Ceramics Handling Procedure

At the end of each dive the foam-padded steel basket containing the plastic grid
baskets was winched to the surface. The grid baskets were carefully removed and taken
directly to the ceramics washing table, which was located on the port bow of the barge under
an awning. A submersible pump supplied a continuous flow of sea water via piping and four
discharge hoses attached to the back of the table. A foam mattress was fixed to the steel
grating table top. Here the ceramics were cleaned of mud, sponges, oyster shells and coral
with scrubbing brushes and plastic chop sticks. The more tenacious marine growth, and there
was a lot of it, was left for removal in the conservation laboratory. Ceramics from each grid
basket were kept completely separate.

Unloading ceramics onto the washing Ceramics cleaning in full swing.

table.

53
After washing, the ceramics were rinsed in fresh water, then sorted according to form, size,
glaze, and condition in preparation for registering. The sorted ceramics were stacked on the
shaded deck in separate areas designated by grid tags. Except for periods of inclement
weather, all ceramics registration took place on deck immediately after washing. One person
would examine the sorted ceramics and call out the pertinent information to another for data
entry into a notebook computer. Information on each piece included the following: grid
square, quantity of pieces with the same attributes, ware, material, form, size, glaze type,
decoration method, internal decoration, external decoration, body condition, and glaze
condition. One additional field was provided to record general traits, such as no footring or
unglazed ring, and another for any further description unique to the piece being registered,
such as spout broken or distorted rim. A small rotating staff of three people was used for
calling and data entry in order to maintain the consistency of logged information. Unique or
exceptional pieces and representative examples of each ceramics type were set aside for
digital imaging.

Ceramics segregated by grid-square.

Logging each find on computer.

Only after registration was completed were ceramics from different grids mixed together
in preparation for packing. Pieces that could be stacked, such as bowls and dishes, were
placed one inside the other with 10 mm bubble plastic between each piece. Large or odd
shaped pieces, such as ewers and kendis, were wrapped in 25 mm bubble plastic. Stacks or
individual large pieces were then sealed in heavy gauge plastic bags containing small amounts
of fresh water to prevent them from drying out while awaiting transport to the conservation
laboratory. The bags were then placed in double-wall cardboard boxes that were lined on all
sides with 20 mm polystyrene sheeting and divided into quadrants with a cardboard partition.
The partition kept the stacks of ceramics separate and increased the vertical strength of the box
to prevent crushing when stacked. In order to prevent shifting, all remaining space in the box
was packed with loose fill polystyrene. Each box was finally taped shut, numbered on all sides,
and stacked in sequence in a secure storage container.

54
 A separate database was used to keep track of each boxs contents, based on form,
size, condition, and quantity. Daily packing lists were printed out each evening for joint signing
by the Indonesian Government Representative and by Pacific Sea Resources.

On arrival at the conservation laboratory in Singapore, the boxes were unloaded from
the container and stacked in the laboratory in sequence, such that processing could start with
the first box that was packed and progress in order. The main priority was to complete cleaning
and commence desalination without allowing the ceramics to dry out. Boxes were unpacked
one at a time, the next one not being opened until the first had been processed. As much of
the marine growth as possible was carefully chipped off with small chisels and chopsticks.
Several pieces were only held together by the marine growth. When the shell and coral were
removed, they fell apart, revealing yellow stained broken edges, a sign that they had been that
way for some time.

Logging the ceramics for packing. The ceramics cargo on arrival in Singapore.

55
 On completion of mechanical
cleaning, pieces that required more work
were left overnight in a very dilute
hydrochloric acid solution. After this
treatment the pieces were scrubbed clean in
fresh water, then carefully stacked in
freshwater baths for desalination, the details
of which are given in Section 3.7,
Conservation. Plastic sheeting was placed
over the baths to prevent evaporation and
the breeding of mosquitoes.

After desalination the pieces were

removed from the baths in the same order in
which they had been placed there, dried,
and sorted for cataloging in accordance
with all the criteria recorded in the original
ceramics database. They were then stacked
in designated groupings on secure shelving in
an air-conditioned storeroom.

The ceramic cleaning and desalination process.

3.6.2 Artifact Handling Procedure

The procedure for registering non-ceramic artifacts was similar to that for ceramics,
although the numbers were far less. As each basket arrived on the surface, artifacts were
removed, labeled according to grid, and placed directly into an appropriate storage medium.
Registration of artifacts generally took place in the evening and was carried out at the PC work
station in the barge office. The artifact database recorded the following fields: artifact number;
grid square; date recovered; quantity; material (ferrous, copper alloy, etc.); form (gong,
balance weight, etc.); classification (cargo, ships structure, etc.); dimensions; condition; and
description. Digital images were usually taken during the same registration session for inclusion in
the database.

3.7 Conservation

3.7.1 Conservation of Ceramics

Conservation of ceramics is fairly straightforward. After cleaning off the marine growth, it
is essential to leach out the salts that have entered the glaze and body of the piece during the
hundreds of years that it rested on the seabed. If this is not done, the pressure created by salt
crystalizing out of solution, particularly between the glaze and the body of an imperfectly glazed
piece, can result in exfoliation and cracking.

To achieve this, the ceramics were placed in desalination baths 2 m in diameter and 0.3
m deep. A few drops of Algaetrene were added to prevent algae growth. The salinity of the
water was indirectly measured with a conductivity probe, and the result graphed over time. As

56
soon as the conductivity started to level out, the water was changed. The first wash increased in
salinity very quickly, particularly the bath containing stoneware jars, which are more porous than
the glazed ceramics. Subsequent washes were typically changed once every 10 days or so.
The total desalination period was typically 3 months.

3.7.2 Conservation of Artifacts

Stable storage and conservation treatment are determined largely by material type. The
different materials that had to be treated on this wreck included iron, copper alloy, glass, ivory,
and wood.

The iron was in the form of concretions, a matrix of iron corrosion products, sand, and
calcareous marine growth that forms around an iron artifact and often grows to several
centimeters thick. In actual fact, there was no iron base metal left after 750 years or so. Some
concretions contained a black material that was quite solid in places, but could readily be
picked out, and other concretions contained slush. Some concretions that were not totally
enclosed were completely hollow. Obviously nothing can be done to save the contents, but
the concretions have become a mold of the original object. They were stored in water, as they
very quickly expand and crack open if allowed to dry.

Selected concretions were cut open with a grinding disk and cleaned out so that a cast
could be taken. The casting material used was Silastic J RTV silicone rubber, a low-shrink, room-
temperature vulcanizing compound that gives highly detailed reproductions.

Copper alloy artifacts tend to be fairly stable after prolonged immersion in seawater.
They can certainly be badly corroded and worn, but do not rapidly deteriorate when removed
from their marine environment. The long-term problem is bronze disease, caused by absorbed
salts. A good storage medium that also removes chloride irons from the artifact is sodium
sesquicarbonate. The copper alloy artifacts were immediately placed in a 3 percent solution on
board the recovery vessel. At the end of the excavation it was impractical for transport in the
solution, so the artifacts were rinsed with fresh water and allowed to dry before being placed in
plastic bags with silica gel to remove any remaining moisture.

In the conservation laboratory in Singapore the large items were put back in sodium
sesquicarbonate for several more weeks of treatment. After that they were dried out as above
and coated with Incralac.

Glass artifacts were treated in the same way as the ceramics. They were simply
desalinated in fresh water. When the increase in solution salinity became minimal, the artifacts
were placed in plastic bags for controlled drying.

The ivory artifacts posed a bit of a problem. The largest piece measures 620 mm long
and 130 mm in diameter, while the smallest piece is a 100 mm long tip of a tusk. All of the pieces
were badly deteriorated and exfoliating. The ivory was soaked in fresh water to remove the salt.
One small piece was put in several acetone baths to remove the water, then in a 40%
acetone/rosin solution in order to bind the layers. The results were quite satisfactory; however, it
is not practical to use this treatment on larger pieces due to the large amount of acetone that is
required. Instead, the larger pieces were simply bound in pantyhose to reduce the risk of
exfoliation, and slowly dried out in plastic bags. So far this simple method seems to have
worked.

57
 Only a few pieces of highly degraded wood were recovered. Two of these had a dowel
hole in them and hence were conserved. As they were only fist sized, the acetone/rosin
method, as described for ivory above, was used. Apart from a few small shrinkage cracks, the
method seems to have worked well.

58
4.0 SITE CONDITIONS by Michael
Flecker

4.1 Site Description

The wreck lies in the Java Sea in Indonesia in latitude 4 14 S and longitude 106 40 E.
The southeast coast of Sumatra lies 50 nautical miles to the west. The islands of Banka and
Belitung are some 70 miles to the north, and the northernmost island of the Thousand Island
Group is situated approximately 60 miles to the south. There are no islands, reefs or sand banks
within 40 miles. The ship sank in the middle of nowhere.

At night the southern horizon is disturbed by a flickering orange glow that marks the
northernmost production platform of the Maxis Oilfield. Fishing boats and ships navigating
through the Banka Straits pass by at regular intervals, but from time to time the sea presents 360
degrees of uninterrupted horizon.

Twenty-six meters below, the seabed is also bleak. A gently undulating silty sandy bottom
stretches for miles in all directions. This loose and mobile sedimentary layer, varying from 0.5 to
1.0 m thick, overlies a stiff marine clay. Environmental and biological weathering have reduced
the clay surface to undulations, holes, and loose chunks.

Projecting some 1.5 to 2.0 m above the surrounding sediments, the wreck mound forms a
significant feature on an otherwise featureless seabed. A series of massive iron concretions
protrudes a further 1.5 m above the mound, creating a thriving artificial reef, an oasis in an
aqueous desert.

There is a pattern to the concretions. The iron cargo evidently concreted together
before the ship completely disintegrated. Compartments and bulkhead locations can be
readily discerned from the present concretion disposition. A long-gone longitudinal bulkhead or
partition has divided the concretions into two rows. Small concretion chunks lie scattered
around the main blocks, possibly the result of fishermens explosives.

It is the longitudinal partitioning of the concentrations that provides the orientation of the
wreck, 300/120, or WNW/ESE. For convenience, wreck north is taken as 30, or NNE, and all
future references to bearings are with respect to wreck north.

Apart from a few scattered shards and some tube worms, the seabed immediately
surrounding the concretions is barren. In fact, there is a relatively clear band of 10 to 15 m
before substantial quantities of surface shards become evident. The outer limits of the wreck are
marked by a carpet of silt-covered, encrusted shards. Again, these form a wide band around
the central concretions, a band with a remarkably well defined outer edge. The oval-shaped
area thus formed is 70 m long and 55 m wide.

To the south, west, and much of the north, the slope of the wreck mound is barely
discernible. However, the eastern slope and some parts to the north are very distinct, as can be
seen in the illustrations of the cross sections (Section 5.0). In various places around the site,
mainly to the north and northeast, there are large holes. They are the result of previous
excavations, reportedly well into the clay, to determine whether any remains or artifacts lay
beneath the clay layer. During the two years since the last excavation, these holes have
accumulated over half a meter of fine silt.

58
4.2 Oceanographic and Meteorological Conditions
The Java Sea is subject to the typical monsoonal weather system of the southern
hemisphere tropics. The southeast monsoon prevails from May through September. The
northwest monsoon blows from November through March and tends to have stronger, more
consistent winds than the southeast monsoon. April and October/November tend to be periods
of transition with variable winds and calms.

Monsoon-induced currents tend to set in a month or so prior to the true onset of the
monsoon, flowing west to northwest during the southeast monsoon and east to southeast during
the northwest monsoon. These currents are weak and are usually overridden by tidal currents.
According to Admiralty weather charts the monsoon currents rarely exceed half a knot.

The tidal current regime in such an exposed location, where there is relatively little
boundary influence, is very complex. The interplay of the monsoon current, tidal movement,
and local meteorological conditions result in currents from all directions. When all factors
contribute to a current in one particular direction, it may reach a velocity of 2 knots. But this is
rare, and currents at the wreck location never exceeded 1 knot during the excavation.

59
 One unusual phenomenon observed during the excavation, which was undertaken from
the beginning of September to the end of October 1996, was that currents tended to be
stronger during the half-moon phase than during full or new moon, a contrast to the norm.
Current shifts were typically diurnal with at least an hour of slack water between shifts and were
rarely diametrically opposed.

The actual weather conditions observed on the site during the excavation were
somewhat different from the expected monsoon pattern. During the first two weeks of
September the wind blew predictably from the southeast through to northeast with moderate
velocity and some heavy rains early on. There were also occasional squalls, the worst gusting to
25 knots for a full day in mid-September. Winds were then consistently light from the southeast
with no rain until the last week of the month, when they swung south and southwest and blew
hard for four days, creating short 2 m seas. A southerly swell preceded this blow by a full day.
The first three weeks of October were typical of the monsoonal transition period, with winds
generally blowing lightly from south to east, and occasional squalls and rain from the south. Only
one day was flat calm. From 20 October there were frequent squalls with very heavy rain from
the south and southwest, lasting four days. Thereafter, winds were light from the south to east.
Only once did the wind blow from the northwest, and that was due to a short-lived local
thunderstorm.

Underwater visibility was moderate, typically ranging from 3 to 6 m. Very fine sediments
on the bottom were easily stirred up by diver activity, and in times of low current, could result in
zero visibility. Moderate currents from the east resulted in an unusually localized milky layer over
the northeast quadrant of the site. Swells formed by the southerly blows also caused reduced
visibility for several days.

4.3 Environmental Monitoring and Marine Life

First, it is interesting to note that the abundant life observed is restricted almost entirely to
the wreck mound. Fifty-meter swim surveys undertaken in the four cardinal directions from the
outer limits of the mound revealed the occasional shoveler shrimp with its attendant goby and a
few worm holes. Otherwise, the gently undulating sandy seabed appeared lifeless.

The shoveler shrimp and goby were also a common sight on the wreck mound. Holes in
the sediment with cleared paths in front indicated their presence. Ceramic shards often formed
the roofs of these burrows. It is very easy to swim over them and see nothing, as the goby acts as
lookout and darts into the hole when danger is near. Shrimp could sometimes be observed
cleaning out the entrance. They are also known as bulldozer shrimp, as they can effectively
push debris forward with their two claws clasped together. They are nearly blind and maintain
contact with the goby through their long antennae. When a goby senses danger, it waggles its
tail and the shrimp waits, retreating if the signal continues. Finally, the goby itself vanishes into
the hole. This is an effective symbiotic relationship with the goby getting a clean home and the
shrimp a forewarning of predators.

Sea snakes were regularly sighted and often caused some alarm by swimming between
the divers legs. Several went for a ride up the airlifts. They are very poisonous, their venom being
much more potent than the most venomous land snakes, but they are docile and were never
any threat. Being reptiles, they need to surface regularly to breath air which resulted in many
sightings from the barge.

60
 Live corals were scarce. There were a few small areas of encrusting corals on the
concretions and several soft whip corals of different color, which often grew on shards scattered
about.

Several species of sponges also grow on the concretions, and on the shards. They
include small tube, barrel, and encrusting sponges.

Live mollusks were relatively scarce, although dead shells were found in and on the
recovered ceramics. Only live sightings were noted. Similarly, only a few live crabs were
actually spotted. No worms were seen although their holes and their casts were evident on the
ceramics.

The concretions provided the main focus for fish life in the area. Rising up to 1.5 m from
an otherwise flat bottom, they served as a refuge for resident fish and a hunting ground for the
predatory fish. The barge also attracted pelagic fish. Large schools of jacks were often sighted
beneath the barge, intermingled with batfish, rainbow runners, and fusiliers. Scuba divers
decompressing on a line under the barge provided a convenient refuge for fry, but fusiliers and
jacks often disregarded the sanctuary of the diver and attacked anyway. Constant dodging
kept the divers blood circulating.

Large groupers up to a meter in length were a common sight around the airlifting. They
watched the divers and waited for any food disturbed from the sediments. They also made
themselves comfortable in the lifting basket and in the diver staging. On days of poor visibility it
was easy to come upon a large grouper unexpectedly, surprising both parties.

A giant stingray measuring 2 m across made its home among the concretions. There was
only one, but he had many large ancestors, evidenced by spines over 200 mm long lying on the
bottom.

One angelfish species was observed, the blue lined angel (Pomocanthus annularis). A
pair of these was always around the concretions, nibbling on the algae. No butterfly fish, parrot
fish, or similar common reef fish, were observed. Other fish observed on the site are listed in
Appendix E.

Far from having a detrimental effect on the marine life, the excavation caused a
temporary increase in activity, largely brought about by the stirring up of food from the
sediments. Sediment from the airlift discharge blanketed corals and sponges on the concretions
at times, but moderate currents prevented significant accumulation. The concretions were not
disturbed, so resident marine organisms were not displaced.

Soft coral (Dendronephthya sp.) Soft coral (Dendronephthya sp.)

61
Nurse shark (Nebrius ferrugineus) Marbled grouper (Epinephelus polyphekadion)

Blue lined angelfish (Pomocanthus annularis) Marbled grouper (Epinephelus polyphekadion)

Yellow tail barracuda (Sphyraena jello) Scrawled filefish (Aluterus scripta)

Hawksbill turtle (Eretmochelys imbricata) Black spotted stingray (Taeniura melanospilos)

Sea snake (Unidentified species) Shrimp goby (Amblyeleotris sp.)

62
Nudibranch (Bornella sp.)

Purple tube sponge (Cribochalina sp.)

Feather star on rim (Unidentified species)

Nudibranch (Chromodoris sp.) Sea cucumber (Stichopus variegatus)

Pair of flatworms (Thysanozoon sp.) Yellow whip coral (Ellisella sp.)

63
5.0 THE WRECKING PROCESS by Michael
Flecker

A description of the site prior to excavation is given in Section 4.1, Site Description. To
recapitulate, the surface ceramics scatter occurs over an oval-shaped area approximately 70 m
long by 55 m wide, which is slightly less than the extent of the so-called wreck mound. The
highest point on the mound projects nearly 2 m above the surrounding seabed. In the center of
the mound are two rows of iron concretions that assumed the shape of the compartments in
which the iron cargo was stored before the ship disintegrated. The concretions cover an area of
approximately 18 m by 10 m in the same alignment as the mound and project a further 1.5 m
above the sediments.

A wreck mound is normally formed by sediment building up over hull and cargo remains.
Typically, sediment carrying currents drop part of their burden in the lee of an obstruction. If
currents travel in several directions, this can lead to a rapid burial of the site, with the result that
hull timbers are protected from marine borers and ceramics are protected from abrasion and
marine growth. Consequently, this vast site was initially viewed with great optimism.

Very early in the excavation, however, a stiff clay was found to underlie a layer of
sediment and ceramics. The elevation of the clay was higher than the surrounding seabed. This
lead to speculation that the clay was part of the cargo, but there was far too much and it was
spread far too wide for this. Then it was suggested that the mound was completely surrounded
by a scour trench, but sounding runs swum for 50 m from the outer limit of the ceramics scatter in
four directions indicated that there was no scour trench at all. The height of the clay relative to
the surrounding seabed is vividly depicted in the longitudinal and transverse cross sections of the
wreck site.

Another puzzling discovery was that absolutely none of the hull had survived intact, not
even under the iron concretions. The only wood that was found was in fist-sized eroded chunks,
and there were only a few dozen of those.

64
 Furthermore, every ceramic item that was recovered, without exception, had some
marine growth on it, whether oyster shells, coral or sponges. So every piece has been
completely exposed for at least some of its life on the seabed.

Therefore, it has been deduced that the ship came to rest upright on a clay seabed with
little or no sedimentary overburden. Over several decades the hull timbers were eroded by
currents and eaten by toredo worms. As the uppermost timbers collapsed, the ceramics cargo,
which was stowed on top of the iron cargo, started to spill out in all directions. With no sediment
covering the timbers they were eventually completely consumed. The iron cargo, consisting of
stacks of woks and bundles of bars, was well concreted together before the wooden
compartments disappeared. Those massive lumps of concreted iron have essentially remained
in their original location ever since and clearly depict the long-gone bulkheads and outline of
the ship.

As the containers holding the ceramics (possibly wooden barrels or bundles of straw)
rotted away, the ceramics spilled out over a wider area. There then followed an erosion regime,
or perhaps it was ongoing. The entire seabed in the region of the shipwreck has been eroded
away by as much as 2 m. But the wide scatter of ceramics protected the clay under the wreck
from erosion. As a slope began to form around the evolving mound, more ceramics would slide
down, increasing the scatter and the resultant protected area. Eventually the entire ceramics
cargo lay in a ring around an essentially bare hill. Perched on top of the hill were the rows of iron
concretions.

Looking at samples of the stiff clay, it is hard to believe that so much could have been
eroded. Two core samples were analyzed by Dr. Tan Thiam Soon of the National University of
Singapore Civil Engineering Department. A core taken from the top of the mound indicates that
the clay is very stiff. Six centimeters from the clay surface, it has a vane shear strength of 63
kN/m2, and 14 cm down it gets as high as 107 kN/m2 (see Appendix F). But when the overburden
is removed to expose the clay surface, extensive weathering and biological degradation
become evident. Rather than a smooth surface, it is pockmarked with holes that have been

65
created by tube worms and is covered in small chunks of clay. On a larger scale, there are
holes and trenches in the clay well over a meter deep and several meters in extent, possibly
where the clay was less stiff and hence more easily eroded (these are not to be mistaken for the
holes dug by earlier salvors). These holes have become traps for ceramics.

It is interesting to note that the clay of the surrounding seabed, and presumably that
underlying the mound, is significantly different to the clay in the mound. The mound clay is
streaked with red. When kneaded it becomes a uniform ochre color. The seabed clay is
generally gray with some brown inclusions, and it is considerably softer, having a vane shear of
25 kN/m2 10 cm below the clay surface. This is similar to the clay found 24 cm below the clay
surface on the mound and implies that most of the 2 m or so that has eroded away was a soft
clay. It was also a fairly silty clay, as indicated by a particle size distribution test, with up to 36%
silt.

In more recent times, with greatly increased agriculture and deforestation, the seabed in
the region of the shipwreck has become a depository environment. Around the mound there is
typically a 0.3 m sediment layer over the clay. On the mound the sediment surface is fairly
uniform. It is the undulations in the underlying clay that cause considerable variation in the
thickness of the layer. In a deep hole there may be more than 1.5 m of sediment, whereas
within the barren band around the concretions it can be as little as 0.1 m thick. There is 0.5 to 0.9
m of sediment around the concretions.

There were virtually no ceramics between the concretions, and very few within a 10 to 15
m radius. The lack of ceramics between the concretions is understandable if the wooden
bulkheads survived until after the ceramics were swept off the iron, thus leaving no gap for them
to enter. It remains a mystery why there were so few ceramics in the immediate vicinity of the
concretions. Given the shallow nature of the overburden, it is possible that fishermen salvors with
no excavation equipment concentrated their efforts here. But there are not many shards in this
area, either, and it is unlikely that fishermen bothered to recover or move the shards.

So, rather than an intact hull filled with iron and ceramics still in their original packing, the
site disclosed a huge scatter of broken ceramics. It is estimated that at least 80% of the
ceramics cargo was broken. Given the distance pieces have traveled from the original wreck
location, this is not too surprising. As if this was not enough, it seems that explosives have been
used on the site for fishing. There is evidence to support this. In isolated areas there is nearly
100% breakage, and the shards are smaller than usual. The stiff clay layer would act as a wall
and bounce the shock waves back, so there was little chance for the ceramics near ground
zero.

66
6.0 INTERPRETING THE SHIP by Michael Flecker

6.1 Hull Structure

There is no surviving hull structure, only a few fist-sized pieces of wood scattered about
the site. As discussed in Section 5.0, The Wrecking Process, the hull and cargo have been
completely exposed to the elements for prolonged periods. There was no layer of fine sediments
to afford protection. Consequently, every piece of the ceramics cargo was encrusted to some
extent with marine growth, and all of the ships timbers have been eaten away by marine borers
and current erosion.

To learn something of the ship, we must instead study the massive blocks of iron
concretion that dominate the site. The iron cargo was stowed in the lower holds and fused
together before the ship disintegrated. The blocks have hardly moved from their original resting
place due to their great weight. It is the shape of the concretions and their disposition that hints
at the vessel layout.

Iron Concretion Plan

When studying the Concretion Plan, the orderly layout of the blocks is immediately
apparent. They are divided longitudinally into two rows by a long narrow gap, and transversely

66
they are similarly separated. These gaps are the positions of bulkheads that rotted away long
ago.

A longitudinal bulkhead is an unusual feature. In fact, to the authors knowledge such a

structural element has never before been recorded on a ship of this era. It is feasible that when
the hull collapsed, the consolidated iron cargo, packed across the full width of the ship, broke at
the keel, but such a uniform break in every compartment makes this highly unlikely. Another
possibility is that there was a nonstructural partition running the length of the ship, but the width
of the gap suggests a substantial structural element.

Conjectural Bulkhead Layout

Transverse gaps suggest a minimum of twelve bulkheads, dividing the ship into thirteen
compartments. The compartment widths varied from 1.2 to 1.5 m in the central portion of the
ship.

Two stone anchors were found lying side by side in Grid K3, a strong indicator that the
bow of the ship is facing east. This supposition is supported by the shape of Concretion 19, the
easternmost block. Looking at this concretion from the northeast at seabed level, one can
clearly see the deep V section of the bow. The block seems to have skewed 45 degrees when
the hull collapsed. In the original configuration the lower levels of the forward hold must have
been filled with iron bars and cauldrons. The longitudinal bulkhead appears to have stopped at
the foremost bulkhead and not continued all the way to the bow. Looking down on Concretion
19, it can be seen that the bow flared out very quickly from a fine prow. The bow compartment
was approximately 2.7 m in longitudinal extent.

67
End view of the bow concretion. Plan view of the bow concretion (the bottom left is the
bow).

Towards the stern the concretions are broken into smaller blocks and are more scattered.
This may partly be the work of fishermens dynamite. But it is also probably due to the iron being
stowed in the raised stern, which must have collapsed before the lower hull. At Concretion 4 the
pattern becomes asymmetric. There is nothing on the port side to balance the weight of
Concretion 4. Either the port side was loaded with another cargo, such as ceramics, or
Concretion 4 has fallen to one side and originally was amidships. If the latter, then the
longitudinal bulkhead stopped short at this compartment.

Concretion 12 has a very interesting

cross section. The outer portion of this
concretion turns up abruptly from the
horizontal, suggesting that it is a molding of the
turn-of-the-bilge. If it is, then the vessel must
have been flat bottomed and hard chined.
Concretion 12, however, is only half as long as
concretions on the port side, so it could not
have extended to the turn of the bilge. An
explanation is that the iron was stacked up on
another form of cargo at this point, a cargo
that has long since disappeared, such as
wood.
Cross sectional view of Concretion 12.

The shape of Concretion 19 indicates that iron was stowed all the way up to the bow.
From the symmetry of the ceramics spread about the centrally located concretion assembly, it is
reasonable to assume that iron was loaded along the full length of the ship. This indicates a keel
length in the order of 18 m. The well consolidated Concretion 11 is 3.8 m long, giving a minimum
hull width of 7.6 m. The stone anchors are 3.2 m east of Concretion 19. Assuming they were
stowed right on the forepeak and did not move far laterally when the ship collapsed, the bow
overhang was approximately 4 m. If the stern overhang was of the same order, the length
overall would have been approximately 26 m, giving an aspect ratio of 1:3.4.

68
 Although very few pieces of
wood survived, two small chunks did
have identifiable dowel holes. One
was only a cross section with a dowel
diameter of 15 mm. The other encased
the entire hole with a dowel diameter
of 27 mm.

Dowel hole in a wood fragment.

6.2 Stowage Pattern and Cargo Capacity

Iron and ceramics made up the bulk of the cargo. Other items, such as ivory and
copper alloy ware, were also present, but the stowage of these less significant trade goods is
discussed in Section 7.0, Artifacts.

Iron was stowed first and appears to have been placed in the lower portion of every
hold. It is very difficult to make out the individual components of the massive concretion blocks,
but in some places it is clear. Many are made up entirely of bundles of iron bars lashed together
in a cone shape. The cones have been stacked end-for-end in a longitudinal direction.
Cauldrons are stacked one inside the other in a longitudinal direction as well, often on top of the
iron bars. One exception is Concretion 6, which is made up of larger cauldrons stacked
athwartships.

By calculating the volume of the iron concretions, making a 30 % allowance for

expansion and packing interstices and using a specific gravity of 7.2 for cast iron, the total
weight can be estimated at 340 tons.

To get an idea of the

stowage of the ceramics, it is
worth looking at the Overall Site
Plan (Section 4.1) and the Overall
Ceramics Distribution Plot
(Appendix C). The ceramics
have spread out 20 to 25 m in all
directions from the iron
concretions. The Ceramics
Distribution Plot must be
interpreted with caution since as
many as 12,000 pieces were
recovered in earlier excavations.
It does, however, indicate trends.
There is a dearth of pieces in the
southeast quadrant, but this is
Bundles of iron bars concreted together.

69
offset by an abundance to the northeast. Overall, the distribution of ceramics is fairly even
along the whole length of the site, suggesting that ceramics were loaded in every hold of the
vessel. They must have been loaded on top of the iron cargo, probably in straw cylinders and
wooden barrels. Nothing remains of the packing material.

It is interesting to study the distribution of the various types of ceramics to determine in

which part of the ship they were stowed (see Appendix D, Ceramics Distribution Plots).
Earthenware kendis were stowed throughout the ship, although there are much higher
concentrations just forward of midships and at the bow. As the kendis are probably not from
China, but rather were picked up at a port en route to Java, it is reasonable to conclude that
they were loaded on top of the Chinese ceramics cargo along the length of the ship wherever
there was remaining hold space. In fact, some of the Chinese ceramics may have been
bartered for the kendis which were very popular in Indonesia.

Octagonal covered boxes appear to have been stowed throughout the vessel. So too
were provincial dishes, but there were more in the bow area. Glazed provincial bowls with no
decoration were essentially stowed from midships aft; whereas provincial bowls with an
unglazed ring in the well were stowed in the stern and in the bow, but not amidships. Fine bowls
with an incised decoration were also primarily stowed in the stern and in the bow with only a
small quantity just forward of midships. Small qingbai vases were stowed from midships forward
with the vast majority right up in the bow compartments. Small dishes with an incised
decoration, on the other hand, were stowed everywhere except in the bow. Small conical
bowls with an incised decoration were stowed amidships and perhaps in a few compartments
just aft of that.

Merchants were generally allotted compartments or portions of compartments for their

trade goods. Unfortunately, the huge scatter of ceramics makes it impossible to determine the
stowage pattern by compartment. So it is difficult to say whether the different types of ceramics
loaded throughout the ship were determined by the preference of individual merchants or were
governed by a few individuals shipping a large consignment.

A very rough estimate can be made of the weight of the ceramics cargo. Based on at
least 100,000 pieces of typical rice bowl size (many pieces are larger and many smaller), the
ceramics cargo weighed between 25 and 30 tons.

The massive iron cargo and huge spread of ceramics strongly indicates that the ship was
loaded to capacity. From the estimated weight of the cargo and from deduced ships
dimensions, it is estimated that the loaded displacement of the ship was 400 to 500 tons.

6.3 Wood Identification

Two wood samples were identified by Dr. Ian Godfrey of the Department of Materials
Conservation at the Western Australian Maritime Museum (Appendix G). These samples, from
their size and appearance, were from structural timbers. Other wood samples were collected;
however, they were mostly in the form of branches. Such material could have been cargo (i.e.
aromatic woods) or dunnage, but it could also be intrusive to the site. Branch wood is also
difficult to positively identify.

Most of the small pieces of wood recovered were of the same species, judging from
macroscopic examination. They included the two pieces with dowel holes alluded to earlier.
The species is thought to be Parastemon, with the features of P. Urophyllus (common name

70
Mandailas) best matching the samples. Parastemon species are native to Southeast Asia and
neighboring regions, being found in Western Iran, Papua New Guinea, Soloman Islands,
Myanmar, Indonesia, and the Malay Peninsula, as well as some adjacent countries and islands.

Another possible match for this timber, although less likely than Parastemon, is Parinari.
This is found throughout Southeast Asia and in southern China to some extent. The drawback on
both of these species is that the trees are not particularly large and the timber is more difficult to
work.

The other sample that was identified was of a light-colored soft wood, a fragment of a
small rectangular beam with a beveled edge. The best match is that of Alstonia scholaris
(common name: white cheesewood or milkwood), which is native to Africa and the Philippines,
as well as the countries listed above. This timber is easy to work and comes from large trees.
Being a soft wood , however, it is not durable. Unfortunately, as a result of the highly degraded
nature of the sample, this identification should not be regarded as conclusive.

6.4 Identifying the Ship

Despite the fact that the main cargo of iron and ceramics is from China, it is unlikely that
the ship is of the same origin.

6.4.1 Construction Materials

The few surviving fragments

of wood provide an important clue
as to vessel origin. The Java Sea
Wreck seems to have been built
largely of mandailas, a
predominantly Southeast Asian
timber, and milkwood, also tropical
but may be found in the very south
of China. It is interesting to note
the woods identified on other non-
European shipwrecks discovered in
Asia.

The Phu Quoc ship (Blake

and Flecker, 1994) has bulkheads
of padauk (Pterocarpus Sp.), a
timber today restricted to
Myanmar and the Andaman
Islands, and hull planks of teak
(Tectona grandis), which occurs
mainly in Myanmar, Thailand, and
Indonesia. The planks were edge-
joined with dowels of sappan
wood (Caesalpinea sappan),
which is widely used by Southeast
Asian boatbuilders, such as the Bulkhead and compartment plans of the Phu Quoc Wreck.
Bugis, to this day (Horridge, 1985).
This wreck was dated to the

71
fourteenth or early fifteenth century and was almost certainly built in Thailand, although
probably under Chinese supervision or influence.

The Pattaya ship (Green and Harper, 1982), had bulkheads constructed from beech. The
particular species was not identified. Beech is found across central and northern Europe, the
Baltic, and throughout Asia. Japanese Beech is restricted to Southern China, Japan, Thailand
and Korea. This wreck is also thought to be of Thai origin.

The Quangzhou ship (Green, 1983) had a pine keel, a cedar hull, and a camphor
transom beam. This wreck from circa 1277 is believed to have been built in Southern China. It is
contemporary to the Java Sea Wreck. The construction appears to be similar, with 12 closely
spaced bulkheads being the most striking parallel, and yet the materials used are very different.
By way of interest, the aspect ratio of the Quangzhou ship was 1:3.1 compared to 1:3.4 for the
Java Sea ship.

From the wood identification alone, it seems unlikely that the Java Sea ship was
constructed in China. Teak was one of the main timbers used for the construction of vessels in
Thailand, due to its availability, workability, excellent durability, and resistance to the toredo
worm. The absence of teak on the Java Sea Wreck suggests that it may not have been
constructed in Thailand either.

6.4.2 Stones Anchors

As previously mentioned, two stone

anchors were found in Grid K3. They were lying
side by side, parallel to the axis of the ship. The
larger anchor measures 1180 mm long, 480 mm
wide, and 280 mm thick. It has been shaped
out of a hard sandstone. In profile and in
section it looks like a ships rudder. Barely visible
lines chiseled into the stone strangely
correspond to the pintel straps of a rudder. The
other anchor is shaped out of a large block of
limestone. It is 1120 mm long, 430 mm wide, 220
mm thick and has a similar shape to the Stone anchor shaped from sandstone, in situ.
sandstone anchor. It is interesting that a porous
and relatively light limestone has been used for an anchor, where weight is a prime
consideration.

A stone anchor stock was found on the

Breaker Shoal Wreck in the Philippines, which has
a very similar cargo and date to the Java Sea
Wreck. That stone is described as a slab,
measuring 3 m in length, in the shape of an
angular column thickening at the middle and
narrowing at both ends (Dupoizat, 1995). It is
similar to anchor stones found on the Quanzhou
Wreck, dated circa 1277. The anchor stones from
the Java Sea Wreck are nothing like them.
Stone anchor shaped from limestone, in situ.

72
 A bas relief on the Bayon of Angkor in
Cambodia clearly depicts an anchor being
lowered from a large seagoing vessel. The
anchor appears to have a large stone slab
fastened near the crown (more like a modern
danforth anchor than an admiralty pattern
anchor) that has the stock well down the shank.
This flat stone slab depicted on the bas relief
bears far more resemblance to the Java Sea
Wreck anchors than to examples on the
Quanzhou and Breaker Shoal Wrecks, which are
both thought to be Chinese vessels.

6.4.3 Fastenings

Chinese ships were fastened exclusively

with iron bolts and nails (Waters, 1946).
Sometimes wrought iron straps were also used to
fasten hull planks to bulkheads. Horridge (1985)
states that the typical Southeast Asian cargo
vessel, which he calls a jonque, is distinctively so
due to the absence of iron fastenings. He
suggests that Southeast Asian shipwrights did not
use iron, as it rusts rapidly in warm tropical waters.
The temperature difference between
temperate and tropical waters would actually
have little effect on the oxidation of iron.
Anchor depicted on the Bayon of Angkor,
Furthermore, the Chinese skillfully covered nail Cambodia.
recesses and sealed edge joints with a lime-
based caulking compound called chu-nam to
inhibit oxidation of the iron fastenings (Li, 1989). Thai ships of the fourteen to sixteenth centuries
have hull and bulkhead planks edge-joined with dowels, and yet the inner strakes are attached
to the bulkheads with iron nails. This is the case with the Pattaya wreck (Green, 1983), the Phu
Quoc wreck (Blake and Flecker, 1994), the Central Gulf of Thailand wreck (Flecker, unpublished),
and the Royal Nan Hai wreck (Sjostrand, pers. comm.). Worcester (1947), in describing methods
employed in building Chinese junks, comments, The side planks are then placed longitudinally
in position and hove down by a Chinese windlass, after which they are firmly nailed to the edges
of the bulkheads. So it may be said that the presence of dowel edge-joining precludes
Chinese construction, but the presence of iron fastenings does not preclude Southeast Asian
construction.

This conclusion can perhaps be further refined. There is no evidence that ancient
Indonesian ships used iron fastenings at all. Some craft are of lashed-lug construction, and some
are fastened exclusively with dowels and treenails. Horridges observations, therefore, probably
refer to Indonesian rather than Southeast Asian vessels. The dowel holes in chunks of wood
recovered from the Java Sea Wreck indicate that it is not a Chinese vessel, but other
characteristics must be examined to determine in which part of Southeast Asia it was
constructed.

73
 6.4.4 Bulkheads

Compartmentalized design has long been attributed to the Chinese, primarily through
the study of early Chinese manuscripts (Needham, 1971). The Thai wrecks mentioned above also
had bulkheads. Green (1978) suggests that as a result of the drastic reduction in the export of
Chinese porcelain caused by the so-called Ming Ban on private overseas trade, which came
into effect in 1371, and the Interregnum (1435-65) calling for the closure of the imperial kilns at
Jingdezhen, Chinese emigrants in Southeast Asia may have been building fleets of trading
vessels to establish a new basis of trade in the region. Such ships would naturally incorporate
aspects of both Chinese and local ship building traditions. Most of that new trade was in Thai
ceramics (which are also heavily influenced by Chinese ware, particularly Sawankhalok ware
which has strong parallels to Longquan ware from northern China), so it is reasonable to assume
that the vessels required to implement the trade were also built in Thailand.

The Java Sea Wreck had at least twelve closely spaced bulkheads. If it was indeed built
in Thailand, then bulkheads have been a feature of Thai ship construction from well before the
boom in the Thai ceramics trade in the fourteen and fifteenth centuries. With the close proximity
of China and continuous interaction over preceding centuries, it would not be surprising if
Chinese techniques were adopted well before this time. It is also quite feasible that Thai
features, such as butt-jointed multi-layered hull planking, were adopted by the Chinese.

Plan of the Central Gulf of Thailand Wreck, a Thai vessel with bulkhead construction.

The attributes of early Indonesian vessels are determined mostly from a very limited
number of temple carvings, such as the ninth century relief at Candi Borobudur in Java, which is
thought to be a coastal trading vessel rather than an ocean-going ship. Obviously, nothing can
be ascertained about internal structure from such sources. According to Manguin (1980),
sixteenth- and seventeenth-century Dutch and Portuguese travelers describe Indonesian vessels
in considerable detail, but nowhere is there mention of bulkheads; whereas early foreign writers
describing Chinese vessels are enthralled by them (Needham, 1971). Furthermore, no modern

74
Indonesian vessel still constructed by the dowel edge-joined method incorporates bulkheads in
their design. So it seems unlikely that compartmentalized construction was an Indonesian
shipbuilding feature.

There is one other Thai wreck with noteworthy similarities. The Ko Si Chang III wreck of the
fifteenth or sixteenth century is estimated to have had sixteen bulkheads, creating
compartments approximately 1.2 m wide. The aspect ratio of 1:3.3 is almost the same as the
Java Sea Wreck, although it is considerably smaller at 20 m LOA.

6.4.5 Conclusion

It is impossible to conclusively determine the country of origin of the Java Sea Wreck.
There was no surviving cohesive hull structure, and there are minimal contemporary Asian wrecks
with which to compare.

The wood identification indicates that the vessel was not built in China. But the absence
of teak may also place doubt on Thailand as the place of origin. The stone anchors seem to be
non-Chinese from the very limited information available. The dowel holes in two surviving pieces
of timber and the absence of iron fastenings, some of which should have survived in concretion
form had they been present, adds to the likelihood that China is not the place. While common
in Chinese and Thai ships, bulkheads are not known to have been utilized in Indonesian
shipbuilding, although there is almost no evidence on the internal layout of such craft.

It is almost certain that the Java Sea Wreck was not built in China. Both Indonesia and
Thailand are possible countries of origin. Hopefully, future discoveries will shed more light on
early construction techniques practiced by these two countries.

6.5 Dating the Ship

Both Roxanna Brown and Dr. John Miksic date the ship to the mid-thirteenth century,
based on a stylistic analysis of the ceramics cargo. In order to check this hypothesis, a resin
sample was sent to Beta Analytic, Inc., a radiocarbon-dating laboratory in the USA, for C14
content measurement and C13/C12 ratio analysis (see Appendix H).

Wood or ivory could also have been used for analysis. However, the wood was perhaps
too degraded to give meaningful results, and ivory has been known to exchange C14 with sea
water over prolonged periods and hence provide misleading data. Wood can also predate a
ship by 100 years or more, if old timber is used for its construction. Resin, on the other hand, is
unlikely to have been kept in storage for any significant period before being shipped, and the
sample has suffered minimal degradation.

The C13/C12 ratio is used to correct the measured C14 age and to calibrate the result to
the conventional calendar. From the analysis there is a 95% probability date range of AD 1215
to 1405. The 68% probability date range is AD 1265 to 1310. So it can be said with some
certainty that the ceramics cargo does not predate the thirteenth century. This is significant, as
several of the finer wares did suggest an earlier date. While it is certainly feasible that such
wares were manufactured over a longer period than previously thought, it is unlikely that
production continued on into the fourteenth century. Thus, the thirteenth century remains the
most likely date for the wreck and its cargo.

75
 There is one other Thai wreck with noteworthy similarities. The Ko Si Chang III wreck of the
fifteenth or sixteenth century is estimated to have had sixteen bulkheads, creating
compartments approximately 1.2 m wide. The aspect ratio of 1:3.3 is almost the same as the
Java Sea Wreck, although it is considerably smaller at 20 m length overall.

6.4.5 Conclusion

It is impossible to conclusively determine the country of origin of the Java Sea Wreck.
There was no surviving cohesive hull structure, and there are minimal contemporary Asian wrecks
with which to compare.

The wood identification indicates that the vessel was not built in China. But the absence
of teak may also place doubt on Thailand as the place of origin. The stone anchors seem to be
non-Chinese from the very limited information available. The dowel holes in two surviving pieces
of timber and the absence of iron fastenings, some of which should have survived in concretion
form had they been present, adds to the likelihood that China is not the place. While common
in Chinese and Thai ships, bulkheads are not known to have been utilized in Indonesian
shipbuilding, although there is almost no evidence on the internal layout of such craft.

It is almost certain that the Java Sea Wreck was not built in China. Both Indonesia and
Thailand are possible countries of origin. Hopefully, future discoveries will shed more light on
early construction techniques practiced by these two countries.

6.5 Dating the Ship

Both Roxanna Brown and Dr. John Miksic date the ship to the mid-thirteenth century,
based on a stylistic analysis of the ceramics cargo. In order to check this hypothesis, a resin
sample was sent to Beta Analytic, Inc., a radiocarbon-dating laboratory in the USA, for C14
content measurement and C13/C12 ratio analysis (see Appendix H).

Wood or ivory could also have been used for analysis. However, the wood was perhaps
too degraded to give meaningful results, and ivory has been known to exchange C14 with sea
water over prolonged periods and hence provide misleading data. Wood can also predate a
ship by 100 years or more, if old timber is used for its construction. Resin, on the other hand, is
unlikely to have been kept in storage for any significant period before being shipped, and the
sample has suffered minimal degradation.

The C13/C12 ratio is used to correct the measured C14 age and to calibrate the result to
the conventional calendar. From the analysis there is a 95% probability date range of AD 1215
to 1405. The 68% probability date range is AD 1265 to 1310. So it can be said with some
certainty that the ceramics cargo does not predate the thirteenth century. This is significant, as
several of the finer wares did suggest an earlier date. While it is certainly feasible that such
wares were manufactured over a longer period than previously thought, it is unlikely that
production continued on into the fourteenth century. Thus, the thirteenth century remains the
most likely date for the wreck and its cargo.

77
7. ARTIFACTS by Michael
Flecker

The positions of individual artifacts are depicted in the Artifact Distribution Plan (see
Appendix C).

7.1 Iron

The most prominent feature of this wreck site is the iron cargo. Concreted blocks of iron
up to 4 meters long, 1.5 meters wide, and projecting up to 1.5 m above the wreck mound mark
the position of a ship that plunged to the seabed approximately 750 years ago. The iron cargo
concreted together in the ships compartments before the hull and bulkheads disintegrated and
have maintained that shape ever since.

The concretions cover an area approximately 18 m long and 9 m wide. Most blocks are
well consolidated. However, there is quite a wide scattering of constituent parts in Grids F4, F5
and G5. This area is thought to be the stern of the ship. The stern, being higher, may have
collapsed before the iron had a chance to fully consolidate. Hence, the scatter of smaller
pieces there.

Several different types of iron cargo have been identified. There are cauldrons of
different sizes, which have been recorded on a number of shipwrecks lost in Asia. But the bulk of
the iron cargo consists of bundles of bars, which were probably blanks intended for forging by
Indonesian blacksmiths.

7.1.1 Cauldrons

Cauldrons are distinguished from woks in that they have a wide rim, whereas woks have
no rim at all and tend to be shallower. Cauldrons of various sizes can be clearly seen in
Concretions 4, 6, 9 and 19 in the Iron Concretion Plan. Several small stacks were also found
loose, and two of those were recovered for measuring.

A stack of cauldrons after the surface concretion layer

Concreted stacks of cauldrons, in situ. has been removed.

77
 After the thick concreting material made up of iron corrosion products, sand, and
calcareous marine growth was carefully chipped off the outside of one stack, 7 individual
cauldrons could be discerned. There was no base metal remaining. The cast iron had
transformed to a black material that in some areas was brittle and in others could be washed
out with a jet of water. This black material did, however, maintain the original shape of the
cauldrons. They were 510 mm in diameter and 2.6 mm thick. The wide rims were decorated
with two sets of three concentric rings, either cast or etched finely into the surface.

A smaller stack of 5 cauldrons was in a similar condition. Each one measured 320 mm in
diameter and 2.0 mm thick. They were decorated in the same fashion as the larger cauldrons. It
is interesting to note that small pieces of rattan were found embedded in the concretion near
the top of the stack. The stacks of cauldrons may well have been bound with rattan, and may
also have had rattan handles to facilitate carrying.

Larger cauldrons can be seen in Concretions 6 and 19. It is estimated that their diameter
is close to 700 mm. Large cauldrons were used to prepare rice wine (Brown, pers. comm.).

7.1.2 Iron Bars

Most of the iron cargo consisted of bundles of bars. Two different types were identified,
although there may be different sizes of each type.

An amorphous iron concretion, measuring approximately 200 x 120 mm, was carefully
dissected to reveal four rectangular iron bars, each 160 x 28 x 28 mm. They were bound
together with three strips of rattan and wrapped in a coarse woven material that looks very
much like hessian. Another contained five bars, each 220 x 30 x 22 mm and rounded at the
ends. This one was also tied with rattan, but wrapped in strips of cane. It is thought that these
iron bars are blanks for making tools, weapons, and agricultural implements for use in Java and
possibly for trade throughout Indonesia. Only the two stacks of rectangular bars were
recovered, although it is likely that at least one concretion block consists entirely of them.
Unfortunately, most of the concretion blocks have fused together so solidly that it is impossible to
distinguish the constituent parts.

Rectangular iron bars bound with rattan. Detail of the material covering the rectangular iron bars.

78
A broken concretion showing bundles of bars stacked Individual concretions of trapezoidal bars.
end-to-end.

Far more common are conical concretions, typically 400 mm long. Most of the
scattering of concretions towards the stern consists of these, and they can be made out in many
of the concretion blocks. A very interesting lump of concretion that has been broken open
(possibly by fishermens explosives) shows clearly how these cones were stacked end-to-end in
the hold of the ship.

Several of these concretions were carefully dissected to reveal their contents. Each was
found to contain a bundle of iron bars with a trapezoidal section and tapering along their
length. The bars were bundled in the same orientation; hence, the conical shape. The bundles
were held together with two loops of twisted rattan, and impressions in the concretion suggest
that they were wrapped in leaves or strips of cane. The individual bars were separated by a
white material that is thought to be a lime compound, such as was used for caulking and
sealing iron fastenings in ships. This compound is known as chu-nam and consists of lime and
tung oil. This would be an excellent method for preventing the iron from corroding in storage
and during the long voyage from China to Java. Iron bars of this unusual shape could have
been forged into large one-sided blades, like a modern parang.

79
Cross section of a bundle of trapezoidal iron bars. Long section of a bundle of bars clearly showing the rattan
binding.

Artists impression of a wrapped and bound bundle of trapezoidal bars (scale in cm.).

7.1.3 Ax

One concretion was formed by a small ax head. It measures 145 mm long, 42 mm wide
at the handle, and in profile varies from 55 mm at the handle to 61 mm at the cutting edge. As
there was only one found, it is likely that this was an item of ships equipment.

7.2 Ivory

80
 In the southwest quadrant of the site,
16 pieces of ivory were recovered. From the
reasonably concentrated nature of the finds
(Artifact Distribution Plan, Appendix C), it
seems that the ivory was stowed in the stern
of the ship on the starboard side.

The ivory is in the form of unworked

elephant tusks. The largest piece measured
590 mm long and 130 mm in diameter.
Unfortunately, ivory deteriorates badly
underwater, tending to exfoliate. As a result,
many of the pieces found are hollow or just
the outer layers of a tusk. In several cases
the solid tip of a tusk has survived. Elephant tusks, approximately 400 mm long.

Two pieces have been cut by hand. One is a 245 mm long tip of a tusk that has had the
very end neatly cut off. There are also cut marks on the surface of the piece, and it is hollow.
The other is the outer layer of one half of a tusk, which had had a portion roughly shaved off.
The tool marks are still quite clear.

7.3 Resin
Eight pieces of resin were recovered. Like the ivory, they tended to be in the southeast
quadrant of the site. However, they were more spread out, probably due to their low specific
gravity. In fact, one small piece was found at the opposite end of the site in Grid N2. As with the
ivory, it is likely that the resin was stowed in the stern of the vessel, probably on the starboard
side.

Hundreds of years of immersion have not had that much effect on the resin. There has
been a bit of erosion, and the outer millimeter or so has become a soft powdery pale brown.
But within, the solidified tree sap remains glassy, and the fragrance remains distinctive. The
largest piece is 410 mm long by 180 mm across. Most of the other pieces are half to quarter that
size. Two of them have very well defined cut surfaces on one end.

81
 A block of resin, 410 mm long.

7.4 Balance Weights and Bars

In all, 14 balance weights were found from at least 4 different sets. This is not sufficient to
conclude that they were an item for trade. It is more likely that they were the personal property
of various merchants on board. Of the 14, 12 are made of a copper alloy. They have been very
finely cast and finished. To obtain an exact weight, holes have been drilled out of the center of
both sides and then plugged with a copper alloy of slightly different composition. There are two
types: one plain and the other with a scalloped edge. Not surprisingly, all of the heavy copper
alloy weights were found in close proximity to the iron concretions that mark the original position
of the ship. Most of them were found to the north of the concretions. However, a scatter of over
25 meters confirms that they were not stowed in one place as a trade consignment.

One unusual balance weight found in Grid H5 has been carved out of milky quartz. As
with the copper alloy weights, holes have been bored out of the center of each side to obtain
the correct weight. Another very finely crafted weight is of composite construction and is only
26 mm in diameter. It appears to be made of marble with copper alloy inserts plugging the
usual central holes. Judging from the perfect fit, the alloy may have been cast in the holes.

82
 Different types of balance weights plain Cu alloy (top left), scalloped Cu alloy (bottom
left and right), composite (bottom center), and quartz (top right).

By weighing all of the balance weights it becomes apparent that there are six distinct
weight categories. These are tabulated below (weights in grams).

Plain Cu alloy Scalloped Cu alloy Composite Quartz

1. - - 27 -
2. 37 - - -
3. 70 71 - -
4. 171, 190 189 - -
5. 340, 363, 364 384 - -
6. 760, 762 - - 727

Discrepancies in the weights may be accounted for by the differing degrees of corrosion
of the copper alloy. The large quartz weight has been eroded and in places penetrated by
marine borers.

83
 Two balance bars were also found.
While being very different in size, they are
exactly the same shape, a round bar tapering
from the center to approximately half the
diameter at both ends. The ends are turned
and there is a flat pivot point in the center that
is broken on both pieces. The small balance
bar was found in Grid K3 and is 150 mm long,
and the large one was found in Grid E2 and is
221 mm long. Presumably the different sized
bars were used for different weight categories
and both types may have been in a set,
although contextual evidence suggests this Balance bars.
was not the case.

7.5 Figurines and Finials

Two very interesting copper alloy

figurines were found on the site. The first was
from Grid J5. It is a man, 65 mm high, sitting
cross-legged and holding with upstretched
arms a corner support for a small platform,
perhaps an altar. He is one of an original set of
four, which would have been legs for the
platform. He is very much stooped under the
load, his back being bent double. His large
head is quite un-Chinese, displaying big eyes,
a small flared nose, a down-turned thick-
lipped mouth, a pointed chin, and long ears.
No parallel has been found.

Various views of platform leg figurine (scale in cm).

Platform leg figurine. Close up of face.

84
 The other figurine was discovered in Grid H2. Measuring 172 mm long, it is a figurine of a
woman riding a sea creature. The creature bears an uncanny resemblance to those spouting
monsters depicted on European sea charts of the sixteenth and seventeenth centuries, having a
large mouth, bulbous nose, and three pronged dorsal, pectoral, and tail fins. The rider sits with
one leg crossed over the other. Her right hand rests on her knee. Her left arm is missing. She has
long hair flowing half way down her back and looks out to the left, her head slightly bowed.

Bronze figurine. Bronze figurine (scale in cm).

Pre-Majapahit folklore explains the role of this fine lady and her unusual steed. An
ancient Javanese text, the Sri Tanjung, describes the afterlife journey of a princess who was killed
before she had made preparations to liberate her soul and therefore could not attain salvation.
Her soul was rescued by a monstrous fish and carried to the next world. This scene is depicted
on a diverse range of Majapahit relics, such as a bronze hanging lamp, a carved stone window
grill, and a gold modesty cover (Miksic & Soekatno, 1995).

The Majapahit and pre-Majapahit religious ceremonies may explain the presence of two
unusual finials. A small one, 90 mm long, was discovered in Grid H2. It is a badly corroded
geometric shape, although some elaborate cast decoration can still be seen on one side.

Bronze finial. Bronze finial (scale in cm).

85
 The other was found in Grid E2. This one has the appearance of a masquerade party
mask with holes for the eyes and distinct upturned eyebrows. It has a wavy cross section,
incorporating a stylized mouth and nose.

Mask-type finial (scale in cm). Mask type finial.

These finials were originally mounted on wooden staffs and were carried by priests during
religious ceremonies. Finials from the Majapahit era have been unearthed in eastern Java and
in Bali. They often depict nagas, a mythical cobra-like creature. Bronze finials with loose rings
that rattled (khakkhara) adorned the top of walking staffs carried by mendicant priests (Miksic &
Soekatno, 1995).

7.6 Trays
Unfortunately, no complete trays
were found. The long exposed site
afforded no protection for these thin
copper alloy artifacts. Only the thicker
rims, and in a few cases the central
portion, of a variety of trays remained. A
tray is defined loosely here as being a
shallow dish with a distinct raised rim. A
few rims have sufficient body still attached
to confirm this identification, and all the
thin sheet fragments are relatively flat.
Other rim types without identifiable
fragments may be from deeper vessels, Tray fragments.
such as copper bowls or cauldrons.

86
 There are four distinct types of rim. The first is a half-round section and is definitely from a
tray. The rim thickness is approximately 1.5 mm, while the tray body is less than 1 mm thick. The
rim is decorated with a faintly impressed double line on the inner surface. These trays would
have had a diameter of approximately 420 mm.

Another definite tray is heavier and unadorned. It has an everted rim, approximately 12
mm wide and 3 mm thick. The body of the tray is 1.5 mm thick. The diameter would have been
approximately 520 mm.

The third type is very solidly made; but as only rim fragments survived, it cannot
definitively be termed a tray. The rim is squared and beveled in tiers and is typically 4 mm thick.
The original vessel would have had a diameter of only 250 mm, suggesting that perhaps it was a
solid bowl rather than a tray.

The fourth type is represented by only two small rim fragments. They are roughly half
round, only 8 mm wide and 4 mm thick, suggesting that they may be cast. The diameter would
have been 160 mm.

Thin copper alloy sheets, heavily corroded around the edges, are thought to be the
central parts of trays. One large piece, measuring 290 x 220 mm and approximately 1 mm thick,
may be the center portion of the first type of tray described above. Three groupings of
concentric circles decorate this fragment: a central group of 30 mm diameter, a middle pair of
160 mm diameter, and a very faint outer pair of 340 mm diameter. One slightly bowed piece
with a thickness of 2 mm may be from the second type of tray described. Like the rim, it is
undecorated.

During the Majapahit era, bronze trays with incised circular decorations were used as
offering trays in religious ceremonies (Miksic & Soekatno, 1995). Some were more elaborate with
incised lotus blossoms and scrolls.

7.7 Gongs
Only two gongs were recovered:
one more than half intact and the other
just a highly corroded central portion.
They were found relatively close
together in the stern area of the ship in
Grids B3 and D3. The better preserved
piece has an overall diameter of 270
mm, and the central raised dome is 60
mm in diameter. The fragment has a
slightly larger raised dome with a
diameter of 65 mm. There is no sign of
decoration on either piece.
A bronze gong.

Bronze gongs have been found on only two other wrecks to the authors knowledge.
One was a Thai vessel lost in the central Gulf of Thailand, which yielded one nearly complete
unadorned gong of 380 mm diameter and one gong fragment (Flecker, 1991, unpublished). This

87
ship was probably bound from Ayudhya to the Philippines. Others with a six-pointed star
decoration were found on a sixteenth century trading vessel, probably Chinese, that was lost on
Royal Captain Shoal in the Philippines (Goddio, 1988). It is postulated that this ship was bound
from China to Borneo with a primary cargo of porcelain.

Gongs were manufactured in China from the sixth century and grew to be extremely
popular throughout China and Southeast Asia. Made from a copper-rich alloy, they were initially
cast, then repeatedly heated, hammered, and filed until the desired sound quality was
achieved. They were used extensively in civil and religious ceremonies, Buddhist rituals in
particular. A poem from the Majapahit era, the Nagarakrtagama, describes a posthumous
ceremony in honor of the poets grandmother. The rite of parisharma was performed to recall
the soul of the departed. Parisharma is a rhythmic, possibly vigorous, ring dance performed by
dancers in a trance-like state, accompanied by the chanting of sutras and the rhythmic beating
of drums and gongs (Miksic & Soekatno, 1995).

7.8 Copper Ingots

Strangely, only three copper ingots were found. Two of them are small conical ingots,
found widely separated in Grids F6 and I3. They flare out slightly at the base and are decorated
with molded concentric circles midway between the base and the peak. They weigh 174 grams
and 231 grams. Similarly shaped but undecorated lead ingots have been found on many Thai
wrecks of the fifteenth and sixteenth centuries and on a few European vessels of the early
seventeenth century. There is some speculation that they were used for currency.

Conical copper ingots. Copper bar ingot.

A single large copper ingot was also found in Grid I3. It is the typical tapering
rectangular shape of a modern gold or silver ingot. The surface is pitted by corrosion, and there
are no markings. The ingot measures 195 x 88 mm at the base and is 52 mm high.

It is thought unlikely that other ingots were removed during earlier excavations as they
lay down at the clay layer and hence could not be reached without excavation equipment. It
remains a mystery why there were so few on board the ship, unless they were the lost remnants
of some previous cargo.

87
7.9 Miscellaneous Copper Alloy Artifacts

There are six artifacts that defy

ready classification. One is a 230 mm
long hook-shaped piece found in Grid
K3. What appears to be a round pan
holder is partly missing from the end of
the hook. There is a possible pivot point
at the top, so this could be a tray
holder for a large set of weighing
scales. There is a decorative knob at
the base of the hook.

Possibly part of a weighing scale set.

Another hook-shaped artifact is possibly a handle of some kind. It has a rivet driven
through one flattened end with evidence of some form of textile under the rivet head. The
wood-like grain structure of the piece indicates that it has been wrought. It was found in Grid I2.

A complete rim of 110 mm

diameter is probably that of a small
copper alloy bowl. None of the
interior has survived.

A small tubular piece of

copper alloy, slightly pinched at the
waist, may be a bushing for a small
wooden pulley. It is from grid G3.

Possibly a handle.

An unusual piece resembles the

base of a large candlestick and is
badly corroded and highly porous. The
base is slightly domed, and is
decorated with concentric circles and
striations. The neck of the piece is
flanged. The copper alloy casting has
been filled with another metal, possibly
an alloy of lead, which is now exposed
where the copper has corroded away.
It was found in Grid D0. Its function is a
complete mystery.
Unidentified copper alloy artifact.

88
A conical hollow piece, open at each end, cannot even be guessed at. The bottom seems to
be the original surface, but the top has been broken off or eroded away.

7.10 Glassware

A total of five glass artifacts were recovered, only two of them of the same type. Those
two are deep olive green shards with a sharp conical peak, very similar to the base of a
European onion bottle. The glass is less than 1 mm thick and has minimal bubble inclusions.
These shards were spread way apart in Grids B6 and I9. Another base type shard was found in
E9. It is a fairly clear dark green glass with many bubble inclusions. The thickness varies from 1 to
2 mm, and the central peak is a relatively flat dome.

The only intact glass artifact appears to be some kind of lid. It has a domed top and
angled sides with a sharp lip between the two. The lid has a diameter of 52 mm and a thickness
varying between 3 and 5 mm. The glass is a very pale green, full of bubbles. It was found in Grid
H7.

A range of glass artifacts lid (bottom left), onion bottle (top left, bottom center), thicker
onion bottle (top right), and stick (bottom right).

The last glass artifact is a small stick, 12 mm in diameter and 44 mm long, found in Grid J4.
It has been formed by rolling very thin sheets of yellow and green glass together, giving a spiral
cross section. The outer surface is deeply pitted.

Glass from the so-called onion bottle and the stick has been analyzed by X-ray
florescence at the Chemistry Department of the National University of Singapore and compared
to glass samples excavated from the fourteenth century stratum at the Fort Canning site in
Singapore; they appear to be of Chinese origin (Miksic, 1996). X-ray florescence is a non-

89
destructive qualitative technique where the first two principal components are plotted using the
relative concentrations of 17 elements in order to compare directly to samples of known origin.
The results are plotted below.

16

14

12
PRINCIPAL COMPONENT 1

10
ONION BOTTLE SAMPLE

8
STICK SAMPLE

FORT CANNING SAMPLES

6

0
0 1 2 3 4 5 6 7 8
PRINCIPAL COMPONENT 2

X-ray florescence analysis of glass.

The onion bottle sample is very similar in composition to the Chinese Fort Canning
samples, to the extent that it can be confidently concluded that it was made in China.

Evidence of glass making in Southern China goes back to the third century, when it was
described by the contemporary authors Wan Zhen (Strange Things of the South) and Ko Hung.
The Tai Ping Yu Lan reports that Emperor Wu-Ti (140-87 BC) sent emissaries to barbarians in
southwestern China to learn glass making techniques (Francis, 1986). Zhao Rukuo, harbormaster
of Canton in 1225, noted that the Chinese merchants exported glass beads and bottles to
Borneo and the Philippines (Hirth and Rockhill, 1911) (from Miksic, 1996).

The stick sample is very different from the Fort Canning samples. Glass was also
manufactured in India and in the Middle East at the time, but there is no firm evidence of glass
manufacture from primary materials anywhere in Southeast Asia during the premodern period. It
is strongly suspected, though, that the Javanese in the late first millennium A.D. engaged in this
activity (Miksic 1996).

90
 7.11 Rocks
Every rock found on the site was brought to the surface. They vary in size from river
rounded stones weighing 5 kg to fingertip-sized pebbles. Some of the rocks have obviously been
hand shaped, and these have been treated separately in Section 7.12. The total number
recovered was 64. Only half of those could be classed a ballast stone size, and, with rare
exceptions, those are well rounded and waterworn. With the massive iron cargo it is plain that
there was no need for stone ballast on the last voyage. The larger stones may be remnants from
a previous voyage where ballast was necessary, and many of the small rocks may be chips and
pebbles mixed with the ballast. The identification of many of these rocks by Gina Rockett, the
curator of the E. de Clarke Geology Museum at the University of Western Australia, and her
associate, John Ridley, provides other possibilities as follows:

a) Eight small and odd-shaped rocks are fossils or fossil-bearing mudstone. One is very evidently
the fossil of a tiny crab. The fossilized inclusions have a dark brown to reddish color and when
tapped give a distinct ping, being light and hard. The associated mudstone is brown to
light brown and is soft enough to have been attacked by marine borers. These rocks are
almost certainly native to the site.

b) Four small, highly porous, gray rocks are readily identifiable as pumice. With Krakatoa
Volcano 120 km from the site, this is not an improbable find.

c) Three small and one large (160 mm x 130 mm) irregular but smooth rocks may also be of
volcanic origin. Dark brown, and in some cases slightly lustrous on the surface, a chipped
piece revealed a rust-colored outer edge with some iridescent dark blue and green
inclusions. They may be iron manganese chert nodules. There is a slight possibility that these
rocks are native to the site.

d) Seven water-worn rocks (the largest being 130 mm x 60 mm) are of sedimentary origin. They
are a form of siltstone known as greywacky and have sulphide crystal inclusions. These rocks
are definitely not native to the site and therefore may be the smaller fraction of a previous
ballast cargo.

e) Three small, angular, thumb-sized rocks have been identified as the mineral corundum
(Al2O3). It is almost certain that they were onboard the ship. Corundum, when occurring
with magnetite, is known as emery and is harder than any other common mineral. It is used
as an abrasive. Rubies and sapphires are gem-quality corundum.

f) Six small, heavy rocks with a reddish-purple surface and crystalline interior have been
tentatively identified as rhodonite, a calcium manganese silicate. They were all found within
meters of each other. When polished, rhodonite is regarded as a gemstone; therefore, this
rock may have been part of the cargo, possible a private trade item judging from the limited
quantity recovered. An attempt to polish one of these rocks was unsuccessful, as it broke
into pieces due to internal weathering.

g) Three rocks, or rather lumps, measuring roughly 120 mm x 100 mm present a bit of a dilemma.
The outer surface is light brown. One example is layered in solidified waves, as if a viscous
flowing material has hardened before another layer flowed over that and hardened in five
distinct episodes. There is much finer layering within the lump. Another has a globular-
domed surface and concentric fine layering. The material is white and chalky, and under a
magnifying glass there appear to be minute quartz crystal inclusions. These rocks were found

91
 close together, in Grids B1, B2, and C1. Their individual characteristics are sufficiently
different to make it unlikely that they derived from one larger piece. On contextual grounds,
they were initially thought to be some form of processed mineral that was part of the ships
cargo, possibly lime or some kind of flux for smelting. It has since been suggested that they
are rhyolite bombs, lava fragments thrown into the air by volcanic explosions. Pumice, which
is a rhyolite lava blown into a sponge-like consistency by the release of gases, has been
positively identified on the site, thereby supporting this theory.

h) An orange-to-brown rock presents another identification problem. Dozens of pieces were

recovered from all over the site, generally down at the clay layer and sometime penetrating
it. These have not been included in the total number of rocks recovered. The outer surface
is very irregular. The largest piece measures 220 mm x 130 mm and is shaped like a bent shoe.
It is completely hollow with a wall thickness of 10 to 20 mm. Other pieces are similarly
convoluted. They appear to have been cast in the odd-shaped wormholes that riddle the
clay seabed. But for the shape of some of the pieces, they could have been classified as
weathered laterite. Angular inclusions in some pieces may indicate a poorly consolidated
calcareous breccia. Contextual evidence, however, indicates that these rocks are native to
the site. Another interesting thought is that they are pieces of kiln slag from the Chinese iron
foundry that produced the cauldrons and bars that make up a large portion of the cargo.

i) Three pieces of water-worn milky quartz were found widely spread out in Grids B1, G3, and
H2. The larger piece, measuring 100 mm x 90 mm, was broken, revealing an angular interior.

j) One raw quartz crystal was

recovered from Grid H8. The main
hexagonal crystal is 60 mm long, and
there is one small offshoot crystal.
The surface is slightly stained with a
brown iron corrosion product.
Although there are some internal
fractures, there are few impurities
within the crystal. This was probably
a curio belonging to one of the ships
company.

k) Only one rock that is slightly

weathered but not water-worn is Quartz crystal.
readily identifiable as granite. It is
reasonably fine grained, and the weathered surface has an orange hue.

l) Most of the ballast-sized rocks are sedimentary, ranging from a course-grained sandstone to
the very fine-grained greywacky mentioned above. Two exceptions appear to be gabbro
and diorite with some quartz content, both igneous rocks. There is only one example of
each.

7.12 Sharpening Stones

Nine stone artifacts have been positively identified as sharpening stones. Five of them
are small and easily held in one hand, while the other four would have rested on the ground.

92
 The hand-held sharpening stones
are shaped out of a hard siltstone. Four
are more or less finger shaped, but one
of a lighter colored stone has the shape
of a flattened L. All have one surface
worn smooth, so they have all seen
active duty. They were well scattered,
lying in grids C1, E9, G0, I8, and J8.

The larger stones tended to be

water-worn, elongated pieces of rock.
One 350 mm long piece is completely
smooth along its full length on one side,
Small sharpening stones.
and has two well-worn hollows on the
other.

A wider rock is roughly chipped flat

on the base and well worn on the working
surface. A smaller stone is worn in two
places on both sides. One of the
sharpening stones is broken in two, and
had been for a long time, judging from
the encrustation of coral and oyster shells
on the broken faces. Three of the larger
stones were found fairly close together, 2
in Grid I3 and 1 in G3, but the other was
well away in Grid F-1.

Three sharpening stones and a stone roller (left).

One other similar stone has been hand shaped, although it is not actually a sharpening
stone. It is like a roller, with both ends tapered slightly as if to fit end bearings. The surface of this
so called roller has been worn smooth. It has a length of 280 mm and a diameter of 80 mm.

The fact that all of the sharpening stones were well used implies that they were part of
the ships equipment or perhaps belonged to soldiers assigned to protect the vessel, rather than
items for trade.

7.13 Organic Finds

Apart from the small chunks of wood that were the only remnants of the ships hull, and
some branches which may have been dunnage or perhaps an aromatic wood, there were very
few organic finds. Given the exposed nature of the site, it is surprising that there were any at all.
In fact, it must be suspected that some of the items that were found are intrusive.

Two organic finds were tentatively identified by Prof. Wee Yeo Chin and Dr. Ian Turner of
the National University of Singapore Botany Department. One nut found in Grid A6 is thought to

93
be canarium, or canary nut, which is typically used for food flavoring. Another very degraded
specimen was identified as the fruit from a nipa palm, which contains what are commonly
known as atap seeds. These seeds are sweet and chewy, and highly edible. The nipa palm is
found throughout Southeast Asia and grows prolifically in swampy river deltas.

An unidentified seed casing was found in Grid J8. It is woody, roughly spherical, and
appears to be internally divided into three cavities. Another unidentified nut is badly
deteriorated, but slightly resembles an almond.

The other organic finds include rattan that was used to bind the bundles of iron bars, and
a small piece of hessian-like material that was used to wrap up a bundle of four rectangular iron
bars. These were preserved by the encasing iron concretion.

7.14 Tin Ingots

Two unusual tin ingots were discovered. Why there were only two remains a mystery.

Badly corroded tin ingots.

One, recovered from Grid J5, is a truncated cone 110 mm in diameter and 30 mm in
height. The base has a scalloped edge, although this decorative feature is highly degraded on
one side. The ingot is badly blistered with yellow, white and crystalline corrosion products. Both
the base metal and the corrosion products have been analyzed by Miss Anna Fletcher, Senior
Material Scientist with the Materials Institute of Western Australia, for Dr. Ian MacLeod of the
Western Australian Maritime Museum. The metal was found to be almost pure with 99.6% tin, and
only a trace of lead (0.019%).

The second ingot was recovered three grid squares away in Grid G5. This one is a
cylindrical rod, 265 mm in length and 45 mm in diameter. It is heavily corroded, with many
blisters showing a radial crystalline structure. The rod ingot was found to contain 98.4% tin,
0.175% lead, and 0.006% zinc.

94
 The detailed results of the metal and corrosion products analysis for both ingots can be
found in Appendix I.

95
8.0 THE IRON INDUSTRY AND TRADE by Dr. Bennet
Bronson

8.1 The History of Iron Production in Asia

Iron is known to have first been made in India by 1000 B.C. (Chakrabarti, 1992), in central
and southern China by 600 B.C. (Wagner, 1996), and in Thailand and Vietnam by 500 B.C.
(Bronson, 1985). Its first appearance in peninsular and island Southeast Asia is later, but there is
no reason to doubt that many parts of the region were making and using iron by 100 B.C.

The raw materials for iron making are distributed quite unevenly within Southeast Asia.
This is especially true of large deposits of high-grade iron ores. Several areas have none,
including Java, Sulawesi and Mindanao (all of which have substantial lower-grade deposits) as
well as southern Vietnam and the Visayas (where even low-grade ores are quite scarce). Other
areas, however, are rich in excellent iron ores, notably Luzon, the Malay Peninsula, and a belt
that runs from northern Thailand through Laos to northern Vietnam and China. Although
available information on smaller high-grade deposits in the 10,000-500,000 ton range is very
incomplete, it seems reasonable on geological grounds to conclude that most of these occur in
the same areas as the multimillion-ton deposits. Useful quantities of rich ores also exist in Central
Thailand, Borneo and the islands between Borneo and Sumatra (Survey, 1970; Bemmelen, 1949;
MRDS-39, 1972; DMR, 1975).

Existing historical records indicate that commercial iron making during preindustrial times
was concentrated in areas with high-grade ores, like those of Phnom Dek in northern Cambodia
(Moura, 1883:43-6), or with unusual ores like the iron-nickel deposits of central Sulawesi
(Abendanon, 1917-8). Whereas almost all parts of Southeast Asia contain enough mediocre
sedimentary ores (many of them lateritic) for a competent early smelter to make sufficient iron
for local use, production on a slightly larger scale seems to have required rich contact-
metamorphic ores. This was not because iron made from lean ores was worse in quality but
because it was more expensive, requiring more fuel and labor to produce a unit of finished
metal. Hence, ore-deficient areas such as southern Cambodia and southern Vietnam may
never have been producers of iron.

It does not follow, however, that all areas with iron ores necessarily used them. Luzon
and West Malaysia, both of which possess excellent deposits of accessible iron ores, never seem
to have been iron producers on more than a strictly local scale. The case of Java is even more
striking. The magnetite sands along its southern coast are not only abundant enough to have
been considered for use by modern steel making firms but are of a type that was often preferred
by early smelters. Similar iron sands accounted for a large proportion of the iron produced in
traditional furnaces in southern India and Japan. Yet the Javanese have never been known to
smelt this or any other iron ore, from the time the Portuguese arrived in the early sixteenth
century through to the mid-twentieth century. Raffles (1830,1:240) is quite clear that in the early
nineteenth century Java produced no iron at all. As far as is known, every kilogram of iron used
for hundreds of years in one of the most densely populated parts of Southeast Asia has come
from other islands, the Asian mainland, or Europe.

This imbalance either of resources or interest in using them resulted in an active iron trade
in and around the Java Sea. In recent centuries, a number of regional centers are known to
have provided raw or worked iron to be fed into the traditional maritime trade network of
western Indonesia. Examples include the Ngaju Dayak of South Kalimantan, whose iron and

95
steel were fabricated at Negara near Banjermasin and exported by sea to other islands
(Hendriks, 1842); the West Toraja and ToBada of central Sulawesi, whose tools supplied the
Moluccas (Kruyt, 1901:148-9) and whose nickel-containing unworked iron, the famed pamor
luwu, was carried as far as northern Sumatra (Groneman, 1910:138); the inhabitants of Belitung
and Karimata, whose bar iron and much-esteemed parangs and axes were used throughout
Indonesia (e.g., Daghregister, v.1679: 121, 137, 178, 222, 229, 342; Reid, 1988: 110-1); and the
Minangkabau of Sumatra, noted kris exporters (Valentyn, 1726,5:2) and suppliers of iron to
several other parts of that island (Marschall, 1968: 130-2).

The same historical sources, the most important of which were produced in Batavia
(modern Jakarta) by the Dutch East India Company (or V.O.C.), make it clear that western
Indonesian sources could not keep up with demand for ferrous metals. The Journals or
Daghregisters of the V.O.C., of which a number of sixteenth-century volumes have been
published, contain numerous references to imports of iron to Java. In the decade 1673-1682, 9
Chinese ships from Xiamen, and Chincheo (either Quanzhou or Zhangzhou) in southern Fujian
province brought 55,214 iron pans, 22 hampers of pans, 92 iron shovels and 6 iron cannon to
Batavia as well as quantities of ceramics and other goods. During the same period, 47 Chinese
and European ships coming from Canton and Macao in Guangzhou province brought a further
4,125 pans, 6,542 nests of pans, 332 kettles, 50 knives and a case of needles. Still more iron pans
came in Chinese ships sailing from Japan. India supplied quantities of bar iron as well as nails
and hammer steel (Bronson 1986: 22). Ships coming directly from Europe brought iron cannon,
cannon balls, miscellaneous hardware and, again, nails. The V.O.C.s records are incomplete.
But they show unmistakably that in the seventeenth century, when Indonesian economies still
were very similar to those of the pre-European period before 1500, the region was by no means
self-sufficient in iron. A very substantial percentage of the iron consumed in Indonesia was
imported from East Asia, South Asia, and even Europe.

That the island region of Southeast Asia simply did not produce enough iron for its needs
seems an inescapable conclusion. Why this should have been so is not obvious. After all, as
pointed out above, the region is not especially poor in ore, and the conversion of ore into iron is
not such a difficult process: indeed, judging from the fact that almost every part of Eurasia and
Africa knew the smelting of iron in antiquity, it is a process that almost every society can (and
did) master. Further, Indonesians of the historical period were not at all primitive in terms of
fabricating objects from already-smelted iron. As Tome Pires noted in 1515, the swords and
daggers of Java were prized enough to be exported to Malacca in Malaysia and even to India
(Cortesao, 1944: 93, 179). But why, then, did ancient Indonesia not smelt enough iron to become
self-sufficient, as was most of Africa, India and Mainland Southeast Asia by the early first
millennium A.D.?

Part of the answer lies in technology. Due to Indonesias excellent maritime

communications and extensive trade links with East and South Asia, would-be iron makers there
must have been exposed to severe competition from low-cost producers in India and China
who were probably the most efficient in the world during preindustrial times. Further, both East
and South Asian iron makers were technically advanced in terms of the kinds of iron they made.
India led the ancient world in the production of steel, employing a variety of crucible processes
to make a uniform, medium-carbon metal that was unmatched elsewhere in the world before
the nineteenth century (Bronson, 1986). And China, beginning about 500 B.C., had evolved a
unique complex of iron-making technologies that would not be discovered in the West until
modern times (Wagner, 1996).

96
 8.2 Iron-Making Technology

The differences between the techniques used for making iron in China on the one hand
and in Southeast and South Asia on the other are crucial to understanding the importance of
the iron cargo on the Java Sea Wreck. For that reason, it is worth devoting a few paragraphs to
technical issues.

The first such issue relates to the fact that until the mid-nineteenth century most of the
iron used in the world was of the low-carbon variety known as wrought iron. This is characterized
by a high melting temperature, which makes it impossible to shape by casting; by relative
softness and ductility at red or white heat, making it easy to shape by forging; and by relative
toughness and low strength. A pure wrought iron contains almost no carbon and is in fact less
strong and hard than many bronzes. However, it has two advantages over any bronze or other
copper alloy: (1) because iron ores are very abundant, it is much cheaper, and (2) through
adding some but not too much carbon by one of several methods it can be converted to steel,
which in terms of combined toughness, strength and hardness is superior not only to bronze but
to most other ancient or modern materials.

Low-carbon wrought iron and its derivative, steel, were the only types of iron used in most
places for most of history. East Asia, however, began at an early date to employ a third type:
the very hard, brittle and moldable high-carbon metal known as "cast" iron which, though a
good material for making objects like kettles and stoves, is unsuitable for functions requiring
shock resistance, as in axes or swords, and which cannot be shaped by forging. The methods of
handling cast iron seem to have been unknown to the other technologically advanced nations
of antiquity, giving China a world monopoly on cast iron production for perhaps a thousand
years. The technology of cast iron was picked up by the Koreans and Japanese sometime in the
early first millennium A.D. and by Central Asians several centuries after that; it was not learned by
Europeans until about A.D. 1300. And yet even in China the market for a brittle, castable iron
was limited. As in other countries a great deal of the iron used by the Chinese was of the
forgeable wrought variety.

The second issue concerns the way in which forgeable iron is made, for it can be
obtained through either of two processes (Rostoker and Bronson, 1990). One is the so-called
"direct" process that was preferred by early Europeans, Africans, and most Asians, including
Indonesians. A mixture of iron ore and fuel (generally wood charcoal) was heated at a relatively
low temperature (between 1200 and 1250 degrees Celsius) in a low furnace, perhaps about a
meter or so high or perhaps a carefully shaped hole in the ground. Most such furnaces were
equipped with two or more medium-sized bellows. The bellows were worked while impurities ran
off in the form of liquid slag, with more fuel being added as needed, until a solid lump or "bloom''
of iron formed at the bottom of the furnace. At that point, the bellows were stopped and the
furnace allowed to cool until the bloom could be extracted. Historians of technology call such
furnaces bloomeries because of the nature of their product. After extraction, the bloom was
hammered while still hot until most of the slag was squeezed out. The result was a piece of
low-carbon (<0.1%) metal, generally of high chemical purity due to the low smelting
temperatures involved, ready for immediate use by a blacksmith.

The other method for making a low-carbon wrought iron, the "indirect" process, appears
to have been used in early times only in China, then later in Korea and Japan, and not until
recent centuries in the West. As its name implies, the process involved more than one stage.
The ore and fuel (either charcoal, mineral coal or coke) were first heated, almost always in a

97
true blast furnace, a larger and taller structure than the bloomeries used in the direct process, to
a temperature high enough actually to liquefy the iron (about 1350 degrees Celsius). Much
larger bellows had to be used, which meant either water power or a large number of bellows
operators. Due to the high temperatures reached, not only the slag but also the iron could be
tapped from the furnace in liquid form through a small opening in the lower part of the furnace.
This made it possible to operate such furnaces continuously, often for periods of weeks or
months, which saved fuel and increased per-day production.

The liquid metal that is tapped from any blast furnace cools quickly, solidifying as a piece
of brittle cast iron with a very high carbon content in the 3-5% range as well as significant
amounts of phosphorus, sulfur and silicon. It is of course possible to remelt this type of iron and to
pour it into molds then and there the techniques and temperatures needed for molding cast
iron do not differ greatly from those needed in casting bronze. Cast iron as a metal is good for
making bells, is extremely hard but easily broken, an only fair conductor of heat (which makes it
heat evenly when used for cooking or industrial purposes), resistant to chemical damage, and
capable of being formed into very complex shapes with only minimal expenditure of labor.

However, if early Chinese iron workers wanted a softer and more malleable metal which
could be shaped in a blacksmiths forge, it was necessary to put this high-carbon cast iron
through a second processing step in a second furnace, a finery. This "fining" step involved
reheating the iron until it again became a liquid and then exposing it to oxygen for long enough
to burn most of the carbon out, yielding a solid mass of low-carbon wrought iron. Fining took
almost as much fuel per unit output as bloomery smelting, and the process was quite labor-
intensive.

Why the complex and seemingly wasteful two-stage method should have been
preferred in ancient East Asia cannot be discussed here. It is enough to say that the indirect
process was more efficient in some ways, that the large blast furnaces involved were capable of
turning out much more metal in a given period of time than the smaller bloomery furnaces used
in the direct process, and that at the present day almost all iron makers use two-stage processes
not greatly different in principle from the one invented by the early Chinese.

Whether made in a bloomery or finery, early low-carbon irons tend to be very similar.
Indeed, most specialists feel that they cannot be distinguished by any currently available
method of analysis except that finery iron, due to the very high temperatures at which it was
originally smelted, tends to contain rather more phosphorus, sulfur and other undesirable
impurities. The form in which the two kinds of iron were traded might sometimes offer a clue. In
some parts of India, slaggy blooms seem to have been the standard form in which iron was
traded (see for instance, Buchanan, 1807, 2: 437). But elsewhere in India and in much of the rest
of early Eurasia, low-carbon iron in historic times was generally traded in the form of rods or bars:
in fact, Europeans have traditionally used the term bar iron as a synonym for wrought iron.

The Chinese too are known to have traded low-carbon iron in the form of bars
(Tegengren, 1923-4:336). Generally speaking, this bar iron did not have a particularly high
reputation among foreigners in the eighteenth and nineteenth centuries, when compared with
either European metal or the bloomery-made iron of the rest of Asia. The Chinese metal often
had a rather high phosphorus content, which made the metal cold-short: liable to break
under stress at low temperatures.

This is why, in the eyes of economic historians and archaeometallurgists, the Java Sea
Wreck finds may seem surprising. Judging from later historical records like those cited above,
one would expect China to have exported quantities of cast iron during the Song and Yuan

98
periods. Chinese merchants should have found cast iron implements and cooking wares to be
ideal commodities for long-distance trade: relatively inexpensive to produce, impossible for
Southeast Asian consumers to copy due to the complex technology involved, and with special
properties (evenness of heat distribution, strength, chemical resistance and sheer size) that
made them irreplaceable by indigenous metal wares. But bar iron, one would think, was quite a
different matter. The Chinese product would have faced real competition from locally made
Southeast Asian bloom iron as well as from Indian iron and steel.

8.3 Historical and Archaeological Parallels

We do not know much about those local producers at the time that the Java Sea ship
went down, but it seems certain that they existed. One of the few contemporary iron-working
areas in Southeast Asia that have been studied in any detail is the Sarawak River delta around
Santubong in northwestern Borneo (Harrisson and O'Connor, 1969). Most datable artifacts
associated with the iron slag and other debris at Santubong were Chinese stoneware and
porcelain of the early Song to early Yuan dynasties. A few coins were also found from the Tang
(earliest: A.D. 618) and Song periods. Thus, the probable peak of activity at the sites occurred in
the twelfth to thirteenth centuries. The original excavator, Tom Harrisson of the Sarawak
Museum, believed that the iron ore worked at the site was brought by boat from upstream; the
site at the river mouth would have been chosen because it provided mangrove wood, which
yields excellent fuel for the furnaces, and was accessible to the sea so that the products could
be readily exported. The site also provided access to fresh water. Other isolated sites with slag
and "crucibles" were found between 6 hours and one day's sail in either direction along the
coast. How much iron was produced at these sites is still unclear due to difficulties in
distinguishing smelting slag from natural concretions, but there might have been enough iron to
support an export trade (Harrisson and O'Connor, 1969: 203-204).

Historical sources too indicate that Indonesians were actively engaged in producing and
shipping iron. Early Arab writers mention that a group, identified by some scholars as
Indonesians, traded for iron from places as distant as east Africa (Ferrand, G. 1922: 66). Zhao
Rugua, writing in the thirteenth century, mentions non-Chinese iron in several places (Hirth and
Rockhill,1911: 46, 61,69, 78). Iron swords, he says, were being made on islands somewhere near
Java, and other swords were imported from India to Java and Sanfoqi (or Srivijaya), probably in
Sumatra. He even says that another Southeast Asian locality, in northern Vietnam, was exporting
iron. He may be implying that this was wrought iron in unfinished form, although Wheatley thinks
that this metal must have been a re-export" (1959: 117).

Zhao also mentions that Chinese tripod cauldrons were imported by two places, one of
which is thought to be in the Philippines and the other in Java. As such items would have had to
be made of cast iron, this is what one would expect in the light of other evidence. But Zhao was
not greatly interested in iron and mentions it infrequently. The best early Chinese source for
international trade in iron is Wang Dayuan, who in the mid-fourteenth century compiled an
important (and partly firsthand) account of the countries of Southeast Asia and the regions
farther west (Zhu Fan Zhi, 1993).

The list below summarizes the places which, according to Wang, imported Chinese iron.
Although many cannot be confidently located on a modern map, there is no doubt that a high
proportion were in Southeast Asia.

99
Pieces of (scrap?) iron, presumably wrought rather than cast: Rili [Pattani in S. Thailand?
Somewhere in W. Malaysia or S. Vietnam?]; Sandao [somewhere in the Philippines]; Mayi [Mait/
Mindoro, C. Philippines?]

Iron, variety not specified: Sian [Siam]; Jiaozhi [N Vietnam]; Suluoli [?]; Duduya or Duduan [?];
Wenlaogu [Maluku Is.?]

Iron ware or objects, variety not specified: Ashili [?]; Bajielajian [Pajekan, near Surabaya, E.
Java?]; Bandali [?]; Boni [Brunei]; Boshili [?]; Dawudie [Orissa, E. India?]; Dongtanmiao [East
Demak, Java?]; Dongxizhu [island near Longyamen or perhaps Pulo Aur]; Lanwuli [Lambri,
Aceh, N. Sumatra]; Panhang [Pahang, Malaysia]; Tanyang [Tamiang, NE Sumatra]; Xiaojunan
[Quilon, SW India?]; Zhaowa [Java]

Iron wire: Xialaiwu [unidentified; somewhere near Malay Peninsula]; Pubin [?]

Iron tripod cauldrons: Baduma [Martaban, S Burma?]; Banzhu [Pancur, Singapore River Estuary];
Dajila [?]; Jianshan [?]; Jinta [in Thailand or Burma?]; Longyabodi [Langkawi Is., W. Malaysia?];
Longyamen [Keppel Harbor, Singapore]; Malilu [somewhere in northeast Malay Peninsula?];
Mayi [Mait/ Mindoro, C Philippines?]; Wuzhipa [?]; Zhenlu [?]

Iron woks: Sanfoqi [Jambi, in Sumatra]; Xiaopan [?]

Iron bars: Banzhu [?]; Dabadan [?]; Ganmaili [Cape Comorin, S. India?]; Huamian [Batak area,
N. Sumatra?]; Jialila [?]; Luowei [Ligor/ Nakhon Si Thammarat, S. Thailand?]; Lingshan [Cape
Varella, S. Vietnam]; Qianlima [?]; Sulu [Sulu Islands, S. Philippines]; Wuniangang [Pontianak,
Borneo];

Wang notes that many of these and other places also imported Chinese copper objects,
and that one or two (for instance, Boshili) were supplied with Japanese as well as Chinese iron.
His impressive list cannot be used to draw maps of the routes followed by metal traders without
much more work by historical geographers in identifying ancient place names. But the list does
show two things of importance to archaeologists working on cargoes of early Chinese ships sunk
in Southeast Asian waters: first, metals played an important role in the international trade of
fourteenth century China, and second, thatcontrary to expectationsChina did export
quantities of low-carbon bar iron as well as utensils made of high-carbon cast iron.

Archaeological data on Chinese-Southeast Asian trade routes and goods has until
recently been even scarcer than historical data, but now relevant information from terrestrial
and, more importantly, maritime excavations is beginning to appear. With regard to trade in
metals, the archaeological and historical data seem to support each other.

Recent excavations on the island of Bangka, just off the Musi River, found five piles of
twelfth-century qingbai bowls on top of five piles of iron woks facing upward, which in turn
rested on three piles of woks facing downward, nearly 2 m deep in a Vishnu temple of the sixth
to seventh century (ACRO Update: 1995). This enigmatic site raises many questions which only
further excavations can resolve. Some fragments of iron excavated at the Parliament House
Complex site in Singapore, of the fourteenth century, on analysis, proved to be cast iron and
therefore Chinese imports.

The Dinghai wreck of the Song period contained two concretions of iron, each 1.5 m
square and 1 m high (Kenderdine, 1995). The Pulau Buaya wrecksite, which contained 30,000
ceramics of the late Song, found near Bintan Island, Riau, yielded large quantities of iron woks

100
concreted in stacks, as well as bundles of iron parangs still tied in bunches with some type of
rattan-like fiber. The Vung Tau wreck, a Chinese lorcha of c. 1690 (Flecker, 1992), had the
midships cargo hold full of large stacked cauldrons and woks with handles. This vessel was on its
way from China to Batavia with a primary cargo of blue-and-white porcelain. Thus the
Southeast Asian demand for cast iron vessels from China continued for many centuries.

By way of interest, several Thai shipwrecks of the fifteenth and sixteenth centuries had
one or more cargo holds filled with iron. The Pattaya wreck (Green, 1983), the Royal Nanhai
wreck (Sjostrand, pers. comm.), and the Phu Quoc wreck (Blake and Flecker, 1994) all had iron
cargoes that interestingly had formed amorphous concretions entrapping several large storage
jars. Some of these jars were completely encased in iron concretion, suggesting that the iron
was stowed around the jars. Worked iron shapes could not be discerned within the concretions,
so it seems unlikely that the iron is a transshipped cargo originating from China. It has been
suggested that the original cargo was in the form of pellets or high grade ore. While long-
distance shipping of raw or beneficiated (pelletized) iron ore is not known to have existed
anywhere in the world before the late nineteenth century, the possibility must be borne in mind
as other shipwrecks are excavated in the future.

In metallurgical terms, the chief importance of the Java Sea Wreck is the presence of the
hundreds of tons of iron bars. This has not only confirmed the accuracy of Wang Dayuans
account but has posed an important new problem to economic and social historians of
Southeast Asia. It has long been known that cast-iron vessels were in demand throughout the
region. But in spite of what Wang recorded, few specialists had suspected that China was a
major supplier of as basic a raw material as wrought iron to Indonesia. Why was it so? How did
Chinese merchants, with a commodity that cannot have been better than locally produced
low-carbon iron, manage to compete with Southeast Asian iron makers? Could they have
made a profit from it, or did the trade in such high-technology goods as cast iron and
inexpensive porcelain produce high enough profits to subsidize the wrought iron trade?

One would also like to know whether there was a link between the ceramic and iron
trades. Tegengren (1923-4: 337-8) notes that in the early twentieth century, the Dehua-
Yongchun-Anxi area not only was still producing small quantities of iron by traditional methods
but that it also has significant reserves of anthracite as fuel; from personal observation, the
present writer can confirm that small-scale smelting in Anxi County (using local hematite ores but
an electric arc for heating) has continued to the present day. Significantly, kilns in Dehua, Anxi
and Yongchun (Ho, 1988) produced a large proportion of the celadon and qingbai wares
recovered from the Java Sea Wreck. It is also significant that later Chinese ships arriving at
Batavia from southern Fujian in the seventeenth century often carried cargoes that included
both ceramics and cast iron vessels.

8.4 Analysis of Iron from the Wreck

One iron bar and a fragment of cauldron were analyzed by Dr. Jonathon Leader, of the
University of South Carolina (see Appendix J), using electron microscopy.

The cauldron showed high concentrations of magnesium, aluminum, silicon, chlorine,

calcium, and sulfur. The high sulfur content is a clear indication of casting. The silicon may have
been a derivative of the casting process, but it could also have been from sand contamination
from the wreck site. Aluminum and magnesium most likely came from clay contamination. The
chlorine is from dissolved salts in sea water.

101
 The results for the iron bar are very interesting. There are high concentrations of silicon,
chlorine and sulfur. The presence of silicon and chlorine have been explained, but the presence
of sulfur in a wrought iron is almost definite proof that the bars were produced in China by the
fining of cast iron (see Section 8.2). Once sulfur is introduced into cast iron by the coal fuel it is
very difficult to remove. Contemporary Southeast Asian, Indian, and European wrought iron
manufacturers used the bloom method which did not introduce sulfur into the end product.

Sulfur is not desirable in wrought iron. It makes the metal brittle and likely to shatter on
the blacksmiths forge. The Chinese product was therefore inferior to more locally produced
wrought iron. Either there was insufficient local product, or the Chinese iron was substantially
cheaper. It would not have been the first choice of the Javanese blacksmiths.

102
9.0 HIGH-FIRED CERAMICS OF SONG DYNASTY CHINA:
THE JAVA SEA CARGO IN CONTEXT by Roxanna
Brown

9.1 Introduction
China held a world trade monopoly on high-fired ceramics for almost two thousand years.
Wherever Chinese traders went, and often far beyond, there are now traces of their ceramics.
China's pottery was so superior, for so long, that its very name came to stand for dishes made of
clay. Until Europeans began making successful imitations of "chinaware" in the eighteenth century,
the market seemed insatiable. Ceramics, along with silk, made China rich.

In order to stop the outflow of currency and precious metals to countries overseas, the
emperors decreed toward the end of the Song dynasty that ceramics be traded instead. Markets
already favored more highly fired wares, or this policy might not have been practical. For example,
the brightly colored earthenware temporarily popular during the Tang dynasty was extremely
decorative, but stoneware, with its greater hardness, could be transported over greater distances
without breakage. Thus high-fired clay, valued for its durability in daily use, acquired still more value.

Export routes and regions where Chinese ceramics were imported during the Song dynasty.

9.2 Chinas Supremacy

Whenever early peoples established settled communities, they somehow discovered that
clay subjected to fire will harden into permanent shapes; that vessels made this way can be used
for cooking and storage, and toys and figures can be modeled from it for play and decoration. Of
all the natural resources available to primitive cultures, clay is probably the most versatile (only
plastic, these days, might be considered comparable). Yet it was only in China that the possibilities
of the material were so fully explored in ancient times. Elsewhere, until modern times, ceramic
technology did not advance beyond the production of decorated earthenware.

The supremacy of Chinese ceramics began with Chinas vast geography of mountains and

103
rivers; clay could be found almost everywhere. There was also practically unlimited energy
available from coal (Marco Polo's story of black stones that burned, it might be recalled, met with
disbelief at home in Europe). Despite the abundance of coal, ceramic technology made its
greatest progress after the Song court's relocation to Hangzhou, in southern China, in 1127, when
abundant supplies of wood largely replaced coal.

In other cultures different materials were developed to answer practical needs filled by
stoneware in China. In Burma, for instance, lacquer-making evolved to answer both utilitarian and
aesthetic needs. In desert environments, animal skins were employed to carry food and liquids. In
the Islamic world metallurgy provided most daily utensils; many of these metal shapes, incidentally,
were copied into clay by Chinese potters. In fact, the history of a great many shapes known in
ceramics lies in the natural development of other materials. Chinese potters, at one time or
another, copied the shapes of their own archaic bronzes, of Middle Eastern pouring vessels, of the
animal-skin bags of the Mongols, and of artifacts of many cultures.

Of course, all these materials were used at one time or another in China as well. But, given
the initial success of a kiln design that could assure sustained high temperatures, plus clays that
could withstand those temperatures, it was the field of ceramics that was continually explored.
Examples of high-fired glazed ware appear as early as the Shang dynasty, in the period of about
1300-1028 B.C. That amazing beginning can perhaps be attributed to luck; nothing similar seems to
have happened anywhere else in the world. Still, for realizing the potentials of this technology, one
must thank generations of Chinese potters and entrepreneurs, who had, of course, a steady
clientele of massive numbers of domestic consumers from all the social classes, and at times a vast
overseas trade. At any rate, the particular combination of circumstances prevailing in China did
spur ceramic technology to heights of art and material science that could only be approximately
imitated anywhere else until comparatively recent times.

9.3 Sources and Properties of Clay

The words pottery and ceramics refer to any product made from clay. Clays vary widely.
Some are too sticky to model easily, some consist of large grains, some have fine and dense grains;
the colors are practically infinite. Whole ceramic industries today are based on one type of clay or
another. Studio potters spend as much time finding and mixing their own personal formulas as they
do actually constructing shapes. The specific qualities of any clay mixture impose limitations even
as they provide opportunity.

Clays can be divided into two general types: primary and secondary. Primary clays are
found in the close vicinity of their parent rock. Without going into their exact chemical structure,
they could be described as rock dissolved in water. They may lie beneath rocky surfaces or are
associated with the running water of streams. Primary clays are white or pale grey, and they have
a large grain size that makes them relatively non-plastic.

River valleys, which were usually the first places to support large populations, yield
secondary clays. Secondary clays have been washed away from the parent rock and have
usually gained a number of mineral impurities along the way. The most common of these is iron,
which, in concentration, can color the clay reddish. The smaller grain size of secondary clays
makes them easier to model than primary clays.

Being able to recognize various fired clay bodies is one of the most important clues in
identifying old trade wares. Ceramics specialists always check the base of any object they handle:
besides looking for a possible mark, or noting the way the foot is cut, they are seaching for an
unglazed part of the pot that shows the clay body.

104
9.4 Firing Temperatures

Clay subjected to heat will become hard. The greater the heat, the harder the pottery.
Earthenware fired at 900 C is harder than earthenware fired at 500. With too much heat,
however, the clay will slump into a viscous mass and eventually liquefy.

Because of the great variety of clays, it is difficult to be precise about the exact
temperature at which changes occur, but they are approximately at 500 and again at about
1200 C. Impurities that act as fluxes can lower the temperature required. The second change,
vitrification, transforms the clay into a glass-like structure. The chemical changes are translated into
three recognizable categories of fired pottery: earthenware, stoneware and porcelain. However,
Chinese terminology recognizes only two categories: dao, low-fired; and ci, high-fired. The
borderline temperature between them falls approximately at 1150-1250 C. Porcelain, a term
devised in the West, referred initially to shell-like characteristics of the end product rather than the
chemistry of its making. Porcelain is translucent and resonant. A strong light can be seen through
its walls and, if tapped, a porcelain vessel "rings."

Concerning the differences in the categories of ceramics, one could say that earthenware
consists of an impure clay fired at low temperature; stoneware consists of purer clay (fewer
impurities) fired at a higher temperature; and porcelain is a pure clay fired at very high
temperature. Earthenware and stoneware are each made with a single clay, even though it may
be a mixture of clays from different sources. Porcelain is made from two distinct materials: kaolin, a
practically pure primary clay, and then petuntse, a prepared form of decayed granite. It is the
structure of the petuntse that holds a pot together at high temperature when the remainder of the
mass is viscous. Stoneware clay fired at high temperatures may be enough like porcelain to be
called porcellaneous.

9.5 Glazes
Another way to categorize ceramics is according to their glazes. Essentially a type of glass
coating, glazes add physical strength to the form, decorate it, and make it impermeable. With
glaze, a pot does not break quite so easily, and glazed jars provide sanitary containers for liquids
and food products. This quality, indeed, is a source for the Middle Eastern myth that a person
would be warned about poisoned food if it were served in Chinese celadon-glazed dishes: the
dish, they believed, would change color. Only the specific link between eating and poison was
misidentified. The poisoning came, certainly, from the locally produced lead-glazed dishes that
were the alternative to Chinese wares; those who ate from them often enough would have
suffered from lead poisoning. At any rate, the solution worked: dining from high-fired Chinese ware
solved the problem.

Tang dynasty China is well known for its lead-glazed earthenware; the famous three-color
horses, camels, and human figures prized by collectors today, for instance.

The Song dynasty, is famed for monochrome stoneware the blue-glazed Ju wares of
northern China, the white-glazed wares of Fujian, and, of course, celadon.

Blue-and-white is associated with the Ming dynasty. Then there are the colorful, intricately
painted, enamel-decorated porcelains that typify the Qing dynasty. A simplistic but helpful
generalization of Chinese ceramic history is thus contained in the phrase "Song celadon, Ming
blue-and-white, Qing polychrome."

105
 Blue-and-white wares, in fact, seemed to have appeared during the Yuan dynasty, several
decades before the beginning of the Ming. With evidence currently available, the closest one
can date its appearance is within the quarter century between 1325 and 1350. This important
innovation was in part the result of intense competition for the wealthy Muslim Middle Eastern
market, a trade actively encouraged by the Mongol rulers who had conquered southern China by
1280.

9.6 High-Temperature Glazes

Together with high-fired clay bodies, Chinese potters searched for high-temperature glazes
to cover them. The lead glazes for earthenware were unsuitable because lead volatizes at about
1150. (This limitation could be overcome, as it was later, by first firing the stoneware or porcelain,
then adding the colors in a second low-temperature firing.)

High-fired Chinese glazes can be called alkaline (meaning in effect, lead-free), or

felspathic (composed largely of feldspars, a natural component of clay), and ash glazes.
Essentially, the technique employed a mixture of clay, ashes, and water. A basic celadon glaze,
for example, might be composed of wood ash kept from a previous firing of the kiln mixed with the
same clay as used for the body, and water. A small amount of iron present in the clay, not more
than 1 to 4%, is responsible for the green color. Sometimes one sees brown celadon, showing a
slightly greater percentage of iron. The type of wood from which the ash is taken may cause the
shade of green to be bluish or green. Potters in northern Thailand, for instance, can produce five
distinct shades of celadon by varying only the type of wood ash. White glazes are made from an
iron-free mixture.

Besides the presence of minerals and the effect of ash, another influence on the color of
glazes is the kiln atmosphere. Two kiln atmospheres are possible, though in practice both will occur
at different points within a single firing. In the thirteenth century, such a firing required about three
days. An oxidizing atmosphere consists of a free flow of air; the heated glaze then acquires oxygen
from the air. Ferrous oxide, for instance, will become ferric oxide. With reduction, the kiln is
suddenly closed, so that the fire burns away free oxygen, then pulls oxygen molecules from the
molten glaze. Induced at varying points in the firing cycle, at certain temperatures and for short or
prolonged periods, reduction can cause a variety of effects. Proper control of the cycle, attained
only with repeated experimentation, could consistently produce blue celadon, the legendary Ju
ware of northern China, for instance. For the famous celadon wares of the Song dynasty, the basic
principles for control of reduction began with the Yueh wares of northern Chekiang province. The
best, most famous of these date to the ninth and tenth centuries. Along with white-glazed Hsing
wares, examples of these have been discovered as far afield as Egypt.

9.7 Glaze Color

Only three minerals can influence color in high-fired ceramics: iron, copper and cobalt. Iron
is the most common. In the oxidized atmosphere of a kiln with free air flow, iron will fire, depending
on its concentration, to a variety of brownish shades to black. If fired in a reduction atmosphere, a
small amount of iron, 3% - 4% of the glaze mixture, will produce shades of green, i.e., celadon.
Without reduction the possible glaze colors on high-fired wares would be hopelessly limited.

Copper, in an oxidized atmosphere, will also produce green; and in reduction, red. But the
firing conditions in either case must be finely honed copper red is notoriously difficult to achieve.
This is one of reasons why red-glazed wares are so very highly prized. In the fourteenth century, for

106
instance, both underglaze blue and underglaze red were attempted, but underglaze-red
decorated wares were quickly discontinued because of the difficulties of production. More often
than not, the glaze fired to a brown or gray color. Cobalt, oxidized or reduced, will produce blue.
It is no accident that a blue-and-white color scheme is associated with porcelain. Of the three
mineral colorants known, cobalt is the least difficult to fire, and the most colorful.

9.8 The Java Sea Ceramics Cargo

The ceramics recovered from the Java Sea Wreck were manufactured in China, except for
a range of unglazed earthenware in unusual ritual shapes. Rare Cizhou type ware was found in
small quantities. The remainder of the ceramics are monochrome wares with green, brown, black,
white or qingbai (bluish white) glaze. The very finest pieces, which are few in number, are qingbai
wares from the Jingdezhen kilns of inland Jiangxi province. All other wares appear to have been
made in Fujian province.

The finely potted earthenware bears a general resemblance to wares associated with the
southern Thai region of Patani. X-ray florescence component analysis shows distinct similarities to
pottery from Kok Moh, in Patani. The discovery of hundreds of these vessels on the wreck clearly
demonstrates the importance of earthenware as in intra-Southeast Asia trade commodity, a fact
absent from any documentary records. Several shapes are completely unprecedented, in
particular toroid kendis, which were previously only known from Ming dynasty China.

4
JAVA SEA KENDI
THAI EARTHENWARE
3.5 JAVANESE EARTHENWARE
SUMATRAN EARTHENWARE

3
PRINCIPAL COMPONENT 1

2.5

1.5

0.5

0
0 1 2 3 4 5 6 7
PRINCIPAL COMPONENT 2

X-ray florescence analysis of earthenware kendis.

107
 Upright shapes decorated with painted designs in underglaze black, and originally
covered in a green lead glaze, constitute an unusual find. In general their clay and decorative
motifs are similar to those associated to the Cizhou kiln complex. They are usually decorated with
floral designs in slight relief, with the background painted solid underglaze black. Designs painted
in black on a flat surface are less common.

There is an interesting group of white to qingbai glazed covered boxes. Many have the
cover sealed to the base by drips of glaze, proving that covered boxes were fired, packed and
shipped with the covers in place. These boxes were sometimes used to hold mirrors in China, and
in Southeast Asia they are known to have been employed as containers for cosmetics and jewelry.
In small and miniature sizes they have served as amulets. The boxes, and similarly glazed small
vases, were manufactured in molds.

108

Kiln sites and major markets in Fujian Province.

 A small group of black glazed tea bowls appear to be southern Temmoku, rather than
the famous Chien tea bowls.

The majority of the wares are green-glazed bowls of varying size and quality. The glaze is
sometimes opaque and sometimes translucent, and the shades of green differ even on pieces that
would otherwise be identical. Such wide variation in color and texture can be caused by firing
conditions that vary within the kiln. Consistent, standardized glaze effects were practically
impossible to attain in ancient wood-fired kilns. The better quality wares are higher fired and are
decorated with cursive cloud and floral designs incised into the clay prior to glazing.

Brown-glazed jars in the cargo are distinct in their structure and purpose. Rather than
being items of trade, they were used to store food and water for the ships crew, and for packing
organic products for trade. To preserve the contents, the mouths of the jars were often sealed with
a disk of wax or wood, secured with tie strings looped through small handles on the jar shoulder.
Stamped Chinese characters adorn the shoulders of the majority of the larger jars.

The late Song dynasty was a prosperous time for the ceramics industry. A great many kilns
were in operation in Fujian province. They served the great port of Quangzhou, the most probable
port of departure of the Java Sea ship.

109
10.0 THIRTEENTH CENTURY POTTING TECHNIQUES:
AS EVIDENCED BY THE JAVA SEA CARGO by Roxanna
Brown

By the late Song dynasty, a ceramics factory was highly organized and capable of mass
production. There could be a dozen or more divisions of labor, depending on the complexity of
the end product.

10.1 Collecting and Preparing the Materials

First came the collection and preparation of materials. As a rule, haulage was minimal
because a factory would be located near a good source of clay. For ready access to unlimited
quantities of water, factories were also normally located near a river or large stream.

Once the clay was collected, workers had to clean and refine it. This process would
consume vast amounts of water. Thoroughly submerged, in a large pot or specially constructed
pit, the clay was then stirred so that larger impurities could be skimmed off the surface. It was
filtered in a variety of ways. Sometimes heaps of clay were left outside for many months to be
naturally weathered before any other preparation processes were begun. Coarser clay was
used for storage vessels, but finer-grained clay was necessary for any wheel potting.

After separating the different grades of clay, workers formed the clay into balls of
workable size for the potters. The balls of clay had to be thoroughly kneaded to extract any
bubbles of air, which could cause a vessel to break during firing.

10.2 Preparing the Glazes

Meanwhile, glaze mixtures were being prepared. Every kiln master undoubtedly had his
own trade secrets, but glazes in the late Song period were generally made from a mixture of
finely ground clay mixed with ash. The particular type of wood from which the ash was
obtained could affect the shade of the fired glaze. Sometimes a type of limestone was burned
together with the wood.

The most important coloring agent for pottery throughout the Song dynasty was iron, a
natural impurity in most clays. A large iron content in a glaze would produce brown and black;
a minute quantity in a reduction firing could produce various shades of green.

If a lead glaze were used, as seems to be the case in the painted wares of the Java Sea
cargo, then a lead-bearing clay was required. For painting an underglaze decoration in black,
a clay with very high iron content was applied with a brush after being mixed with water to a
brushable consistency.

Once formed, dried, and decorated, all the Chinese pottery of the Java Sea cargo was
glazed. The glazing appears to have been accomplished by dipping the pieces into the glaze,
although there are signs that it may have been applied by brush as well. A heavier glaze could
be achieved by dipping a vessel into the mixture more than once. The celadon-like olive-green
glazed pieces were probably dipped at least twice.

108
10.3 Furnishing the Kiln

A third group of workers manufactured such kiln furniture as spacer disks, stands, and
saggars, the tall round boxes, sometimes with covers, into which finer pieces are set for firing.

Fine wares had to be kept spotlessly clean prior to firing. A single speck of dirt or a bit of
grit, perhaps falling from the kiln walls, could mar them. They could be damaged by direct
flame or blasts of hot air. Saggars, designed to protect them, were made from a rough, durable
clay and themselves fired before use.

The saggars were formed on a potters wheel, left to dry for a time, then shaped with a
knife. The saggar makers sometimes used the same rough clay to make common earthenware
vessels for their personal daily use at home.

Cruder wares could be fired without being packed within saggars, and bits of grit caught
on the surface of the glaze during firing is evident on them. For the coarser varieties of bowl, an
unglazed ring round their interior bottom allowed them to be stacked directly one upon the
other, the footring of the bowl above set into the unglazed ring of the bowl below.

Sometimes kiln stacking supports for separating wares in the kiln are also required.
Apparently these were not consistently used for the wares of this shipwreck; only a few pieces
show traces of circular scars on their bases from the tubular supports, or pontils, onto which the
pots were set.

10.4 Making the Molds

Another group of more skilled workers prepared molds for such shapes as the covered
boxes, ewers and vases. The molds were made from a white plaster into which liquid clay could
be poured or damp clay could be pressed. On the box covers from the Java Sea Wreck, one
can sometimes see finger impressions on the inside where the clay has been pressed into the
mold. The boxes were formed in two pieces, the upper and lower halves. Upright shapes were
more complicated. Ewers, for example, with their handles, spouts and covers, required the
greatest skill; the many different parts had to be formed in separate pieces, then luted together.
In luting, the dry clay edges are slightly roughened or scored with a pointed tool, soaked with
water, then joined. Join marks are usually visible on the interiors of such pieces. Among wares
from the Java Sea Wreck, joins can be seen on the exterior surfaces as well. This is probably a
consequence of the pressures of mass production.

Molds had to be made continuously, as they would become worn with constant use. The
mold makers also had to be very familiar with the shrinkage properties of the clays, for the
finished pieces would always be smaller than the mold after firing. This was particularly
important for ensuring that box covers fitted on bases. At any one factory only three or four
workmen might have the skill to produce similar molds consistently. Another set of workers would
join the finished sections of pottery.

10.5 The Potter's Wheel

Certain shapes, such as those that are ribbed, lobed or angular, are more easily and
economically produced with molds. The round bowl shapes, on the other hand, were more

109
easily and quickly made on a fast-turning potter's wheel. For this function yet another set of
skilled workers were employed. Some were trained in turning smaller pieces, while others made
larger shapes. The wheel, essentially a round wooden table set on a perpendicular axle, is
made to turn continuously. The potter raises the shape from a prepared ball of clay, using his
fingers and hands. A skilled potter can make the basic form of a dish shape in a few minutes.

Vases, bottles, and ewers of some shapes could also begin on the potter's wheel. After
drying, the vessel would be returned to the wheel for final touches.

10.6 Finishing the Form

The finishing work, the final cleaning and shaping of the dried clay, was carried out by
another set of workers with their own special tools. Knives, for instance, were used to pare away
superfluous clay. This paring of dried pieces accounts for the chatter marks seen around the
lower walls of some of the Java Sea cargo's undecorated green-glazed bowl shapes.
Craftsmen also used knives to cut out a footring from the solid foot of clay left by the wheel
potter. The footring could not be cut until the dish or vessel was dry enough to be set upside
down. With this final cutting and polishing the potters' work was done.

10.7 Decoration
In the Song dynasty factories, another group of specially skilled workers would apply
incised or carved decoration to the surfaces of round shapes begun by wheel. Various hand
tools, such as knives and bamboo combs, were used for this work. A floral decoration, for
example, will first have a heavy outline carved, and then the petals and leaves will be
enhanced with strokes of a comb.

Compared to later designs in underglaze blue, the underglaze-black designs of the Song
dynasty, as evidenced by the Java Sea cargo, were not very complicated. It is possible that,
although a few decorators may have been specially trained in brushwork, those who became
skilled in incised and carved design may have also handled the painted motifs.

10.8 Firing
Stacking the kiln could take a day or more, depending on its size. The finer wares would
first be set in saggers, then carried in stacks to the kiln. Inside the firing chamber, the pots were
arranged to take advantage of every square inch of space without inhibiting the flow of heat
from the firebox to the chimney. Old Chinese kilns were long and narrow, and built up an incline
toward the chimney. The heat would be most intense at the upper end, so wares requiring more
or less heat were positioned accordingly.

A sufficient fuel supply had to be maintained for the kiln firings, which usually required
three days of continuously sustained heat. In northern China, coal was normally used, but in the
lands south of the Yangtse River, wood-burning kilns were more common.

The most exciting moment in the long involved pottery process was the opening of the
kiln. This had to be timed perfectly. If it was opened too soon after firing ended, the sudden
coolness would cause uneven contraction between the clay body of a pot and the glaze. The
glaze would consequently craze or crackle, an effect that was sometimes intentional. Many
other kiln accidents were possible, ruining weeks of work. When beautiful pots emerged, all

110
those that contributed to their creation must have felt intense pleasure and relief.

Any celebrations at the completion of a successful firing, however, provided only a short
break. After the ceramics were taken out of the kiln, they had to be packed for transport. In the
mean time, all the departments of the factory would have been simultaneously preparing for
the next firing, and the one after that.

10.9 Illustrating the Potting Process

1. CLAY

As the primary material involved in pottery-making, clay

undergoes a series of purification and preparation
processes before being introduced to the potter.
Although their opinions on the suitability of any one
clay mixture are taken into account, the wheel potters
first see the material in prepared lumps of appropriate
size for immediately throwing onto the wheel. Here,
one of the lumps is being kneaded a final time to make
certain that its texture is uniform and that it contains no
tiny air pockets. Any bits of grit in the clay would scar
the surface of a pot as it is being thrown, and air
pockets might explode during firing.

2. WHEEL-THROWING

Potters able to throw shapes on a potter's wheel are considered skilled

labor. They are creators whose training may take years. Here a potter
draws a lump of clay upward into a jar shape. The potter often holds a
wet rag in one hand to keep the clay pliable. By applying pressure with
his fingers and hands, the potter controls the contours of the vessel.

3. MOLDING

Elsewhere in the factory liquefied clay is

poured into molds made from white plaster. Depending
on the skill of the mold designers, the variety of shapes
made by this method is practically unlimited. Separate
molds may provide multiple parts for a single final form. Besides providing a way to utilize clays that lack
enough plasticity to be easily formed on a potter's wheel, the method is economical. With molds, an order
for a hundred jars all in the same shape and size could be quickly and efficiently filled with a mostly
unskilled labor force.

111
4. FINISHING

Once the basic shape is formed, whether by wheel or by

mold, it is allowed to dry and thus slightly harden for a day
or more. Mouthrims are shaved smooth, extraneous bits of
clay are removed, and flat bases are cut or footrings
carved. For objects made from piece molds, the join lines
are smoothed away.

5. INCISED DECORATION

One of the oldest forms of pottery decoration

is to cut designs onto the surface of the clay.
The designs, incised with a variety of tools,
are mostly floral or vegetal in nature.

6. UNDERGLAZE DECORATION

Once the clay is completely dry and hard, underglaze

decoration can be added with a brush. Iron black and cobalt
blue are the most common colors. Straight lines around the
body of a pot can be applied easily if the pot itself moves on a
turning table.

112
 7. GLAZING

After the vessel has been decorated, it is dipped into a vat of

glaze, a viscous liquid. The most common mixture is made
with clay and wood ash. After dipping, the glaze is allowed
to dry.

8. FIRING

Finally the clay pots are baked in kilns. Kilns have been
made in all shapes and sizes and the technology of
applying the appropriate amount of heat, the period of
firing and controlling the kiln atmosphere is complicated.
The exact combination of firing conditions is also usually
considered part of the special trade secrets associated with
fine ceramics. In the past, the fuel for kilns was wood or
coal. Today it is gas or electricity.

113
11.0 CERAMICS INVENTORY by Roxanna
Brown

11.1 Painted Ware

No kiln site has been documented as a
source for the once colorful ceramics
described in this section. They might come
from the Cizhou kilns, or from the Quanzhou (or
Cizao) kilns, where lead-glazed kendi were
produced some of them with the same
green glaze that originally covered the
painted wares (see Hughes-Stanton and Kerr,
1980, 32; and Adhyatman, 1987, Nos. 58-61).
The variety of these pieces in Southeast Asia
may indicate that they were made only for a
short period. From collections of trade
ceramics, only one example in similar style has
been published (Southeast Asian Ceramic
Society, 1983, no. 180), and it was tentatively
identified as being from Northern China. Some
equivalent examples were recovered from the
Breaker Shoal shipwreck discovered in the
Philippines (Dupoizat, 1995, Figs. 40-41). The
shapes from the Java Sea Wreck range from
bowls and cups to ewers and covered boxes.
The decoration is sometimes in simple
underglaze black, but more often the design is
either incised, or molded in relief, with the
Painted ewer on the seabed.
ground painted black. The whole piece was
then covered in a bright shade of green lead glaze, a low-temperature glaze that would have
been applied after the first stoneware firing of the pottery. A second low-temperature firing
would mature the glaze.

The painted ware designs are basically floral, usually featuring a peony spray, sometimes
along with leaf scrolls. There are also examples of the phoenix frozen in flight on the covers of
boxes, but standing when they appear on upright shapes. There are two box covers with a
sinuous dragon drawn in underglaze black.

The clay body of the ware is hard and grayish-white, although it can be buff-gray on
exposed surfaces where the original glaze has eroded. The potting of the upright shapes is
sometimes so thin as to seem fragile; many of them were constructed from parts formed in
molds. The dish shapes and covered boxes have a more substantial feel about them.

BOWL SHAPES

The dish shapes range from a small conical bowl to a large flared dish. The designs vary in
detail but include foliage and peony scrolls incised in outline and reserved against an
underglaze black background. Usually the exterior is plain but it may also be painted with a
design in underglaze black of scrolling foliage. The entire dish, including the base (presumably
with the exception of the footring on which it would have set for firing), was then covered in
green glaze. The present exposed surface of the clay, where the glaze has eroded away, is buff-
gray.

116
 Small Bowl

Conical, with a short carved footring. The interior

has a scrolling floral spray incised below an
undecorated band at the mouthrim. The exterior
has a painted design of scrolling foliage.

Example: 1
Diameter: 14 cm

Large Bowls

Several somewhat similar fragments are from larger size bowls with a straight mouthrim but
curved walls. The largest of these shows traces of a circular stacking scar on the interior bottom,
perhaps from the footring of a dish set upon it for firing.

Examples: 4 damaged
Estimated diameter of largest bowl: 25 cm

117
Large Dishes

Dishes with a narrow everted

mouthrim, decorated on the interior
with scrolling foliage, the exterior
plain.

Examples: 3 damaged
Largest diameter: 30 cm

COVERS

Cup-shaped, heavily-potted, and decorated with

incised scrolling foliage around the exterior
reserved on an underglaze black ground then
originally covered in green glaze.

Examples: 3
Diameter: 6.8 - 7.5 cm

118
 Concave center area set with a tiny ring handle at one side
and wide, flat mouth flange, covered on the upper side with
bright green glaze.

Examples: 11
Diameter: 8 cm

Dome-shaped with a wide mouth flange and two tiny

perforations at one side. They are molded with the design
of a plum blossom reserved on an underglaze black
medallion at the center. There are traces of bright green
glaze on the flange.

Examples: 14
Diameter: 7.2 - 7.5 cm

Concave depression around a short knob handle, and wide

flat mouth flange with two small perforations at one side.

Smaller Examples: 17
Diameter: 7 cm
Larger Examples: 2 with a retaining ring to one side in
place of the perforations
Diameter: 8 cm

Cover with a ring of overlapping lotus petals in slight relief around

a knob handle, the underside with a short vertical flange; with
traces of green glaze on both the upper and lower surfaces.

Examples: 3
Diameter: 5.2 cm

119
EWERS

The ewers are mostly pear-shaped with a bulbous lower body curving upward into a graceful
neck flared at the mouth. They are unusual in being mold-made, in parts, rather than wheel-
thrown. A strap handle is applied to one side and a tall curved spout opposite. They are
decorated either with painted floral designs in underglaze black, scrolling foliage incised in
outline against an underglaze black design, or with the white designs in relief (the clay having
been pressed against a mold to obtain the motifs) that include the phoenix amidst scrolling
foliage, all on an underglaze black ground, then covered in bright green glaze. The various
styles show slightly differing proportions. The foot of these vessels is particularly unusual. Rather
than being carved out with a knife after preliminary shaping, it was made by the clay being
pressed into a mold that also formed the lower walls. Thus it imitates wares with a carved foot,
but the footring is uncharacteristically rounded and the base concave rather than being flat.
This potting technique is seen clearly on several ewer fragments intricately modeled with eight
vertical panels of floral and phoenix motifs. One ewer also has Chinese characters in relief on
the base.

Examples: 27 in damaged condition plus fragments of more

Height: 23.5- 24.5 cm

120
121
122
123
WINE POTS

These have lobed walls, an inset base, a strap handle applied to one shoulder and a
short curved spout at the other. A circular depression around the mouth suggests the
shape was meant to be combined with a cover.

Small example: 1
Height: 10 cm

Larger examples: 2 damaged

Height: 15 cm

124
COVERED BOXES

Examples are in various sizes, with rounded

walls or with flattened panels on the sides.
The domed covers are sometimes slightly
flattened. Most covers are decorated with
incised or molded decoration (usually floral)
on an underglaze black ground, then
originally covered in bright green glaze.
Very occasionally the cover features an
underglaze painted design. Two show a
sinuous dragon, another (a complete,
sealed box 7.5 cms in diameter) features a
peony spray; the lower portions of the boxes
are undecorated except for the green
glaze.

Diameter: 9 - 13.5 cm

125
BOTTLES

The main body is roughly spherical, with a tall flared neck

and carved footring with tiers on its exterior side;
decorated with scrolling foliage on an underglaze black
ground. The clay is unusually thin, and the glaze is eroded.

Example: 1
Height to break at neck: 19 cm

The main body is roughly spherical, with a tall neck flared toward the mouth, and a carved
splayed footring with tiers on its exterior side. It is decorated with peony scrolls reserved on an
underglaze black ground, and with a sculptural representation of a long sinuous dragon
wrapped round the lower neck. The modeled dragon attached to the lower neck of this vessel
is the same as that seen on similar bottles with qingbai glaze described later.

Example: 1
Height to break at upper neck: 24 cm

126
Pear-shaped bottle, with a carved splayed
footring and sharply everted mouth flattened on
its upper side, decorated in underglaze black with
sprays of foliage. Its original low-temperature,
bright-green lead based glaze is completely
gone.

Example: 1
Height: 21.5 cm

JAR

Barrel-shaped jar, with in inset base and

depression round the wide mouth meant to be
fit with a cover; decorated in relief with incised
peony scrolls on a ground of underglaze black
round the upper body and with a band of tall
lotus petals round the lower walls; with some
traces of the original bright green glaze.

Example: 1
Height: 12.5 cm

127
 VASES

Vase in meiping shape, sturdily potted, with an

everted flattened mouthrim; the carved foot
with an inset base; decorated round the upper
body with incised peony sprays, then covered in
bright green glaze; the clay body dark gray.
(Whether underglaze black was originally
applied is unclear in the vessels present
condition; only traces of the green glaze
remain.)

Examples: 1 complete and a shoulder

fragment of a second.
Height: 28.4 cm

Vase, with a more whitish, more compact clay

body than apparent on other painted ware
shapes; the main body almost spherical, the
foot flared, and with a tall graceful neck that
opens outward into a lotus leaf shaped
mouthrim; decorated round the body with an
incised peony scroll set on an underglaze black
ground.

Example: 1
Height: 17 cm

Vase, upper portion only, with rounded shoulders

128
and fairly straight sides, the neck straight, slanted inward and unglazed, as though it were meant
to have a cover probably one of the cup-shaped covers also recovered; decorated on the
body with incised peony sprays reserved on an underglaze black ground; the clay body dark
gray.

Examples: 1
Present height: 14 cm

AMBROSIA BOTTLES

Potted solidly, with a bulbous lower body, a tall neck and bulbous mouth. The shape formed to
be hand-held at the neck. With traces of cursively painted underglaze black decoration, but
with no traces of glaze presently apparent; the flat base sometimes with three spur marks or, on
one, traces of a circular kiln support scar; the present surface of the clay orange-gray in color.

This shape is rare among existing pottery in Southeast Asia, but it is well-known from Indian-style
sculpture where such bottles are often held in the hand of gods and said to hold the elixir of
immortality. It was perhaps a container for especially treasured medicines. A few similar
examples are recorded from the Breaker Shoal shipwreck in the Philippines (Dupoizat, 1995, Fig.
38), and a single white-glazed one was found in a Yuan-dynasty (1280-1369) context in Sarawak
(Chin, 1988, Plate 37).

Examples: 2 intact, 3 damaged

Height: 10 cm

129
11.2 Black-Glazed Ware
There are only bowls in this category,
and they are a standard size and shape
valued as a tea-drinking bowl. The glaze is
temmoku black with brownish patches, thin
at the mouthrim but flowing thickly onto the
lower exterior walls. The mouth is potted with a
slight curve to form a lip from which a liquid
could be sipped. The footring is roughly
carved leaving a shallow but thick base.
Some bowls were made with a rough grainy
gray clay; others display a similarly grainy
brown clay body .

A similar bowl in gray clay, illustrated in

an exhibition by the Southeast Asian Ceramic
Society [1983, No. 159] is identified as made at
Jizhou, Jiangxi. It is said to have originally had
clear glaze applied to the mouth.

Examples: 10 mostly intact,

36 damaged
Diameter : 12 - 12.5 cm
Height: 5.5 - 6 cm

130
 11.3 Molded White/Qingbai Ware
This group of ware made excellent use of a native white Fujian clay that was apparently
relatively non-plastic; the problem of its not being easily shaped on a potters wheel was solved
by thinning the clay to a viscous consistency and pouring or pressing it into molds that formed
both the shapes and decoration in relief. Small vases and covered boxes, particularly, could be
made easily in quantity. The vases were probably used at altars to contain flowers or perhaps
sticks of incense, and the boxes could be used for storing cosmetics, perfumes and medicines
or for the many miscellaneous uses for which boxes might be used even today. In Indonesian
burials, boxes sometimes contain jewelry. In later years sculptural figures would be added to the
Fujian white ware repertoire. The glaze color varies from white to gray to bluish-green. Probably
the varying shades were chance results from individual placement in kilns and naturally varying
firing conditions. Among this ware, the specific color was mostly fortuitous. At other kilns these
effects could be planned and attained with proper glaze proportions and strict control of firing
conditions. All the white-glazed ware of the cargo represent ceramics made in Fujian province,
at both the Dehua and Anxi kilns (Hughes-Stanton and Kerr, 1980).

VASE

Made with four mold-produced sections that were subsequently luted together: the neck with
flaring mouthrim, upper and lower halves of the main body, and a flared foot domed inward on
its underside. The clay is whitish and somewhat grainy, with a sugary feel to its unglazed surface.
Decoration is in relief, with ribbing on the main body and lotus petals round the footring. The
interior is unglazed except at the flaring mouth.

Examples: 5 intact, 65 damaged

Height: average 11 cm; additional broken miniature, 6 cm in height

131
BOTTLE

Upper body fragment, with a spherical

main body and curved neck,
decorated on the shoulder in relief with
two rings from which a border of petal
design hangs.

Example: 1
Present height: 3.5 cm

COVERED BOXES

Mold-made in upper and lower halves, some undecorated except for relief striations down the
walls, from a ring round the outer cover, that divide the sides into curved panels. Others are
decorated with a floral spray on the upper cover and/or with ribbed walls. The flat base most
often is concave; and the interiors are splashed with glaze that also covers the exterior to the
lower walls. All the boxes are circular (rather than square), though some have flattened panels
around the outer walls.

132
Two hundred and sixty-eight boxes were recovered sealed. Many, maybe even most, of them
were shipped in this state. They are sealed by glaze having run from the cover onto the lower
portion. Why they were shipped in this way is a mystery. Perhaps it was a sort of guarantee seal,
one that could be broken by a judicious hammer blow, and maybe local brokers provided this
service, in the way that certain persons become skilled at cracking open coconuts. Mud and
water did gradually seep through open cracks between the two halves during their time
underwater, so that these sealed examples are presently heavy; otherwise they were shipped
empty.

Examples: 268 boxes were recovered sealed; 6 very small, 80 small, 60 medium,
and 120 large, with approximately 600 separate covers and 900 bases
in various sizes

Very Small Covered Boxes

These are decorated with fine, lush sprays of a variety of flowers at the flattened center of the
cover and with bands of ribbing at the shoulder and round the lower section. The unglazed
base is flat rather than concave as on larger examples.

Diameter: 5 cm

133
Small Covered Boxes

With a peony or lotus spray in relief on the cover center and separate bands of vertical ribs
round the cover shoulder and round the lower box section. The base is slightly convex.

Diameter: average 6 cm

Medium Covered Boxes

With walls divided into eight sides, and decorated with a peony spray in relief on the center of
the slightly domed cover encircled with a band of striations, sometimes together with a ring of
dots.

Diameter: average 8.5 cm

134
Fruit-Shaped Covered Boxes

In medium size, with eight lobes and the

design of a fruit stem at the depressed center
of the cover; the base slightly inset.

Diameter: average 8 cm

Melon-Shaped Covered Boxes

In medium size, with ribbed walls extending to

a floral design at the depressed center cover;
the base flat.

Diameter: average 8.5 cm

Large Covered Boxes

The outer walls either 8-sided or with

bands of ribbing over the shoulder of
the cover and round the lower section,
most with a floral spray (often a peony)
in relief on the slightly domed center
cover.

Height (with ribbing): 6.5 cm

Height (octagonal) : 4.5 cm
Diameter: average 11 cm

135
Unusual Examples of Covered Boxes

One small undecorated, perhaps

wheel-thrown box, with a hand-
modeled and detailed sculptural
representation of an entwined,
copulating couple attached to
the inside bottom.

Diameter: 5.6 cm

Various covers with unique designs, including one with a Buddhist swastika, another with fish,
and one with a geometric pattern, all from large boxes.

Diameter: 11.5 cm

136
Chinese Characters and Marks

A number of examples from the more standardized box

shapes are made unique by having designs in relief on
their base. These are sometimes simple floral motifs, and
sometimes Chinese characters that seem to refer to family
names, or which may be kiln factory names, something like
a trademark (see following pages). In China boxes have
been documented with marks such as mirror box or
sometimes they have names of medicines written on them
in black ink.

11.4 Qingbai Ware

A rather small but definite group of ware could be called qingbai. Their bluish-green
glaze was an intentional effect, deliberately and consistently achieved. A small number,
recognizable by the sureness of their potting, their whitish compact clay and a thick evenly
applied glaze must come from the rightfully famous Jingdezhen kilns of inland Jianxi province.
Others, with a brownish-gray grainy clay probably come from the same areas of Fujian as the
green-glazed ware. A range of ewers displays the same grayish-white clay as some of the finer
green-glazed ware. Structurally complicated shapes, they have a short foot with a smooth
concave base that appears to have been hand-formed rather than carved with a tool. The
main body was wheel-thrown, then neck-and-mouth, handle, and spout sections were
attached separately. All the ewers were originally meant to have covers that would have had a
matching retaining ring handle such as seen on the strap handles. A string would have been
tied through them so as to secure the cover to the vessel.

DISHES

Dish, with a foliated rim, very short foot and flat

unglazed base. Decorated with an incised floral
design on the interior and covered in mostly
uncrazed bluish-green qingpai glaze. From the
Jingdezhen kilns.

Examples: 1 damaged
Diameter: 15 cm

137
Zi Indecipherable molded Indecipherable molded on the Li a surname molded on
on the base of a qingbai box base of a Cizhou-type ewer the base of a qingbai box

Chen Zhe Ying seal of the Chen

family molded on the base of a
qingbai box

Liu Shi 60 molded on

the base of a Cizhou-type Zheng a surname
decagonal box molded on the base of a
round Cizhou-type box

138
 Yi Lang - a surname molded
Zheng a surname on the base of a qingbai box Chen a surname
molded on the base of a molded on the base of
qingbai box a qingbai box

Da big, Tong brass/copper, only

two characters decipherable molded Wu a surname
Indecipherable molded on on the base of a qingbai box molded on the base of
the base of a qingbai box a qingbai box

139
Dish, or perhaps a cover, Jingdezhen qingbai ware, with conical walls and glazed all over
except at the straight mouthrim. It is plain with no incised or molded decoration.

Example: 1
Diameter: 9 cm

COVERED BOX

Oval-shaped and undecorated, with a neatly carved

shallow inset base. Covered to the lower walls in an
opaque very pale bluish glaze. Probably Jingdezhen
ware.

Examples: 2, one of them sealed

Diameter: 5.4 cm

COVERED BOX COVER

Finely potted with compact white clay,

decorated with a spray of two peonies in
high relief within a ring encircled by a
band of classic scroll on its upper surface.
Probably from Jingdezhen.

Example: 1
Diameter: 11 cm

140
VASES

Small and compact, with a bulbous body, flared neck and an everted foliated mouthrim and
slightly splayed carved footring. Covered to the interior of the neck with thick opaque glaze
presently eroded to a rough finish but probably originally extremely fine. The clay body is fine,
smooth and gray on the interior but displays a rough texture on the unglazed base. Possibly from
Jingdezhen.

Example: 1
Height: 9.5 cm

With a bulbous lower body, flared foot and tall curved

neck. Made in four horizontal sections subsequently
luted together the foot plus three body sections. The
whole is divided into eight lobes, each containing a
floral spray in relief. Covered in an unusually bluish
opaque glaze. The clay body pale gray.

Example: 1
Height: 20.5 cm
(broken at upper neck)

141
 A mouth section from a vase or
perhaps bottle, made with eight
lobes on the exterior which at the
flared mouth are drawn outward
to make a foliated rim; the bluish-
green glaze crackled; the clay
body gray.

Example: 1
Height: 5.5 cm
Mouth diameter: 7.4 cm

The following pieces are made from brownish-gray or buff clay.

DISHES

Small and thin, with a flat unglazed base and a wide ring round the interior bottom, otherwise
plain.

Examples: Two base fragments only

Diameter at widest point: 7.5 cm

142
With an incised ring at the interior bottom set directly above
the circumference of a very short carved footring. The base
is slightly concave and unglazed. Faintly impressed
decoration at the center is no longer clear. The bluish glaze
is heavily crazed. The clay body is gray and grainy.

Example: 1 damaged
Diameter: 18 cm

A damaged dish made with fine grayish clay, with a

slightly everted foliated mouthrim and a finely carved
footring. Decorated on the interior with a stylized
swirling floral motif with an incised ring at the bottom
and incised curls near the mouthrim. A small knob of
clay is visible at the interior center. The glaze is thick,
opaque and, in some places, more pale greenish than
blue in color.

Example: 1 damaged
Diameter: 20 cm

143
Fragments only, of a saucer-like dish with a
short carved footring and everted mouth
with foliated rim. Elaborately decorated
with a swirling floral motif on the interior in
such high relief that the dish must have
been formed by pressing the clay into a
mold. A smaller incised or molded flower
spray decorates a ring at the interior center.
Covered in thick bluish-green glaze. The
exterior walls are plain. Fragments similar to
this and the following dish have been
discovered at the Jingdezhen kilns, (Oriental
Ceramic Society of Hong Kong, 1984, No. 11.)

Examples: 7 base fragments

Diameter: 16.5 cm

Further fragments, in the same shape as

the previous group, display carved petals
on the interior radiating outward from an
encircled incised floral design at the
center. Similar petals are also carved on
the exterior walls.

Examples: 2
Diameter: 16.5 cm

144
BOWLS

Sturdily potted, with the footring carved so that the base is only slightly inset and so unusually
thick. The interior is incised and combed with dense floral scrolling. The glaze is bluish-green and
the clay body pale gray. All examples have traces of a circular kiln stacking scar on the base.

Examples: 3 fragments
Diameter: 22 cm at widest point

EWERS

Very thinly potted, with eight flattened panels around the walls, each of the panels incised with
a floral design accented with combing. At the uppermost point of the long strap handle sits a
small retaining ring. A similar ring on the original cover would have allowed the two to be tied
together.

Examples: 3 damaged
Height: 25 cm

145
Gourd-shaped and lobed, these vessels were
constructed from the parts formed in piece
molds. Three types: (1) plain, (2) with a single
band of decoration round the lower shoulder,
and (3) with vertical floral sprays in relief. Its
curved handle is intricately modeled into the
shape of a leaping sea creature, with a
retaining ring at its upper side (for fastening a
string tie from cover to handle) and
sometimes a perforation just below.

Examples: 10 damaged
Height: 20 cm

146
 BOTTLES

With a bulbous body, a carved splayed

footring, and long sinuous dragon in sculptural
relief set around the tall neck. The dragon,
unfortunately broken at the head and tail, is
almost identical to those on the painted
bottles of this same shape described earlier in
this inventory. The body of the vessel is
decorated with scrolling stylized foliage in
incised and short combed lines.

Examples: 2
Height: 23 cm (broken at neck)

Gourd-shaped and mold-made, this

miniature bottle is decorated with vertical
rows of knobs on the lower portion above a
band of tiny lotus petals at the base and
with upright petals on the upper portion.
The flat base shows traces of a circular
stacking support scar. The glaze is fairly
thick and bluish, and the clay body is
brownish-gray.

Example: 1
Height: 6.5 cm

147
JAR

Small, with a bulbous upper portion, narrow

mouth with everted rim and a base that is
slightly inset. It is decorated with horizontal
potting grooves. The runny bluish glaze
ends in a thickened tier at the lower body.

Example: 1
Height: 10 cm

COVERS

This group is similar in size and construction but varied in design. The upper exterior portion,
which is slightly domed, sets over a vertical flange that would have fit into the mouth of the
parent vessel. Each has an attached retaining ring to one side. Some have floral designs in
relief, with a small twirled clay handle. Another type features a sculpted model of a four-legged
animal, possibly a ram.

Examples: 4
Diameter: 5.5 cm

148
 11.5 Celadon and Olive-Glazed Ware
While most of the bowl shapes excavated from the shipwreck could be described as basically
green-glazed, this group makes an obvious attempt to imitate wares traditionally known as
Northern Celadon. Many art historians would prefer to discard the term celadon altogether.
However, in order to distinguish this ware from the many other green-glazed wares in the cargo,
the traditional term has been maintained. Some pieces in this category are better described as
olive-glazed. All are relatively heavily potted and most are decorated with cursive carved and
incised designs. The glaze varies in shades of green but tends to dark rather than light.

BOWLS

Plain, with a straight mouthrim, a sturdily carved

footring, a single ring incised around the upper
interior walls, and a carved ring round the
bottom interior. The green glaze is heavily crazed
and ends unevenly at the lower exterior walls.

Examples: 3 damaged
Diameter: 17 cm

Deep with a slightly everted mouthrim and a carved footring beveled on its inner side to a
relatively shallow base. Inside, an undecorated center is delineated by a carved ring.
Decorated in carved and combed lines with cursive cloud forms on the interior walls, and with
carved, stylized lotus petals around the exterior. Dark green opaque glaze runs unevenly to the
lower walls. Tiny bits of grit sometimes mar the interior.

Examples: 2 intact, 3 damaged

Diameter: average 12.3 cm

149
Heavily potted, with a slightly everted mouthrim and a carved footring bevelled on its inner side
to a shallowly-cut base. Incised on the interior with a cursive, stylized floral design including
three blossoms accented with wisps of combing, and on the exterior with striations down the
walls. The glaze is thin at the mouthrim and ends unevenly just above the foot. The buff-gray
clay is grainy.

Examples: 5 intact, 20 damaged

Diameter: 18 cm

Heavily potted, with an unusual mouth that curves inward then out, and a heavily-carved
footring showing silver-gray grainy clay on the unglazed base. Decorated on the interior with a
carved ring around the bottom and with cursive cloud-like designs of incised and combed lines,
and on the exterior with curved striations. The clay body is silver-gray and grainy.

Examples: 7 intact, 20 damaged

Diameter: 17.5 - 18.5 cm

150
Wide and flaring, heavily potted, with a short carved footring, these are decorated on both the
interior and exterior with cursive incised and combed designs perhaps imitating cloud forms or
stylized vegetal scrolls. The center interior is delineated with a carved ring.

Examples: 3 intact, 30 damaged

Diameter: 24.5 - 25.5 cm

Deep and wide, these have a somewhat distinctive mouthrim that is short and curved.
Decorated in a cursive style, with vegetal and cloud forms in carved and combed lines on both
interior and exterior walls. There is a carved ring around the interior bottom and a short carved
footring.

Examples: 5 intact, 25 damaged

Diameter: 24 - 25 cm

151
Wide, shallow and heavily-potted, this group is decorated on the interior with wispy carved and
combed swirls and with an impressed Chinese mark within a circle at the center. The mark is ji, a
commonly used character meaning auspicious. Similar dishes found elsewhere often have
two characters, da ji, a wish for good luck or good fortune, instead of ji alone (e.g., Southeast
Asian Ceramic Society, 1979, No. 92, for the single character; and Dupoizat, 1995, Fig. 6, for
examples with both characters).

Examples: 2 intact, 5 damaged

Diameter: average 22.5 - 24 cm

Carved with in inset base, and with prominent ribs and tiers of deeply incised lotus petals on the
exterior. The interior is plain except for a small floral motif at the encircled inside bottom.

Example: 1 damaged
Diameter of fragment: 21 cm

DISHES

Widely flared dishes, intricately decorated with a

carved dense design of peony blossoms and
foliage on the interior and short carved curls on the
exterior. With a short finely carved footring,
unglazed on its rim and base. The clay is pale
grayish.

Examples: 4 damaged
Diameter: largest fragment 23 cm

152
 11.6 Decorated Green-Glazed Ware

These pieces do not fit into any clear categories of glaze type. Often they are simply
called transitional ware, a convenient reference to the thirteenth century change from Song to
Yuan dynasties. Although one basic glaze mixture was probably in use, varying firing conditions
produced a variety of color shades from creamy white to pale to dark green. At some kilns,
especially in later times, some of these variations would be codified and deliberately attained
with more careful monitoring of material quality and control of kiln atmosphere, especially with
the fine art of reduction control. Variations that appear accidental here include qingbai
(bluish-white) and what would later be called shufu (cream white). Because the glaze is
transparent, incised and carved decoration is more common than on the Longquan green-
glazed celadon that became popular in the fourteenth century. In exhibition catalogs
examples of the following ceramics have been dated from the tenth to the thirteenth century
(Southeast Asian Ceramic Society, 1979 and 1983). The glaze on most of them ends at the lower
walls; none has a glazed base. These ceramics are from several Fujian province kilns, all of which
made fairly similar products. Five sites have been documented: Anxi, Tongan, Nanan,
Quanzhou, and Putian. Most of the dish shapes are from Tongan and Nanan.

DISHES

Small and eggshell thin, with a straight mouthrim and flat

unglazed base; decorated on the interior bottom with lightly
incised stylized four-petal star-like designs accented with
combed marks; the clay body whitish with a sugary feel, a
clay similar to the type used for the molded white wares; the
glaze crazed and bluish-green in color.

Examples: 7 damaged
Diameter: 9 cm

Wide and flat, decorated on the interior with a cursory incised floral design accented with
curiously distinctive zig-zags of short combed lines. The exterior walls are sharply angled to a
small flat base. The green glaze pools on the interior bottom and at the sides of the base.
Probably from the Nanan kilns.

Examples: 4 damaged
Diameter: 11 cm

153
Small, with gently curved walls rising to a straight mouthrim. The base is flat, unglazed and very
occasionally has traces of a kiln support scar. Decorated within an encircled interior medallion
with a stylized, cursively incised floral spray. Covered with a somewhat runny glaze that usually
fires without crazing to various shades of green, some very grayish and others more blue. The
clay body is pale silver-gray and somewhat rough. In the fifteenth century dishes of exactly this
shape and size, except for a base carved into the bottom, would appear among blue-and-
white ware as the hole bottom dishes.

Examples: 600 intact, 300 damaged

Diameter: Average 10 cm, 1 larger example 12.5 cm

154
BOWLS

Small tea-bowl size, displaying the same designs, clay

and glaze as the previous dishes, with conical walls
and a short carved footring. The glaze runs unevenly
onto the exterior of the foot. The interior floral design is
encircled. They ring when tapped, clearly a sign of
firing at an high temperature.

Examples: 150 intact, 50 damaged

Diameter: Average 12.5 cm

Relatively shallow, with a slightly everted mouthrim. A

cursive floral spray on the interior runs over an encircled
bottom upwards to a carved ring around the upper
walls. The exterior is plain, and the base inside the
footring is thick.

Examples: 100 intact, 50 damaged

Diameter: Average 18

155
LOBED DISHES AND BOWLS

These display the same glaze and body characteristics as the previous decorated categories.
The lobes of the mouthrim are further accentuated by single striations at each rim nick that
appear to be made with thin lines of slip beneath the glaze; otherwise they are undecorated.

Dishes

Small, with flared walls and a wide flat interior bottom

encircled by a single carved ring.

Examples: 10 intact, 10 damaged

Diameter: 11-12 cm

Wide and flared, with an encircled flat interior bottom that

corresponds to an angled tier at the lower exterior walls.
Wide dishes such as this, with a sharply angled exterior
wall, are sometimes identified as brush washes, for use in
Chinese caligraphy.

Examples: 20 intact, 10 damaged

Diameter: 16.5 - 17.5 cm

156
Bowls

The curved everted mouthrim of these is nicked to form five

lobes that are accented with a striation between each
down the interior wall to a carved ring that encircles the
bottom. The glaze is pale grayish-green. The relatively
heavy footring is carved and unglazed. The clay body is
pale silver-gray, with a rough surface. Some examples
have an unglazed stacking ring carved on the interior.

Examples: 20 intact, 20 damaged

Diameter: average 12.5 cm

Another group is characterized by a straight mouthrim and six

lobes.

Examples: 18 intact, 22 damaged

Diameter: 18 cm

157
BOWLS WITH SIMPLE COMBED DECORATION

With a carved footring and plain exterior; the interior decorated with combed curls of
decoration within a band between the carved ring encircling the interior bottom and an incised
ring at the upper walls; the clay and glaze similar to the previous categories, with chatter marks
often evident on the lower exterior walls and pin holes in the glaze.

Small-sized with a straight mouth and a wide ring at the bottom interior.

Examples: 20 intact, 10 damaged

Diameter: 12- 13 cm

Medium-sized, with the mouth everted only at its lip. Some have
a broader ring at the interior bottom.

Examples: 400 intact, 250 damaged

Diameter: average 15-16 cm

158
Large-sized, with a slightly everted mouthrim. Some have an unglazed stacking ring cut onto the
interior, and some have bits of grit caught on the surface of the glaze.

Examples: 2 intact, 10 damaged

Diameter: 23 cm

BOWLS WITH CLOUD DECORATION

Bowls in a standard medium size, decorated on the interior with combed cloud-like motifs. The
mouthrim is straight and the footring is carefully carved. The glaze is pale green to gray, and the
clay silver-gray and slightly rough.

Examples: 1,000 intact, 250 damaged

Some have a ring of broad diameter at the

interior bottom.

Diameter: 17 - 21 cm

159
 Others display a small button-size encircled bottom
and slightly higher walls. One example has four fish
motifs scattered among the cloud designs.

Diameter: 17-21 cm

11.7 Undecorated Green-Glazed Ware

The glaze characteristics of these pieces are identical to the previous, decorated types.
They probably come from the same kilns. Three to four qualities of ware were often produced
from any single kiln firing. The following dish shapes seem to have been made for mass sales.
The clay is a little more rough, there are often chatter marks on the exterior walls and pin holes in
the glaze, and the footrings are not carefully carved.

BOWLS

With a slightly everted mouthrim, usually shaved

flat on its upper surface, an unglazed stacking
ring cut onto the interior bottom, incised rings at
the upper and lower interior walls, and carved
footring; often with bits of extraneous clay
adhering to the surface at odd places, and with
pin holes in the glaze.

Examples: 1200 intact, 400 damaged

Diameter: 14-17 cm

160
 Much like the previous category except with a more
smoothly everted mouthrim, and without the unglazed
stacking ring.

Examples: 120 intact, 60 damaged

Diameter: 14 -17 cm

Similar to the two previous categories but with a

narrower ring carved in slight relief at the interior
bottom.

Examples: 1800 intact, 600 damaged

Diameter: 14-17 cm

Wide and shallow, the bowls in this group are so heavily potted that they produce very little
ringing sound. They have an everted mouthrim, a carved ring on the upper interior walls, and
often chatter marks on the lower
exterior walls.

Examples: 180 intact, 30

damaged
Diameter: average 15 cm

BOWLS WITH EVERTED MOUTHRIM

161
Heavily potted and plain, with a somewhat
carelessly applied glaze that ends unevenly at the
lower exterior walls. Chatter marks are often
evident on the exterior. Unevenly carved footring.

Small-sized, some with an unglazed stacking ring

on the interior.

Examples: 100 intact, 30 damaged

Diameter: 12-13 cm

Medium-sized.

Examples: 100 intact, 50 damaged

Diameter: average 17 cm

162
COVERS

With a short solid stopper on the underside,

and made in three vertical tiers, with a
single perforation through the cover's wide
flat mouth flang. Glazed on the upper
surface only.

Examples: 1
Diameter: 6 cm

With a depressed center, the lower portion of which was

designed to fit into a vessel mouth, a wide flat mouth
flange and a ring handle set to one side in the
depression. Glazed on the upper surface.

Examples: 3
Diameter: 7.7 cm

With a short solid stopper below a depressed center

portion, a flared mouthrim with a ring handle to one
side, and a small cross-hatched knob handle at the
center. The greenish glaze is crazed.

Examples: 6
Diameter: 7.5 cm

163
With a wide flat mouth flange, a lotus-bud-shaped
handle, and a retaining ring to one side.

Examples: 3
Diameter: 8 cm

Jar cover, heavily potted, with straight sides and

a flared knob handle; covered on the exterior in
pale green opaque glaze. The clay is silver gray
and mottled, with beige undertones.

Example: 1
Diameter: 13 cm

Dome-shaped, with a wide flat mouth

flange and short flat knob handle; glazed
pale gray-green on its upper surface.

Example: 1
Diameter: 10.7 cm

Cover, probably from a covered box,

sturdily potted from silver-gray clay, with a
flat cover center angled downward to
meet the straight sides.

Example: 1
Diameter: 5.6 cm

164
 11.8 Brown-Glazed Ware
This group of ceramics contains a variety of shapes, although the majority of them are
containers of one sort or another. They share several basic structural techniques and could
perhaps be best described as a kind of utilitarian peasant pottery. Probably they are made
from a mixture of clays, for the basic medium gray color of their fired body is faintly mottled with
beige. Most of them were probably fashioned on a turntable, rather than on a true potters
wheel. Structurally, they begin with a solid disk of clay from which the walls are built upward with
coils of clay. If the first coil was laid atop the clay disk, it became a solid foot; if the coil was
begun from around the exterior of the disk, a separation between the initial disk and the walls is
visible on the base. Because a turntable would not rotate as quickly as a wheel, the coils are
never completely smoothed flat. These pieces were not, however, meant to be prized for their
beauty; they have more practical purposes as containers for other products. The bowl is an
exception. The glaze is usually thin, brown or greenish brown, and it rarely covers the whole
exterior body. The interiors of the closed shapes are only occasionally glazed.

Most of these pieces possibly came from the kilns at Quanzhou itself, the port from which
the ship most probably sailed. Eleven kilns, known for having produced low-fired utilitarian ware,
were sited at the east gate of the city.

BOWL

Lotus-pod shaped and thinly potted, with a

carved footring and straight mouthrim.
Glazed brown or blackish to the middle or
lower exterior walls. The clay is silver-gray
with beige undertones.

Examples: 4 intact, 2 damaged

Diameter: 11.5 cm

165
BASIN

Heavily-potted, with a flat, usually slightly

concave base and thickened mouthrim.
The walls are roughly conical, with
horizontal potting grooves. Originally
glazed brown on the interior and onto
the upper exterior wall. The clay body is
mixture of silver-gray with beige
undertones. One basin has impressed
stylized floral motifs on the inside bottom.

Examples: 7 intact, 7 damaged

Diameter: 24.4 - 26 cm
Height: 10 cm

BOTTLE

With bulbous body, a tall conical neck, a

rolled mouthrim, and flat base; brown-glazed
from a couple of centimeters within the
mouth to the lower body; the clay body
relatively lightweight.

Examples: 15 intact, 14 damaged

Height: 24 cm, one 30 cm

POURING VESSEL

This one is an unusual shape, perhaps for

wine, documented elsewhere only
among twelfth-century Cambodian
stoneware. Thinly potted, with a relatively
short but wide, bulbous body and flat
base. The short neck curves outward into
a cup-shaped mouth, into which a cover
was probably meant to sit. Brown-
glazed to the lower exterior walls.

Examples: 5 damaged
Diameter: one 17 cm, others 14 cm

166
JARLETS

Bulbous, with a short narrow neck and rounded

mouthrim. The flat base is slightly concave. The
clay is lightweight and silver-gray, and the glaze is
completely eroded away.

Examples: 2
Height: 6.5 cm

Squat and bulbous, with a mouth cut directly into

the upper surface, covered in an uneven line to
the lower walls in brown glaze. The clay body is
gray. This single example is chipped at one side
of the mouth, perhaps deliberately in order to be
used as a lamp, with the chip providing a place
for the wick.

Example: 1
Height: 3.8 cm

JARS

Relatively small in size with a flat base, narrow

mouthrim, and potting grooves evident around
the walls. Glazed only on the upper shoulder
and mouth. The clay is silver-gray with beige
undertones.

Examples: 1
Height: 11.6 cm

167
Tall and conical, with a narrow rounded
mouthrim and flat base; glazed only at the
upper shoulder and mouth. The clay is silver-
gray with beige undertones. These small-
mouth jars are believed to have held
mercury, a substance said to have had
medicinal uses.

Examples (medium): 4
Height: 18 cm

Examples (large): 2 intact,

1 damaged
Height: 30 cm

Bulbous, with a curved neck and everted

rounded mouth; the foot flat. Some are glazed
greenish-brown on the interior, otherwise they
may be glazed only on the exterior to the lower
walls; in various sizes.

Examples: 8
Height: 8.6 - 16 cm

Squat, with a wide body drawn into a short

narrow neck with flared mouthrim; the base
flat; the clay pale silver-gray; glazed dark
brown.

Examples: 60 intact, 20 damaged

Diameter: 9-16 cm

168
These jars appear in various sizes and proportions but all of them have four applied strap handles
around the shoulder, and a flat base or flat rudimentary foot. They were originally glazed from
inside the mouth to the lower walls, usually in greenish-
brown glaze. The clay is medium gray, sometimes
heavily potted and sometimes lighter in weight. Many
examples show a circular stacking scar on the base.
The largest size sometimes has a single incised wavy
line around its mid section, and a series of carved rings
at the neck. Occasionally there is a Chinese mark
within a vertical rectangular border stamped at
places around the upper shoulder (see following
page).

Examples: 80, some with minor damage

Height: 12 - 37 cm

169
Gong Zi Hao trademark of Chuan Xing name or trademark Bing Sheng a cyclical date/60
Gong stamped on a storage stamped on a storage jar year cycle stamped on storage
jar jar

Bing Zi Nian a cyclical date/60-year cycle stamped

on storage jar shard

Chang Zi Hao trademark of

Chang stamped on storage
Zi Shou a surname Zi Hao trademark of... stamped
stamped on storage jar on storage jar

170
COVER

With a slightly domed center

surmounted by a handle in the form
of a strip of curled clay, a wide cover
rim, and a vertical flange below
which would fit within the mouth of a
vessel.

Examples: 2
Diameter: 6-7 cm

EWERS

Bulbous, with a wide neck and everted

rounded mouthrim. A strap handle is applied
to one shoulder and a curved spout at the
opposite side. A single wavy line at the upper
body is similar to ones seen on some of the
four-strap handled jars. Roughly potted and
originally glazed to the lower exterior walls.

Example: 1 damaged
Height: 15.6 cm

Large and intricately modeled with an

unidentifiable animal at the spout, possibly
with a long snout and the horns of a ram,
and with a wide sturdy strap handle at the
opposite shoulder. The wide flared neck
and ringed everted mouthrim were made
separately and luted to the main body.
Covered in runny brown glaze.

Example: 1, a fragment
Height: 15.5 cm

171
 11.9 Indian-Style Earthenware
All the vessels of this unusual ware are elaborately modeled. Their style suggests a
southern Thai origin. This is strongly supported by comparative X-ray florescent analysis (see
Section 9.8). They were probably used in Hindu-Buddhist rituals, for similar shapes are sometimes
seen in religious sculptures and bas-reliefs. Some of them, like the wide-bodied kendi, may also
have been employed in the brewing of medicines and medicinal teas. Each of the pieces is
unique in its exact proportions, an indication that each was made individually. Some are
kundika, an ancient Indian shape that was copied into some Chinese wares during the early
Tang dynasty (A.D. 618-906). The kundika shape is believed to precede the kendi, which
probably evolved in Southeast Asia, most likely in Indonesia, the destination of this cargo. The
original surface of these vessels appears to have been burnished to a brownish-red, traces of
which remain. Now they have a mostly bare surface and are colored orange-brown; recent
breaks in smaller examples show a slightly yellowish ivory-colored clay, slightly porous, with a feel
something like fresh-carved ivory. Where the clay is more thick, breaks show a black core in
profile. The clay may have originally been gray. Altogether there are about 360 examples of
various kendi and kundika, all of them with some damage.

KUNDIKA

This basic shape features an oval, sometimes egg-shaped body set on a flared flat foot.
Sometimes there is a carved band at the lower body that appears to simulate an original pot set
onto a stand. A straight conical spout is set, pointing dramatically upward, at the upper
shoulder. The tall neck, where one would grasp the vessel, is ornamented with horizontal bands,
bulges and flanges.

Height: 12-17 cm

172
One kundika has a more angular body and cylindrical shape, with traces of faintly incised
(possibly mystic) designs around the lower walls.

KENDI

Wide-bodied, with a wide straight neck and everted

flanged mouthrim; decorated with carved rings and
with striations around the main body. The spout is
short and bulbous with a flange at its tip. The foot is
flat.
Height: small 11 cm
large 14 cm

173
Wide-bodied, with a wide straight neck, an everted flanged mouthrim, a long bulbous spout
with a flange at its tip and a flat foot. This type is undecorated except for a small flange at the
lower neck.

Height: 15 cm

Gourd-shaped, with a wide flared foot, a flange and multiple bands at the mid-section, and a
bulbous spout set into the upper section.

Examples: 10 damaged
Height: average 16 cm

174
Toroid shape, with a spout at the upper shoulder of the
tubular ring that forms the body, and an attached foot.

Examples: 13 damaged
Height: approximately 25 cm

Kendi Maling or Thieves kendi in Indonesian. The body is squat, with a wide banded neck,
and the cover is totally sealed. A funnel leads into the interior of the vessel from the flat base.
This is a traditional shape in Indonesia, where it is still made today (Adhyatman, 1987, No. 141-
142).

Example: 1 intact, 2 damaged

Height: 12.7 cm

175
HOOKAH

This shape has an ovoid body, flared foot and flat base. It resembles an Indian or Middle Eastern
vessel used for smoking. The shoulder is surmounted by a long neck with bulbous sections and
flanges that seem to terminate in a small narrow mouth (its tip broken). The spout-like
appendage is sharply curved upward. If this were in fact a smoking device, the smoked
substance would sit (probably together with a bit of charcoal to provide heat) within the top of
the spout and the smoke would be drawn through the water-filled body up into the mouth tip.

Examples: 1 intact, 3 fragments

Height: 28 cm

BOTTLE

With a wide, flared flat foot, hollow on its inside;

and multiple carved bands and flanges between
two bulbous body sections and at the tall wide
neck.

Example: 1
Height to broken neck: 19.6 cm

176
RICE POT

Wide and squat, with a rounded base, and tall

wide neck with a wide-flanged everted
mouthrim, decorated with carved rings, angular
bands at the neck and striations down the walls
of the mid-section.

Examples: 10
Diameter: 20 cm

COVERS

Probably for a rice cooking pot. It is heavily

potted and with a circular vertical flange as a
handle, curved on its underside, and one
centimeter thick.

Example: 1
Diameter: 12.5 cm

Wide and tall in a roughly gourd shape, with

both angular and bulbous tiers rising upward
to a lotus-bud-shaped, knob-like handle; the
interior is hollow.

Example: 1
Height: 13 cm (broken at the mouth
section)

177
UINIDENTIFIED

Probably for religious ritual

use, perhaps a stand upon
which another vessel was
set. The hollow within the
wide, tall flared footring
extends upward to the wide
mouth at the upper
cylindrical portion, with a
bowl-like section surrounding
the center column.

Example: 1
Height: 13 cm

11.10 Miscellaneous Ceramics

Table or stand, probably made

from a heavy, now gritty
earthenware clay (or possibly
carved from a stone such as
sandstone). Circular, with three
solid legs (one missing). Carved
on the underside with an
unidentifiable animal (perhaps a
lion) in a center medallion
encircled with two bands of
sawtooth design separated by
another band with lotus panels.
The central portion of the upper
side is plain and flat, with
carved borders around its
circumference.

Example: 1
Diameter: 31.5 (in
fragments, with one
portion missing)

178
179
OIL LAMP

Oil lamp, thickly potted and pinched

outward at one side, with a flared foot.
Similar ware is known from Majapahit-Java
(Miksic & Soekatno, 1995). The find of a
single example implies this was used
onboard the ship and was not an item of
trade. Sandy dark grey clay.

Examples: 1
Diameter: 8 cm

RICE MEASURE

The interior is glazed brown with no glaze on

the exterior. Decorated with a carved
fingerprint-like design and appliqu dots
around the neck. It copies an archaic Chinese
shape.

Examples: 1
Height: 5 cm

180
DISK

Low-fired and made from clay pressed into a mold. It depicts four women, each with a basket
in the left hand and stick in the right, walking over a bridge.

Example: 1
Diameter: 9 cm

181
12.0 CONCLUSION
The nature of the Chinese ceramics cargo, and radiocarbon dating of a resin sample,
suggest that the Java Sea wreck should be dated to approximately the middle of the 13th
century. Although other wrecks of approximately the same age are known in Indonesian
waters, no data on the precise location, composition of cargo, or nature of structural remains
has ever been published.

The location of the ship, in the Java Sea, half way between Bangka and Jakarta, is the
first archaeological evidence for the use of what is now a busy modern shipping lane. From the
position alone, the ship may have passed through either Bangka or Gaspar Strait. But given the
nature of part of the cargo, it is most probable that Bangka Strait was used.

The array of artifacts, representing non-ceramics trade goods, constitutes the first reliable
information on pre-European trade in Indonesia. Only with such information, and more like it, is it
possible to evaluate the relative importance of maritime trade in the economies of Indonesia
and her trading partners during this early period. At present, scholars are unable to draw precise
conclusions on the contribution of foreign trade to the ancient Indonesian economy, in
comparison to agriculture and cottage industry.

Given the complete absence of cohesive hull structure, it is impossible to establish the
origin of the ship firmly. However, wood identification, internal layout (as determined from the
iron concretion disposition), fastening details, and anchor type, provide sufficient information to
conclude that the ship was not constructed in China. Whether the ship was built in Thailand or
Indonesia remains open to conjecture. The existence of bulkheads tips the scales in favor of
Thailand, whilst the presence of Javanese-type bronze artifacts might indicate an Indonesian
origin.

The wreck contains a wide variety of ceramics, as well as other commercial

commodities including iron, ivory, resin, glass and bronze ware. Most of the ceramics were
manufactured in China (mainly Fujian). The iron cauldrons and bars that made up the bulk of
the cargo were produced in China. The earthenware kendis are probably from southern
Thailand. Resin and ivory are likely to have been obtained from Sumatra. Glass is mainly from
China. The bronze ware is typical of Majapahit and pre-Majapahit craftsmanship, and may
have belonged to the Indonesian crew, or remained on board from a previous trip to Java. This
range of artifacts from various ports is consistent with theories regarding the "peddling nature"
of some Southeast Asian commerce during this period, somewhat reminiscent of an early
twentieth-century tramp steamer.

Trade in iron is mentioned in several ancient texts, but until now, there have been no
documented archaeological sites to confirm just how important the Chinese iron trade was. We
can now begin to judge the relative importance of iron in the two-way trade and the precise
forms of the artifacts involved.

The vast majority of the ship's ceramics cargo was from China. Interestingly, none of the
many varieties of ceramics seems to have come from Longquan, one of the more productive
kilns of this period. Many of the forms (the qingbai ewers, conical bowls, covered boxes, and
vases), are well known to students of Chinese ceramics of this period. Some varieties, however,
such as the qingbai dishes with heavily molded bodies, are quite unusual. The high frequency of
boxes with inscriptions on the bases are another particularly interesting find, since vessels of this
type are not common in Southeast Asia, although they have been studied in China. The green
porcelain seems to belong to well-known types from Fujian, principally the Tongan kiln complex.

182
The larger storage jars are also in general from known types, but the high frequency of jars with
stamped characters on the shoulders is unusual. The elaborate ewers, boxes, and bowls with
white floral motifs on a background of brown glaze are very rare. In general their clay and
decorative motifs are similar to those associated with the Cizhou kiln complex, but examples in
which brown floral decorations were placed over a white background are thought to be
unknown. The reverse process, with the possible addition of green overglaze, in complex
decorative compositions, is a most surprising and intriguing discovery which will excite the
attention of ceramic connoisseurs.

The large numbers of kendis from southern Thailand are the first indication that
earthenware was traded between Southeast Asian countries, an item completely absent from
any surviving documentary records. There are several unprecedented shapes. The toroid
kendis, in particular, were previously only known from Chinese versions made during the Ming
period.

In summary, the archaeological results of this project have yielded considerable data,
both qualitative and quantitative, which will add to the knowledge of many aspects of early
trade between China and Indonesia, and, equally important, the intra-Southeast Asian trade.
This significant topic has received little attention compared to trade between Southeast Asia
and other regions. It is hoped that the discoveries made on this site will stimulate a
reconsideration of this field of study.

183
184
 BIBLIOGRAPHY

SECTION 2.0 by Dr. John Miksic

An Jiayao, Glass trade in Southeast Asia, in Ancient Trades and Cultural Contacts in Southeast
Asia. Bangkok, Office of the National Culture Commission, 1996, pp.127-138.

Arasaratnam, V.S., Merchants of Coromandel in Trade and Entrepreneurship circa 1650-1700, in

Emporia, Commodities and Entrepreneurs in Asian Maritime Trade, C. 1400-1750, edited by
Roderich Ptak and Dietmar Rothermund, Stuttgart, Franz Steiner Verlag, 1991.

Chang Pin-tsun, Chinese Maritime Trade: the case of 16th-century Fu-chien (Fukien), Ph.D.
Thesis, Princeton University, East Asian Studies, 1983.

Ch'en, J., Sung Bronzes - an Economic Analysis, BSOAS 28/3, 1965: 613-626.

Christie, J.W., Trade and settlement in early Java: integrating the epigraphic and
archaeological data, in I. Glover, P. Suchitta, and J. Villers, eds., Early Metallurgy, Trade and
Urban Centres in Thailand and Southeast Asia, Bangkok, White Lotus, 1992, pp. 181-197.

Christie, J.W., On Po-ni: the Santubong sites of Sarawak, Sarawak Museum Journal 34, 1985: pp.
77-89.

Crawfurd, J., Descriptive Dictionary of the Malay Archipelago, 1856.

Ferrand, G., L'Empire Sumatranais de Srivijaya, Journal Asiatique 20, 1992, pp. 1-104.

Harrisson, T., O'Connor, S.J., Excavations of the Prehistoric Iron Industry in West Borneo, Ithaca,
Cornell University, Southeast Asia Program, Data Paper Number 72, 1969, two Volumes.

Ho Chuimei, The Significance of West Asian Ceramics in East and Southeast Asia in the 9th -10th
Century, Trade Ceramic Studies No. 14, 1994, pp.35-59.

Kenderdine, Sarah, Bai Jiao I-the excavation of a Song Dynasty shipwreck in the Dinghai area,
Fujian Province, China, 1995, International Journal of Nautical Archaeology 24,4, 1995, pp. 247-
266.

Kwee Hui Kian, Dao Yi Zhi Lue as a Maritime Traders' Guidebook, HT, 1996-97.

Li Hui-Lin (ed.) Nan-fang Ts'ao-mu Chuang, A Fourth Century Flora of South-East Asia, Hong
Kong, Chinese University Press, 1979.

Ma Jun-chao, Lawrence, Commerce Development and Urban Change in Sung China, Ann
Arbor, Ph.D. Thesis, University of Michigan, 1971.

Manguin, Pierre-Yves, Trading ships of the South China Sea: Shipbuilding techniques and their
role in the history of the development of Asian trade networks, Journal of the Economic and
Social History of the Orient 36, 1993, pp. 253-280.

Needham, J., The Development of Iron and Steel Technology in China, London, 1958.

185
Ptak, R., China and the trade in tortoise-shell (Sung to Ming periods), in Emporia, Commodities
and Entrepreneurs in Asian Maritime Trade, C. 1400-1750. Edited by Roderich Ptak and Dietmar
Rothermund, Stuttgart, Franz Steiner Verlag, 1991.

Reid, A., Flows and seepages in the long-term Chinese interaction with Southeast Asia, in
Sojourners and Settlers: Histories of Southeast Asia and the Chinese, edited by Anthony Reid, St
Leonards, Allen and Unwin, 1996, pp. 15-49.

Rockhill, W.W., Notes on the relations and trade of China with the Eastern Archipelago and the
coast of the Indian Ocean during the fourteenth century, Toung Pao 15:, pt. 1, pp. 419-447; pt. 2,
pp. 61-159, 236-271, 374-392, 435-467, 604-626, 1914.

Treloar, F.E., Chemical analysis of some metal objects from Chandi Bukit Batu Pahat, Kedah:
suggested origin and date, JMBRAS 41, 1968, pp. 193-198.

Wang Gungwu, The Nanhai trade. A Study of the Early History of Chinese Trade in the South
China Sea, Journal of the Malayan Branch of the Royal Asiatic Society 31,2, 1958, pp.1-135.

Wheatley, P., Geographical Notes on some Commodities involved in Sung Maritime Trade,
Journal of the Malayan Branch of the Royal Asiatic Society 32,2, 1959 pp. 1-140.

Wicks, Robert S., Money, Markets, and Trade in Early Southeast Asia. The development of
indigenous monetary systems to AD 1400.

Wicks, R.S., Ancient coinage of mainland Southeast Asia, JSEAS, 1985.

Wong, Grace, A comment on the tributary trade between China and Southeast Asia, Southeast
Asian Ceramics Society, Transaction No. 7, 1979.

Wong, Grace, Chinese blue-and-white porcelain and its place in the maritime trade of China,
in Chinese Blue and White Ceramics, 1978.

SECTIONS 6.0 & 7.0 by Michael

Flecker
Blake, W., Flecker, M,. A Preliminary Survey of a Southeast Asian Wreck, Phu Quoc Island,
Vietnam, International Journal of Nautical Archaeology, Vol 23(2), 1994.

Flecker, M., Excavation of an Oriental Wreck of c. 1690 off Con Dao, Vietnam, International
Journal of Nautical Archaeology, Vol 21(3), 1992.

Goddio, F., Discovery and Archaeological Excavation of a 16th Century Trading Vessel in the
Philippines, World Wide First, 1988.

Green, J., The Shinan Excavation, Korea: an Interim Report on the Hull Structure, International
Journal of Nautical Archaeology, Vol 12(4), 1983.

Green, J., The Song Dynasty Shipwreck at Qhanzou, Fujian Province, Peoples Republic of China,
International Journal of Nautical Archaeology, Vol 12(3), 1983.

186
Green, J., Harper, R., The Excavation of the Pattaya Wreck Site and Survey of Three Other Wreck
Sites in Thailand, 1982, Australian Institute of Marine Archaeology, Special Publication No. 1, 1983.

Green, J. et al, The Maritime Archaeology of Shipwrecks and Ceramics of Southeast Asia,
Australian Institute of Marine Archaeology, Special Publication No. 4, 1987.

Guy, J., Oriental Trade Ceramics in Southeast Asia, 9th to 16th Century, Oxford University Press,
1990.

Hoffman, P. et al, The 14th Century Shinan Ship - Progress in Conservation, International Journal of
Nautical Archaeology, Vol 20(1), 1991.

Horridge, A., The Prahu - Traditional Sailing Boat of Indonesia, Oxford University Press, 1985.

Horridge, A., The Lash-Lug Boat of the Eastern Archipelagoes, Maritime Monographs and Reports,
No. 54, National Maritime Museum, Greenwich, 1982.

Horridge, A., The Design of Planked Boats in the Moluccas, Maritime Monographs and Reports,
No. 38, National Maritime Museum, Greenwich, 1978.

Keith, D.H., Buys, C.J., New Light on Medieval Chinese Seagoing Ship Construction, International
Journal of Nautical Archaeology, Vol 10(2), 1981.

Kenderdine, S., Bai Jiao 1 - The Excavation of a Song Dynasty Shipwreck in the Dinghai Area,
Fujian Province, China, 1995, International Journal of Nautical Archaeology, Vol 24(4), 1995.

Li, G.Q., Archaeological Evidence for the use of Chu-nam on 13th C Quangzhou Ship, Fujian
Province, China, International Journal of Nautical Archaeology, Vol 18(4), 1989.

Manguin, P-Y., The Southeast Asian Trading Ship: an Historical Approach, Journal of Southeast
Asian Studies, Vol 11(2), Singapore University Press, 1980.

Miksic, J.N., Seokatno, (Ed), The Legacy of Majapahit, National Heritage Board, Singapore, 1995.

Miksic, J.N. et al, X-Ray Florescence Analysis of Glass from Fort Canning, Singapore, Buletin de
lEcole Francais dExtreme-Orient, Vol 83, 1996.

Needham, J., Science and Civilization in China, Vol 4(3), 1971.

Shaffer, L.N., Maritime Southeast Asia to 1500, M.E. Sharpe, New York, 1996.

Van Tilburg, H., Chinese Maritime History and Nautical Archaeology: Where have all the Ships
Gone?, The Bulletin of the Australian Institute of Maritime Archaeology, Vol 18(2), 1994.

Wang Gung-Wu, The Nanhai Trade, a Survey of the Early History of Chinese Trade in the South
China Sea, Journal of the Malayan Branch of the Royal Asiatic Society, Vol 31(2), 1958.

Waters, D.W., Chinese Junks: the Antung Trader, Mariners Mirror, Vol 24, 1938.

Waters, D.W., Chinese Junks: the Pechili Trader, Mariners Mirror, Vol 25. 1939.

187
Wong, G., A Comment on the Tributary Trade Between China and Southeast Asia, Southeast
Asian Ceramic Society, Transaction No. 7, 1979.

Worcester, G.R.G., Junks and Sampans of the Yangtze, Vols I & II, 1947.

Worcester, G.R.G. Sail and Sweep in China, Science Museum, London, 1966.

Zim, H., Shaffer, P., Rocks and Minerals, Golden Press, New York, 1957.

SECTION 8.0 by Dr. Bennet

Bronson
Abendanon, E. C. , Midden-Celebes Expeditie: Geologische en Geographische Doorkruisingen
van Midden-Celebes (1909-10), Leiden, Vol 4, 1917.

Bemmelen, R. W. van, The Geology of Indonesia, The Hague, Vol 2, 1949.

Bronson, B., Notes on the History of Iron in Thailand, Journal of the Siam Society 73, pp. 205-225,
1985.

Bronson, B., The History and Making of Wootz, A Crucible Steel of India. Archeomaterials 1, pp.
13-51, 1986.

Buchanan, F., A Journey from Madras through the Countries of Mysore, Canara and Malabar, 3
vols. London, 1807.

Chakrabarti, D.K., The Early Use of Iron in India, Delhi, 1992.

Corteso, A., (trans. and ed.) The Suma Oriental of Tom Pires, London, 1944.

Daghregister 1887-1931 Dagh-Register gehouden in 't Casteel Batavia, 1628-1682, 41 vols,

various editors. The Hague and Batavia, 1887-1931.

Hummel, C. L. and Prachuab Phawondon, Geology and Mineral Deposits of the Phuket Mining
District, South Thailand, Report of Investigation no. 5, Department of Mineral Resources, DMR-5,
Bangkok, 1967.

DMR Maps, Mineral Resources of Thailand (1:2,500,000 Map Series). Department of Mineral
Resources, Bangkok, 1975.

Groneman, J., Der Kris der Javaner, Internationales Archiv fr Ethnographie 19: pp. 90-109,
123-161, 178-211, 1910.

Hendriks, A., Lets over de Wapenfabrikatie op Borneo, Verhandelingen van het Bataviaasch
Genootschap 18, 2: 1-30, 1842.

Hirth, F. and Rockhill, W.W., (trans. and ed.), Chau Ju-Kuas Chu-Fan-Chi, A Description of
Barbarous Peoples. St. Petersburg, 1911.

Ho, Chuimei, Minnan Blue and White, An Archaeological Survey of Ancient Kiln Sites in South
Fujian, 230 pp. British Archaeological Research, International Series, vol. 428. Oxford, 1988.

188
Kruyt, A.C., Het Ijzer in Midden-Celebes, Bijdragen voor Taal-, Land- en Volkenkunde 53: 150,
1901.

MRDS-39, Mineral Resources of the Lower Mekong and Adjacent Areas of Khmer Republic,
Laos, Thailand and Republic of Viet-Nam, Mineral Resources Development Series, No. 39, United
Nations, 1972.

Marschall, W., Metallurgie und Frhe Beseidlungsgeschichte Indonesiens, Ethnologica 4: 29-263,

1968.

Moura, J., Le royaume du Cambodge, Vol. 1. Paris, 1883.

Raffles, T.S., The History of Java, Vol 2. London, 1830.

Reid, A., Southeast Asia in the Age of Commerce, Vol 1: The Lands below the Winds. New
Haven, 1988.

Rostoker, W. and Bronson, B., Pre-Industrial Iron, Its Technology and Ethnology, Archeomaterials
Monograph No. I, Philadelphia, 1990.

Survey, Survey of World Iron Resources, Department of Economic and Social Affairs, United
Nations, 1970.

Tegengren, F. R., The Iron Ores and Iron Industry of China, Memoirs, Geological Survey of China,
Ser. A, No. 2. Peking, 1923-4.

Valentyn, F., Oud en Nieuw Ooost-Indien, Dordrecht and Amsterdam, 1726.

Wagner, D.B., The Earliest Use of Iron in China. To be published in Paul T. Craddock and J. Lang
eds., Aspects of Early Mining and Extractive Metallurgy, 1996.

Zhu Fan Zhi, (ed. and pub.), Wang Dayuans Dao Yi Zhi Le, in Shanghai Antique Books ed. and
pub., pp 71-104. Shanghai [in Chinese], 1993.

SECTIONS 9.0, 10.0 & 11.0 by Roxanna

Brown

Adhyatman, Sumarah, Kendi, Himpunan Keramic Indonesia, Jakarta, 1987.

Chin, Lucas, Ceramics in the Sarawak Museum, Kuching, Sarawak, 1988.

Dupoizat, Marie-France, The Ceramic Cargo of a Song Dynasty Junk Found in the Philippines
and its Significance in the China-South East Asia Trade, in South East Asia & China: Art,
Interaction & Commerce, London, Percival David Foundation of Chinese Art, 1995, pp. 205-224.

Hughes-Stanton, P. and Rose Kerr, Kiln Sites of Ancient China, Recent Finds of Pottery and
Porcelain, London, The Oriental Ceramic Society, 1980.

Oriental Ceramic Society of Hong Kong, Fung Ping Shan Museum, Jingdezhen Wares, The Yuan
Revolution, Hong Kong, 1984.

189
Southeast Asian Ceramic Society, Chinese Celadons and Other Related Wares in Southeast Asia,
Singapore, Arts Orientalis, 1979.

Southeast Asian Ceramic Society, Song Ceramics, Singapore, 1983.

Ye Wencheng, Qingbai kilns in Yiyao and Qingyao, Mingqing Xian, Fujian in Ho Chuimei, ed.,
Ancient Kiln Technology in Asia, Hong Kong, Centre of Asian Studies, University of Hong Kong,
1990, pp. 60-73.

Zainie, C., Harrisson, T., Early Chinese Stonewares Excavated in Sarawak, 1947-67, The Sarawak
Museum Journal, Vol. XV/30-31 (July-December 1967) pp. 30-87.

190
 PROJECT PERSONNEL

Bill Mathers Mike Flecker Nigel Kerr Rahman

Jeff Nicklason Anthony Holley John Miksic Iwan Sumantri

Mustar Basri Rohili Sanimat

Junaidi Nasan Yusup Mugeni

256
Mustafa Ismunanto Herianto Rasidi

Mat Dolla Bahroni Hasballah Tjipto

Sukardi Amin Dayat Yapih

257