Hindu Gods of Peninsular Siam Author(s): Stanley J. O'Connor, Jr. Reviewed work(s): Source: Artibus Asiae. Supplementum, Vol.

28, Hindu Gods of Peninsular Siam (1972), pp. 35+7-9+11-76 Published by: Artibus Asiae Publishers Stable URL: http://www.jstor.org/stable/1522663 . Accessed: 06/02/2012 00:18
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Artibus Asiae Publishers is collaborating with JSTOR to digitize, preserve and extend access to Artibus Asiae. Supplementum.

http://www.jstor.org

STANLEY

J. O'CONNOR,

JR.

HINDU
O F P E NI N

GODS
S IAM

SULAR

M CMLXXII

ARTIBUS ASIAE PUBLISHERS/6612

ASCONA

SWITZERLAND

This book has been published with the aid of a grant from the
HULL MEMORIAL PUBLICATION FUND OF CORNELL UNIVERSITY

TABLE OF CONTENTS

Map of PeninsularSiam . . . . . . . . . . . . . . . . . . . . . . . . Map of India and SoutheastAsia . . . . . . . . . . . . . . . . . . . . ChapterI ChapterII

inside cover inside cover

II
i9

A Bridge and a Barrier . . . . . . . . . . . . . . . . . . . . . . Conch Shell on Hip: The "Aberrant"Statuesof Visn.u . . . . . . . . . Vis.nufrom Jaiyi . . . . . . . . . . . . . . . . . . . . . . Visn.uat Nagara Sri Dharmardja. . . . . . . . . . . . . . . . . Present Dating . . . . . . . . . . . . . . . . . . . . . . . . Inscriptionsfrom Nagara Sri Dharmardja. . . . . . . . . . . . . Visnu from Oc-Eo . . . . . . . . . . . . . . . . . . . . . . Iconography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I9 2i
26
27

28
29

ChapterIII Indian Comparisons . . . . .. . KusdnaImages from Mathuri . Visnu from Bhinmal, Gujarat. The Y8lswaram Excavations . Summary. . . . . . . . . .

32

33 34 37 38
41

ChapterIV The Visnu from Takuapa . . . . . . . . . . Present Dating . . . . . . . . . . . . . PallavaAnalogies . . . . . . . . . . . . Sichon (Sijala) . . . . . . . . . . . . . . Sating Pra . . . . . . . . . . . . . . . Hua Kao Village, Surd~tradhani Province . Petburi ..... . . ... . ... . . . . . Conclusion ..... . . . . . . . . . . . ChapterV

. . . . . . . . . .

. . . . . . .

. . . . . . . . .

. . . . . . .

. . . . . . . . .

. . . . .

. . . . . . .

. . . . .

. . . . . . .

4I 43
45

46 46 47 48
52

Late Pallavaand Cola Style on the Peninsula . . . . . . . . . . . . . Three C6oaImages . . . . . . . . . . . . . . . . . . . . . . .

6o 64 73

BibliographyAbbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . Index. 5

LIST OF ILLUSTRATIONS

Fig.

a. Visnu. Stone. Found at Jaiyl, southern Thailand. BangkokMuseum. Front; b. Side; c. Rear

Fig. z Visnu. Stone. Found at Nagara Sri Dharmaraja(Ligor). Wat Mahidhitu, Nagara Sri Dharmaraja Fig. 3 Visu. Stone. Found at Nagara Sri Dharmaraja(Ligor). Wat Mahidhitu, Nagara Sri Dharmaraja Sai Fig. 4 Visnu. Stone. Found at Oc-Eo, South Vietnam. National Museum, gon Fig. 5 Lihga. Stone. Nagara Sri Dharmaraja Fig. 6 Visnu. Stone. From RaniwalaWell at Palikhera,Mathura,India. Kusan.aPeriod. Mathuri Museum (accession register No.933) Fig. 7 Visnu. Found near Mathurd.Kusina Period. Mathbur Museum (accession register No.I729) Fig. 8 Visnu. Stone. Early Kusana Period. MathurdMuseum Fig. 9 Visnu. Stone. Found in a well at Baghichal Birhal near Mathurl. Kusana Period. Mathbur Museum (accession register No.956). a. Front; b. Rear Fig. io Fig.
II

Visnu. Stone. Late Kusina Period. MathuraMuseum Visnu. Stone. Found at BhinmAl,Gujarat.BarodaMuseum. Photograph from: Bulletin of Baroda Museum, Vol. XII (I95 5-I956)

Fig. i z Visnu. Stone. Excavated at Y8lswaram, Andhra Pradesh,India. 4th or 5th century A.D. Fig. Fig. Fig.
I3 I4 I5

NarasirhhaRelief. From CoastalAndhra Pradesh. 3rd century A.D. (After M.W.Khan) a. Visnu. Stone. Found at Takuapi, Thailand. BangkokMuseum. Front; b. Side; c. Rear a. Visnu I. Stone. Found at Vieng Sra, Thailand. BangkokMuseum. Front; b. Side II. a. Visnru Stone. Found at Vieng Sra, Thailand. BangkokMuseum. Front; b. Rear

Fig. i6 Fig.
x7

a. Visnu. Stone, Found at SrIvijayaHill, Suristradhini, Thailand. BangkokMuseum. Front; b. Rear a. Fig. x8 Visu. Stone. Found at Sating Pra, Thailand.BangkokMuseum. Front; b. Rear Fig.
i9

a. I. Visnru Stone. Found at Petbuti (Bejrapuri),Thailand BangkokMuseum. Front; b. Rear Visnu. Stone. Found at Sating Pra, Thailand. Wat Majjhimivisa, Songkla. a. Front; b. Side Visn.u. Stone. Found near Wat Srivijaya,Hua Kao village, Pun Pin district, Suristradhini, Thailand. a. Front; b. Side. Photograph: A. B. Griswold Visnu. Stone. Found near Wat Srivijaya,Hua Kao village, Pun Pin district, Suristradhini, Thailand. a. Front; b. Side. Photograph: A. B. Griswold Visiu. Stone. Found near Wat Srivijaya,Hua Kao village, Pun Pin district, SuristradhAni. Photograph: A. B. Griswold 7

Fig. zo Visnu II. Stone. Found at Petburi, Thailand. BangkokMuseum Fig. zi Fig. 23 Fig. Fig.
24

Fig. zz Visnu. Stone. Found at Sichon, Thailand. PrivateCollection, Nagara Sri Dharmardja

25

Fig. z6 UnidentifiedFigure. Stone. Found nearWat Srivijaya,Hua Kao village, Pun Pin district, Suristradhini. Photograph: A. B. Griswold. a. Front; b. Rear Fig. 27 Visnu. Stone. Collection H.R.H.Prince Chalermpol of Dighambara.Bangkok. Photograph from: Exhibition Masterpieces PrivateCollections, of from National Museum, Bangkok.i968 Fig. z8 'Pra NarAi' (Nirayana) Group. Stone. Located on Takuapi River, approximately iO miles upstream from the estuary. Photograph: A. B. Griswold Fig. 29 CentralMale Figure, 'Pra Nar~i' Group, Takuapi. Photograph: A.B. Griswold Fig.
30

SmallerMale Figure, 'Pra NarVi'Group, Takuapi. Photograph: A. B. Griswold Female Figure, 'Pra Nar"i' Group, Takuapi. Photograph: A. B. Griswold Visnu. Stone. Found at Vieng Sra, Thailand. BangkokMuseum. Front; b. Side a.

Fig. 3' Fig.

32

Fig. 33 Batuka-Bhairava. Stone. Found at Vieng Sra. BangkokMuseum. Front; b. Side a. Fig. 34 Sfirya. Stone. Found at Jaiyi. Bangkok Museum. a. Front; b. Side; c. Rear

PREFACE

his book is aboutthe ancientpast of a compactand sea-flanked of Southeast Asia. part

More preciselyit is a view of the past as it can be recreatedfrom a study of monumental statues of Hindu gods discovered in PeninsularSiam. In this group of shaped stones, I believe, there is an indistinct, but partly recoverablemessage, a record of lost kingdoms and city-states in a little-studiedtract of earth which is dense with the debris of early civilization. I firstthought of trying to decipherthis messageduringvisits to PeninsularSiamin i963-64, and I wrote my findings in a doctoral thesis for Cornell University in i965. This book is a shortenedversion of the thesis, revised in the light of additionalfield work performedin i966. Readers who take the trouble to look at footnotes will be aware of my indebtedness to three specialistson SoutheastAsia: the late PierreDupont, and ProfessorsA. B.Griswold and Oliver W.Wolters. Without Dupont's studies of Pre-Angkorianand Isthmian sculpture the present work would have been impossible. Professor Griswold, whose studies are basic to our knowledge of Buddhist sculpturein PeninsularSiam, has contributedmuch time, encouragement and helpful criticism.ProfessorWolters' studies of the maritimehistory of SoutheastAsia are fundamentaland formativein my conceptions of developmentsin PeninsularSiam. Finally, I should record my intellectualdebt to my Cornellcolleague, ProfessorMartieW. Young. Ithaca,N.Y. StanYn lJ. O'Connor, Jr.

JuneI97I

CHAPTER I

A BRIDGE

AND A BARRIER

Peninsular Thailandbegins just south of I z 50' north latitude and extends in a crescentfor
about 6oo miles to the present border of Malaysia.This is the narrow tract, the isthmus, that joins the MalayanPeninsulato the continentalmass of Asia. In some places the land between the Bay of Bengal and the Gulf of Thailandis only zo miles wide; and even at its greatest width the isthmus is only a mere I 35 miles wide.' Down the spine of the Peninsulahere mountains fold into a series of ranges that march generallynorth-south. Seen from the air, they are intensely green; the tropical rain forest spreadsits thick moist canopy over the rock. The thin margins of the coast are punctuatedby plains of alluvial soil washed down from the hills by rivers that strike longitudinallyacross the Peninsula. The plains of the east coast are broader than on the west and the coast line is much more regular. Along portions of the west coast, jungle-cladhills come down to the water's edge where the tigers and wild pig of the forest hear the roar of the surf. But much of that coast is marked by mangrove formation. This is a mix of mud and salt water held in suspension by vegetable matter. Only a few specializedorganisms can survive in these melancholyzones: chief among which stand on their stilt roots at the outermost them are the mangrove trees (rhiZophara) margins of the swamp.2 Behind them come the other bushes and trees of this strand comand most useful in a variety of ways to munity, of which the nipa palm is both characteristic coastal peoples.3 While the west coast is indented with bays, cut by mangrove swamps and studded with off-shoreislands, the east coast has been smoothed and scoured by sea and wind. For much of the distance between the Band&nBight and Songkla, the coastline is a dazzling, almost unbroken, beach. Borderingthe beach there is a thin line of beachforest where the filmy casuarina trees bend in the wind. And furtherlandwardthere are the diked padi fields that spreadalong region, on the strip of land between the Tale Sap lake and the plains in the Songkla-Pattalung the sea, and along the extensive plains of Nagara Sri Dharmaraja(Nakhon Si Tammarat), and Band&n Chumpon.4Other clustersof padi fields spread up the valley floors some distance inland and smallerpockets of rice land are scatteredalong the west coast.
I 2

and Robert L. Pendleton, Thailand:Aspects of Landscape Life (New York: i962), pp. 46-47. For a discussion of the major types of strand landscapesalong the Malaysiancoasts see David E. Sopher, TheSea Nomads (Memoirs of the National Museum, 5, i965, Singapore), pp. 3-I9. The many uses of the products of the swamp including nipa palm are discussed fully by Tom Harrisson in TheMalays SarawakBeforeMalaysia(London: I970), pp. 272-273. of Southwest Siam, Technical Bulletin no. 5, Department of Pendleton, Thailand..., p. I 52. See also his SoilsandLand Use in Peninsular Agriculture (Bangkok: I949).
I I

Life in this narrow portion of the earth is persistentlygenial, with enough fish, rice, fruit, water and a relatively unvarying climate.sThese attractionswere sufficientto induce men to settle in this region, and by the beginning of the Christianera severalof the small estuaryplains were, according to inferencesdrawn from Chinese records, the sites of developing city states. In the rise of civilization in the isthmian tract, two factors played an important role: mineral wealth, and a strategiclocation on the international sealanesjoining China,India and the Persian Gulf in one great web of exchange. When Ralph Fitch sailed in I599 from Pegu to Malacca,he passed Tavoy Island "from whence cometh great store of tinne, which serveth all India."6 Below Tavoy in the MerguiTenasserimareathere are the remainsof ancient tin-workings, several of them on a scale that suggests the activity of thousandsof men.7These depositsarepartof the bandof tin that extends from Bangka and Billiton Islands up through the entire Peninsula.In the isthmiantract, tin is in concentrated the graniticrangesof the west coast which run from Ranong to Puc (Phuket) Get Island and up the flanksof the TakuapdRiver. There is also an importantgraniticrange on the eastern side of the isthmus with tin deposits occurring at Si Chon which is just south of the Bight of Bandon.Other significanttin ore deposits arefound near Songklaand againat Pattnii.8 The lure of tin in ancient times was great, and although the date when tin-working began on the Peninsulais not known, one author has stated that alreadyby the 5th centuryA. D. tin in Malayawas being mined and exportedby Indians.sBy the 9th century,the metal was known to the sailing mastersand crews thronging Sirafand other bustling ports of the PersianGulf.lo There is good reasonto believe that some of this tin was exportedfrom one of the ports on the west coast of the isthmus. The fame of the rich deposits in the isthmian tract would probably have circulatedabout maritimeSoutheastAsia long before written recordsor before it came to the notice of the courts of India or the trading centers of the PersianGulf. It is instructiveto learn, for example, that tin ores were apparentlybeing traded over the open sea from the PhilippineIslands to the workshops of Fu-nanin the Mekong Delta by the 3rd centuryA.D." The other factor which appearsto have favored the rise of complex economic and social units on the isthmus is its strategiclocation. A glance at a map of Asia will show that the Malay Peninsulalies astridethe open sea lanes linking Chinawith India, the Middle East and beyond. The Peninsulais an inescapablefact in the strategic calculationof Chinaand of all the maritime kingdoms of SoutheastAsia.I2 The isthmus offeredthe possibility of a trans-peninsular portage,
5 The daily life of a Malay community living in a coastal village in PattAnI Province has been fully published in Thomas M. Fraser, Rusembilan: Malay Fishing Villagein Southern A Thailand (Ithaca: i960). 6 Ralph Fitch, "The Voyage of M. Ralph Fitch ... to Pegu, to Lamhay in the kingdome of Siam, and back to Pegu and from thence to Malacca, Zeilan, Cochin and all the coast of the East India," Hakluyt's PrincipalNavigations,I 599, vol.II, pp.250-268, citedin H. L.Chhibber, Mineral The ResourcesBurma of (London,I934), p. I 8i. 7 Capt.G. B. Tremenheere, "Report on the Tin of the Province of Mergui," Journalof the Asiatic Society Bengal,vol. io of
(i84I), 8 p.845.

Pendleton, Thailand..., pp.236-237. 9 Wong Liu Ken, "The Malayan Tin Industry" in K. B. Tregonning (ed.), Paperson MalayanHistory (Singapore, i962),
p. I0.

10

" 12

Arab literatureon trade with the Peninsula Asia is summarizedin G. R. Tibbetts, "The Malay Peninsula as known to the Arab Geographers," The MalayanJournalof TropicalGeography, vol.9 (1956), pp.2i-60. See also A. Lamb, "A Visit to Siraf,"Journalof the Malaysian Branchof the RoyalAsiatic Society (JMBRAS), vol. XXXVII, i ( 946), pp. I-I 9. R. A. Stein, "Le Lin-yi," Han-Hieu, vol.2 (1947), pp.I20-I22. The characterof diplomatic and commercial relations between China and maritime Southeast Asia is examined in 0. W. Wolters, "China irredenta: the south," WorldToday,vol. 19, I2 (i963), pp. 540-5 52.
I2

a short cut that would save a thousand miles sailing through unpredictablewaters where each for river headand mangroveswampwas a potentialsanctuary pirateships. Rivers strikinglongitudinallyacrossthe isthmus could be followed up to the low watershedand then, with a portage of a few miles by foot, goods could be loaded by river to the opposite coast. The Tenasserim valley has long served as a convenienttranspeninsular route, and MauriceCollis, when he served as British residentat Mergui was able to find many traces of ancient objects, including Chinese pottery dating from the Sung dynastyand 1 5o bronze images, that were the residueof the brisk traffic along this passage.'3At the Thailand-Burmaborder, where the isthmus reaches its narrowestpoint, interest in developing a canal system linking the Bay of Bengal and the Gulf of Siam has been a topic of interest in modern times.14 Perhapsthe most frequentlymentionedroute leads from Takuapaon the west coast by river and portage acrossto the Bay of Bindon. The discoveryof many antiquitiesat both ends of this route suggests that this was an importanttrade route in ancient times.is Other possible routes with the West coast at Kantan,and follow out the Trang River linking Nagara Sri Dharmardja the well-known river passagesfrom Kedah to Pattdn.I6 The commercialimportanceof the isthmus was reinforcedby the periodic alternationof the monsoon wind system which set the rhythm for sea movement in Asian waters prior to the advent of steam. From the ancientports of India, such as Barygazain Gujaratand the Gangetic from the ports in Andhradesasuch as Alosygni, and from the great trading port of Thmralipti, emporium of Ceylon, Mahdtittha,men set course for the Malay Peninsula and sailed eastard on the favorablewinds of the southwest monsoon. Other ships, plying the Nanhaitrade from southernChina, reachedthe Peninsulaon the northeastmonsoon, and the pattern of trade set by the periodicityof the winds offeredthe possibility of developing relay stationsin the web of east-west trade. Ships coming on the monsoon could off-loadtheir cargoes and pick up luxury goods brought from distant lands, perhapshauled by portage across the neck of the isthmus. Sailing vessels, waiting in shelteredharborsfor the change in the wind systems before starting their returnjourney,needed victualing and refitting; warehouseswere necessaryfor storage of goods to be transshipped.As Singaporeis now, and Malaccawas formerly,so earlierentrepots sprang up along the Peninsulato facilitatea commercethat stretchesinto pre-history. becausethe wake of a vessel leaves no scaron the surface Becausethe sea is truly "trackless", of the water, the early voyaging of the coastal peoples of SoutheastAsia is little known to us. involved in the spreadof pottery types throughIntellectuallywe may recognizethe seamanship out the coasts of SoutheastAsia and the islands of Oceania,or the shipmentof IndonesiancmI'

Maurice Collis, Into HiddenBurma(London, I95 3), pp.2i8-233. F. G. Tremenheere, "Report of a visit to the Pakchan River and of some tin localities in the southern portion of the Tenasserim Provinces," Journalof the Asiatic Societyof Bengal,vol. XII, 2 (I 843), pp. 523-5 34, A. J. Loftus, Notes on a JourneyAcross the Isthmusof Kra (London, i883). A. Kerr, "Notes on a trip from Prachuap (Kaw Lak) to Mergui," Journalof Siam Society, vol. XXVI, 2 (I 933), pp. 203-2I4, and E. Hutchinson, "Journey of Mgr. Lambert, Bishop of Beritus, from Tenasserim to Siam in i662," Journalof the Siam Society,vol.XXVI,2 0933), pp.z2xI5-2zi8. 5 F. H. Giles, "Remarkson Land Routes across the Malay Peninsula,"Journal theSiamSociety, of vol. XXVIII, i (I 935), and H. G. Q. Wales, "A newly explored route of ancient Indian cultural expansion," IndianArt andLetters,vol. IX, i (I 93 5),
pp. I-3 5.

There are several accounts of the Kedah-Pattdn!journey: A.W.Hamilton, "The Old Kedah-Pattani Trade Route," Journalof the Straits Branchof the Royal Asiatic Society,vol.86 (1922), pp.386-392, and W.E.Maxwell, "A Journey on Foot to the Patani Frontier in I 876," Journalof theStraits Branch the RoyalAsiatic Society, (i 882), pp. I-67. See W. G. of 9 Solheim, "Pottery and the Malayo-Polynesians," Current Anthropology, V, 5 (Dec. I 964). vol.

I3

We namon to East Africa and thence to Rome in the time of Pliny.17 may follow the circulation of a common fund of materialculture as it spreadsby sea around the welcoming coasts of the South China Sea.18It is possible to read of the yearly voyages of Bugis vessels from Makassar to the tiny Aru Islands off the coast of New Guinea, and to relive that journeyto ultimaThule Even today one can see those same Bugis ships with the great scientistAlfred Russel Wallace.19 coming into Singaporeroads. But it is difficultto seize in the imaginationthe implicationof all this voyaging as it is reflectedin the openness to the sea, the capacioushorizons and the cosmopolitanism of the coastal peoples of ancient SoutheastAsia. In order to capturesome perception,however dim and refracted,of the lived world of the peoples of the isthmiantract, it is essentialto see it set in a streamof humanmovement by land and sea. Few areasin SoutheastAsia are as impactedwith the press of history. Perhapsthe first historical referenceto the isthmus is to be found in a Chinese record of a mission which the Han emperorWu Ti (140-87 B.C.) apparentlysent to India, or at least some Rather than risking the sea passage through the locality on the shore of the Bay of Bengal.2o the narrow waist of the Peninsula.Whateverits diploStraitsof Malacca,the mission crossed matic or strategic objectives were, it was also charged with a searchfor pearls, glass and rare stones which were obtainedin exchangefor gold and assortedsilks. The Chinesewantedluxury goods from India and Rome, as well as the "strange and precious" objects of the teeming tropicalrainforest. Camphor,rhinoceroshorn, the diseasedand fragrantgharuwood, aromatic laka wood and beeswax, all abundanton the isthmus, found a marketin China.2I In responseto the quickeningimpulses coursing along an extendednetwork of international trade, certain favored villages on the Peninsulabegan to develop into city-statesin the early centuries of the Christianera. With the growth of population centers, and shifts in the complexity of social roles and economic exchange,new legitimizingmyths, legal norms and religious sanctionswere accepted.Many of these new impacts on the intellectuallife of the isthmus came from India as SoutheastAsian rulersbegan to bring Indian scribesinto their service.22 Humble villages, little more than clustersof cultivators,fishermenand pirates,came to international,or

The consequences of these Indonesian-East African contacts are reflected in language, boat-types, rice cultivation and the music. See A. M. Jones, Africa and Indonesia: Evidence the Xylophone of (Leiden, i964). In a recent contribution to the problem, J. Innes Miller examines the botanical and textual evidence to demonstrate the existence of the cinnamon trade in Roman times: TheSpice Tradeof the RomanEmpire (Oxford, i969). 18 L. Malleret, L'archbeologiedeltadu Mdkong du (ADM), vol.III, La Culturedu Fou-nan(Paris, i962), pp.41I-408. 19 TheMalay Archipelago (New York: i962), pp. 308-375. 20 The relevant passage is translated and analyzed in Wang Gungwu, "The Nanhai Trade: A Study of the Early History of Chinese Trade in the South China Sea," Journalof theMalaysian Branch the RoyalAsiatic Society, of vol. XXXI, 2 (I958),
pp. I9-22.
2!

17

22

For a survey of commodities involved in the early maritime trade of Southeast Asia, see Paul Wheatley, "Geographical Notes on Some Commodities Involved in Sung Maritime Trade," JMBRAS, vol. XXXII, 2 (I959). This is, of course, the process of "Indianization" brilliantly discussed by J.C. van Leur, Indonesian Tradeand Society (The Hague, 2nd ed., i967), pp. 107-I I0. See also A. Christie, "The political use of imported religion, an historical exde des ample from Java," Archives sociologie religions, 9, I 7 (i 964), pp. 53-62. For a study of this process of social change vol. on the Malay Peninsula, see Paul Wheatley, "Desultory Remarks on the Ancient History of the Malay Peninsula," in J. Bastin and R. Roolvink (eds.), Malayanand Indonesian Studies:Essays Presented Sir RichardWinstedt to (Oxford, i964),
pp.41-42.

'4

at least Chinese, notice as kingdoms such as Tun Sun and Langkasuka.23 strategicrole of The these little states was of sufficientimportanceto early internationalcommercefor Fu-nan, the earliestmajor maritimepower in SoutheastAsia, to find itself obliged to extend its hegemony over the isthmus in the 3rd century A.D. Even after the fifth century when the textual evidence indicates that at least part, possibly the bulk, of internationalcommercebetween India and Chinano longer utilizedthe long portage but was directlysea-bornethrough the Straitsof Malacca,the arearetainedsome of its entrep t functions and its importanceas a supplierof tin and gharuwood. Certainlythe small city-states of the area experienceda period of prosperityduring the sixth and seventh centuriesafter Funan's control had ended. This is clearlyindicatedby the brisk pace of tribute missions sent to China by states of the isthmus during the period.24 Such missions, whatever else their function may have been, were a mechanismin facilitatingcommerce.2s With the consolidation of internationaltrade attendanton the prosperityof a strong China underthe T'ang dynasty,the isthmusbecameonce againa focal areain the geo-politics of Southeast Asia. By the last quarterof the seventh century,the dominantmaritimepower of the period, the Sumatran-based empire of Srivijaya,had extended its control over the Hindu-Buddhist settlementsstrung out along the Merbok Estuaryof Kedah. The Ligor inscriptionreveals that the northern reaches of the isthmus, probably as far north as the Bay of Bandon, had passed under Srivijayancontrol by A.D. 775. The degree of Srivijayancontrol, the fluctuationof its power, and its culturalimpact on the areaare open questions; but Srivijayahas given its name to an art style which is associatedwith many objects found in isthmian sites. The Indianshad manifesteda commercialinterestin the areaduringPallavatimes and before. An Indian mercantile company, the Ma~utgrdma, was active at Takuapl, an early entrepot located on the west coast of the isthmus above PuiGet, during the ninth century.26 The successors to the Pallava,the Cola, reactingagainstthe restrictivetradepolicies and impositions of Srivijaya,raided the isthmian tract possibly in IOI7 and also in IO25 and io68.27 Among the materialevidence of the Co1aimpact in the areaare statuesat Takuapi and Vieng Sra (written Viaft Srah and pronounced Wieng Sa) in C6la style, and an inscription at Nagara Sri Dharmarija.28 Burmesekings turnedtheir attentionto the isthmus as earlyas the reign of King Aniruddha in the mid-eleventhcentury.29 And it was control over the commercialroutes of the area that brieflyembitteredrelationsbetween Burmaand Ceylon, leading to a war in i I65-66.30 Indeed,
23

24

25

26

27
28 29

30

On the basis of the textual evidence, it is believed that Tun-sun was located in the northern reaches of the peninsula ratherthan on the isthmian tract. For an analysisof the problem, see Paul Wheatley, TheGolden Khersonese (Kuala Lumpur, i96i), pp. I5-2I. It is believed that Langkasuka was located on the isthmian portion of the peninsula. See ibid., pp.67 and 252-267. Gungwu, "Nanhai," Appendix A, pp. I I 8-i23, lists the tribute missions from Southeast Asia to China up to the Sung dynasty. The characterof diplomatic and commercial relations between China and Southeast Asia is examined in 0. W. Wolters, i "China irredenta: south," World the Today,9, i z (I 963),pp. 540-5 52. K. A. Nilakanta Sastri, "Takuapd and its Tamil Inscription," JMBRAS, vol. XXII, I (I949), pp. 25-30. R.C.Majumdar, "The Overseas Expeditions of King RdjendraCola," Artibus Asiae, vol.XXIV, Pts.3/4, pp.338-342. G. Ccedes, Recueil inscriptions Siam, deuxieme des du partie, znd ed. (Bangkok), p. 38. G. H. Luce, "Some Old References to the South of Burma and Ceylon," in FelicitationVolumes Southeast-Asian of Studies Presented PrinceDhaninivat,II (Bangkok, i965), p.270. to Sirima Wickremasinghe, "Ceylon's Relations with Southeast Asia with Special Reference to Burma," The Ceylon Journal of HistoricalandSocialStudies,vol. III,I (I 960), pp. 44-49.

'5

this incidentis merelya punctuationmarkin the long continuumof contactsbetweenthe isthmian area and Ceylon. Professor A.B. Griswold has recently identified Buddha images found at Sungai Golok, in the province of Naridhivisa near the Malayanborder, and at Prinapuriprovince furthernorth, as productsof Ceylondating to about the fifth century;and Ceyloncultivated during especiallyclose contacts, religious and political, with the isthmian state of Tambralifiga the eleventh through the thirteenthcenturies.3I In the rich mix of competing influencesto which the isthmian tract was subjectin ancient times, the Khmer empireplayed a prominentrole. Fu-nanhad left its impact on the area;even after its collapse, culturalexchangesappearto have persistedaround the shores of the Gulf of At Siam, involving the kingdoms of Chen-laand the city-statesof the isthmus.32 the beginning of the eleventhcentury,a pretenderto the throne of Angkoritself appearsto havebeen a member of the ruling family of a kingdom at Ligor (NagaraSri Dharmar-ja),an indicationthat the elite in the areamoved easilyin the culturaland politicalambienceof the Khmerworld at that time.33 the Following his victory over Jayaviravarman, latter now consideredto be the usurperwith the isthmian background, StryavarmanI extended Khmer power over the Menam basin as attestedby an inscriptionat Lopburi of 3022-245 .3 In the view of L. P. Briggs, Khmer power at Recent researchon an inscriptionfrom Jaiy! on the this time extended as far as the isthmus.3s Bay of Bindon indicatesthat Khmer culturalinfluencewas still strongly felt in the isthmus as late as the last three decades of the thirteenthcentury.36 It would be necessaryto add to the flux of these contacts: the presenceof Arab sailorsfrom the PersianGulf; the unrecordedflow of shipping in small craft from ancient ports of Burma, Sumatraand Java; and the passage of Buddhist monks en route from Chinato India to collect texts and to make pilgrimagesto the sites in northeasternIndia made holy by associationwith the historical Buddha.37 This is merelyto block out some of the majordevelopmentsthat markedthe isthmusbefore the Thai imposed their control over the tract in the late I3th century. It should be possible to trace the shape of these events through the recovery and analysis of artifactslost, strayed or to buried in the red earthof the isthmus. As event piled on event, merging imperceptibly form a past, so the great ash-heapof history, the materialresidue of action, should form in strataaccessible to the archaeologist. Unfortunately,it is impossible to point to any systematic,wellpublished, archaeologicalexcavationof an historicalsite on the isthmiantract. Nearly all the sculptures of bronze or stone found on the isthmus are "floating" objects, random disgorgements of the past, most of them discovered by accident. They are without proof that historicalcontext in the sense of absolutecalendardates or the chain of documentary links artisan,patron and object. There are no archivesto consult, almost none of the sculptures
31 32 33 34 35 36 37 pp.41-55;

"Imported Images and the Nature of Copying in the Art of Siam," in Essays Offeredto G. H. Luce (Ascona, i966), II, and Senarat Paranavitana,Ceylon Malaysia(Colombo, i966), pp.74-8i. S. J.O'Connor, "Satiiprah: An expanded Chronology," JMBRAS, XXXIX, (July, i966), pp. I 37-I44. du See G.Cced~s, "Stele de Prasft Ben (K. 989)," Inscriptions Cambodge, VII (Paris, I 964), pp. I 64-I 89. vol. d'Indohine et d'Indonesie, ed. (Paris, i964), pp.252-253. 2nd G.Ccedes, Les Etats hindouises L. P. Briggs, "The Khmer Empire and the Malay Peninsula," Far EasternQuarterly,vol.IX, 3 (I950), p.286. J.G. de Casparis,"The Date of the Grahi Buddha," Journalof theSiam Society,vol. LV, I (i967), pp. 30-40. of J.Legge, A Recordof Buddhistic Kingdoms (Oxford, i886); H. A. Giles, The Travels Fa-Hsien (London, I956); J.Takain as kusu, A Record the BuddhistReligion Practiced Indiaandthe Malay Peninsula of (A.D., 67i-695) by I-Tsing (Oxford, Translations the BuddhistTripitaka(Oxford, i883). i896); B.Nanjio, A Catalogue the Chinese of of

i6

themselves bear inscriptions,and those inscriptionsthat do exist on the Peninsulado not refer specificallyto invidual works of art.38 If then these objects are undocumentedthey are equallyunfixed in any archaeologicalcontext. Not one of them was uncovered as a result of controlled excavation.Not one of them was found in any direct associationwith a dated monument. In the first sustainedstudy of a group of isthmian sculptures,the late PierreDupont noted that the examplesof sculptureon the isthmiantract displayeda great variety of styles and that each style was often representedby only one or two examples. Since most of them could be integrated,in his opinion, into known styles in India or Indonesia,he concludedthat it could be It assumed they were imported.39 is now possible to qualify Dupont's conclusion as the result of a study made by Professor A. B. Griswold on the problem of discriminatinglocal copies of Buddhistsculpturein SoutheastAsia from those images that have actuallybeen importedto the area from India. He has demonstratedthat the small Indianizing states in the Peninsula had developed workshops and that images were being made in the areabefore the 6th centurythat are sufficiently distinctive in the inflection of their dress and modelling for them to be distinguished from their Indian prototypes.40 His conclusions are importantin two ways: first, becausethey indicate that art actuallydid have a history of its own on the isthmus, and this would accord well with the culturallevel in located in the area;and the areaas deducedfrom the textualreferencesto city statespresumably second, because the early date he ascribes to this artistic activity, some time before the 6th century, may give us an approximatedate for the beginning for Buddhist art on the isthmus. As a contributionto the emergingpictureof culturallife on the isthmus, our study will focus on problemsof style and dating for the Hindu sculpturesfound there. These are objects with a common religious purpose, occupying a spatiallycompact area. There are actuallyquite a few of them when they are all grouped together, and it is a virtual certaintythat only a portion of the extant pieces are known to me. Many must be privatelyowned in the towns and villages of the isthmus, or kept in Buddhist monasteriesthere, or else, splashed with gold leaf, honored with candlesand joss sticks in dimly lit caves high in the limestonebuttes that rise like chimneys from the plains of the isthmus. But there is a significantnumber of Hindu sculpturesavailable to us now, which, taken together, indicate that monumentalstone sculpture must have been produced in abundancein the isthmus in ancient times, that an archaeologicalstudy of the materialpast of these coastalstateswould be rewarding,and that excavationwould provide the beginning of some congruence between the literaryrecordnow availableand the testimony of things.4I
It is true that the inscription dated 775 A.D., found at Nagara friDharmarfja, commemorates the foundation of three monuments; but typically, it offers little information useful for an understanding of the style, construction, or even the proximate location of the monuments. The inscription has been edited and translated by George Ccdes, Recueildes du inscriptions Siam, Part II (Bangkok, 2nd ed., n. d.), No. XXIII, p. 20. 39 Pierre Dupont, "Variet~s archeologiques, II: Le Buddha de Grahi et l'cole de Chaiya," Bulletinde l'Ecole Franfaise
38

d'Extrime-Orient (BEFEG),
40

vol. XLII

(I942),

pp. I05-I

I 3.

41

A. B. Griswold, "Imported Images and the Nature of Copying in the Art of Siam," in Essays Offered G. H. Luce to (Ascona, I 966), II, p. 57. Chinese records provide the earliest and most reliable evidence for the development of the Malay Peninsula, and they are supplemented by early Western geographies and Indian literary allusions; later Arab sailing records too afford some precision on the information supplied by Chinese pilgrims, historians, commercial attaches and diplomats. There is also

'7

a small corpus of epigraphy available; and amidst the fictions of local chronicles there are historical truths. As the study of these materialsby historical geographers has been under way for at least a century, there is now available an almost staggering body of literature, of which one of the principal aims is to anchor the texts to the topographic reality of the Peninsula. Professor Paul Wheatley has recently brought some measure of order to the subject by collecting, collating and analyzing these materialstogether with the work of the commentators: TheGolden Khersonese (Kuala Lumpur, i96i). This body of material has recently been surveyed and correlated with commercial and pharmaceuticalrecords in a magisterial study that has given economic substance to the study of early Southeast Asian history: 0. W. Wolters, Commerce Early Indonesian (Ithaca, i967).

i8

CHAPTER II

CONCH SHELL ON HIP: THE 'ABERRANT' STATUES OF VISNU

hree rathersmall stone images of Visxnu (figs. I, 2, 3) found on the isthmus share certain singularities of dress and iconography that immediatelyset them apart from the other sculpturesof the god found in that region. They all project a quality of bristling astringency. While one of them is on display in the National Museum in Bangkok, the other two are propped up casuallyin the dim clutter of a treasureroom at Wat Mahadhatuin the isthmian town of Nagara Sri Dharmaraja.Only the first has received any written notice, and the date generallyattributedto it appearsto be several hundred years in error. Vi~srngfromJagyd Jaiydis a large village in the northernmostreachesof the isthmian section of the Malay Peninsula. It is on the route of the SouthernRailway, but the expresstrain stops only briefly and the railwayhere is used mostly by farmersfrom neighboring villages bringing their produce to the small marketat Jaiya. Their arrivalon the local train in the morning and departurebefore noon is a rhythmicflurrythat punctuatesthe otherwise placid calm of the station-master's day. But though Jaiyd today is a quiet provincial town, there is an abundanceof archaeological evidence to indicate that in the past it was an extensive, wealthy and cosmopolitan city. The collections of the National Museum in Bangkok have been enrichedby many bronze and stone statues found in and around the ancient structuresstill standing at Jaiyd. It was at Jaiyd,near the Monasteryof the GreatRelic (Mahadhdtu), PrinceDamrong, the founder of that Siamese archaeology,found the bronze bust of the Bodhisattva Lokesvarathat is one of the great treasuresof the museum and a leading representativeof the "Art of Srivijaya".'There areworks in stone that clearlypredatethis bronzeby some centuries;and despitethe vicissitudes of history, the erosion of the power of the Empire of Srivijaya the coming of the Siamesein and the i3th and I4th centuries, Jaiyd produced sculpturesfrom the I5th to the i9th that reflect local preoccupationsand traditions,managingto maintaina degree of artisticautonomy in the face of widespreadacceptanceof the canons of the National School of Ayudhyd.2 Jaiyd, because of the reputationof its antiquities,has been visited by a number of archaeologists during the past sixty years.For the most part, however, their visits were brief and, with the exception of severaltrial trenchesmade by J. Y. Claeysand QuaritchWales, there have been
I George Ccdes, Les collections archeologiques Musie National de Bangkok(Ars Asiatica, vol. XII, Paris: du
i928),

plates

Xv-XVI.
2

A. B. Griswold, oral communication.

'9

Most of the monumentsnow standinghave been recorded, no controlledexcavationsat Jaiyd.3 and Claeys has published excellentplans of several.These buildings considered as a group are remarkablefor their eclectic architecture.The monument of Wat Mahidhdtu, despite some accretionsdating from the period of Siameseoccupation,has been relativelywell preservedand the original characterof the building is still readable.It is strikinglydifferentfrom any Siamese monument,and it has been likenedto constructionsseen in bas-reliefon the Borobodurin Java.4 Wat Keu (writtenKev and pronouncedGeo), althoughin ruins,has two fairlywell preserved faces, the south and the north. It is obviously relatedto early Chamarchitecture,and further, according to Ccedes,its plan is analogous to that of Chandi Kalasan in Java, while in his opinion it also bears some resemblanceto archaicpre-Angkorianarchitecture.He would not date any of the structuresat Jaiydprior to the eighth century.s Dr. QuaritchWales has envisioned Jaiydas a center of diffusionfor Indian culture, and in his view the architectureof Wat Keu reflectsa primitive non-specializedIndian colonial architecture. In this sense it would be a survival of a common ancestraltype from which the early Cham, Javanese, and pre-Angkorianarchitectureis derived.6He argues that, if the influence were actuallycoming to Jaiyafrom Java, one would expect to find the typicallyJavanesekalaarch and tracesof Indo-Javanesearchitectural ornament.The force of Dr. Wales'sargumakara ment is somewhat diminished,however, when, after acknowledging that the bronze Mahdyand Buddhist sculpturesfound at Jaiyd and elsewhere on the Peninsula are significantlydifferent from their alleged Pdlaprototypes from the famous monasterycomplex at Ndlanddin Bengal, he himself arguesthat the lack of a numberof specializedPdlafeaturesis not sufficientevidence to deny the primaryrole of Pgla influencein shaping this aspect of Hindu-Javaneseart.7 The vestiges of ancient monuments, which are apparentto the most casual observer, bear testimonyto the importanceof Jaiydas a centerof Indianizedsettlementin earlytimes;8 though whether or not it played a prominentrole in the diffusionof Indian cultureis not, at this point, fruitfulquestionwhen our knowledge of the site remainsso meagre.At least it is a particularly encouragingto note that thereis a very active body of local scholarswho have takenan intensive interestin the history of the area. Severalof them have publishedmaterialon aspects of Jaiyd's past which is of special interest becauseit contains referencesto local traditions.s One of the Brahmanical images found at Jaiydis a standing figure carvedin a greyish limestone (fig. i). It is now in the National Museum in Bangkok, and according to the museum label, its provenanceis Wat Sala Tung at Jaiyd.It is a four-armedfigure of Visnu in sthdnakamzirti,27 inches in height. The posterior left arm is missing, the anterior left hand holds a on conch-shell (Jawkha) the hip, the posterior right hand holds a heavy club (gadd), and the
vol.XXXI (I93I), pp.386-387. H.G.Quaritch Wales, "A Newly Explored Route of Indian Cultural Expansion," IndianArt and Letters,vol.IX, no. i, pp. I9-22. 4 H.Parmentier, "Origine commune des architectures hindoues dans l'Inde et en Extreme-Orient," Etudes Asiatiques, vol.11, p. 2I. 5George Ccdeds,"A propos d'une nouvelle theorie sur le site de Srivijaya,"JMBRAS, vol. XIV, Pt.III, Dec., I963. 6JAL, vol.X, no.i, p.26; and, TheMakingof GreaterIndia (2d ed. rev.; London: I96I), pp. 50-5I. 7 Wales, Makingof Greater India, pp. 54-5 5. See also A. J. Bernet Kempers, TheBronzesof NalanddandHindu-Japanese Art (Amsterdam: I933). 8 Erik Seidenfaden, "Recent Archaeological Researches in Siam," JSS, vol.III, Pt. II (April I93 8), pp. 24I-247. 9 Brah GariuIndapafifnicariya, Brief Accountof the AntiquitiesSurrounding Bay of Bdndon(Jaiya: I950; in Thai). the A in Dhammadasa Banij, A History of Buddhism Jaiyd andthe Bay of BandonRegion(Jaiya: I96I; in Thai and English).
20

3 J.Y.Claeys, "L'archeologie du Siam," BEFEO,

anteriorright hand appearsto be in abhayamudrd. the statuerepresentsVisnu, the attributein As the missing hand must have been eitherthe lotus (padma),the round symbol of the earth (bhzi), or the discus (cakra). with an elaborate The figure wears a tall mitre,decoratedin bas-relief patternof leaf andvine, with rosettes at the corners.The face is round and relativelysmall; the almond-shaped eyes are sharplyincised, flatand without any interiordefinitionof pupil or structure.They are extremely long and spreadto the outer limit of the face. The ears are constructedin a remarkable fashion, presentedin frontal view, and joined to the mitre to secure them against fracture.In addition the sculptorhas greatly exaggeratedtheir size, apparentlyin order to include the interior detail of the ear. From these large ears, and in the same charged scale, hang heavy rectangularearrings (kuwdala). The torso is nude except for a flat torque, with an ornamentof opposed S-shapedelements at the center. This takes on a configurationreminiscentof the srialsa mark,a symbol of beauty and fortune often associated with Visnu. In addition to the torque, the image wears simple armletsand bracelets. The costume is a dhoti,a large rectangleof untailoredcloth which can be drapedin so many different ways that it is difficultto analyzethis particulardispositionwith certainty.The same is true of the kamarband waist-sashwhich holds the dbotiin place, and the broad sash which is or looped through it. As well as we can make out, the patternof circles and ellipses at the waist is the decorated upper edge of the dhofi; the lower edge encircles the legs between knees and ankles; and the heavy verticalfold of cloth that falls down in front between the legs is a pleated lateraledge of the dhoti(the other lateraledge is either combined with it, or more likely should be understoodas forming a similarverticalfold in the rear).The more or less horizontalcreases or pleats in the dhotiacross the legs are renderedby means of incised lines. A narrow sash or encirclesthe waist twice; and as its two ends can be seen hanging down over the left kamarband thigh for a short distancewe assumethat it is distinctfrom the broad sash which is held in place by it. This broad sash appearsto begin at the base of the statue near the left foot; though it is now broken, it evidently rose from there to the left hip, where it bulges out before passing under (?) the kamarband; it falls in a semicircular in front of the thighs, passesunderthe then arc at kamarband the right hip, re-emergesto form a puffed-out loop at the right hip, continues and (acrossthe back?) to the left hip, where it may have been tucked under the kamarband, then presumablydescendedagainto the base of the statuein the samemassof cloth in which it began. This mass of cloth formed one of the three supportsof the statue.On the right side the now broken club (gadd) filled the same function. The third support is the main block of stone in which the legs and the heavy vertical fold of the dhotibetween them are carved. Visnuat NagaraSri Dharmarajia Roughly sixty miles southeast of Jaiya is the coastal city of NagaraSri Dharmar-ja.It is located some five miles inland from the sea, although its location in an area in which the beaches are building up would suggest that it was a good deal closer to the sea severalthousandyearsago.Io
IO Major P. D. R. Williams-Hunt, "An

Introduction to the Study of Archaeology from the Air," Journalof the Siam Society
p.i o8, pl.9.

(JSS), vol.

XXXVII, pt. 2

(June

I949),

2I

From the town, the horizon to the west is closed by high wooded mountainswhich protect the coastalplain from the force of the southwest monsoon. According to Claeys,"1 of the sumone mits reachesa height of almost 6,ooo feet. It is altogether a pleasant situation for a town and there is a fairly extensive area of plain availablefor the cultivation of rice. The town is laid out on a north-southaxis. What appearsto be the oldest part of the city lies within a walled enclosure. Situatedalong a road which runs down the middle of this enclosed section of the town are a great many religious structures,many of them in ruins. During his visit to Nagara Sri Dharmardja, Lajonquiereestimatedthat there were fifty monasteriesin the town, most of them at that time in ruins.I2It is not surprisingthat there are many religious It edificesto be seen in NagaraSriDharmardja. was a leadingcenterof Theravdda Buddhismfrom which, as we deduce from Ram Kamheng'sinscriptionof I292,I3 the Thai kingdom of Sukhodayareceivedits initialindirectcontactswith the fountainheadof the Theravdda, Ceylon. Prince Damrong speculatesthat monks from Ceylon made an appearancein Nagara Sri Dharmaraja some twenty years before the date of Ram Kamheng'sinscription,and that subsequentlysome Siamesemonks went to Ceylon for ordination.Upon their return,they set up headquarters in Nagara Sri Dharmardja and, with the help of Sinhalesemonks alreadyresidentthere, undertook the architecturalmodificationof the old monument of the Great Relic to a Ceylonese style.14 This is the famous Wat Mahddhdtu. There is some other evidence, though rather tenuous, which suggests that Nagara Sri Dharmardja may have had connectionswith the southernIndian Buddhistcenterof Negapatam as well as Ceylon. The P?li Buddhistliteratureof Ceylonofferssome referencesto the monastery complex at Negapatamand to the state of religion in south India. One of the religious reformers of south India, a monk named Buddhappiya,was either a native of Tambarattha (Nagara Sri or had spent some time there. This informationis containedin an inscriptionfound Dharmardja) at Polonndruva,which may date from the first half of the twelfth century.I5 It would not be surprisingfor Nagara Sri Dharmardja have links with Negapatam,which to was one of the last Buddhist centers surviving in India. Since the earliesttimes Negapatamhad been one of the chief Indianports for tradewith China.The Srivijayan empire,of which Nagara Sri Dharmaraja a part at least by 775 and with which it retainedcontact until the beginning was of the eleventh century,'6 was a leading supporterin the effortto maintaina Buddhistpresence in south India. Srivijayanmonks, sponsored by the Sailendrakings and patronizedby early Cola monarchs,built at least two temples there at the beginning of the eleventh century.'7 The largest and most imposing monasteryin Nagara Sri Dharmardja Wat Mahddhdtu, is a walled complex located within the city walls. Its chief architectural featureis a large bell-shaped
BEFEO, vol. XXXI, no. 3-4, p. 373. BCAIC (i91 2), p. I48. 13 Ccedes, Recueil desinscriptions, I, Inscription I, pp. 37-48. vol. '4 Prince Damrong Rajanubhab, A History of Buddhist Monuments Siam, trans.S. Sivaraksa (Bangkok: i962), p.6; and in A. B. Griswold, "Siam and the 'Sinhalese' Stupa," Buddhist Annual(Inauguralissue, Colombo, I964), pp. 76-77. Griswold gives reasons for dating the remodeling around the first quarter of the 13th century. IS S. Paranavitana,"Negapatam and TheravddaBuddhism in South India," Journal the Greater of IndiaSociety, vol. XI, no. i I (I 944), pp.- 7-2 5. 16 For the political history of the period see 0. W. Wolters, "Tdmbralinga," Bulletin of the Schoolof Orientaland African Studies,vol. XXI, Pt. 3 (195 8), PP. 587-607. 7 T. N. Ramachandran,"The Ndgapattinamand Other Buddhist Bronzes in the Madras Museum," Bulletinof the Madras Government Museum,vol. VII, no. I (I954), p. I 3.
II
12

2z

stupa, which is surroundedby i6o smaller stupas.I8 It was apparentlythis large stupa that the Siamesemonks and their Sinhalesecompanionsbuilt in the thirteenthcenturyencasingan older monument. The custom of enclosing older monuments in a series of reconstructionswas practicedby the Mons and continuedby the Thai.When undertakingthe alterationof an earlier structure,they often constructeda replicaof it on a reducedscale nearby.Examplesof this can be found, accordingto Claeys,I'at Bejrapurl, NagaraPathamaand Chieng Mai. Being awareof this practice, Claeys was interestedin a small masonry structurestanding in the courtyardof which in no way resembledany of the other stupasin the area.He made very Wat Mahddhdtu, carefuldrawings of its elevation and plan, and concluded that it resemblesthe Mahdyana Buddhist monument Chandi Kalasan of central Java, and the Cham towers at Dong Diiong and
Mi-S6n.20

In several other structuresin Nagara Sri Dharmardja both Claeys and QuaritchWales detected vestiges of "Indo-Javanese"architecturalstyles.2, There are three small Brahmanical sanctuaries which are ratherlost in the midst of this proliferationof Buddhistmonuments.One is of modern construction.The second, a Siva Temple, is a ruinedbrick structureshelteredby a in wooden shed.22 QuaritchWales consideredit to be "Indo-Javanese" style, and saw the same affinitiesin another small ruined Siva sanctuarynearby where he was able to cut several trial trenches.He found two distinct brick floors under the Siva structure,on the lower of which he found a reliquarycontaining a coin dating from one of the earlierreigns of the Ayudhydkingdom. Below this were many potsherds of Sung type. He concluded that the site did not date from a period earlierthen the ioth or i ith century, and was probablya good deal later.23 Beyond the southernend of the present town there are the ruinedwalls of an old city. This town which is oriented on the same north-south axis as the present town, is probablya much earliersettlementthen the present Nagara Sri Dharmardja, it may be the source of some of and the very ancient objects to be seen in Wat Mahddhdtu. There is one piece of evidence overlooked by WalesandClaeysthat may prove of some value in assessing the antiquity of Brahmanical practicein Nagara Sri Dharmardja. his report on In the Siva temples, Lajonquierementions five lingas,four of which he says were designed in three sections, with a squarebase, an octagonal middle portion, and a rounded upper portion with a ridge in the middle representingthe frenum.During a visit to NagaraSri Dharmardja had an I opportunity to study some of these liNgasand confirm the description given by Lajonquiere. One of them is especiallyinteresting(fig. 5). There are obviously pitfalls in any effort to draw chronological conclusions from such a highly conventionalizedand simple form as the linga. This is especiallytrue in an area such as the MalayPeninsulawhere there are not a great numberof liugasavailablefor study. Any effort to set up typologies on the Peninsula would be unrealistic,but it is possible to draw some
18 '9 20 21
22 23

For plan and elevation see: BCAIC BEFEO, vol. XXXI, p. 374.
Ibid., p.377.

(i9g2),

p.

149.

24

Ibid., p.374; and, IAL, vol.IX, no.I, p.23. IAL, vol.IX, no. I, plate VII. Ibid., p. 23. BCAIC (I 9 I 2), p. I 6o, and sketch in fig. 37,

p. I 5 9.

23

conclusions from recentwork on this problem in connection with the excavationsat Oc-Eo,2s The and the development of styles in the sculptureof Champa.26 lingaplays an importantrole in both Champaand the successivepolitical entities on Cambodiansoil from the time of Fu-nan. It is mentioned in many inscriptions,but there is no obvious correlationbetween extant litigas found a and ancient inscriptions.In his pioneer study of pre-Angkorianart, Henri Parmentier in that appearedto be rathernaturalistic conception, which he was inclined to number of litigas His consider earlierin date than the more conventionalizedrepresentations.27 opinion has been given added force by the discovery in the Transbassacarea of a number of lingasthat are distinguished by their greaterrealism.Their frequencyin the territoryof the lower Mekong, the area consideredto be the center of the political community of Fu-nan, has argued in favor of as viewing the more realisticrepresentations earliestin time. ProfessorMalleretwould date the areafrom the end of the 5th to the beginning of the most realisticligas from the Transbassac 6th century.28 The criteriafor greateror less realismdepend on the treatmentof the upper cylindricalor and hemisphericalportion of the litiga(Rudrabhdga), its relative size in relation to the middle which is a cube. Among the most octagonal section ( Visnubhdga),and the base (Brahmabhdga), in region, there is an unequalemphasisin realisticand presumablyearliestlitigas the Transbassac the parts, or a total suppressionof one of the lower parts, and an exaggerationof the Rudrabhdga.29 There is, too, a tendency towards anatomicalfidelity in rendering.On the other hand, that are divided into three well-definedunits of approximately equal length, and in those litigas which there has been an attenuationof the relativerealismof the precedingseries, are classified by Malleretas "conventional emblems."30 found at Oc-Eo is consideredto be the with a face (mukhalitga) Within this last series,a litiga oldest.3'Its realismconsists of the swelling ovoid form of its top section, the strongly marked gland, and the disproportionbetween the size of the top section and that of the octagonal and square sections. The measurementsof the three sections beginning with the base are: 2i cm., 2I cm., and 23 cm. It would thus appearthat this emblemwould fit with ease into the late Fu-nan or early pre-Angkorianchronology, though of course without sufficientevidence for any firm date. The Nagara Sri Dharmaraja lifigaexhibits the same pronounced ovoid swelling of the top section, and the same raised markingof the contours of the gland and the frenum. It also exhibits a disparitybetween the relative size of the three sections. It is not possible to draw any very definite conclusions from these similarities,but they suggest an early date, and point to or contact between the Oc-Eo area and Nagara Sri Dharmardja, at least to a common artistic and iconographic traditionfrom which both drew their models. Despite the ratherdisappointingevidence provided by Dr. Wales'sexcavationsat the site of the abandonedSiva temple, it would appear,on the basis of the litigastill to be seen in Nagara
25 Malleret, L'archeologie deltadu Mekong(ADM), vol. I, pp. 377-3 88. du 26 Boisselier, La statuaire du Champa,pp.410-415. 27 Henri Parmentier, L'art khmer primitif (Paris: I 927), pp. 3 1-3 1 2. 28 ADM, vol.I, pp.379-380.
29 30
3

Ibid., pp. 380-382;

nos. ioz-io6.
nOS. I07-

Ibid., pp.382-385;
Ibid., p.383,
no.I07,

1 5.

plate LXXXI.

24

that Brahmanical practicethere may date from at least as early as the sixth or Sri Dharmaraja, seventh century. The question obviously cannot be settled by the rough dating of an object like a linga. Indeed, there is no reason to assume that the linga was not moved to its present location from an area outside the walls of the present city. This serves to underscorethe very meagre knowledge that we possess of what is very likely one of the most promising areasfor archaeologicalsurvey and excavationon the MalayPeninsula. Within the precinctof Wat Mahadhatuat Nagara Sri Dharmar-jathere is a most interesting museum. Among the objects on display in it are two stone sculptures(figs.2 and 3) that bear a very close resemblancein style and iconographyto the JaiydVisnu (fig. i). These images have received almost no attention from visitors to the sites of early Indianized settlement on the Peninsula. Both images have small round faces with button-likeeyes; both are four-armedfigureswith the posterior arms missing. In the remainingleft hand of each is a conch shell held against the hip. The remainingright hand of fig. z, which has the palm up and the fingers extendedat hiplevel, must have held either the lotus or a symbol of the earth. Both images wear high mitres, which are similarin shape and decoratedwith patterns of leaf and vine; both wear heavy earringswhich were originallyattachedto the shouldersas well as the earlobes.Apartfrom the earringsthe only jewelryis an armleton fig. 3. The chest is nude. The upper edge of the dhoti,seen at the waist, is unornamented.As in fig. i, the lower edge encirclesthe legs below the knees and the heavy vertical fold of cloth falling between the legs is made up of the lateraledge or edges of the dhoti.The diagonalfolds or creasesalong the legs passes once around the waist, its two are rendered by incised lines. In fig. 2 the kamarband tasseled ends can be seen near the left hip, and the verticalfold of the dhotiemergesfrom under which is twisted, passes twice aroundthe waist; its ends are invisible; it. In fig. 3 the kamarband, and the vertical fold of cloth seems to start at the upper edge of the dhotiand pass in front of as both rounds of the kamarband, if the lateraledge or edges of the dhotiwere pulled up and a in behind the kamarband. Both images wear a sash falling in an arc in front of portion tucked the thighs, knotted or looped at the left and right, and forming a tight spiralat the left; in fig. 3 as it is uncertainwhether or not this sash is distinct from the kamarband, the ends of the latter are not visible. If we compare the Nagara Sri Dharmarija sculptures with the Jaiyd image (fig. i), their relationshipin terms of technique,dress, and iconographyis immediatelyapparent.Most striking is the fact that all three wear the same sort of high, decoratedmitre. True, the figuresfrom have taller mitres, which, unlike that worn by the Jaiyd image, have Nagara Sri Dharmardja inclined sides. But the differencesare only relative and the similaritiesbetween the two are manifest.Again, all three images have earringsthat rest on the shoulders,renderedon an almost oppressively heavy scale. All three images are nude to the waist, and all wear a similar long dhoti,markedwith incised lines, and with a heavy vertical fold. All three have a similar,if not identical, system of sashes. All three images are totally frontal. Each is designed to be seen in a controlled experience, program.Their overridingcharacteristic probablyas the occupantof a niche in an architectural is their flat, two-dimensionalpresentation.The surface of the stone pier is unmodified except for the superficialattachmentof linearpatternsof dress and adornment.AUlthree images have
25

a harsh, broken outline which is in tense opposition to the rigidity and staticimmobility of the vertical axis. These unresolved tensions would be even greaterif the images had retainedtheir upper arms. Despite minor iconographicaldifferences,the threeimages sharea common scheme: they all present Visnu in terms of a system of agreed formal conventions. They are not only similarin type but also similar in style, in the handling of the surface planes and in the harsh angular rhythm of the contours, while at another level they could be said to sharea community of expressive content. At every level then, the relationshipbetween the images seems evident. There are, however, enough differencesto indicate that these statues of Visnu were produced either in different workshops or at differenttimes when the schemahad undergone some slight alteration. PresentDating The Jaiy! Visnu has aroused some interest because of its unusual characteristics, such as the form of its mitre and its unusualiconographywith the conch held in the left hand on the hip; but it has been discussed only in relation to some other problem, and never as the primary object of study. When Ccdedspublished it in i928, he noted that it was one of a group from Jaiy! and Vieng Sra that were close to Indian prototypes, but without specifying which prototypes he had in mind.32 Reginald le May notes its "ratherclumsy make", and the "peculiarheaviness" of the ears and headdress.33 stimulatedonly a passingcommentfrom the Indianhistorian,K. A. Nilakanta It Sastri, who describedit as "stiff and inelegant [ ...] a product of late art, valuableas showing the persistenceof Indian influencesto a late period."34 PierreDupont found a relationshipbetween it and the extremely crude Visnu found at Tuol Koh in Cambodia,which occupied a somewhat marginal position in his scheme of pre-Angkorianart. Regarding the Tuol Koh Visnu, Dupont writes: "On peut d'ailleursse demanderd'abord s'il s'agit d'une production d'un archaismeexceptionnel ou au contraired'une imitation faite par des artistes maladroits. Un des axiomes de l'archeologiedu Sud-Est asiatiqueest que les statuesles plus proches de la traditionindienne dans ce qu'elle a de classiquesont aussi les plus anciennes.C'est le cas notamment des statues du Phnom Da. L'archalsmese traduitdonc par la plus grande perfection des formes. Ceci constitue une premiere presomption defavorablea la haute anciennetedu Visnu de Tuol Koh, mais d'autress'excercentdans le meme sens. Elles tiennentau fait que nombredes particularites propresa cette image s'expliquentseulementpar l'alterationde detailsdejareleves sur les autres Visnu a vetement long ... [La statue] est extraordinairement plate et evoque davantageune dalle de schiste decoupee qu'une veritableronde-bosse."Dupont then goes on to compare this statue with the Visnu of Jaiya (our fig. i), and concludes: "Ces comparaisons ont pour consequencede montrer que le Visnu de Tuol Koh n'est pas une statue particulierement archakque mais une production assez fruste derivant d'images locales. Son anatomie est d'ailleurssinguliere ... Elle suppose un art autochtone dont le Visnu de Chaiya [our fig. i] et
32
33

34

Ccedes, Les collections archdologiques plate X and p. 25. ..., Le May, The Culture..., p.8o. SouthIndianInfluences the Far East (Bombay: I949), p. in

92.

26

quelques autres statues inedites, originaires de Malaisie centrale, constituent egalement des specimens. Cet art est caracterisepar la reproduction, singulierement deformee, d'apports indiens qui peuvent etre reconstituesmais ne sont pas tous de meme epoque. Le centre de cette production reste a determiner,mais il semble exterieurau Tchen-la."35 Some of Dupont's views were questioned by the English art historian P.S.Rawson in 1957.36Rawson consideredthat some of the long-robed Visnu images, such as our fig. i, were older than Dupont thought, and he believed that this group of images had affinitieswith the pre-Guptaart of the Kusanas.These views were contested by Professor Boisselier,who stated that the iconography of Visnu with the conch shell on the hip, as representedby the Jaiya Visnu and severalpre-Angkorianimages, was unknown in India until the eighth century. In a long and wide-rangingarticleaddressedto problemsposed by the discovery of an early Visnu image at Tjibuajain west Java, Boisselier expandedthese views.38He placed the Jaiya Visnu in the context of a group of images of diverse iconography, religious intention and geographic provenance - from Bali, Java, Cambodia, Annam, Thailand and the isthmian section of the MalayPeninsula.He saw these images as sharinga mixed stylistic character,the result of southern Indian influences of the 7th and 8th centuries combined with earlierlocal traditions.The Visnu of Jaiya,accordingto Boisselier,is the work of a mediocreartisanwhose techniqueis a regressionfrom earlierwork on the Peninsula.39 Inscrzptions NagaraSri Dharmaraja from In attempting to assess the antiquity of Hindu practice in PeninsularSiam the testimony of epigraphyis of interest. Inscriptionsfrom Jaiya are not of much assistancebecause (with the possible exception of the Ligor inscriptionwhich is of contestedprovenance)none of them are is earlierthan the izth century.The situationat NagaraSri Dharmaraja quite different.A line of is inscriptionon the stairwaynear the door to the small museum at Wat Mahadhatu of interest The because of the archaismof its characters,which Cacdesdated to the 5th or 6th century.40 of eight lines, a Tamil inscription probably same museum possesses an illegible inscription dating to Cala times, and an inscribed granite slab which is not included in the collection of but inscriptionsedited by ProfessorCcdeds, which was photographedby Professor Lamb when On is he visited Nagara Sri Dharmardja.4' the basis of the photograph,Dr. de Casparis inclined to read a referenceto Siva, and to date the writing to the 6th century or earlier.42 This would tend to confirmthe existence of Hindu worship at Nagara Sri Dharmaraja an earlyperiod. at Our examinationhas been confinedto materialevidence actuallyfound on the isthmus. The record is meagre, and in a sense it may be misleading. Paralleltestimony from Chinese texts indicates that Brahminsfrom India were a rather common sight on the isthmus by the third
35 36 37 38

La statuaire preangkorienne (Ascona, Switzerland: I95 5), pp. 133-135. P.S. Rawson, Review of La statuaire preangkorienne, Pierre Dupont, OrientalArt, Spring (I957). by "La statuaire preangkorienne et Pierre Dupont," Arts Asiatiques,vol. VI (I959, pt. i), p. 67. "Le Visnu de Tjibuaja (Java Occidental) et la statuaire du Sud-Est asiatique," Artibus Asiae, vol. XXII
2IO-226.

(I959),

pp.

39 40 4' 42

Ibid., p. 224. Ccedes, Recueildesinscriptions part.II, no. XXVIII, p. 39. ..., Lamb, "Miscellaneous Papers," FMJ, vol.VI (I96I), plate I 17. Ibid.
27

century A.D., so it would of course follow that Hindu cult practice was a feature of life, at least for an elite, at that time. The difficultyis that the Chineserecordsare not securelyanchored in the topographic reality of the isthmian earth. Tin-sun,according to Chinese reports which can ultimately be traced to the third century A.D., was an important trading center and the home of more than iooo Brahmans.Unfortunatelyit has not yet proved possible to find sufficient precision in the Chinese text to link the name firmly to any narrowly circumscribed portion of the Peninsula.Although most scholarsagree that the geographicalindicationsin the text all point to the peninsula, various locations have been proposed.43 is clear too from It Chinese texts that in the third century there were Brahminsat P'an-p'an,another small state almost certainlylocated on the isthmus. But until the archaeologicalcontext is much richer,it would be rash to attempt to link up any of the isthmian sculptureswith specific place-names mentioned in Chinesetexts. Oc-Eo Vi~snufrom Directly relatedto the three isthmianVisnu images with the conch shell on the hip (figs. I, 2, 3) is a statuefrom Oc-Eo in the Mekong Delta. Here, between the ist and the 6th centuriesA.D., was the port city of Fu-nan.Excavationsand aerialphotographyhave revealeda web of complex canalsthat served to drainthe water-logged soil of the delta and createdan earlyAsian parallel to Venice. It is not surprisingthat objects found on the isthmus should resemble one from Oc-Eo, since Fu-nan had a powerful political impact on the neighboring kingdoms. The Liang-shu, the Annals of the Liang Dynasty in China,containsmaterialcollectedin Fu-nanby the third-century Chinese envoys, K'ang T'ai and Chu-Ying. From this and other records of the period, it is known that the MalayPeninsula,or several of the kingdoms on the Peninsula,were under the hegemony of Fu-nan.44 The Transbassacsite of Oc-Eo has yielded only four statues. The first, a wooden statue of the Buddha, has been lost. The second, a small figure of Gane'a, is probablynot earlierthan the second half of the 7th century.The third is a stone torso of a god which is remarkable for the relativenaturalismof its forms and the subtle transitionsof its surfaceplanes; Malleretwas undecided about its date, though he consideredit to be an earlywork; but its incised sampot shows no parallelwith the alreadyestablishedFunanesestylisticgroup of Phnom Da. The fourth image is a Visnu with a conch shell at the left hip (fig.4), which Malleretbelieved might date from the first half of the 6th century.45 Despite its evidence of commercialsuccess and cosmopolitan contacts, as exhibited in the many small items of personal adornmentfound at Oc-Eo, no stone statues that can be dated with assuranceto the earlyperiod of Funaneseprosperityhave been found there. It is not until almost the end of the Fu-nan empirethat stone statuaryappears,and then it is found not in the maritimesite of Fu-nan,but ratherat the uplandsite of Angkor Borei in southeastern Cambodia. The earlieststatuaryfound in this latter region appearsto date from the reign of King Rudra43Wheatley,
44

45

..., GoldenKhersonese pp.I 4-22. The standardcompilation of those Chinese materialsrelating to Fu-nan is: Paul Pelliot, "Le Fou-nan," BEFEO, vol.III 3 (I903), pp. 248-3 27. ADM, vol. I, p. 363.

28

varman (5I4 until after 539 A. D.).46 Pierre Dupont has establisheda chronology for this late Fu-nan sculpturefrom its beginnings around 5 I4 until its terminationat the end of the sixth century. This is the group of images known as "Phnom Da styles A and B". As there are no statuesin the Oc-Eo region that can be fitted into the Phnom Da style, it now appearsthat the problems of stone sculpture were not attempted by the Funanese until after changes in the course of the Bassac River, or pressurefrom Chen-la,had forced the re-location of the major centers of population on to higher ground where stone was readily available. The stone statuaryof the Transbassac region, when it does appearafterthe triumphof ChenIt la, is markedby two characteristics. is almostalwaysquite small,and most often it is assembled and joined ratherthan carved from a single block of stone. Both of these featuresare probably relatedto the scarcityof stone in the delta region. The Visnu with the conch on the hip (fig.4) displaysboth these characteristics, which helps confirmits provenance,for the image was not excavatedby Malleretbut sold to him in 1945 by Its inhabitantsof the Oc-Eo areawho said they found it at the site.47 resemblanceto the Jaiyd Visnu (fig. i) did not escape Malleret,who has developed a convincing case for linking the two images.48 Oc-Eo image is even more like the two from NagaraSri Dharmardja The (figs.2 and 3) both in iconography and in style. There are a number of similaritiesin the dress. All three images wear tall mitreswhich are variantsof the same basic shape, though the Oc-Eo image has no decoration on its mitre, in contrast to the patternsin bas-reliefon the others. All three images have the same sash falling and in an arc down the front of the dhoti, the heavy verticalfold of draperyfalling down between the legs. In contrastto the Peninsularimages, however, this fold in fig. 4 seems to fall down in front of the arc of the sash, which is difficultto explain;but there are so many possible ways of drapinga voluminous dhotiwe may well hesitate to accuse the sculptor of a solecism. Another oddity of the Oc-Eo Visnu is an indistinctloop under the conch shell which, Malleretobserves, could be the simplificationof the bulging knotted sash on the hips of the Peninsularimages.49 These oddities suggest that the Oc-Eo image is a copy, laterin time and removed in distance,of a Peninsularmodel. The loop may be a vestigial remnantof the free-hangingend of the knotted sash which is present in the same position on the left side of both the Jaiyd and Nagara Sri Visnus. All the images retainthe earringsreachingto the shoulder.Otherwisethey Dharmardja are all relativelyunadorned. Ironography All four of these four-armedstanding images are representationsof one of the twenty-four forms (mfirti) or emanations (vyuha)of Visnu, which are distinguished by the possession of certainaspectsof the god's attributessuch as energy, power, knowledge, or strength,and which are invoked by the supplicantin accordancewith their special capabilitiesto meet the prayers and requirementsof the moment.50 These manifestationsof the varied nature of the supreme god were limited to four when the concept was firstformulatedaroundthe second century B.C.
46

Dupont, La slatuaire priangkorienne, 22. p.

volI1, P.363.

47ADM,
48 49

Ibid. PPT393-395p

Ibid-, P.3 94. 50 J. N. Bancfjca, The Development Hindu Icono~graphy of (Calcutta: I 946), pp.- 387-3 88.
29

By Gupta times, however, the number has grown to twenty-four,which is still the number in use today in the iconographic manuals.Each of the mfirtisholds four attributes,one in each hand. These are listed in the Indian texts as the conch shell (satkha), the wheel or discus (cakra), the club (gadd) and the lotus (padma). In the Cambodianinscriptions the lotus is replacedby a symbol of the Earth (bhul),which, as Pierre Dupont noted, takes a variety of art: forms in Cambodian a sphere,a disk, a segment of a sphere,or a disk that has been hollowed I out. The Sanskritinscriptionfrom the templeof Ta Keo, datingfrom the reign of Siuryavarman An u (Iooz-c. 105 0), lists the attributesof Visn. as the club, the conch, the wheel and the earth.sI inscriptionin Old Khmer, dating from the middle of the 7th century, lists the earth as one of his attributes.52 Dupont observes that in Indianreligious traditionthe form of the earthwas not a sphere,but a disc, and it is thus necessaryto considerthe sphere,when encounteredin Southeast Asia, as less ancient than the disc. This may be the case, but a thorough survey of the iconographyof pre-GuptaVisnu images needs to be made before there is any certaintyon this point. There was no standardof practicein the formative period of Hindu iconography. For apparentlya Kusdnaimage from Mathura,carriesa slightly example, the Visnu of Hinkrail,53 rounded attributewhich could be either a flattenedsphere or a disc. A late Gandharabronze Visnu, 4now in the Museum fuirVolkerkunde, Berlin, holds a realisticlotus with a stem. The well-known Visnu from Taxilassalso carriesa realisticlotus. According to Agrawalathe early images of Visnu at Mathurdexhibit a considerablediversity in the range of attributes,and the In the padmawas the last attributeto be evolved.56 southernIndia, accordingto Sivaramamurti, it lotus occursinfrequentlyin earlyPallavasculptures,but when it is reprensented is naturalistic. In some earlyCdlukyan sculptures,it is representedas "a small bud which may be mistakenfor a frUit."57This diversityin the form of attributescould point to the existence,at an earlyperiod, of of paralleliconographictraditionscalling for the representation either the lotus or the earth. was standardized, after the Gupta period, the time the iconography of Brahmanical images By the lotus may have emerged as the more usual and "legitimate"attributefor Visnu images. A certainlack of precision in the identificationof four-armedVisnus arises from a lack of uniformity in the texts available to the iconographers. The key to the discriminationof the individual mfirtifrom the other twenty-threeemanationsis the disposition of the attributes.In Rao compiling his lists of the various combinationsand the identities of the mzirtis, cites several as of the Purdanas his guide. These versifiedtexts deal with a varietyof subjects,secularas well as of religious. They are primarilydevoted to a descriptionof the characteristics one of the great Much of the informationin the Puranas ancient. divinities, and to the elements of his cult.58 is containhistoricalinformation They give evidence of a long period of elaboration.Some Purdnas the 6th or 7th century, and so they would have been compiled no earlierthan that relating to This time. One of the texts cited by Rao is the Pdtdlasection of the Pddma-purdna.59 section of
pp. Dupont, La statuaire preangkorienne, I 43-I 44. Ibid. 53 S. K. Saraswati, Early Sculpture Bengal(Calcutta: I 962), pp. I 3-I 5, and fig. 4. of 54 C. Sivaramamurti,Indian Bronzes (Bombay: i962), fig. 2. 55 Banerjea, Hindu Iconography, plate XXI. 56 Vasudeva S. Agrawala, A Catalogue the Brahmanical qf Imagesin MathurdArt (Lucknow: I95 57 Ancient India, no.6 (January I950), p.5 I. 58 Louis Renou (ed.), Hinduism (New York: i963), pp. I39-I40. 59 T.A. Gopinatha Rao, Elementsof Hindu Iconography (Madras: I9I4), vol.I, part.I, pp. 230-232.
5I 52

I).

30

the text is consideredto be of somewhat recent origin because of its referencesto tantrismand It its pronouncedreligious syncretism.6o is in some cases in conflict with the other text used by Rao, the Rfipama~ndana, which, although not of great antiquity,is regardedas a compilation of greaterauthority.61 any event, it is apparentthat these Indiantexts were compiledin Medieval In times, so they may reflecta traditionthat was stabilizedconsiderablyafterthe productionof the Visnu images from Nagara Sri Dharmaraja, Jaiydand Oc-Eo. This time-lag, together with the conflicts in the two texts cited by Rao, makes a positive identificationof the various mfirtisa most uncertainendeavor. In the great majorityof Khmer images of Visnu, from the beginning of the Pre-Angkorian period to the end of the Angkorian, the attributes are arrangedaccording to the following scheme: the conch in the upper left hand, the wheel in the upper right, the earth in the lower right, and the club in the lower left.62 This mfirti Visnu, accordingto the Rfipamayzdana, of would be Janardana;while according to the Pddma-puradna it would be Vdsudeva. The Oc-Eo Visnu does not fit this preponderanticonographicpattern. The conch shell, instead of being held in the upper right hand, is held in the lower left hand againstthe hip. Whatevermay have been the distributionof the attributesheld in the upper, and now missing, hands, it is apparentthat the Oc-Eo image stands outside the mainstreamof iconography in Cambodianart. The same is true of the isthmianimages when comparedto any other Visnu image createdin Siam. In orderto find parallelsfor this "aberrant" iconographictraditionand to resolve the contradiction between the dating for these images, 6th centuryfor the Oc-Eo image and 8th century for the very similaristhmian images, it will be necessaryto turn to a considerationof several images of Visnu in India.

60 61
62

L. Renou, L'Inde classique (Paris: 1947), pp. 47-419. Rao, Hindu Iconography, vol.1, part I, p. 23 I; and Banerjee, Development Hindu Iconography, of pp. ADM, vol.I, p. 39I 3I

22-23.

CHAPTER III

INDIAN

COMPARISONS

he problemof the 'aberrant' Visnu images is difficultbecausethey stand outside the iconographictraditionof the regions in which they were found, they are completely unfixed in terms of stratigraphyor of documented associationwith materialthat might be dateable,and there is such a small sample that internal comparativeanalysis will not carry very far. Confronted with such a situation, it is sometimes possible to interpolateproperly ordered objects into the chronology of a contemporaryand better documentedculture. Scholarshave used the frameworkof Indian art in order to draw conclusions about objects found in SoutheastAsia; problemsof Etruscanart have been solved by referenceto Greek and earlyRoman art; and the comparativeuse of the chronology of Roman, Parthian,and Sassanianart has been helpful for dating many objects found in CentralAsia, Afghanistan,and northwesternIndia. Though this has been a very useful technique,it has had some unfortunateconsequencesin Southeast Asia. The constant referenceto Indian styles has led some scholarsto give an undue prominence to Indian influence without reference to the autonomous elements of Southeast Asian culture and its powers of assimilation, adaptation and innovation.' While caution is necessaryin drawing conclusions, Indian art styles do in some cases offer an importantbenchmark,especiallyfor locating the prototypesfrom which earlySoutheastAsianimages developed. As Buddhist art, especiallyin the Theravida with its severely circumscribediconography, changes very slowly, it is possible to draw the closest parallelsbetween certain Buddhaimages in SoutheastAsia and their prototypesfrom the workshops of Sarnath,Mathurd Amardvati. and on this problem is now at an advancedstate as a result of the researchof PierreDupont, Work Jean Boisselier and especiallyA. B. Griswold.zIt has proved possible, if somewhat more diffibetween Hindu images madein SoutheastAsia and cult, to draw a fairlyprecisecorrespondence those fashioned in India. For example,PierreDupont, while noting some general affinitiesbetween the Phnom Da seriesin the art of Fu-nanand the post-Gupta style of Ellora, pointed also
T

It should be noted that there has been a vigorous reaction to this tendency. For example, see D. G. E. Hall's treatment of the problem in Chapter I of the new edition A History of South-EastAsia (3rd ed., New York: i968), and J. G. de Casparis,"Historical Writing on Indonesia: Early Period," in Historiansof South-EastAsia, ed., D. G. E. Hall (London: Earlier work by the economic historian J. C. van Leur, and the archaeologists W. F. Stutterheimand 196i), pp. I 2i-i63. F. D. K. Bosch, among others, had provided the stimulus for a greater awareness of elements of autonomy in Southeast Asian culture. Pierre Dupont, Archiologie manede Dvdravati(Paris: I959): J. Boisselier, Le Cambodge (Paris: i966), pp. 266-27I; and A. B. Griswold, "The Santubong Buddha and Its Context," SarawakMuseum Journal,vol. X (July-December, i962), pp. 363-37 I, "Prolegomena to the Study of the Buddha'sDress in Chinese Sculpture,"ArfibusAsiae, vol. XXVI, 2, pp. I 25 f.; and "Imported Images and the Nature of Copying in the Art of Siam," in A. B. Griswold, et al. (eds.), Essays Offered to G. H. Luce, vol. 11 (Ascona: I 966), pp. 37-7 3.
32

between the styles in many details so that the analogieswere on the order to the dissimilarities Another instructive exof general tendencies rather than a point-for-point correspondence.3 ample would be the inconclusive discussion of the prototype for the recentlydiscoveredbronze at Jalong in Malaysia.4 It will be possible, however, to trace the most direct relationship between the 'abberant' and Visnu images from Jaiyd,Nagara Sri Dharmardja Oc-Eo and their prototype in Indian art. Images from Mathurd Kufdna is Mathurd an ancientpolitical,commercialand religious center.It servedas a politicalcapitalfor the kings of the Kusdnadynasty,a group of Indo-Scythiannomads that ruled Taxila,Peshawar and SurkhKotal from this city in the upper D6ab. Although the datesfor this dynastyare much debated, they apparentlyconsolidated their political position in India some time in the ist centuryA. D. and ruled until the mid-thirdcenturyA. D. when they were defeatedby ShapurI, of the SassanianDynasty of Persia. Innumerablesculpturesof the Buddhist, Jain and Hindu pantheons have been recovered for from the soil at Mathurd, the city was hospitableto all the majorreligions.sIt was at Mathurd during the period of Kusana rule that many of the iconographic forms of Hindu art were developed. In the earliestforms at Mathurd,the representationsof Visnu are almost identical one with those of the BodhisattvaMaitreya;one right hand is raisedin abhayamudrd, left hand (nectarflask), and the other two hands hold the gaddand the Cakra.6 holds the am~ritaghata The god 6 is an early Kusana image of Visnu of this sort, at Palikheranear Mathurd.7 Fig. while the other is wrapped wears a foliated turban. One right hand is raised in abhayamudrd, aroundthe club (gadd),which is almost exactlyas tall as the figure, taperedwith the broad end at the top, and encircledby bandsat intervals.One left hand holds the cakra,the other holds the nectar flask associatedwith Maitreya. Fig. 7 is anotherearlyVisnu of similartype. The image wears a large torque decoratedwith is a design of rosettesand leaves. The dhoti held in place by a kamarband, most of which is hidden by the sash tied around the waist.8 The Visnu from Jaiyd(fig. i) resemblesfigs. 6 and 7 in severalrespects.All three images have the posterior right hand carved in relief against the gaddand the anteriorright hand in abhayathe mudrd; gaddis encircledwith bands at intervals; and the disposition of jewelryis the same: in and torque, earrings,wrist and arm bands. There are, however, significantdifferences mukuta dhoti;and the attributeheld on the left hip in fig. i is a conch shell instead of a nectarflask. In
3Dupont, Prdangkorienne p. 26. ..., A. Lamb, "Treasure Trove among the Tapioca," Malayain History, vol.VIII, I (Dec. i962), pp. II-I4. 5 Basic general references for an understanding of the art objects and their context at Mathurdare: J. E. van Lohuizen-de Leeuw, TheScythian Period(Leiden: I949); J. Ph. Vogel, La sculpture Mathurd,vol. XV: Ars Asiatica (Paris and Brude xelles: I930); and John M.Rosenfield, The DynasticArts of the Kushans (Berkeley: i967). 6 The basic guide to the Brahmanicalsculptures in the MathurdMuseum is V. S. Agrawala, A Catalogue theBrahmanical of Imagesin MathuraArt (U. P. Historical Society, Lucknow; I95I). Valuable also for the study of early Visnu images at Mathurd is D. B. Diskalkar, "Some Brahmanical sculptures in the Mathura Museum," Journalof the United Provinces HistoricalSociety,vol.V, I(1932), pp. i8- 7
7 8

Agrawala, Catalogue ..., Ibid., pas 5.

p.4.

33

fig. 8, however, which is consideredto be later in the KusanaVisnu series, the flaskis replaced by a conch shell, a significantdevelopment in the emerging iconographic traditionfor Visnu images. In figs. 6, 7 and 8 the club is approximatelythe same height as the figure. According to a recent survey of the changing forms of the weapons held by Visnu in India, this is a diagnostic clue to images of the god in the Kusanaperiod.sThe club in fig. i is broken, but it too must have This would have been necessaryfor aestheticas reachedabout as high as the top of the mukuta. well as iconographic reasons; a posterior left hand raised to that level would requirea visual on counterbalance the right side; and as the posteriorright handholds the club at shoulderlevel, the only thing that could serve the purpose would be the top of the club itself. It will now be apparentthat in the iconography of Visnu the conch shell on the left hip is not a development of the 8th century.'oOn the contraryit seems to be the original position of that attributein the Kusanaart of Mathura.It is also the position in which it is most often seen on images found in northernIndia and Nepal.", Before leaving Mathura,we must note the change from the foliated turbanto the tall mitre, which was typicallyworn by images of Indra. It was a shift of considerablemagnitudebecause the tradition of regal or divine headdressin Indian sculpturehad traditionallybeen a turban, with a flat plate, feathersor some other device attachedto the front. Professor Rosenfieldsuggests that the stimulus for the change may have been the example of the high, jeweled crown worn by Kusdnakings.I2 period at Mathurd,illustrate this deFigs. 9 and io, both of which date from the Kusdn.a velopment. Both have the same disposition of attributes, and both wear crowns which are relatedto the usual headdressof Indra. In fig. io, which is ascribedto the late Kusdnaperiod, the crown is more cylindricalin shape.13 On the basis of analogieswith the images from Mathura,figs. i, 2 and 3 would appearto be among the most ancient images of Visnu in SoutheastAsia. It will be possible to confirm this by making comparisonswith some early Visnu images recently discovered in India, which are lineal descendantsof the traditionof style and iconographydeveloped at MathuraunderKusana patronage. Gujarat Vi4su from BhinmdA, about sixty miles northwest of Mount Abu. It has Bhinmalis in the Jalor districtof Rajasthan, He been the subjectof a study by UmakantP. Shah of the Universityof Baroda.'1 notes that the itself covers an old habitationmound, while there are several mounds to the town of Bhinmgl which are right of the presententranceto the town "full of pottery and ruins of brick structures is being dug out by local Chilsfor recoveringold bricks." In Shah'sopinion these ruins go back
9 C. Sivaramamurti,"The Weapons of Vishnu," Artibus Asiae, vol. XVIII, 2, p. 13 5. Jean Boisselier, "La statuairepreangkorienne et Pierre Dupont," Arts Asiatiques,vol.VI, i (I959), p.67. II Aschwin Lippe, "Vishnu's Conch in Nepal," Oriental Art, vol. VIII (Autumn, I 962), pp. 1I7-I I 9. 12 Rosenfield, Dynastic Arts ..., p. I 89. 13 Agrawala, Catalogue p. 5. The image is no. 2007. ... andPictureGalleryBaroda,vol. XII I4 "Some Early Sculptures from Abu and Bhinmal," Bul'etinof the Museum
IO

(I955-56),

pp. 42-5 6. '5

Ibid., p. 54.

34

at least to Gupta times. His most importantfind was the Visnu in fig. i i, which is twelve inches high and carved in a greenish-blueschist. On the grounds of dress, modeling, proportionsand configurationof the facialfeatures,Shah is "inclinedto assign the Bhinmil Visnu to the period of transitionfrom the art of the Kshatrapaage to that of the classicalGupta art. The sculpture may date from c.4oo A.D."16 to A comparisonof fig. i i with fig. i offerssufficientcorrespondences indicatetheir relation to a common stylisticcurrent.Both images arepresentedin total and unrelievedfrontality.They share, too, the same harsh, segmented outline, the same dominanceof the linear contour over the volumetric potential of form, and the same flat board-like surfacesaccented by raised or incised linear patterns. They share, too, a common distribution of attributes; both hold the conch shell on the left hip and both grasp the club with the posterior right hand. The broad, round faces are strikingly similar,and so are the long eyes spreadingacross the entire face. between the two images are in the attachedpatternsof dress and confinedto The differences the rind or epidermallayer of the stone. Consideringthe thousandsof miles that separatetheir in provenance,it is not surprisingthat there are some differences dress. Even so, it is possible to relateindividualfeaturesof the dress of fig. i to the traditionsof western India. A twisted ropelike sash falling in an arc in front of the rhighs, and knotted at either side, is found on a number in of early Gupta images from Sdmaldji North Gujarat,in which the treatmentof the dhotiwith vertical fold and its system of incised radiatingpleats, is much the same as in fig. 1.17 its heavy The incised geometric patternon the upper edge of the dhotiin fig. i recallsthe patternon the And the decoration belt of an earlystanding Buddhaof the kapardin in the Musee Guimet.18 type of the mitre in fig. i, with vines, leaves and rosettes,is ratherlike that of the headdressworn by one of the Skmal'jiVisnus.19 That there should have been artisticcontact between western India and the MalayPeninsula is not at all surprising.Ccedesconsidersit a certaintythat the great western ports such as Bharukaccha(modernBroach),Sarpdraka (modernSopara)and Muchiri(modernCranganore) were in contact with the Golden Khersonese.zo is well known that these ports did a flourishing It trade with the Roman Empire during the early centuries of the Christianera. There is now a body of archaeologicalevidence that links SoutheastAsia withe the early Mediterranean trade, and it is reasonableto assume that some of the Roman objects showing up at SoutheastAsian marts were transshippedfrom the west coast of India. It is, of course, very difficultto connect most of them with any specificport in western India. But it is possible to point to a direct link in the case of the bronze figure excavatedat Tra Vinh in Cochinchina,which is a local copy of

16 '7

Ibid., P-54.

i8

'9

20

U. P. Shah, "Sculptures from 8dmaldjiand Roda (North Gujarat) in the Baroda Museum," Bulletinof the Museumand PictureGallery,Baroda,vol.XII (Special Number, i960), pls.41, Standing 8iva; 47, Kumdra (Skanda); 48, Ndrfyana or Vi~varflpa-Visnu;and, 50, Nrfyaria of Vi'variipa Visnu. J. E. Van Lohuizen-de Leeuw, The "Scythian" Period (Leiden: 949), pl. XIX, fig. 30. See also: Moti Chandra, "The History of Indian Costume from the ist century A. D. to the Beginning of the 4th Century,"Journalof the IndianSociety of OrientalArt, vol. VII (I 940), p. 206. For another Vigiu image dating from the same period as the Bhinmnl image and with conch shell on the hip see: M. R. Majumdar(ed.), Chronology Gujarat(Baroda: i960), p. 209 and Plate XLIV. The type was apparentlyspread widely in of northwestern India. Les S8tatsbindouise's., p. 63. ...

35

a Poseidon by Lysippus,21 while another Poseidon figure of the same type was excavated at Kolhapur in the Deccan south of Bombay.22 At this point, however, the trade in Roman objects is less relevantto our inquiry than the establishmentof a chain of relations between Southeast Asia and northwesternIndia. In the DjakartaMuseum there is a small seal which Professor Louis Mallerethas relatedto a plaque found at the Kusdnasite of Begrdm,fifty miles west of Kabil.23His excavationsof the Funanese city of Oc-Eo revealeda number of links with northwesternIndia. He discovered a blue tourquoise seal bearingwhat looks like the figure of a Sassanian nobleman.24 site yielded a great The many gems, some of which appearto have affinitiesto jewels found at Sirkapand Taxila.2s Many of the gems are inscribedin Sanskritin a script which, according to Professor J.Filliozat,26 is used in north and centralIndia during the period from the second through the like the brdhmi fifth centuriesof the Christianera.At Bathe,nearOc-Eo, a Buddhahead of Gandhdran type was The pre-Angkorianimages of Siirya,dressedin a short tunic, wearingboots and a high found.27 mitre, are perhaps a distant reflection of Indo-Scythianinfluence.28 There may be a dynastic link between Fu-nan and the Indo-Scythiankings who bore the royal title of Chandan.29 the In 4th century, according to the Annals of the Liang Dynasty, a foreigner bearingthis title made himself king of Fu-nan. There is thus a body of evidence to link north and northwesternIndia with SoutheastAsia. Most of the materialevidence was found at the Funanesesite of Oc-Eo, a trading station of the earliest commercialempire in the region. The evidence for the Malay Peninsula, though less abundant,is not altogetherlacking. Among the early Buddhaimages found on the Peninsula, some appear to be imported from the classical Gupta school of Sdrndthand the post-Gupta schools of western India. As an importanttrade route connected Mathurdwith the west coast it port of Broach,30 is entirelypossible that products of the Mathurd workshops were reaching the Peninsulavia Broach. There is a small Buddhaimage from Ipoh in Perak, Malaysia,which A. B. Griswold considersto be very close to work done at Sdrndth.31 Another, seven inches high, carved in relief on a small stone stele, which was found by Dr. QuaritchWales at Vieng Sra on the isthmus, is believed by Mr.Griswold to be eithera product of a north Indianworkshop or a very close copy made on the isthmus.32
21

22

23 24
25

CharlesPicard, "A Figurine of Lysippan Type from the Far East: The Tra Vinh Dancer," ArtibusAsiae, vol. XIX, Pts. 3 and 4, pp.342-352. H. D. Sankaliaand M. G. Dikshit, Excavations Brabmapuri at (Kolhapur)I 945-46 (Deccan College Monograph Series: 5; Poona, India: I952), pl.XXXIII B. Louis Malleret, "Pierres gravees et cachets de divers pays du Sud-Est de l'Asie," BEFEO, vol.1, no. I (I963), p.99. Louis Malleret, L'Archdologie deltadu Mekong,vol.111, Texte: La culture du Fou-nan (Paris: I962), pp. 372-373. du
Ibid., pp.375-376.

26

G. Ccedes, "Fouilles en Cochinchine, Le site de Go-Oc-eo, ancien port du royaume de Fou-nan," Artibus Asiae, vol. X,
pt.3 (1947), p.197.

27

28

29 30

J. Boisselier, La statuaire khmere p. 86. ..., Dupont, La statuairepreangkorienne, 63-64, pl. XII B; and V. Goloubew, "Les Images du Suiryaau Cambodge," pp. Cahiersdel/'ico/e Franfaised'Extreme-Orient, no.22 (I 940), pp.38-42. L. Malleret, "Une nouvelle statue preangkorienne de Suirya dans le Bas-Mekong," in A. B. Griswold (ed.), Essays Oflered G. H. Luce,vol. 11 (Ascona: I 966), pp. I09-I 20. to See: G.Ccedes, Les 1Statshindouises pp.9I-94, and references. ..., M.D. Desai, "Some Roman Antiquities from Akota near Baroda," Bulletinof theBarodaState Museum, VII, pts.I and vol. II (I949-I9So), p. 22.
A. B. Griswold, "Imported Images
...," p.6i.

31 32

Ibid.

36

These images serve to bring the Peninsulainto the streamof contacts that left such a thorsites of Indo-China.If the archaeologicalevidence offered ough impressionon the Transbassac the isthmus is ratherslight for this period, it should be rememberedthat very little of the by area has been explored systematicallyby trained archaeologists.In any event high humidity, torrential rainfall and dense forest all conspire to reduce the body of evidence potentially availablefor study, and to createdifficultyin locating those materialswhich have withstood the ravages of the environment. No better illustrationof the random nature of discovery on the Peninsula can be offered than Colonel Low's comment that he discovered the Cherok Tekun inscriptionunder heavy undergrowthin Kedah at a place that he had passed often for a period of years.33 It should be observed that our laboredquest for existing evidence of materiallinks between the Peninsulaand northwesternIndia has been arbitrary its focus. It has ignored entirelythe in rich body of Chinese historical writings which would enable us to see the city-states of the Peninsulafunctioningas entrepotsin the tradethat reachedfromChinato IndiaandwesternAsia. Any doubts about the existence of links between the Peninsulaand western and northwestern India would be dispelled by an examinationof the early pattern of the Po-ssiutrade between western Asia and China,in which, as earlyas the fifth centuryA. D., Indonesianshipperswere alreadyproviding a key middlemanlink.34But the problem in this study has been to provide links in the transmissionof an art style, and since art styles are either transmittedby artists traveling from one point to another, a subject about which we have no evidence during this early period, or by the movement of art objects, we have concentratedour searchon the latter. The YtkeIwaram Excavations Recent excavationsnear the famous Buddhist site of Ndgdrjunakonda Andhra Pradeshhave in contributeda key document to the discussion of chronology and possible prototypes for early Visnu images in SoutheastAsia. Becauseof the threatof flooding posed by dams rising on the River Krishna,the importantsite of Ydleswaram was subjectedto a programof salvage archaeology during the period i96o-62. Habitationat the site was tracedthrough six culturalperiods ranging from about the znd century B.C. up to the late medieval period.35 In level IV, Dr. Khan, who was in charge of the excavation,uncovered a two-armedstatue of Visnu (fig. iz). The right hand holds a large columnar club (gadd), while the left holds a conch shell at the left hip. Dr. Khan dates the image to the 4th or 5th century, noting that in He style it is close to late Amardvatd. believes, however, that it was carved under the early Pallava kings who succeededthe Iksvdku dynastyat end of the 3rd centuryA.D.36 The dressis of specialinterestwhen comparedto the Visnu images of the isthmusand Oc-Eo (figs. I-4). The sash that falls in an arc in front of the legs is sharedby all the images. This is not found on any Kusdnaimages from Mathurd which I have been able to study. The inference
33 34 35 36

Low, Miscellaneous Papers Relatingto Indo-China, vol.I (i886), P. 223. . W. Wolters, Early Indonesian . Commerce. ., pp. I 29-I 5 8. M. A. W. Khan, A Monograph YtIlfwaram on Excavations,Andhra Pradesh Government Archaeological Series no.I4 (Hyderabad: i963). Ibid., p. I4. I am grateful to Dr. Khan for the photograph of the Vigiu from Yaldgwaram.

37

is that the immediateprototype for the SoutheastAsian images under examinationis likely to be found in one of the sites of the Andhradesa. Subsequentto the publicationof the Y6l6waram report, Dr. Khan producedan interesting reporton a relief (fig. I 3) recovered in coastalAndhraPradesh.It is an earlysculptureof Visnu who in his man-lionincarnationas Narasithha,37accompaniedby the "Five Heroes" (pahcavira) personifiedcertainqualitiesof Visnu. The second from our left, a standingfigure of Visnu with which derives from that of Indra and matchesthe one worn by two arms,wears a kiritatrikuta and the Visnus of Jaiyd,Nagara Sri Dharmaraja Oc-Eo. This figure holds a conch shell on the left hip, and wears a sash that falls in an arc over the dboti. Stylistically,Dr. Khan associates this slab with the sculptures carved under the Iksvdku unakonda.He tracesthe religiousinfluenceback to Mathurl dynastyin the 3rd centuryat Ndgdrj where the Mora Well inscription,dateableto the ist century B.C., refersto a temple of the Five This Vaisnava development in the Heroes (paficavira)where actual images were installed.38 by north was being felt in the religious practiceof the Andhradega the last quarterof the third as revealedby inscriptionsat Ndgdrjunakonda.39 century A.D., was It is not difficultto believe that the new force of religious conviction in the Andhradega communicateddirectly to the small kingdoms developing on the isthmian tract of the Malay Peninsula.There is a considerablebody of evidence provided by Buddhist sculpturethat links that the heritage the two areasin a system of artisticexchange.A. B. Griswold has demonstrated in the workshops of the isthmus in the 3th centuryderivedoriginallyin large of artisticpractice part from Amardvatior Ceylon, and that even after Gupta and Post-Gupta styles were superimposed, echoes of it can still be discerned.40 or The Amardvat! earlyCeylonstyle had a strong impacton SoutheastAsia, which appearsin numerous places including 0 Tong and other sites in Thailand which go back to the preDvaravatiperiod. It is now clearthat a vital civilization,developed in the Menambasin around the second or third century A.D., was subsumedin the kingdom of Dvdravatiin the late 6th work in the isthmus might well revealproto-historiccontact Intensive archaeological century.4I with Andhradesaof the kind now proved for 0 Tong. Epigraphy also points to early contact between the Malay Peninsulaand South India. The earliestinscriptions on the Peninsula are in South Indian script. They were not found in the isthmus but at Cherok Tekun in Province Wellesley, and at Bukit Meriam in Kedah about halfway between the Muda and Merbok Rivers. There is also the famous inscription of the

37

M. A. W. Khan, An Early Sculpture Narasi#ha, Andhra Pradesh Government Archaeological Series no. i 6 (Hyderabad: of
i964).

38 39
40

Ibid., p. 4; also Agrawala, Catalogue...,p. Ibid., p. 5.

A.B. Griswold, "Imported Images

III; Rosenfield, DynasticArts...,

pp. I5I-I52.

41

p. 57. ...," These new finds and their significance are discussed in the following: Jean Boisselier, "R6centes recherches arch~ologide ques en Thallande," Arts Adatiques,vol. iz (i965) and Nouvellesconnaissances archdologiquesla villed'U T'ong(Bangkok: x968), and H. G. Quaritch Wales, Dvaravati(London: i969). In Burma,a similarheavy impact of the Andhradegais apparentin the period between the ist and 5th centuries as revealed at by the important excavation of the early Pyu site at Beikthanomyo. See: Aung Thaw, ReportonExcavations Beikthano (Ministry of Union Culture, Rangoon: i968).

38

sailing-master found in the north of ProvinceWellesley.42 of these were found Buddhagupta, All by Colonel Low in the first half of the nineteenthcentury. The CherokTekun inscriptionis so worn that it is illegible, and the Bukit Merriamhas been lost, although "a not very satisfactory" hand copy of it survives.43 The Buddhaguptainscription,now in the Indian Museum,Calcutta, has been examinedby Professor B.Chhabra,who dates it in the 4th or yth century.44 A fourth inscription, found by QuaritchWales at Bukit Choras,in Kedah north of Kedah Peak, 45was ascribed to the 4th century by a curator of the British Museum on paleographic evidence. This inscription has been accepted in support of Dr. Wales's dating of a structure whose remains he uncovered there.46 But the early date for the inscription has recently been contested by Professor F. D. K. Bosch and Dr. J. G. de Casparis,and if their arguments are sustained,an 8th or 9th century date would be more likely.47 Summary On the basis of the evidence adducedabove, it would appearthat the Visnu from Jaiya(fig. i) is probably the most ancient Hindu image discovered in Southeast Asia. The 8th-centurydate generally ascribed to it must be too late by at least three centuries. It can be traced back to ultimate prototypes from the period of Kusana rule at Mathurd,though the most immediate stylistic influence seems to be from the 4th century art of the Andhradesa. In the iconography of the four-armedVisnu as it developed at Mathurd,the earliestimages have the anteriorright arm raisedwith the hand in abhayamudrd. padmawas evolved lateras The a distinctive symbol to be held in Visnu's hand.48 the basis of the disposition of the arms, On the Jaiya image (fig. i) may be earlierthan those at Nagara Sri Dharmar-ja(figs.2, 3), as the latter appearto have held the padma. Since Gupta images of Visnu are differentin style and iconographyfrom the Kusanaprototypes, the Jaiya image, which is unaffectedby these changes, should be dated no later than 400 A.D. The images in this style from Nagara Sri Dharmar-jaand Oc-Eo should then be dated in the yth century.49 The conclusionthat the Jaiyaimage is older thanthe closely similarVisnu from the Funanese port city of Oc-Eo is of considerableinterest. In the first half of the yth century,Fu-nanappears to have undergone a culturalrevolution which is described as a "second period of Indianiza42Colonel James Low, "An Account of several inscriptions found in Province Wellesley, on the Peninsula of Malacca," Miscellaneous Papers Relatingto Indo-China, vol.I (i886), p. zz3-zz5; reprinted from TheJournalof the Asiatic Societyof

43

44

45

46

47
48

Bengal(JASB) vol. XVII (I 848), pp. 6z-66; also, "On an Inscription from Keddah," Miscellaneous PapersRelating Indoto China,vol.1 (i886), pp.232-234, reprinted from JASB, vol.XVIII (i849), pp.247-249. Alastair Lamb,, "Kedah and Takuapa: Some Tentative Historical Conclusions," Federation Museums Journal (FMJ), vol.VI (196I), pp.78-79. "Expansion of Indo-Aryan Culture During Pallava Rule as Evidenced by Inscriptions," Journalof the Royal Asiatic Societyof Bengal,vol.I (I935), pp. i6-20. H. G. Quaritch Wales, "Archaeological Researches on Ancient Indian Colonization in Malaya,"Journalof the Malayan Branchof the RoyalAia/ic Society(JMBRAS), vol. XVIII, pt.I (I940), p. 7. Roland Braddell, "Most Ancient Kedah: Part I," Malayain History,vol.IV, no.2 (July I95 8), pp. 2 i-22. Alastair Lamb, "A Note on a Small Inscribed Stone Tablet from Dr. Wales' Kedah Site No. i," FMJ, vol.VII (i962),
pp. 67-68. Agrawala, Catalogue..., p. VII.

49

Malleret, ADM, vol.I, p. 395, had dated the Oc-Eo image to the first half of the 6th century.

39

tion". According to the Chinese accounts this shift in the intellectuallandscape came about who because the people of Fu-nan accepted as king a Brahmanfrom India named Kaund.inya He changed all the laws to conform to the system of India.5o is said to have come from P'anp'an, a kingdom that scholarsplace on the isthmus, on the Bay of Band n.5 This is, of course, where Jaiyi is located and where the image of Visnu in fig. i was found. Altogether, this constellationof evidence, text and image indicatesthat intensive explorationof the areaaroundthe Bay of Bindon might lead to the recovery of one of the most ancient kingdoms in Southeast Asia.

50 George Ccedes, The IndianiZedStates of Southeast Asia (Honolulu:

I968),

p. 56.

40

CHAPTER IV

THE VISNU FROM TAKUAPA * 0

Visnu from Takuapi (fig. I4) is a very impressive statue, well over six feet in height. When Lunet de Lajonquierefirst came upon it in I909, it was lying on a small hillock near the mouth of the Takuapi River on the west coast of the isthmian tract.' It was broken into several pieces, which had been collected and grouped around the base. Even in such a dismal condition it struckhim as possessing an aestheticqualitysuch as he had never seen in any of the nine hundred monuments he had visited in Cambodiaand Siam. His judgment is still valid. Despite the loss of the forearmsand most of the facialfeatures,and despite the fracturesof the body, it is perhaps the most powerful and emphaticallymonumentalsculptureyet discovered in Siam. The existence of such a statue raises a number of obvious questions. When was it made? Does the work reflect in any way the level of culturalachievementon the isthmus during the period of the earlycity-states,or was it imported?Unfortunatelythere is no text, inscriptionor archaeologicaldata that can be brought into definite associationwith the image. During his visit to Takuapl some twenty-fiveyearsafterLajonquiere's exploration,Dr. Quaritch Wales made an examinationof the find-site of the image. After clearing the rank undergrowth on the hillock, he was able to trace out the foundation of a small platform made of earthand laterite.It measuredabout twenty-fivefeet square,with a height of two feet. He also excavatedwhat appearedto be the remainsof a brick stairwayleading a short way down the hillside.zAside from this clearingoperation and tentative exploration,there has been no systematic excavationin the area. Whether it is possible to amplify the meagre record now availableis problematic.Takuap! is in the rich belt of tin stretchingalong the west coast of the isthmus, where the earthhas been worked by local miners since ancient times, and extensive tractshave been scoured by dredges in recent decades. It is almost certain that potentially valuable archaeologicalsites have disappearedunder the dredges. PresentDating In view of its surpassinggrandeur,it is surprisingthat the Takuapi Visnu (fig. I4) has received so little attention.The generallyaccepteddate for it, c. 6th centuryA. D., dependslargelyon the
I

the

Lunet de Lajonquiere,"Le domaine archeologique du Siam,"Bulletindela Commission Archiologiquede l'Indo-Chine (BCAI),

vol.1, I909, pp.I70-I7I.
8-9.

IAL, vol. IX, no. I, pp.

4I

work of Pierre Dupont whose opinion, formed almost twenty years ago, has found wide acceptance. He first discussed it in a discerning article which appearedin I94I, in which he studies a number of stone statues which had come to light in Siam, Cambodiaand the Mekong Delta region of what is now South Viet-nam,and which had severalcharacteristics style and iconoof graphy in common.3They were all standing figures of the god Visnu with four arms and bare chest; they all wore a plain dhotiwrapped tightly about the hips and securedat the waist by a knot; and they all wore a ceremonialheaddress,the kirtamukuta "mitre". or Dupont's views were refined somewhat in an article the following year. In it he stated his belief that the Takuapi image could be directlyattachedto the Pallavastyle of southernIndia, and suggested that its presence on the isthmian tract was due to the accidents of trade rather than to the development of any artistic tradition that could be identified as distinctive to PeninsularSiam.4 Indeed he took it to be the prototypefor the other mitredVisnus of Siamavailable to him for study at that time. These were from scatteredsites on the isthmiantract, as well as from Petburi (Bejrapuri)at the very head of the Peninsula,and the site of Dong Si Mahapot (Sri Mahibodhi) in Prdcinapuri Province in easternSiam. They were so closely analogousto the long-robed Visnu images of Cambodiathat artisticcontacts could be assumed. In his book on pre-Angkoriansculpture,publishedin I95 5, Dupont returnedto the problem. He took the TakuapaVisnu to be contemporary with the most ancientHindu images known in Cambodia,those belonging to the style of Phnom Da, which he dated in the first half of the 6th century. The tradition of the long-robed Visnu image spread from Siam to Cambodia,being propagatedin the latterareaat the beginning of the 7th centuryby the Khmer artistsof Chen-la, the kingdom that was the successorto Fu-nan.While there were successivemodificationsin the style, images of this basic type were produced throughout the 7th and 8th centuries.sDupont succeeded in establishing an orderly and convincing picture of stylistic development for the long-robed Visnu images, which he divided into three groups.6 In Group A he placed the images which (apartfrom fig. I4) he consideredto be the oldest. They wore a low mitre, which in most cases widened toward the top; the dhotiwas held in place by a "buckle" (really the upper hem gatheredinto a knot) at the waist; and a sash was wrapped horizontallyaround the hips. In this group he placed the following images found in Siam: two from Dong Si Mahapot,two from Vieng Sra (our figs. I 5 and i 6), one from Sating Pra (fig. i8) and one from Petburi (fig. 19).7 In Group B the mitre was cylindrical,or nearly so, and taller than in Goup A; in addition to the "buckle",the dhotiwas securedby a narrow belt tied in a bowknot; while a sash passed diagonally from the right thigh to the left hip in front, and back again in the rear. Group B included three images from Siam: two from Dong Si Mahapot,and one from SrivijayaHill in Province (fig. I 7). Surdstradhani In Group C there was no hip-sash; and the hands, instead of being attachedto the hips for
3 Dupont, "Visnu mitres de l'Indochine occidentale," BEFEO, vol. XLI,
4 2 (I94I),

pp.

233-254.

5
6

"Le Buddha de Grahi et l'Fcole de Chaiya," BEFEO, vol.XLII (1942), pp. I05-Io6. Dupont, Priangkorienne, I 28. p. BEFEO, vol. XLI, p.244. The provenance of the Vi~nu from Sating Pra (fig. i 8) was erroneously given as Dong Si Mahap6t (gr! Mahabodhi).

42

reinforcement,were held away from the body. Dupont considered Group C the latest of the from 650 to 8oo A.D.8 three, dating approximately The TakuapdVisnu (fig. I4) did not fit into either Group A or B. As Dupont knew of no other images of Group C from Siam it appearedto stand apartin type and qualityfrom all the images discovered in Siam. When Dupont wrote, only a limited number of them was known to him. Since then there have been developments that raise doubts about the antiquity of fig. I4. As we have seen in Chaptersz and 3, the traditionof four-armedstatues of Visnu in SoutheastAsia originatednot with fig. I4 but with statues like those in figs. I-4. This traditionhad a long life before the 6th century.9 PallavaAnalogies Let us reexamineDupont's view that the TakuapaVisnu is an almost pure reflectionof Pallava style, perhaps even the work of a sculptor trainedin southern India. He drew this conclusion because of similaritiesof costume. In Pallava art, Visnu wears a cylindricalmitre and a long dhotiquite like that of the Visnu from Takuapd;but a comparison of specific images shows that these similaritiesin the attachedpatternsof dress tend to divert attention from the very real plastic differencesbetween the isthmian sculptureand the Pallava tradition. south The anatomyof the armsin fig. I4 is differentfrom the prevailingpracticein pre-C61a India. At least on those four-armedstone images known to me, the arms are not separated above the elbow.xoThis factor can perhapsbe relatedto the Indian sculptors' concern for the support and stability of projecting parts. Fig. I4, on the contrary,has a striking separationof the arms at the shoulder,and the sharplyopposed angles of the armscreatea spatialambience that is much greater then that of the Indian images noted. The formal division and surface patterningof the torso is one of the salientstylisticfeaturesof fig. I4. The medianline, dividing the upper torso in half vertically,is clearlymarked.The torso is also divided horizontallyinto two distinctly demarcated zones. A slowly swelling convex plane, beginning at the clavicle (or where the clavicle would be if the sculptor had been interestedin articulatingthis feature)and extending in an arc to a point above the abdomen, is echoed by the swelling of the abdomen itself in a convex planewhich ends at the top of the dhoti (fig. I4b). Such surfacepatterningand concern with tectonic structureare not a mark of Pallavasculpture,in which the torso is commonly flat.",
8 Dupont, Priangkorienne, 132. p.

9 While this study is focused on the Hindu sculptures of peninsular Thailand it inevitably has connections with similar images found in the pre-Angkorian art of Cambodia. On the basis of the information now available, it seems to me that the Vigiu from Tuol Koh, ascribedby Dupont to the eighth century, is closely related to the ancient traditions of Viaiu with the conch shell on the hip. It still retains the sculptured jewelry that is characteristicof this earlier tradition, and probably dates no later than the early sixth century. Similarly,the Vigiu of Bathe, published by Malleret in ADM, I (pl. LXXXIII), is, as he suggested, also an early representation of the god. Both are clearly earlier than the image from Takuapd, which therefore cannot be the prototype for the long-robed Viaiu images of pre-Angkorian art. I0 National Museum, New Delhi, India: Museum Accession Nos. 59.I 5 3/I I4, 59.I 5 3/I I7, 59.1 5 3/i60, and Madras45 I I.I0. This holds true for all the four-armed images at the late Pallava temple at Tiruttani illustated in: Douglas Barrett, The Templeof Virattaneivara Tiruttani(Bombay, i958). This appears also to have been the standard practice among the at related sculptures produced by the Nolambas and the Pandyas. See: Douglas Barrett, Hemavati(Bombay, I958), and C. Sivaramamurti,Kalugumalai Early PandyanRock-CutShrines(Bombay, i96i). and I F. H. Gravely and C. Sivaramamurti,Guideto the Archaeological Galleries(4th ed., Madras, I960), p. 2 3.

43

A puzzling featureof fig. I4 is the sudden reductionin thickness of the upper right forearm at the wrist, as if it were cut down in order to be fitted into some attachment.Noting this singular feature, Lajonquiereconjectured that the hands may have been attached by metal collars.I2It has gone unremarked,and perhaps unnoticed, by subsequent writers. It is not apparentlya common feature in Pallava art, nor, for that matter, is it met with again on the Peninsula. Other dissimilaritiesin style may be mentioned in passing. Fig. I4 is severely unadorned, while Pallavaimages of Visnu usuallyshow the god with the sacredthreadand neck ornaments. The legs and arms of fig. I4 are taut with musculartension, and the deltoid, pectoral and calf muscles are even given a somewhat naturalisticmodelling. Pallava images are organized as a composite of abstractvolumes unified by a svelte and flowing contour line. The legs and arms tend to be tubular,without any definition of musculature. It could safely be said that fig. I4 would look out of place in the family of Pallavaimages. There may be some affinitiesbetween the two styles, but they would be the result of purposeful adaptive responses on the part of the Takuapdsculptor to artisticinfluencesfrom south India which seemed useful and relevantto his own pre-occupations.There is no evidence that he was committedto Pallavastyle in any sense, and certainlythere is nothing to indicatethat fig. I 4 was made by an Indian sculptor. If, as Dupont originallythought-and it should be noted that he implicitly revised this view in his last work on pre-Angkorianart-the image had actuallybeen made in India and later taken to PeninsularSiam,it should be a simple matterto place it in one of the Indian schools; but it is not apparentwhat Indian style it might be attachedto. On the other hand a number of reasons will be advancedbelow for including it within the art of the Peninsula,and regardingit as a signaldocumentin the explorationof plasticand spatialproblems on the part of the Peninsularsculptor. Most writers have taken it for granted that the style of the mitred Visnus from the PenOne reason seems to be the cylindricalform of the mitre or insula derives from the Pallava.13 This is a crown worn by Indra and Kuvera in early Indian art. It is met with in kiritamukuta. in representationsof Indra at Amardvatiand Ndgdrjunakonda the pre-Pallavaart of southern India.14In the Buddhist art of Gandhdra,Indra wears a basket-like kirita.'s In the Kusana to notes the presthe was sculptureof Mathurd kiritamukuta transferred Visnu. Coomaraswamy ence of the cylindricalkiritamukuta both Pallavaand Gupta art. For the Gupta, he cites the in Visnu images outside the Candragupta Cave at Udayagiri,Gwalior. He adds that this form of headdress"styled by French scholarsthe coiffure mitre [...] has reachedSiam and Cambodia, en where in 'Khmer primitif' art, in the sixth and seventh centuries,it is characteristic both for
Visnu and Indra."16

An example of the cylindricalmitre in northernIndia in early post-Gupta art is seen in the

I2

'4

I5
6

Lunet de Lajonquiere, "Le domaine arch~ologique du Siam," BCAI, I9I2, p. I70. Le May, TheCulture South-EastAsia, p. 8i. Groslier, TheArt of Indochina, 84. Wheatley, TheGolden of p. Khersonese, i96. p. Wales, TheMaking of GreaterIndia, p. 44. Philippe Stern, Mireille B~nisti, Jeannine Auboyer and Madeleine Hallade, Evolutiondu styleindiend'Amaravati(Paris, 196i), p. 54. For AmaravatIsee Plate XXVI in: Douglas Barrett,Sculpturesfrom Amaravatiin theBritishMuseum (London, 1954). For Nagarjunakonda,see Plate XII in: P.R. RamachandraRao, TheArt of NdgdrjunakoV4a (Madras, I956). See "The Visit of Indra," Fig.246 in: Alfred Foucher, L'artgreco-bouddhique Gandhdra du (Paris, I905). Ananda K.Coomaraswamy, "Early Indian Iconography," EasternArt, vol.I, no.i (July, I928), pp.40-4.

44

standingVisnu from Kanauj,now in the BarodaMuseum,which is said to date from about the 6th or 7th century.'7 The form of this headdressis quite similarin shapeto that of fig. I4, in fact much more so than those of early Pallava art, which are taller. I do not mean to argue that there is any relationshipbetween fig. I4 and the KanaujVisnu images, but only to point up the dangerin isolating a single feature,such as the headdress,without referenceto the stylisticcontext of the whole work of art, and drawing chronologicaland culturalconclusions on the basis of a similarfeatureon an image thousandsof miles away, existing in a differentsocial, political and ethnic context. Compellingargumentsfor a reassessmentof the date of fig. I4 are provided by a number of sculpturesthat have come to light in recent years. It will be apparentthat some of them are almost identical in type to the TakuapdVisnu. Dupont, of course, did not have this evidence before him when he formed his views about fig. I4. Sichon(Sdala) In i966 I was able to pay a brief visit to several interesting sites scatteredabout the village of Sichon, located on the east coast of the isthmus about thirty-eightmiles due north of Nagara The Sri Dharmardja. objects I examined,some in situand others said to have been found around Sichon, indicate that the area possessed an old and important culture. The pre-Thai objects show influences from a variety of traditions, including Dvdravati and Cambodia, and it is possible to date them from at least the 8th centurythrough the I 3th, afterwhich the traditions of Thai Buddhist art can be discerned.18 While the sites would probably yield a wide range of materialif properly excavated,I'one statue (fig. 22) is alreadyknown which sheds new light on fig. I4. It is a male figure with four arms,wearing a long dboti,and a headdresswhich resemblesthat of fig. I4. The armsare broken, but it is almost certain that the lost attributeswere the conch (sankba), club (gadd), wheel Like fig. I4, fig. 22 has no (cakra) and eitherthe lotus (padma)or the symbol of the earth(bbz7). reserves of stone at the hips, which in the images of Groups A and B were disguised by the knot of the sash and the handleof the club, and which functionedas part of a systemof supports to protect the projectingpartsof the image from being broken off. Instead,the lower armswere held away from the body, and apparentlysupportedby reservesof stone, the remainsof which may be seen on the base near the feet. This image, then, would fit into Dupont's group C, those long-robed Visnu images which do not wear a sash and whose arms are held away from the body.20Since the latterare the latest in the series of long-robed Visnu images in Cambodia,and since they are believed to date from between 650 and 8oo, we may assign a similardate to the Sichon image.
'7

18

U. P. Shah, "A Few BrahmanicalImages in the Baroda Museum," Bulletinof theBarodaMuseum, X-XI (April, 195 3 vol. March, 195 5), pp. I9-23, and Fig. io. For a detailed analysis of some of the objects from the Sichon area, and speculation on the historical background of the area, see: Stanley J. O'Connor, "Si Chon: An Early Settlement in Peninsular Thailand,"JSS, vol. LVI, i (January, i968), pp. i-i 8; see also, Suchit Wongthet, "WanderingArchaeology" (in Siamese), ChaoGrung, vol. XVI, 2 (November,
I 966).

I9

20

Already in May i966, this obscure site had been visited by persons who allegedly removed several objects presumably for commercial advantage. Dupont, Priangkorienne, I30-I32; pp. Malleret, ADM, IV, pp. I45-I46; Jean Boisselier, Le Cambodge (Paris, i964), p. 239.

45

SatingPra A second importantsite on the isthmusthat has only recentlycome to notice is the smallhamlet of Sating Pra located on the west coast some 20 miles north of Songkla. It was at one time an important trading and administrativecenter but was superseded by the modern town of Songkla. A wide range of objects, dug up from the soil in the region, have been transportedto Songkla where they are now under the supervisionof the Abbot and monks of Wat Majjhimavdsa.21 Among them are bronze images of Mahdyana and Hindu iconographyperhapsreflecting influencesfrom Sumatra,Java, Bengal and south India; a stone Ganesadateableto the 7th or 8th century; several long-robed Visnu images of similar date; and a lot of ceramics,many of which appearto belong to the late Sung and the Yuan dynasties.An object of specialinterestis a reliquarycontainerwhich, since it was used in sacredarchitecture, would indicate a tradition of monumentalarchitecturein the area of Sating Pra. In the present context, the most important object at Wat Majjhimdvdsa a broken stone is torso (fig. 2zIa,b). The head, arms and feet are missing. The upper torso is nude. The tightly fitting dhotifalls almost to ankle level. Its upper hem is fastened by being twisted on itself two or three times at the waist: what looks like an elliptical buckle is actually the twisted hem. Around the hips is a girdle renderedas a pair of narrowbands in relief which completelycircle the body, and tied in a figure-of-eightbowknot in front. Falling from the twisted hem at the waist is a verticalfold of cloth, made up of the pleatedlateraledge of the dhoti.It passes under the hip-girdle,gaining in mass and projectionas it descendsbetween the legs, doubtless to join the base which is now lost. In images of the long-robed Visnu, this verticalfold usually serves as one of the points of support to assurethe stability of the image on its base.22Fig. zi has or had several other features in common with the long-robed Visnu images: bare chest, double shoulder, four arms, long dhoti.It seems most probablethat it belongs to this series. It shares an importantcharacteristic with fig. I4: the absenceof fracturesat the hips shows that the lower arms were held away from the body. This then would extend the number of images qualifying for membershipin Dupont's Group C. No less importantare the controlled naturalismof fig. 2I and the sophisticationof its modeling, which place it nearfig. I4 in quality. Its archedback and swollen abdomen,while conforming to the Indian Yogic concept of inner breath or prcna, also show the depth to which the sculptorhas worked his block of stone. The facility of handling and the pronouncedpatterningof the musculatureon the chest are all suggestive of the preoccupationsthat reach full statementin fig. 14. Hua Kao Village,Surditradhjni Province The evidence availablefor the study of early Hindu sculpturein the isthmus was considerably augmented in i966 when Professors A.B.Griswold and Jean Boisselier photographed four previouslyunstudiedimages at Wat Srivijaya Hua Kao village nearthe Bay of Blndon. They at
21 The following articles deal with objects said to come from Sating Pra: S. J. O'Connor, "An Early BrahmanicalSculpture at Songkla," JSS, vol. LII, 2 (July, I 964), pp. I 63-I 69, and "Satingphra: An Expanded Chronology," JMBRAS, vol. XXXIX, i (i966), pp. I 37-I44; A. Lamb, "Notes on Satingphra,"JMBRAS, vol. XXXVII, i (July, I 964) and H. G. Q. Wales, "A Stone Casket from Satingphra,"JSS, vol. LII, 2 (July, i964), pp. zI 7-223. 22Except for the Sichon image (fig. z2).

46

are importantin showing that Visnu images of all three types (Groups A, B, C) were produced at one workshop, although presumablynot all at the same period. The badly mutilatedstatuein fig. 2 5 appearsto be a long-robed Visnu of Group A. It is bare chested; the closely-wrappeddhotiis tied at the waist; and fragmentary reservesof stone remain at the hips, aroundwhich a horizontalsash of the sort associatedwith Group A is wrapped.The statuewould thushaveaffinities with figs.I 5, i 6, i 8 and i 9. Two VisnuimagesfromDong Si Mahapotbelong to the same group.23 A second image at Hua Kao (fig. 23a,b) is clearly a Visnu of Group B, wearing a double girdle, and a diagonallyplacedhip-sash.It is very close in style to fig. I 7 from the sameprovince of Surdstradhani, was probably made in the same workshop. It is also similarin type to and two images from Dong Si Mahapot.z4 A third image at Hua Kao village (fig. 24za, is almost certainlya Visnu of Group C. It b) clearlyhas no reserves of stone at the hips to support the arms as in Groups A and B. It does not wear a sash at all, which is not needed as a pretext for a puffed-outknot on the right hip. It joins figs. zi and 2z in providing analogies of type with fig. I4. Anothet stone torso at Hua Kao (fig. z6 a, b) is unidentified.It is not a Visnu, and is reproduced here so as to bring it to the attention of scholarswho may wish to study it. Represented on the right leg is the head of a tiger and on the rear of the image is a paw. It would appear thus to be either a representationof Siva or Avalokitesvara.Dr. Lamb has suggested that representationsof Avalokitesvarawearing the tiger head and pelt may belong to the art of the empire of Srivijaya.2s Petburf Another image that ProfessorDupont did not have before him when he wrote is a four-armed Visnu from Petburi (fig. zo). It is in very good condition, being almost entirely presentedin a single plane without disengaging the limbs from the matrixof the stone block. As there are no reserves at the hips to reinforcethe hands, the image should be assigned to Group C. Against this classification may be objectedthat no such reservesareneeded becausethe figureis carved it almost completelyin relief; but it is only necessaryto note that in fig. I 5, which is carvedin the same manner,all the points of reinforcementare carefullyincluded, showing that the sculptor was making a rituallycorrect copy of the type of Visnu image which was orthodox in his day. The sameconsiderations would surelyhave governed the authorof fig. zo if he were not working at a time when the sash and reservesthat characterized Groups A and B were no longer a matter of prevailing style. It will be recalled that another long-robed Visnu was found at Petburi (fig. i9). This one clearlybelongs to Group A. It wears a heavy mitre that widens towards the top; and a sash is wrappedhorizontallyaroundthe hips. The hands, held at the hips, are attachednot only to the body but also to two of the supports- the club at the left and the cascadingend of the sash at the right. An unusualfeatureis the second sash, which falls in an arc in front of the body, and
23 24 25

Dupont, BEFEO, vol.XLI (I94I), pl.XXVII, A and B. Ibid., pl. XXIX A and B. A. Lamb, "A Note on the Tiger Symbol in Some Southeast Asian Representations of Avalokitesvara," Federation Museums Journal,vol. VI (I 96I), pp. 8 9-90.

47

which appearsto have been inspired by the draperyof the "aberrant" Visnu images which carried the conch on the hip (figs. I-4).26 The mitre too is unusual; though it has analogies with the large heavy mitres of Group A, it is convex on top and crowned with a protuberance.This is not found on any of the other long-robed Visnu images from Siam. It is, however, a feature of pre-Angkorianart, where it can be identifiedin the lintels of Tuol Ang, Vat Eng Khna, Tang Kasang and Vat Po Veal. Dupont dates the traditionto the 7th century.27 is also found on a It free-standingimage of Umd, which Dupont consideredto be the oldest image of the goddess in pre-Angkorianart.28 Fig. zo is later in time than fig. i9. There has been a loss of naturalismin the treatmentof the conch shell in comparisonwith fig. I 5. The mitre,which narrowstowards the top insteadof widening, sits low on the craniumand has pointed flapsat the temples. This type of mitre, and the presence of a moustache, are in conformity with developmentsin later pre-Angkorianart. By drawing analogies with the styles of PrasatAndet and Kompong Prah we may date fig. zo to the 8th century. As this image is one of the long-robed Visnus of Group C from Siam, we can be reasonablysecure in dating that group between 650 and 8oo, which is the same dating Dupont gave this group in Cambodia. Conclusion Recent discoveries have permitteda reassessmentof the date of fig. I4. It no longer appearsto be a work of the 6th centurypossibly importedfrom PallavaIndia, but insteada fully developed product of an isthmianworkshop between 650 and 8oo. If we follow the changes in the traditionof the long-robed Visnu image, we can discernan original impulse from the earliest Hindu art of the Andhradega,which was later modified by Gupta and post-Guptainfluences.This is particularly evident in figs. I 5 and i 6, which are close in style to the Funanesetraditionof Phnom Da and can be dated c. 6th century. It must be rememberedthat we look at the mitredVisnxu images without seing them with the realjewels and other detachableadornment which they almost certainlyonce wore. We have no inscription from Siamthat sets forth this custom, but the epigraphyof Cambodiashows that by the mid-7th century the practicewas common there.29 is, however, possible to find one miitred It Visnu of Group A in Thailandin which the elements of adornmentare shown in relief. It is in the collection of H. R. H. PrinceChalermpol Dighambara(fig. 27).3O When seen with its adornment,the mitre loses the superficialresemblanceto the Pallava style that is initially so persuasive an analogy. Its affinitiesare close to Gupta and post-Gupta art.3I The Visnu images of Group B, those with the tall cylindricalmitre, appearto reflecta response to Pallavaart of the 7th century, assimilatedinto an earliertradition.
In addition to the Vi$nu from Oc-Eo, this convention is also found on several images of Vishiu in pre-Angkorian art, for example those from Tay Ninh and Trapah Ven; see Malleret, ADM, IV, fig. 3 and PI.IV. It is also found on the Ganesa image from Triton, which Professor Malleret felt could not date any later than the first half of the eighth century: ADM, I, pp. 4I 5-417 andpl. LXXXIX. 27 Dupont, Priangkorienne, XXVI B, XXVI A, XXVI and pp. I 3 5-I 38. pl. 28 Ibid., pl. XXXVII A and pp. i 6 5-I 66. 29 Inscription of Ponhea Hor analyzed by Dupont, ibid., pp. 143-I46. 30 Masterpiecesfrom Private Collections, National Museum, Bangkok, I968. 31 For a Hindu sculpture from the isthmus exhibiting Gupta influence see: S.J.O'Connor, "An Ekamukbalihga from Peninsular Siam," JSS, vol. 54, I (i966), pp. 43-49.
26

48

It is possible to see the images of Group C, including fig. I4, as a development in which local preoccupationsare expressedby opening up the image so that it standsfreely and actively in space. The importanceof all this is threefold.First, it is possible now to see an exchangeof artistic conventions around the Gulf of Siam from the 4th or 5th century through the 8th. Second, in the development of these Visnu images we see the same kind of accretionof styles that Pierre Dupont traced in the development of the Dviravati images of the Buddha. Into a tradition traceableto the art of Amaravatiand Ceylon, Gupta and post-Guptaelementswere assimilated, while the memory of earlierstyles and iconography gives the Dvaravati images some of their unique characteristics.32 Third, we see in fig. I4 a prime document that suggests the qualitative achievement of the isthmian sculptor. It is the culmination of a local school rather than, as originally thought, the achievementof a Pallava sculptor of southern India, from which the more and more distant replicason the isthmiantract could be read as a devolution.33 Both the qualitativeachievementand the heterogeneousinfluencesto be discernedin a work like fig. I4 can be ascribedto the cosmopolitannature of the commerciallyoriented city states that sprang up at favored locations on the isthmus. The sophisticated forms of fig. I4 are tradecenteredat TakuaPa matchedby evidence which suggests that an importantinternational in ancient times. The areais endowed with severalattractivefeatures:abundanttin, one of the communication better anchorageson the west coast, and the possibilityof earlytrans-peninsular with areas of early settlement on the east coast. In earliertimes, the TakuapaRiver was navigable over the greaterpart of its course. From the head of navigationa short portage, perhapsfive miles, led to the headstream tributaryto the river systems flowing down to the Bay of Bandon. Local residents of the Takuapaarea were accustomedto making the round-tripfrom the town up to the watershedand over to Ban Sok on a branchof the Bandon River in one day.34 Lajonquierereportedthat the wrecked hull of a boat fourteen meters long was found near Pong at the upper reachesof the TakuapaRiver.3s It is not clearwhether it is identicalto the "anchorsand debrisof a sea-going vessel of respectable size" which W. W. Bourke reportedhad been found on the upper reachesof the river some fifty years prior to his visit in 1902.36 In I934 QuaritchWales made the journeyfrom Takuapa to the Bay of Bandonfollowing the rivers, and publisheda very full accountof his experience.37 He found the upper reachesof the river so silted that they could be navigatedonly by the smallest boats, and then only at the height of the rainy season. Ko Kao Island, at the mouth of the TakuapaRiver, provides evidence that Takuapdplayed an importantrole in the early internationaltrade between Western Asia, India and China. The island, the southernmostof three which affordshelterto the TakuapdRiver estuary,is roughly
32

33

34

Dupont, L'archeologie m6nede Dvdravati(Paris, 1959). This picture is at variance with the view of H. G. Quaritch Wales. He has examined my revised dating for the Takuapa Vianu as put forward in a short article, and he objected to it at some length. For his view see: "A Note on the Takuapa Visnu," JMBRAS, vol. XL, I (July, i967), pp.153-154. My reply is in JMBRAS, vol. XLI, i (July, i968). F. H. Giles and George Scott, "Remarks on the Land Route Across the Malay Peninsula," JSS, XXVIII, Part I (July,
1935), p.83.
(I 909), p. 2 57. i904), pp.

3 5BCAIC 36

G.E.Gerini, "Siamese Archaeology: A Synoptical Sketch," Journalof the Royal Asiatic Society(London,
245-246.

37

H. G. Q. Wales, Towards Angkor (London, 1937), and IAL, vol. IX, no. i, pp. i6-17. 3.

49

eight hundredyards off the mainland,and on its estuaryside it is borderedby a narrowband of mangrove swamp. The chief archaeologicalsite on K6 Kao Islandis at the southernend, where a smallfresh-waterstreamcuts through the growth of mangrove on the estuaryside, and at low tide its sandy bottom provides easy access to the site. Here, scatteredover a sandy plain several acres in extent, is a profusion of pottery and glass fragments,broken bricks, ceramicand glass beads. The pottery and glass fragmentsoccur in such quantitiesand are concentratedin such a small areathat they would far exceed the domesticneeds of any communityliving on the island. They are of diverse age and origin. Chinesestonewaresand porcellaneouswares are thoroughly and Middle Eastern of jumbled together with earthenwares either local or Indian manufacture, blue-glazedwares. Mixed in with the ceramicsis an amazingquantityof glass fragmentswhich glitter in the sun. During his visit to the island in I934, Dr.Wales collected some ceramicsand sent them to the British Museum where they were examinedby Mr.R. L. Hobson. The latter identified one ware as belonging to the Six Dynasties period (220- 589 A.D.), while the rest of the Chinese wares, in his view, fell within the limits of the T'ang dynasty, and the blue-glazed Middle Eastern wares were roughly contemporarywith them.38Dr. Lamb, who visited the site in February i 96i, collected some ceramicsherds, glass fragmentsand beads from the surfaceof the islandat the samesite which Dr. Waleshad explored.The sherdswere examinedby Mr. Basil Gray and Mr.Pinder-Wilsonof the British Museum, and, in their opinion, the Chinese sherds were not later than the T'ang period and not earlierthan the 7th century. Sherdsof blue-glazed ware appearedto be Middle Easternin origin, and were assigned tentativelyto the 8th or 9th century.39 In i963 I made a small collection of sherdsat Ko Kao Island and sent them to the Sarawak Museumfor Mr.Tom Harrisson,then Curator,to comparewith the excavatedmaterialfrom the great trading sites in the SarawakRiver delta. He was able to match almost all the sherdswith excavatedspecimensfrom Sarawak.He found some Ytieh sherdswhich could be pre-T'ang,but most of the wares could be fitted into types associated with the T'ang and Sung periods.40 Neither the blue-glazedware of the Mediterranean the abundantfragmentsof glass vessels nor occur in the Sarawaksites. It is clearthat Takuapa- an importantentrepot,with tradingconnectionsreachingto the was Middle East, India and China.Ships callingtherewould presumably offload cargo, revictualand refit. Quite likely they waited in the estuaryfor favorablewinds. All this activity would require warehousing facilities, at least a limited agriculturaldevelopment in the hinterland,and developed forms of economic and social organization.The Ko Kao site has offeredlittle evidence beyond the abundanceof trade goods. The only structurebrought to light is the foundation platform of a building which Dr. Wales concluded was a temple built of perishablematerials.41 Presumablywarehousesand domesticbuildings, if indeed therewas any settlementon the island beyond that essential for the maintenanceof commerce, were all constructed of perishable materials,and have left no trace.
38

IAL, vol. IX,

p. Io.

39 Alastair Lamb, "Some Glass Beads from Kakao Island, Takuapa, South Thailand," FMJ, vol.VI (i96i), p.48. 40 S. J. O'Connor and Tom Harrisson, "Western Peninsular Thailand and West Sarawak:Ceramic and StatuaryComparison," SarawakMuseum Journal,vol. XI, 23-24 (July-December, I964), pp. 562-566. 4' JAL, vol.IX, pp. II3.

50

Unfortunatelythe materialson the surfaceof the site have been thoroughly disturbed,and everything of intrinsic value that may have come to light has disappeared.There is thus no evidence of the kind of trade goods such as seals, inscribed gems, cloths with designs, or figurines, which could have had an impact on the formation of styles on the isthmus. Nevertheless, the Peninsula and especially a west coast entrepot called Kalah were well known to Most of these writtensourcesare subsequentto 85o, and the Arab geographersand navigators.42 trade at Takuapd,on the basis of the ceramics,was alreadyunder way by that time. And of course tradersfrom WesternAsia had been active in the trade centeredon Tun-sun as early as the middle of the 3rd century.43 Po-sst or Persiantradehad been a continuingfactorin early The These extensive Asian commerce,and in the rise to power of the Srivijayan Southeast Empire.44 contactswith WesternAsia may have stimulatedthe artistsof the Peninsula.At the very least, it is apparentthat the cultural ambience of the isthmian artist was subject to a wide range of influences. Any interpretationof the early culturalhistory of the Peninsulawhich ignores the importance of reciprocalrelationshipsbetween the neighboring countries of Southeast Asia tradewould and the subtle flow of influencesresulting from its pivotal position in international not be sufficientlycomprehensiveto have any prospect of accounting for the development of art styles.

et au G. Ferrand, Relations voyages textesgdographiques de arabes, persanset turksrelatifsa I'Extreme-Orient Vyule si'cle(Paris, 2 vols., I9I3 and I914). Paul Wheatley, "Arabo-Persian Sources for the History of the Malay Peninsula in Ancient Times," in MalaysianHistoricalSources (Singapore, i962). G. R. Tibbetts, "The Malay Peninsula as Known to the Arab Geographers," Malayan Journalof TropicalGeography, vol.VIII (I956), pp. zi-6o, and "Early Muslim Traders in SouthEast Asia," JMBRAS, vol. XXX, Pt. I (I95 7), pp. I-44; "Pre-Islamic Arabia and South-East Asia," JMBRAS, vol. XXIX,Pt. 3 (I956), pp. i82-2o8. 43 Wheatley, Golden Khersonese, p. i6. 44 See, 0. W. Wolters, "The Po-ssfi Pine Trees," Bulletinof theSchool Oriental of andAfrican Studies,vol. XXIII, Pt. 2 (1960).
42

5I

CHAPTER V

LATE PALLAVA AND COLA STYLE ON THE PENINSULA

Brahmanical sculpturein PeninsularSiam from the 9th to the i ith centuryis dominatedby
South Indian styles. The few images availablefor study are so close in style to known Indian examplesof late Pallavaand earlyCola art that they seem to have been made by South Indian artists,either on the Peninsulaor in India. This is an unexpectedphenomenon, since the Brahmanical sculptors of the Peninsulahad slowly developed both a technicalmasteryof their mediumand a stylisticvocabularywhich reflectedlocal and regional,not Indian,preoccupations. Although this intrusion of south Indian style representsa violent discontinuityin the course of Brahmanicalsculpture on the Peninsula, there is no indication that any scholar has ever regarded it as such., Of these later Brahmanical sculpturesthe only ones that have attractedmore than passing notice are the group on Pra Narai(Narayana) Hill at Takuapd(figs. 28-3 I)*. Over the past sixty years a small literaturehas grown up about them. Indeed their situationis so unusualthat they could hardlyfail to commandattention. They are embeddedin the trunk of a large tree located along the north bank of the TakuapdRiver about ten miles upstreamfrom the estuary. They are much revered by the local people, who have grouped other fragmentsof sculpture,and a stone slab bearing an inscription, around the tree, and enclosed the whole arrayin a low masonry wall. In December i963, Alec Moon, the local managerof the SiameseTin Syndicate,was supervisingthe erection of a permanentshelterfor these objects neartheir presentlocation, but out of the path of the approachingtin dredge. The statues and inscription first came to European attention in I902, when W. W. Bourke came across them in the course of his duties as a mining expert.2Colonel Gerini wrote a notice of them in I 904, but he had not personallystudiedthem, and was merely summarizing information received from Bourke.3It was not until the visit of Lunet de Lajonquiere,in i909, that a sketch was published.4Dr.Wales visited the site in I934, and subsequentlypublished the first photograph of these sculptures.sThey were illustratedand briefly discussed by Dr. le May in
Dr. Quaritch Wales, "Recent Malayan Excavations," JRAS (I 946), pp. 142-149, does refer briefly to several of these images, but only to support his conviction that sculpture had no development on the Peninsula, and that, as direct Indian influence waned, sculpture reflected a corresponding decline. Not recognizing the previous level of autonomy in the Peninsular images, the resurgence, or intrusion, of south Indian style does not exist as a problem for him. 2 W. W. Bourke, "Some Archaeological Notes on Monthon Puket," JSS, vol.11 (1905). 3 Colonel G.E. Gerini, "Siamese Archaeology: A Synoptical Sketch," JRAS, I904, pp.233-248.
I

4 BCAIC,

1909, p.

23

5.

5 IAL, vol. IX, no. i, Plate IV, and Towards Angkrr, p.48. * See p. 63.
52

1938.6 In i96i, AlastairLambtook severalsharplydetailedphotographsof them.7By comparing the sketch and the photographs,it is apparentthat the three images have undergonesome modification since Lajonquierefirst saw them. The smallermale figure has lost one of its arms, and its lower body is no longer visible. The centralfigurehas gained a head-dressmodelled afterthe one worn by the smallermale figure. One of its four armsis missing now, but the loss may have occurredbefore Lajonquiere's visit. In any event, one of its handsis now detachedand propped up against the tree. The female figure apparentlylost one of its breasts temporarily.It was replaced,but upside down. An identificationof the figures, all three of which have lost whatever attributesthey may have carried,is most difficult,perhapsimpossible. Lajonquiereconsideredthe centralfigure to be Siva. QuaritchWales agreed,adding that the three figures could be comparedto the Gaingadhara group at Trichinopoly.8Professor K. A. Nilakanta Sastri, though disagreeingwith Dr. Wales on a number of other questions relating to Takuapd,accepts the Gailgadhara comparison.9In any case it seems likely that the three figuresdo in fact constitute a group. They are all carved in high relief in the same material,a schistose limestone; the style is similar; and the differencein scale between them is doubtless hieratic,intended to emphasizethe importanceof the central figure. In the absence of any iconographic clue to the contrary,the Gaingadhara designation seems plausible. More importantthan the iconography of these images is the question of their style. Most scholarswho have written about them have agreed that they are similarto images from South India. They are usually said to be of Pallava style.IoIt is not evident however that they have been carefullyanalyzedto determinewhether they date from the Pallavaor earlyC6oaperiod. We shall base our comparisonson the centralfigure of the group (figs. 28, 29), since the other two present few clues to chronology. The sacred thread (yajiopavita) provides one indication of a Pallava date. It is decorated with a double bell-clasp and ribbons (fig. 29), an iconographic feature of the early medieval period in South India., Theyajgjpavita, however, does not flow over the right armas it does in Pallavaimages.I2It is composed of bands of pearls, stemming ultimatelyfrom the many early who succeededthe Calukyas the westernDeccan in 753 A. D. in Calukyanstyle. The Rastrakiitas carriedon the rivalry of the former dynasty with the Pallavas.During campaignsagainst the Pallavasin 775 and 803, the Rdstrakfitas introduced the style of intricatelyworked decoration into the dominions of the Pallavas.I3 The style of sacred thread in fig. 29 suggests the image no earlierthan the 8th century. should be dated A Visnu recently acquiredby the British Museum, attributedto the late Pallavaof the 8th
Reginald le May, A Concise Historyof Buddhist Art in Siam, Cambridge, 1938, p.42 and figs.41, 42. "Three Statues in a Tree: A Note on the Pra Narai Group, Takuapa," FMJ, vol.VI (i96i), Plates XXXVIII-XL. 8 IAL, vol.IX, no.i, p.15. o "Takuapa and its Tamil Inscription," JMBRAS, vol. XXII, Pt. I, p. 27. I Le May, The Culture of South-EastAsia, pp. 8o-8i; Wales, The Makingof GreaterIndia, p.45; Dupont, "Le Buddha de Grahi et l'6cole de Chaiya," BEFEO, vol. XLII (1942), p. Io. Sivaramamurti,"Geographical and Chronological Factors in Indian Iconography," AncientIndia,no.6 (January, 195o),
6 7

p.24.
I2

13

C. Sivaramamurti,IndianBronzes(Bombay, I 962), p. 9, and figs. 29 and 30. A. K.Coomaraswamy, Historyof IndianandIndonesian (New York, 1927), p. 102; Sivaramamurti,AncientIndia,no.6, Art P.37.

53

or 9th century, has fairly close analogies with fig. 29.I4 The yajnopavita and the double-bell clasp with ribbons are similar.The dress, while entirely differentin configurationfrom that of

fig.i9, has the samestiffappearance.

Another rathersimilarfigure, the Sarya from Kiveripikkam, probably dates from the 9th century.'sIts more elaboratecostume sets it apart from the stone sculpturesin the rock-cut temples at Ndmakkal.I6 These caves, located in the SalemDistrict of MadrasState, are believed to have been constructedaroundthe beginning of the 8th century.Fig. 29 is much closer to the gth-centurySiiryathan to the 8th-centuryfigures from NMmakkal. Whetherin fact fig. 29 belongs to the earlyCo1a,ratherthan the Pallava,will ultimatelyhave to be settled by a specialistin South Indian sculpture.Tentatively, a late Pallavadate seems to be indicated. The stiff draperyis a typical Pallava feature,17 the bow above the girdle is and another.18 The two swags of cloth falling from the hips are seen on many Pallavabronzes. The girdle of fig. 29 is simple comparedwith the profusion of detail and festoons of ropes of pearls which adorn the girdles of early C6la bronzes.i9The presumption of a Pallava date is given addedforce by a comparisonwith the sculpturesof the Virattenesvara temple at Tiruttani,which is dated by inscription to the last quarterof the 9th century.20 was built during the reign of It the last of the Pallava kings, Aparajita,who was overthrown by the Cola king, Aditya I, in 890 A. D.zzThe necklaceand armbands worn by the figuresat Tiruttaniare more elaboratethan those worn by any of the Pra Narai figures (fig.28-31). It would appear that the Pra Nargi group is earlierthan the Tiruttaniassemblage,and that a date between 73o and 850 should be assigned to it, though without any great confidence. Next to the statues is a roughly circularstone slab inscribed with Tamil characters.The inscription was first translatedin I914 by E.Hultzsch,22whose version Professor Cede's included in his collection of inscriptionsfrom Siam.23 According to Hutzsch the inscriptionrefers to the building of a tank by a person of royal descent whose first name ended in "-varman" 24 The tank was placed under the "protectionof the membersof the Matuigrdmam of the men and of the vanguard and of the cultivators."2s subsequent reading has been provided by K. A. A Nilakanta Sastri.26 believes that the inscriptionincludes the title of one of the later Pallava He kings, NandivarmanIII, who reigned from 826 to 850. The greaterpart of the inscription,according to this later version, would read as follows:27
Illustrated on the cover of OrientalArt, vol. VIII, no. i. Illustrated in: F. H. Gravely and C. Sivaramamurti,Illustrations IndianSculpture, of Mostly Southern (Madras, I960), P1. XXIX. 16 P. R. Srinivasan, "Sculptures in the Two Rock-Cut Vaitiava Cave Temples of Namakkal," Artibus Asiae, vol. XXIV, pt. 2, pp. I07-I I6. 17 Gravely and Sivaramamurti,Guideto the Archaeological Galleries,p. 23. I8 James C. Harle, "The Early Cola Temple at Pullamanigai," OrientalArt, vol. IV, no. 3, p. I03.
14

I5

I'
20 2I

Ajit Ghose, "A Groupof EarlyCola Bronzes,"Ostasiatische Zeitschrift (I934),

pp. 176-I86.

22

23 24 25 26 27

Jouveau-Dubreuil, Pallava Antiquities,Vol.II (Pondicherry, I918), p. 17. Barrett, The Templeof Virattanesvara pp. 3-4. ..., "Supplementary Note on a Tamil Inscription in Siam," JRAS (1914), Pt.I, pp.397-398. See also JRAS 337-339, for a previous version. Recueildesinscriptions Siam, vol.II (second edition, Bangkok, n. d.), Inscription XXIV, pp. 32-3 3. du Ibid., p. 33. Ibid.

(I9I3),

pp.

"Takuapa Its TamilInscription," and JRASMB, vol. XXII, pt.I (March,I949),

Ibid., p. 29.

pp. 25-30.

54

(and) called The tank dug by Naingfir-udaiyan Avani-narinan(is placed under) the protection the of the Manikkiramam, residentsof the militarycamp and ... If Professor Sastri has correctly untangled the referencesto Avani-ndrinan(Visnu on the earth)by associatingthis title with King NandivarmanIII, then the inscriptioncan be assumed to date from the second quarterof the 9th century. This date would fit comfortablywith the ceramicevidence on Ko Kao Island. Although there is no firm evidence that the inscriptionis relatedto the statues,its date would correspondroughly to their late Pallavastyle. The inscription may also provide a context for them, since the tank it refersto might possibly be associated with a temple. The most importantquestion, from the point of view of the art historian,has so far gone unasked in the discussion of the Pra Narii group. As we saw previously, Peninsularsculptors createdan image of Visnu at Takuapain the 7th or 8th centurywhich is regardedas one of the greatestachievementsof stone sculptureto be found on the mainlandof SoutheastAsia (fig. I4). It was the final solution of stylisticproblemswhich had occupied the isthmiansculptorsat least since the late 6th century.Why then, in the 8th or 9th century, should three stone relief figures appearat Takuapa,which in style and costume are almost classic statementsof laterPallavaart? If Peninsularsculptorswere so skilled, why were they not commissionedto make the Pra Narai images? They almost certainlydid not make them. There is nothing that leads up to the Pra Narai group, and no other Hindu image in the Peninsula which would appear to be even remotely relatedto it. Since the group can be fitted into a known Indian style with some ease, and since it shows no idiosyncrasiesof style or handling that would certainly accompany a copy of a Pallava original by a Peninsularsculptor, we may assume either that it was made in India, or that it was madeat Takuapaby an Indianartist.Eitherway, the sculpturessuggest that there was a significantIndian presenceat Takuapa. The inscriptionnow associatedwith the statues on the Pra Narai Hill sheds some light on a (Man. the problem.It refersto the founding of a tank by the Mavzigra-mam ikkiramam), powerful south Indian merchantguild, and to the existenceof a militarycamp. Professor Sastri,who was the first to identify the merchantguild in the inscription,has suggested that the inscriptionindicates that a good number of Tamils, including soldiers and merchants,were present at Takuapa,and that they probablyhad a permanentstakein the area.He furtherconjecturesthat the 9th centuryPallavaking, NandivarmanIII, may have extendedhis power to "partsof the west coast of the Malay Peninsula,at least for some years."28 Professor Sastrihas elsewhereassembleda body of inscriptionsrelatingto the South Indian merchantguilds.29 we placethe PraNaraiinscriptionin this largercontext, it seemsimprobable If that it indicates any political or militaryrole for the Pallavas on the Malay Peninsula. The inscriptions assembled by Professor Sastri show that the guilds were granted a measure of autonomyat home, and quite probablyreceivedsuch privilegesabroad.In additionto regulating trade, the guilds made donations of public works and religious monuments. The earliest of
28 29

JRASMB, vol.XXII, Pt. I, p. 30. K. A. Nilakanta Sastri, The Colas (2d ed., rev., Madras, I 95O), pp. 595-5 98.

55

these inscriptions in India, which dates from about 870, describes a grant for the support or maintenanceof a tank.30 this case there is no doubt that it was an irrigationtank, not a ritual In one, because the inscriptionwas found in situ at the northern end of an irrigationtank.3' Many of the other inscriptions relating to the merchant guilds refer to matters of local the administration, articles of merchandisein which the corporationstraded,the international character the trade,and, most often, the grantof money for the supportof religiousestablishof ments. An inscription of I05o A. D., in the reign of the C6la king Rajadhirdja records beneI, factions madeby one of the guilds to a South Indianvillage, and statesthat the guild was served Since it is certainthat the guilds were subject by regiments of foot-soldiers and swordsmen.32 to the authorityof the Cola kings, who were then almost at the zenith of their power, we must assumethat the soldierswere employedby the guilds to protect the securityof goods and warehouses. In this respectthey would be analogousto the industrialpolice found at modern manufacturing plants, the Pinkerton detectives, or the railroadpolice. This may suggest the nature of the Indianpresenceat Takuapa,and the significanceof the PraNaraiinscription.It is unlikely that the mercenariesof the Mavigrdmam were charged with any duties beyond protecting the propertyof the guild; Professor Sastrisays that the state did not give the guilds strong backing in their foreign trade ventures.33 statementmight be subjectto qualification view of the His in i ith-centuryC6la raidson the Srivijayan Empire; but in the context of many other inscriptions relatingto the guilds, nothing in the Pra Narai inscriptionis out of the ordinaryrange of guild activities,and it would not of itself indicatea political-military extensionof Pallavapower to the MalayPeninsula. This conclusionis reinforcedby the fact that the guilds endowed templesand left inscriptions abroad in areas in which, clearly, they were not allowed any more latitude than a limited autonomy over their daily commercialtransactions.An inscription found at Pagan in Burma, writtenin I 3th-centuryTamil characters, recordsdonationsmadeby an Indianfrom the Malabar coast to a Visnu temple built by one of the guilds at Pagan.34 there can be no question of Cola As sovereignty over Burma,the significanceof the inscriptionis probablythat, in preponderantly Buddhist Pagan, the Hindu merchantsfound it necessary and desirable to build their own temple. Similarly,the i2th or I3th-centurySouth Indian statuesdiscoveredin a Chinesetemple at Ch'uan-chou,on the Chinacoast opposite Formosa, do not indicateany Indianpolitical role 3 in China.But they suggest a parallelto the Pagantemple, showing that the guilds erectedHindu temples and statues in countries where the prevailing religion was Buddhism or some other non-Hindu religion. Erecting temples in foreign lands was by no means solely an Indian practice.During the II reign of the Pallavaking Narasirfhavarman (69-72Jo A. D.), a Buddhist temple was built at the South Indianport of Negapatamfor the use of ChineseBuddhistswho camethere to trade.36
30

S. R. Balasubrahmanian, "The Tisai Ayirattainurruvar and the MunicandaiRecord," Tijdschrift Indische voor Taal Land-en Volkenkunde, Deel LXXIV (I934), pp.6I3-620.
Ibid., p. 6 I 5. Sastri, The Colas, pp. 5 96-5 97.

3I
32

33 Ibid., P.598.
34

35 36

E. Hultzch, "A Vaishnava Inscription at Pagan," EpigraphiaIndica,vol. VII (I 902-I 903), pp. I 97-I 98. Sastri, The Ce~las, p.6o8. T. N. Ramachandran,"The Nagapattinam and Other Buddhist Bronzes in the Madras Museum," Bulletinof the Madras Museum,vol.VII, no. I (I954), p. I4.

56

It is well known that the Mahdyina Buddhist Pdla ruler of Bengal recorded in 850 A.D. the dedicationof five villages for the upkeep of a temple built at the famous college and monastery About a centuryand a half later, a Prince Bdlaputradeva.37 complex of Ndlanddby the Sailendra Since Negapatamwas one of the most built a temple at Negapatam.38 Sailendrarulerof Srivijaya important ports in South India, as well as one of the last remainingcenters of Buddhism in India in medieval times, it may be assumed that commerce, at least as much as piety, had atof to tractedthe Malaysand Sumatrans Srivijava form a colony there. They are believed to have been responsiblefor an ancient tower at Negapatam, which served as a landmarkfor vessels approachingthe Negapatam roadsteaduntil it was pulled down by European missionariesin 1867.39 to It is not at all surprisingfor the Mayigrdmam have had a tank dug at Takuapdin the 9th century, or to have maintainedmercenariesto guard their warehouses. The latter was very It likely a privilege grantedthe guild by the Empire of Srivijaya. seems likely that Srivijayacontrolled Takuapdin the 9th century,though we have no direct evidence to that effect.Kedah was a dependency of SrIvijayain the late 7th century, as we know from the Milasarvdstivddawritten by I-Tsing after his returnto Chinain 695,40but before that it had been ekasatakarman independent,at least it seems to have sent an embassyto Chinain 639.41 After seizing this outpost on the Straitsof Malacca,Srivijayaextendedits influenceto the east coast of the Peninsula, policy was as attested by the Ligor inscription of 775.42 The animatingprinciple of SrIvijayan to win and maintaina commandingposition in the Straitsin order to dominate the rich trade flowing between the ports of China,India, and the Middle East. It would scarcelybe compatible with this policy for SrIvijayato allow Takuapdto operate in uncontrolled competition with Kedah. The archaeological evidenceindicatesthat Takuapafar overshadowedKedah as an international trading center before the ioth century, for example the surface finds of ceramics at Takuapdsome of which appearto fall within the limits of the T'ang dynasty.43 The situationat the Kedah sites aroundthe MerbokEstuaryis different.Severalof them may date from the period when the entrepot at Takuapdwas flourishing, including some on the upper reachesof the Bujang River excavatedby Dr. Wales a little before World War II;44but for there is no evidence of any significantinternationaltradethere before the Sung period,45 the sites have yielded only an odd piece of T'ang pottery or earlyMiddle Easternglass; these places seem to have been engaged mainly in agriculture,though also carrying on some measure of
37

38K.V.Subrahmanya
I934), pp.28i-284.

Hirananda Shastri, "The Ndlandl Copper Plate of Devapdladeva," EpigraphiaIndica,vol.VII (1924), pp. 3I0-327. Aiyer, "The Larger Leiden Plates," EpigraphiaIndica, vol.XXII (I933-I934), also, pp.2I3-28i; R.C.Majumdar, "Note on the Sailendra Kings Mentioned in the Leiden Plates," EpigraphieIndica,vol.XXII (I933Ramachandran,Bulletinof the MadrasMuseum,vol. VII, no. I, p. I4. For a translation of this passage see: 0. W. Wolters, Early Indonesian Commerce, 227-228. pp. O. W. Wolters, "SrivijayanExpansion in the Seventh Century," Artibus Asiae, vol. XXIV, Parts 3-4, p.4I9. Ccedes, Recueildesinscriptions, vol.11, Inscription XXIII, pp. 20-24. See above, Chapter IV, p. 50. Dr. Quaritch Wales, "Archaeological Researches on Ancient Indian Colonization in Malaya," JMBRAS, vol. XVIII, Pt. i (February, I940). Alastair Lamb, "Some Notes on the Distribution of Indianized Sites in Kedah," Journalof theSouthSeasSociety, vol. XV, Pt.II (December, I959)' pp.99-Iii. Lamb has raised a number of objections to Wales' chronology, especially to the early dates he assigned the bulk of the sites on the upper Bujang. Several of these sites Wales assigned to fifth-sixth centuries, while the bulk of the others he dated from the seventh to the tenth centuries.

39
40 4I

42
43

44

45

57

trade with the shifting communities of the interior.46 Some 300 years after the beginning of internationaltrade at Takuapd,however, an entrep t pattern developed at Pengkalan Bujang (as the name indicates, a landing place on the River Bujang). In the area around it and in the river bed, archaeologistsfrom the University of Malayahave excavatedthousandsof fragments of porcelain,stonewareand earthenware, Middle Easternglass, fragmentsof glass bracelets,and glass beads; so there is little doubt that it was an entrepot with a trade involving goods from China, India and the Middle East;47but the pattern of internationaltrade there, though it is similarto that found at Takuapd,does not seem to have starteduntil the latter came to an end. These considerationsindicate that Srivijayacontroled the port of Takuapain the second quarterof the 9th century when the Pra Narai inscription was written. Any other conclusion would conflict with the known facts regardingthe structureof power and the policies of the Sumatran-based About the monopolistic objectives of Srivijaya thalassocracy. there can be little doubt. Its capabilities variedsomewhatthroughoutits history,but its intentionsremainedstable. In the 9th century, on the evidence supplied by Arab geographerssuch as Abu Zaid and the author of the Akhbaras-Sinwa'l-Hind,Srivijayacontrolled Kalah,the entrepot best known to the Arabs.48 Thereis a great deal of latitudein the variousArab referencesto Kalah,and nothing in the texts specificenough to allow an equationbetween Takuapdand Kalah.49 Srivijaya If was able to control the leading entrepotof the period,it is likelythat it maintained control, or at least surveillance,over any secondaryentrepots.Whetheror not Takuapdwas Kalah,it was certainly not a fugitive station for sporadicsmall-scaleattemptsto evade the monopolisticimpositions of Srivijaya.If it were, it is unlikely that stone statues of Hindu divinities of monumentalproportions, would have been erectedthere, or that an Indianmercantileguild would have invested money there in public works. This conclusion may help us to decide why, after some hundredsof years of sculpturalproduction on the isthmus and the development of stylistic autonomy, three purely Indian works should appearat Takuapd the 9th century.The answermaywell be that Mahdyana in Buddhism, under the patronageof Sailendras,monopolized the bulk of the artistic talent, the workshops and elite support on the Peninsula.This may mean that the workshop which producedfig. I4 in the 7th or 8th century was now exclusively occupied in producing sculptureto serve the Mahayana.It may be for this reasonthat the Man.igrdmam forced to build its own Hindu shrine, was and bring its statuesfrom India or have them made by an Indian sculptorat Takuapd;in other words, to do the same sort of thing the guilds did later in the non-Hindu communities at Ch'uan-chouand at Pagan. There is plenty of evidence to supportsuch a view. In the 8th centurytherewas a remarkable expansion of MahdyanaBuddhismin several of the countries of SoutheastAsia, which Ccedes describesas the dominantculturalfact of the century.so the isthmus itself the ratio of Hindu On statues,as opposed to Buddhist,swings heavilyin favor of the latter.In terms of artisticquality,
46 47

48

A. Lamb, "Kedah and Takuapa: Some Tentative Historical Conclusions," FMJ, vol. VI (i96i), pp. 79-81 . A. Lamb, "Pengkalan Bujang, An Ancient Port in Kedah," Malayain History,vol. VII, no. I (September, I 96i), pp. I 2-17. Translations from G. R. Tibbetts, "The Malay Peninsula as Known to Arab Geographers," Malayan Journalof Tropical

Geography, vol.VIII (2956), pp. zi-6o. 49 Historical geographers are in disagreement about the location of Kalah. For a summary of the literature see: Wheatley, The Golden Khersonese, 222-224. pp.
50

Cced~s, The IndianiZedStates ..., p. 96, and sources cited.

58

of as we shall see, Hindu sculpture,afterthe fortuitous appearance the Pra Narai group, takes a sharpdownturn. QuaritchWales says that at the sites he exacavatedaroundthe MerbokEstuary in in Kedah there seems to have been a trend away from Hinduism toward the Mahayana the The Ligor inscriptionof 775 commemorates foundationsof the second half of the 8th century.s5 i the erectionof shrinesto the Buddhaand two Mahayana divinities,the BodhisattvasPadmapdn. The main monumentof Wat Mahddhdtu Jaiyd,which may date from around at and Vajrapdni.S2 is the 8th century,53 thought to have been originallydedicatedto the Mahayana, so is the old and monument now hidden inside the main stupa of Wat Mahddhdtu Nagara Sri Dharmardja. at A great many bronze images with Mahayana iconographymust have been produced in the Peninsulafrom the 9th to the I 3th century,and some a great deal later.In the absenceof a more exact designation, they are all classifiedby the Siamese museums in a basket category called "Srivijayastyle". There are several dozen in the display cases of the National Museum at at Bangkok, about a dozen in the museum of Wat Majjhimavdsa Songkla, and others in monasteries at Jaiya and Nagara Sri Dharmaraja, to mention a quantityin private collections in not Bangkok and elsewhere.Most of them were found in the isthmianpart of the MalayPeninsula. Some of them are masterpieces,such as the two famous bronze statues of the Bodhisattva but Lokesvarafrom Jaiyd;s4 the majorityare small in size and indifferentin qualityas could be expected of cheaply-producedobjects of piety. In addition there are the Mahdydnistbronzes dredged up in tin mines of the Kinta Valley in Perakin Malaya,which may date from roughly the same period.s5s Most of the images, whether discovered on the Siameseor the Malayside of the border, and particularlythe best ones, have affinitieswith Indo-Javaneseart or with the bronzes producedfor the monasteryof Ndlanddin Bengal under the patronageof the Pala and Sena kings (73a-II97 A.D.). A bronze Guru figure, discovered a few years ago at Jalong in Perakand tentativelydated to the i ith century,might equallywell have been commissionedby a Hindu or a MahayanaBuddhist.s6 The religious syncretismprevailingin Java and Cambodia
5I

Wales, JMBRAS, vol. XVIII, pt.I (I940), p. 7I. Ccedes, Recueil inscriptions, II, Inscription XXIII. des vol. 53 Luang Boribal Buribhand and A. B. Griswold, "Sculpture of Peninsular Siam in the Ayuthya Period," JSS, vol. XXXVIII, pt. 2 (January, I 95 2). 54 Ccdeds,Les collections archeologiques Musie National du Bangkok,Plates XV-XVI, and XVII. du 55 Wales, JMBRAS, vol. VII, pt. I, Plates: 79, Avalokitesvara, from Bidor, Perak; 8o, Avalokite'vara from Sungai Siput, Perak; 8i, Avalokiteivara, from Sungai Siput, Perak. 56 A. Lamb, "Treasure Trove Among the Tapioca," Malqyain History, vol. VIII, no. i (December, i962), pp. I I-I 3; and, A.B. Griswold, "The Jalong Bronze," FMJ, vol. VII (i962), pp. 64-66. There is actually some uncertainty about both the style and the very simple iconography of this bronze. Part of the problem resides in the fact that-the Brahmin lacks the customaryjatdmakutaassociated with Guru figures, and there are almost traces of portraiture in the features. It is quite possible that it is not a straight-forwardpresentation to be seen in terms of Hindu iconography, but, instead, the portrait of an individual who has been deified as Agastya. Both Professor de Casparisand Mr. Griswold have arrived, independently, at the same conclusion. Since expert opinion has not yet resolved the problem of the Jalong bronze, it may be of some interest to point out the visual analogy, in the coiffure and simple adornment, between the Jalong figure and the very similar Brahmin portrayed on the extreme right of relief number XIII at the east Javanese bathing place at Jalatunda.This is the most ancient monument of the East Javanese period of art, and is dated 977 A. D. This monument, according to Dr. Bosch, has political-dynastic significance in the tangled web of tenth century diplomatic relations between Java, Cambodia, and Srivijaya. See: F.D.K.Bosch, "The Old Javanese Bathing-Place Jalatunda," reprinted in Selected Studiesin Indonesian (The Hague, i96i), pp.49-107; and Plate I 3. Dr. O.W.Wolters has explored the Archaeology configuration of inter-dynastic relationships of this period in connection with Cola raids of the eleventh century. In his reconstruction of events, the eastern Javanese would be the beneficiaries, and possible allies, in the Cola thrust at a Srivijaya-Cambodia-Tdmbralinga bloc. See: O.W. Wolters, "Tdmbraliiga," BSOAS, vol. XXI, Pt. 3 (995 8), pp. 596-5 97.
52

59

in the i ith century was probablyparalleledin the religious life of the Malay Peninsulaat that time.57 There is no reason to doubt that the Brahmanicalgods were still accorded a position of respect in the Peninsula, just as they have always been in Buddhist Burma and Siam; but it foundationsand statuary,doubtappearsthat the bulk of official support went into Mahdyanist less reflectingthe personalpreferenceof the elite and its control of patronage.It is possible, too, that one of the motives of the rulersin encouragingthe Mahayinawas to bring some measure votive tablets, Great quantitiesof Mahayanist of culturalunity into an areaof ethnic diversity.58 dating from the 8th or 9th century to the I3th, have been found in limestone caves at widely scatteredpoints in the interiorof the isthmiantract,and not confinedto religious establishments This patternof difmaintainedby the elite in the populous commercialcenters on the coast.59 fusion may suggest that the Mahayanahad a wide base and considerablesocial importance. ThreeColaImages Only three stone images of Hindu gods of later date than the Pra Narai group have been found on the isthmus. All three are small portablepieces, very close in style to South Indian sculpture of the i oth or i ith century. Two of them, a Visnu (fig. 3z) and a manifestationof Siva in the guise of Batuka-Bhairava (fig. 33),were discovered at Vieng Sra, the same site that yielded the two mitred Visnus of the 6th or 7th century (figs. I 5, i 6), as well as a small Buddha image, discovered by Dr. Quaritch Wales,60 which Griswold considers the work of a Peninsularartist working in the Sarnath traditionin the late 5th or early 6th century.6' The discovery of the Buddha image and of figs. I 5 and i6 at Vieng Sra suggests that the place was of some importancebetween the late 5th and the 7th century,while the discovery of figs. 32 and 33 suggests the samething for the ioth or iith. It is hardto say what its importance was based on. The site, about twenty miles southwest of Bandon, would be on the transPeninsularroute from Takuap! to the settlementson the Bay of Bandon, so the town may have functioned as a way-stationon it; but we have no evidence that this route was of any economic importanceas early as the iith century, not to mention the 6th. The site itself has yielded little else than the sculptures.Lunet de Lajonquieremappedit in I909.62QuaritchWales made a survey of it in I934, which confirmedthe general accuracyof
57 Ccrdes, TheIndianiZed ..., States
58

p. 146.

Inscriptions of ii83 and 1230 found at Jaiydare in the Khmer language but they are written in a script which is foreign in Khmer paleography. The first seems to be written in a Sumatran-likescript while the second has affinities with a Javanese script. George Ccedes,"Le royaume de ;rIvijaya," BEFEO, vol. VIII, no.6 (i9i8), pp. 33-34. Pierre Dupont remarkedthat Khmer may have been the language of diplomacy, in an area marked by a variety of languages and ethnic JSS, vol. 20. I (I 926), pp. 1-2 5. Votive Tablets," groups.See: BEFEO, vol. XLI (I 942), p. i o6. G.Ccdes, "Siamese 59 The tablets under reference are those of Mahaydnaiconography which Ccedesclassified as type II. Ccedes'ssuggestion of a tenth century date for this type may be somewhat conservative. Mr. Boisselier has drawn attention to the similarities between the tablets and Indo-Javanese sculpture of the seventh to the ninth centuries. See: Boisselier, La Statuairedu Champa,p.83. 60 Wales, IAL, vol.IX, Plate V. 6i Griswold, "Imported Images and the Nature of Copying in the Art of Siam", Essays Offered G. H. Luce, II, p. 62 f. to and fig, 22.
62

BCA IC, Lajonquiere,

i9i2,

pp. I39-I44,

andfig.29.

6o

The area of the old settlementis bounded on the south and east sides by Lajonquiere'splan.63 a narrow moat and a mound, and on the other sides by a lake and a stream. Both Wales and Lajonquiereconcluded that the only ancient site within the enclosurewas at a point in the very center of the area. Wales clearedthe site at this point and cut several trenches. He found the brick base of what he consideredto be a small shrine. There was no pottery present other than local earthenware.64 Figs. 32 and 33 are both about zo inches high and made of sandstone. Dr.Wales ascribed As them to the 8th century or earlier.65 they appearto be closely relatedto the C6la art of the ioth and i ith centuries,this dating seems to be in need of revision. a The Bhairava(fig. 33) wears a vanamdld, long string of bells, reachingdown almost to the The aureole of hair fanning out ankles in the fashion found from Early Co1atimes onward.66 of from the head is characteristic EarlyCola figuresof Bhairava.According to one authoritythis date between 850 and 1000.67 A similarhair style is worn by some images hair style indicates a I found in a temple built during the reign of the Cola king Paranataka (907-9 55).68 A rather close analogy with fig. 33 may be seen in an image excavatedin Ceylon, which dates from the period of C6oaoccupation of the island (993-1o70).69 Fig. 32 was apparentlycarved by the same sculptor as fig. 33. Both have the same sharply beveled planesin the articulationof facialfeatures;in both, there is the same over-alllinearand essentiallygraphicpresentation;in both, the sculptor has confined his modificationof the surface to a system of linearpatternsin relief, renderedin partby rows of small knobs. While there of is a purely descriptive reason for renderingthe vanamdld Bhairavain fig. 33 in this manner, the tail of the dog that standsbehindhim, as well as severalother portions of the samesculpture, are in part renderedin similarfashion. The same sort of knobs are deployed on the surfaceof the the the Visnu image (fig. 32), in the kiritamukuta, necklace (hdra), the bracelets(kauikana), spiral sash-format the right hip, and so on. Besides these similaritiesof treatment,the iconography of fig. 3z agreeswell enough with our ascriptionof a Co1adate to fig. 33. One indication of post-Pallavadate in fig. 32 is furnishedby Visius attributes,the iankhaand the cakra,which anotheris are displayedfrontallyratherthan edgewise, and perched on two upraisedfingers;70 The the sinuously-curvingyajgopavita, composed of three strands.7" post-Pallavafeaturesvisible in both figs. 32 and 33 suggest that there is no furtherreasonfor acepting the date proposed by Dr. Wales. Tentatively, until they are studied by a specialiston South Indian art, both may be ascribedto the ioth or ii century. The thirdimage to be discussedis a small Siiryafound at Jaiyd(fig. 34), which has apparently not been previously published. It is about 26 inches high, and carved in a reddish stone. The feet are missing, but the anklets show that the figure was not wearing boots. North Indian images of Suryausuallywear boots; and in Pdlaand Senaart this god is generallyrepresentedin
63 64 65 66 67 68 69
70 71

IAL, vol. IX, pp. I 7-I 9. Ibid. Ibid. in James C. Harle, TempleGateways SouthIndia (Oxford, I 963), footnote no. I, p. i 8. J. E. van Lohuizen-de Leeuw, "The Protector of the Mountain of Truth," Artibus Aiiae, vol. XX, Pt. I, p. I 5. Ibid. Archaeological Survey Ceylon,i960, PI. i A and B., and pp. 74-7 5. of Sivaramamurti,Ancient India, no.6, p. 24. Ibid., p.29.

6i

relief againsta stela, which is often crowded with accessoryfigures.72 South India, by way of In contrast, Sirya usually stands barefoot on a pedestal, without attendants,and with his hands raisedin front of his chest-all of which is, or was, true of fig. 34. There is sufficientcorrespondence between this image and an earlyCola bronze of the i ith centuryin the MadrasMuseumto indicate a similardate.73 The presence of figs. 32-34 on the isthmian tract of the Malay Peninsulawhere they were discovered should be associatedwith the continuum of culturaland commercialrelationsthat existed for many centuriesbetween the isthmus and the Coromandelcoast of India.74 The comwere well structuredin the entrepot trading centers of the 9th century,as mercialrelationships the Takuapdinscriptiondemonstrates.With the steady volume of internationaltrade geared to the demands of the Sung dynasty,'7these contacts must have intensified.We know, from the thatone of the greatIndianmerchantguilds was operatinscriptionof Loboe Toewa in Sumatra, ing at that place in io88 A. D.76Professor Sastri conjecturesthat the Tamils of Loboe Toewa may have erecteda temple like those of South India of the Colaperiod and placedinscriptionson its walls.77We have alreadynoted the activities of the Manigr mamat Takuapd. Certainevents which occurredbetweenthe erectionof the 9th-century inscriptionat Takuapa and that of the Loboe Toewa inscription in io88 illustrate the importance of the economic contacts of which those two inscriptionsare symptomatic.The temple built by Srivijayaat the port of Negapatamin I005 was doubtless intended to serve the Srivijayantrading community It there, but it may also have been an instrumentof diplomacy.78 would be an advantage,perhaps even a strategicimperative,to neutralizethe growing sea power of the Colas, so that Srivijaya could allocateits resourceseffectivelyin its prolonged struggle againstJava. The Colas, on the other hand, must alreadyhave been restive under the continuing commercialimpositions and monopolistic practicesof Srivijaya.In any event they sent their first mission to Chinain IOI5, perhapssignalling their interest in a greaterrole in the developing Sung maritimetrade. Two years later they may have launcheda raid on Srivijaya.79In I025 Rdjendra Co1asent a force to attack the capital of Srivijayaand many of its Peninsularpossessions. Having made his point, he apparentlywithdrew his forces without attemptingto impose a territorialadmninistration.8o According to a South Indian inscription, Virardjendra conqueredKaddramaround i o68 A.D. on behalf of a king who sought his aid and protection.81 This event remainsobscure, and there
Banerjea, TheDevelopment Hindu Iconography, of p.439. Illustrated in Sivaramamurti,IndianBronzes,Pl139. 74 N. K. Sastri, "Kataha,"JGIS, vol.V, Pt.2 (1938), pp. i28-I46. 75 During the eleventh century, the Sung dynasty became increasingly dependent on the profits from maritimecommerce, and made efforts to stimulate the level of international commercial activity. See: Paul Wheatley, "Geographical Notes on Some Commodities Involved in Sung Maritime Trade," JMBRAS, vol. XXXII, pt. 2 (June, I959) pp. I-I40, and, especially, the table on p. 24, illustrating the growth in revenue derived from maritime trade from the founding of the Sung dynasty in 960 to ii89 A.D. 76 K.A. Nilakanta Sastri, "A Tamil Merchant Guild in Sumatra," Tijdschrift, Deel LXXII (I932), pp.3I4-327. 77 Ibid., p. 326. 78 Ayer, "The Larger Leiden Plates," EpigraphiaIndica,vol. XXII, p.229. 79 R.C. Majumdar, "The Overseas Expeditions of King RdjendraCola," Artibus Asiae, vol. XXIV, pts. 3/4, pp.338-342. Majumdar'sinterpretation is disputed by Sastri, who considers that Rdjendralaunched only one raid, that of the fourteenth year of his reign, which would be I025 A.D. 80 Sastri, The Colas,pp.2i8-220. 8i Ibid., pp. 27I-272; Wolters, BSOAS, vol. 2I .3 (I958), pp. 587-607; David K. Wyatt and John Bastin, "MainlandPowers on the Malay Peninsula, A.D. IOOO-I500," Paper presented to the International Conference on Asian History, 5-IO August I968, University of Malaysia, Kuala Lumpur.
72 73

6z

is no reasonto believe that it was anythingmore than a passing affair.The tangledinternational rivalries on the isthmian tract between the iith and the i3th centuries could hardly fail to involve the interests of Ceylon, Burma and Cambodia.Eventually relationsbetween Srivijaya and the C6las returnedto a more orderlypattern.This we gatherfrom the smallerLeiden Grant of the C6la king KulottuAgaI (r.I070-I I20), which records the arrival of two ambassadors around i090, who were given a copper-platecontainingthe names of the villages from Srivijaya granted by the C6la kings to the Buddhist temple at Negapatam.82 The memoryof the Cola Empire survivedlong afterthe empireitself disappeared. According to Chinese sources the Kings of Siam associatedthemselves with the past glories of the Cola kings.83 The isthmianpeople, in their turn, may have had some impacton South India. Standing images of the Buddhawith the two handsperformingseparategestures,which occur in the arts of Dvdravat!and Srivijayaaroundthe 8th or 9th centurythough unknown in India at that time, and a monk who was active in the reform of occur later on among the Negapatambronzes;84 Buddhismin South India in the i 2th centuryappearsto have had some connectionwith Nagara Sri Dharmardja.85 In this context figs.32, 33 and 34 provide further evidence of the close and continuing contacts between the isthmus and South India in Cola times. While it is possible that these images were carried to the isthmus by Co1araiders, it seems much more likely that Indian merchants,possibly middlemen in the lucrative trade in gharu-wood, commissioned them for their own temple, just as Indianmerchantsdid at Ch'uan-chouand at Paganin the i 2th and I 3th centuries;or else they may have been made for the communityof Brahminswho lived in these in city-stateslike the Brahminswho performedritualsfor the rulersof Nagara Sri Dharmardja the i9th century, and whose descendants still live there.86 These images, wherever they may have been made, are stylisticallya product of Indian art, not isthmian.After the 8th century,so far as we can tell on presentevidence, the Hindu gods were no longer a primaryconcern of the isthmian sculptors.

82

Cced~s, The IndianiZedStates . . ., p. I 58. Paul Pelliot, "Encore a propos des voyages de Tcheng Houo," T'oungPao, vol. XXXII, pt.4 (i936), pp. 2x 6-217. 84 Dupont, L'Archdologie m6ne,p. x 84, and footnote no. i. 85 S. Paranavitana,"Negapatam and Theravdda Buddhism in South India," JGIS, vol. XI, pp. I7-25. 86 JRASMB, vol. XXXII. There is a line of inscription on a piece of granite at Nagara Sri Dharmardja(see Chapter I, footnote 28) and, according to Mrs.Devahuti, a bronze Gane'a has been discovered there with a Tamil inscription: India andAncientMalaya (Singapore, I 965), p. 59.
83

* After this book had gone to press, it was learned that thieves had recently carried off the heads of the three statues in

the tree at Takuapd. The remaining fragments of the statues have been removed to the regional office of the Department of Fine Arts at Nagara Sri Dharmardja.

63

BIBLIOGRAPHY

AND ABBREVIATIONS

AA: Arts Asiatiques.Musee Guimet, Paris AAS: ArtibusAsiae. Ascona du ADM: Archeologie deltadu Mekong,3 Vols, by L. Malleret.Paris, 1959-1960 BEFEO: Bulletin de l'Ecole Franfaise d'ExmrdmeOrient.Paris de Archeologique BCAI: Bulletin de la Commission Paris l'Indo-Chine. tot Bidragen:Bijdragen de Taal-, Land-en Volkenkunde. The Hague of BSOAS: Bulletinof the School Orientaland African Studies.London publi6es a l'occasion du vingtEA: Etudesasiatiques cinquieme anniversairede l'Ecole Frangaised'Extreme-Orient,z Vols. Van Oest, Paris, I925 Indica.New Delhi EI: Epigraphia FEQ: Far Eastern Quarterly. New York. After September, I956, known as TheJournalof Asian Studies Kuala Lumpur Museums Journal. FMJ: Federation IAL: IndianArt andLetters.London JA: JournalAsiatique.Paris Ann Arbor, Michigan of JAS: Journal Asian Studies. JAOS:Journal of the AmericanOrientalSociety.New Haven, Connecticut Calcutta of of JASB: Journal the Asiatic Society Bengal. Malay States MuJFMSM: Journalof the Federated seums,Taiping and Kuala Lumpur. JMBRAS: Journalof the MalayanBranchof the Royal Singapore Asiatic Society. of the Royal Asiatic Societyof Great JRAS:Journal BritainandIreland.London JRASMB:Journal of theRoyalAsiatic Society, Malayan Kuala Lumpur Branch, Bangkok of JSS: Journal theSiamSociety. Geography. of Journal Tropical MJTG: TheMalayan
Singapore

Berlin Zeitschrift. OZ: Ostasiatische des RAA: Revue Arts Asiatiques.Musee Guimet, Paris Journal.Kuching SMJ: SarawakMuseum Taal-, Land- en Volkenvoor TBG: Tddschrift Indische Genootkunde uitgeven hetKoninkljkBatavia-asch door Batavia, 's Graen schapvanKunsten Wetenschappen. venhage Ackerman,James S. "WesternArt History," Art and Archaeology, J. S. Ackerman and R. Carpenter. by New Jersey, I963. the Admiralty Pilots. Bay of BengalPilot: comprising the coastof coastsof Ceylon, eastern and southern eastern coastof ThaiIndia,thecoastof Burma,andthewestern landfrom the PakchanRiver to Goh Puket; also the 7th and Andaman NikobarIslands. ed., London, I 940. the - China Sea Pilot: comprising western shoreof the on side ChinaSeafrom Tonjong Penawar, theeastern of Jahore,to Fokai Point, Bias Bay; HainanIsland;also the Singaand bordering mainroutefrom theislands banks

znd Kong. ed., London,I 9 5I. poreStraittoHong

- Malacca Straitandits Malacca Strait Pilot: comprising Strait, andthewestcoast Singapore northern approaches, of Sumatra.3rded., London, I 946. Aiyer, K. V. Subrahmanya. "The Larger Leiden pp. 2I3-28i. Plates," EI: Vol. XXII, I933-I934, of Agrawala,Vasudeva S. A Catalogue theBrahmanical
I in Art. Images Mathura Lucknow, 9 5 I .

Art. London OA: Oriental

Album of Art Exhibits, National Museum Bangkok, Vol. I. Bangkok, (n. d.) Annual Reports. Survey Ceylon, of Archaeological Colombo. Aurousseau, L. "Nouvelles fouilles de Dai-Huu (Quang-Binh, Annam); une fouille au village de Trung-quan,"BEFEO, Vol. XXVI, pp 358-3 69. en Bagchi, Prabodh Chandra. Le Canon bouddhique et Vol. I, Paris, Chine:les traducteurs les traductions. I927. Vol. II, Paris, I938.

64

S. Balasubrahmanian, R. "The Tisai Ayirattainfirruvar and the MunicandiRecord," TBG, Deel LXXXIV,
I934, pp. 6I3-620.

Peninsula," FEQ, Vol. IX, No. 3, I950, pp. z56305- TheAncient Khmer I Empire. Philadelphia,95 I. Burkill, I. H. A Dictionary the Economic Products of of

Banerjea,J. N. The DeveloppmentHindu Iconography of Calcutta,I946. in Banij, Dhammadasa.A Historyof Buddhism Chatya andtheBay of Bandon Region. Chaiya,I96I. Thai and English. Barrett,Douglas. Hemavati.Bombay, I958. - Sculpturefrom Amardvat! theBritishMuseum. in London, I954. - The Temple Virattanesvara Tiruttani.Bombay, at of
I95 8.

theMalay Peninsula.Vols. London,I93 2

5.

New Batten,Mark.Stone Sculpture theDirectMethod. by York, I9 5. Buddhist Beal, S. Travelsof Fab-hian and Sung-yun, Pilgrimsfrom ChinatoIndia(400A.D.andy i 8 A.D.).
London,, 1911.

Carpenter,Rhys. GreekSculpture: Critical Review., A Chicago, i960. Casparis, J. G. de. Inscriptiesuit de 7ailendra-tjd, Pralasti Indonesia Bandung, I950. I, - Pralasti Indonesia Selected II, Inscriptions the7th from to the 9th Century A.D. Bandung, 1956. - "Historical Writing on Indonesia: Early Period," Historiansof SouthEast Asia. Edited by D. G.E. Hall. London, i96i, pp.i2i-i63. - "New Evidence on Cultural Relations Between Java and Ceylon in Ancient Times." AAS, Vol. XXIV, I96I, pp. 241-248. - "The Date of the Grahi Buddha,"JSS, Vol. LV, i,
i967, PP
30-40.

Bhattacharya,Kamaleswar.Les religions brahmaniques et dansl'ancienCambodge d'apresl'6pigraphie l'iconographie.Paris, 196I. Blumel,Carl.Greek Sculptors Work. at New York, I9 5. Boisselier, Jean. La Statuairekhmereet son 6volution Paris and Saigon, 195 SS - "La statuaire pr~angkorienneet Pierre Dupont", AA, Vol. VI, Pt. I, I959, pp. 59-69. - "Le Visnu de Tjibuaja (Java occidental) et la statuaire du Sud-Est asiatique," AAS, Vol, XXII,
I959, pp. zio-zz6.

Casson, S. The Technique Earbl GreekSculpture. of Ox-

ford, 1933.
Chatterji,S. K. "A BrahmanicalDeity in Indo-china and Indonesia," J.N. BaneroeaVolume.Calcutta, 1960, pp. 86-92. Chhabra,B.Ch. "Yiipa Inscriptions," India Antiqua,
I947, pp.77-82.

- La statuaire Champa. du Paris,I963. Bonner, Alice. Principles Composition HinduSculpof in

- "Expansion of Indo-Aryan Culture during Pallava Rule as Evidenced by Inscriptions,"JRASB, Vol. I, I935, pp. I-64. Chhibber,H. L. TheMineralResources Burma.Lonof

don, 1934.
Christie, Anthony. "The Ancient Cultures of IndoChina," Asian Culture,Vol. II, No. 2, April-June,
pp.5I-71.

ture.Leiden,I962.
Bosch, F.D.K. "The Old Javanese Bathing-Place Studies Indonesian in Jalatunda",reprintedin Selected The Archaeology. Hague,I96I, pp. 49-107. Bourke, W.W. "Some Archaeological Notes on Monthon Puket",JSS, Vol. II, I 90 5. Braddell, Roland. "Notes on Ancient Times in Malaya (I): Prehistoryand Protohistory",JMBRAS, Vol. XX, Pt.I, I1947, I 6I-I8. pp. - "The Ancient Bead Trade,"JMBRAS, Vol. XX, Pt. 2, 1947, pp. I-I0. - "Takola and Kataha,"JMBRAS, Vol. XXII, Pt.i,
I949, pp. i-i6.

- "Malayadvipa: A Study in Early Indianization," MJTG,Vol. IX, I956, pp. I-20. Braddell, Roland. "Most Ancient Kedah: Part I," Malayain History, Vol. IV, No. 2, July, 1958, pp.
I8-40.

Briggs, P. L. "The Khmer Empire and the Malay

"An ObscurePassagefrom the Periplus,"BSOAS, Vol. XIX, I957, pp. 345-353 . "Some Writings on South-East Asia Pre-History," Historiansof South-EastAsia. Edited by D. G.E. Hall. London, I 96 I, pp. I 07-I 20. "The political use of imported religion: an historical example from Java," Archivesde Sociologie des Vol. 9, I7, i964, pp. 53-62. religions, Claeys, J. Y. "L'archeologie du Siam," BEFEO, Vol. XXXI, I93I, pp. 36I-448. Coedes,George. "Le royaumede rivijaya," BEFEO, Vol. XVIII, No. 6, I 9 I 8, pp. I-36. "Tablettes votives bouddhiques du Siam," EA, Vol. I, I925, pp. 144-i67. "Excavations at P'ong Tfik in Siam," Annual of Bibliography IndianArchaeologv, 1927, pp. i6-zo.

65

du archdologiques MusdeNational de Les collections

Paris,I928. Bangkok.
- "Excavations at P'ong Tuk and their Importance for the Ancient History of Siam,"JSS, Vol. XXI, Pt. 3, I928, pp. 195-210. - Recueildes Inscriptions Siam. Vol. II. Bangkok, du
1929.

- "lRtudes cambodgiennes:deux inscriptionssanskrites du Fou-Nan," BEFEO, Vol. XXXI, 1938, pp. i-8. - "Notes sur quelques sculptures de grideb (Siam)," Linossier.Vol. I. Paris, 1932, ]8tudesd'orientalisme,
PP.I59-i64.

continental,"Bulletinde la SociNt6 .tudes Indochides noises, Vol. XXX, Pt. 4, 1955, pp. 365-3 82. D'Ancona, Mirella Levi. "Amaravati, Ceylon and Three Imported Bronzes," Art Bulletin,Vol. 34, March, 1952, pp. 1-17. Desai, M.D. "Some Roman Antiquities from Akota near Baroda," Bulletinof the BarodaState Museum, Vol. VII, Pts. I and II, 1949-1950, pp. 2I-24. Devahuti, D. India and Ancient Malaya. Singapore,
I965.

- "A propos d'une nouvelle theorie sur le site de JMBRAS, Vol. XIV, Pt. III, December, ?;rivijaya,"
1936,pp.i-9.

Diskalkar,D. B. "Some Brahmanical Sculpturesin the Mathura Museum," Journalof the UnitedProvinces HistoricalSociety, 5, Pt. I, 932-1934, pp. i 8-5 7. Vol. Diskul, M.C. Subhadradis. "Mueng Fa Daed, an Ancient Town in Northeast Thailand,"AAS, Vol.
XIX, Pts. 3 and 4, pp. 362-367.

- Review of H. G. Q. Wales, Toward Angkor, BEFEO, XXXVIII, 1938, pp.309-314. - "Quelques precisions sur la fin du Fou-Nan," BEFEO, Vol. XLII, 1942, pp. 1-11. - "Fouilles en Cochinchine. Le site de Go Oc Eo, ancien port du royaume de Fou-Nan," AAS, Vol. X, Pt. 3, 1947, pp. 190-199. - Les EStatshindouisis Paris, et d'Indochine d'Indondsie. 1948; New Edition, I964. - "Un Yantrarecemmentdecouvert a Angkor,"JA, Vol. CCXL, Pt. 4, I952, pp. 465-477. - "Nouvelles donnees epigraphiquessur l'histoire de l'Indochine centrale,"JA, Vol. CCXLVI, Pt. 2,
1958,pp.125-142.

Dupont, Pierre. "La statuaire en ronde-bosse dans l'Asie du Sud-Est," Revue Arts Asiatiques,Vol. des X, No. II, June, 1936, pp. 97-106. - "Visnu mitres de l'Indochine occidentale," BEFEO, Vol. XLII, Pt. 2, 1941, pp. 233-254. - "le Buddhade Grahiet l'ecole de C'aiya,"BEFEO, Vol. XLII, 1942, pp. 105-io6. - "La dislocation du Tchen-la," BEFEO, Vol. XLIII, I943, pp. I7-5 5. - "Recherchesarcheologiquesau Siam,"Bulletin la de Socidt6 t8tudesIndochinoises, XXIV, No.3, des Vol.
1949,pp.79-92.

- "L'inscription de la stele de Ligor. 1tat present de son interpretation," Oriens Extremus, Vol. VI, March, I959, pp.42-48. - Review of Chandi Bukit Batu Pahat by Alastair Vol. XXI, No. I, Noof Lamb,Journal Asian Studies, vember, i96i, pp. 9 5-96. Collis, Maurice.IntoHiddenBurma.London, 195 3, pp.
2i8-233.

Combaz,Gisbert. "La Loi de frontalitedans la Sculp-

tureindienne," RAA, Vol. VII,No. II, 1931-1932,

pp. I05-III.

Coomaraswamy,A. K. Historyof IndianandIndonesian

Art. New York, 1927.

- "Early Indian Iconography," EA, Vol. I, No. I, July, i928, pp. 33-41. Dale, W. L. "Wind and Drift Currentsin the South GeograChinaSea," TheMalayan Journalof Tropical phy, Vol. 8, June, I956, pp. I-3I. Damais, Louis Charles. "Les 1critures d'origine indienne en Indonesie et dans le Sud-Est asiatique

- "Les premieresimagesbrahmaniques d'Indochine," Bulletin la Socidtd tudesindochinoises, XXVI. de des Vol. Pt. 2,I95I, pp.I31-140. - "A propos d'une image de Visnu en bronze," Bulletin Vereeniging Van Vrienden Der A.iatsche Kunst, No. 2, December, 1953, pp. 2I-26. - La statuaire priangkorienne.Ascona, 1955. - L'archdologie de mone Dvjravatf. Paris, 1959. - "Buddhas dits d'Amaravatien Asie du Sud-Est," BEFEO, Vol. XLIX, 1959, pp. 632-636. - "Tchen-la et Panduranga,"Actes du XXI Congris desOrientalistes, 1949, pp. 242-243. Evans, I. H. N. Ethnology Archaeology the Malay and of Peninsula. Cambridge, 1927. - "On Ancient Remains from Kuala Selinsing, Perak,"JFMSM, Vol. XII, I928, pp. I28-I3 2. - "FurtherNotes on Remainsfrom Kuala Selinsing," JFMSM, Vol. XII, I929, pp. I39-I42. - "A Further Note on the Kuala Selinsing Settlement,"JFMSM, Vol. XII, I929, pp. I86-I96. - "Notes on Recent Finds at Kuala Selinsing," JFMSM, Vol. XV, I930, pp. 25-27.

66

of Feroci, C. (Silpa Bhirasri). TheOriginandEvolution

ThaiMurals. Bangkok,I959.
et Ferrand, G. Relationsde voyages textesgdographiques du arabes, 2 persans,et turksrelatzifs I'Extreme-Orient VIIe auXVIIe si'cles.z Vols. Paris,I9I3 andI9I4. Fine Arts Department, Thailand. Brab Buddharfipa lebBrabBimbandi kru brab prjaga Jdt RJjapiranah (BuddhaImages andVotive Tabletsin the Ayudhyd. Crypt of Wat Rijapfirana, Ayudhyd). In Thai.

chives the Chinese Society America,Vol. VII, of Art of I953, pp. 5-4I. - Dated BuddhaImages of Northern Siam. Ascona,
I957.

Bangkok,I959.
du Foucher, Alfred. L'Art grdco-bouddhiqueGandhara. Paris,I 905. Paris, I895-I908. Fournereau,Lucien. Le Siamancien. Fraser, Thomas M. Rusembilan:A Malay Fishing Thailand. Ithaca, I960. Villagein Southern Gangoly, O.C. "Relations between Indian and Indonesian Culture,"JGIS, Vol. VII, No. I, January,
I940, I957.

"The Jalong Bronze," FMJ, Vol. VII, i967, pp. 64-66. - "The Santubong Buddha and Its Context," SMJ, Vol. X, July-December, i962, pp. 363-37I. - "ImportedImages and the Nature of Copying in the Art of Siam, " in EssaysOffered G. H. Luce.Ascoto na, i966. Groeneveldt, W. P. Historical Notes on Indonesia and Malaya: Compiled from ChineseSources.Djakarta,
i960.

pp. 5I-69.

- and Goswami, A. TheArt ofthePallavas. New York, Gardner, G. B. "Ancient Beads from the Jahore River,"JRAS,I937, pp. 467-470. Gerini, G.E. "Siamese Archaeology: A Synoptical Sketch,"JRAS, I 904, pp. 245-246. Ghose, Ajit. "A Group of Early C6la Bronzes," OZ, I934, pp. I76-I86. of Gibson, James J. The Perception the Visual World. Cambridge,Mass., I950. Gibson-Hill, C.A. "Further Notes on the Old Boat Found at Pontianin SouthernPahang,"JMBRAS, Vol. XXV, Pt. I, I952, pp. III- I13. Giles, F.H., and Scott, George. "Some Remarks on the Land Route Across the MalayPeninsula,"JSS, Vol. XXVIII, Pt. I, I935, pp. 79-84.

Translated Groslier, BernardP. TheArt of Indochina. by George Lawrence.New York, i962. Grousset,R. "L'artpila et senadansl'Inde exterieure," Linossier. e8tudes d'orientalisme Paris,I 932. Gubiand, Maude. "Les Caravaniersasiatiques et les riverainsde l'Ocean Indien vus par les coroplastes de la Smyrne romaine," AAS, Vol. X, Pt. 4, pp.
324-3 33.

Asia. New York, Hall, D. G. E. A HistoryofSouth-East

I962.

Hamilton, A. W. "The Old Kedah-Pattani Trade Route," Journal of the Straits Branchof the Royal Asiatic Society, Vol. 86, I922, pp. 389-392. Harle, James C. "The Early C6la Temple at Pullamafngai," OA, Vol. IV, No. 3, pp. 96-io7. Harrisson, Barbara."Stonewares: Marco Polo Ware in South-EastAsia," SMJ, July-December, I962. - and Harrisson, Tom. "Indian Pioneers in Borneo: c. 5ooA.D. on...," SMJ, Vol. VI,No. 6 N.S., I955,
pp.
5II-5I7.

of Giles, H.A. TheTravels Fa-hsien: 399-4I4 London,I 956.

A. D.

Harrisson, Tom. "Export Wares Found in West Borneo," OA, Vol. V, No. 2, Summer, I959, pp.
42-5 2.

Goloubew, V. "Les Images de Sarya au Cambodge," Cahiers l'Pcole Franjaise de No. d'Extre'me Orient, 22,
I940, pp. 38-42.

- The Malays of Southwest SarawakBefore Malaysia.

London,
I970.

Gombrich, E. H. J. Art andIllusion.New York, I 960. C. Gravely, F. H., and Sivaramamurti, Guideto the Galleries. ed. Madras,i960. 4th Archaeological Grimes, A. "The Journey of Fa-hsienfrom Ceylon to Canton," JMBRAS, Vol, XIX, Pt. I, I94I, pp. 76-96. Griswold, A. B. and Luang Boribal Buribhand, "Sculptureof PeninsularSiam in the Ayuthya PeI riod,"JSS, Vol. XXXVIII, Pt. 2, January, 95I, pp. i-6o. Griswold, A. B. "The Buddhas of Sukhadaya,"Ar-

- and O'Connor, Stanley J. Excavations the Prehisof in toricIronIndustry WestBorneo, Vol. II, Associated ArtifactsandIdeas.Ithaca,N. Y., CornellUniversity, Southeast Asia Program Data Paper No. 72, i969. His Hirth, F., and Rockhill, W. W. ChauJu-kua, Work on the Chineseand Arab Tradein the Twelfthand Thirteenth Entitled Chu-fan St.Peterschi. Centuries, burg, I 9 I I. Hodder, B. W. Man in Malaya.London, I 959. Hultzsch, E. "A VaisnavaInscriptionat Pagan," El,
Vol. VII,
I902-I903,

pp.

I97-I98.

67

"SupplementaryNote on a Tamil Inscription in Siam,"JRAS, Pt. I, I94, pp. 397-398. Hutchinson, E. "Journey of Mgr. Lambert, Bishop of Beritus, from Tenasserim to Siam, in I662," JSS, Vol. XXVI, Pt. 2, I933, Ipp. 2I5-2I. Brah Garu. A Brief Account of the Indannfiicariya, the AntiquitiesSurrounding Bay of Bandon.Chaiya,
-

"Three Statuesin a Tree: A Note on the Pra Narai Group, Takuapa,"FMJ, Vol. VI, i96i,pp. 64-68. - "Some Glass Beads from Kakao Island, Takuapa, South Thailand," FMJ, Vol. VI, i96i, pp. 48-5 5. - "Pengkalan Bujang-An Ancient Port in Kedah," in Malaya History,Vol. VII, No. I, September,i96i,
pp.I2-I7.

I950.

Janse, Olov. R. "Quelques reflexions propos du bol de type megareen,trouve au Vietnam,"AAS, Vol. XXV, Pt. 4, pp. 280-292. - "Dionysos au Vietnam," Viking, (Oslo), I958, pp.
38-50.

"The Hermits of Perlis: Clay Tablets from Caves near Kangar," Malayain History, Vol. VII, No. i,
i962, Pp. I4-I9.

"A Note on a Small Inscribed Stone Tablet from Dr.Wales' Kedah Site No. I," FMJ, Vol. VII,
i962,pp.67-68.

- "On the Origins of TraditionalVietnameseMusic," Vol. Asian Perspectives, VI, Nos. i and 2, SummerWinter, I962, pp. I45-I62. of The Evidence the Jones, A.M. Africa andIndonesia: Leiden, I964. Xylophone. du Jouveau-Dubreuil, G. Iconographie sud de l'Inde Paris, I9I4. Jouveau-Dubreuil, G. "Les antiquites de l'epoque de Pallava," Revuehistorique l'Inde Franfaise,Vol. I
I9I6-I9I7.

- "A Note on Some Ceramicsin Two PrivateCollections at Buantan, Pahang," FMJ, Vol. VII, i962, pp. 68-69. - "Notes on Satingphra,"JMBRAS,Vol. XXXVIII,

Pt. i, July, i964.

Legge, J. A Recordof Buddhistic Kingdoms. Oxford, i886. le May, R. The Cultureof South-EastAsia. London,
i965.

- Pallava Antiquities. Translated by Swaminadha Dikshitar, Vol. II. Pondicherry, I9I 8. Kempers, A.J.Bernet. The BronZesof Ndlanda and Art. Hindu-Javanese Amsterdam, I93 3. - AncientIndonesian Art. Cambridge,I959. Khan, Mohamed G. Historyof Kedah.Penang, I95 8. 2 Kramrisch,Stella. TheHindu Temple. Vols. Calcutta,
I946.

TradeandSociety. Leur, J.C. van. Indonesian Essaysin Asian Social and EconomicHistory. The Hague,
I955.

- Indonesian TradeandSociety.The Hague, z nd ed.,

i967.

- "Palaand SenaSculpture,"Rupam, 40, October, No.

I929,

pp. io8-I26.

Kubler, George. The Shapeof Time: Remarkson the New Haven, I962. Historyof Things. Lajonquiere, Lunet de. "Le domaine archeologique du Siam," BCAI, Vol. I, I909, pp. i88-z6z. - "Essaid'inventaire archeologiquedu Siam,"BCAI, pp. I9-I8I. I9I2-I9I3, Lamb, Alaistar. "Some Notes on the Distribution of of Indianized Sites in Kedah,"Journal theSouthSeas Vol. XV, Pt. II, December, I959, pp. 99-III. Society, - Chandi BuketBatuPahat.Monographson Southeast Asian Subjects,No. I, Singapore, I960. - "Chandi Buket Batu Pahat: Some Additional Notes," FMJ, Vol. VI, I96I, pp. I-9. - "Kedah and Takuapa: Some Tentative Historical Conclusions," FMJ, Vol. VI, I96I, pp. 69-88. - "Research at Pengkalan Bujang: A Preliminary Report," FMJ, Vol. VI, I96I, . 2I-37.

Levi, S. "Ptole'me, le Niddesa et la Brhatkatha," EA, Vol. II,1I 925, PP.I-5 5 arid43I- 432 Lippe, Aschwin. "Vishnu's Conch in Nepal," OA, Vol. VIII, Autumn, i962,pp. II7-II9. Issues in the History Loehr, Max. "Some Fundamental of Chinese Painting," JAS, Vol. XXIII, No. 2, February,i964, pp. i85-I94. AcrosstheIsthmus Kra. Loftus, A. J. Notesonajourney of London, i883. Lohuizen-de Leeuw, J. E. van. The"Scythian" Period.

Leiden,I949.
- "The Protector of the Mountain of Truth," AAS,

Vol. XX, Pt. I,

pp. 9-I7.

- "The Dikpalakasin Ancient Java," B /dragen, Deel III, Pt. 4, PP 356-384. Low, Colonel James. "An Account of Several Inscriptions Found in Province Wellesley, on the Peninsulaof Malacca,"Miscellaneous PapersRelating to Indo-China, Vol. I, i886, PP. 223-22 5. Reprinted fromJASB, Vol. XVII, Pt. II, i848, pp. 6z-66. - "On an Inscription from Kedah," Miscellaneous

PapersRelating Indo-China, to Vol. I, I886, PP. 68

232-234.

Reprinted fromJASB, I848, pp. 247-249.

VOL. XVII, Pt. I,

Luce, G. H. "CountriesNeighboring Burma,"JBRS,

Vol. XIV, Pt. 2, I925, pp. I38-205.

"A Note on the Sambas Finds," JMBRAS, Vol. XXII, Pt. IV, I 949, pp. I6-I9. - "Takuapa and its Tamil Inscription,"JMBRAS,
-

Vol. XXII, Pt. I, I949, PP

I949. East. Bombay,

25-30.

- "The Early Syam in Burma's History, a Supplement," JSS, Vol. XLVII, Pt. I, June, I959, pp.
59-I0I.

in Indian NilakantraSastri,N. K. South Influences theFar O'Connor, Stanley J. "An Early BrahmanicalSculpture at Songkla,"JSS, Vol. LII, No. 2, July, i964,
pp. i63-i69.

- "Some Old Referencesto the South of Burmaand Ceylon," in FelicitationVolumesof Southeast-Asian Studies to Presented Prince Dhaninivat. Bangkok, I965.
pp. 270.

Majumdar,R. C. Hindu Colonies the Far East. Calin

cutta,I 944.
-

Ancient in Southeast IndianColonirZation Asia. Baroda,

I944.

"The Overseas Expeditions of King Rajendra Cola," AAS, Vol. XXIV, Pts. 3 and 4, pp. 38-342. Malleret, Louis. "Apersu de la Glyptique d'Oc-Eo,"

- "An Ekamukhalihga from Peninsular Siam," JSS, Vol. LIV, No. i, i966. pp. 43-49. - "Satingphra: An Expanded Chronology," JMBRAS, Vol. XXXIX, No. i, July, I 966, pp. I 37-I 44. - "Si Chon: An Early Settlementin PeninsularThaii968, pp. land,"JSS, Vol. LVI, No. i, January, i-i 8. - and Harrisson,Tom. "WesternPeninsularThailand and West Sarawak:Ceramicand StatuaryCompari964, ison,"SMJ,Vol. XI, 23-24, July-December,

BEFEO, Vol. XLIV, I 947-I 95 0, pp. I 89-I 99. - L'archdologie deltadu Mdkong. Vols. Paris, I 959du 4 I963.
- "Classificationet nomenclature des 'perles' archeologiques en fonction de la symetrie minerale," BEFEO, Vol. LI, No. I, I963, pp. I I7-I24. - "Pierresgravees et cachetsde divers pays du sud-est de l'Asie," BEFEO, Vol. LI, No. I, i963, pp.
99-II6.

Pp. P62-566. Pacht, Otto. "Art History and Art Critics-VI: Alois April, i963, pp. i88Reigl," Burlington MagaZine,
I93.

Colombo. I 946. S. Paranavitana, TheStfpa in Ceylon. - "Negapatam and Theravada Buddhism in South

India," JGIS, Vol. XI, No. I,

I 944, pp. I 7-2 5.

PrivateCollections, NationalMuseum, Masterpiecesfrom Bangkok, i968. d Ma-Touan-Lin. Ethnographie peuples6trangers la des Chine. Translated by Marquis d'Hervey de Saint
Denys.
2

Vols. Paris, i876.

Maxwell. W. E. "Journey on Foot to the PataniFrontier in I876," Journal of the Straits Branch of the
Royal Asiatic Society, Vol. 9, i88z, pp. I-67.

- "The Significance of Sinhalese 'Moonstones'," AAS, Vol. XVII, Pts. 3/4, pp. I 97-2 3 I. - Ceylon Malaysia. and Colombo,i966. Parmentier,H. "Nouvelles notes sur le sanctuairede Po Nagar a Nhatrang," BEFEO, Vol. VI, Pts. 3 and 4, I906, pp. 29 I-300. - Inventaire Monuments Camsde l'Annam. Vol. II. des

Paris,I 9 I 8.
-"Origine
I925,

Miller, J. Innes. TheSpice Tradeof the RomanEmpire. Oxford, i969. Medway, Lord. "Archaeological Notes from Pulau Tioman," FMJ, Vol. VII, i962, pp. 55-63. Moens, J. L. "grivijaya,Yava en Kataha,"JRASMB, Vol. XVII, Pt. II, January, I940, pp. i-io8. An abridgedtranslationby R. J. de Touche from TBG, Deel LXXVIII, I937. Nanjio, B. A Catalogue theChinese of Translations the of Buddhist Tripilaka.Oxford, I 883. NilakantraSastri,K. A. "A Tamil Merchant-Guildin Sumatra,"TBG, Deel LXXII, I932, pp. 3I4-327. - The Colas. 2 Vols. Madras, I 935 -I 93 7. - "Agastya," TBG, Deel LXXII, I936, pp. 47I-545. - "SriVijaya,"BEFEO, Vol. XL, I940, pp. 239-3 I 3.

commune des architectures hindoues dans l'Inde et l'Extreme-Orient," EA, Vol. II,
pp. I99-24I.

L'Art Khmer Paris,I927. primitif.

"La construction dans l'architecturekhmere classique," BEFEO, Vol. XXV, Pt. 2, I935, pp. 283284.

Pelliot, P. "Le Fou-nan," BEFEO, Vol. III, No. 3,

I903, pp.248-330.

- "Deux itinerairesde Chine en Inde a la fin du VIIIe siecle," BEFEO, Vol. IV, I904, pp. I 3 I-4I 3. - "Encore a propos des voyages de Tcheng Houo," Toung Pao,Vol. XXXII, I 93 6, pp. 2 I0 - 222. Pendleton, Robert L. Soils andLand Use in Peninsular Siam, Technical Bulletin No. S, Department of

Agriculture. Bangkok,I949. 69

- Thailand:Aspectsof Landscape Life. New York, and i962, pp. 46-47. Picard,Charles."La lampe alexandrinede P'ong Tik

Schnitger, F.M. The Archaeology Hindoo Sumatra. of Supplement to Vol. XXXV, Internationales Archiv

fur Ethnographie. Leiden,I937.

- Forgotten Kingdoms Sumatra.Leiden, I939. in Tonn, See Fong. "The University of MalayaArchaeological Society's Survey of the Kuala Muda Area (South Kedah), in July, I95 6," JMBRAS, Vol. XXXII, Pt. I, May,I959, pp. 209-2I 3. Seidenfaden,Erik. "Recent Archaeological Research Work in Siam,"JSS, Vol. XXX, Pt. II, April, I 93 8,
pp. 24I-247.

(Siam)," AAS, Vol. XVIII, Pt. 2, pp.

I37-I49.

- "A Figurine of LysippianType from the Far East: the Tra Vinh Dancer," AAS, Vol. XIX, Pts. 3 and
4,P 342-352.

Rajanubhab,Prince Damrong. A History of Buddhist Monuments Siam. Translated by S. Sivaraksa. in Bangkok, i962. Ramachandran, N. "The Nigapattiniam and Other T. Buddhist Bronzes in the MadrasMuseum,"Bulletin of the MadrasGovernment Vol. VII, No. I, Museum,
I954.

Rao, P.R. Ramachandra.The Art of Nagirjunakonda Madras, I 9 56. Rao, T. A. Gopinatha. Elementsof Hindu Iconography. z Vols. Madras, I 9 I 4-I 9 I 5. Rawson, P. S. Review of La statuaire pr6-Angkorienne by Pierre Dupont,OA, Spring,I 9 57, pp. 33-43. - "The Methods of Indian Sculpture,"OA, Vol. IV, Winter, I 9 58, pp. I 36-I44. Ray, Amita. "House Building Rituals in Ancient India,"J. N. Banerjea Volume. Calcutta,i960. Read, Herbert. TheArt of Sculpture: A. W.Mellon The Lecturesin FineArts 19y6, NationalGallery Art, the of New York, I 9 56. Washington. Renou, Louis (ed.). Hinduism. New York, i963. - etal. L'Inde Classique. Paris,1947. Rentze, Anker. "Gold Coins of the Northeastern Malay States,"JMBRAS, Vol. XVII, Pt. I, October, I939, pp. 88-97. - "A Historical Note on the Northeastern Malay States,"JMBRAS, Vol. XX, Pt. I, June, I 947, pp.
23-40.

- "A Mystery Temple in Surat Province,"JSS, Vol. XXXIV, Pt. I, April, I943, pp. 63-65. Sellman, R. R. An OutlineAtlas of EasternHistory. London,I 954. Shah,UmakantP. "A Few Brahmanical Images in the BarodaMuseum,"Bulletin theBaroda of Museum, Vol. X-XI, April, I 9 53-March, I 95 5 pp. I 9-2-3. - "Some Early Sculpturesfrom Abu and Bhinmal," Bulletin of the Museumand Picture GalleryBaroda, Vol. VII, SpecialNo. I, i960. - "Sculpture from Samallji and Rhoda (North Gujarat) in the Baroda Museum," Bulletin of the Museum PictureGallery and Vol. VII, Special Baroda, No.1, i960. Shapiro, M. "Style," Anthropology Today.Edited by A. L. Kroeber.Chicago, I 9 53. Singaim, S. Durai Raja. India and Malaya Through the I Ages.KualaLumpur, 9 50; 3rded.,Kuantan,9 54. I Sivaramamurti,C. "Geographic and Chronological Factorsin Indian Iconography," AncientIndia,No.

6,I956, PP 44-46.
- "The Weaponsof Visnu," AAS, Vol. XVIII, Pt. 2, pp. I28-I36. - Kalugumalai and EarzlyPindyan Rock-Cut Shrines. Bombay, I96I. - IndianBronrZes. Bombay, I962. Skeat, W. W. "Reminiscenceof the Expedition by the late W. W. Skeat, M. A., Leaderof the Expedition," JMBRAS, Vol. XXVI, Pt. 4, December, I 9 53. Smith, Vincent A. "Indian Sculpture of the Gupta Period A.S. 300-650," OZ, April-June, I914, pp.
I-28.

Rowland, Benjamin.TheArt andArchitecture India. qf

Baltimore, 59. 19
Sankalia, H. D. The Archaeology Gujarat.Bombay, of
I94I.

- "Imported Mediterranean Amphorae from Kolhapur,"JRAS, I957, pp. 207-208. - andDikshit, M. G. ExcavationsatBrahmapuri(Kohla-

pur),I94J-I946. Deccan College Series: Monograph I S*Poona, 9 52.

Saraswati, S. K. Early Sculpture Bengal. Calcutta, of
I962.

Sastri, H. Krishna. South IndianImages of Gods and Goddesses. Madras,I9I6. Peaches Samarkand. Schafer,Edward H. The Golden of Los Angeles, i963.
70

SolheimW. G. "Potteryandthe Malayo-Polynesians," Current Vol Anthropology, V, S, December, I964. David E. The Sea Nomads. (Memoirs of the Sopher, National Museum, S, I965). Singapore,pp. 3-I9. Stein, R.A. "Le Lin-yi," Han-Hieu, Vol. 2, I947, pp.
I 20-I 22.

Stern, Philippe. Le Bayon d'Angkoret l'volutiondePart Khmer.Paris, I927. - "Ajantd, Ellora et revolution des styles Gupta et post-Gupta de l'Inde," Actes du XXIe Congres International Orientalistes. des Paris, I949, pp. 2z8229.

- etal. r'volution dustyle indien d'Amardvatf. Paris, i96i. Sullivan, Michael."Notes on ChineseExport Waresin South-East Asia," TransactionsOriental Ceramic Vol. XXXIII, i963, pp. 6I-77. Society, Stutterheim, Willem. "The Meaning of the HinduJavanese Candi,"JAOS, Vol. 5i, No. i. March,
I93I,

pp. i-i6.

as Takakusu,J. A Record theBuddhist of Religion Practiced in India and the Malay Archipelago(671-69; A.D.) byI-Tsing. Oxford, I896. Temple, R. C."Notes on Antiquitiesin Rdmafifiadesa," IndianAntiquary, Vol. VIII, I 956, pp. 2i-6o. Tibbetts, G. R. "The Malay Peninsula as Known to Arab Geographers," MalayanJournal of Tropical Vol. Geography, VIII, I 956, pp. 2i-6o. - "Pre-Islamic Arabia and Southeast Asia," JMBRAS, Vol. XXIX, Pt. 3, I956, pp. i82-2o8. - "Early Muslim Traders in South-East Asia," JMBRAS, Vol. XXX, Pt. I, I957, pp. I-44. Tremenheere,F. G. "Report of a Visit to the Pakchan River and of some Tin Localities in the Southern Portion of the TenasserimProvinces,"JASB, Vol. XII, 2, I 843, pp. 523-5 34. Tremenheere,Capt. G. B. "Report on the Tin of the Province of Mergui,"JASB, Vol. X, i841, p. 845. Tweedie, M. W. F. "The Stone Age in Malaya,"

- TheMakingof Greater India. London, I 9 5 I . - "An Early Buddhist Civilizationin Eastern Siam," JSS, Vol. XLV, Pt. I, April, 1957. - "A Stone Casketfrom Satingphra," JSS, Vol. LII, Z, July, I964, pp. 2I7-223. Wang, Gungwu. "The Nanhai Trade," JMBRAS, Vol. XXXI, Pt. 2, I95 8. Its Wells, Kenneth. ThaiBuddhism: RitesandActivities, Bangkok, i960. Wiener, Sheila. "From Gupta to Pala Sculpture," AAS, Vol. XXV, Pts. 2 and 3, pp. i67-I92. Wen Fong. "Chinese Painting: A Statement of Method," OA, Vol. IX, No. 2, i963, pp. 73-78. Wheatley, Paul. "GeographicalNotes on Some Commodities Involved in Sung Maritime Trade," JMBRAS, Vol. XXXII, Pt. 2, June, I959, pp.
I-I40. -

Khersonese. Kuala Lumpur, i96i. TheGolden "Arabo-Persian Sources for the History of the Malay Peninsula in Ancient Times," Malayan HistoricalSources. Singapore, I962. "Desultory Remarkson the Ancient History of the Malay Peninsula," in J. Bastin and R. Roolvink Studies:Essays Pre(eds.), Malayanand Indonesian sentedto Sir RichardWinstedt.Oxford, i964, pp.
4I-42.

JMBRAS, Vol. XXVI, Pt. 2, I 95 3.

Vogel, Jean Philippe. La Sculpture Matburd. de Paris,
I930.

Wales, H. G. Quaritch."A Newly Explored Route of Indian CulturalExpansion," IAL, Vol. IX, No. I,
I93 5, pp- I-3

- Towards Angkor.London, I 937. - "Archaeological Researches on Ancient Indian Colonizationin Malaya,"JMBRAS, Vol. XVIII. I,
I940, pp. i-85.

- "RecentMalayan Excavations," JRAS, 1946, pp.

142-149.

Wheeler, R. E. M. et al. "Arikamedu:an Indo-Roman Trading Station on the East Coast of India," AncientIndia,No. 2, July, I946, pp. 17-I24. Wickremasinghe, Sirima. "Ceylon's Relations with South-EastAsia with SpecialReferenceto Burma," Vol. and TheCeylonJournal Historical Social Studies, of 3, No. I, January-June,i960, pp. 38-5 8. Williams-Hunt, Major P.D.R. "An Introduction to the Study of Archaeology from the Air,"JSS, Vol. XXXVII, Pt. 2, June, I 949, pp. 8 5-I I 0. Winstedt,R. 0. "The Ligor Inscriptions,"JMBRAS, Vol. XXII, Pt. I, I 949, pp. I 76-I 77. - A History of Malaya.Singapore, i962. Wirjosuparta, R. M. Sutjipto. "The Second VisnuImage of Cibuaja in West Java," MadjalahIlmuI/mu Sastra Indonesia(Djakarta), Vol. I, No. 2, September, I963, pp. 170-187. BSOAS, Vol. XXIII, Wolters,W. 0. "Tdmbralifiga,"

- "Culture Change in Greater India," JRAS, 1948, PP. 1-32. "A Note on Takola, Langasuka, and Kataha," JMBRAS, Vol. XXIII, Pt. I, 19 50, pp. 152-15 3. - "A Note on the 'Four Main Waves',"JMBRAS,

Pt. 2,
-

1960, pp. 323-3 50.

"Chen-li-fu,a State on the Gulf of Siam at the Beginning of the I3th Century,"JSS, Vol. XLVIII,

Pt. 2,
-

i960, pp. 1-36.

Vol. XXIII, Pt. I, 1950,

PP 153-I55.

"rlvijayan Expansion in the Seventh Century," AAS, Vol. XXIV, Pts. 3 and 4, i96I, pp. 417-424.

7I

- Review of Wheatley's GoldenKhersonese, BSOAS, Vol. XXV, Pt. 3, i962, pp.63 8-639. - "The Po-ssu Pine Trees," BSOAS, Vol. XXIII, Pt. 2, i960, pp. 3233-350. - Eary Indonesian and Commerce the Originof Srfvjqya. Thesis. London University, I962. - "China Irredenta: the South," World Today,Vol. XIX, I2, i963, Pp. 540-552.

Wongthet, Suchit. "BorankhadiiPharechon (Wandering Archaeology)," Chao KrUng,Vol. XVI, 2, November, i966. Zainal-'Abidinbin Ahmad. "Some Malay Legendary Tales," JRASMB, Vol. XXIV, Pt. I, February,
19 51, pp.

80-84.

Zimmer, H. TheArt of IndianAsia. 2 vols. Edited by J.Cambell. Bollingen Series XXXIX. New York,
I95 5.

72

INDEX

21, abhayamudrd 33, 39 Aditya I, King 54 Akhbaras-Sinwa'l-Hind58 Amarivati 32, 37, 38, 49 33 amritaghata Borei 28 Angkor Aniruddha, King I5 Annam 27 Aparajita,King 54 Arab geographers 5i, 58 Arab sailors i6 armlet 25 Aru Islands 14 Avalokitesvara47

Buddhappiya22 Bugis 14 Bujang River 57 Burma xs, I6, 56,63

2i, cakra 30, 33, 45, 6i

Calukya30, 53 Cambodia27, 4i, 42, 45, 48, 63 camphor I4 Ceylon I5, I6, 38, 49, 6I, 63
Cham
20, 23

Ayudhya

19,

23

Balaputradeva,Prince 57 Bali 27 Bandon River 49 Bangka Island I2 Bassac River 29 Bathe 36, 43 n. 6o Batuka-Bhairava of Bandon II, I2, 13,I5, i6, 40, 46, 6o Bay beads so, 58 beeswax 14 Begram 36 Bejrapuri,see Petburi Bengal 20, 46 Bhairava6i Bharukaccha(Broach) 35, 36 Bhinmal 34 bhfl21, 30, 45 Billiton Island i2 Borobodur 20 24 Brahmabhaga brahmi 36 Brahmins 27, 28, 63 Broach, see Bharukaccha

Champa24 Chandi Kalasan 20, 23 Chen-la i6, 29, 42 Cherok Tekun 38 Chieng Mai 23 China I2, I4, I5, I 6, 22, 37, 49, 50, 57, 58 Chinese records 27, 28 Ch'uan-chou 56, 58, 63 Chumpon I I Chu-Ying 28 cinnamon 13-14 C61aI 5, 22, 27, 52, 54, 56, 6i, 62, 63 Cranganore,see Muchiri Damrong, Prince i9 dhoti21, 25, 29, 33, 38, 42, 43, 45, 46, 47 Dong Diiong 23 Dong Si Mahapot 42, 47 Dvaravati 45, 49, 63 earrings 25, 33 Ellora 32 Fu-nan 12, 15, i6, 24, 28, 32, 36, 39, 40, gadd 20, 30, 33, 37, 45 Gandhara44 Ganesa 28, 46
42

73

gharu wood 14, 15, 63 glass 14, 50, 57, 58 gold i4 Gulf of Siam i6, 49

kiritamukuta 38,

42,

44, 6I

Gupta 35, 36, 38, 44, 48, 49

hara 6i

Hua Kao 46, 47 Iksvaku 37, 38

India i2, 14, 15, i6, 50, 57, 58 Indonesia 17, 37
17, 22,

Ko Kao Island 49, 50,55 Kolhapur 36 Kompong Prah 48 KulottunigaI, King 63 kundala2I Kusna 27, 30, 33, 34, 37, 39, 44 Kuvera 44 laka wood 14 LangkasukaI5, I5 n. Liang-shU 28 Ligor, see Nagara Sri Dharmaraja
lifiga
23, 24, 25

27, 31, 34, 35, 37, 46, 49,

Indra 34, 44
Inscriptions
I 5, 27

Buddhagupta 39 Bukit Choras 39 Cherok Tekun 37, 39 Ligor I 5, 57 Loboe Toewa 62

Lopburl i 6

Lokesvara 59 lotus 25
Mahdtittha
I3

Mahayana23, 46, 57, 58, Maitreya33 Makassar14

Malacca 62
i2, 13

59,

6o

Nagara 8ri DharmarajaI5 Pra Narai 56, 58

Ram Kamheng 22 Ta Keo 30 Tamil 27 Wat Mahadhatu 27

Manigrdmam (Manikkirdmam)

I 5,

54, 55, 56, 57, 58,

Ipoh 36 Jain 33
Jaiya i6, 19, 20, 2i, 25, 26, 31, 33, 38, 39, 40, 59, 6i Jalong 33, 59 Janardana 31 Java i6, 20, 23, 27, 46, 62

see Manikkiramam, Manigramam Mathura30, 32, 33, 34, 37, 39, 44 Merbok Estuary of Kedah I 5, 57, 59 Mergui I3
Middle East I2, 50, 57,s8

Mi-Son 23 mitre 2I, 25, 35 Mons 23 Muchiri 35

mukhalihga24 mukzia 33, 34 Mzlasarvdstividaekasatakarman57

Jayaviravarman, King i6 Kabfil 36 Kalah 5I, 58

kailamakara20 kamarband 25, 33 K'ang T'ai 28 kaukana 6i

Kantan 1 3
kapardin 35 Kaun.dinya, King 40

furti 29, 30 Museums Baroda Museum 45 DjakartaMuseum 36 MadrasMuseum 62 Musee Guimet 35 Museum fur Volkerkunde, Berlin 30 National Museum, Bangkok 19, 20 SarawakMuseum 50 Nagara Pathama23 Nagara Sri Dharmaraja (Nakon Si Tammarat),I I,
I5,

Kdveripdkkam54
Kedah 13, 37, 39, 57

Kev, see Wat Keu Khmer i 6, 3 1 Kinta Valley 59 74

3,

16, 19, 2I, 22, 23, 25, 27,

31, 33, 38, 39, 45, 59,

63

Nagarjunakonda37, 38 Nakon Si Tammarat,see Nagara Sri Dharmaraja Nalanda 20, 57 Namakkal 54 NandivarmanIII, King 54, 55 Narasirthha 38 Narasixhhavarman King 56 II, Narayana,see Pra Narai Negapatam 22, 56, 57, 63
Nepal 34

pre-Angkorian 20, 24, 26, pre-Dvaravati38 PuiGet (Phuket) 12, 15 Purinas 30 Rajadhiraja King 56 I, RajendraC6ja 62 Rastrakfitas53

27,

42,

44,

48

rhinoceros horn 14
Roman Empire 35

Nolambas 43 n. Oc-Eo
24, 28, 29, 31, 33,

36, 37, 38, 39

Rome14 Rudrabhdga 24 Rudravarman, 28-29 King 31 Ripama~y1ana gailendra 57 22,

saiikha 20,
30,

'i Padmapan 59 padma 2, 30, 39,

Pjdza-purina 30 Pagan 56, 58, 63

45

45, 6i

Pala20,57,59,6i
Pallava 15, 30, Pandyas43 n.
P'an-p'an 28,
42, 43, 44, 45, 48, 49, 5!, 53, 54, 55

Sarawak 50 Sarnath 6o 32, sash25, 29, 33,

42

40

I, Paranataka King 6I Pattami12, 13 pearls 14 Pegu 12 Persian Gulf I 2, i 6 Peshawar 33 Petburi (Bejrapuri)23, Philippine Islands i2 Phnom Da z6, z8, 29, Phuket, see PuiGet Pliny 14 Polonnaruva 22 portage 13 Poseidon 36
Po-Ss'U37, 51

Sassanian33, 36 SatingPra42, 46 Sena 59 Shapur I, King 33

Siam 41,
silks 14
42,

42

Sichon 12, 45

47

Singapore 13, Siraf 2 I

14

32, 42, 48

post-Gupta 32, 36, 38, 44, 48, 49 Pottery Chinese 13, 58 Six Dynasties 50

Sung 13,

23, 46, 50

Sirkap 36 Siva 23, 27, 53, 6o Songklaii, I12, 46, 59 Sopara, see 8firpiraka Sri Mahabodhi, see Dong Si Mahapot 21 Jrlvatsa Srlvijaya I5, 19, 47, 51, 5 7, 58, 63 20 sthinakamifrti Straitsof Malacca 14, 1 5, 57 Sumatrai6, 46, 62 Sungai Golok i 6 SurdstradhiniProvince 42 Surkh Kotal 33
grarpiraka 3 5 Stirya 36, 54, 6i

T'ang 50, 57 Yuan 46 Yiueh 50 earthenware58 Middle Eastern 50 Province 42 Pracin.apurl Pra Narii 5,55, 5 6, 58

Suaryavarman King i6, 30 I, Takuapa I2,

6o
13, 15, 41, 43, 44, 49, 51, 52,

566,58,

Prasat Andet48
75

Tambarattha22 I Tambraliuiga 6

Timralipti 13 Tang Kasang 48 tank 56 tUntrism3 '

Trichinopoly 53 Tun Sun I5, 15 n., 28 Tuol Ang 48 Tuol Koh 26, 43 n. UmA48 Tong 38 Vajrap1i 59 vanamili 6 i Vasudeva 3I Vat Eng Khna 48 Vat Po Veal 48 Viai Srah, see Vieng Sra

TavoyIslandi2
Taxila 30, 33, 36
I Tenasserim 3

Thai i6, I9, 27 theravida 22, 32 tin iz, 1I5,49 Tiruttani 54

27 Tjibuaja

torque zI, 33 Trade Goods beads 50, 58

Vieng Sra(ViaftSral.),I5, z6,36,42,6o

24 Virsnpbhdga votive tablets 6o 29 vyuha

beeswaxI4 I4 camphor
ceramics, see pottery
I cinnamon 3-I4

gharu wood I4, I5, 63

glass I4, 50, 57, 58

lakawood I4 pearlsI4
rhinoceros horn I4

silks I4 TrangRiver I 3 24, Transbassac

Tra Vinh 35

Wat Keu (Kev) zo Wat Mahldhitu I9, 20, 22, Wat Majjhimivisa 46, 59 Wat Sala Tung 20 Wat Srivijaya46 Western Asia 49 Wu Ti, Emperor I4 yajiopavita 53, 6i YelsAwaram37, 38

23,2 5, 27, 59

29

76

Fig. Visnu. Stone. Found at Jaiya, southern Thailand. Bangkok Museum

s~~~~~~~~~~~~~~~~

Figi Sid

Fii

Rea

Fig. i a Front

Fig. 2 Visnu. Stone. Found at Nagara 8ri Dharmardja (Ligor). Wat Mahadhatu, Nagara 8r! Dharmardja

Fig. 3 Visnu. Stone. Found at Nagara Sri Dharmardja (Ligor). Wat Mahadhdtu, Nagara Sri Dharmardja

l. Figt 5..f]

.'

Stne

i; 0,XVg ;a.;fXC,, Nagar Xr Dha rma'r'V'ja

r;<~~~~ig
Vn Stn ouda Vita._ainluemSio OcB, Sout

Fig. 6 Visnu. Stone. From Raniwala Well at Palikhera, Mathurd, India. Kusdna Period. Mathurd Aluseum (accession register No. 933)

Fig. 7 Visnu. Found near Mathurd. Kusdna Period. Mathurd Museum (accession register No. I729)

Vsnu

Stn.EryK Mueu

aPeid

i~~~ahr

Fig. 9 a Front

Fig. 9 b Rear

Fig. 9 Visnu. Stone. Found in a well at Baghichal Birhal near Mathurd. Kusdna Period. Malburd Museum (accession register No. 956)

Bulgarian,

lit

Fig. i z Visnu. Stone. Excavated at YZdExwaram, Andhra Pradesh India. 4th or 5th century A.D. Photograph: MI.A. W. Khan

Fig. 13 Narasithha Relief. From Coastal,, Andhra Pradesh. 3rd century A.D. (After M. W. Khan)

At~~;

Fig

4bSdFi

aFrnFg.

:C

Rea

Bangkok Museum~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~sl

Fig. Isa Front

Fig. I 6a Front

Fig.i~~~~~~~b Re~~~ar Fi.i_

Fig.i

ie VegSa t hiad V II.Soe_~ Fig 1.' on tVegSa hiad

VinuI Soe.Fun

Bago Mueu

Fig. I7a Front

Fig. i 8 a Front

Fig. i0b Rear

Fig

l~Ra

Fig IS Viu Stn.Fuda aigPahiad

Fig. I 9 b Rear

Fig.

2o

if
Fig. Front Ia

A~~ ~~~ ~~~ ~~~~~~~~~~~~~~~~~~~~~~ ~ _'

B M

Fon Vi~~u Stone.~~~ II.

at Petbur%'I, hiad

j~~~Fg

_9

VIn

tn.Fuda

ebui(erpr)

hiad

OWN-

Fig. 2 i b Side Visnu. Fig. 2 i Stone. Found at Sating Pra, Thailand. Wat -t\fajjhimdvdsa, SongklA.

Fig. 2 i a Front

aid

A,

Fig.

2zz

Fig.

25

Visnu. Stone. Found at Sichon, Thailand. Private Co//cc/ion, Nagara Sri Dharmardj'a

Visnu. Stone. Found near Wat SrIvijaya, Hua Kao village, Pun Pin district, Surdsradhani. Photograph: A. B. Griswold

Fig. 26a Front

I~~~

Fig. 27 Visnu. Stone. Collection of H. R. H. Prince Chalerrnpol Dighambara. Bangkok. Photograph from: Exhibition of Masterpiecesfrom Private Collections, National Mllsellm, Bangkok. I1968

Fig. 28 'Pra Nar~i' (Ndrdyania) Group. Stone. Located on Takuapd River, approximately io miles upstream from the estuary. Photograph: A. B. Griswold

jw~

~~ ~ ~ ~ ~ ~ ~ ~ ~

~~~~~~htorp:A

.Giwl

Fig. 30
4A

Male Figure, ~~~~~~~~~~~~~~~~~~~~~~~~~~Sm 'Pra Nar~i' Group, Takuapd. Photograph: A. B. Griswold

Fig.

31~~~~~~~~~~~~~~~~~~~~~~~~~

Fig. 32 Visnu. Stone. Found at Vieng Sra, Thailand. Banigkok Museum.

Fig. 32b Side

Fig. 3za Front

*o

l -

X-

K
e

i -

-gEtiS4>:
mu:

--

:: :;0:00

:;S

Fig.

3 4 a

Front

j
_
. .

Fig.

3 4 b

Side

A
An: ,0 X
f

a_0;
.

7t A:

00.