South Indian Influences in the Far East K. A. Nilakanta Sastri   First published in 1949, this seminal work by one of the greatest Indian historians is a gold mine of information on the spread of Indian civilization in South East and Far East Asia. Long out of print, it was reprinted in 2003 by the Tamil Arts Academy and is reproduced here with the kind permission of Dr. R. Nagaswamy. (The text was scanned and proofread by IFIH in 2008.) Prof. K. A. Nilakanta Sastri, besides teaching for many years at the Madras University, authored several important works on Indian history, the best known of which are perhaps A History of South India (Oxford University Press) and The Cholas (University of Madras), both of which remain unrivalled classics. Except for long vowels, diacritical marks could not be preserved in the scanning; we hope to release later an updated version with all diacritics. The map referred to by the author in his preface is unfortunately missing; we will try to include it in a later release. The book is also available for download as a pdf file (550 kB).
Contents
	Preface
I	Introduction: Early Culture Movements
II	Indo-China
III	Burma
IV	Fu-Nan
V	Kambuja
VI	Campā
VII	Siam
VIII	Malay Peninsula
IX	Takua-Pa and Other Places
X	The Southern Islands
XI	Borneo
XII	The Philippines
	Abbreviations
	Bibliography
References and Notes

Preface

The manuscript of this small work was completed in 1942 and submitted for publication in the first instance to the University of Madras. War conditions prevented the publication of the work by the University before my retirement from service in 1947. With the permission of the University the work is now published by Messrs Hind Kitabs Ltd.

The only change in the original manuscript has been the inclusion of a notice of the additional Yupa inscriptions of Borneo at pages 137-138. The finds of Buddhist images in gold and silver in Sambas in the same island has not been noticed as they have been included in a recent publication of mine, History of Srī Vijaya published by the University of Madras.

My obligations to the French and Dutch archaeologists who have worked in Indo-China and in Java, in particular to the learned contributions of Professors G. Cœdès and N. J. Krom, will be apparent on every page as also in the map that accompanies the book. Other works mentioned in the Bibliography have also been found very helpful.

K.A. NILAKANTA SASTRI

‘Nilesvar’
Mylapore, Madras.
10^th October 1949

***

Prof. K.A. Nilakanta Sastri was the most outstanding historian of South India, whose contributions are well known. His book on The Cholas is a classic. He brought to bear on his writing absolute objectivity and thoroughness. Among his various writings this work South Indian Influences in the Far East received the admiration of international scholars. This has been long out of print and not available for scholars for long. The Tamil Arts Academy is happy to bring out this reprint on the occasion of the International Conference on Mahabharata that focuses on artistic and cultural contact India had with South East Asian countries.

R. NAGASWAMY

23-12-2003

I. Introduction: Early Culture Movements

Our aim in the following pages is not to offer a systematic account of the Hindu colonization of the East, much less a history of the various kingdoms that were established as a result of that movement. It is much more limited; we propose to consider only one particular aspect of the movement, and of the early history of the States, their art and social life with a view to estimating the role of Southern India in their evolution.

But before we enter upon this task, it is necessary to have some idea of the state of culture attained by the peoples of the lands to which our colonizers went. However far we go back in pre-history, the evidence from skulls shows distinctly that the races had become thoroughly mixed, and it is therefore safer to speak of a people rather than of a race as the authors of any particular culture. And for the limited purposes of our study, it is not necessary for us to go further back than the latest phase of the New Stone Age. The most characteristic feature of this period is the different forms of adzes with quadrangular sections. This quadrangular adze-culture, says Heine-Geldern, “probably came to the Malay Peninsula and Indonesia by way of China and Central Indo-China (Laos and Siam) between 2000 and 1500 B.C. Its bearers introduced to Indonesia the Austronesian languages, the outrigger canoe, rice cultivation, domesticated cattle or buffaloes, head-hunting, and the custom of erecting megalithic monuments. Beaked adzes belong to this culture as also four-cornered ones. Their gradual development can be traced archaeologically along the trail of the Austronesion migration from Upper Laos through the Malay Peninsula to Indonesia.” Of the adzes of Sumatra and Java in particular, the same writer observes: “Both the quadrangular as well as the beaked adzes of Sumatra are very similar to those of Java. The neolithic cultures of both islands show the same preference for semi-precious stones and coloured varieties of silex, and the same wonderful perfection of stone cutting. Some of these adze blades are real works of art.”[1]

By a careful study of the linguistic evidence, Kern showed in 1889 that in the cradle-land from which the Indonesians began to expand there were grown sugarcane, cocoanut, banana and bamboo—all tropical products. Rattan was also known, and rice was the staple food. The people were also a sea-faring folk. This cradle land he located in Campā, Cochin-China, Kambuja and the neighbouring lands, although he did not rule out the possibility of a still earlier home. This earlier home Heine-Geldern finds in Yunnan and Southern China.

These neolithic men thus received from Yunnan the people whose speech later grew into the Indonesian languages and who occupied the coastal lands on the east whence they began to practise sea-faring. Their river boats—hollowed tree trunks—developed later into the outrigger boats used in the archipelago. The neolithic migration which reached the islands by way of the Malay Peninsula does not seem to have gone beyond Moluccas; it did not reach New Guinea. The relatively high culture attained in this period is attested not only by the cultivation of rice and the fine adzes already mentioned, but also by the pottery and weaving of the time which appear to have attained a high level of excellence. [2]

The next stage in the cultural development of these lands was marked by the extensive use of bronze, coupled with a knowledge of iron, and a greater skill in the arts of navigation and ship-building. Ships of considerable size, manned by large crews, are portrayed on the bronze kettle-drums which were also a remarkable trait of this period. This culture is often designated Dong-son culture; Dong-son is a village on the right bank of the Song-ma, in the Tanh-hoa province of Annam, and many bronze drums were found here.

Recent investigations have traced the origins of this culture to the Yueh people who inhabited the coastal regions of China about 2000 B.C. Eberhard says of the culture of the Yueh people: “As typical of this coastal culture we may mention the following traits: A developed navigation; the practice of holding boat races, with its outgrowth, the dragon boat festival; the use of bronze drums decorated in a way showing connection with that rite; and the concept of the dragon as river god ... Elements of this culture were the worship of serpents, of sacred mountains (the latter destined to develop into important temple festivals), and of certain trees.” [3] He adds significantly that the whole subject of the affiliations of these early cultures of the Pacific lands can yet be handled only in a tentative manner; it also raises the question of the existence of the many parallels between the Central American civilizations on the one hand and those of Eastern Asia and Farther India on the other.

Besides the bronze kettle-drums, which surprise us by their huge size and by the thinness of their walls, swords, daggers and helmets, household utensils and small statuettes, all of bronze, ornaments of shell and semi-precious stones have also been unearthed near Dong-son. [4] This late Bronze Age culture probably began to penetrate South-East Asia from the north not later than 300 B.C., possibly even as early as 600 B.C., and must have lasted till about A.D. 100. Bronze casting was practised à cire perdue. The big drums are hollow and exhibit no casting-seams inside; clay cores must have been used and the distance between the core and mould, i.e. the thickness of the wall, must have been held by small bronze pieces of which relics are still discernible in some of the drums. The binding between the body of the drum and its ears is finely achieved—it is not known by what process. The alloy is not copper and tin, but copper and lead as in the mints of China in the Han period, and this alloy could be made into a thin flowing liquid which was easy to cast into the thin-walled drums. Later, in Hindu times, the usual copper-tin alloy came into use. [5] The find spots of kettle-drums, glass beads, stone cist graves, etc. in the Malay Peninsula furnish the necessary links indicating the route taken by this culture in its migration from Tonkin to the islands. [6]

We are thus enabled to distinguish two main element in the culture-complex of Indo-China and Indonesia as it had developed before the advent of the Hindus—one a late neolithic element, and the other of a late Bronze Age both developing in Indo-China and spreading south and south-east, though the roots of both may possibly go back to a much earlier time and to remoter districts of China. There is enough evidence at hand to show that it is wrong to think of the pre-Hindu population of these lands as utter savages to whom civilization first came with the Hindus. There is in fact accumulating evidence of a widespread Austric culture, as it is sometimes called, not confined to the archipelago but spreading across the peninsula to portions of north-eastern India.

It is possible, however, that influences from China and India had begun to operate in this area much earlier than is generally believed, though they were not yet strong enough to make much difference to the content of its culture. We have gained from the few excavations that have taken place very valuable evidence of trade contacts in the Bronze Age between South China and Indo-China on the one hand and the mountain lands of Sumatra which attracted gold-seekers on the other. [7]

But evidence taking us still farther back comes to hand from the Philippines where Prof. Beyer conducted a remarkable series of excavations at his own cost during the years 1926-1930. That part of the evidence which most concerns us is thus summed up by R. B. Dixon who visited the Philippines and examined the objects brought to light by Beyer's excavations: “It is from finds in the Iron Age strata which overlie the neolithic deposits that immediate conclusions can legitimately be drawn. These comprise pottery of a considerable range in quality and types of decoration and a very large variety of forms. Secondly, iron implements and weapons such as knives, axes, daggers and spear-points; thirdly, glass beads and bangles, both green and blue, and finally beads of semi-precious stones such as agate, cornelian, amethyst and rock-crystal. It is certain that some at least of the iron objects were of local manufacture, since deposits of iron slag and evidences of iron smelting have been found. It is uncertain as to glass, but unfinished beads adhering to each other in series of half a dozen or more are found, and clear evidence of the repairing of broken bangles. In the earlier Iron Age strata only green glass, whose colour is due to iron, occurs; in the later both this and a blue glass whose colour is due to copper.

“Now both the iron and glass objects are similar to and in some cases identical with the prehistoric glass and iron finds in the south of India. These occur in the dolmen tombs and urn burials which are found by hundreds of thousands, and which almost certainly antedate the historic Chera, Chola and Pandyan kingdoms, whose history goes back to the beginning of the Christian era or before. As finds of similar glass beads and bangles have recently been made in the Malay Peninsula, in dolmen tombs in Java, and in North Borneo, the inference is inescapable that we have clear evidence of a trade contact between the northern Philippines and southern India, running well back into the first millennium B.C. The extensive trade and colonization and later conquests of the South Indian kingdoms, in Sumatra and Java as well as in Indo-China in the early centuries of the Christian era are of course well known. This new material, however, seems to make it clear that this was far from being the beginning of such contacts, but rather the last stages in an association reaching as far as the northern Philippines which had begun many centuries before. In Chinese historical sources, there are a few references to maritime traders bringing typical Indian products to China as far back as the seventh century B.C. These accounts have generally been regarded with incredulity or strong suspicion at least. In view of this evidence from the Philippines the probability of these accounts is greatly increased, with consequences for the history of Chinese culture which are obvious.” [8]

I need not apologize for the length of this quotation; it deals with an important aspect of pre-history which no summary could reproduce exactly. If the facts and arguments of Dixon are accepted as correct—and I see no reason why they should not be—it would follow that South-East Asia was touched by cultural streams not only from the North but to some extent from the West as well, and it seems possible that the sources of some of these reach farther west than India. A Hittite stone bead of about 700 B.C. was found some years ago among a large collection from the Johore river, the bulk of which belong to a date about the first century A.D. when the Roman Empire came into active contact with India and the Far East. [9]

China and India were thus the two main sources from which higher cultural influences kept flowing into south-eastern Asia in prehistoric as well as historical times; the movements were by no means always only in one direction and Indonesian influences can be traced on some aspects of Indian life. For a general estimate of the respective spheres of Indian and Chinese influences, we may well accept the following statement from Bishop at the conclusion of his illuminating paper on the “Origin of the Traction Plow”: [10]

“From China, again, the traction plow travelled to the East India Archipelago, occupation of which it shared with the type from India. Generally speaking, the line of demarcation between the two fields of cultural influence extends, though with many interpenetrations, from east-central Tibet southward through the Indo-Chinese peninsula, thence swinging off in a south-easterly direction into Indonesia. Formosa, the Philippines and North Borneo remain on the Chinese side, while Sumatra, Java and their nearer neighbours fall within the Indian sphere.”

Even within the Indian sphere so defined, the Chinese did establish themselves at selected points from olden times for purposes of trade and formed colonies in course of time; but they always remained colonies of foreigners with little inclination to mix with the local populations, and in contrast to what the Hindus achieved, there is nowhere any trace of the taking over of Chinese culture by the children of the soil. [11]

II. Indo-China

About the beginning of the Christian era when the historical period may be said to commence in Indo-China, the peninsula was inhabited by a number of peoples. [12] The Burmans, still in touch with their congeners, the Tibetans, already occupied the upper course of the Irawady, and the Peguans (Mon) its lower valley and possibly also the valley of the Lower Menam where we find them later. The Thai were still in Yunnan, their original home, where they preserved their independence till the thirteenth century. The Khmers (Cambodians or rather their ancestors) inhabited not only the present-day Cambodia but a good part of Laos and Cochin-China. The Chams, who spoke a language closely related to Indonesian tongues, occupied southern Annam, from Cape St. Jacques in the south up to Tourane in the north. Lastly, the Annamites held the north of Annam and Tonkin. Indo-China was already divided between two civilizations: the Chinese sphere comprising the Annamites, and the Indian sphere embracing Pegu, Kambuja and Campā.

The age of Indian colonization in Indo-China is by no means certain. There is no reference to it in Kautilya, though the contrary opinion has been expressed occasionally. [13] But the third or even the fourth century B.C. is not an improbable date for the beginning of this movement. Buddhist legends relating to the conversion of Suvarnabhūmi seem to afford a valuable clue; and Blagden has succinctly summed up the position in the following words:

“The precise position of Suvarnabhūmi is not beyond doubt but its early missionaries, Sona and Uttara, have long been claimed by Burma as the founders of their branch of the church; and though the tale has been embellished with many legendary accretions in the course of ages, it can hardly on that account be dismissed as being altogether devoid of foundation. Evidence is gradually accumulating from various different quarters which tends to show that Indian influence made itself felt in Indo-china from about the beginning of the Christian era, or possibly even two or three centuries before that date; and there seems to be nothing antecedently improbable in the story of a Buddhist mission being sent there at a relatively early period, though it may well be hazardous at present to fix that date precisely.” [14]

This Indianization of the southern and south-eastern parts of Indo-China must be looked upon as a pacific penetration, proceeding by slow imperceptible stages just like the similar movement that preceded it in the Deccan. Whether these missionaries of Indian civilization came by land by way of Burma, or by sea (possibly after crossing the Isthmus of Kra), their culture prevailed wherever they went. Sanskrit became the official language of the Khmers and the Chams; Hindu beliefs, from Vedic sacrifices down to sectarian beliefs, particularly Saivism, were adopted by them; and with Brahminism came also Buddhism.

III. Burma

Ptolemy's Golden Chryse doubtless included Burma and must have been a translation of Suvarnabhūmi, the classical name of Burma. The central region of the country was called Sonāparānta in Pāli, from Sona (Sk. Suvarna) meaning gold, and prānta or aparānta meaning ‘frontier country’. “Sonāparānta was regularly used in the record of the titles of the kings of Burma, and it was the name given to the territories round the capital in all State documents. It is also to be noted that up to the end of the monarchy, Tampadipa (copper island or region) figured among the royal titles, and this is no doubt Ptolemy's Chalcitis.” [15]

Burma is easily approached only by sea, and indications are not wanting of early maritime connexions of the outside world with Burma. (There were land routes across the north but they do not concern us here.) The author of the Periplus of the Erythraean Sea records that very large ships, called Colandia, sailed to Chryse from the ports on the eastern coast of South India, and Ptolemy mentions fleets from Ceylon following the same course. Musicians and jugglers are said to have arrived in Burma as early as A.D. 120 from the distant Roman province of Syria (Ta-Ts'in). [16] The sea coast of Burma must once have lain much farther north than it now is. “Cables and ropes of sea-going vessels have been dug up at Ayathema, the ancient Takkala, or Golamattika, now quite twelve miles from the sea-shore, and not many years ago remains of foreign ships were found near Tunte (Twantay, close to Rangoon) buried eight feet beneath the surface of the earth.” [17]

The name Talaing, often applied to the Mon people, is said to be a memento of Telingana, the original home of some late arrivals among these people, if not of all of them; this view, however, is by no means universally accepted. Tradition credits to the Telingas the foundation of Thaton in 543 B.C. [18] It also states that disputes between the Brahmins and the Buddhists marked the early years of the new kingdom, and that as a result of the Third Buddhist Council convoked by Asoka at Pātaliputra c. 250 B.C., Sona and Uttara were sent as missionaries to Burma, to ‘revive’ Buddhism there. [19] But all this is highly doubtful, and Asoka himself has nothing to say in his inscriptions about Burma or his mission to it. Equally devoid of foundation are the stories relating to Buddhaghosa, his birth in Burma, his crossing over to Ceylon and his return to Burma with a complete set of the Tripitaka. The Cambodians also claimed that Buddhaghosa came to them. There is nothing in the more authentic sources of Buddhaghosa's life that supports the claims of these two countries.

In historical times we find the Peguans, Mon or Talaing, related to the Khmer and occupying the coastal districts of Lower Burma, and their country, Rāmanyadesa, had for its capital Hamsavati (Pegu). Tradition places the foundation of this kingdom about A.D. 573, but the Indianization of the country appears clearly to have started much earlier. The Mon, who once occupied a wide area in Lower Burma and in the Malay Peninsula, were the earliest peoples of Burma known to history.

To the north of this kingdom lay the land of people who were more nearly related to the Tibetans, and who had Prome (Srīksetra)—Hmawaza—for their capital till the ninth century when it was shifted to Arimaddana (Pagan). During the Prome period this kingdom was also subject to strong Indian influences, though the language of the inscriptions is an archaic Tibeto-Burman idiom, otherwise, unknown; it is convenient, following the Chinese name of P’iao for this early Burman kingdom, to call this idiom employed in the official records of the Indianized Prome kingdom Pyū. In fact an early Mon inscription (A.D. 1101-2) mentions side by side the three ethnic terms Tircul (indigenous name for the Pyū), Mirma (Burmese), and Rmen (Mon), [20] thus clearly distinguishing the three elements of the population. The people who spoke Pyū are best regarded as the forerunners of the Tibeto-Burman movement into the southern parts of the Irawady valley; they had reached the neighbourhood of Prome, where all their known records are found, long before the Burmese came down from the north. [21]

The Pyū seem to have received their veneer of Indian civilization at second hand from the Mon people of the delta. When this happened is not known; but the alphabet of the Pyū records is archaic and contains forms which were going out of use in India even in the fourth century A D. and the Pyū word for gold seems to be borrowed from Mon [22] They seem to have been completely absorbed by the Burmans and left nothing behind except short funerary inscriptions and some bilingual records in Pyū and Sanskrit or Pāli containing mostly extracts from the Buddhist canon. But the Pyū version of the Myazedi inscription shows that the nation was still sufficiently important at the beginning of the twelfth century for its language to be recognized as of equal importance with Burmese and Mon.

The effect of the Burmese conquest which began about the middle of the eleventh century and came to an end five centuries later was less disastrous to the Mon people only because there was Siam to which they could go at first; and in later times, when the Alompra dynasty actively pursued a policy of annihilation of what was left of the Mon, even the indifference of the British government in Tennesserim was found to offer a welcome refuge. But the language no longer maintains a literary standard and has sunk to the level of a patois comprising a congeries of local dialects.

After this preliminary sketch of the background, the reader will be in a position to appreciate the details of the evidence on the Indian influences at work at different times in the early history of Burma. The earliest epigraphical text so far known is engraved on two thin gold plates discovered in 1897 at Maunggun village near Hmawaza in the Prome district. [23] The inscription comprises quotations from Pāli Buddhist scriptures written in a clearly South Indian alphabet of the fifth or sixth century A.D. Pāli Buddhism was quite strong in South India all along the east coast; and the Krsnā valley, Kāñcīpuram and Ceylon, not to speak of less-known places in the Kāvēri and elsewhere, were well-known centres of Buddhism, most favourably situated for intercourse with Burma across the Bay. Dharmapāla, the great rival and contemporary of Buddhaghosa, lived in Kāñcī in the fifth century. The particular connexion of some of the colonies with the Pallavas and their cultural traditions is well known and traced in detail elsewhere in this book. We may note here that Srīksetra, the kingdom of Prome, appears to have had an alternative name Vanavāsi which reminds one of the capital of the Kadambas in the western part of South India; this becomes clear from an old inscription on a metallic image of the Buddha from Prome, which begins: idam Vanavāsīrattha-vāsinam pūjanatthāya, meaning, this is for the worship of the residents of the kingdom of Vanavāsi. [24]

The conclusion suggested by these facts is strengthened by the discovery in 1910-1911 at Hmawaza of part of a stone inscription, also written in Pāli, in characters very similar to those of the Maunggun plates; the full inscription must, it has been calculated, have covered a space of 1.4 metres by 1.75 metres and constituted a large panel in the wall displaying the selected text from the Vibhanga for the edification of the faithful. [25]

Further excavations in the neighbourhood of Hmawaza in the year 1926- 1927 brought to light striking and valuable evidence pointing in the same direction. [26] At a site known as Khin-bha-gōn near the Kalagangon village, a relic chamber of a stūpa containing many finds of great interest was exposed. The chamber was found closed by a stone slab bearing a representation of a type of stūpa having a cylindrical dome with a rounded top and five umbrellas above the hti; “the prototypes of these forms must be sought for in South India” (Duroiselle). In the relic chamber itself, there was a silver gilt stūpa, cylindrical in shape, supporting on its flat cover the trunk of a Bo tree, of which the branches and leaves had broken off and lay scattered about the chamber; the stūpa with the tree is 26” high and has a diameter of 13” at the top and 16” at the base. “Around the drum of the stūpa are four seated Buddhas, each with an attendant monk standing on one side. The stūpa itself is hollow, with no bottom, and is of silver plate with the images repoussé in high relief. The top, forming the cover, is removable and has, round the rim, a line of inscription in Pyū and Pāli, in an early Telugu-Canarese script of South India, very closely allied to that of the Kadambas of Vanavāsi and that of the Pallavas of Kāñcīpura. The character is practically the same as the script of the Maunggun plates.... Each of the passages in Pyū gives the name of the Buddha immediately below it; and after each of these names comes a short extract consisting of a few words from the Pāli scriptures.” The four Buddhas named are those who have appeared in the present Kalpa, viz. Konagamana, Kakusandha, Kassapa and Gotama; the attendants are the four disciples of Gotama, viz. Kassaba (Kassapa), Maulana (Moggalāna), Sari (Sāriputta), and ... da (Ānanda). “Around the lower rim of the same stūpa is another line of inscription, also in Pyū, of which some letters are missing owing to the rim, which is very thin and brittle, having broken off.” The inscription is difficult to interpret, but contains two names Srī Prabhuvarma and, separated from it by a few words, Srī Prabhudevī, possibly the reigning king and queen. Notice the -varman ending of the king, a South Indian feature common to most of the colonies in the East.

Stūpas, images of Buddhas, Bodhisattvas and others in gold and silver, bowls, caskets, symbolical coins of various sizes, beads and other ornaments, and gold and silver plates with inscriptions have been discovered in considerable numbers. Prominent in the last category of inscribed plates is “a manuscript in every way similar to the palm leaf manuscript so common in India and Burma but with leaves of gold, twenty in number, with writing incised on one side. These leaves, within their two gold covers, were found bound together by a thick wire with its ends fastened to the covers by sealing wax and small glass beads. There are two holes in each leaf and cover, through which the gold wire was passed, to keep the whole in position and proper order. It was necessary to cut this wire in order to free the leaves.” Each leaf, 6½ x 1½, contains three lines of writing, except the last but one with four lines and the last with only two. The characters are similar to those of the Pyū inscription round the rim of the large silver stūpa described above, and of the same date (sixth or early seventh century). The manuscript is made up of short extracts in Pāli from the Abhidhamma and Vinaya Pitakas, the Dhamma as preserved in the Tripitakas being an object of worship among the Buddhists. Another gold plate (part of it missing) bearing the text of a well-known formula of the Vinaya and Sūtta Pitaka in two lines of the same early South Indian script was found in the Kyundawzu village in Old Prome. [27] Lastly numberless terracotta plaques have also been found carrying the effigy of the Buddha in the bhūmisparsa mudrā on the obverse, and extracts from the Abhidhamma or the ye dhammā formula on the reverse in South Indian characters of varying dates from the fifth to the seventh century A.D. [28]

In the fifth and sixth centuries Prome was thus a centre of Southern Buddhism—though Mahāyāna is also known, witness the Bodhisattvas—where doctrinal Pāli texts of an abstruse character were studied and the writing employed was of South Indian origin. What is true of Prome is the more so of Pegu. Finot has observed: “It is not impossible that Siam borrowed it from Pegu and transmitted it to her eastern neighbours, and that the inscriptions of Maunggun and of Hmawaza are thus indirectly the beginnings of modern Buddhism in Cambodia. [29]

To a slightly later period belong the seven short Pyū inscriptions on funerary urns bearing dates from 35 to 80 presumably in the Burmese era starting from A.D. 638; these inscriptions reveal the name of three kings with Indian names in mixed Sanskrit and Pāli form, viz. Sūriyavikrama, Harivikrama, and Sihavikrama, and record the dates of their deaths. [30] It may be noted in passing that neither these -vikrama kings nor the -varman ruler noticed earlier find any mention in native tradition. [31] To the same age must be ascribed a broken Buddha from Hmawaza with a bilingual inscription on the pedestal in South Indian characters of the seventh or eighth century; the Buddha is in dhyāna mudrā and the treatment of his dress seems to show Gupta influence; the inscription is in Sanskrit and Pyū, the words in the Sanskrit version being apparently arranged according to Pyū syntax, [32] and not always correct. Another small headless Buddha from the same place bears the Buddhist formula in Gupta characters of the seventh century. [33] Another inscription in Gupta characters of the seventh or eighth century is found engraved in two lines on a bronze bell from VesāIi (Arakan); the language is mixed Sanskrit as in some Jaina inscriptions from Mathurā; the inscription records the name of the Caitya to which the bell was presented and of the donor who made the gift. [34]

We have thus evidence of Sanskrit at Prome from a fairly early period, besides Pāli and Pyū and of the working of North Indian influences side by side with those from South India. Buddhist texts in Sanskrit, however, do not all of them necessarily belong to Mahāyāna; the Mūla-sarvāstivādins, a sect of the Hīnayāna, had also a Sanskrit canon; and “they spread themselves very early over a vast extent of Asia, having settled in Turkestan, China, Indo-China and the Indian Archipelago.” [35]

To complete this account of the Buddhist antiquities of Indian origin at Hmawaza we must add that the clay votive tablets with Pyū and Sanskrit inscriptions range up to the ninth or tenth century A.D., some going up to the eleventh century as well, and some bearing Nāgarī legends possibly under the influence of Nālandā. [36] These and the numerous bronze figurines of Bodhisattvas are distinctly Mahāyānist in character. It has been noted also that the figures on some of these later tablets exhibit decidedly un-Indian facial features; they were clearly of local make and the features must be taken to be typical of the Pyū physiognomy of which we have no other specimens.

Though not strictly relevant to a study of South India influences in Burma, a brief mention of the Arī and their place in the early religious history of Burma may well be considered necessary and useful. [37] The name does after all seem to be related to Ārya; nevertheless, the suggestions that its correct form is Arañ, an abbreviation Pāli Arññaka (Sanskrit Āranyaka), that in early Buddhist literature this term is used to describe the purest member of the Sangha, and that we must suppose that its significance suffered a change for the worse in Burma, turn out to be untenable on phonetic grounds. [38] The Arī wore black, worshipped Nāgas, enjoyed a sort of the jus primae noctis and practised animal sacrifice. Their cult was definitely suppressed by royal edict in the middle of the fifteenth century. They are usually taken to have represented a form of corrupt Mahāyāna Buddhism mixed up with Tantrism and Saivism, and first introduced into Burma from north-east India about the sixth century A.D. But they are not strictly confined to Burma, and some of their practices can be traced in Cambodia, Laos and Siam; it seems possible, therefore, that some widespread, primitive, indigenous cults might have contributed their share to the make-up of the bizarre cult of Ari. In any event, this is not the best side of the Indian influence on Burma, and South India had little part or lot in it.

Hinduism comprising the worship of Siva and Visnu was also known and practised in Lower Burma in early times of which we have been speaking. Vestiges of Hinduism are however not so common in Lower Burma as in some parts of Indo-China. “At Thaton have been discovered three fine bas-reliefs, one representing Visnu caturbhuja, seated, the two others Nārāyana recumbent on Ananta, with a lotus coming out of his navel and supporting the three gods of the Trimūrti.” [39] Prome glories in the name of Pissanumyo, the town of Visnu, and excavations there have yielded a Visnu caturbhuja standing on Garuda. At Pagan itself “we have found a small temple which was evidently consecrated to Visnu; many statues of this god survive and the external walls are decorated with bas-reliefs of the ten avatārs.” Another standing Visnu and a Visnu recumbent on Ananta and supporting the Trimūrtis on lotuses issuing from his navel, both sandstone sculptures, were discovered at Kalagangon, near Hmawaza, in 1920, [40] and these have been assigned to an eighth century date, and the style of art is probably of Gupta inspiration. [41] Hinduism, it has been suggested, counted its followers in ancient Burma mostly among the foreign settlers and colonizers who hailed generally from South India, while the bulk of the Pyū population must have been Buddhists. And not only in Prome, but in the whole of Burma, Saivite remains are rare as compared with the relics of the Vaisnava creed. [42] Coins of the eighth century bearing the Saiva symbols of the Nandi and trident have been found in Vesāli (Arakan), and local tradition seems to point to Bengal as the source of the people who used them. [43] A linga 14” in height was found in the village of Kalagangon meaning ‘the village near the mound by the Indian tank’, near Hmawaza, very near sites which yielded some of the Vaisnava statues and the most important Buddhist relics already noticed. [44]

Lastly, we must make mention of an interesting terracotta plaque first noticed in 1935 and the first of its kind so far discovered in Burma. It is I’ 6” square and 2” thick and made of hard clay. It bears a large sunken medallion in the centre bounded by a circle of beads and portraying “a party of musicians, of whom there are five arranged in two rows. In the upper row are two figures, the one on the left blowing a kind of French horn and the other on the right playing on some uncertain instrument which has broken off. In the lower row, the two figures on either side are beating drums, and the one in the centre, probably the worse for liquor, is dancing, steadying himself on the shoulders of his companions. The figures are well portrayed.... Their style, dress and features are purely Indian. They wear each a necklace of beads, armlets and a dhoti, and the dancer has in addition a piece of linen across his chest. The hair is parted in the centre, and formed into two big tresses falling just over the shoulders and covering the ears. Their bodies are plump, and their faces round.” [45] The half-medallions at the corners were doubtless meant to be completed by similar ones in adjacent plaques, all adorning the base of a fairly large-sized monument. The plaque is said to have be found by a Buddhist monk at the bottom of a tank in Kyontu, Pegu District, while the tank was being cleaned. There is now no means of fixing a definite date for this interesting find.

Besides these sculptures, the large number of Sanskrit words in Talaing and Burman is another strong proof of Hindu influences. [46] Besides the short Pāli sentences in Talaing records attesting the influence of Pāli Buddhism and the pedantry of the composers of the records, there are Indian loan-words in considerable numbers in the Mon text itself; “these form an integral part of the language and are not merely tacked on like the Pāli sentences. They are very common in the early inscriptions, and many of such loan-words have survived through the mediaeval into the modern form of Mon. A remarkable proportion of these words is of Sanskrit origin, not Pāli. Sometimes we find mixed forms, partly Sanskrit and partly Pāli. The Sanskrit forms include some of the commonest religious terms, such as dharma, swar (from svarga), and the like. As to the reason for their presence in early Moil, allowance must be made for the fact that Brahmans, who are often mentioned in the inscriptions, played a great part at all Indo-Chinese courts from Burma to Campā.” Again, in the earlier Mon inscriptions the proper names are Sanskrit, Pāli, or mixed, and this feature persists in mediaeval times when other names of native origin begin to appear. The kings, both Burmese and Mon, “indulged in a double nomenclature: an elaborate Indian name, sometimes of stupendous length, was used by them as their royal style, though they had shorter native names as well, by which (as a rule) they are known in the histories. In the inscriptions the Indian style is given the preference, presumably because it sounded grander and was the specifically royal name, the other one being personal. Certain conventional phrases based on Indian originals are also sometimes used; for example, the people in general are styled ‘the four castes’; although there is no real reason to believe that, apart from the Brahmans, who were of foreign introduction, any real division into castes was recognized in Burma.”

Prof. Blagden has compared [47] the Burmese conquest of the Mon country to the Roman conquest of Greece. Prior to this conquest, the Burmese were no more a race of savages than the Romans were when they conquered Greece. “But just as Rome became in a great measure Hellenized, so the Burmese adopted much from the Mons.” The southern form of Buddhism with its Pāli canon, the particular variety of South Indian script in use among the Mons, and certain useful crafts and ornamental arts were all taken over by the Burmese from the conquered people; and much that is now supposed to be distinctively Burmese was derived from the Mons who had themselves got it from India and Ceylon.

Direct contact with India was also actively maintained during the mediaeval period and, in fact, has continued with interruptions down to our own day. In his account of the reign of Alaungsithu (A.D. 1112-1187) Scott [48] has observed. “The connection with India was still maintained and the forms of many of the Pagan temples suggests architects from the Dekkhan, along with others which certainly point to Singhalese models. Many of the images and their attitudes are quite South Indian, and the square structures with mandapas or porches, instead of the round tumulus, to say nothing of the vaulted chambers and corridor passages, all suggest Indian influence rather than the present conical style of pagoda.” And the presence of a considerable number of South Indian Tamils is attested by the well-known Grantha-Tamil inscription of Pagan attesting the existence of a Visnu temple built there by the Nānādesi merchants, and a gift to the temple made in the thirteenth century by a merchant from one of the port towns on the Malabar coast. [49]

IV.  Fu-Nan

In the early Christian era, the country that later became Kambujadesa was divided into two political entities known to us only by their Chinese names of Fu-nan and Chen-la. Both the States claimed an Indian origin and cherished foundation legends of a very similar character—the rulers of Fu-nan tracing their descent from the union of Kaundinya of the Somavamsa (lunar line) with the nāgī Somā, and those of Chen-la from that of Maharsi Kambu of the Sūryavamsa (solar line) with the apsara Merā. [50]

Fu-nan, it has been pointed out, is the Chinese representation of the word which has survived to this day as Phnom, meaning hill or mountain; [51] the underlying idea is that the capital city of a State, the residence of its king, occupies the same place in the kingdom as Mount Meru, the abode of the Gods, does in the Universe.

Fu-nan occupied the lower valley of the Mekong, the area now designated Cambodia and Cochin-China. Its capital was probably Vyādhapura, or modern Ba Phnom. [52] It was a strongly Hinduised land from the earliest times in which we begin to hear of it. Here is the oldest account of the introduction of Hindu culture into Fu-nan given by K’ang T’ai, a Chinese writer who visited Fu-nan about A.D. 245-250.

“In the beginning Fu-nan had a woman named Lieou-ye (Willow-leaf) for ruler. In the country of Mo-fou there was a man Houen-chen by name who offered worship to a spirit with great love and ardour. The spirit was touched by his extreme piety, and one night Houen-chen dreamt that a man gave him a divine bow and asked him to embark on a boat and set out on the sea. Next morning, Houen-chen entered the temple and found a bow at the foot of the tree which was the home of the spirit. He then got into a large boat and set sail. The spirit so guided the wind that the boat reached Fu-nan. Lieou-ye wished to rob the boat and capture it. Houen-chen raised the divine bow and shot; the arrow pierced the barge of Lieou-ye through and through; she became afraid and submitted, and Houen-chen thus became master of Fu-nan.” [53]

The divine bow is part of the folklore which has been traced by Goloubeuw to Herodotus’ account of the Scythians. The cult of the spirit is the Chinese way of referring to Brahmanism. The location of Mo-fou is unknown, though the cast coast of the Malay Peninsula has been suggested; if this is correct, Fu-nan received its Hindu culture at second-hand from one of the earlier colonies of the peninsula. It seems possible in any event that this story preserves the name of the leader of the first band of Hindu colonists to reach Fu-nan, a leader, whose name and country of origin might well have been preserved by tradition two or three centuries after the event.

This early kingdom has left some traces behind in the form of inscriptions and monuments which are being discovered and identified by the progress of modern research. As in many other sections of Indo-Chinese archaeology and history, Cœdès leads here also. Pelliot has collected all the Chinese texts on Fu-nan and provided an illuminating commentary on them. [54] With their assistance, let us review the early history of Fu-nan from our standpoint.

Under the successors of Kaundinya, Fu-nan seems to have become a great kingdom commanding several vassal states. It is fairly certain that from the second century A.D. at the latest, relations were established between India and China by way of the Isthmus of Kra and of the Malacca Strait; Fu-nan was on this route and must have served as a necessary stage in this long voyage.

Fan-che-man, at first commander of the troops of Fu-nan, and later king, was the founder of the greatness of Fu-nan. He subjugated neighbouring kingdoms and reduced them to vassalage; he fitted out a navy and conquered a good part of the Malay Peninsula; he was the first to assume the title of ‘Great King of Fu-nan’. He fell ill in the course of an expedition against Suvarnabhūmi, doubtless Lower Burma, and died soon after, sometime in A.D. 225-230 or perhaps a little earlier.

The celebrated Sanskrit inscription of Vo-Canh from South Annam, engraved in a definitely South Indian alphabet has been assigned, on palaeographical grounds, to an age not later than the third century A.D. And recently Cœdès has suggested that this most ancient inscription of Campā must be taken to have been the work of a ruler of Fu-nan, and that Srī Māra mentioned therein as the ancestor of the king was no other than Fan-che-man, Fan being the -varman ending which the Chinese took to be a family name. [55] The Vo-Canh record is Buddhist in inspiration, but we shall see that Buddhism was known and practised in Fu-nan under the successors of Fan-che-man.

Fan-che-man was followed on the throne, according to the Chinese sources, by Fan-Tchan, his elder-sister's son who murdered the legitimate heir Fan Kin-cheng. The reign of this usurper is important because it witnessed the commencement of direct official relations between Fu-nan and the princes of India. From a Chinese who had travelled from the west across India to Fu-nan, Fan-Tchan heard of the glories of India and sent one of his relations, Sou-wou by name, as ambassador to India. He embarked from Takkola, which is evidence of the authority of Funan over the west coast of the peninsula, reached the mouths of the Ganges, met the king of the Murundas in the interior, and returned with a Hindu companion and a present of four horses from the king of the Indo-Scythian country. Sou-wou was absent for four years on this mission (c. A.D. 240-4) and these years witnessed many political revolutions in Fu-nan.

Fan-Tchan was assassinated by the second son of Fan-che-man who had come of age, and was in his turn removed by General Fan-siun. It was in the reign of Fan-siun that the Chinese mission of K’ang T’ai and Tchou Ying visited Fu-nan (A.D. 245-250), and from this time regular missions were sent from Fu-nan to the court of China. Fan-siun is credited with a long reign, but a period of confusion seems to have followed. In 357, the Hindu Tchou Tchan-t’an, we learn, ‘called himself king’ and sent an embassy to China. Another three-quarters of a century passes before we hear of the next embassy in 434. But this interval is said to witness another complete transformation of Fu-nan by the arrival of a Kaundinya from Pan-pan who reformed the institutions of Fu-nan on the model of those of India, and completed the Hinduisation of the land; this occurrence may be placed at the end of the fourth or the beginning of the fifth century A.D.

Of the reign of one of the successors of Kaundinya, Jayavarman, we are somewhat better informed. He sent some merchants to Canton who, on their return, were shipwrecked on the coast of Campā together with the Hindu monk Nāgasena who then gained Fu-nan by land. In 484 Nāgasena was sent by jayavarman with presents to China and a request for aid against Lin-yi (Campā); the emperor received the presents thankfully but declined to help against Campā. From the account of the embassy we learn that Nāgasena told the emperor that there was in Fu-nan a mountain called Motan on which Mahesvara descended incessantly and where the plants never withered. To this cult of Siva must have belonged the images with two heads and four arms, or four heads and eight arms, and holding an infant, an animal, the sun and the moon. It is possible that some were Vaisnava images; for the presence of that creed in Fu-nan is attested by inscriptions of the time, as we shall see later. Buddhism was also practised side by side. This becomes clear also from the fact that two monks, Sanghapāla and Mandrasena, who were employed in translating Buddhist texts into Chinese at this time are said to have come from Fu-nan. Let us note in passing that though the Sanskrit texts translated into Chinese by these monks were Mahāyāna texts, we have no reason to conclude from this fact that the Buddhism practised in Fu-nan was of that variety. [56] Jayavarman sent another embassy to China in 503 and got in turn the title ‘General of the peaceful South, King of Fu-nan’.

Jayavarman died in A.D. 514 and was followed on the throne by Rudravarman, his eldest son by a concubine. Rudravarman put to death the younger son of Jayavarman by his legitimate queen. He sent many embassies to China.

The legitimate queen and her unfortunate son have left behind one inscription each, both Vaisnavaite in character, and both engraved in correct Sanskrit verses in South Indian characters of the fifth century A.D. or so. The queen's inscription calls her Kulaprabhāvatī, the chief queen (agramahisī) of Jayavarman; its purpose is to record the foundation by her of a hermitage, tank and temple (ārāmam satatākam ālayayutam). The opening verse of the inscription is a fine invocation of Visnu anantasāyin in Sārdūlavikrīdita metre. [57]

The prince who was deprived of his rights by Rudravarman may well be identified, as Cœdès has suggested, with Gunavarman of the Thāp-muoi record. In this inscription, Gunavarman is said to have been appointed by the king to a religious office in spite of his tender age (bālo’pi) on account of his character (guna), an allusion to his name, and valour (gunasauryyayogāt). In this capacity, Gunavarman consecrated the feet of Visnu under the name Cakratīrthasvāmi, with the aid of Brahmins who were versed in Vedas and Vedāngas and were equal to the gods (vedāngavidbhir amarapratimair dvijendraih, Srutisu Pravināh), and performed an eight days’ ceremony for the purpose. The part of the mother in the function which is alluded to is not clear owing to a gap in the record; we have only the phrase: ātmajananīkarasampra …… How completely the technical phraseology of Vaisnavism is adopted in this record is clear from some other words like padam Vaisnavam, bhāgavataih, and Visnoh paramam prāpya padam. [58]

Close upon these Vaisnava records of Kulaprabhāvatī and her son Gunavarman comes the Buddhist inscrition of Rudravarman himself. The inscription, a long record of eleven Sanskrit verses in different metres, is too damaged for us to understand even the general import of the matters recorded. But enough of it survives in the beginning to attest its Buddhist character (the first two verses are in praise of the Buddha), its authorship (the third and fourth verses praise Rudravarman), and the relation of the author to jayavarman (tat pitrā Jayavarmanā, v. 5). [59]

The last embassy of Rudravarman to China was in 539 when he sent a Buddha relic in the form of a hair twelve feet long.

After Rudravarman, we hear of no other king of Fu-nan, but the annals proceed to narrate the conquestof Fu-nan by Citrasena, the king of Chen-la, whose son Īsanasena sent an embassy to the Souei court of China A.D. 616-17. And the inscriptions of Kambuja reveal us a predecessor of Citrasena, by name Bhavavarman who has left a number of inscriptions, all undated, but most probably belonging to the second half of the sixth century.

The art of this early Hindu state of Fu-nan has not survived in a definitely identifiable form; in fact, the history of Fu-nan is itself a subject of recent discovery and the differentia of the art of this period are still in the process of tentative formulation. Parmentier, whose knowledge of early Khmer art is unrivalled, has succeeded in isolating some characteristic features of the art of Fu-nan; he has done this by looking for motifs that are rare in clearly primitive Khmer monuments and disappear altogether in classical Khmer art, and by adopting the rule that the monuments in which such motifs dominate may well be ascribed to the Fu-nan period. The geographical distribution of the monuments goes far to support this assumption. The two motifs that fully satisfy these conditions are the kūdu, and the somassūtra or the water-spout in the form of a makara. Both these are decidedly South Indian in origin, and the general appearance of the buildings that may be assigned with more or less certainty to Fu-nan recalls the well-known Pallava and Indo-Javanese forms of pyramidal vimānas on a square base characterized by diminishing stages culminating in a sikhara, each stage being ornamented by kūdus, pilasters, etc. The lingas of the period are also seen to have an ovoidal shape with a small face of Siva, features which are continued in early Khmer art for a time. A number of pesanis, or grinding stones for preparing sandal paste, may also be ascribed to this age. Lastly, the considerable number of Visnu statues with the characteristic cylindrical headdress may also without hesitation be ascribed to this period. The wide diffusion of their provenance shows their close association with the earliest Hindu colonies and attests the extension of the empire of Fu-nan. [60]

On the admittedly deceptive grounds of style and general appearance, Parmentier counts among the products of this archaic art of Fu-nan two fragments of beautiful statues, one from Mahā Rosei with an uplifted arm, the other a fine bust with four arms and a striking coiffure and facial appearance, found by M. Dalet at Vat Ari Roka to the north of the province of Ta Kév.

V.  Kambuja

Among the feudatory states of Fu-nan was the land of the Kambujas which the Chinese called Tchen-la. This vassal kingdom had its capital at Sresthapura near Vat Phu. The Kambuja princes traced their descent from rsi Kambu and the apsaras Merā, another version of the recurrent motif of foundation myths of Indian royal families in South India and the colonies. The Kambuja rulers were steadily aggrandizing their power, but we know little of the history of this period. Srutavarman and his son Sresthavarman are mentioned in many later inscriptions as having secured freedom from tribute for their people (apāsta-vali-bandhakrtābhimānāh, v. 13 of Baksei Camkron inscription). [61]

At the death of Rudravarman, the last king of Fu-nan, the succession to the throne seems to have been disputed, and Bhavavarman, king of the Kambujas, who was perhaps a grandson of Rudravarman, [62] seized the occasion for the overthrow of Fu-nan; and in this task he was greatly assisted by his brother Citrasena. This campaign did not result in the total destruction of Fu-nan but only in a diminution of its power and a change of capital for its rulers. Fu-nan no longer held the first place as an Imperial power in the eastern and central portions of Indo-China as it had done for some centuries, and according to the Chinese annals its rulers were forced to migrate from To-mou (Vyādhapura, Ba Phnom) to Na-fou-na, more to the south, to escape the incursions of Tchen-la.

Of Bhavavarman we learn a good deal from the inscriptions. Only one of them, however, may be taken to belong to his reign—the beautiful single line record of Phnom Banteai Neang [63] in Sanskrit verse, announcing the consecration of a linga by the king with the aid of riches won by the use of his bow (sarāsanodyoga-jitārtha-dānaih). The script of the record falls in line with that of the Fu-nan inscriptions and admirably fits its age.

Two other records mention Bhavavarman in greater detail but are of a slightly later period. Only one of them is well preserved and is known as the Han Chey [64] inscription; the object of this long record is to commemorate the consecration of a Sivalinga under the name Bhadregvara at Ugrapura by a loyal and highly favoured servant of two kings, Bhavavarman and his son. In this record Bhavavarman is described as king of kings, who was impregnable in his strength, and like unto Mount Meru:

Rājā srī-bhavavarmmeti patirāsīn mahībhrtām
apradhrsya-mahāsattvah tungo merurivāparah. (v. 2)

More particularly he is said to have overthrown the mountain kings (jitvā parvatabhūpālān, v. 10), a clear reference to the rulers of Fu-nan. [65] Another record of a slightly later date, A.D. 668, states that Bhavavarman took the kingdom by force, svasaktyākrānta-rājyasya. [66]

Bhavavarman had a sister; her name is not given, she being called simply the daughter of Viravarman. She married a learned Brahmin, Somasarman, and had a son, Hiranyavarman. Somasarman was first among the knowers of the Sāmaveda, and he established images of Tribhuvanesvara and the Sun with great éclat, and arranged for the permanent exposition of the Rāmāyana, the Purānas and the entire Bhārata, copies of which he presented to the temple. [67] The record is undated, but its script would place it between the inscription of Bhavavarman and the two others in which he is mentioned.

If now we turn to the Chinese annals, they seem to tell a slightly different tale of the rise of Tchen-la. The Souei annals of China say that Citrasena, the king of Tchen-la, overcame Fu-nan which was formerly suzerain of Tchen-la; after his death, his son Īsānasena succeeded him. The same source also mentions an embassy to China from Tchen-la in A.D. 616. The New History of the Tang ascribes the conquest of Fu-nan to Īsāna himself in the period 627- 649. [68]

There are some inscriptions of Citrasena, all bearing very close resemblance to the South Indian Pallava inscriptions of the early seventh century. One of them from Thma-kre, meaning stone-bed, from a large level rock in the bed of the Mekong between Sambok and Kratié, is a single anusthup verse recording the erection of a linga by Citrasena after obtaining the permission of his parents. [69] The other record is found in two places, Phou Lakhon in Laos [70] and Khan Thevada in the province of Ubon. [71] It comprises three verses in the same anusthup metre. It opens with the statement that the grandson of Sārvabhauma, the younger son of Viravarman, was not inferior in prowess to his elder brother, Bhavavarman; then it says that this younger son was Citrasena who took the name Mahendravarman at his consecration, and after having conquered the entire country set up a linga of Girisa (Siva) on the mountain as a symbol of his victory.

Jitvemam-desam [72] akhilan Girisasyeha bhūbhrti lingannivesayāmāsa Jayacihnamivātmanah

These events, the liberation of Kambuja and the erection of the linga, must have taken place a little before A.D. 616; in fact, the nearly contemporary Souei annals cited above place them between A.D. 589 and 618, and this is in perfect accord with the date unmistakably revealed by the palaeography of the inscriptions of Citrasena. It is clear that at the time of the first record he had not yet become king.

It will be recalled that about the same time another Mahendravarman, the first of that name and most talented among the Pallava rulers of South India, erected a shrine to a linga on the rock of Tiruchirapalli overlooking the Kāverī river. Considering the very close resemblance in the lettering of the inscriptions of the two Mahendravarmans, one is tempted to ask whether this is not more than a mere coincidence. Separated by several hundreds of miles of land and sea, the records of these two rulers are evidence of exactly the same type of culture, same in almost every detail that can be thought of.

Let us now see what the inscriptions reveal of Īsānavarman, doubtless identical with Īsānasena of the Souei annals. A number of inscriptions have come down to us from his reign, not to speak of references to him in the records of his successors. The script of all of them belongs to the same South Indian variety whose spread and growth in the various colonies constitutes one of the chief attractions for students of the subject in South India. Some are not dated but mention the king by name; for instance, the Svai Chno inscription [73] (in the province of Phnom Penh) recording the foundation of an āsrama by Arya Vidyādeva, and the Ang Pou (in the province of Trēang) [74] record commemorating the consecration of Harihara image and an āsrama to Bhagavat by Muni Īsānadatta. The reference to Harihara in the second record is in these words:

Sankarācyutayor-arddha-sarīrapratimāmimām, i.e. this image of which one half is Sankara and the other Acyuta. The record contains a Khmer part also rather carelessly engraved.

There are, however, two records of Īsānavarman clearly dated Saka 548 and 549 (A.D. 626 and 627). The first from Vat Chakrat (in the province of Ba Phnom) [75] praises the king's valour and fame; it refers to a vassal ruling over Tāmrapura who is said to have long enjoyed the privilege of subjection to the three cities of Cakrānkapura, Amoghapura and Bhīmapura; this vassal king obtained the permission of his suzerain and installed an image of Harihara (haritanu-sahitam sthāpayāmasa sambhum). The wide popularity of the Harihara cult in this period in Indo-China is very well attested by its epigraphy and sculpture. The second record comes from Sambor [76] and comprises fifteen Sanskrit verses in an excellent state of preservation. The opening verse is an invocation to Kadambesvara; five verses follow in which the valour, policy and fame of Īsānavarman are praised; the next three verses introduce Ācārya Vidyāvisesa appointed by the king for the supervision of all his spiritual affairs; this ācārya was well-versed in many spheres of learning, particularly in Sabda (grammar), Vaisesika, Nyāya, Sāmkhya (Samīksā) and Buddhism; he was an eloquent poet and knew the ways of the world. In his great devotion to Īsāna—note the double reference here to the king, his master, and to Siva, his deity—Vidyāvisesa erected a lingam (v. 10), and presented to the shrine the village of Sākatīrtham with its servants, cattle, fields, etc. (v. 11); a Brahmin pāsupata appointed by the king was to be in charge of the worship in the temple to the end of time (v. 12). The record closes with an exhortation for the continued maintenance of the foundation (v. 13), the date with full astronomical details (v. 14), and the mention of some fresh dignities and gifts bestowed by the king on Vidyāvisesa (v. 15).

The inscriptions attest the power of Īsānavarman and the prosperity of his reign, but say little directly on his part in the completion of the conquest of Fu-nan. In fact, it is a little difficult to determine exactly the parts played by Bhavavarman, Mahendravarman (Citrasena) and Īsānavarman in the elevation of Tchen-la at the expense of Fu-nan. We may suppose that Citrasena probably assisted his brother Bhavavarman as commander of his forces and that the first conquest of Fu-nan in which the brothers took part meant the capture of the northern provinces of the empire of Fu-nan, a surmise supported by the absence of any inscriptions in the Ba Phnom area of a period earlier than the reign of Īsānavarman. But there was perhaps a raid or raids on the capital itself which may have accounted for the change of capital to Navanagara recorded in the Chinese sources. The definite overthrow and occupation of the Ba Phnom region must have been the work of Īsānavarman. [77]

It is not necessary to follow the history of Kambuja any further; for our aim in this study is just to draw attention to the most significant factors in the Hinduisation of Indo-China and assess the role of South India in this process. But there are still a few inscriptions of this early period which remain to be noticed. The first part of the inscription of Ang Chumnik, [78] dated Saka 551 (A.D. 629), records the reconsecration of a Sivalinga and the temple called Rudrāsrama by Ācārya Vidyāvinaya, and though it does not mention the name of any king, Barth thinks that it may be assigned to the reign of Īsānavarman.

More interesting, if somewhat enigmatic, is the earliest inscription from the temple of Bayang. [79] It bears two dates in the Saka era, 526 and 546, corresponding to A.D. 604 and 624. Though the lettering of the record is wonderfully conserved and looks as if it were fresh from the hands of the engraver, the stone being of very fine grain has peeled off in many places causing gaps in the inscription which seriously hamper its proper interpretation though its general sense is clear. It refers to the erection in A.D. 604 of a temple where the feet of Siva were worshipped: girisasya padam (v. 5), sambhoh padasyedam (v. 8), pasupati padabhāk (v. 10), padam aisam (v. 11), sivapādāya (v. 12), and of a tīrtha (salilasthāna) [80] attached to the temple, both by a Brahmin named Vidyābindu. [81] This inscription, which most probably spans the reign of Bhava, Mahendra and Ïsāna by the two facts recorded by it, is of great interest in many ways. Like all the other inscriptions of the time its characters are unmistakably South Indian, and if its provenance were not known, no epigraphist could distinguish it from, say, a Pallava inscription of the seventh century. Moreover, its language is flawless Sanskrit, and there are employed many terms of technical import in Pāsupata lore. As this is of some importance, no apology is needed for transcribing here the second half of the opening verse and the succeeding lines where these terms mostly occur and of which no translation can reproduce the impression created by the original.

Yam antaran jyotir upāsate budhā

niruttaram brahma param jigīsavah (1)

tapassrutejyāvidhayo yadarppanā

bhavantyanirddesyaphalānubandhinah

na kevalam tat phalayogasanginām

asanginām karmaphalatyajām api (2)

nisarggasiddhair animādibhirgunai-

rupetam angīkrtasakti-vistaraih

dhiyām atītam vacas (ām agocaram) [82]

(anā)spadam yasya padam vidurbudhāh (3)

The entire apparatus of the religious experience and philosophical thought of India as specially adapted by the Pāsupatas is here, and the particular reference to yoga is noteworthy as making an interesting phase in the development of saiva religious practice. [83] The dominant position of the Pāsupatas in Kambuja and even earlier in Fu-nan is well attested from several sources. We have noticed above that the Sambor inscription of Vidyāvisesa provides that a Brāhmana Pāsupata should offer worship to the linga set up by him. Besides the other ācāryas noticed above, we find another mentioned in the inscription of Phnom Prah Vihar of the time of Bhavavarman II—a royal preceptor who was a Pāsupatācarya of the name Vidyāpuspa, a poet who was an adept in Sabda, Vaisesika and Nyāya. [84] There is no need to gather here all the instances from the inscriptions of Kambuja; but the atmosphere of saivism which prevailed in the court and dominated the minds of the court-poets is best illustrated by a verse in an inscription of the commencement of the reign of Jayavarman V, A.D. 968. The king is compared to Srīparvata in a manner that brings out forcibly not only the saivite leanings of the author of the verse but also the source of Kambujan Saivism. The verse reads as follows:

Daksināpatha vinyasta-sarassiddhiprado’ rthinām Yuktam yo yuktinipunais-srīparvata itīritah [85] '

“Distributing his wealth by way of daksinā (having his essence in the Deccan), giving success to those who come to him with solicitations (giving siddhi to those who desire it), he received logically from experts in logic the name of Mountain of Prosperity (Srīparvata).”

This striking allusion in the tenth century to the famous centre of Saivism in the Deccan shows the strength and continuity of South Indian influences on the culture of the colonies.

Tchou-ta-kouan who visited the capital of Kambuja in the thirteenth century found Saivism still flourishing there.

Lastly, the Bayang inscription marks the modest beginning of a celebrated shrine, the history and architecture of which have been the subject of an illuminating study by Henri Mauger in recent years. [86] He thinks that the Girisa of our inscription was a large bronze image of Siva of which only the feet have been recovered in recent excavations on the site. It must have been in every way a remarkable work of art to which these beautifully modelled feet belonged. Judging from the size of the feet, the entire statue must have been more than life-size, say, nearly two metres in height. “By the naturalism of its modelling," writes Mauger, “this work is clearly pre-Angkorian; to judge it by the care for detail and by the delicacy of execution of this humble fragment, the divinity must have been of such beauty, assuredly, as to cause the construction of a prasat as imposing as this original sanctuary.” He thinks that the words in the inscription, padam aisam, vinibaddham istakābhih, taken along with anāspadam yasya padam vidurbudāh of verse 3 cited above, imply that this large bronze had no pedestal, and that it was stood on a brick platform together with a large slab of stone with rounded corners into which the legs had been fitted and which did not permit of the heavy tenon below usual in such statues. I must say that while I have cited the corresponding Sanskrit words of the original inscription, M. Mauger seems to have relied throughout on Barth's French translation which gives an air of plausibility to Mauger's views. But the Sanskrit original is shattering in its effect on Mauger's airy structure, and I think we are in the presence of one of the clearest examples of the dangers of rearing far-reaching theories on the basis of translations. There is nothing in the original to indicate that anything more than the feet of Siva formed the original object of worship in the shrine, and the words padam aisam vinibaddham istakābhih mean simply that an image of Īsa's feet was made of bricks (and mortar), not stone. Consider also the number of times the word for ‘feet’ is repeated in the course of the inscription, as pointed out above, and it becomes certain that we have here an instance of Sivapāda just like Buddhapāda and Visnupāda which we find in many other shrines. And it is very doubtful if at the beginning of the seventh century the technique of the art of bronze-casting had attained the efficiency required to manipulate the mass of metal needed to cast so perfectly such a large statue as that indicated by these feet. It would be more reasonable to suppose that this statue was a work of the eleventh century or the tenth at the earliest, the date of the apogee of bronze-casting in Southern India under the Cōlas. The beginnings of the Bayang temple were therefore much more modest than Mauger supposes; it began with a brick image of the feet of Siva.

Two facts of particular interest remain to be mentioned before we take leave of early Kambujan epigraphy. First is the direct reference to the rulers of Kāñcī, i.e. the Pallavas, in a eulogy of Jayavarman I (latter half of the seventh century) in a context which is unfortunately not easy to make out on account of a break in the stone; the phrase is ā-Kāñcīpura-nrpā. [87] The other is the reference to Bhagavān Sankara, the great South Indian teacher of Advaita Vedānta, in an inscription of the reign of Indravarman I, dated Saka 80x, i.e. between A.D. 878 and 887. Sivasoma, the royal guru, is thus described in this record:

Yenādhītāni sāstrāni bhagavac-chankarāhvayāt
nissesasurimūrdhāli-mālālīdhānghripankajāt [88]

“He learned the sāstras from him who is known as Bhagavan Sankara, and whose lotus feet are licked by the row of bees, i.e. the heads of all scholars." There can be no doubt that for many generations, in fact, for centuries after they first established themselves in the lands of the East and began the work of civilizing and Hinduising these lands, the leaders of Hindu society in the colonies eagerly kept up a live contact with the original springs of the great culture of which they were the carriers into distant lands.

Some of the old Hindu ceremonial has survived in Cambodia to this day, and a European observer has recorded in much detail the elaborate formalities attending the Cūlā-kantana-mangala (the auspicious tonsure) of a prince royal at Phnom-Penh at the beginning of the current century. The Cūdākarma, as is well known, is one of the samskāras of the ancient Indian manuals of domestic ritual; it is performed in the royal household of Cambodia today by court Brahmins called Bakus under their ācārya, and the ceremony as it is now practised contains a large admixture of Buddhist forms. But the Khmers still say that this tonsure at the age of puberty was instituted by Prah Iso (Siva) who himself shaved the head of Prah Kenes (Ganesa) when he was eleven years old, at Mount Kailās. [89]

It has been observed with justice that while ancient Brahmanism has left many vestiges of a material nature in the form of temples, images and so on, not much of its influence in the moral or spiritual sphere has survived in modern Cambodia. Hindu deities have been absorbed by Buddhism and relegated to subordinate positions in its system, and the beliefs and ceremonies which are not of Buddhist origin in modern Cambodia are related not to Brahminism but to old animistic conceptions widely spread among the savage tribes of Indo-China. It seems probable, therefore, that the strong Hindu influences that came into the land in the most ancient days of which we have spoken above were effective only with the aristocracy and the court circles, and that to the masses at no time did they give anything more than a superficial veneer. [90]

The most important survival of Brahminism at Phnom-Penh today is the existence of the Brahmins of the Court, the Bakus, whose part in the tonsure ceremony we have just mentioned. The name Bakus or Bako has not been satisfactorily explained. The Bakus are distinguished from others around them by their long hair and their Brahminical cord (upavīta). It is from this class which practises some abstinences and enjoys certain privileges that the priests who play an important part in ceremonials are recruited. “At the royal palace they are in charge of certain old cult images in metal which they guard together with the sacred sword, and carry behind the king when he takes command of his armies. They prepare the lustral waters, and take them to the king in gold-tipped conchs when the king performs his ablutions and purifications for the new year and on other important ceremonial occasions, as also both before and after a battle; they recite the mon (mantra) and akom (āgama) or mystic formulas; they light a dozen sacred candles and carry them in the pradaksina path the prescribed number of times; at the cremation of princes, they light the pyre if the king does not himself perform this last duty; in a word, they conduct all the sacred ceremonies of the palace, or rather they assist the king who is the supreme priest.” [91] To this account given by Aymonier, of the functions of the Bakus, Cœdès adds other details. The court Brahmins still play a part in the administration of oaths to officials, in ploughing the first furrow, and at the ‘Feast of the Waters’. The prayers they recite or chant on such occasions are in corrupt Sanskrit, often unintelligible, but still written in the grantha characters of South India. The writing is palaeographically much later than that of ancient Kambuja; this proves, Cœdès thinks, that these Brahmins are not direct descendants of the ancient Brahmins; but this is not a necessary inference.

The Brahmins of Cambodia are also Buddhists like the other Cambodians, and frequent Buddhist temples during festivals. The small chapel in the royal palace where they jealously guard the sacred sword, the palladium of Khmer royalty, contains, besides Brahminical idols, images of Buddha and even a magnificent Lokesvara dating from the days when the cult of this Bodhisattva was popular.

VI.  Campā

Campā on the east coast of Further India (11° to 18° N.L.), the present Annam, formed a half-way house between Java and China, and had a large part in the spreading of Hindu culture in the Far East. This name has generally been held to have come from Campā, the capital of the Anga country in the lower Ganges valley; [92] but it may be recalled that this was also the name of the ancient capital of the Cōla country, Kāvēripatnam, which was also a famous seaport. The oldest inscription in the region of Campā so far known, the Vo-Canh rock inscription, is decidedly South Indian in its script, and the name Campā may well have come directly from that quarter. Some Chinese authors place the foundation of Campā in A.D. 137; Marco Polo mentions it at the close of the thirteenth century; it was overrun by the Annamites (Yavanas of the late Campā inscriptions) at the end of the fifteenth century. Today the Cams are few in number, about a hundred-thousand, confined to the province of Phan-rang, the ancient Pānduranga. Ancient monuments [93] are present only in the provinces of South Annam. They are all in brick, stone being used only for the gates and for decoration.

The inscription of Vo-Canh dates from the third century A.D., or even the second. It is only partly legible and mentions the line of Srī-Māra to which the king belonged. The inscription is clearly Buddhist in inspiration, though its author was no adept in the doctrine. Fifty years ago Bergaigne [94] compared the script of this inscription with that employed in the Girnār inscription of Rudradaman and the contemporary Sātavāhana inscriptions at Kānheri, and reached the conclusion that the Vo-Canh record was anterior to the fourth century A.D. and might even go back to the second. The third century, he said, would be a good date for it, and it would be one of the most ancient records in Sanskrit. The progress of Indian epigraphical studies since then has confirmed the estimates and fixed the definitely South Indian origin of the earliest phases of Hindu culture in those distant lands. This result again should cause no surprise if we recall that even Ptolemy knew of geographical names of Sanskrit origin belonging to this region and to the archipelago.

The Vo-Canh record and the Srī-Māra line, however, do not belong to the history of Campā; they are, as we have seen, relics of the time when lower Campā at least formed part of Fu-nan, [95] one of the oldest Hindu kingdoms of Indo-China of which we have any knowledge.

The earliest inscriptions from Campā proper are the inscriptions of Bhadravarman of about A.D. 350. [96] One of them, the Cho’-dinh inscription, mentions a sacrifice performed on behalf of the Dharmamahārāja Bhadravarman or possibly one of his descendants. The title of this king is clearly derived from South India where the Pallavas and Kadambas are known to have employed it; it means ‘the great dharmic ruler’. The -varman ending of his name, henceforth a regular feature in the names of the rulers of Campā, also recalls the practice of several South Indian dynasties like Sālankāyanas, Kadambas and Pallavas. Again, the record is engraved in bold box-headed characters very similar to those of Vākātaka inscriptions on the one side, and the Kutei inscriptions of Mūlavarman on the other. In fact, as Bergaigne pointed out, [97] the comparison of the alphabets of the Indian and Farther Indian inscriptions leads us to two conclusions of considerable importance. First, the stage of development exhibited by the letters of an inscription and its general appearance furnish a very reliable datum for determining its age; secondly, the closely parallel development of writing in the mother country and in the colonies implies incessant intercourse among them in those far off times.

The Cho’-dinh inscription is a very short record; its age, the excellence of its preservation and its value as evidence of prevailing religious beliefs and practices justify the reproduction of the text of the record.

(I) namo devāya bhadresvara-svāmi-pāda-prasādāt agnaye tvā justam karisyāmi (2) dharmamahārāja srī bhadravarmmano yāvaccandrādityam tāvat putrapautram moksyati (3) Prthivīprasādāt-karmmasiddhir-astu. This means: “Homage to God! By the favour of the feet of Lord Bhadresvara, I shall make thee pleasant to Agni. As long as the sun and moon endure, he will release the sons and grandsons of the great dharmic king, Srī Bhadravarman. By the favour of the Earth, may the sacrifice be successful.” This is followed by a short inscription in smaller letters of the same type: Sivo dāso baddhyate, meaning “propitiatory dāsa is bound (to the sacrificial post).” [98] This inscription, like the Yūpa inscriptions of Mūlavarman of Borneo, attests the prevalence of faith in the Vedic religion of sacrifice; and it is unique in its employment of liturgical formulae. To cite Bergaigne once more: [99] “The formula agnaye tvā justam karisyāmi, for instance, appears to be borrowed from a ritual very similar to those of the Srauta- and Grhya-sūtras, while the addition of Bhadresvarasvāmipāda-prasādāt places the ceremony under the auspices of Siva, and also attests an advanced stage of Saivism, the deity being adored, according to a custom we shall find perpetuated at Campā, under a name recalling that of the king who raised the temple to him. There is no introduction other than the invocation namo devāya, and no conclusion other than the formula prthivīprasādāt-karmasiddhir-astu. Yet this inscription incised with an admirable regularity, in deep and large-sized characters must be something more than the simple fancy of an idle priest.” The postscript seems to imply human sacrifice, and there is perhaps nothing to prevent our accepting this for a fact seeing that offering human sacrifices to propitiate Siva is mentioned in the Atharvaveda and the Mahābhārata. [100]

Another inscription of the same age is engraved on a rock called Hon-cut about 28 kilometres south-south-east of Tourane and contains only an invocation of the Lord Mahādeva Bhadresvarasvāmin. This short record contains a bad error in Sanskrit grammar, employing one dative between two genitives, and other examples of a similar nature occur also in the inscription to be noticed next. But the record is so much like the Cho’-dinh inscription that they must both be assigned to the same ruler.

The shrine of Bhadresvara alluded to in these two records is the earliest royal linga of the Far East and is represented today by an imposing array of ruined structures in the village of My-son, eight kilometres to the south-east of the Hon-cut rock. [101] In front of the main temple in this group was discovered a stele of a type generally found among the early antiquities of the Malay Peninsula and the archipelago. The slab which is two metres high and one metre broad is inscribed on both sides in the same characters as the two preceding ones. Parts of the inscription have suffered damage, but enough remains to reveal the language of the inscription and its purpose. It records the grant to Bhadresvara of the entire valley enclosed by three mountains in which his temple is located and the gift is described as an aksaya-nīvī after the Indian manner. There is no doubt that this inscription is the foundation charter of the temple; but it bears no date. The language of the record is faulty Sanskrit, and one peculiarity in its orthography may well be taken to point to the Telugu country as the original home of the colonists: Prithivī is written for prthivī, and duskritam for duskrtam, and we know that to this day these words are pronounced by the people of the Andhra country nearly as they are found written in the My-son record, though sometimes a u sound takes the place of the infixed i. Finot has rightly said: [102] “The fact that the three inscriptions are all in the name of Bhadravarman proves, besides, that the Chams formed a unitary state and ‘not a series of independent petty kingdoms’.”

Contemporary with the Sanskrit inscriptions of Bhadravarman I is an inscription of three lines engraved in large characters on a rock face two metres in length and one metre in height at a place about a mile to the west of the ancient city of Tra Kieu. It was discovered in 1935; the script is the same as that of the Sanskrit inscriptions, but the language is Cham, though the record opens with Sanskrit siddham and contains the words nāga, svargga, paribhū, naraka and kula, all Sanskrit. It mentions a king but not his name. Its purpose is to invite attention to a holy nāga of the king, perhaps enshrined in a neighbouring temple of which some relics are still traceable, and to invoke the joys of heaven for those who treat it with respect and threaten any one who insults the shrine with a thousand years in hell for himself and seven generations of his family. The interest of this record is two-fold. It attests the early prevalence of the nāga cult in Campā. And it is the earliest text known in any Malayo-Polynesian dialect. It is three centuries earlier than the earliest Malay inscriptions of Srī Vijaya, which belong to the close of the seventh century A.D. [103]

The fine bronze Buddha of Dong-düöng and the inscribed vessels from the ‘treasure’ of La-tho, [104] both from the province of Quang-Nam go to confirm our opinion on the original home of the early colonists of Campā. The Buddha statue over a metre in height is a beautiful work of art in the true Amarāvati style; it is a finely modelled standing figure, with the right hand (of which the palm alone has survived as a broken fragment) in cinmudrā and the left in kataka; the treatment of the robe which leaves the right shoulder bare and falls in a straight fold at the back is unmistakably inspired by Amarāvatī art. It is surely no accident that the region of Dong-düöng also bears the name of Amarāvatī. From La-tho we have a platter and a pitcher both inscribed in South Indian characters of a very early age, not later than the sixth century A.D. The platter is made of an alloy in which silver predominates, and the inscription on it reads: Sri-vanāntesvara. The pitcher is of silver and bears a sloka:

Vanāntaresvarāyāsmai Srīmate divayakīrttaye

Campāpurapati raupyam kalasam sraddhyātmanah

Here is a king of Campā presenting silver vessels to a Siva temple and recording his act in a correct Sanskrit verse.

Campā seems soon to have embarked on a policy of expansion northward and come into rather sharp conflict with China. She seems to have sought in vain the aid of Fu-nan in this adventure. [105] The story of the war that followed as given by the Chinese is of interest to us as giving some clue to the conflict between the two civilizations contending for supremacy in these regions and to the considerable wealth that the Hindu temples of Campā had already accumulated in gold and otherwise. Campā suffered terribly in the wars against China (A.D. 431-46); not only did Fan Yan-mai, for that was the name of the king, fail to realize his ambition of extending the power of Campā northward at the expense of China, but he lost everything; the whole of his country was occupied by the Chinese, and his capital and all the temples in the kingdom were pillaged. The idols alone when melted yielded, we hear, a hundred thousand pounds weight of pure gold. This subjection of Campā was only temporary; how it came to an end we do not know.

Another stele from My-son, broken and mutilated, gives the first dated inscription [106] of Campā and one of considerable importance for the further history of the Bhadresvara shrine. The record must have contained three dates at least: the date in which the temple of Bhadresvara was burnt down in a fire, that of the death of Rudravarman and that of the reconsecration of the new temple by his son and successor Sambhuvarman. Only the first of these dates has been preserved, and in it only the figure for the hundreds, thus¾yuitaresu catursu varsasatesu sakānām vyatītesv agnidagdham devadevālayam. This places the occurrence in the fifth century of the Saka era, between A.D. 479 and 577. The new temple bore the name Sambhubhadresvara, prefixing that of the renovator to the original name of the shrine. The inscription speaks of Campādesa and is the earliest to do so.

Campādesa janayatu sukham Sambhubhadresvaroyam.

The writing in this record exhibits some traits common with that of Bhhadravarman's inscriptions, but has undergone several modifications. [107]

About 15 kilometres from My-son, in the village of Tra Kieu in the province of Quang-nam (Annam), recent excavations have led to the definitive location of the most ancient capital of Campā, called Simhapura in the inscriptions. This location first suggested by Pelliot and Aurosseau on the basis of a Chinese description of the Cham citadel, has now received striking confirmation from the field work of J. Y. Claeys. [108] Dominating the town towards the east was an important group of shrines devoted to the Saivite cult, although Vaisnavism also seems to have been held in honour, as is shown by the inscription of Prakāsadharma (A.D. 650-679) recording the construction and dedication of a temple (pujāsthānam) to Vālmiki; the sage, says the record, was an incarnation of Visnu, and in his grief, he uttered a verse that was highly respected of Brahmā:

Yasya sokāt samutpannam slokam Brahmābhipūjati Visnoh
pumsah purānasya mānusasyātmarūpinah. [109]

The numerous lions and elephants sculptured in relief and in the round in the principal group of temples distinctly recall the Kailāsa temple of Ellorā. This group comprised eight temples. “The principal shrine must have been a building remarkable not only for its vast dimensions but also on account of the quantity and quality of sculptures which supplied its plastic decoration. In the middle of this sanctuary there stood a sandstone altar of imposing size, adorned all round with a frieze in high relief, representing a succession of musicians and female dancers. The eight temples were raised on platforms decorated with raised ornaments and mouldings. They were built of brick, as is indeed the case with all the monuments constructed by the Chams.” The temples were easily accessible from the sea by way of the adjoining river. The palace of the king and the residences of palace servants must have adjoined this group of temples, and the whole city was surrounded by a massive wall which protected it against damage from the annual floods in the river.

The inscriptions of Prakāsadharma are of great interest from several points of view. We have just noticed his foundation of a temple of Vālmīki, and the reference in the foundation charter to the story of the meeting between Brahmā and Vālmīki found in the opening cantos of the Bālakānda. In another inscription of this king at My-son we find another episode of the Rāmāyana, this time from the Uttarakānda summed up similarly. The occasion is furnished by the foundation of a temple to Kuvera Ekāsapingala. [110] These two inscriptions establish beyond doubt the vogue in Campā of the seventh century A.D. of the text of Vālmīki's Rāmāyana in the form in which we now have it. And if we recall that the temple of Vālmīki was not a new foundation but a renovation of an already existing shrine¾pūjāsthānam punastasya krta … is the inscription, it becomes probable that the currency of the epic goes back much further. In another of his inscriptions Prakāsadharman is himself compared to Rāma, the son of Dasaratha, for his nobility and valour, and for the prosperity of his reign: [111]

aviratanaradevabrahmavasyas svatejah samitaripusanātha(h) srīsamutsekahetuh

Dasarathanrpajo’yam Rāma ityāsayā yam srayati vidhipurogā srīr aho yuktirūpam.

Surely, no greater proof could be needed to show that Vālmīki's great poem enjoyed the same hold on the imagination of the literati in Campā as in Kambuja and India.

The worship of Visnu was, Mus has suggested, [112] introduced into Campā by Prakāsadharma from the Kambuja country. The inscriptions say that his father Jagaddharma somehow reached Bhavapura where he espoused the princess Sarvānī, daughter of Īsānavarman, and the several Vaisnavite inscriptions of Kambuja in this period reviewed in the section on that country sufficiently attest the Vaisnava persuasion of its royal family. [113] But it is difficult to believe that Īsānavarman, with his capital at Īsvarapura, or his daughter bearing the name of Siva's spouse, were Vaisnavas in the sense of being exclusive worshippers of Visnu. But there is no doubt that the relics of Visnu worship are more numerous and date from an earlier time in Kambuja than in Campā. The cult of Brahmā and of Harihara mentioned in inscriptions though no image is known, as also Buddhism, were known in Campā. [114]

The My-son inscription of Prakāsadharman (A.D. 657) contains also the legend of the foundation of the Kambujan kingdom by Kaundinya in the form it had taken at the date of the inscription. After the mention of Bhavapura in verse xv, we read:

(tat)ra sthāpitavāñ-chūlam Kaundinyastaddvijarsabhah
Asvatthāmno dvijagresthād Dronaputrādavāpya tam (16)
.… kulāsīdbhujagendrakanyā Someti sā vamsakarī prthivyām
āsritya bhāvetivisesavastu yā manusāvasam uvāsa (17)
Kaundinyanāmnā dvijapungavena kāryārthapatnītvam anāyi yāpi

Bhavisyato’rthasya nimittabhāve vidher acintyam khalu cestitam hi (18)

Here Kaundinya got a trident from one of the heroes of the Mahābhārata, the Brahmin Asvatthāma, son of Drona, and in some mysterious manner this enabled him to espouse the Nāgī maiden Somā who had then taken to a human mode of life and enabled her to become the founder of a royal line on earth (vamsakarī prthivyām); the whole episode is represented as the inexplicable result of the working of Fate. We are thus in the full flood of the cycle of Nāgī legends that are known very well to Tamil literature and to the relatively late Amāravatī stone inscription of the Pallavas where Asvatthāma's liaison with Madanī, an apsaras, gives rise to the Pallava line of kingsso called because the offspring of the alliance was cradled in a litter of sprouts (pallava). [115] We may note in passing that Kaundinya, Kambu, Bhrgu and Agastya were names warmly cherished in the colonies as the symbols of the great work of Hinduising and civilizing these extensive lands in which learned Brahmins, actuated by a high sense of the duty they owed to their fellow-men to give them of their best, took the leading part.

The cult of Bhagavatī held an important place in the minds of the ancient Chams, and in a study of South Indian influences on the colonies, this fact deserves more than a passing mention. The sanctuary of Po-nagar, the Lady of the City, as the Umā-Bhagavatī of this shrine is called in the inscriptions, survives in part to this day, though not in its original form, and it has recently been renovated by the Ecole Française d'Extrême-Orient, Hanoi. The original structure of wood was burnt down in A.D. 774 by pirates from the south who raided the country in some strength, and we have no means of knowing what it looked like. The subsequent history of the temple can be gathered from its numerous inscriptions, and it is worth noting that this temple and that of My-son are the most important Cham monuments containing a large number of the most valuable inscriptions of Ancient Campā. The present image of the goddess was installed in A.D. 965 by Jaya-Indravarman; but even this image has undergone a remodelling as the head is clearly of ugly Annamese workmanship that ill suits the rest of the image. The worship of Bhagavatī is a cult very popular on the west coast of South India. [116]

One of the Po-nagar inscriptions (A.D. 918) mentions the Kāsikā, the well-known work on Sanskrit Grammar, and Jinendra's commentary on it, the Nyāsa or Kāsikāvi-varana-pañjikā. This attests not only the cultivation of grammatical studies in Sanskrit but also contact with Bengal, the home of Jinendra. [117] “The Chams,” says Finot, “have to this day the custom, in certain festivals, of smearing on the face of the deities a thin layer of paste.” This is without doubt a paste of sandal mixed with scents, and the practice is well known in South India. The paste was apparently prepared with the aid of a quern-stone, and it is interesting to note the existence of an inscribed stone of this character bearing two letters pu vya, a Cham expression meaning ‘Her Majesty the Queen’; doubtless, the stone was a gift from the queen to some temple. [118] These stones are usually called rasang batau (Cham); Cœdès, however, considers pesanī a better name for this utensil, and he draws attention to the domestic and ritual uses to which it is put elsewhere¾for grinding spices for curry and in some domestic ceremonies in South India, and for mixing medicines in Siam. He also rightly points out that sandal paste is got by rubbing sandalwood against a round stone of a particular type. But there is nothing in all this to prevent our supposing that in the temples of Campā sandal paste used to be made in the past as evidently it is being made today with pesanī.

Let us conclude this part of our study with a reference to a curious survival from ancient times in modern Indo-China. The Muslim priests of Cham are called ācār (Sanskrit ācārya), and among them the head of the community, priests and laity taken together, is known as gru (Sanskrit guru). [119] The Brahminist Chams who still survive call their priests baseh. The derivation of this last word is not quite clear; it has been taken to come from upāsaka or upajjhāya. It has been treated as a shortened form of Pāsupata on the assumption that the baseh are modern representatives of the pa-sseu or pa-sseu-wei mentioned by the Chinese traveller Tchou-ta-kouan in the thirteenth century. Cœdès has argued that the pa-sseu-wei were tapasvins and cited a Kambujan inscription of the eleventh century which mentions ‘the holy assembly of the tapasvins of the Sivasthāna’ (vrah sabhā tapasvi sivasthāna). [120] This is perhaps the best view of the question. Durand, who has an intimate knowledge of modern Chams, says decisively that the baseh cannot be upāsakas, for as he rightly observes, the upāsaka and the bhiksu go together in Buddhism, the bhiksu holds out his alms bowl, and the upāsaka fills it. The baseh are the priests of Brahminical Chams. Durand gives details of the hierarchy in which they are organized and of the functions performed by the higher grade of these priests at royal coronations. [121]

Tchou-ta-kouan in his tract on the customs of Kambuja records this interesting fact about the religious texts of that country in the thirteenth century: “The texts they recite are very numerous. All are on palm leaves put together very regularly. On these leaves they write black letters, but as they do not employ either brush or ink, I do not know with what they write.” [122] Of course they wrote with an iron style as they did in South India till recently, or possibly with styles made of hardwood like those which are in use even now among the Chams. After writing, the palm leaves were treated with some black substance to make the letters stand out for easy reading.

VII. Siam

Palm-leaf scriptures of a relatively modern though uncertain date have preserved traditions of Indian migration into Siamese territory, and these have been summarized by a modern writer in the following terms: [123] “In the year 687 of the Maha or Great Era (A.D. 765), great political disturbances took place all over India, and the inhabitants finding it impossible to make a living, were forced in large numbers to leave their home and country and settle amongst other nations.... At that time four tribes of Brahmins, consisting of a considerable number of persons, made their way eastwards from ‘Wanilara’ to Burma, Pegu (then independent), the Laos States, Siam and Cambodja. Those coming to Siam went partly to the northwest and settled in Sukotairajatani and Lawo (the present Lopburi), others went from Pegu to Tanawassi (Tennassarim) and across to Pechaburi, and still others came to Lakhon (at that time called Sai Pet or Kai Pet) where they built a temple and erected their Sao ching cha or posts for the swinging ceremony. These pillars still exist in the town as a proof that the Brahmins came to Lakhon before they reached Bangkok.” The absence of the descendants of Brahmins in Lakhon today is explained by the transfer to Ayuthya of the inhabitants of Lakhon who were vanquished in war by a Siamese emperor in A.D. 1769. This is not history, but the shape taken in men’s minds by genuine historical events attested by the more trustworthy evidence of archaeology.

Among the earliest relics of Hindu culture found in Siam are the objects found in 1927 [124] in the village of P'ong Tuk on the right bank of the Meklong or Kanburi river in the province of Ratburi, near the point “where the railway from Bangkok turns south for the Peninsula and Penang” (Le May). These finds include a Graeco-Roman lamp of definitely western Mediterranean make of the first or second century A.D. This gives a rough date for the finds and reminds us of the so-called embassy in A.D. 166 from An-tun (Antonine) to China mentioned in the history of the Han dynasty. The Chinese annals also mention that in A.D. 120 a company of musicians and acrobats from Ta-Tsin (i.e. the eastern provinces of the Roman Empire) reached China by sea. This date is also confirmed by a fine bronze statuette of Buddha, clearly of the Amarāvatī school, not later than the second century A.D. There were also several votive tablets [125] and bronzes and Buddha images, in the later style of Dvāravatī showing traces of Gupta influences and not later than the sixth century A.D.

Excavations have brought to light the foundations of two small buildings—one being the basement of a circular stūpa 9 metres in diameter, the other of a square building 6 metres each way, which contained the pedestal of a fairly large statue and whose walls must have been almost entirely faced with stucco decoration as may be judged from the fragments recovered in the course of the work. These structures recall some early Buddhist buildings at Anurādhapura in Ceylon, also known to have been subject to strong influences from Amarāvati.

The Graeco-Roman lamp throws light on the fascinating problems of the relations between Farther India and the Roman Empire. But the possibility remains that it is an Indian copy rather than an original brought from the Mediterranean; the Arnarāvatī style of the image of Buddha and the mention of a similar lamp in an early inscription from Allūru in the Krsnā valley give support to this possibility. The find-spot of these articles is not so removed from the highways of commerce as might appear at first sight; for traders are known to have avoided the strait of Malacca and chosen one of the land-routes across the Peninsula or the isthmus of Kra, while a more northern land-route from Burma passed along the Kanburi river “exactly in the vicinity of P’ong Tuk, if not by P’ong Tuk itself”.

It is clear then that this site was the home of an early South Indian colony in the first and second centuries A.D. This colony flourished up to the sixth century A.D. and came under North Indian influences of Gupta origin. It is probable that in this later phase it was part of the Buddhist kingdom of Tu-ho-lo-po-to (Dvāravati) mentioned by Huien Tsang between Srīksetra (Burma) and Īsānapura (Cambodia). Further excavations on the site carried out eight years later [126] have generally confirmed these results. Le May reproduces several other early bronzes of the Buddha from Siam of unknown provenance; it is quite possible, though this is by no means established, that some of them belonged to “well-defined Indian or Sinhalese schools”. [127]

It is a pity that of this very early period of Hindu colonization in Siam, the Amarāvatī period as it has been called, no inscription seems to have survived; at least none has been discovered so far.

Of the next succeeding period we have remains, still scanty, but sufficiently varied-structures, statues and inscriptions—which enable us to see that the people of Central Siam were of the Mon-Khmer race, practising the Buddhist religion of the Hīnayāna and maintaining a live contact with the Indian sources of its cultural advancement. These remains fall into two groups—Indo-Kambujan, and Indian non-Kambujan.

We shall consider the purely Indian relics first as they are doubtless the earlier ones. In the Pachim (Sanskrit Pascima = West) valley, a large enclosure called Muang Phra Rot (the City of the Sacred Car) is formed of a spear-shaped embankment of earth for a length of about 2.5 kilometres surrounded by a wide moat; there are ruins of laterite and brick structures scattered outside the enclosure on all sides, but they are not enough to lead to any reasoned conclusions on their origin and nature. At the centre of the enclosure, however, in the Vat Na Prasat and its adjuncts, two small fragments of a statue have been found; one contains a socket between two heels, and the other a part of the crown of the heads of a nā-ga, doubtless shading a Buddha; the workmanship, particularly of the legs, is very fine and totally different from the ugly and hideous treatment of legs common to Kambujan statues. [128]

Another Muang Phra Rot in Dong Srī Mahābodhi about 60 kilometres to the north seems to have been connected with its namesake by a causeway of which some traces still survive. This is also a large trapezoid enclosure behind a moat of about 5 kilometres long, and the material employed is laterite. The only traces of construction found inside are more or less heavy wheels of laterite which might have served as bases for wooden pillars in some kind of a light structure. Outside the enclosure is a rectangular trench cut into the laterite bank which bears on its walls a series of animal figures in relief, elephants, makaras, lions and so on, of a design so correct and an execution of which the excellence is limited only by the hardness of the material employed. Such figures are not found anywhere else in Indo-China. [129] A linga of shape unlike those of Kambuja and snānadroni found in the neighbourhood attest the Saiva cult once practised there. A very finely modelled statue of which the head with a tall cylindrical cap and thoroughly Indian facial features was discovered by Lajonquière helps to distinguish the school of art to which the ruins belonged and determine their age. [130]

Within a few kilometres to the north-west is the Dong Lakhon, a smaller square enclosure of 500 metres, behind an earth embankment 4 or 5 metres high and a moat of a width of 40 metres. Its entrances are pierced in the embankment near the angles and not at the centre of the sides, as usual in Kambuja. There are no structures, statues or inscriptions here; but from the neighbouring temples have been got a fine stone Buddha head like that of the Muang Phra Rot in which the upathisat is replaced by a tonsure adorned with a cakra cut in hollow, and grinding stones with rollers (rasang batau) used for grinding colours like the curry stones still in use in South Indian homes. [131]

Phra Pathom is today an important railway station about 30 miles due west of Bangkok. The temple, reached by an avenue of trees from the railway station, comprises a vast circular stūpa with four vihāras round it and a terraced platform. [132] The temple has been remodelled quite often and all styles of construction are found mixed up in a confused manner. But possibly the stūpa retains its original shape, and legend assigns a hoary antiquity to the temple. One tradition takes it back to the time of Sona and Uttara, Asoka's missionaries to Suvarnabhūmi, while another ascribes the foundation of the stūpa to an ancient king of Ratburi who expiated an unconscious patricide by this pious foundation. [133] With the exception of a portable linga on one of the terraces, everything about the place is of Buddhist origin, and there is no reason to think, as Fournereau did, that this was originally a Brahminical shrine. A large number of sculptures found on the site in the course of successive remodellings and deposited in one of the local vihāras, attest the antiquity of the city. These include (a) statues of the Buddha, entire or fragmentary, representing him as standing or seated on a throne in European fashion, a manner unknown to current Siamese statuary; (b) a stele representing the Buddha standing between two women with high chignons and holding fly-whisks, a group found also in certain votive sculptures in the caves of the Malay Peninsula; (c) a sculptured panel showing the Buddha seated on a throne being fanned by two apsarases, one on either side, preaching (right hand raised in cinmudrā) to a group of ten disciples seated on the floor, five on either side, one set comprising the indigenous people and the other, the disposition of whose dress recalls the sculptures in the Madras museum, evidently representing Indians; (d) fragments of ornamental panels, little resembling those of Kambujan art; (e) very realistic statues of couchant deer with necks turned back; (f) large stone wheels supported on stands modelled like those in many panels in the archaeological section of the Madras museum; [134] (g) some terracotta debris including a vase without its base and carrying an apparently Buddhistic inscription. Some Brahminical relics like a linga, pedestals like snānadronis, and a grinding stone which may belong to a neighbouring temple or even a Brahminical temple in the same enclosure. None of the sculptures has any Kambujan trait and they are generally of the same type as those of the Pachim province, and such inscriptions as have been got contain Pāli Buddhist texts inscribed in characters more or less the same as in the other inscriptions noticed before. [135] They are all attributable to the Dvāravatī period of the seventh century A.D. and earlier; [136] the wheels and the deer, which are probably earlier than the sixth century, as also the preaching Buddha, show that the shrine commemorated the First Sermon in the Deer Park of Benares.

At Suphan about 80 kilometres north of Phra Pathom there is a colossal Buddha seated à la mode Européenne and two rudely sketched sculptures of four-armed Brahminic deities, a male and a female. The rock-cut Buddhas of the Phu Khao Ngu cave at Ratburi on the Meklong, particularly the emaciated ascetic form, recall early Indian types.

In the valley of the Nam Sak, a mountain torrent in the midst of a highly wooded country, are two sites Si That or rather Srideb and Sap Xamphra; a fragmentary stone statue from the former, now in the museum of Ayuthia, gives the clue to the origin of these monuments. [137] In the fine form and studied modelling of the body, as also in the shape of the head-gear, cylindrical at the base and octagonal at the upper end, the statue clearly stands apart from the usual run of Kambujan sculptures. Again, unlike in such statues, this figure has no ornaments whatsoever in the ears, on the neck or on the waist; arms and feet are missing, but presumably they were also unadorned. We may suppose that in spite of their eccentric situation, these monuments were Hindu in origin, and later remodelled by the Kambujans who preserved the older divinities in the new sanctuaries.

Srideb has come to be recognized as a site of very high importance in recent years. [138] Lajonquière's surmise on the importance of this site has been confirmed by the discovery of many early statues and the fragment of a very early inscription. The inscription comprises fragments of six lines and in its present condition yields no continuous sense. [139] But the language is Sanskrit; and the stone on which the inscription is engraved was once taken to be a linga, and is said to be explained by the Siamese as being a foundation stone (a lak mu’ang, une pierre de fondation de ville); but I think that the inscribed stone was chiselled into its present shape at a later time by someone who had no regard for the inscription and that we should not assume that the stone always had its present shape, much less seek to determine its use from that shape. But there can be no doubt whatever that the very clear lettering that has survived goes back to the sixth or even the ninth century A.D., and that it is decidedly an alphabet of South Indian variety.

The sculptures comprise, among others, the magnificent torso of a Yaksinī, two very fine statues in tribhanga with cylindrical caps (very similar to the statue that Lajonquière saw in the museum at Ayuthia), and a fragment of a nandi; there is also a dvārapāla of clearly Khmer origin, which, while differing altogether from the other pieces, evidences the later Khmer occupation of this area. The other statues, unfortunately not easy to identify because the arms are broken and therefore the symbols lost, are thoroughly Indian in inspiration; the head-dress, the features, the massive neck, the treatment of the legs and clothing, and the tribhanga must be noted. The torso of the Yaksinī is a masterpiece of technical perfection. Cœdès has mentioned Gupta art and its canon as nearest allied to this; but the epigraphy of Srideb points to South India, and I am inclined to place the art of Srideb as a transition from Amarāvatī to the later forms of Pallava art of the time of Mahendravarman and his successors. We know that in epigraphy the colonies supply transitional forms of the South Indian alphabet not so well represented in the home country; [140] something similar in monumental art need therefore cause no surprise.

The results obtained by Mr. Quaritch Wales in a recent expedition to grideb in 1935 confirm these conclusions. He has identified a Vaisnava shrine with a ruined brick tower 40 feet high on a laterite base 20 feet from ground level; the inner vault of the tower is constructed by means of successive encorbelments, and a fragment of an inscription on the bell capital of a pillar is of the same alphabet as the other fragmentary inscription noted before. The statues, a headless four-armed figure, and “a large and very noble head,” are also of the fifth or early sixth century A.D. [141]

The art of Srideb has rightly been described [142] as “the most ancient link now known in the history of art in the entire peninsula” of Indo-China. It is remarkable that this superb school of art flourished so early and in so inaccessible a valley so far from the sea. “Such was the force of expansion,” says Cœdès, [143] “of Indian civilization that it did not merely touch the coast, as we may be tempted to believe, but penetrated to the centre of the peninsula with its language of learning, its writing, its religion and its art.”

What was the political position of Srideb? Was it a vassal of Fu-nan, and did Fu-nan extend its sway so far to the north-west? Or was it an independent state? We have no means of deciding this.

To turn to the later Indo-Kambujan remains, inscriptions in Mon Khmer and Sanskrit ranging from the sixth century A.D. to the thirteenth are found; they are mostly religious in import, though it is possible that some of the buildings where they occur might have been palaces; after all, the distinction between temples and palaces was not very sharply drawn either in the thought of the people or in the structures themselves. The Brahminical cult predominates in the art of this group, and images of Brahmā, Indra, Visnu, and Siva adorn the entrance of almost every temple; Buddha images are not unknown, but usually only as an avatār of Visnu. There is, however, one exception; in the sanctuary of Phimai, the image of Gautama takes the place of honour on the lintel of the principal entrance, the Brahminical gods being relegated to the subsidiary entrances. But this temple is unfinished, and belongs to the last days of Khmer prosperity. In any event, this temple in honour of the Buddha reconstructed in the midst of an old Visnu temple is clear proof of the rising importance of Buddhism at the cost of Hinduism. [144]

In the beginning, the Hindu colonies doubtless arose in favoured spots as more or less independent units of moderate size; more powerful kingdoms must have been formed later in the course of several generations, by alliances, wars of conquest and so on. Many names of the states together with descriptions of their people, government, manners, trade and so on are found in the Chinese accounts relating to the period; but the names are not easy to identify, and the descriptions vague. We see enough to recognize that we have before us a picture of several states, all of them Hinduised, and in more or less active communication with one another, and with China on the one hand and India on the other. But on the restoration of the original names of states from their Chinese forms, and on their geographical location, wide differences of opinion are still unfortunately possible and prevalent.

Traces of an ancient Hindu settlement occur at a spot three miles to the east of modern Chantabun (Candanapuri) on the banks of a navigable river of the same name and commanding fertile country all round, rich in rice, pepper and other products as also in precious stones like rubies and sapphires. The soil is red in colour, a matter of interest in the location of the country of Tche-t'ou (red earth) of the Chinese geographers. Acadra of Ptolemy's maps may very well be located here. [145] Fragments of inscriptions in Sanskrit and Khmer have been found, Sanskrit being used generally for praising gods or the founders of religious edifices, Khmer being employed for edicts or other records meant to be understood by the common people. One of them, [146] of about the end of the tenth century A.D., is a royal order in Sanskrit and Khmer communicated by the king's guru to the civil or religious dignitaries designated Vāp, Steñ, and Ācārya. Another fragmentary inscription noticed by Lajonquière is part of the digraphic inscriptions of Yasovarman (end of the ninth century) set up in different parts of his kingdom. The provenance of these inscriptions is uncertain; Lajonquière rejects the slopes of Mount Sabab suggested by Schmidt as there are no traces of ancient monuments there, and thinks that the hamlet of Phamniep, near the village of Bau Narai, 3 kilometres east of Chantabun, about half way to Mount Sabab, their more likely source. To this group belongs a sculptured slab of red stone, 80 cm x 60 cm, that must have formed part of a decorative lintel; the sculpture shows part of the façade of a palace with five women seated in front and clad in short sampots with vertical stripes, and wearing conical head-dress and ornaments. Another slab of red stone, meant doubtless for a spandrel, exhibits an unfinished sculpture which is a replica of another found at Phamniep very near Chantabun and described below.

The ruins of Phamniep have long been used as a quarry for extracting building material, and not much is now left of them; but there is the village with a Hindu name, Bau Phra Narai—the village of Visnu; there are the foundations of a double structure, possibly a palace, standing in the open and not enclosed together with other buildings by a surrounding wall, as obtains in Cambodia in similar instances; and above all, there has been found in the pepper garden of the village a finely sculptured slab which must have been the lower half of the tympanum over the doorway at the entrance of a temple; it shows Garuda carrying Visnu on his shoulders, flanked by two open-mouthed makaras with short trunks; only the upper part of the Garuda's body is seen, the lower half being hidden behind ornamental motifs, and of Visnu only the legs carried on the shoulders of Garuda and held by his hands are seen; the bust of Visnu and whatever else was sculptured on the upper slab are lost. The sculpture from Chantabun noted at the end of the last paragraph is an unfinished copy of this.

The date of these relics is a matter of conjecture; they resemble Kambujan art, but there are also striking differences; they must have had the same source of inspiration as Kambujan art; makaras similar to those we have noted here appear in the earliest phase of Kambujan art. We seem to have here the relics of an original Hindu colony established in the valley of the Chantabun at some indeterminate, but early epoch; towards the ninth century it became part of the Kambujan kingdom, as the inscriptions testify.

Lobpuri (Louvo) in the valley of the Menam, was a centre of Mon-Hindu culture from very early times. Later it became part of the Kambujan kingdom and the seat of a Khmer viceroy for Central Siam from the beginning of the reign of Sūryavarman I (1002). Its monuments, sculptures and inscriptions, particularly the earliest among them, are of great interest to us. The city is located on a flat plain liable to inundation in rainy weather, but it commands the more salubrious highland adjacent to it on the eastern side.

Several statues of the standing Buddha were discovered in 1924 in Vat Mahādhātu and its environs; statues in bluish limestone have nothing in common with Khmer statuary and evidently belong to an earlier art; one of them bears a Sanskrit inscription in characters similar to those of the most ancient epigraphs of Kambuja, [147] and clearly of a South Indian variety of the sixth or seventh century. The inscription reads:

tangurjanādhipatinā sāmbūkesvarasūnunā nāyakenārjjaveneyam sthāpitā pratimā muneh.

I. e. the Nāyaka Ārjava, chief of the people of Tangūr and son of the king of Sāmbuka, has set up this image of the Muni. The two regions named cannot now be identified; but the glimpse afforded by this correct Sanskrit record into the political organization of the land in this early period and the purely Indian designation of the offices mentioned are noteworthy. Another of the Buddha statues bears a single line in equally ancient characters, and probably in the Mon language. [148] Vat Mahādhātu, as it is at present, has a strong affinity to the architecture of Angkor; but these early Buddhas of the Mon-Hindu period are a clear indication that this Khmer temple replaced an earlier temple in another style.

Four singularly archaic Mon inscriptions are found engraved on an octagonal stone pillar with an ornamental cubical capital; the pillar which comes from the neighbourhood of San Sung [149] is identical with some others found in the gallery surrounding the great stūpa of Brah Pathamacetiya to be mentioned presently. “Without doubt,” says Halliday, [150] “this inscription of Lobpuri is the most ancient Mon text deciphered and published till now.” The inscription contains some Sanskrit and Pāli words, and records gifts of slaves, betel, carts, and a flag to a Buddhist temple by different persons whose names are given. Some of these names are indigenous like Cāp Sumun; others are Indian like Prajñavanta, Sīlapāla, Sīlakumāra. The characters of the record are those of a South Indian alphabet of the sixth or seventh century A.D. The importance of the Mon element in the population of the valley of the Menam and in the colonization up to Haripuñjaya is being revealed for the first time by these new and still rather obscure inscriptions.

The triple shrine of Phra Prāng Sam Yot [151] (the temple with three Sikharas) was doubtless at first a Hindu structure turned later to Buddhist uses. The central shrine is slightly larger than those on the sides; all face east and are connected with one another by covered passages along the north-south axis; they are built of limonite, stone being used for doorways, pediments and so on. “The design is certainly not Buddhist,” says Le May, “and the three towers ranged alongside one another invariably bring to the mind the Hindu Trinity of Brahmā, Siva and Visnu. Non-Buddhist figures, too, have been found on the towers¾bearded figures with their hands resting on clubs¾which also points to an originally Brahman construction.” Two other features are noted by Lajonquière as marking the age and origin of this temple; the shape of its openings, the windows and passages terminating in an ogive, is unknown to purely Kambujan monuments; again, the decorative sculptures are barely sketched and the details are picked in stucco, a procedure extremely rare in Khmer art, though not altogether unknown, Some vestiges of ancient snānadronis attest the original character of the shrine. The Phra Prāng Khek, [152] also a triple shrine without the connecting passages and with the lateral shrines definitely smaller than the central, and the San Sung, the Vaisnava shrine, [153] in the neighbourhood of which the pillar with the Mon inscription was found, are other early monuments also worthy of note.

Some temples in Saxenalai-Suk’otai in Central Siam, particularly the Vat Pr'a Pai Luang and the Vat Sisawi in old Suk'otai, seem to have been originally built for Brahminical worship and later adapted to Buddhist uses in the Tai period; [154] this is clear from the plan of the structures as also from the survival of the older decorative sculptures on their walls.

The lower valley of the Mekong and the valley of the small stream Pechaburi offered the most favourable conditions for the establishment of colonies; accordingly we find relics of a number of old states with Ratburi on the Mekong at their centre, Muang Sing farthest inland, Kanburi, Phra Pathom between the Mekong and the western arm of deltaic Menam, and Pechaburi more to the south nearer the sea. It is possible that at one time these centres were united under a single state; but we know nothing certain of their actual history. [155]

The Vat Kampheng Leng at Pechaburi is another temple, Brahminical in origin, as its plan and the surviving images of dvārapālas and of Visnu on Garuda, testify, but turned later to Buddhist purposes. [156]

These monuments of the Indo-Kambujan group, as Lajonquière has called them, show that the Kambujan kingdom extended its sway to the west into Siam and tried to spread its own form of religion and architecture among the conquered states; but in these outposts of Kambujan culture enough remains yet to reveal their original condition and to show that the early colonists from India had selected most of these sites as favourable for their settlement and occupation.

Siam, like Cambodia, maintained a number of court Brahmins at Bangkok until recently when as the result of a revolution a republic came to be established. Joseph Dahlmann who travelled in Siam in the twenties of this century gives the following account of them: [157] There are about 80 families. Their dwellings are erected round a poorly temple comprising three insignificant structures enclosed by a wall. The Brahmans differ from the Bonzes by the long flowing hair on their heads. The white ceremonial gown and the conical cap vividly bring to our minds the Brahmans of the island of Bali. Small as is their number by the side of the thousands of Buddhist Bonzes, they have still many privileges conceded to them, as, in spite of all the changes due to Buddhism, the memory of the old Brahmanical royalty is still so deeply rooted in Siamese tradition. To the Brahman community is reserved the consecration of the new king, and royalty is held to be properly transmitted to the new ruler only by the completion of such consecration. Simply and solely for this end is this small group of Brahmans preserved in the midst of the large community of Buddhist Bonzes. At their head stands a guru bearing the proud title Mahārājaguru. With the consecration of the king goes the consecration of the royal elephant, also reserved to the Mahārājaguru; for what is the Siamese king without his white elephant?

There is a published official account in English of the details of ceremonies and mantras employed on the occasion of the coronation of His Majesty King Prajādhipok in February of B.E. 2468 (A.D. 1926). We have only to note that unlike the Brahmins of Cambodia the Siamese Brahmins are not relics of a once powerful religious caste, as Father Dahlmann seems to think, but appear to have been brought in at a later time from Ligor and elsewhere to conduct the court ceremonies, in imitation of other courts, with an Indian ceremonial. The Thai conquerors of Siam sought thus to legitimatize their rule in the eyes of the people by observing the same forms as the ancient Khmer monarchy of the land. In 1821, one of the Brahmins told Crawford that he was fifth in descent from his ancestor who first settled in Siam and had originally been an inhabitant of Rāmesvaram, the sacred island adjacent to South India on the east, to the north of Ceylon. Quaritch Wales [158] says that some Brahmins today have a tradition that their ancestors came from Benares, that both these accounts may be true and that there may be now in Bangkok descendants of Brahmins from both North and South India. These traditions are surely evidence of late immigrations; but the modern Bakus of Cambodia have no such tradition, and the head priest at Phnom-Penh is said to have claimed very recently and quite seriously that his ancestor came from Mount Kailāsa!

VIII.  Malay Peninsula

In the Malay Peninsula the early colonists from India founded a number of independent states. Our knowledge of these states is still very limited and we have to depend on Chinese notices which are not always easy to interpret. About the eighth century A.D. these states began to attract the attention of their more powerful island neighbours in the south; almost to the end of the thirteenth century the whole region may be said to have been under the political tutelage of the Sailendras, who at first made their appearance in Central Java and later became masters of the maritime empire of Srī Vijaya. Ruins of the Javanese and Sumatran periods of Malayan history are found scattered throughout the peninsula. When Srī Vijaya fell from power, the Malayan states fell an easy prey to the Siamese, though the southern states passed under the Javanese empire of Majapahit for a time.

          Malacca must have been an early colonial centre. Lajonquière has drawn attention to a makara fragment built into a retaining wall near the ancient Portuguese church containing the corporal remains of St. Francis Xavier; doubtless this came from an ancient temple destroyed by the Christian conquerors. [159]

The village of Kuala Selinsing on the coast of the Matang district of Perak has been identified by Mr. I. H. N. Evans as an ancient Hindu settlement on the strength of a cornelian seal bearing the incorrect Sanskrit inscription Srī Visnuvarmmasya in box-headed characters of a South Indian variety of about the sixth century A.D. or earlier. The level at which the seal has been found justifies this date also. [160] There are also beads of shell and opaque glass, besides a gold ring bearing a group identified by Mr. Evans with reservations as Visnu borne on the shoulders of Garuda. There is nothing improbable in this, and though I share to some extent the doubts regarding the Hindu character of the ring and the figure on it, I think that Mr. Quaritch Wales carries his scepticism too far in doubting the presence of Hindus in the settlement. [161] It is true, however, that no definitely Hindu cult object has so far been found on the spot.

In ancient Kedah we have an important and unmistakably Hindu settlement which has been known for about a century now from the discoveries reported by Col. Low and has recently been subjected to a fairly exhaustive investigation by Dr. Quaritch Wales. [162] Among Col. Low's discoveries was an inscribed slate slab found near Bukit Meriam in a ruined brick house 12 feet square, possibly the hut of a Buddhist monk, as Kern was inclined to think. The inscription comprises two stanzas—the yedharmmā formula and the verse

ajñānāccīyate karma janmanah karma kāranam

jñānānna kriyate karma karmmābhavānna jāyate

which means: Karma is accumulated through Ajñāna; Karma, is the cause of birth. Jñāna leads to desistance from Karma, and in the absence of Karma there is no birth. [163] We have no means now of judging the age of the record from its palaeography, as the original is lost and there is no mechanical copy.

But it seems hardly likely that this inscription differed much in age from others from Kedah found by Dr. Quaritch Wales, to be noted presently, and from the other discovery of Col. Low, viz. the inscription of Mahānāvika Buddhagupta from the northern district of Province Wellesley. This record, also on a slate slab, is engraved on both sides of a stūpa with a chattrāvati (umbrella series) of seven members, and is in characters very similar to those of Pūrnavarman's inscriptions in Java of the early fifth century A.D. Besides the verse ajñānāt, etc., the inscription contains a short prose passage of benediction wishing success in all ways and everywhere to the enterprises undertaken by the Mahānāvika, the sea captain Buddhagupta, resident of Raktamrttikā. It contains the interesting word siddhayātrā, which is found in some other early Indonesian inscriptions also, where it is seen to refer to a pilgrimage to a holy place for the attainment of spiritual merit or potency leading to success. [164]

To return to the antiquities of ancient Kedah, Dr. Wales investigated no fewer than thirty sites round about Kedah. The results attained show that this site was in continuous occupation by people who came under strong South Indian influences, Buddhist and Hindu, for several centuries. We need mention here only some of the most conclusive and significant links in the chain of evidence brought to light by these valuable investigations, leaving the other details to be gathered from Dr. Wales' work by the interested reader. [165] An inscribed stone bar, rectangular in shape, bears the ye-dharmmā formula in South,Indian characters of the fourth century A.D., thus proclaiming the Buddhist character of the shrine near the find-spot (site I) of which only the basement survives. This inscription naturally recalls the Bukit Meriam (site 26) inscription of the same formula noted above. A more interesting find from site 2 brings it into line with the colonies in Lower Burma; it is a sun-dried clay tablet measuring 5^1/8” x 1^1/8” x 1^1/8” in the centre and slightly tapering towards either end; it is inscribed on three faces in Pallava grantha of the sixth century A.D., possibly earlier; each face carries two lines making a complete sloka. The three Sanskrit verses embodying Mahāyānist philosophical doctrines have been traced together in a Chinese translation of the Sāgaramati-pariprcchā, the original of which, is not forthcoming; two of these three verses occur also in a number of translations of other works, all of the Mādhyamika school. This inscription which, as Dr. Wales rightly points out, precedes the earliest Mahāyānist inscription from Sumatra (Talang Tuwo A.D. 674) by about a century, brings Kedah into the same class as Prome in the same period where also some Sanskrit Buddhist texts have been found in the midst of several from the Pāli canon.

On a low spur of the Kedah peak to the south are traces of a Siva temple (site 8); its plinth and lower courses built of small granite blocks have survived, as also a fragment of a bronze trident and two curious nine-chambered reliquaries of a type unknown in India, but common in more elaborate forms in Java in the ninth and tenth centuries; this temple may be considered, for several reasons, to be an important link in the transition from the sepulchral shrines of South India with lingas in them to the developed Candis (tomb-shrines) of Java enshrining the portrait figures of particular monarchs. From site 10 have been recovered foundation deposits of a type unknown so far in India or Java; they comprise one gold and six silver discs, each 1 ½” in diameter, inscribed on one side in South Indian characters of a cursive type which may be assigned to the ninth century A.D. The inscriptions are generally either names of Bodhisattvas, whose images were perhaps set up in the shrine or possibly of devotees who took part in the consecration, though in one case there is only one syllable, Om. Among the foundation deposits of a Buddhist shrine in site 14 were two silver coins of the Abbasid Caliphate, one of them bearing a clear date 234 A.H. (A.D. 848). A large Siva temple has been identified as such (on site 19) by a four-armed Ganesa figure in terracotta and a bronze Sakti weapon of Kārttikeya, and the temple is assigned to the eleventh or twelfth century.

Kedah was identified by Cœdès with Kadāram of the Cōla literature and inscriptions and Katāha of Sanskrit literature; this has however been questioned subsequently by other writers, I think on insufficient grounds, and the explorations of Dr. Wales seem to me to go far to confirm the identification made by Cœdès on other grounds.

IX. Takua-Pa and Other Places

Takua-pa [166] at the mouth of the river of that name was identified by Gerini with the Takola of Ptolemy and of the Milinda Panha. Lajonquière's investigations brought to light a number of antique sculptures and monuments which taken along with the Tamil inscription discovered earlier (in 1902) by Mr. Bourke, a mining engineer of the Siamese Government, makes it quite certain that Takua-pa was in the early centuries of the Christian era a well-known harbour and trading centre often resorted to by ships coasting along the Golden Chersonese.

The hinterland is rich in tin-mines and there are old mining shafts here which are clearly distinguishable from those sunk by the Chinese and Europeans in later times. There are also the ruins of an old brick structure in the isle of Thung Tu’k (‘the plain of the monument’), but they are not enough to warrant any inference being drawn from them. Not far from Thung Tu'k is a small conical hill known as Phra Noe in a small island lost in a labyrinth of canals winding through a forest of mangrove trees; on this hill was found a statue, broken but with all the parts in situ, of a four-armed figure with a cylindrical tiara. [167] The figure is of natural size and presents one of the finest examples of artistic modelling; the cylindrical head-dress and the long sarong, together with the total absence of ornaments, place the statue in the same class as that of Muang Phra Rot in the Pachim valley in Siam.

The top of the hill has been cleared and levelled, but bears no signs of any construction. Lajonquière considers the statue to be one of Siva; but there is no means of sure identification.

Farther in the interior is the hill of Phra Narai at the confluence of the Khlong Pong and the Khlong Ko Srok which unite lower down with the Khlong Phra Va to form the river Takua-pa. “This is a small hill conical in shape, 40 metres high and covered by forests. On the summit we find only debris of bricks and two large flat unwrought stones. These few vestiges appear to be all that remains of a small square sanctuary which measured three metres from side to side in the interior and opened to the east.

“On the opposite side of the Khlong Ko Srok, an ornate stele of three Brahmanical figures which came from this small shrine is deposited on the bank. Local tradition says that it had been brought from there by the Burmans during their last invasion and deposited at the spot where it now is, to await its transport to Burma. But when they were about to proceed with this operation there fell such violent rains, that the spoliators were constrained to abandon it. It was since broken by the wild elephants which were for long the masters of this region devastated by wars. When the inhabitants returned, they found it in the present state and supported the debris against a tree. This tree with a trunk divided in two parts and now 20 metres high, has framed the debris in the growth of its trunks, and the folds of the bark cover the figures in part.

“It appears to have been cut in a large slab of schistose limestone, on which three figures come out in reliefs exceeding, at certain points, 90 centimetres in thickness.

“The most important, at the centre, represents Siva standing. The head (of which the cover has been removed), disappears under a fold of the bark; the feet are broken; they form one piece with a small plinth and tenon lying on the side of the tree. The bust is nude; collars made of gold adorn the neck; above (these) hangs a necklace of pearls increasing (in size) as it descends; a girdle of rectangular plaques of gold goes round the chest above the breasts; a thick ribbon woven of many rows of pearls is attached by a wrought buckle on the left shoulder and falls on the right hip; a girdle with a large wrought buckle holds round the edge of a long sarong with many folds; along the thighs fall the folds of embroidered cloth and cordons from which hang button-like ornaments. The god is represented with four arms; we see only the right front and left front arms; the right posterior arm is only indicated by the lines of its fracture. The fore-arm of the right front arm is raised, the wrist is adorned by three bracelets; the hand, open and raised, is adorned with rings for the little and ring fingers, the thumb appears to hold a cord, which, passing above the right shoulder, comes to attach itself to the thumb of the left front arm. The hand of this arm is supported on the hip; its wrist has three bracelets of which two are of pearls; a large wrought bracelet adorns the biceps.

“The figure which, in the stele common (to all of them), is placed to the right of the god is without doubt that of his wife Pārvati. Probably the goddess is represented as seated, but the lower part of the body is missing, or is masked by the ligneous developments of the trunk which form a natural niche round the bust. The head is dressed in the form of a high cylindrical chignon formed of tresses gathered up in front and held in position by golden ornaments; a golden crown with two large earlaps sustain this edifice of hair. The countenance is round, the eyes lightly turned up, the nose is broken, the mouth with a thick lower lip is well designed; the bust is nude, a large necklace of pearls hangs on the chest between the well-marked breasts; a cordon passed over the left shoulder falls above the right hip; the right arm encircled by pearls at the level of the biceps is lowered and covered largely by the bark; the left arm is raised, the hand supporting the head-dress; we can only see the upper part of the sarong and the knot of the belt.

“The figure to the right represents a danseuse; resting on the left knee with the left hand on the hip, she stretches towards the god her right arm which is broken; her head, inclined to the right, is dressed as a high conical chignon held in position by a crown of gold; the face is round, with eyes half-closed with very curved and slightly upturned eyebrows; the nose, the mouth with thick lips, the accentuated chin, are of a pretty design; a necklace hangs on the chest; a cordon passed over the left shoulder hangs between her two well-developed breasts; the bust is nude, the waist supple and elegant; the left arm is adorned at the biceps by a bracelet worked in a rosaceous pattern, and at the wrist by three bangles; the pelvis and the legs are very tightly draped in a long sarong which descends in multiple folds; the feet with anklets of metal are nude.

“These three figures of natural size are very superior as sculpture to what we have so far found in Indo-China.”

Lajonquière's identification of the figures as Siva, Pārvatī and a danseuse need not be accepted, for it is more likely a representation of Vishnu and his two consorts; but his careful description of the sculpture is so valuable and so forcibly brings out its South Indian inspiration in all its details that I have not hesitated to reproduce it here in extenso. In fact, the same writer observed earlier and more summarily: “The costumes, in numerous folds treated with details, the profusion of jewels, the elegant movements of the body, recall very nearly the oldest sculptures of Dravidian India.”

The inscribed stele by the side of this sculpture carries, appropriately enough, a Tamil inscription clearly of the ninth century A.D. It records the construction of a tank, named Avani-nāranam, evidently after Nandivarman III (826-850) Pallava, by a person who described himself as the Lord of Nāngūr; the tank is placed under the protection of the members of the Manigrāmam, the residents of the cantonment (Senāmukham) and one other group of which the nature is obscured by a gap in the inscription. This record is valuable and conclusive proof of the active contact maintained in the ninth century between the two shores of the Bay of Bengal. But there is much in the record that we are not in a position to explain. Was the Lord of Nāngūr a military chieftain of South India or just a merchant prince? Was he actually present in Takua-pa when the tank was dug and the record of it engraved? If so, was his mission peaceful or warlike? And who maintained a Senāmukham at Takua-pa and for what purpose? Did the troops have any connexion with and were they under the employ of the Manigrāmam (Sanskrit Vanigrāmam), the large and influentlial guild of merchants of which we hear in diverse connexions? Questions like these which leap to our minds and which we are unable to answer indicate the large gaps in our knowledge of those remote times. We have to be grateful to the scientific zeal of the explorers from Western lands whose labours, undertaken often under conditions of great discomfort, have brought to light vestiges of long forgotten chapters of the efforts and achievements of Indians in ancient times.

Jaiya [168] on the southern shore of the Bay of Bandon was a dependency of Srī Vijaya for several centuries and contains several monuments some of which at least must be taken to date from a much earlier time. The Vat Pra That is surely a construction of the Srī Vijaya period having much in common with constructions depicted on the bas-reliefs of Borobudur and following the canons of the Silpasāstras of Indian origin.

Among the numerous statues found in this neighbourhood, belonging to different periods and styles, the admirable bust of Lokesvara, discovered by Prince Damrong and now in the museum of Bangkok, [169] deserves special notice. It is one of the most magnificent bronzes of the Srī Vijaya art of the ninth century. “The benevolent serenity of the face, the noble bearing of the shoulders and the magnificence of dress and adornment,” says Cœdès, “class this statue, badly mutilated, among the masterpieces of Indian sculpture in Indo-China.”

The Vat Keu is a brick structure on a plan similar to that of Candi Kalasan of Central Java, though its architecture recalls the ‘cubic’ art of Campā. [170] In a small vihāra to the east of this ruined temple, there are some interesting sculptures including a Buddha statue clearly of the Dvāravati art. There is also a statue of Visnu [171] described by Cœdès thus: “Image of Visnu standing with a sort of decorated mitre on the head and wearing huge earrings of a peculiar style. The deity has four arms; the right back arm rests on a large mace, and the front one is raised holding the disc; on the left, the back arm is broken, and the front one rests on the hip and holds a conch.

Similar statues, wearing an identical costume, are still found in situ at Nagara Srī Dharmarāja (Ligor).” The statue is stiff and inelegant; it is a product of late art, valuable as showing the persistence of Indian influences to a late period.

On an isolated hillock, 3 to 4 kilometres to the south of Vat Keu, are relics of an old structure similar in plan and style to Vat Keu, from which a Bodhisattva head has been recovered and preserved in the museum at Bangkok. [172] This site is called Khau Nam Ron (‘hillock of warm water’) from a hot spring at its foot.

The Vat Hua Vieng comprises the debris of a large brick vihāra; from its neighbourhood comes the fragment of a statue without head, arms or feet, but notable for the modelling and the treatment of the dhoti and ornaments; the Jaiya inscription of Chandrabhānu and another on a bronze Buddha dated A.D. 1283 come from the same place. [173]

The Vat Sālā Tung is another ruined shrine containing some fragments of Brahminical sculptures, and a fine stone statue of Lokesvara, arms and feet broken. [174] The simple treatment of the body contrasts with the complicated jatāmakuta.

Some shrines in actual use today show, in spite of repeated remodellings, their original dependence on Indian silpasāstras for their design; Vat Palelai and Vat-To are noted by Claeys as good examples of this class. [175]

Jaiya was an important centre of pilgrimage for the Buddhists and numerous votive tablets in clay attest this fact. Such tablets which are also found in many caves in the mountain ranges of Malaya used as Buddhist residences, have been studied in some detail and classified according to their fabric, locality and age. Those of the Malay Peninsula are of terracotta, circular in shape, and bear representations of the Buddha or Bodhisattvas and the formula of the creed. Palaeography and fabric alike lead to the conclusion that they may be dated about the tenth century A.D. and taken to follow on similar tablets from Pra Pathom. [176] The practice, however, continued through centuries, and at Ligor metallic tablets of modern make are in use to this day.

Between Jaiya and Ligor there are a number of old sites of interest, particularly Khao Srīvijaya whence a beautiful Visnu with cylindrical head-dress and remarkable for the peculiar knot of a scarf on the right hip attests the early age of the site and the prevalence of Brahminical faith at the time.”

Nagara Srī Dharmarāja, in the middle of the peninsula within twelve miles of its east coast, was doubtless the capital of Tāmbralinga, one of the vassal states of Srī Vijaya. [177] It is a walled city in the midst of a fertile plain of rice fields well protected from the monsoon by the neighbouring mountains. [178]

In this centre of Buddhism, Brahminical antiquities are not very important or striking, but are not altogether unknown. In the interior of the town are three sanctuaries, Bot Prahm, with a number of lingas, San Pra Isuon, containing bronze images of the dancing Siva, Pārvatī and Ganesa, and Na Pra Narai, with its statue of Visnu clearly showing Indian influence like all statues of this region. The Bot Prahm was called Na Phra Narai by Lajonquière by mistake. [179] The bronze Ganesa [180] bears a Tamil inscription which reads:

Ma- Jha- pi- ci- de- sa

i.e. the prosperous country of Majhapi (Majapahit), in modern characters. In fact these temples are still in use and some of them have been recently remodelled. [181]

The Vat Phra That or Mahādhātu is considered by Lajonquière to have been the model of many similar edifices in Indo-China. It is a bell-shaped monument on a square basement; elephants are sculptured issuing from it and seem to carry on their backs the weight of the superstructure, a motif borrowed from India and copied again in many old northern sites of Siam, like Kampheng Pet, Sukhotai and Saxanalai. [182] The rest of the structure bears sculptures representing Buddhistic scenes sketched on laterite with the details picked out in stucco. The extant structure is an enlargement of an earlier and simpler building the pattern of which is presumed to be preserved in a reduced model within the precincts of the temple; the tradition of making a small-scale model of an existing structure before enveloping it in a new and larger edifice seems to have been commonly followed. [183] Of the model of Phra That this is what Claeys says: “We find here again the plan of a large monument, reduced to the dimensions of a small structure, recalling the Candi Kalasan of Central Java or the Cham towers of Dong-düöng and My-son. The base of the bell of the upper stūpa seems to rest on a lotus flower, symbolized by a cylindrical moulding adorned by petals lanceolated vertically.”

There are three inscriptions from this temple. One of them from the Vat Sema jaya comprises eight lines and is illegible. Another engraved in large letters on a step in a staircase is palaeographically of the class of South Indian script of the fifth or sixth century A.D.; the whole inscription is only one line of seven letters which I would read somewhat differently from Cœdès thus:

bha ta ma ra yya lba ge sva rah

This reading is not much easier to interpret than Cœdès’

bha ta mā yya ai ge spah or (ai gra sthāh)

though the ending of my reading may appear to be an incorrect—īsvara ending. [184] The third is a Tamil inscription dated in a Saka year in words, but the word for the hundred figure is unfortunately lost, though one may guess from the palaeography and the expression (n) ūrru (a) ñju that it was eight, giving a date in the last quarter of the ninth century A.D. This, if correct, would bring the record very near the age of the Takua-pa Tamil inscription noticed elsewhere. The record mentions some charity in favour of Brahmins instituted according to the orders of a Dharmasenāpati. [185]

Before leaving Ligor, we must notice another inscription in six long lines, much damaged, each line comprising a verse in Sanskrit. The record comes from Vat Maheyang in the province of Nagara Srī Dharmarāja; it is engraved in characters ultimately South Indian in origin but closely allied to those in the Khmer empire in the seventh to ninth century A.D. It records prescriptions relative to the internal discipline of a Buddhist monastery. [186]

It may be noted here that on this coast of the peninsula bordering on the China sea there are no traces of early Hindu settlements south of Patani; it is by no means easy to explain this, though it has been suggested sometimes that the full force of the north-east monsoon might have had something to do with it. [187]

The temple of Vieng Sra (‘the fortress of the lake’) in the upper valley of the river Bandon about 80 kilometres to the south of Jaiya contains some early relics. [188] The antiquities of the place, described in considerable detail with a plan by Lajonquière, comprise a walled enclosure with the relics of a small Buddha temple at the northeast of which a small stone model has survived (0.60 metre), an elegant square pedestal with simple horizontal mouldings symmetrically inversed and a square mortice in the centre for the reception of the image, and fragments of red stone statues including one of a pot-bellied Buddha called Mek Thong by the Siamese. But the most important find from the place is a fine statue of Visnu, with cylindrical head-dress, which was found almost in the centre of the enclosed space and which is now in the Bangkok museum. [189] Nothing is now left of the temple in which this fine image must once have been enshrined, and it seems possible that the material from it was employed in the construction of the ruined vihdra or of the more modern pagoda which is still in use.

A standing Visnu and a Vatuka-Bhairava form of Siva in stone figured by Cœdès [190] also deserve to be noticed. Every detail in these figures is decidedly South Indian except the facial features which are indigenous. These figures may be of the ninth or tenth century, in any case much later than the Visnu with cylindrical head-dress.

It is perhaps worth noting that the famous stele inscribed on both sides—one a Srī Vijaya record and the other an incomplete Sailendra inscription—which was at one time believed to come from this place has since been traced to Vat Sema Muang in Ligor. [191] Both are in South Indian script and one of them bears a Saka date corresponding to A.D. 775.

The Malay Peninsula continues to be in the debt of South India to this day and the contact between the two lands is being actively maintained along many channels, primarily economic. The results on the cultural side of these long-established contacts have struck all close observers; and Annandale [192] says: “There are many similarities between the Muhammadanism of the Labbies of the Indian shore of the Gulf of Manaar and that of the Malays, and I think it would not be impossible to find striking parallels between objects in daily use, and especially in the patterns with which these objects are adorned, among the two races.” Evans has studied the persistence of an old type of Indian water vessel, the kendi with a spout, and reproduces a Chinese porcelain kendi as an example of non-Chinese ware made in China for export to Malaya. [193] “The importance of Rāma and Hanuman in the folklore of the Malays, Buddhists and Muhammadans alike, agrees with legends which link these with the region round Adam's Bridge, the region whence came the bulk of the ‘klings’ resident in Malaya.” “I would even hazard a suggestion,” continues Annandale, “that it is largely owing to the commercial activity of the Labbies and their ancestors that the Malays of the mainland were first converted from pure Shamanism to Hinduism, and then from Hinduism to what they call, in phraseology of curiously mingled derivation, the āgama Islam.” This is a just estimate on the whole, though perhaps the emphasis on commerce and the ancestors of the Labbies may be considered a little too strong in the light of facts known to Annandale himself, and the more so in view of new facts that have come to light since this estimate was written.

We may conclude this sketch of South India's part in the making of Malayan history and culture with some living examples of the results still seen today. The Sanskrit word ‘Srī’ which begins all auspicious formulae persists today in Malay in Muslim kingdoms long after the advent of Islam, and serves as the name of an oath of allegiance in Perak as well as in Borneo. The word is found, of course, only in a much altered form as ‘chiri’; but its definitely Hindu origin, possibly from the days of Srī Vijaya, may be inferred from some Malay traditions of Perak recorded by Maxwell in 1881. [194] The Malays of Perak say that the chiri was introduced in the time of the first Malay Raja, who came down from the mountain Saguntang Maha Meru, and appeared suddenly in Palembang, in Sumatra, riding on a white bull.” Ronkel has traced several common Malay words like those for washerman, kind or sort, marriage pledge, leaf, couple, and so on, to indubitably Tamil origins.” [195] We cannot be quite certain of the age of any of these words in Malay as contact with the Tamil country has been unbroken throughout the centuries that followed the early period of colonization with which we are particularly concerned.

X.  The Southern Islands

To gain a correct idea of the extent of the influence of Hindu culture in the islands that came under it, one must contrast Sumatra, Java and Bali with the islands farther east which were not touched by this influence. It will then become clear that all the elements of higher culture, the form of organized state-life, trade and industry, art and literature were practically gifts of the Hindus to these islands, and that the archipelago falls easily into two divisions¾one which accepted the new culture and advanced with it into civilization, and the other which lagged behind. We shall naturally be concerned most with the first.

The earliest inscriptions from the islands attesting the establishment of Hindu culture belong to the end of the fourth century A.D.; but external evidence is sufficiently clear that this movement must have begun very much earlier. Strangely enough, this evidence is more Chinese and Greek than Indian for the earliest phases of the colonization. Trade at first, and later religion when the Buddhist pilgrims began to use the sea-route to and from India, stimulated the interest of the Chinese in the islands of the southern seas, and their dynastic chronicles and travel books have preserved in one way or another much that is of interest regarding the conditions prevailing in those lands from very early times. The Chinese were good observers and faithfully recorded what they saw and heard in these strange lands, though it is quite probable that they often enough derived erroneous ideas of these things. On social, economic and religious conditions, nevertheless, they tell us much that is sound, precise and authentic. Though in course of time numbers of Chinese came to live in these islands, unlike the Hindus, they always remained colonies of aliens whose presence had little or no influence on the culture of the surrounding inhabitants.

Western evidence on these lands is naturally even more vague and difficult to interpret than Chinese. Though the trade of Hellenistic and Imperial Roman times brought the Graeco-Roman world into active contact with India proper, its notions of the lands farther east were more often derived at second-hand than based on direct observation. Chryse with its gold-mines and tortoise-shell and Thinai with its silk and silk products are mentioned in the Periplus of the Erythraean Sea; [196] though the inland silk route is traced fairly accurately, there is no evidence of any direct or detailed knowledge of the sea-route to China on the part of the writer.

A Chinese source states that in A.D. 132, a king of Yetiao named Pien or Tiao-pien sent an embassy to China and received a present of a gold seal and violet ribbon in return. Ye-tiao has been taken generally to stand for Yavadvīpa, and perhaps less plausibly, the king's name has been restored in Sanskrit as Devavarman. If these restorations are correct, this would be evidence of Hindu influences already at work in Java, the island being known in China by its celebrated Sanskrit name and as being ruled either by a Hindu ruler or a native with a Hindu name.

About the middle of the second century A.D., Ptolemy, the Alexandrine geographer, gave a geographical description of the world as it was known in his time, with maps and tables of latitudes and longitudes of important places; the lands of the Far East are included in his account under the name ‘India beyond the Ganges’. It is, however, by no means easy to identify his names on modern maps. He says: “The island Iabadiou or Sabadiou signifies barely ‘island’; this island is said to be very fertile and much gold is also got there; its capital named Argyre (‘silver city’) lies at its western extremity.” Iabadiou has been taken generally to be a representation of the Prākrt form of the Sanskrit name Yavadvīpa, and Ptolemy’s account is of interest not only for the Hindu name of the island, but for its recalling a well-known verse in the Rāmāyana, although a relatively late one, and perhaps of the same period as Ptolemy, describing that very island for the benefit of the apes that were to set out in search of Sītā:

Yatnavanto yavadvīpam saptarājyopasobhitam
Suvarnarūpyakam caiva, suvarnākara-manditam
Yavadvīpam atikramya Sisiro nāma parvatah
Divam sprsati srngena devadānavasevitah [197]

Here Yavadvīpa is said to comprise seven kingdoms, to abound in gold and silver and to have gold-mines as well. Beyond the island lay Mount Sisira touching the sky and frequented by devas and dānavas. The identity of names, and the occurrence of the phrase sampannam kanakākaraih in the description of ‘the noble island of Java,’ dvīpavaram Yavākhyam, in the Cangal inscription of Sañjaya (732 A.D.) seem to justify the view that Ptolemy’s Iabadiou is no other than the island of Java, possibly Sumatra also being included. And as barley does not grow here, the name has been held by some to refer to the shape of the island. But other views have been held especially by those who consider the mathematical data of Ptolemy more important than phonetic similarities between the names mentioned by Ptolemy and those on our maps. In any event, it is certain that the geographer knew several places in the Archipelago and Indo-China, right up to the borders of China, under their Hindu Sanskritic names, and this is full of significance for the date of the first establishment of Indian colonists in these lands. And Ptolemy's mention of cannibals in several parts here and of men with tails, though possibly exaggerated and distorted in part, must be accepted as some evidence on the state of savagery that prevailed here before the arrival of the Hindus.

The Sanskrit names for cinnamon and nutmeg imply that they were brought to India at an early date from across the seas and may well be taken to attest the most ancient trade relations between India and the Archipelago. But the ignorance of even Indian astronomers of the fifth and sixth century on the proper configuration of the eastern lands—witness their legendary references to Yavakoti surrounded by golden walls—shows that the knowledge gained by Hindu mariners of the coastal towns did not spread inland and was itself, possibly, by no means accurate or extensive.

From the third century A.D., at the latest, begins a series of stone inscriptions scattered over the various parts of Indo-China and Malaysia, which, amidst the differences of time and place, are characterized by an undeniable family likeness. They cover some centuries and are found in Burma, Malay Peninsula, Java, Borneo, Kambuja and Campā. They are usually composed in Sanskrit and written in a script which though often called ‘Pallava’ is perhaps best described as ‘South Indian’ using the term so as to include Deccan also. The princes mentioned in these inscriptions have usually names with a -varman ending. We thus see that the whole of Sout-east Asia was touched by this vast movement of culture which must have been slow and steady and a gradual and peaceful penetration rather than the result of military expeditions and violent conquest. That the earliest of these inscriptions so far known comes from Vo-Canh in Campā should be held to be more an accident; it might have come from any other part. At any rate it should not lead us to infer that Campā was the earliest region to come under Hindu influences; the chances are that the western coast of the Malay Peninsula and the islands had been occupied some time before the lands bordering the Gulf of Siam and the China Sea were reached. Time, the action of natural forces and the vandalism of man have destroyed irretrievably many of the traces of these ancient occurrences and what is now left enables us to see only a part of the story, and that only in its broad outline. There is little doubt that in every case, the evidence now available belongs to a period much later than the commencement of this movement of colonization.

West Java: Among the earliest traces of Hindu culture in Java now known are the Sanskrit stone inscriptions of Pūrnavarman from the West Java. “It is significant,” says Vogel, “that these earliest records of Hindu settlement are found exactly in that part of the island where the Dutch traders first established their ‘factories’ and which became the centre from which the power of Holland has spread over the whole of the Indian achipelago. The geographical position of the Batavian coast with regard to the continent of India and Sumatra and the special advantages its figuration offers to shipping and trade are circumstances which will easily account for a coincidence that is certainly not due to mere chance.” [198]

Though the extant inscriptions of West Java are, as we shall see, of a later date than those of Borneo, there can be no doubt that Hindu culture must have reached Java, if anything, a little earlier from South India than it reached Borneo.

The routes taken by Hindu colonists can only be a matter of surmise in the absence of direct evidence. There is no reason to believe that there was any particular centre, Caiya, as Dr. Wales has suggested, or any other place to which a greater importance attaches as the basis of further advances in the movement of colonization. It seems much more probable that every area which was Hinduised during the early centuries of the Christian era became in its turn a centre of diffusion of the new culture among its neighbours. Java and Sumatra, however, attained great celebrity in the arts about the eighth century, and the evidence of Indo-Chinese epigraphy and Javanese traditions taken together attests the rather widespread influence of Indo-Javanese culture in these eastern lands. [199]

The inscriptions of West Java are engraved in the distinctly South Indian type of characters to which the names ‘Vengi’ and ‘Pallava’ have been applied by epigraphists; the letters show a stage of development which would place them in the middle of the fifth century A.D., about half a century later than the Kutei inscriptions of Mūlavarman of Borneo.

These inscriptions are four in number. Apparently the earliest of them is the one known as the Ci-Aruton record; it is a single anusthup verse engraved in four bold lines, each line comprising a pāda (quarter of the verse), under a pair of human feet in front of which are two additional carvings which have been described as ‘lotuses’ or ‘spiders’.

The inscription just says that these footprints which are like those of Visnu belong to the illustrious Pūrnavarman, the lord of Tārumanagara, a valiant ruler of the earth. We have a repetition of Pūrnavarman’s footprints in another place in the same district; “they are partly broken off with the top of the rock” (Vogel); here we have a verse in the sragdharā metre engraved in two long lines in elegant characters of the same type as the Ci-Aruton record. This inscription is usually called the Jambu rock-inscription, and as it constitutes the nearest approach we have to an account of the king's reign, it may be reproduced here:

Srimān dātā krtajño narapatir asamo yah purā Tārumāyām
nāmnā srī-Pūrnnavarmā pracura-ripu-sarā-’bhedya-vikhyāta-varmā
tasyedam pādavimbadvayam arinagarotsādane nitya-daksam
bhaktānām yannrpānām [200] bhavati sukhakaram salya-bhūtam ripūnām

i.e. “Illustrious, munificent and true to his duty was the unequalled lord of men—the illustrious Pūrnavarman by name—who once (ruled) at Tārumā and whose famous armour (varman) was impenetrable by the darts of a multitude of foes. His is this pair of footprints which, ever dextrous in destroying hostile towns, is salutary to devoted princes but a thorn in the side of his enemies” (Vogel). A third inscription, the Kebon Kopi (coffee garden) rock-inscription, accompanies the footprints of the elephant of the lord of Tārumā [201] which is compared to Airāvata, the divine elephant of Indra.

The inscription comprises just one anusthup verse, illegible in part, and engraved in one line between two enormous elephant footprints covering almost the whole of the flat surface of the rock. This was no doubt Pūrnavarman's elephant. [202]

In all these three instances the inscriptions stand in definite relation to the footprints near them. The worship of the footprints of gods, prophets and saints is well known in India and Ceylon; and in the Rāmāyana, Rāma is said to have given Bharata his pādukas (sandals) to represent him in the rule of Ayodhyā during the period of his exile. The exact import of the footprints of Pūrnavarman and his elephant has been the subject of some discussion. Vogel suggests that the Ci-Aruton rock marks the spot of the king's cremation and that the Jambu rock was more or less worshipped as a posthumous shrine of magic potency; he admits that it is even more difficult to explain the motives which prompted the engraving of the elephant's footprints and the inscription accompanying it. [203] Stutterheim suggests that the footprints, the king's as well as the elephant's, are marks of occupation after the conquest, and he recalls the practice of the conqueror placing his foot on the neck or head of the vanquished rulers to signify their subjection. [204]

I think that all the inscriptions are posthumous and probably put up in the reign of Pūrnavarman's successor, most probably his son. Let us note the word purā in the Jambu record and the absence of anything to indicate that any inscription was actually engraved in Pūrnavarman's reign. The footprints of the king and his elephants are no more than mementos of the valour and heroism of a great king and his state elephant that played a notable part in his wars of conquest. Pūrnavarman was thus, so far as we know, the first conqueror-king of Hinduised Java as the Jambu inscription clearly shows, and all the inscriptions mentioning him turn out to be memorials of his rule raised by a pious and grateful successor who inherited a considerable kingdom with many vassal states. These impressions receive further confirmation from the remaining record in the same neighbourhood and of the same time¾the Tugu rock inscription.

Before considering this inscription, however, a word must be said on the spider-like attachments to the footprints in the Ci-Aruton record. The most plausible of the numerous explanations offered by scholars [205] is that of Finot which treats these marks as actual representations of spiders, and refers them to the practice, common in Indo-China and Insulindia, of representing the souls of men, which are supposed to leave their bodies when they are asleep, in the form of insects, particularly spiders. The padadvayam and the spiders then represent respectively the body and the soul, the nāmarūpa of the king. If this view is correct, it would go to show that the blending of Indonesian and Hindu cultures, of which we have many tangible instances in later monuments and institutions, began to be effective and successful at a very early stage in the contact between them. And this may well have been so. We know for instance that even the Rgveda is now seen to be at least in part the production of a composite culture.

“Another puzzle” related to the Ci-Aruton rock is the single line of cursive writing “which is written over the inscription proper but in a different direction along the right-hand side of the royal footprints.” None of the attempts so far made to read this line, including that of Jayaswal to treat them as ‘shell-characters’, [206] can be pronounced successful or convincing. It is by no means certain that this line was a part of the original inscription.

Let us turn now to the Tugu rock inscription. This is engraved in five lines running round “a natural, undressed rock, conical in shape and measuring about one metre in height and a little less in diameter.” [207] It comprises five anusthup verses, and its bold characters closely resemble those of the other three inscriptions already noticed. The first verse mentions the Candrabhāgā, dug out of old (purā) by the famous king of kings (rājādhirāja), the strong-armed guru, and flowing into the sea after skirting the famous city—evidently Tārumā. The remaining verses form one long and complex sentence, but their construction is not so difficult or obscure as others have maintained. They state that Pūrnavarman, who was prominent among kings by the height of his prosperity and virtue, carried out the excavation (of the channel) of a beautiful stream (nadī ramyā) of clear water (nirmalodakā) in the twenty second year of his reign; the work was begun on the eighth day of the dark fortnight of the month of Phālguna and completed on the thirteenth day of the bright half of the month of Caitra, i.e, in a period of twenty-one days; the length of the stream was 6122 bows, [208] and it cut across the camping-ground of the grandfather who was a royal sage (pitāmahasya rājarseh); [209] and the opening of the stream was marked by the gift of a thousand cows to Brahmins.

This record therefore also commemorates an act of Pūrnavarman, the digging perhaps of an irrigation channel. Obviously this inscription is the work of a successor of Pūrnavarman, and the reference to the grandfather must be to the grandfather of the author of the record; if the author was Pūrnavarman's son, the grandfather would be the father of Pūrnavarman; but of this we cannot be sure. It seems clear, however, that though all the West Javanese inscriptions of this time refer to Pūrnavarman directly or indirectly, thereby indicating the high place he filled in the history of the Hindu colony in this part of Java, still Pūrnavarman had predecessors and successors, and the kingdom of Tārumā flourished for some generations, though we know little of the details of the story. Pūrnavarman, we learn, had a reign of at least twenty-two years. Let us note also the reckoning of the month from the new moon (amānta system) which is characteristic of the South Indian calendar. [210]

Thus, we have clear evidence of a settled Hinduised society flourishing in West Java in the fourth and fifth centuries. Pūrnavarman rules for over twenty-two years [211] in which he apparently effects some conquests and lays the foundations of a durable kingdom; this king, who is compared to Visnu and makes a gosahasra dāna, was doubtless a Hindu colonist from South India or a Hinduised Indonesian. That Hinduism was the prevalent faith at the time in Java is borne out by Fa-Hien who came to Java in A.D. 414 at the end of a storm-tossed voyage on his way from Ceylon to China. He says that the land was full of Brahmins and heretics (Pāsupatas?) [212] and the lore of Buddha little known. He must have had occasion to know as he had to wait in Java for five months before he could embark again, which he did on a large ship sailing to China and manned by Indians. Fa-Hien's account of the religious condition of the Hindu-Javanese society of his time accords with what we learn from other Chinese accounts of the mission of Gunavarman in Java and of his preaching of Hīnayāna (Mūlasarvāstivāda) there with the support of the queen-mother, between the years A.D. 396 and 424.

It is necessary to note that the identification of Ye-p'o-ti of Fa-Hien and Chop'o of the Gunavarman story with Java, though probable, is not accepted by all scholars.

Early Hindu colonies in Java were not confined to the western part of the island, as we were apt to think till recently from the state of the evidence at hand. For in December, 1933, Dr. Stutterheim had his attention drawn to a short rock inscription at Rambi-poedji in the eastern corner of Java near the Loemadjang-Djember road. [213] The script of the inscription is clearly early South Indian, and there is not the slightest doubt that this record falls in the same class with the inscriptions of Pūrnavarman of West Java and Mūlavarman of East Borneo. The whole inscription comprises only five letters very clearly engraved on a megalith with its upper sides smoothened and a big knob on one of its sides. On the relatively rough underside is found the inscription which reads:

pa rvva te sva ra,

the Lord of the Mountain, a name of Siva. Was this big boulder a primitive Indonesian object of worship and did the incoming Hindus continue to recognize its sanctity in the new order by treating it as a linga or symbol of Siva? If that was so, the blending of the old and the new began very early and went on in the happiest conceivable manner.

We must also note the one-line inscription below a number of religious symbols such as cakra, sankha, tridanda, parasu, and kamandala engraved on the rockface of Toek-Mas at Merbaboe in Central Java. The record speaks of a tīrtha and may be dated about A.D. 650. It is the earliest record giving a clue to the state of Hindu-Javanese religious observances. [214]

Other Kingdoms: The Chinese annals mention embassies from a number of other minor kingdoms in the southern islands. Their identification is not easy. But the general impression we get is clearly that of a set of minor states in active intercourse with one another and with China on the one side and India on the other.

Kan-t'o-li is one of these kingdoms located somewhere in Sumatra by the best modern opinion; its rulers were in communication with China during the period A.D. 450 to 563. [215] The names of these rulers, to judge from their Chinese transcriptions which alone are now available, were typical Hindu names, and the manners and customs of the country are said to have been similar to those of Campā and Kambuja.

We may note at this point another trace of South Indian influence in Sumatra, though we are by no means sure that it dates from this very early period. It is the presence of certain names of tribal sub-divisions which are unmistakably South Indian among the Simbiring, a branch of the Karo-Batak race. These names are—Coliya, Pāndiya, Meliyāla, and also Pelavi (Pallava if not Melawi or Malay), as well as Tekang (Tekkanam, Deccan). Though there were many occasions in later history when these names might have been introduced, it is not altogether impossible that they came in early, or at least the ground was prepared early for their reception at a later time. [216] The social organization of the Karo-Bataks seems to date from a very remote past and it is quite probable that these names were taken over when they were still powerful realities in South India. Some support for this view may be derived from archaeology; no temples seem to have survived from these early days when Indonesian society was being transformed by the advent of Indian Hindu influences; and it is reasonable to suppose that these changes took place at a time when temples were still built of wood or other perishable material, and we know that this was so in the early centuries of the Christian era. That a continuity was maintained between the older Indonesian religious institutions and the later Hindu-Javanese temples, as we may call them, is seen from the preservation of and worship offered to large bronze kettledrums in such temples, and these drums are known clearly to have belonged to the pre-Hindu phase of Indonesian religious life.

Sumatra certainly came into contact with Hindus and Hindu culture during the first two centuries A.D. at the latest, and the contact thus established never wholly ceased and was kept up through varying fortunes for well over a thousand years. And this was not confined to contacts with South India, though doubtless proximity gave it a dominant part. I have dealt with Hindu Sumatran history in some detail elsewhere; [217] what happened in Sumatra is typical of the history of almost every one of these colonies, and the following sketch in which Heine-Geldern briefly sums up this history and gives a fair estimate of the strength of the agencies concerned will be read with interest. [218] “I need only point out,” he says, “the Buddhist establishments founded at Nālandā in the ninth century and at Negapatam about A.D. 1000 by kings of Srī Vijaya; the reproduction of Sumatran Buddhist idols in a Nepalese manuscript of the eleventh century; the prominent part played by Srī Vijaya in the history of later Buddhism, and the manifold threads of Buddhist activity and learning spreading from Sumatra to China, India, and even Tibet. The invasions of Sumatra by a king of Cōla in the eleventh century, the Tamil inscription of Luba Tua from the year A.D. 1088 and the Dravidian tribal names still to be found among the Batak are also not to be forgotten. So we can safely assert that Sumatra has not only once been colonized by Hindus, but that, owing to more than a thousand years of close connection, it became an integral part of the Greater Indian cultural area. It is natural that other cultural elements reached Sumatra from the Tamil region and Malabar than those that came from Bengal, and again, influences coming from South India in the time of the Cōla kings of the eleventh century, must have differed remarkably from those of the Pallava period in the seventh. Moreover, material as well as spiritual influences did not make their way always directly from the Indian mother-country but were also transmitted by way of various Indian colonies, specially by Java, thus being subjected more or less to changes and assimilations before reaching the island.”

Then we have fairly long notices of the kingdom of P'o-li from which embassies reached China between A.D. 518 and 630. These so-called ‘embassies’ were doubtless visits of groups of traders, and the ‘tribute’ they offered were articles of merchandise which were exchanged by way of presents for other articles of more or less equal value; these embassies often produced letters real or faked, purporting to be addressed to the Imperial court by the ruler of the country from which they came; even these letters may be right in so far as they give the impression that Buddhism was spreading in the archipelago.

P'o-li is said to be on an island in the sea to the southeast of Canton. It has been located by some in Sumatra, and identified by others with Bali; but some of the data do not suit either and lead one to think of Borneo (Puni); we are told for instance that it took fifty days to traverse the island east to west, and twenty days north to south, and there were 136 villages in it. “The functionaries are called Tu-ka-ya-na, and those of lower rank Tu-ka-si-na. The people of this country are skilled in throwing a discus-knife; it is the size of a (Chinese metal) mirror, in the middle is a hole, and the edge is like a saw; when they throw it at a man they never fail to hit him. Their other arms are about the same as in China. Their customs resemble those of Camboja, and the productions of the country are the same as of Siam. When one commits a murder or theft they cut off his hands, and when adultery has been committed, the culprit has his legs chained for a period of a year” (Sui Annals). [219]

These particulars have a familiar ring and may well apply to many a Hinduised land in the islands at the time. But then we are told that the family name of the ruler of Fo-li was Kaundinya, and that the queen of Suddhodana, the father of the Buddha, was a woman from his country. This shows that the royal family was actually a line of Indian princes or considered itself to be such. The name Kaundinya naturally leads us to think of the other Kaundinya, the Brahmin founder of Fu-nan (later Kambuja). We may complete the data on this tantalizing kingdom of P’o-li by another citation, this time from the Leang annals. [220] “The people of this country use cotton for their clothes, and also make sarongs of it. The king uses a texture of flowered silk wrapped round his body; on his head he wears a golden hat more than one foot high, its shape resembling the one called pien in China, and adorned with various precious stones. He carries a sword inlaid with gold, and sits on a golden throne with his feet on a silver footstool. His female servants adorn themselves with golden flowers and all kinds of valuables, and some of them carry white feather-dusters or fans of peacock feathers.

“When the king goes out, his carriage, which is made of different kinds of fragrant wood, is drawn by an elephant. On the top of it is a flat canopy of feathers, and it has embroidered curtains on both sides. People blowing conchs and beating drums precede and follow him.”

Whether P'o-li was Bali or not, modern Bali is of great interest to us as the only island where Hindu culture has survived to this day; and to this survival we owe the preservation of manuscripts which have so much to tell of the early history, literature and culture of Java from which these books disappeared soon after the advent of Islam. Bali is, as it were, a living museum of mediaeval Java. Neither Balinese traditions nor Balinese monuments carry us back to a very early period. A persistent tradition ascribes the incoming of Javanese culture to a mass flight of the Hindu-Javanese after the fall of the empire of Majapahit; but we are sure that Bali possessed a Hinduised population and culture many centuries before, the earliest date on a Balinese charter being S. 818 (A.D. 896). There is no need to follow the mediaeval history of Bali here. But we may note that the so-called Vedas of Bali are compilations of secret tantric mantras, and that the only part of the real Veda known seems to be a corruption of the gāyatrī-mantra. In religion the most curious development is the Sivāditya cult, a combination of Saivaism and Sūrya-worship. Siva was considered at once as the highest member of the Hindu triad and as a form of Sūrya, and this peculiar tenet appears to have formed the chief deviation from Indian Hinduism for which the Hindu Javanese were responsible. [221]

The Balinese village has preserved its cultural character as a Hindu organization to this day; though doubtless some of its characteristics may be derived from pre-Hindu Indonesian institutions, it is now not easy to identify them as such—a trait which finds its parallel in South India where we are hardly able to distinguish the Aryan from the pre-Aryan elements in the culture of the Tamils in historical times. The villager in Bali entertains a vivid consciousness that he is a member of a religious community to which it is a privilege to belong. The three essentials of every village are first, a place for the common meetings of the villagers, which is usually in the centre; secondly, places for the worship of the Lord of the Soil and the ancestors (pitarah) who were the founders of the society, generally located at an elevation; and lastly, a place usually below the village for the disposal of the dead (preta) who became pitarah after purification in due course. Balinese temples are of various types. Worship spots where animistic sacrifices were performed are derived probably from Indonesian origins. These may be said to correspond to the grāmadevatas of an Indian village. Then there are Hindu and Buddhist shrines of the usual type, and occasionally a shrine with Islamic associations. Lastly, there are spots where men of earlier generations, and historical or legendary celebrities are commemorated in piety and gratitude. Each village had its complement of officers whose number varied according to needs. Business is transacted at periodical meetings of the people of the desa held on occasions of a quasi-religious character attended with feasting and entertainments. This sketch of village life as it obtains today in Bali [222] may be taken to give a fair clue to life in all the Hinduised colonies of the eastern lands as it obtained in the early centuries of the Christian era.

The most interesting and instructive evidence of South Indian influences at work in the colonies is furnished by the Buddha statues found in various places in the islands; these are not numerous, but enough to enable us to see clearly their mutual relations and also their resemblances with similar statues from Siam, Kambuja and Annam, and to point beyond any doubt to the celebrated art-centre of Amarāvatī on the banks of the Krsnā as the source of their inspiration. A bronze from South Djember, 42 centimetres high and therefore larger than the usual run of Javanese statuettes; another, still larger (75 centimetres) found in Sikendeng on the west coast of Celebes; and the colossal stone Buddha of Bukit Seguntung at Palembang recently restored almost to its original form by a head from the Batavia museum being successfully tried on its trunk¾are all in the characteristic Amarāvatī style, even the differences noticeable among them exactly reproducing similar difflerences in the Amarāvatī images. It is probable that the bronzes were brought from Amarāvatī by the colonists, or imported from there by colonists already established overseas; the transport of the large stone Buddha of Palembang must have been more difficult, though by no means impossible. If that image was made locally, it must have been the work of an artist who went to school at Amarāvatī. The art of Amarāvatī, it should be noted, reached its high watermark in the latter half of the second and early third centuries A.D., and the Buddha of Palembang shows affinities with the earliest phase of this art. It is thus very likely that this Buddha image is the oldest relic of Hindu culture in the archipelago. And Palembang deserves to count among the oldest centres of this culture; which is in good accord with the statement contained in late Chinese authorities that this was the region where the early state of Kan-t’o-li flourished.

The Celebes Buddha has been studied in detail by Dr. Bosch, [223] and he has demonstrated in a most convincing manner that this bronze has little in common with early Sumatran (Srī-Vijaya) Hindu-Javanese art, and that it must have been imported directly into Celebes from the Amarāvatī region sometime during or after the blossoming of Amarāvatī art and before the rise of Srī Vijaya, i.e. between the second and seventh century A.D. No closer dating is possible when we have no certainty whether we have before us an original art-piece from Amarāvatī as suggested above, or a local copy. Bosch rightly observes that it is now futile to guess the nature or the duration of the Buddhist settlement on the west coast of Celebes to which this fragmentary Buddha image bears solitary witness. Still less is it possible to say how far this culture penetrated into the interior of the island. Yet one fact deserves mention; an ancient bell and a pair of cymbals were recently presented to the musical collection of the Batavian Society of Science and Arts by an official of the district of Loewoe in Celebes; these were used till then by the Boeginese bissoes [224] for chasing evil spirits during and after child-birth. The bell and cymbals are very similar to those still in daily use in South India in domestic worship and otherwise. Thus we may suppose that South Indian Buddhism, received at first on the west coast of Celebes, penetrated along the valley of the Karama river into the neighbouring province of Loewoe, and the people of that region preserved these ancient ideas and usages until the time came for them to mingle with a fresh but allied stream of culture that came in with the spread of the empire of Majapahit; for the Nāgarakretāgama counts Luwak among the dependencies of that empire.

What then are the general conclusions that emerge from the data briefly reviewed so far? We see that the movement of colonization was in full swing in the second century A.D., and its beginnings may well be put at the beginning of the Christian era. The Hinduisation of the archipelago did not take place all at one time, and must have been a gradual process with different beginnings and results in different places. The Buddha of Palembang was perhaps set up in the second century and Devavarman ruled in Java about the same time. On the other hand, the beginning of Hindu rule in Borneo (Kutei) can be traced only to one generation before Mūlavarman c. A.D. 400. The relics of this movement are naturally traceable in the Malay Peninsula also where the oldest epigraphical records date from about A.D. 400.

It would of course be wrong to imagine that Hinduism in the archipelago was confined only to the spots that have yielded relics; this is largely a matter of chance, and but for the Palembang Buddha recovered in so fortuitous a manner and the Yūpas at Moeara Kaman (Borneo), we should have known nothing of two of the most ancient Hindu centres in Sumatra and Borneo. We cannot also be guided in our conclusions by the abundance of Hindu relics on a site; for the chances are that these date from the period of the spread of Hindu-Javanese power which came long after the initial period of colonization and by which a culture long since strongly Hinduised spread itself practically over the whole archipelago. Even trained archaeologists sometimes find it difficult to separate the remains of the earlier culture from those of the later. Palaeography and art-style are the two unmistakable marks of the antiquity of objects belonging to really early times and attesting direct contact of these lands with India in those days.

These tests, as we have seen, point to a time much earlier than that of the rise of the Pallavas as the age of colonization par excellence, and the country farther north of the Pallava kingdom of history as the original home of the colonists. We have indeed evidence of a somewhat later date attesting the part of the rest of the cast coast of South India and even of its west coast in the movement; but the primacy in this expansion movement belongs to the Andhra country, to its great centres of Buddhism and its trade marts on the coast.

In what manner did Hindu influences spread in the eastern lands, and what was the motive of the migration of the Hindus to the eastern countries? In a general way, this movement may well be looked upon as just a continuation of the process by which the Deccan and South India were Aryanised and Hinduised by the inflow of northern influences. Having secured the prevalence of their culture in the whole of Jambudvīpa, the apostles of Aryan culture turned their attention to the neighbouring lands (dvīpāntara) [225] across the sea. But the question remains: What exactly was the means of propagation and who were its agents? This question has been discussed at considerable length by Krom [226] and the following observations are based on that discussion.

Of political conquest and empire-building, of the holding down by force of vast populations and their exploitation to the economic benefit and political advantage of a distant foreign power, there is no question here at all. All our sources agree in presenting a picture of a number of autonomous Hinduised states, each going its own way and living its separate life, all having direct but by no means very brisk trade relations with India and China—witness for instance Fa-Hien's long halt in Java. But of the political influence of India there is no trace whatsoever. [227]

To the question, who first brought the elements of Hindu culture to these lands, there are a number of possible answers: the merchant, the adventurer, the priest or the exile. We shall consider all of them in turn.

The Merchant. The existence of trade relations between India and the East, and the considerable share of Indians in the carrying trade of the Indian Ocean are alike attested by the early literature of the Tamil country and of Pāli Buddhism. The splendid description of the sea and its riches given by Varāhamihira at the beginning of the chapter on Agastyacāra in his Brhat-samhitā has been interpreted as an indirect reference to the large gains already enjoyed for many years by Indian merchants as a result of their trade with the archipelago. [228]

Wherever the merchants found a suitable market for their wares or articles which they wished to bring over to India, in those places they would naturally spend a considerable time, establish lodges and factories, and perhaps enter into marital relations with the women of the land and raise progeny. Thus might grow up gradually a half-Indian, half-Indonesian population which became the means of spreading Hindu ideas and institutions among the indigenous population. Parallels to this development may be found in what happened a thousand years later when Islam came into the archipelago or in the early stages of the establishment of European trading companies in the East.

The Adventurer. The second possibility is that of robber chieftains raiding and plundering the coasts with their followers and where possible holding the population under by force, and thus securing for themselves footholds in the new countries from which they might extend their depredations farther and farther. More friendly relations with the local population and something like a new society might grow up in such places in course of time. In such cases, conceivably, the foreign adventurer might marry into the local royal family and thus legitimatize and strengthen his position in the eyes of the people. The memory of such occurrences seems to be preserved in the later Malay and Javanese traditions; but they need not necessarily have been confined to the beginnings of colonization and might well have happened long after the Hinduisation of these lands was completed.

The Priest. Tradition is eloquent on the role of the priest in the spread of Hindu culture. The Brahmin founder of Fu-nan in Cambodia has his peer in the Javanese tradition relating to Tritresta who is said to have introduced the Hindu mode of divine worship and the Hindu calendar into Java and whose son became king there. Most of the inscriptions we have dealt with before attest the great importance attaching to religious ceremonial in the colonies and are framed in the correct metrical Sanskrit idiom of scholarly Brahmins; such verses could hardly proceed from merchants or soldiers. After all it is the Brahmin who can alone secure to society the protection of the Higher Powers and had the knowledge required to assist the state with advice in a difficulty; no wonder then that whether by hearsay or by observation of neighbouring states, Indonesian chiefs came to realize the worth of such magical protection and did their best to procure it for themselves. There was much competition for the services of such Brahmins and they began to appear in increasing numbers; and besides performing the particular services for which they were invited, they became active agents in the further propagation of Hindu culture.

And lastly, the Exile. In this category we have to reckon groups of people forced by circumstances to leave the land of their birth. Krom states that at the beginning of the Christian era the Tamil country was in a restless condition and that Tamil texts narrate the story of the siege and fall of Tiruchirapalli after which a hundred families migrated in ships to ‘an island’. I am unable to trace any authority for these statements and Krom does not cite any. Again, Moens has sought to find in the confusion following Samudragupta's raid into the Deccan and the political revolutions in the Andhra country preceding and following Pulakesin's conquest of Vengi, circumstances which compelled considerable sections of the population to leave the country and migrate to the islands. Others have turned to the period following the death of Harsa, in northern India, for a similar reason. These suggestions, however, relate to a much later time than the beginnings of colonization; and they lack evidence in their support. There is nothing to indicate that political unsettlement in India drove people to abandon the country and migrate to other lands. And on the whole, the role of political exiles in the furtherance of this movement of colonization could not have been anything so extensive or significant.

The part of the learned Brahmin priest might appear at first sight to be the most important of all; but then his services would be required and appreciated only in a society that has already gained acquaintance with Hindu culture and institutions, and it seems extremely unlikely that Brahmins went out in any numbers in a missionary spirit to preach their creed and commend their practices to peoples who were utter strangers to both. Even Buddhist monks, who were far more eager to preach their gospel, often awaited a call before they started on a preaching mission. The case of a Brahmin founder of a kingdom like Kaundinya of Fu-nan is of course quite another matter. The best course then would be to suppose that the merchant's role was the most important at the outset, and when success attended his enterprise and a mixed society arose, the priest came in to consolidate it and make it a centre from which the process of Hinduisation could be extended further into fresh lands.

In every instance in which we are able to follow the history of the new states, we find the native elements of the population holding their own by the side of the newcomers, and the culture and society resulting in course of time are seen to be decidedly of a mixed character. If there was any exclusiveness among the Hindu colonists at the beginning, it must have broken down under the pressure of time and circumstance.

We can hardly believe that the actual course of events was the same in all places or that the use of force against natives was totally unknown. Hindu institutions might have been forced on an unwilling people in one place and eagerly welcomed by them in another; but everywhere the newcomers seem to have been few and the natives many; in course of time this naturally led to the growth of a mixed culture to which both elements of the population had made their contribution. The mixture was sometimes symbolized at the outset by intermarriages among the people in some cases, in the ruling families in others.

The language of the inscriptions has much to tell us in this matter; the earliest of them are in good Sanskrit, certainly not the language of the people, and represent perhaps the phase of relative exclusiveness of the Hindu element in the newly-settled lands. When the new mixed society comes into its own some generations later, inscriptions begin to appear in Old Javanese or Old Malay; the structure of the language is fully Indonesian, but a large number of words of Indian origin have got in, and these are mostly the names of the higher culture goods that have come in with the Hindus.

Let us also note this: the Indian words are borrowed not from any of the spoken idioms of the Hindus, but usually in their Sanskrit form. Thus the new language was an Indonesian idiom adapted by liberal borrowings from Sanskrit to suit the newly-growing culture. [229]

The javanese tradition relating to Adi Saka, the introducer of a new religion, a new social order, a new script and a new calendar, viz. the Saka era beginning with A.D. 78, is a transparent fiction which personifies the name of the era employed in Javanese inscriptions (the era was used elsewhere in the colonies also) and preserves the memory of the times when much of the higher culture of South India overflowed into Java in the early centuries A.D. This tradition also clearly points to a pre-existing community in Java which received the new culture with more or less readiness when it was first introduced among them.

Everywhere the court and the nobles might have sought distinction by affecting the new culture more or less thoroughly; but in the interior, the people would have kept to their traditional ways and their mode of life would have been but little touched by the new practices. There is a Javanese proverb—Nagara mawa tata, desa mawa cara (the court has its culture, the people their custom)—which stresses the sharp difference between the higher Hindu-Javanese culture of the land and its folk-ways; this surely is based on a real difference between the Hindu-Javanese and the purely Javanese elements of the population. [230] Yet in course of time their pagan religion forced its way into the new society and compelled a compromise between the old and the new. The great popularity in Java of Siva as Bhatāra Guru, the divine bestower of wisdom, and the worship of Srī as the special protectress of the rice-crop, may well hide the essence of Indonesian deities under a Hindu garb. Likewise in political life, while the king might go on issuing his edicts based on the Hindu law-books, the local officials stuck tenaciously to the rule of custom. And in course of time when contact with the mother-country became infrequent and finally ceased, the Hindu culture of the colonies necessarily underwent a radical transformation, lost its capacity for further expansion and became merged in a culture which was neither Hindu nor Indonesian but a synthesis of the two.

To trace in any detail the impulses derived by these islands from South India would be to embark on an elaborate review of the whole range of Hindu-Javanese history together with that of the empire of Srī Vijaya, and that is beyond the scope of my present purpose. Some observations of a general character calculated to illustrate the sustained contact between South India and the islands, and its results on the religion, administration, literature and art would, however, not be out of place here.

The role of Agastya as the promoter of Hinduisation in Java and the preacher of Saivism is well attested by epigraphy, literature and sculpture. [231] The famous Dinaja inscription (A.D. 760) records how the ruler of East Java, who was a great devotee of Agastya, made a fine abode for that great sage, and installed in it a beautiful Agastya image of black stone in the place of the wooden image set up by his ancestors—clear evidence of the importance and continuity of Agastya worship in Java. And Bosch has argued in a convincing manner that Agastya was supposed to have played the role of an intermediary between the Deity and Royalty in Java, just as other sages did in Campā and Kambuja. [232]

This is just a continuation of the Agastya tradition from South India in which he is represented as the pioneer of Indo-Aryan culture in peninsular India south of the Vindhyas. We may note in passing that the Cangal inscription (A.D. 732) contains a reference to Kuñjarakuñjadesa in a line from which a few letters have disappeared. This desa was till recently taken to be the borderland between Tirunelveli and Travancore where the Tāmbraparni river takes its rise in the Agastyakūta hill. Recently the suggestion has been made by Stutterheim that the line should be read:

Srimat kuñjarakuñjadesa nihitam Gangādi tirthāvrtam

and that the desa should be looked for in Central Java, where the Toek Mas river is found compared to Gangā in another inscription noticed already by us. [233] But this makes little difference, for the possibility remains that the desa in Java got its name from a district in South India.

The same scholar has argued with great force that the Hindu-Javanese Candi, though its shape and ornaments are Hindu in origin, is really a purely Indonesian monument based on purely Indonesian conceptions. The close similarity in architectural features between the Hindu-Javanese temple and the Pallava temples of South India has often been noted and closely studied in a well-known memoir by Dr. F.D.K. Bosch. [234] And Stutterheim is not oblivious of this relation; he says: “the entombing of old Javanese kings was not a Hinduistic practice grown in course of time more and more Indonesian, but a thoroughly Indonesian ceremony, which in Java and Bali took a Hinduistic form and should be considered as a higher form of the analogous ceremonies of the Dayaks and other Indonesian peoples not influenced by the Hindus.” [235] Temples dedicated to dead kings and warriors are not so entirely unknown in South India as Stutterheim seems to think, [235] though one sees quite clearly that the practice of worshipping actual statues of dead kings and queens as gods and goddesses carried the South Indian practice much farther than in the mother country. And one may doubt if Stutterheirn's references to the Indonesian conceptions of ‘the land of the souls’ has any real bearing on this development. Again, it is far from certain that all Candis are sepulchral in character; in Bali a temple is exclusively the residence of a god, though the Javanese Candi like the South Indian temple, seems to have had a dual character—a temple as well as a mausoleum. [237]

The frequency with which Ganega images occur in Java is paralleled only by the innumerable shrines to that urbane godling in the whole of South India, and it has been very properly explained by Stutterheim that, as the guarantor of safety in all enterprises and protector against vighnas (obstacles), he was honoured in ferries, forests, mountains and other dangerous spots. [238] A peculiar Bhīma-cultus of ancient Java and the posthumous name of Krtanāgara of Singhasāri (A.D. 1268-1292), viz. Sivabuddhaloka may be noticed in passing as witnesses to interesting features of later Hindu-Javanese religious development. The Bhīma cult is noteworthy for the features it derives from a confusion between Bhīma as a name of Siva and Bhīma, the second of the five Pāndava brothers, heroes of the Mahābhārata. [239] The name Sivabuddhaloka shows the very close connexion between Saivaism and Buddhism prevailing in Java unlike in India; indeed, an easy-going tolerance among the cults, and borrowing and blending of originally distinctive traits, mark the entire religious atmosphere of ancient Java. Kings are at times represented by images of Visnu bearing Saiva emblems—a feature which only reproduces the engraving of Saiva and Vaisnava symbols side by side above the Toek Mas inscription already noticed. In one instance, the conception of Ardhanārī is effectively adapted to the representation in one figure of Krtanāgara with his queen Bajradevī. [240]

The indebtedness of old Javanese literature and sculpture to Indian originals is well-known; the literary forms including metres, the literary dialect often called Kawi (though the term Old Javanese is to be preferred, to avoid confusion with the Javanese literary dialect of today) and the choice of subject-matter are all more or less completely Indian. Many books are direct translations, others adaptations to Javanese needs. The Mahābhārata, particularly the Ādi and Virāta parvas, is the main source of the new Javanese Wayang tales. The comparison of the available texts, however, leads to no definite results regarding the particular recensions employed by the Old Javanese translators; the present South Indian text of the epic differs from the Javanese, and the suggestion has been made that these differences are more modern than the times when the epic was taken over by Java. Ksemendra's summary of the epic and the Old Javanese Mahābhārata, both seem to be derived from a common text. [241] Likewise the Rāmāyana sculptures of Prambanam exhibit differences from the story as depicted in Vālmīki; Stutterheim, who has made a detailed study of the various versions of the Rāma legend prevalent in India and in the colonies, has not found it easy to trace the source of the Javanese version; he only reaches the rather vague conclusion that in this matter our eyes should not be turned exclusively on South India but Western and Eastern India also. [242]

The Javanese theatre and dance have attracted great attention and formed the subject of many studies, some of a controversial nature. Java and Bali know to this day many forms of the Wayang: shadow-play, mask-play, pantomime and something like the legitimate theatre where the actors speak and act, though even here the dalang (announcer) has not been quite eliminated. The main point of the discussion has been the extent to which the Wayang is Hindu or Indonesian in its origin. The Indonesian view had a strong vogue until in 1906 Pischel drew pointed attention to the chāyānātaka (shadow-play) of India; since then it has been recognized that Hindu influences must be allowed a considerable share not only in providing the subject-matter of the Lakons (stage versions of stories) but in the technique of the whole art. At the same time, in the form in which we now know it, there is no doubt that several features of a primitive Indonesian ancestor-worhip have taken a secure place in the apparatus of the Wayang. Two facts of undoubted significance to the history of the Wayang are: First, it is not found in Indonesia as a whole but exclusively in Java and Bali, the lands that came strongly under Hindu influences. [243] Secondly, the shadow-play comes into dominance in later Javanese culture to the extent to which it departs from Hinduist tradition. Both these facts go far to confirm the hypothesis originally formulated by Krom, and confirmed after detailed study by Rassers, that a Hindu shadow-theatre was first introduced into court and higher society in the early days of colonization, and that it then slowly penetrated popular culture and there became mixed up with the traditional rituals of ancestor-worship carried out by means of images of ancestors. The Wayang as we know it is thus neither Hindu nor Javanese but Hindu-Javanese. We cannot escape the impression that the Indian shadow-play is in Java rejuvenated and renewed. An evolution takes place which, as Rassers puts it, brings it nearer its source and presents it with unusual clarity in its full religious significance. [244]

There are two scales in Javanese music—the slendro with 5 tones in the octave and pelog with seven. The former is popular in Central Java, the other in the eastern and western parts of the island; slendro is connected with the Rāmāyana and Mahābhārata while pelog goes with Indonesian stories of the Pañji. The origin and relative antiquity of the two scales have been subjects of debate. Kunst holds that the slendro scale is younger and foreign in origin, probably an introduction of the Sailendras from Sumatra immediately, and ultimately from India. Stutterheim [245] points out as against this that the slendro is older as a scale of court music, just as Middle Javanese courts flourished much earlier than East Javanese. Tradition treats the slendro as a gift of Girinātha, the lord of the mountain, which is at once a name of Siva and a synonym of the name Sailendra; Stutterheim thinks the Saiva Sailendras of Matarām must be taken to be indicated by this tradition, and that consequently the scale was as indigenous in origin as the other scale. But the Sailendras were doubtless a race of Hindu-Javanese rulers and not without South Indian affiliations of their own. [246]

In the organization of rural economy and village administration, again, Java presents the same unmistakable blend between pre-Hindu Indonesian institutions and ideas and those borrowed from South India. The ideas of common property in some part of the village lands and of unrestricted individual property, including the right to sell or transfer it to persons who did not belong to the tribe, seem to have been developed by Indonesians themselves; [247] this is an inference based on modern observations of those parts of Indonesia which never came under Hindu influences as Java and Bali did, and must be accepted with some reserve when it is applied to Indonesian society of the early historical period. But Javanese tradition is quite clear that the existing organization of villages as more or less autonomous townships each with a separate individuality of its own, dates from the advent of the Hindus upon the island, and the institutions of village government are either unknown or quite different in their nature in the non-Hinduised parts of Indonesia. [248] The Hindu regard for desadharma, its tendency to treat the family rather than the individual as the unit of social system, the regard for the married man (grhastha) who had a secure position in the social order (only married men being full burghers in Java), and the grouping of four, eight, or ten adjacent villages into a larger local unit—all these features Java shared in common with India. And the proceedings at village meetings in Java even today strongly remind one of the conditions of village administration in South India in ancient days as it is vividly portrayed in the numberless inscriptions of the Cōla monarchs. This may be seen from the account [249] of such a meeting from the pen of a modern observer.

Our attention has been given mainly to the early cultural movements and the role of South India in their promotion. But this is only due to our initial intention to restrict the scope of our study to this aspect of the subject; influences from other parts of India flowed into Java and other lands in the East, though as I understand the matter, they were, on the whole, weaker and less persistent than those from South India. The discovery of a Gupta gold coin of Candragupta II near Batu Baka in Central Java, the early nāgari script and Mahāyāna Buddhism characteristic of the early Sailendra records of Java, which bring them into direct relation with Nālandā and the Pālas of Bengal, are sufficient testimony to the presence of other Indian influences in Java than those of South India.

We should not also overlook the continuous contact maintained by Java with South India in later times. The relations between the Sailendra empire of Srī Vijaya and the Cōla empire at the end of the tenth century and in the eleventh century form an important chapter of their history and Java was certainly not unaffected by them. The Nāgarakretāgama mentions that Buddhāditya, a bhiksu of Kāñcīpura, sang slokas in praise of the Javanese ruler Hayam Wuruk in the fourteenth century—a testimony to those renewed and fresh contacts with South India which students of Hindu-Javanese art history have found it necessary to postulate. [250] Jayanagara adopted the characteristic Pāndyan title Stindarapāndya at his coronation early in the fourteenth century and adopted the Pāndyan emblem of mīnadvaya (two carps) for his seal. [251] And there is literary evidence of an embassy from Malaya to Vijaynagar in the days of the celebrated Krsna Deva Rāya. [252]

XI.  Borneo

Large rivers are often the channels of commerce and colonization, and in Borneo, the Kapoeas, Barito and Mahakam rivers have promoted intercourse between the interior of this large island and the outside world. At Kutei at the mouth of the Mahakam, and at Moeara Kaman are found the oldest inscriptions of the Archipelago. Though not dated, these inscriptions may from their script be taken to belong to the end of the fourth century A.D.

They are engraved on four stone pillars, sacrificial posts (yūpas), and are in Sanskrit verse. The first mentions that the celebrated king Kundunga had a famous son, Asvavarman, the founder of a dynasty (vamsakartā); he had three sons, the best among them being Mūlavarman who performed the bahusuvarnaka sacrifice, commemorated by this yūpa set up by the best of Brahmins. The second post was set up by priests who had come there to receive gifts of twenty thousand kine [253] in the most sacred ksetra of Vaprakesvara. The third, also set up by priests, commemorates the great gifts of the same king including Kalpavrksa and land among other things. The fourth inscription is just a fragment; it compares Bhagīratha born of Sagara evidently to the son of Mūlavarman.

Besides these four pillars, three other similar inscribed yūpas were found in the same place in 1940 and they also record the gifts of Mūlavarman. The first of these inscriptions comprises two verses, an anustup and an āryā, and mentions gifts of jaladhenu, ghrtadhenu, kapilā, tila, and vrsabhaikādasa, meaning respectively water-cow, ghee-cow, tawny-cow, sesamum, and eleven bulls. Dhenu is explained by Monier-Williams as a gift in lieu of or in the shape of a cow. The second inscription, a single anustup verse records the gift of a tilaparvata (sesamum mountain) with a row of lamps (dīpamālā). The last is longer, comprising four verses in anustup, but has many gaps. It says that Mūlavarman conquered many kings in war and made them his vassals just as Yudhisthira had done; it also records the gifts at Vaprakesvara of 40,000 and 30,000 (gold coins?) and of Jīvadāna of different kinds (prthagvidham); it mentions an ākāsadīpa set up in the capital city and ends:

yūpoyam sthāpito viprair-nnānā (desād) ihā (gataih)

i. e. this yūpa (post) has been erected by Brahmins who came here from different countries. [254]

The Sanskrit language, the script and the contents of these inscriptions are fully Hindu, and decidedly South Indian. The twice-born (dvijendraih, vipraih), the sacrifices named, the kalpavrksa and all the other dānās, the vamsakartā, the genealogy, and the reference to the story of Bhagīratha and Sagara, are all typically Indian. But the name of the ksetra where the gift is made, Vaprakesvara, though Hindu in its appearance, is hard to explain. The term recurs as Baprakesvara in Javanese epigraphy.

The suggestion has been made that in V(B)aprakesvara we have to recognize an Indonesian institution in an Indian (Sanskrit) garb. Each Indonesian village had its own shrine in the form of an enclosed space (vapra) in the centre of which stood a wooden sacrificial post at which many an animal was slaughtered ceremonially. Possibly it was also a burial ground at which again animal sacrifices were common, the prototype of the well-known Candis of Java. [255] But all this sounds rather far-fetched and unconvincing. True, monumental yūpas are rare in India and those of Borneo differ considerably in size and form from the most typical of the yūpas so far known in India, e.g. those of Īsāpūr, a village on the bank of the Jumna opposite Muthurā. [256] But we cannot expect any close conformity in details between Mathurā of the first century A.D. and Borneo of the fourth, and we should remember also that yūpas of different sizes and shapes were prescribed for different sacrifices and by different schools of Srauta-sūtras. Let us also note that only one of the yūpas of Borneo is sacrificial (yajñasya yūpa), two others are donative and describe as tasya punyasya yūpoyam and tesām punyaganānām yūipoyam, and we lack the data for determining the nature of the fourth stone, the inscription on it being only legible in part.

There is absolutely no indication in the inscriptions of any Indonesian religious influences. They are all in correct Sanskrit, and fully devoted to yajña and dāna in which blue-blooded Brahmins who came from all the Hindu settlements (ihāgataih) in the neighbourhood took part. There is indeed the name Kundunga; he is described as a mahātmā and has Asvavarman for his son and Mūlavarman for grandson; yet his name sounds un-Indian. [257] But he too, as Kern pointed out long ago, [258] “must have been an adherent of Hinduism. Otherwise he would not have given his son an Indian name.” Clearly we have here to do with an indigenous royal family in the process of being Hinduised, and accepting the Hindu culture and religion just as they came to them from across the seas. But whether this culture was just a veneer adopted by the upper classes for its novelty and distinction, or permeated the details of everyday life and practice and soaked down to the common people, we have now no means of determining.

In any event V(B)aprakesvara is by no means an Indonesian term. The -īsvara ending is typically Indian, not to say South Indian, and Vapra figures among the twenty-eight Veda-vyāsas named in the Visnu Purāna and is represented by an image in the Brahmā temple at Prambanam. There is also reason to think that this sage was held to be identical with Agastya, so that the name of the ksetram in Borneo attests the role of Agastya, the mythical apostle of Hindu culture in South India and the Eastern lands. Some early names and forms of Hinduism which in the homeland became overlaid by later aspects might well have been preserved in their original form in the colonies, and Baprakesvara may well be considered one of them. This and some other names of a like formation, Pūtikesvara, Malankusesvara and so on, are known in Java also.

The provenance of the inscribed yūpa stones is not altogether clear; they have sometimes been stated to have been found at Kutei, but after considering all the evidence now available, Vogel came to the conclusion that Moeara Kaman, at the confluence of the Mahakam river with its tributary the Kaman, has the strongest claim to be considered the findspot of the yūpas. [259] He quotes H. von Dewall, who visited Moeara Kaman in 1847 and wrote: “Here are found a number of stone slabs piled up underground. It was on this spot also that the idol of massive gold, weighing 8 thail, was discovered, which the Sultan wears round his neck on State occasions. This image is four-armed, well proportioned and of good workmanship and seems to represent some god of Hindu mythology. The youthful prince, moreover, wears beneath this idol another golden, box-shaped object on which various mythological figures of the Hindu religion are shown in alto-relievo. The same appears to be of higher antiquity than the golden image, but, like the idol, it was discovered beneath the stone slabs in the reign of Sultan Muhammad Salih-ud-Dīn.” The golden statuette is a representation of Visnu; [260] a gold tortoise is also known to have come from the same place.

Hindu images of the Saiva pantheon have been found in various other localities in the island. A small-sized nandi and a linga come from the spot where the Rata river joins the Mahakam. [261] The caves of Mount Kombeng, another site in the Pantoen valley in Kutei, contain a collection of Hindu images which seem to have been stowed away here for safety at some time or other, and must originally have belonged to some temples in the neighbourhood. They have tenons underneath the base, indicating that they originally stood in some niches or receptacles. A Ganesa surely, and possibly a Visnu, a Brahmā, a nandi and a Kārttikeya seem to be among them. [262] They have not yet been fully and accurately described. Three of these images, all Saiva, are said to have been removed to the Batavia museum.

The exact age of most of these finds, other than the yūpa inscriptions, has not been studied and can hardly be settled without further exploration of the still virgin field of the archaeology of Borneo. Dr. Bosch [263] has pointed out that the style and grouping of the Saiva images of the Kombeng caves show clear Hindu-Javanese influences at work and concluded thence that it is improbable that Hindu colonists migrated directly to Borneo. But to say this is to fly in the face of the evidence of the yūpa inscriptions. It is best to suppose either that the Hindu colonies of Borneo in the later phases of their development lost touch with India and naturally fell more and more under influences of Hindu-Javanese origin, or that these images are the relics of a new and later Hindu-Javanese settlement in Borneo. There are some Buddhist images also among them, mostly female deities, which “show grave errors and misconceptions”; the stone-masons who fashioned them must have lost touch with the authentic tradition of their co-religionists elsewhere.

Buddhism is not otherwise unrepresented in the island. A bronze Buddha from Kota Bangoen is much earlier and nearer the true Buddhist tradition; it recalls the Amarāvatī style, though Hindu-Javanese influences also seem to be present. This image was destroyed by fire in the Paris exhibition of 1931. [264] In West Borneo in the Kapuas region and in Batoe-Pahat at the source of the Tekarek are found stūpas engraved on rocks with inscriptions on their sides in somewhat late Pallava script, containing the Buddhist ye-te formula and another verse ajñānāccīyate jñānam etc., nearly as well known. These verses also occur in the Kedah inscription of Nāvika Buddhagupta. There are other parts of these inscriptions of which the import is far from clear as yet. [265]

A Ganesa image from Sarawak, North Borneo, a linga and yoni in the upper Malawie in West Borneo, [266] and a Pallava inscription from near Sang-betrang on the cast coast, are other relics to be noted. There is also a mukhalinga of the sarvasama type in which the square Brahmabhāga (below), the octagonal Visnubhāga (middle), and the cylindrical Sivabhāgha (above) are of equal length; the linga comes from Sepaoek in the Sintang division of West Borneo. [267] We thus find unmistakable traces in different parts of Borneo of the settlements of Hindu colonists who had come directly from South India; they are most strikingly seen in the valleys of the Kapuas and Mahakam rivers, the relics of the Mahakam valley being among the earliest known and dating from about A.D. 400. These must be distinguished from the later monuments of a Hindu-Javanese character of the Majapahit period when Borneo was subject to strong cultural influences from Java. Such influences might have come from Java also at an earlier time, say, in the second half of the tenth century under Dharmavamsa who adopted a policy of active expansion of the Javanese state. Of the later history of the Hindu colonies of Borneo we know nothing at present.

XII.  The Philippines

The Philippines must be held on the whole to have remained outside the range of the early cultural movements with which we are concerned. Two images of deities are the only known Hindu antiquities from these islands. One is a solid gold image of a goddess found accidentally after a storm on the left bank of the Wawa river in Mindanao. The gold is of 21 carats, and the image which is 15.2 cm. high and 9 cm. broad from knee to knee, weighs 1791.5 grammes and is estimated at $ 1003.15. The area from which the image comes was before A.D. 1500 under a chief known as the Rāja of Butuan. The image is of fine workmanship and shows clear evidence in its tall pointed head-dress and other ornaments of the influence of Hindu-Javanese art of the tenth century A.D. It is not easy to decide if the goddess belongs to the Hindu or Buddhist pantheon. The other is a copper image, 8 cm. high, found on the island of Sibu in 1820 most probably of Siva. [268] The extent of the influence exerted by the Hindu civilization of Java and Sumatra even in late historical times on the Philippines is not beyond dispute; while American scholars working in the Philippines are inclined to rate it rather high and to derive several features of Philippine culture from the colonies established by Srī Vijaya on these islands, Krom is somewhat sceptical of the far-reaching inferences drawn from the name Visaya current in the Philippines and in Borneo. The subject requires far more detailed study before any final judgement can be formed. [269]

ABBREVIATIONS

AA           Ars Asiatica

ABIA        Annual Bibliography of Indian Archaeology (Kern Institute, Leyden)

ARB         Archaeological Reports, Burma

ASI          Archaeological Survey of India, Annual Report

BCAIC      Bulletin de la Commission Archaeologique de l’Indo-Chine

BEFE0      Bulletin de l’Ecole Française d'Extreme-Orient

BKI          Bijdragen tot de Taal¾, Land en Volkenkunde Van Nederlandsch Indie

EI            Epigraphia Indica

Et. As.     Etudes Asiatiques, 2 vols. (Hanoi)

IAL          Indian Art and Letters

IHQ          Indian Historical Quarterly

ISC(S)C   Inscriptions Sanstrites du Campā et Cambodge

JA            Journal Asiatique

JGIS        Journal of the Greater India Society, Calcutta

JOR         Journal of Oriental Research, Madras

J(P)ASB   Journal (and Proceedings) of the (Royal) Asiatic Society of Bengal

JRAS        Journal of the Royal Asiatic Society

0V           Oudheidkundige Verslag, Batavia

QJMS       Quarterly Journal of the Mythic Society

TBG         Tijdschrift voor Indische Taal¾, Land en Volkenkunde, Bataviaasch Genootschap

TP         T'oung Pao