THE BOOK WAS DRENCHED CO u< OU_158628>5 CO itp library. . Collection Accession No.. Call No.... Q. ANCIENT INDIAN COLONIES IN THE FAR EAST VOL. II SUVARNADVIPA PART I Political History. BY Dr. R. C. Majumdar, M. A., Ph. D., Professor, Dacca University, Author of Corporate Life in Ancient India, Outline of Ancient Indian History and Civilisation, Qurjara-Pratlharas, Early History of Bengal, The Arab Invasion of India, etc. DACCA 1937 Published ? Asoke Kumar Majumdar, Ramna, Dacca. All Rights Reserved 1937 To be had of the following book-sellers : 1. Chuckervertty Chatterji & Co. 15, College Square, Calcutta. 2. Book Company 4/4A, College Square, Calcutta. 3. Punjab Sanskrit Book Depot. Saidmitha Street, Lahore. 4. Greater India Society 21, Badnr Pagan Eoiv, Calcutta. 5. Asutosh Library 5, College Square, Calcutta, or Patuatuly, Dacca (Bengal). PRINTER TRA1LOKYA CHANDRA SUR A8UI08H PKJS88, DACCA To The Dutch Savants whose labours have unfolded a new and glorious chapter of the History of Ancient Culture and Civilisation of India this volume is dedicated in token of the respect, admiration, and gratitude of the author. PREFACE The first volume of Ancient Indian Colonies in the Far East, dealing with the colony of Champa, was published in 1927. Various causes have delayed the publication of the second volume. One of them is a change in the planning of the different volumes. Originally I had intended to deal with the history of Kamboja (Cambodia) in the second volume. As the wonderful monuments of this kingdom were to constitute an important part of the volume, I paid a visit to Cambodia in order to obtain a first-hand knowledge of them. There, in my conversation with the Archaeological authorities, I came to learn for the first time that many novel theories were being advanced regarding the age and chronological sequence of the monuments of Angkor Thorn. I was advised to put off the publication of my book until these had been fully explored. Acting upon this advice I took up the history of Malayasia which was to have formed the third volume. My knowledge of Dutch being very poor at the time, I had to spend a long time in mastering the contents of relevant books and Journals which are mostly written in that language. Hence it has taken me nearly nine years to prepare and bring this volume before the public. The interval between the first and the second volume has further been prolonged by several urgent pre-occupations. It is needless to dilate on the difficulty of working on the subject in India, without any possible help or advice from any competent authority, and without any adequate library. It would be hardly an exaggeration to say that the small collection of books on the subject, which I have patiently acquired for Dacca University during the last seven years, is the best in India, but it is still very far from being adequate or satisfactory. The study of the Indian Colonisation in the Far East is still at its very infancy in this country. The Greater India Society and its Journal are notable recent PREFACE. enterprises, but the establishment of a Central Institute with facilities for the study of the subject is still a great desideratum. At the time when I took up the task of writing a series of studies on Indian Colonisation, the Society had not yet come into existence, and there was no book, big or small, on the subject in English language. As regards Java, the remark still holds good, save for a small pamphlet published by the Society, and a book on Indo-Javanese literature, published by one of my pupils after the first draft of this book was ready. I state these facts, not with the motive of claiming any special credit, but with a view to craving the indulgence of the readers for the many shortcomings which will be found in this pioneer work. When the book was completed, it proved too bulky for one volume, and hence I thought it advisable to divide it into two parts. The first part, now published, deals with the political history and the system of administration. The second part, now in press, deals with law, society, art, religion, literature, and the economic condition of Suvarnadvlpa. I have experienced considerable difficulty in the spelling of proper names. As regards the Javanese names of persons and places, I have followed the Dutch spelling, substituting y, ch y and ^^ respectively, for dj\ ij, and oe. I have also used y and v respectively for j and ?/>, except where these occur at the beginning of a word. The modern Javanese personal names are spelt exactly as in Dutch. As regards the Chinese names, I have followed the English, French, and Dutch spellings, according to the source from which I derived my knowledge of them. Originally I intended to insert in this volume a complete collection of j Javanese inscriptions on the lines followed in Volume I. But while this volume was in progress, my pupil Mr. Himansu Bhusan Sarkar, M. A., a research-scholar working under me, took up this work, and has now practically completed it. I hope his 'Collection of Javanese Inscriptions' will shortly be published, and hence I do not think it necessary to add a third part dealing with the Javanese inscriptions. As at present planned, the Second Part of this volume, referred to above, will be published before the end of 1937. The Third Volume, dealing with Kamboja (Cambodia and Siam), will be published in two separate parts, one containing the history, and the other the collection of inscriptions. I hope these will be out before the end of 1939. Volume IV, forming the sixth book of the series, and containing a general review of Ancient Indian Colonies in the Far East, will, I hope, be published by 1941. The task of writing these volumes has been a painful and laborious one, particularly as I have to work, for the most part, in a remote Mofussil town, under heavy pressure of administrative and other duties. I can only crave the indulgence of my generous readers for the many errors which must necessarily have crept into this book. My sole excuse for the choice of this difficult undertaking is the general apathy and ignorance in this country about this important branch of study. If I succeed in removing them even to a small extent, I shall consider my labours amply rewarded. Ramna, Dacca. 1 > R. C. MAJUMDAK The 7th of December, 1936. ) TABLE OF CONTENTS. Page Introduction ... ... ... i Abbreviations ... ... ... x i Additions and Corrections ... ... xvii Maps 1. Malayasia ... \ f . 2. Central and Eastern Java J lacln P' L Book L The Dawn of Hindu Colonisation. Chapter L The Land ... ... ... 1 II. The People ... ... ... 9 III. Prc-Hindu Civilisation in Malayasia ... 26 IV. Suvarnadvlpa ... ... ... 37 V. Early Hindu Colonisation in Malay Peninsula 65 VI. Early Hindu Colonisation in Java ... 91 VI T. Early Hindu Colonisation in Sumatra ... 116 VIII. Eurly Hindu Colonisation in Borneo ... 125 IX. Early Hindu Colonisation in Biili ... 132 X. Hindu Civilisation in Suvarnadvlpa up to the end of the Seventh Century A.D. ... 138 Book II. The Sailendra Empire. I. The Sailcndra Empire ( up to the end of the Tenth Century A.D.) ... ... 149 II. The Struggle between the Sailendras and the Colas ... ^ ... ... 167 III. Decline and Fall of the Sailendra Empire 191 Appendix ... ... ... 204 ii 11 Book III. Rise and fall of the Indo-Javanese Empire. Chapter Page I. The Kingdom of Matarfim ... 229 II. Else of Eastern Java ... ... 255 III. The Kingdom of Kadiri ... ... 276 IV. The Dynasty of Singhasari ... ... 292 V. The Foundation of Majapahit ... 308 VI. The Javanese Empire ... ... 319 VII. Downfall of the Empire ... ... 339 VIII. Sunda ... ... ... 356 Book IV. Downfall of Hindu Kingdoms in Suvarnadvipa. I. End of Hindu Rule in Sumatra ... 363 II. End of Hindu Rule in Malay Peninsula ... 378 III. End of Hindu Rule in Java ... ... 401 IV. End of Hindu Rule in Borneo ... 412 V. The Bali Island ... ... ... 419 VI. Political theory and public administration in Java 429 INTRODUCTION I propose to deal in this volume with the Hindu colonisation in Malay Peninsula and the Malay Archipelago. For this entire region, now known as Malayasia, I have used the name Suvarnadvipa. My authority for the use of this Indian name in this wide sense is set forth in Chapter IV. In this volume I have followed the same plan as was adopted in the case of the earlier volume on Champa. I have tried to bring together such information as we possess of the political history of the different regions constituting Suvarnadvlpa, and have also dealt with the various aspects of civilisation of their people, viz., religion, literature, law and administration, social and economic conditions, and art. I have not discussed such general themes as the nature of Indian civilisation, the influence of the Pallavas or of South India on the civilisation of Sumatra and Java, the origin of art and alphabet of these regions, and similar other questions which are pertinent to the subject. These will be discussed in a subsequent volume. Although Suvarnadvlpa is a mere geographical expression and a congeries of states, it came to be on two occasions, at least, almost a political entity. First, under the Sailendra kings from the end of the eighth to the beginning of the eleventh century A.D., and, secondly, in the palmy days of the Empire of Majapahit. Even in other periods, there has almost always been a close political relationship, be it friendly or hostile, between its constituent parts, such as we do not meet with between any of them and the outside world. Even now the predominance of the Malay-speaking people all over the area serves as a bond of unity, which is also artificially maintained to a large extent by common subjection to the Dutch. Those considerations would be a further justification of the choice of Suvarnadvipa as a historical unit, 11 INTRODUCTION Our knowledge regarding the Hindu colonies in the various small islands which dot the Pacific is very meagre, and this volume primarily deals with the Indian colonies settled in the Malay Peninsula, Sumatra, Java, Bali, and Borneo. The sources of information on which the accounts are based will be found in detail in the body of the book, but it may be convenient to give a general idea of them at the very outset. The sources may be broadly divided into two classes, indigenous and foreign. Among indigenous sources, again, the two most important sub-divisions are (1) archaeological, and (2) literary. The archaeological evidence consists mainly of inscriptions and monuments, as coins play but little part in unfolding the history of these countries. As regards inscriptions and monuments, Java offers the richest field, and those in tho other regions are far inferior both in quality and quantity. The Sanskrit inscriptions of Java were studied by Kern, and may now be conveniently consulted in his collected works (Kern V. G.). The Kavi inscriptions have been collected in two works by Cohen Stuart (K. O.) and Dr. Brandes (O. J. O.). Other inscriptions have been noticed or edited in the publications of the Dutch Archaeological Department, particularly in O. V. The monuments of Java are principally described in three series of archaeological publications, m%^ (1) Rapporten (2) O. V. and (3) Arch. Ond. The last named series really consists of three monumental works on Candi Jago, Candi Singasari, and Barabudur. While one volume is devoted to each of the first two, that on Barabudur consists of five big volumes. Two of these contain only plates, and of the three volumes of texts, two give the archaeological, and one, the architectural description of the great monument. It may be noted that the two volumes on archaeological description have been translated into English, INTRODUCTION Ul As regards the island of Bali, we have a collection of inscriptions in Epigraphia Balica, Vol. I, by P. V, Stein Callenfells. The results of more recent archaeological investiga- tions are given by Stutterhcim in 'Oudheden Van Bali 7 . The monuments of Sumatra and Borneo, which are in Dutch possession, have been described in O. V. For those of Malay Peninsula we have got a preliminary account by M. Lajonquierre in B. C. A. I, 1909 and 1912. As regards the literary sources of history, there are two works in Java which may claim the highest rank : The first is Nagara-Krtiigama, a poem written during the reign of Hayam Wuruk, by Prapanca, who held the high office of the Superintendent of the Buddhist Church in the court of that king. It was composed in 1365 A.D., and, although primarily concerned with the career of the king, gives other historical informations of high value. It has been translated by Kern (V. G., Vols. VII, VIII) and re-published by Krom. The second is a prose work called Pararaton. It is a sort of historical chronicle beginning with the life of Ken Angrok, and continuing the history of Java down to the end of the Hindu rule. It gives dates for most of the events, but these have not always proved to be correct. The book has no doubt a genuine historical background, but the incidents mentioned in it cannot always be regarded as historical without further corroboration. The book was originally edited and translated by Brandes (Par.), and a revised edition has been published by Krom. There are other modern historical works in Java arid Bali, called Kidung, Babads, and Sajara which have preserved traditions regarding their ancient history. These have been referred to in detail in the chapter on Literature, Similar works exist in Malay Peninsula, e.g., Sajarah Malayu. Besides historical works, Java and Bali are rich in literature of all kinds to which a detailed reference will be found in the chapter on Literature. IV INTRODUCTION A very large part of this literature still exists in manuscripts alone, but a few important texts have been ably edited, some with a Dutch translation. There are very learned and comprehensive catalogues of Javanese manuscripts by Vreede, Brandes, and Juynboll. Among the published texts may be mentioned, Rainayana, Mahabh&rata (portions only), Bharata- yuddha, Arjunavivaha, Kunjarakarna, VrttasaScaya, Bhoma- kavya, Galon Arang, Tantri Kamandaka, Megantaka, Dreman, Lingga Peta, Nitisara, and various Kidung works, in addition to several religious texts and one law-book. The former include Sang hyang KamahaySnikan, a MahaySnist text, and Agastya Parva, Brahmanda Purana, and Tantu Panggelaran, all works of the nature of Purana, containing theology, cosmogony and mythology. The law-book is Kutara-manava, edited with notes and translation by Jonkcr. A fuller account of these will be found in the chapters on Literature and Religion. The foreign sources may be subdivided into two classes, the eastern and the western. To the former category belong the Chinese, and to the latter, the Indian, Greek, Latin, and Arabic texts. The Indian, Greek, and Latin sources contain stray references to Malayasia and its constituent parts, and occasionally, as in the case of Ptolemy's Geography and Marco Polo's accounts, some valuable geographical information. The Arab texts, consisting principally of travellers' accounts, arc also very valuable for a knowledge of the trade and commercial geography of the whole region. But these western sources do not offer much material for reconstructing the history of Malayasia. For this we have to turn to the Chinese texts which contain very valuable data for the political and cultural history of the entire region. The Chinese possessed special opportunities for obtaining first-hand informations about the different regions of Malayasia, as these had diplomatic and trade relations with China. The envoys from these lands to the imperial court f and th$ accounts of the Chinese ambassadors who visited them, must have furnished excellent materials to the official Chroniclers who incorporated accounts of these foreign lands in the histories of the Imperial dynasties. A number of Chinese travellers also visited these far-off lands and recorded short accounts of the countries visited by them. The traders from these lands also imparted valuable information to Chinese officials. Thus the Chinese annals possess a store of information about Malayasia, which in quality and quantity far exceed, in importance, what we know from other foreign sources. In view of this, and as frequent references have been made to these Chinese sources in the text, we give here a short account of the Chinese texts on which we have principally relied. First, we have the famous Dynastic Histories. As is wellknown, there are twenty-four official Histories which deal with the history of China from the earliest time up to the end of the Ming dynasty (1643 A.D.). The first book, Che-ki, deals with the history of the country from the earliest time up to 122 B. C. The other books deal separately with the history of every dynasty which has since reigned in China. The history of each dynasty was written after its downfall with the help of the Government archives. It contains accounts of foreign countries "which have always been drawn up from the materials at hand, and may therefore be considered to refer to the time when the dynasty still existed, even if the time of their compilation and publication falls considerably later *". The following is a list of the Dynastic Histories, principally referred to in this book. The date, given within brackets, refers to the period covered by each. 1. History of the First Sung Dynasty (420-478 A. D.) 2. History of the Liang Dynasty (502-556 A. D.) 3. Old History of the T'ang Dynasty (618-906 A. D.) I, Groeneveldt Notes, p. VII. VI INTRODUCTION 4. New History of the Tang Dynasty (618-906 A. D.) E. History of the Sung Dynasty (960-1279 A. D.) 6. History of the Yuan (Mongol) Dynasty (1206-1367 A.D.) 7. History of the Ming Dynasty (1368-1643 A. D.) Among the non-official accounts, those of Fa-hien and I-tsing (Record, Memoire) belong respectively to the fifth and seventh centuries A. D. After a long interval we come across regular accounts from the twelfth century onwards. These are enumerated below with brief notes. 1. Ling-wai-tai-ta, by Chou kii-fei, Assistant Sub-Prefect in Kui-lin, the capital of Kuang-si. It was composed in 1178 A. D. 2. Chu-fan-chi by Chau Ju-kua, Inspector of Foreign Trade in Fu-kicn. The date of this work has been discussed on p. 193. The author had special facilities for obtaining information on the subjects treated by him from the foreign sailors and traders who frequented his port. Though he has relied on Liug-wai-tai-ta for several sections of his work, those dealing with San-fo-tsi and its subordinate states (which alone are mainly used in this book) seem to be based exclusively on the information gathered by him from Chinese and foreign traders 1 . 3. Tao-i Chih-lio or "Description of the Barbarians of the Isles" by Wang Ta-yuan with the cognomen of Huan-Chang. He visited, for purposes of trade, a considerable number of foreign localities during the period 1341-1367 and recorded what he had seen in this work. It is a personal and, conse- quently, trustworthy record. There are two dates in the work from which we may conclude that the author was already travelling in 1330, and that he probably put the last touches to his work after the summer of 1349. 4-5. Ying-yai Sheng-lan by Ma Huan and Hsing-Cha Sheng- lan by Fei Hsin. Both Ma Huan and Fei Hsin accompanied I. Chau Ju-kua, pp, 22, 36. INTRODUCTION Vll the famous eunuch Cheng Ho in some of his voyages. These voyages were undertaken at the command of the Emperor with a view to exploring foreign lands for commercial purposes and demonstrating to them the might and prestige of the Chinese Empire 1 . Some idea of these voyages may be obtained from the fact that in one of them Cheng Ho is said to have taken forty-eight vessels and 27,000 Imperial troops with him. Cheng Ho made altogether seven voyages between 1405 and 1433 AJX, and visited thirty-six (or thirty-seven) countries, in Malayasia, India, Arabia, and Africa. Both Ma Huan and Fei Hsin must have gathered materials for their work from the voyages they undertook. Ma Huan was attached to the suite of Cheng Ho as "Interpreter of foreign languages and writing to the mission ." Fei Hsin was 'presumably a secretary or clerk'. Both of them had thus splendid opportunities of gaining first-hand knowledge about these foreign lands, and this invests their chronicles with a special importance. The original text of Ma Huan was revised by Chang Sheng, and Rockhill has made a confusion between the original and the revised text. The whole matter has, however, been clearly set forth by Pel Hot. Rockhill assigned the first publication of Ma Huan's work to a date between 1425 and 1432 A. D. Pelliot is, however, of opinion that the first edition of the work really appeared in 1416, the date given in the preface to the work, soon after Ma Huan's first voyages in 1413-15 A. D. The work was evidently enlarged after the two subsequent visits in 1421-2 and 1431-3, and completed about 1433. But the book probably appeared in its final form only in 1451 A.D.* It is not necessary to refer in detail to the other Chinese works to which occasional reference has been made in the following pages. 1. For full discussion on this point cf. T'oung Pao. 1934, pp. 303 ft. 2. Cf. T'oung Pao, 1933, pp. 236 ff. iii viii INTRODUCTION Excepting the Indian texts, it has not been possible for me to consult the other sources in original. Fortunately, reliable translations by able and competent scholars are available for most of them. The Greek and Latin texts have been translated by Ccedfcs ( Ccedes-Textcs ), and the Arab Texts by Ferrand ( Ferrand-Textes ), both into French. For the Chinese sources the following deserve special mention. I. Translation. 1. Translation of Fa-hien's account by Legge. 2. Translation of I-tsing's works by Takakusu and Chavanncs ( I-tsing-Record, I-tsing-Memoire ). 3. Translation of Chan Ju-kua's work by Hirth and Rockhill ( Chau Ju-kua ). II. Translation of Extracts. 4. W. P. Groeneveldt Notes on the Malay Archipelago and Malacca ( Batavia 1877 ). [ Supplementary Jottings T'oung Pao, Scr. I, Vol. VII, pp. 113 ff. ]. 5. W. W. Rockhill Notes on the relations and trade of China with the Eastern Archipelago and the coasts of the Indian ocean during the Fourteenth Century. T'oung Pao, Serie II, Vol. XVI (1915), pp. 61 ff., 236 ff., 374 ff., 435 ff., 604 ff. III. Critical Discussion. 6. P. Pelliot Deux Itineraires de Chine en Indio & la fin du VIII 8 sifccle ( B.E.F.E.O., Vol. IV, pp. 132-413 ) 7. Schlegel Geographical Notes. T'oung Pao, Ser. I, Vol. IX. ( pp. 177 ff, 191 ff, 273 ff, 365 ff ) ; Vol. X ( pp. 33 ff, 155 ff, 247 ff, 459 ff ) ; Ser. II, Vol. II ( pp. 107 ff, 167 ff, 329 ff. ) INTRODUCTION IX 8. J. J. L. Duyvendak Ma Huan re-examined ( Verhand. der. Kon. Ak. van Wetensch., Afd. Lettcrkunde, N. R M d. XXXII, no 3, Amsterdam, 1933. ) 9. P. Pelliot Lcs Grands Voyages Maritimes Chinois au Debut du XV e sifecle ( T'oung Pao, Vol. XXX, 1933, pp. 236-452 ; Vol. XXI, pp. 274 ff. ) [ This is by way of review of the preceding book. ] Before leaving this subject we must also mention the works of Ferrand who has collected all the sources of information about Sri-Vijaya and Malayu-Malakka in two articles in Journal Asiatique ( J.A. II, XX ; J.A. II, XI-XII ). Of the modern historical works dealing with the subject, those by Raffles, Fruin-Mees, With, and Veth have all been cast into shade by Kroni's 4 Hindoe-Javaansche Geschiedenis' which is bound to remain the standard work on the political history of Java for many years to come. Krom's other work, Inleiding tot de Hindoe-Javaansche Kunst, is equally valuable for the history of Javanese art. It is with pleasure and gratitude that I recall the fact that these two books formed the foundation of my study of Javanese history, and I have freely utilised them in the following pages. The second edition of the first named work reached my hands after the first draft of this book was composed. Although I have utilised the new edition in the revision of my book, references given are mostly to the first edition. For the religious history of Java the works of Goris and Pigeaud (Tantu), and the numerous articles by various scholars, have been of the greatest assistance to me as they are sure to prove to others. As regards Literature, the Catalogues of Manuscripts ( Cat. I, II, III ), and the works by Berg (specially Hoofdlijrien, Mid. Jav. Trad, and Inleiding), Pandji Roman by Rassers, and several articles, notably the one by Berg in B. K. I., Vol. 71 ( pp. 556-578 ), have been most useful to me. Not being acquainted with the Kavi language, I had to derive my X INTRODUCTION knowledge of Javanese literature mainly from these and the few translations of texts that have been published so far. Of the secondary sources for the history and civilisation of Malayasia, by far the most important are the learned articles contributed to T. B. G. and B. K. I, the organs of the two famous institutions that have done yeoman's work in rescuing from oblivion the glorious past of Java and the neighbouring islands. These articles touch upon every aspect of the subject and are of inestimable value to anyone who seeks to study the history of Indonesia. It will be seen from the above that our data regarding the history of the different regions is very unequal. While we possess, in an abundant degree, evidences, both literary and archaeological, for the history and civilisation of Java, these are very meagre when we come, for example, to Borneo. Between these two extremes we may place, in order of adequacy of historical materials, Bali, Malay Peninsula, and Sumatra. It would be wrong, however, to imagine that the degree of importance, which should be attached to the different regions, is in any way proportionate to the extant evidences regarding them. The absence of evidence available to us may be quite accidental. The archaeological evidence is mostly perishable, save in the case of massive monuments such as we find in Java. As to the Chinese evidence, the Chroniclers could only record events when there was any intercourse with one of these states. A state would come in and go out of their history according as it sent any embassy to China ( or rice versa ) or ceased to 'do so. The absence of evidence, therefore, should not lead us to infer the political insignificance of a state, far less its non-existence, ABBREVIATIONS A. B. I. A. = Annual Bibliography of Indian Archaeology (Published by Kern Institute, Leyden). Ann. Rep. Arch. Surv.= Annual Report of the Archaeological Survey ot India. Arch. Ond.=Archaeologisch Onderzock op Java en Madura (By the Commission appointed in 1901), 3 volumes dealing respectively with Tjandi Djago, Tjandi Singasari, and Barabudur. Arch. Surv.= Archaeological Survey Report (Provincial). B. C. A. I.=Bulletin dc la commission Archaeologique de Wndochine. Beal= Buddhist Records of the Western World, Translated from the Chinese of Hiuen Tsiang by Samuel Beal (London, 1906). B. E. F. E. O.=Bulletin de 1'Ecole Frangaisc d'Extr&me-Oricnt, (Hanoi). Berg-Hoof cllijnen = Hoof dlij nen der Javaansche Litteratuur- Geschiedenis by C. C. Berg (1929) Berg-Inleiding=Inleiding tot de studio van het Oud-Javaansch (1928) Bib-Jav=Bibliotheca Javanica B. K. I^=Bijdragen tot dc taal-, land- en Volkenkunde van Nederlandsch-Indie, uitgegcven door het Koninklijk Instituut voor de Taal-, land- en Volkenkunde van Nederlandsch Indie Cat. L= Supplement op den Catalogus van de Javaansche en Madoerecsche Handschriften der Leidsche Univer- siteits-Bibliotheek by Dr. H. H. Juynboll. Leiden, VoLI (1907), Vol. II (1911), x ABBREVIATIONS Cat. EE. - Supplement op den Catalogus van de Sundaneesche Handschriftcn en Catalogus van Balineesche en Sasaksche Handschriften der Leidsche Universi- teits-Bibliotheek by Dr. H. H. Juynboll (1912). Cat. III. = Catalogus van de Maleische en Sundaneesche Hand- schriftcn der Leidsche Universiteits-Bibliotheek by Dr. H. H. Juynboll (1890). Cat. IV.=Juynboll-Catalogus van 's Rijks Ethnographisch Museum. Chau Ju-kua=Chu-fan-chi* by Chau Ju-kua. Translated by P. Hirth and W. W. Eockhill. St. Petersburgh (1911). Coedes-Textes=Textes d'auteurs Grecs et Latins relatifs a 1' Extreme-Orient depuis le IV e stecle Av. J. C. jusqu'au XIV sifecle. Kecueillis et traduits par George Coedfes (Paris-Ernest Leroux, 1910). Cohn-Ind.=Indische Plastik von William Cohn (Berlin, 1923) Congres I=Handelingen van het eerste Congres voor de taal-, land- en volkenkunde van Java, 1919 (Albrecht & Co-Weltevreden). Coomaraswamy=Ananda K. Coomaraswamy History of Indian and Indonesian Art (1927). Crawfurd-Dictionary=A Descriptive Dictionary of the Indian Islands and Adjacent Countries by John Crawfurd, London (1856) Encycl. Ned. Ind.= Encyclopaedic van Nederlandsch-Indie, Second Edition (1919). Ep. Carn.=Epigraphia Carnatica. Ep. Ind.=Epigraphia Indica. Fa-hien=A record of Buddhistic Kingdoms by Fa-hien. Tran- slated by J. Legge (Oxford, 1886). Feestbundel=Feestbundel uitgegeven door Koninklijk Bataviaa- sch Genootschap van Kunsten en Wetenschappen bei gelegenheid van zijn 150-jarig Bestaan 1778- 1928 (G.KolflFA Co, 1929), Ferrajid-Texte8=r Relations de Voyages etTextes.Geographiques Arabes, Pereans et Turks relatifs a PExtrSme- Orient du VHI* au XVHP sifecles by Gabriel Ferrand, 2 Vols (Paris-Ernest Leroux 1913-14). Foucher-Etude, I=tude sur 1'Iconographie Bouddhique de l ; Inde by A. Foucher (Paris, 1900). Foucher-Etude, II.=Do-Paris, 1905 Friederich-Bali=An account of the Island of Bali by Dr. R. Friederich (Miscellaneous Papers relating to Indo- China and the Indian Archipelago, Second Series, Vol. II, London, 1887). Fruin-MeesGeschiedenis van Java by W. Fruin-Mees. Part I (2nd Edition, Weltevreden, 1922). Gerini-Researches= Researches on Ptolemy 's Geography of Eastern Asia by Colonel G. E. Gerini, London (1909). Goris=Bijdrage Tot de kennis der Oud-Javaansche en Balincesche Theologie by R. Goris. Leiden, 1926. Groenevcldt-Notes= Notes on the Malay Archipelago and Malacca compiled from Chinese Sources by W. P. Grocneveldt. V. B. G. Vol. XXXIX, Part I. (Batavia, 1877) I. A. L.= Indian Art and Letters. I. C.== Indian Culture (Calcutta). I. H. Q.=Indiaii Historical Quarterly (Calcutta). Ind. Ant, = Indian Antiquary. Indian Art=The Influences of Indian Art. Published by the India Society 1925. I-tsing-Memoire=Memoire compost a 1'epoque de la grande dynastic T'ang stir les Religieux Eminents qui allerent cherchcr la loi dans les pays d'occident par I-tsing. Translated by E. Chavannes (Paris 1894). XIV ABBREVIATIONS I-tsing-Record=A Record of the Buddhist religion ad practised in India and the Malay Archipelago (A.D. 671-695) by I-tsing. Translated by J. Takakusu (Oxford, 1896). J. A.=Journal Asiatique. J. A. S. B.=Journal of the Asiatic Society of Bengal (1865- 1904). J. A. S. B. N. S.-Do, New Series (1905-1934). J. A. S. B. L.=Do (Letters, from 1935). J. Bo. Br. R. A. S.= Journal of the Bombay Branch of the Royal Asiatic Society. J. B. O. R. S.= Journal of the Bihar and Orissa Research Society. J. F. M. S. M.= Journal of the Federated Malay States Museum. J. G. I. 8.= Journal of the Greater India Society. J. I. H.= Journal of Indian History. J. Mai. Br. R. A. S.= Journal of the Malay Branch of the Royal Asiatic Society. Jonker Wetboek=Een Oud-Javaansch Wetboek vergeleken met Indischc Rcchtsbronnen by J. C. G. Jonker (Leiden, 1885). J. R. A. S. Journal of the Royal Asiatic Society of Great Britain and Ireland. J. Str. Br. R. A. S.= Journal of the Straits Branch of the Royal Asiatic Society. Kernpers=The Bronzes of Naltmda and Hindu Javanese Art by Dr. A. J. Bernct Kempers (Leiden) (Originally published as an article in B. K. I. Vol. 90, pp. 1-88). K. O.=Kawi Oorkonden'in Facsimile Met Inleiding en Trans- criptie by Dr. A. B. Cohen Stuart (Leiden, 1875). Krom Geschiedenis=Hindoe-Javaansche Geschiedenis by Dr. N. J. Krom (Martinus Nijhoff, Hague, 1926). The second Edition ( 1931 ) is indicated by Krom- Geschiedenis 1 . ABBREVIATIONS xv Krom-Kunst=Inleiding tot de Hindoe-Javaansche Kunst by Dr. N. J. Krom (Martinus Nijhoff, Hague, 1923). Levi-Texts= Sanskrit Texts from Bali by Sylvain Levi (Gaekwad Oriental Series). Mid. Jav. Trad.==Bcrg De Middeljavaansche Historische Traditie (1927). Nag. Kr.=Nagara-Krtagama Edited by H. Kern (V. G. Vols. VII-VIII). N. I. O. N.=Nederlandsch Indie, Oud en Nieuw Not. Bat. Gen.=Notulen van dc Algemecne en Bestuurs- vergaderingen van het Bataviaasch Genootsehap van Kunstcn en Wctcnschappen. O. B.=Oudheden van Bali by Dr. W. F. Stuttcrheim (Singradja, 1929). O. J. O.=Oud-Javaansche Oorkonden. Nagelaten Transcrip- ties van Wijlen Dr. J. L. A. Brandes. Uitgegeven door Dr. N. J. Krom ; V. B. G. Vol. LX. (Batavia and the Hague, 1913.) O. V.==Oudhcidkundig Vcrslag (Rapporten van den Oudheid- kundig Dienst in Nederlandsch Indie, Series I, 1912-1919 ; Scries II. 1920 etc.) Par=Pararaton of Hot Bock der Koningen van Tumapcl en van Majapahit door Brandes ; Tweede Druk door Dr. N. J. Krom (V. B. G. Deel LXII), 1920. Poerbatjaraka-Agastya=Agastya in den Archipel by Poerba- tjaraka (Lesya) (Leiden, 1926). Raffles- Java The History of Java by Sir Thomas Stamford Raffles, 2nd Edition (London, 1830). Rapporten=Rapporten vande Commissiein Nederlandsch-Indie voor Oudheidkundig Ondcrzoek op Java en Madura, 1901 etc. Rum-Serams=Dc Rum Serams op Nieuw-Guinea of Het Hinduismc in het Oosten van onzen Archipel door Dr. D. W. Horst (Leiden, 1893). iv ABBBEVIATIONS Sarkar-Literature= Indian Influences on the Literature of Java and Bali, Calcutta, 1934. Sastri-Colas='The Colas' by Prof. K. A. Nilakanta Sastri, Madras, 1935. S. I. Ep. Rep. = Annual Report on South Indian Epigraphy. S. I. L= South Indian Inscriptions. Tantu=De Tantu Panggelaran by Th. Pigeaud (Hague, 1924). T. B. G.=Tijdschrift voor Indische Taal-, Land- en Volken- kunde van het Bataviaasch Genootschap van Kunsten en Wetenschappen (1853 cfr.), Batavia. V. B. G.=Verhandelingen van het Bataviaasch Genootschap van Kunsten en Wetenschappen, Batavia. V. G.=Verspreide Geschriften van Prof. Dr. H. Kern. ADDITIONS AND CORRECTIONS Page 7, 1. 5. Mr. Oldham has definitely identified Paloura with the "existing village of Paluru at the northern extremity of the Gaiijam district, about 6 miles N.E. of Ganjam town/' (J.B.O.R.S., Vol. XXII, pp. 1 ff.). Page 25, f.n. 2. Reference may be made to the following statement : "The Malay Peninsula is the fatherland of the Malays who colonised centuries ago Sumatra " ( Toung Pao 1898, p. 370.). Page 27, 11. 8-16. For a recent example in the neighbourhood of Vanasari ( Jogyakerta ), cf. T. B. G., 1935, pp. 83 ff. Page 29, f.n. 1. Add at the end : "and 'History of Malaya (1935) Chapter I." Page 81, 1. 5. The scholars are now inclined to refer the seal to about 600 A.D. Cf. J. Mai. Br. R.A.S., Vol. XII ( 1934 ), p. 173 ; Vol. XIII ( 1935 ), p. 110 ; J.G.I.S., Vol. II, p. 71. Pages 96-7. Mr. H. B. Sarkar suggests (J.A.S.B., Vol. XXIX, pp. 17-21) that as a result of the conquests of Skandagupta, a large body of Sakas from Gujarat, under a local chieftain, probably Aji Saka by name, emigrated to Java and introduced the Saka Era. The arguments in support of this theory do not appear to me to be very convincing. Page 99. paras 1-2. Dr. J. Przyluski holds that 'the most ancient travellers did not make a clear distinction between the islands of Java and Sumatra, and these two great islands formed the continent of Yava. Probably for Ptolemy and for all the ancient geographers Yava is Java-Sumatra/ (J.G.I.S., Vol. I, p. 93) XVlii ADDITIONS AND CORKECTIOtf & Page 106, 11. 9-11. A Shell inscription is engraved at Ci-Aruton below the foot-prints of king Parnavarman. Dr. K. P. Jayaswal reads it as "Sri Purnna- varmanat" ( Ep. Ind., Vol. XXII, p. 4 ), but it is, at best, doubtful. Mr. F. M. Schnitger draws attention to a reference to Tarumapur in an inscription of Kulottunga (S.LI., Vol. Ill, Part 2, p. 159). It is about ten miles north of Cape Comorin, the region from which Agastya worship spread to the Archipelago. Schnitger finds in the name Taruma an additional argument for the southern origin of Purnavarman (T.B.G., 1934, p. 187.). Page 142, 11. 4-5. Cf. also Schlegel's views ( Toung Pao, Ser. II, Vol. II, pp. 109 ff. ). Book I THE DAWN OF HINDU COLONISATION SUVARNADVlPA Chapter I. THE LAND The Malay Peninsula and the Malay Archipelago constitute together the region known as Malayasia. Although this name is not in general use, we prefer to adopt it as it very nearly coincides with the group of ancient Indian colonies in the Far East with which we propose to deal in this volume. The Malay Peninsula forms the most southerly part of the mainland of Asia. It is a long narrow strip of land projecting into the China sea and connected with the mainland by the Isthmus of Era. In spite, however, of this connection with land, the peninsula belongs, geographically, to the Malay Archipelago and not to the Asiatic continent. The Malay Archipelago is also designated Indian Archipelago, East Indies, Indonesia, Asiatic Archipelago or Insulinde. It begins with the large island of Sumatra which lies to the west of the Malay Peninsula and is separated from it by the Straits of Malacca. The narrow Sunda Strait parts Sumatra from the neighbouring island of Java to its south-east. Java is the beginning of a series of islands lying in a long chain in the direction from west to east. These are Bali, Lombok, Sumbawa, Flores and a number of small islands which almost stretch upto New-Guinea. A little to the south of this line are the two important islands, Sumba and Timor. A similar chain of islands lies to the north, along a line drawn through the centre of Sumatra towards the east. It 2 THE LAND begins with Borneo, the largest island in the archipelago. Next comes Celebes, and then the large group of islands known as the Moluccas or Spice islands. Beyond all these islands, numbering more than six thousand, lie the large island of New Guinea to the east and the group of islands known as the Philippines to the north. The Archipelago is separated from Indo-China in the north by the South China Sea and from Australia in the south by the Timor Sea. To the west there is no large country till we reach the shores of India and Africa, the intervening sea being dotted with hundreds of islands. The most important of these, beginning from the cast are Andaman, Nicobar, Ceylon, Maldives, Laccadives and Madagascar. As Wallace has pointed out, it is seldom realised that the dimensions of the Archipelago are really continental. "If transferred to Europe and the western extremity placed on lands' End, New-Guinea would spread over Turkey." It extends over 50 degrees of longitude (100 to 50) and nearly 25 degrees of latitude (10S. to 15N.) It is a very singular characteristic of the Archipelago that one part of it, including Borneo, Celebes, Sumatra, Java and Bali is separated by shallow sea from Asia, and the other part, including New Guinea, Flores, and Lombok is similarly separated from Australia. Between these two parts, however, the depth of the sea has been found to be from 1000 to 3,557 fathoms, although in some places, as between Bali and Lombok, the two regions are separated by a strait not more than 15 miles wide. The study of the fauna corroborates the natural difference between these two regions, and we might accordingly divide the Archipelago into an Asiatic and an Australian Zone. Wallace, who has gone more deeply into this question than any other scholar, postulates from the above premises that Sumatra, Java, Bali and Borneo formed at one time a part of the continent of Asia. He describes their evolution into separate islands as follows : SUVAENADVlPA 3 "Beginning at the period when the Java Sea, the Gulf of Siam, and the straits of Malacca were dry land, forming with Borneo, Java and Sumatra a vast sothern extension of the Asiatic continent, the first movement was probably the sinking of the Java Sea as the result of volcanic activity, leading to the complete and early separation of Java. Later Borneo, and afterwards Sumatra, became detached and since then many other elevations and depressions have taken place." Similar observations are made by Wallace regarding other parts of the Archipelago. As we arc mainly concerned with that part of it alone which includes Sumatra, Java, Bali and Borneo, we need not pursue these interesting investigations any further. A detailed account of the more important islands will be given separately when we deal with them individually in subsequent chapters. Here we need mention only a few general characteristics of the Archipelago. The equator passes almost through the centre of the Archipelago, and, excepting the northern half of the Philippines, nearly the whole of the Archipelago lies within ten degrees of latitude on cither side. In consequence warm summer prevails throughout the year and the only change of seasons is that from dry to wet. The whole of this region is within the influence of the monsoons but free from hurricanes. The Archipelago is eminently a mountainous region and a volcanic band passes through it "in a sweeping curve five thousand miles long, marked by scores of active and hundreds of extinct craters. It runs through Sumatra and Java, and thence through the islands of Bali, Lombok, Flores to Timor, curving north through the Moluccas, and again north, from the end of Celebes through the whole line of the Philippines. The zone is narrow ; and on either side the Malay Peninsula, Borneo, Celebes, and New Guinea have np 4 THE LAND known volcanoes, and are apparently not subject to serious disturbances/' 1 The geographical position of Malayasia invested it with a high degree of commercial importance. Situated on the highway of maritime traffic between China on the one hand and western countries like India, Greece, Rome and Arabia on the other, it was bound to develop important centres of trade and commerce. The route to China from the west lay either through the Straits of Malacca or along the western coast of Sumatra and then through the Sunda Strait. Thus Sumatra and Malay Peninsula, and, to a certain extent, Java also profited by this trade. The main volume of this trade must always have passed through the Straits of Malacca, and sometimes, perhaps, the goods were transported by land across the Isthmus of Kra in order to avoid the long voyage along the eastern and western coasts of the Malay Peninsula. Malayasia has been famous in all ages for its timber and minerals and almost enjoyed the monopoly in spices. This was undoubtedly the main reason why the western nations were attracted to this corner of Asia from very early times. This was particularly true of India and China which were the nearest countries to the Archipelago that possessed a highly developed civilisation from an early period. There was a regular maritime intercourse between India and the Far East as early at least as the first century A. D. This is definitely proved by the statement in the Periplus that ships from Indian ports regularly sailed to Chrysc, and there I, The preliminary account of the Archipelago is based on the following works : (a) Major C. M. Enriquez Malaya ( Hurst and Blackett, 1927 ). (b) John Crawfurd Dictionary of the Indian Islands and adjacent countries ( London, 1856 ). (c) A. Cabaton Java, Sumatra and the other islands of the Dutch East Indies (T. Fisher Unwin, 1911 ). SUVARNADVlPA 5 was a brisk trade relation between the two. 1 As we shall see later Chryse was a vague name applied to Malayasia. The further statement in the Periplus, that after Chryse "under the very north, the sea outside ends in a land called This," is of singular importance, inasmuch as 'This' undoubtedly stands for China. As Clifford has pointed out, this tends to prove "that the sea-route to China via the Straits of Malacca even though it was not yet in general use, was no longer unknown to the mariners of the cast." This is confirmed by the fact that not long afterwards the sailor Alexander sailed to the Malay Peninsula and beyond ; for, to quote again from Clifford, "it may safely be concluded that the feasibility of this south-eastern passage had become known to the sea-farers of China long before an adventurer from the west was enabled to test the fact of its existence through the means of an actual voyage." 2 The author of the Periplus docs not seem to have possessed any definite information or accurate knowledge of the Far East. The reason seems to bo that there was no direct communication between the Coromandcl coast and the Far East, but the voyage was made from the Gangctic region either direct or along the coast of Bay of Bengal. This follows from the fact that whenever Chryse is mentioned in the Periplus it is invariably associated with the Ganges. As this aspect of the question has not been generally recognised I may quote below the relevant passages from SchofPs translation. 1. Referring to the Chola country the author says ; "Among the market-towns of these countries, and the harbours where the ships put in from Damirica (Tamil land) and from the north, the most important are, in order as they lie, first Camara, then Poduca, then Sopatma ; in which there are ships of the country 1. The Periplus of the Erythraean sea ( edited by W. H. Schoff Longmans, 1912 ) pp, 45-48- 2, Ibid. p. 260. 6 THE LAND coasting along the shore as far as Damirica ; and other very large vessels made of single logs bound together called sangara ; but those which make the voyage to Chryse and to the Ganges are called eolandia^ and are very large." (p. 46) 2. "After these, the course turns towards the east again, and sailing with the ocean towards the right and the shore remaining beyond to the left, Ganges conies into view, and near it the very last land towards the east, Chryse." (p. 47) 3. "And just opposite this river (the Ganges) there is an island in the ocean, the last part of the inhabited world toward the east, under the rising sun itself ; it is called Chryse." (p. 48) Thus there is hardly any doubt that to the author of the Periplus Chryse is closely associated with the Gangetic region. The last sentence in the first passage may be taken to imply a direct voyage to Chryse, but it is at least very doubtful. Besides, it is to be remembered, that the author of the Periplus himself says that the coasting voyage was the order of the day, and he narrates the striking discovery by Hippalus of a direct voyage to the west coast of India from African shore. 1 It is difficult to believe that the author would not have referred to a direct voyage from the Coromandcl coast to the Far East, if such a course was known in his time, at least in passage No. 1. quoted above. This view is confirmed by Ptolemy. He refers to the aplieterium, immediately to the south of Paloura, where the vessels bound for the Malay Peninsula "ceased to follow the littoral and entered the high seas"*. 8. L6vi has shown that the city of Paloura, which played such an important part in the eastern ocean trade of India was the same as the famous city of Dantapura, in Kalinga, which figures so i. Ibid. p. 45. 3. Ptolemy. ( M'c. Crindle ), pp. 66, 69. SUVARNADVfPA 7 prominently in the Buddhist literature. Thus even in Ptolemy *s days there was no direct voyage from the Coromandel coast, but in addition to the coastal voyage along the Bay of Bengal from Tamralipti, a direct voyage to the east was made from Paloura near modern Chicacole. 1 It is difficult, therefore, to accept the view, generally held on the authority of the Periplus, that there was a direct voyage between South India and the Far East in the first century A. D. 8 It cannot, of course, be maintained that a direct voyage between South Indian ports and Malay Peninsula was an impossible one. All that we learn from Ptolemy is that the usual point of departure for the Far East was near Paloura. It is possible, however, that occasionally ships sailed direct from Coromandel coast to the cast, or via Ceylon and Andaman Islands to the coast of Sumatra. 8 The fame of Paloura or Dantapura, in Kaliiiga ( the coastal region between the Mahanadl and the Godavarl ), was no doubt due, at least to a great extent, to its importance as the point of departure for the Far East. That probably also explains why the Chinese referred to Java and other islands of the Archipelago as Kling, no doubt an abbreviation of Kalinga. All these point out to Kalinga as the particular region in India which was more intimately connected, through its port Paloura, with the Far East in the early period. There were important ports on the opposite coast also. In the Malay Peninsula we have reference to Takkola in classical writings, to Kala by Arab writers and to Singapore and Malacca by the Portuguese. In Sumatra the most important 1. J. A. 1925, pp. 46-57. G. Jouveau-Dubreuil held the view that the apheterium was situated near the mouth of the Godavarl (Ancient History of the Deccan pp. 86-88) 2. Cf. e. g. Krom's emphatic opinion in Geschiedenis, p. 53. 3. For the probability of such voyage in pre-historic times cf. Chap. II. 8 THE LAND port was Srl-Vijaya. Others will be referred to in due course. On the whole, therefore, we can easily visualise Malayasia as a fairly extensive region between the continents of Asia and Australia, enjoying peculiar advantages of trade and commerce, both by its geographical position as well as by its native products. From a very early period it had intercourse with China on the north, Australia and the Pacific islands on the south and east, and India and various islands in the Indian ocean on the west. It was more intimately connected with Burma and Lido-China on the north, as their inhabitants were allied to its own. Chapter II THE PEOPLE A detailed discussion o the people or peoples that inhabited Malayasia before the advent of the Hindus belongs to the domain of anthropology. It is beyond the scope of the present work to dwell upon this question at length and I propose, therefore, merely to give in broad outline the salient facts on which there is a general agreement among scholars. It is usual to divide the population into three main strata : (1) The primitive races (2) the Proto-Malays and (3) the Malays. 1 (1) The Semang and the Sakai of the Malay Peninsula may be taken as fair specimens of the wild tribes that inhabited the region in primitive times. The Semang Negritos belong to the earliest stratum of population which has survived in the peninsula. They now occupy "the wooded hills in the north of the peninsula, in Kedah, Pcrak and northern Pahang : with occasional communities like the Temo in Ulu Bera and Ulu Rompiii in south Pahang". "They are dark, with woolly hair, and flat, spreading noses, feeble chins, and lips often everted : and sometimes they are almost pigmies in size. But for a bark loin-cloth, they are naked They have no form of agriculture whatever, and live upon jungle produce and by hunting, fishing and trapping. Their distinctive weapon I. The account of the tribes is taken from Major C. M. Enriquez 1 excellent book "Malaya an account of its People, Flora and Fauna" (Hurst and Blackett 1927) Chs. V-VIII. The quotations are also from this book. A detailed account of the manners and customs of the primitive people is given by I. H. N. Evans in "Ethnology and Archaeology of the Malay Peninsula" (Cambridge, 1927) and by R. J. Wilkinson in "A History of the Peninsular Malays" 3rd Edition Singapore (1923). 2 10 THE PEOPLE is the bow and poisoned arrow. They live under over-hanging rocks or leaf-shelter and build no houses." The Sakai occupy the mountains of south-east Perak and north-west Pahang. They resemble the Semang in many respects and the two have interbred to a considerable extent. "In colour the Sakai vary from brown to yellow, and are lighter even than Malays. The hair is long and black, the nose finely cut and tilted, the eyes horizontal and half-closed and the chin sharp and pointed. They tattoo the face in certain districts and sometimes wear a ring or a porcupine's quill through the nose. Their distinctive weapon is the blow-pipe with which they arc extremely skilful. As a rule they live in huts sometimes placing them up trees at a height of 30 feet from the ground." (2) A number of wild tribes to be found all over Malayasia are called Proto-Malays, as their languages are distinctly Malay. The Jakun who occupy the south of the Malay Peninsula may be taken as a fair specimen of this type. "They are coppery in colour, with straight smooth black hair of Mongolian type. The cheek-bone is high, the eyes are slightly oblique. Though inclined to be nomadic, they usually practise some form of agriculture, and live in fairly good houses." The Proto-Malay type is met with all over Malayasia. The Batak, Achinese, Gayo and Lampongs of Sumatra, the Dayaks, Kayan, Kenyah, Dusun and Murut of Borneo, and the aborigines of Celebes, Ternate and Tidore all belong to this type. Some of them are cruel and ferocious. The Batak, for example, are said to be cannibals who eat prisoners and aged relatives. . The Kayan and Kenyah are noted for their frightful cruelty and their women seem to have a genius for devising tortures for captives, slaves and strangers. Others are more civilised. The Dayaks of Borneo, although head-hunters for ritualistic purposes, are described as 'mild in character, tractable and hospitable when well used, grateful for kindness, industrious, honest and simple ; neither treacherous nor cunning, and so SUVARNADVlPA 11 truthful that the word of one of them might be safely taken before the oath of half a dozen Malays/ (3) The Malays, who now form the predominant element of the population of Malayasia, have been divided by Wallace under four great heads 1 : (1) The Malays proper who inhabit the Malay Peninsula and the coastal regions of Sumatra and Borneo ; (2) the Javanese of Java, Madura, Bali and parts of Lombok and Sumatra ; (3) the Bugis of Celebes ; and (4) the Tagalas of the Philippines. Wallace describes the Malay as follows : "In character he is impassive. He exhibits a reserve, diffidence and even bashfulness, which is in sonic degree attractive, and leads observers to think that the ferocious and blood-thirsty qualities imputed to the race arc grossly exaggerated. He is not demonstrative. His feelings, of surprise or fear, arc never openly manifested, and arc probably not strongly felt. He is slow and deliberate in speech. High-class Malays are exceedingly polite, and have all the quiet ease and dignity of well-bred Europeans. Yet all this is compatible with a reckless cruelty and contempt for human life, which is the dark side of their character." Having given a short description of the various peoples, we may now proceed to trace their origin and affinities from racial and linguistic points of view. Both these questions are beset with serious difficulties and the views of different scholars are by no means in complete agreement. We must, therefore, content ourselves by merely quoting the view of one eminent authority in each line of study, referring the readers, who seek further information, to special treatises on the subject. Mr. Roland B. Dixon has summed up as follows the racial history of the Malay Peninsula. 9 1. Wallace Malay Archipelago, Vol. II, p. 439. 2. Roland B. Dixon The Racial History of Man (Charles cribner's Sons, New York, London, 1923) p. 275. 12 THE PEOPLE "The oldest stratum of population was the Negrito Palae- Alpine which survives to-day in comparative purity only among the Andamanese. With this was later blended a taller Negroid people, of mixed Proto-Australoid and Proto-Negroid types, to form the Semang. This Negroid population is still represented among some of the hill-folk in Burma, such as the Chin, is more strongly present in Assam and dominant in the greater part of India. Subsequently to the formation of Semang a strong immigration came into the Peninsula from the north, of the normal Palae- Alpine type, of which perhaps some of the Karen may be regarded as the last survivors. From the fusion of these with the older Semang was derived the Sakai and some, perhaps, of the Jakun ; the later and less modified portions of this wave forming the older Malay groups of to-day. Finally in recent times came the Mcnangkabau Malays from Sumatra who have overlain the earlier group throughout the south." The statement in the last sentence that the Malays came from Sumatra, is, perhaps, no longer valid, as we shall see later. As regards language, it has been recognised for a long time that the language of the Malays belongs to the same family as that of Polynesia and the name Malayo-Polynesian was applied to this group. Since then, however, Mclanesian, Polynesian, Micronesian and Indonesian (Malay) languages have all been proved to belong to the same family to which the new name Austro-nesian has been applied. The discovery of human skulls and the pre-liistoric stone implements in Iiido-China and Malay Archipelago has demonstrated the racial and cultural affinity between many of the races speaking these languages. 1 There is hardly any doubt that the primitive wild tribes of Malayasia belonged to the palaeolithic age. Discoveries of human skulls and other pre-historic finds establish a sort of racial and I. For pre-historic finds cf. Tijdschr. Aardr. Gen. Vol. 45 (1928) PP- SS 1 "^; O. V. 1924 (127-133); 1926 (i74-i93) 1929 (pp. 23 ff). SUVARNADVIPA 13 cultural affinity among large groups of them spreading over Indo-China, Indonesia, Melanesia, and as far as Australia. They were gradually ousted by the peoples speaking Austro-nesian group of languages and belonging to the Neo-lithic period. The time and nature of contact between all these races we have no means to determine. So far as we can judge from the analogy of similar events and the few facts that present themselves to us, the result of the conflict seems to be, that the original inhabitants were partly exterminated, partly incorporated with the new-comers, and partly pushed back to hills and jungles where some of them maintain a precarious existence upto the present day. Whether the conquering peoples all belonged to one race cannot be definitely determined. This view is at least in accord with the fact that their languages were derived from one stock, and it is also supported by prc-historic finds, as noted before. Be that as it may, there is hardly any doubt that they must have lived together in close bonds of union, before they were scattered over the islands in the Pacific ocean. "We can thus easily postulate a common home for this Austro-nesian group of peoples. Kern made a critical study of the question by considering the fauna and flora of this home- land as revealed by the common elements in the various languages of the group. By this process of study he placed the home-land of the Austro-nesians on the coast of Indo- China. 1 This view is corroborated by the fact that human skulls which are purely Indo-nesiari and prc-historic finds which are undoubtedly Proto-Indo-nesian have been found in Indo-China. I. Kern V. G. Vol. VI, pp. 105-120. Kern calls it "Secundaire stamland" (Second home), for he traces their origin further back to India, as will appear later (V. G. Vol. XV, p. 180). R. O. Winstedt has further supported this view by noting the occurrence of identical tales in the Indo-nesian and Mon-Khmer languages (J. Str. Br. R. A. S. No. 76, pp. iiQff). 14 ffiE PEOPLE Ferrand has traced the early history of these peoples still further back, mainly on the authority of an account preserved by Ibn Said (13th cent.). He thinks that they originally lived in upper Asia as neighbours of the Chinese, and being driven by the latter, about 1000 B. C., came down to Indo- China along the valleys of the Irawadi, Salwin, Mekong and Menam rivers. Nearly five hundred years later they migrated again from this region to Malay Peninsula and various islands of the Indian Archipelago. 1 Of late, another theory has been advanced by Van Stein Callenfels. He infers from the remains of. their metallic objects that the original home of the Austro-nesians lies in the region of the Altai mountains.* It must be remembered, however, that considering the scanty and uncertain data on which the above conclusions are necessarily based, they can only be regarded as provisional. Nor should it be forgotten that the settlement of the vast region of Malayasia could not possibly have been a simple process of migration of a body of people from the mainland to each of the islands. There must have been currents and cross-currents from different quarters that swelled the tide, and we have to postulate migrations and emigrations, not only many in number but probably also varied in character. It will be outside the scope of this book to pursue the ramification of this fascinating problem any further. But there is another point of view regarding this question which is virtually connected with the subject-matter of this book and must be treated at some length. Recent linguistic researches have established definite connection between the languages of some primitive tribes of India such as Munda and KhiTsi with Mon-Khmer and allied languages including those of Semang and Sakai. The great philologist Schmidt has thus established the existence of a linguistic 1. J. A. Il-XII (1918) pp. 120-123 ; I 9 I 9> P- 201. 2. T. B. G. Vol. 64 (1924), p. 604. SUVARNADVlPA 15 family, which is now called Austro-Asiatic. 1 Schmidt believes that 'the linguistic unity between these peoples which is now definitely established, points to an ethnic unity among them as well, though positive and satisfactory evidence on this point is lacking yet. )a "Schmidt has extended his studies even further and proposed to connect the Austro- Asiatic family with the Austro-nesian" to which, as stated above, the Malays belonged. Schmidt thus seeks to establish a "larger linguistic unity between Austro- Asiatic and Austro-nesian and calls the family thus constituted 'Austric* " Here, again, Schmidt indicates the possibility of an ethnic unity among the peoples whose linguistic affinity is thus definitely assured. Schmidt thus regards the peoples of Indo-China and Indo-nesia as belonging to the same stock as the Munda and allied tribes 3 of Central India and the Khasis of North- eastern India. He regards India as the original home of all these peoples from which they gradually spread to the east and south-east. The following passage sums up his views in this respect. 'In the same way as I have presented here the results of my investigations on movements of peoples who, starting 1. Die Mon-Khmer-Volker etc. (1906) pp. 35 ff, I have used the French translation in B. E. F. E. O. Vol. VII. (pp. 213-263), VIII (pp. 1-35). A good exposition of Schmidt's view, so far as the linguistic aspect is concerned, is given in the introductory chapter in ''Pre-Aryan and Pre-Dravidian in India" by Dr. P. C. Bagchi (Calcutta University, 1929) from which I have freely quoted. (The page marks within bracket in the text refer to this book). 2. Schmidt, op. cit. cf. specially, p. 233. 3. The Muncla group of language includes Kol, the more eastern Kherwari with Santali, Muncjarfc Bhumij, hirhor, Kocla, Ho, Turi, Asuri, and Korwa dialects and the western Kurku ; Kb aria ; Juang ; and the* two mixed languages Savara and Gadaba. (Dr. P. C. Bagchi, op. cit. p. VI.) 16 THE PEOPLE from India towards the east, at first spread themselves over the whole length of Indo-Chinese Peninsula, and then over all the islands of the Pacific Ocean upto its eastern extremity, my attention has for long been drawn to another current which, in my opinion, also started from India, but turned more directly towards the south and touching only the western fringe of the Pacific Ocean proceeded, perhaps by way of New Guinea, towards the continent of Australia' 1 Schmidt's views, like those of Ferrand and others noted above, must be regarded as only provisional. 2 But several other scholars have supported this view on entirely different grounds. Among them may be mentioned the names of 8. L6vi, J. Przyluski and J. Bloch. The relevant articles on this subject by these eminent scholars have been published together in English version by Dr. P. C. Bagchi. The following summary is derived almost entirely from this book entitled "Pro-Aryan and Pre-Dravidian in India/' 'Prof. Thomson first maintained that Munda influence can be traced in the formation of Indian vernaculars. Recent studies have tried to establish that this influence can be traced further back. Prof. Przyluski has tried to explain a certain number of words of the Sanskrit vocabulary as fairly ancient loans from the Austro-Asiatic family of languages. Prof. Jules Bloch has proved that the question of the Munda substratum in Indo-Aryan can not be overlooked (pp. XI-XII) 'But the problem has other aspects too, and it has been further proved that not only linguistic but certain cultural and political facts also of the ancient history of India can be explained by admitting an Austro-Asiatic element. In 1. Schmidt, op. cit, pp. 248-249. A critical summary of Schmidt's view is given by Blagden "From Central India to Polynesia" (J. Str. Br. R. A. S. No. 53 p. 63). 2. Recently Schmidt's view has been challenged by W. F. de Hevesy who denies the existence of the Austro-Asiatic family of languages (J. B. O. R. S. Vol. XX pp. 251 ff). Slf^AKNADVlPA 1? 1923 Prof. S. L^vi tried to show that some geographical names of ancient India like Kosala-Tosala, Auga-Vanga, Kalinga-Trilinga, Utkala-Mekala, and Pulinda-Kulinda, ethnic names which go by pairs, can be explained by the morphological system of the Austro-Asiatic languages. In 1926 Przyluski tried to explain the name of an ancient people of the Punjab, the Udumbara, in a similar way and affiliate it to the Austro- Asiatic group. In another article, the same scholar discussed some names of Indian towns in the geography of Ptolemy and tried to explain them by Austro-Asiatic forms (pp. XII-XIII). 'In another scries of articles, Prof. Przyluski is trying to prove a certain number of Indian myths by the Austro- Asiatic influence. He studied the Mahabharata story of Matsyagaiidha and some legends of the nagl, in Indian literature, compared them with similar tales in the Austro-Asiatic domain, and concluded that these stories and legends were conceived in societies living near the sea, societies of which the civilisation and social organisation were different from those of the neighbouring peoples, the Chinese and the Indo-Aryans/ (p. XIII) The bearing of all these interesting investigations on the question under discussion has thus been admirably expressed by S. L<$vi. "We must know whether the legends, the religion and philosophical thought of India do not owe anything to this past. India has been too exclusively examined from the Indo-European standpoint. It ought to be remembered that India is a great maritime country, open to a vast sea forming so exactly its Mediterranean, a Mediterranean of proportionate dimensions which for a long time was believed to be closed in the south. The movement which carried the Indian colonisation towards the Far East, probably about the beginning of the Christian Era, was far from inaugurating a new route, as Columbus did in navigating towards the West. Adventurers, traffickers and missionaries profited by the technical progress of navigation, 3 IS THE PEOPLE and followed under the best condition of comfort and efficiency the way traced from time immemorial by the mariners of another race whom the Aryan or Aryanised India despised as savages." (pp. 125-6) In other words, the cumulative effect of all these researches is to push back the first phase of Indian colonisation in the Far East to a time prior to the Aryan or Dravidian conquest of India. It will not perhaps be rash to imagine that, that colonisation was, at least partly, the result of Dravidian and Aryan settlements in India which dislodged the primitive peoples and forced them to find a new home across the seas. 1 It may be noted, however, that conclusion of an almost opposite character has been arrived at by certain scholars. Krom, for example, believes that the Indo-nesians had colonised India in primitive times, and the later Aryan colonisation of the Far East was merely the reverse of that process. 2 This is in flagrant contradiction to the views of Schmidt and Lvi, and seems to be based mainly on the theory of Mr. J. Hornell. In his Memoir on u thc Origins and Ethnological significance of the Indian Boat Designs" Mr. Hornell "admits a strong Polynesian influence on the Prc-Dravidian population of the southern coast of India. He thinks that a wave of Malayan immigration must have arrived later, after the entrance of the Dravidians on the scene, and it was a Malayan people who brought from the Malay Archipelago the cultivation of the Coco-palm." (p. XVII) Two other observations by different scholars probably lend colour to this view. In the first place, Prof. Das Gupta "has brought out the striking analogy between some sedentary games of India (specially of the Central Provinces, Bengal, Bihar, Orissa and the Punjab) and those of Sumatra." (p. XVII) 1. Kern also held similar view; cf. V. G., Vol. XV, p. 180. He held that they came from India, their ultimate home being Central Asia. This is not in conflict with his original view that the home- land of the Malayo- Polynesians was the eastern coast of Further India, 2. Krom Geschiedenis, p. 38, SUVARNADVlPA 19 Secondly, we have the following remarks made by Dr. J. H. Hutton with reference to some pre-historic monoliths of Dimapur near Manipur. "The method of erection of these monoliths is very important, as it throws some light on the erection of pre-historic monoliths in other parts of the world. Assam and Madagascar are the only remaining parts of the world where the practice of erecting rough stones still continues The origin of this cult is uncertain, but it appears that it is to be mainly imputed to the Mon-Khmer intrusion from the east." In his opinion these monoliths take the forms of Ufigam and yoni, and he thinks that they possibly originated in Indo-nesia. (pp. XVII-XVIII) In all these cases the similarity that undoubtedly exists may be explained by supposing either that India derived the practices from Indo-ncsia or that Indo-nesia derived them from India. The recent discoveries at Mohenjo-daro, 1 however, prove the existence of the cult of Linga and Yoni in the Indus Valley at least in the beginning of the third mUlenium B. C. Thus the migration of the cult towards the cast seems most probable. Considering the whole course of Indian history it seems more probable that the migration of the people and ideas was generally from India towards the east, and no tangible evidence has yet been obtained that the process was just the reverse. On the whole, therefore, the views of Schmidt and Sylvain Lvi appear far more reasonable than those of Horncll and Hutton. In view of a possible pre-historic connection between India and Malayasia, it is necessary to say something on the word Malaya which has given the name to the dominant race and the dominant language in Malayasia. It is a well-known fact that an Indian tribe called Malava ( var. Malava ) or Malaya (var. Malaya) is known from very ancient times. The common form, of course, is Malava, but the form 'Malaya' also occurs on their coins. In a discussion of these coins Mr. Douglas maintained that Malaya is the older form of the j. Marshall The Indus Civilisation, pp. 58 ff. 20 THE PEOPLE tribal name. His conclusion rests chiefly on the Greek form of the name. "The Greeks" says he "called them the Malloi. Had the name Malava been in common use at that time, I feel sure that the Greeks would have transliterated the word as the Malluoi. This seems to me to show that the commoner form of the tribal name at the time of the Greek invasions was Malaya." 1 Whatever we may think of this view, there is no doubt that both the forms were in common use. The form Malaya occurs in Mudrfi-Raksasa 9 and Malaya in an inscription found at Nasik. 8 The interchange of y and v is also attested by the alternative names of a Satavahana king as Pulumayi and Pulumavi.* The antiquity of the Mftlava Malaya tribe is proved by Panini's reference to it as a clan living by the profession of arms (ftyudhajivin). There is no doubt also that the Malavas were widely spread in different parts of India. Alexander met them in the Punjab, but their settlement in Rajputanfi is proved by the discovery of thousands of their coins at Nagar in Jaypur State 5 and the reference in the Nasik inscription mentioned above. The Indian literature also makes frequent references to the Malavas. The Mahabharata knows of various Malava tribes in the west, north and south. 6 The Eamayana and Matsya- purana include the Malavas among the eastern tribes 7 while various other texts refer to them as a people in one or other parts of India. i. J. A. S. B., N. S., Vol. XIX (1924). Numismatic supplement No. XXXVII, p. 43. 2. Canto I, verse 20. 3. Rapson Catalogue of the Coins of the Andhras etc , p. LVII. 4. Ibid, fn. I. 5. V. Smith Catalogue of Coins in the Indian Museum, pp. 161 ff. I7off. 6. cf. Mahabharata 11-32, III-si, VI-g, 87, 106. 7. RamSyana IV-4O, V-22. Matsyapurana Ch. 1 14 V. 34. SUVAKNADVlPA 21 The wide spread of the Malavas may also be guessed from Indian dialects or toponyms connected with them. Mr. Grierson has referred to a Malavia dialect extending from Perozcpur to Bhatinda in the Punjab, and we have also the well-known Malayalam language of southern India. The well-known Indian provinces of Malava in northern India and Malaya-bar or Malabar in southern India still testify to the influence of that tribal name. The Malaya mountain, the source of Sandalwood, is referred to in Purftnas and other ancient literature as one of the seven Kulaparratas or boundary mountains in India. Lastly the famous era, beginning in 58 B. C., has been associated with the Malavas from the earliest times. The Buddhist literature also refers to Malaya country. The famous Lankavatara Sutra is said to have been delivered by the Buddha in the city of Lanka on the summit of the Malaya mountain on the border of the sea. The Buddhist reference to Malaya has been regarded by some as purely imaginary but the existence of a Malaya mountain in Ceylon is proved by Ptolemy and MahSvarnsa. That of a Malaya country and a Malaya mountain in the south of India also rests on definite grounds. The great Buddhist Vajrabodhi who came to China in A. D. 719 is described as a native of the Malaya country adjoining mount Potalaka, his father being preceptor of the king of Kancl. Hiuen Tsang places the country of Malakuta, 3000 li south of KaSci, and refers to its mountains Malaya and Potalaka. Alberuni also places Malaya 40 farsakhs (about 160 miles) south of KaSci. Thus we have both a Malaya country and a Malaya mountain in the extreme south of the Indian Peninsula. 1 There is no doubt that this name is i. S. L6vi in J. A. CCVI, pp, 65 ff. Walters On Yuan Chwang, Vol. II, pp. 229-231. Ptolemy (M'c. Cr indie), p. 249. Geiger Mahavarhsa, p. 60. Sachau- Alberuni, Vol. I,, p. 200; cf. also B. E. F. E. Q. Vol. IV, p. 359. 22 THE PEOPLE preserved in modern Malabar which the Arab Geographers call either Malaya-bar or simply Malay. 1 While the Malava and Malaya can thus be traced as tribal or geographical names all over India, upto its north-western, eastern and southern extremities, the spread of this name across the sea is no less conspicuous. On the east, the famous Malays of Malayasia, the place names Malay and Malacca in the Peninsula, Malayu in Sumatra, 8 Mala or Malava for Laos and perhaps even Molucca islands in the eastern extremity of the archipelago, and on the west Maldives (Maladvlpa), and Malay the ancient name of Madagascar 8 testify to the spread of the name in Indo-China and along the whole range of the southern ocean. Now Ferrand has- drawn our attention to the fact that the Indo-nesian language, mixed with Sanskrit vocabulary, was current in Madagascar. Combining this fact with other traditional evidences he has come to the conclusion that Madagascar was colonised in ancient times by Hinduiscd Indo-nesians.* It is not necessary for the present to discuss the further implications of this theory as enunciated by Ferrand, and I must rest content by pointing out the bearing of the account of Malava Malaya, as given above, on this as well as several other theories. Now the theories of Schmidt, Lvi, Homell and Hutton (as modified by the discoveries at Mohenjo-daro) referred to above, all presuppose, or are at least satisfactorily explained by 1. Ferrand Textes, p. 38 fn. 5, pp. 204, 340. 2. "The name Malayu is very common in Sumatra. There are a mountain and a river of that name ; there are five villages called Malayu and a tribe of that name." T'oung Pao, series II, Vol. II, p. 115. 3. Ferrand Textes, pp. 389, 396, 4. J. A. II-XI1 (1918) pp. 121 ff. J. A. 1I-XIV (1919), pp 62 ff., pp. 201 ff. Krom, however, thinks that the Indo-nesian people colonised Madagascar before they came into contact with the Hindus. He attributes the Indian element in the language of Madagascar to later intercourse (Geschiedenis, pp. 38-9). SUVABNADVIPA 23 a stream of migration of Indian peoples towards the east and south-east, to Assam, Burma, Lido-China and Malay Archipelago, both by land and sea. The migrations of the Malava tribe, so far as we can judge from the occurrence of geographical names, follow, as we have seen above, exactly this course, as we can trace them from the Punjab to Assam on the one side and to Malabar on the other. From Malabar we can trace the name in the east through Ceylon (Malava mountain in Lanka) and Sumatra (Malayu) to Malay Peninsula, perhaps even to Moluccos. On the west we can trace it from Malabar to Maldives and Madagascar. It is, no doubt, more reasonable to explain the linguistic facts observed by Ferrand in Madagascar by supposing a common centre in India, from which the streams of colonisation proceeded both towards the east as well as towards the west, than by supposing that Hindu colonists first settled in Malayasia and then turned back to colonise Madagascar. The people of Madagascar have a tradition that their ancestors came from Mangalore. 1 This place is located by Ferrand in the south of Malaya Peninsula, but it should not be forgotten that Mangalore is the name of a well-known place in Malabar Coast and is referred to by Arab writers as one of the most celebrated towns of Malabar. 2 I do not wish to be dogmatic and do not altogether reject the views of Ferrand. But the known facts about the Malava- Malaya tribe in India seem to me to offer quite a satisfactory explanation not only of the problem of colonisation of Madagascar but also of the racial, linguistic and cultural phenomena observed by Schmidt, Hutton and Hornell. It is interesting to note in this connection that various words inscribed on the coins of the Malavas which have been provisionally explained as names of tribal leaders, are non- Sanskritic. Thus we have Bhapafnyana, Majupa, Mapojaya, 1. J.A. Il-XiV (1919), p. 64. 2. Ferrand Textes, p. 204. 24 THE PEOPLE Mapaya, MagajaSa, Magaja, Magojava, Gojara, MaSapa, Mapaka, Pacha, Magacha, Gajava, Jamaka, Jamapaya, Paya. Whatever the language may be, it shows one peculiar Austro-nesian characteristic, which has been traced by Sylvain L6vi in certain geographical nomenclatures of ancient India, viz., the existence of a certain number of words constituting almost identical pairs, differentiated between themselves only by the nature of their initial consonants. Among the terms on the Malava coins noted above we may easily select two series of this type. 1. Paya, Ma-paya, Ja-ma-paya. 2. Gajava, Magojava. The tribe Malava-Malaya has played great part in the history of India. Its name is associated with an old language, the most ancient era and two important provinces of India. The Malaya tribe has played an equally dominant part in the Indian seas. It has been the dominant race in the Indian Archipelago and its name and language are spread over a wide region extending almost from Australia to African coast. I have shown above enough grounds for the presumption and it must not be regarded as anything more than a mere presumption that the Malava of India may be looked upon as the parent stock of the Malays who played such a leading part in Malayasia. It may be interesting to note here that Przyluski has shown from linguistic data that Udumbara or Odumbara was the name of an Austro- Asiatic people of the Punjab and also designated their country. 1 The Odumbaras were neighbours of the Malavas and the coins of the two peoples belong approximately to the same period. 3 Thus, prima facie there is nothing inherently objectionable in the assumption that the Malava-Malaya may also be the name of an Austro-Asiatic people. I. P. C. Bagchi Pre- Aryan and Pre-Dravidian in India, pp. 149-160. a. V. A. Smith op. cit, pp. 160 ff., p. 166. SUVABNADVIPA 25 If the presumption be held a reasonable one, we may refer to Ptolemy's account as an evidence that the Malays had spread to the Far East before his time. Ptolemy refers to mountain Malaia in Ceylon and cape Malcou Kolon in the Golden Khersonesus. Regarding the latter, M'c. Crindlc remarks as follows : "Mr. Crawford has noticed the singular circumstance that this name is pure Javanese signifying "Western Malays." Whether the name Malay can be so old is a question : but I observe that in Bastian's Siamese extracts the foundation of Takkhala is ascribed to the Malays." Thus indications are not wanting that various branches of the Malay tribe had settled in Malayasia before the second century A. D. There is a general tradition among the Malays of Minankabau that their parent stock came from India and settled in the western coast of Sumatra. 1 Thus while it is impossible to arrive at any definite conclusion in this matter, pre-historic migrations of Austro-nesian tribes from India to Malayasia appear very probable, and if this view be correct, we may regard the Indian Malaya-Malava people as one of these tribes. 8 1. Cf. Ferrand in J. A. il-XII, p, 77. 2. Although I have arrived at the theory of the Indian origin of the Malays quite independently, it is only fair to note that Gerini made the same suggestion in his 'Researches on Ptolemy's Geography of Eastern Asia' (pp. 101 ff). I have not referred to his views as they are mixed up with a great deal of extraneous matters and some amount of fanciful etymological derivations. So far as I can see, his views rest primarily on the resemblances of geographical names. Gerini explains Maleou-Kolon as referring to two prominent Indian tribal names- Malay and Kola (Cola) of south India, and he traces many other south Indian tribal names in the Malay Peninsula (cf. pp. 102-3). He holds that Malacca was either a modification of Malaykolam or Malayaka (meaning the country of the Malays) or identical with Malaka, the name of a southern Indian tribe mentioned in the Mahabharata (p. 105). I have tentatively adopted this view in respect of both Malacca and Moluccos, With the exception of this and the statement that 4 Chapter III. PRE-HINDU CIVILISATION IN MALAYASIA The Austro-nesian races must have occupied Malayasia for a pretty long time before they came into contact with the Hindus and imbibed their civilisation. In order, therefore, to estimate properly the influence of this new element we must have some idea of the civilisation which these indigenous races possessed before the arrival of the Hindus. Unfortunately the materials for such a study are very scanty. The actual remains left by these races do not difier very much from what is usually termed as 'pre-historic' and met with in various other countries. We may start with a brief account of them, beginning from Java 1 , where a more systematic study has been made of these materials than in other places. Laos is referred to as Malava (p. n?) I have not borrowed from Gerini any views or statements recorded in this chapter. I must also state that it is usually held, though without sufficient reason, that the term Malaya as designating the Malay Peninsula came into use only in the seventeenth century A.D. ( J. Mai. Br. R. A. S. 1930, p. 85), presumably in consequence of the migration of a large number of Malays from Sumatra, in the fifteenth century A. D. ( B. C. A. Iv I Q09, p. 184 ) Blagden refers to I-tsing's Malayu and infers that Malaya country par excellence* was in Central Sumatra, a fact agreeing very well with native Malay tradition on the subject which derives the origin of many of the Malays of the Peninsula from the old Central Sumatran state of Minangkabau ( J. Str. Br. R. A. S. No. 32 pp. 211- 213). This view admits the possibility of the name Malaya being applied to the Peninsula at an earlier date. (cf. Crawford Dictionary pp. 250-252). i. The following sketch of the pre-historic remains of Java is based on (a) Krom Kunst Vol. i. pp. 121-26 ; (b) Krom-Geschiedenis, pp. 42-45. SUVAKNADVlPA 27 The pre-historic archaeological remains in Java may be classified as follows : I. Palaeolithic and neolithic implements such as axe-head chisel, pole and various weapons. II. Megalithic monuments for burying the dead. These are of three kinds. (a) Rock-cut caves, either rude or well-shaped. (b) The stone coffins, consisting of a long and deep rectangular chest with a cover curved like an arch on the outer side. Both the chest and the cover have thick walls, which are rough outside (probably due to long exposure) but polished within. The dead body was introduced through a hole at one end which was then closed by a flat stone. The hole was sometimes surrounded by decorative designs. The chest was also sometimes painted with straight and curved lines and primitive pictures of men and animals (tiger, birds etc). (c) The dolmens which were constructed by placing one big long stone over several other stones set upright in the ground. Various articles arc found in these graves, such as beads, neolithic stone implements, copper rings for arms and legs, iron lance-point or short swords. Bronze articles are not, however, found in these tombs, though sporadic finds of chisels and axe-heads, made of bronze, by their likeness with neolithic implements of the same kind, indicate a knowledge of bronze before the period of later Hindu colonisation. III. In some places in western Java are found rough scratchings under human figures, engraved on rock. These scratchings have been regarded as pre-Hindu Inscriptions. IV. In certain places are found rows of pointed stones, occasionally along with very rude and almost monstrous human figures in stone, known as Pajajaran or Polynesian images. Although all these monuments are properly ascribed to the people or peoples who settled in Java before the Hindu 28 PBE-HINDU CIVILISATION IN MALAYASIA colonisation, it should not be imagined that they are all to be dated before the introduction of that civilisation. They continued to be built throughout the Hindu period, particularly in those regions where the Hindu influence was comparatively weak. It may not be without interest to note that even to-day the megalithic tombs of the types II (b) and II (c) described above are in use among the people of Sumba. On the whole, therefore, while the monuments described above may justly be regarded as characteristic of the pre-Hindu settlers, they cannot all be described as remains of the pre-Hindu period. Attempt has been made to classify the pre-Hindti settlers in Java into distinct groups on the basis of the different types of monuments described above. But as sometimes the different classes of monuments arc found together in the same locality, such attempts cannot lead to any satisfactory conclusion. The pre-historic remains of Sumatra mostly belong to the same classes as those of Java and need not be referred to in detail. We meet with megalithic dolmens and menhirs as well as rock-scratchings or inscriptions with human figures. In respect of this last alone Sumarta offers some striking peculiarities as we occasionally come across a unique type of human figures in stone. These are characterised by large eyes, broad jaws and thick lips. They have got a head-dress of the form of a cap, and a bag hanging from the shoulder. Their wrists and legs are covered and they are represented as either riding on elephants or engaged in. fighting with them. The rectangular back-pieces of some of these figures show that they were used to support a structure. 1 I. Krom-Geschiedenis, p. 44 ; O. V. 1922, pp. 31-37- It is not possible to prove definitely that these figures are really pre-historic and not influenced by the later Hindu civilisation. From the evidence at our disposal it would, perhaps, be safer to regard these figures as belonging to the megaiithic period of culture. SUVABNADVlPA 29 The pre-historic remains of the Malay Peninsula have not yet been studied to the same extent as those of Java and Sumatra. But enough has been discovered to show their general nature. 1 A number of caves containing palaeolithic implements, some of them of Sumatran types, have come to light. But the great majority of the stone implements hitherto discovered are neolithic. Most of them are axe or adze heads, and there is a total absence of knives, spear-heads or arrow-heads. Probably bamboo and hardwoods were used for these purposes. Among implements of rare type may be mentioned a hand-axe and quoit-shaped objects figured in plates XXXVI- VII of Evan's book. Rough cord-marked pottery, in imitation of ware made in a basket, and often with diamond- shaped reticulations, has also been found with the stone implements. Certain tools of bronze or copper have been discovered, but they are distinctly rare, and it is doubtful whether there was any bronze age in the Peninsula. Ancient iron tools are also occasionally discovered, and we have some specimens of graves built of large granite slabs and 'cists' closely resembling the dolmen. On the whole the remains afford us the picture of a very primitive civilisation. Mention may also be made of what are popularly known as Siamese mines. These are circular pits, sometimes more than hundred feet deep and about two feet apart, and connected with one another by galleries at the base. Primitive stone implements have been obtained from various islands in the archipelago such as Borneo, Celebes, Timor, and i. The account of the pre-historic remains of Malay Peninsula is based mainly on "Papers on the Ethnology and Archaeology of the Malay Peninsula" by Ivor H. N. Evans M. A. (Cambridge, 1927). For a detailed account of the neolithic and palaeolithic implements, cf. R. O. Winstedt 'Pre-history of Malaya', in J. Mai. Br. R. A. S. 1932 pp. I ff, 30 PRE-HINDU CIVILISATION EST MALAYASIA Moluccas. These afford us the picture of a primitive people such as we meet with in other parts of the world before the dawn of civilisation. On the whole the actual archaeological finds in different parts of Malayasia lead to the conclusion that at the time of the first contact with the Hindus the people of Malayasia were in a primitive state of civilisation, and that in some regions they had not yet emerged from the state of barbarism. But the very fact that they had spread over so many different islands in the Archipelago forces us to admit that some of them had developed a high degree of skill in navigating the open sea, and it is only reasonable to hold that a people who could do this must have passed beyond the elementary stage of civilisation. Kern has made a serious attempt to form some idea of this civilisation. By a comparative study of the different Indo-nesian languages he has hit upon a number of roots or words common among them all. These may be reasonably regarded as having been in use when the Austro-nesian races lived together in Indo-China. With the help of these words, as well as by a study of those islanders who have been least affected by foreign intrusions, Kern has drawn a picture of the life led by the common ancestors of the peoples of Malayasia. It cannot, of course, be maintained that the civilisation which they had developed in Indo-China remained unaffected after they had moved to the various islands, for, according to local circumstances, it must have made further progress, or even received a set-back. But the picture of civilisation drawn by Kern may be regarded as a general background of our study. Without going into unnecessary and controversial details, we may give the following sketch of this civilisation on the authority of Kern. The Austro-nesians cultivated banana, sugarcane, cucumber etc. and were also acquainted with cocoanut and bamboo. Whether the cultivation of rice was known to the whole group is doubtful, but the section which peopled Malayasia SUVAENADVlPA 31 certainly acquainted with it Among other articles of food may be mentioned lobster, prawn and turtle, which they got from the sea. They tended buffaloes, pigs, and probably also cows, which were employed for cultivation and supplied them with meat and milk. Hunting and fishing were very popular with them, and they were acquainted with iron weapons. Their clothes were made of barks of trees and they knew the art of weaving. They built houses of bamboo, wood and rattan. About their intellectual attainments it may be mentioned that they could count upto a thousand and possessed an elementary knowledge of astronomy, indispensable for navigation in open sea. Their religious beliefs, like those of all primitive tribes, may be characterised as Animistic. Everything in nature which excited their curiosity or apprehensions and before which they felt themselves powerless to act, such as storm, thunder, earthquake, conflagration etc., were conceived as work of spirits who must be satisfied with proper worship. They also regarded trees, rocks, rivers and other natural objects as abodes of spirits. But the most important classes of spirits were those of the ancestors who were regularly worshipped and were supposed to exercise great influence on the lives of their descendants. The dead bodies were either thrown to the sea, or left in the forests, to be devoured by wild animals, or to undergo a natural decomposition. For it is only when the bones alone were left could the soul of the dead leave the body and go back to its proper realm, there to enjoy an eternal life very much in the same way as on the earth below 1 . It may be noted that the picture drawn above is in full accord with what we know of the primitive tribes in India. It i. The summary is taken from "Fruin-Mees", pp. 5-6. For the linguistic discussion on which it is based, Cf. V. G., Vol. VI, pp. 107-120. As to navigation and the knowledge of astronomy Cf. V. G., Vol. VI., p. 24. For general account cf. V. G., Vol. XV, pp. 180-81. 32 FEE-HINDU CIVILISATION IN MALAYASIA may also be reasonably held that the Indo-nesian settlers in Java and other islands, although mainly clinging to the old habits, introduced certain modifications therein. The monuments, described above, undoubtedly show that their method of disposing of the dead bodies had undergone a great change, and that they had made remarkable improvements in the art of stone- cutting. 1 It may also be presumed that the people of Java made further notable progress. It appears that the Javanese had developed various industries and excelled in making various articles of iron, bronze, copper, silver, gold, ivory, tortoise-shell, and horn of rhinoceros. It is to be remembered that tortoise and elephant are not to be found in Java and that gold, too, was found there only in small quantities. The work in ivory, tortoise-shell and gold, therefore, indicates active trade-relations with foreign countries from which they must have been imported. The rich fertility of the soil must also have made Java an emporium of grain. It is perhaps for this very reason that the Hindu traders who probably replenished their store of food from this fertile country on their way to China named the island 'Yava-dvlpa' or 'Island of Barley', a name which completely superseded in later times the indigenous name Nusa Kendeng. Thus we must hold that on the whole the Javanese possessed a high degree of civilisation. As to their religious beliefs and practices, the worship of spirits and ancestors seems to have played a dominant part in their every- day life. They built statues of these ancestors either of wood or stone, and also suitable temples to house them. A class of men called Zaman was believed to have been possessed of a peculiar faculty which enabled them to serve as a means of communication with the spirits of the ancestors. By suitable ceremonies in which dance, music, and burning of incense I. Fruin-Mees believes that some of the graves described above may belong to the primitive races who settled in Java before the immigration of the Indo-nesians (p. 7). This may be true, and in that case the Indo-nesian colonists in Java may be regarded as having learnt the art from these primitive peoples of Java. SUVARNADVlPA 33 formed the chief part, the Zamans became the medium through whose mouth the spirits of the ancestors gave their blessings to, and directed the undertakings of, their descendants. The Zamans also were, therefore, held in great veneration. The Javanese also made distinct progress in astronomy. They calculated a month of 30 days according to the phases of the moon, and their year consisted of 12 months. The year was again divided into two parts, ten months of work and two of rest. Five days, or rather nights, formed a unit, and two such units formed the week, of which there were thirty in the working period of ten months or 300 days. While admitting that the people of Java had attained to a much higher grade of civilisation than their neighbours, it is difficult to accept the highly exaggerated picture which is sometimes drawn of it. We may, for example, refer to the views of the great scholar Brandes who held that the pre-Hindu Javanese had the knowledge of the following. 1 1. The Wajang, a kind of shadow-play well-known in modern Java. 2. Gamelan, modern Javanese music accompanying Wajang. 3. Metre. 4. The art of weaving Batik cloth. 5. The metal industry. 6. Monetary system. 7. Sea-voyage. 8. Astronomy. 9. Cultivation by means of artificial irrigation. 10. State-organisation of a high order. A knowledge of some of these, for example, nos. 5, 7, and 8, may be accepted without discussion and has already been referred to. The others are, however, open to serious objection, i. T. B. G., Vol. 32 (iSCg), pp. 122 ff. 5 34 PBE-HINDU CIVILISATION IN MALAYASIA and Brandes' views in respect of them have been adversely criticised by eminent scholars. 1 As regards Wajang, I have discussed the question in some details in an Appendix to Bk. V., Chap. III. It is admitted by all that this has never been known to any other Indo-nesian tribe outside Java (except where it was imported in later times from Java), that we first come across it in Java when the Hindu colonists were established there for centuries, that similar play called Chaya-nataka was undoubtedly known to the Hindus, and that the plot of the earliest type of Wajang in Java is invariably derived from the Hindu epics. Against this it is pointed out that the technical terms in Wajang are Javanese and not Sanskrit, and that Wajang is very closely connected with the ancestor-worship of the Javanese. But it is to be remembered that when a people adopt a foreign custom, or import a foreign article, they not only sometimes give them their own names but also adapt them to their own peculiar needs. Although Wajang is closely associated with ancestor-worship in Java to-day, there is nothing to show that it always has been so. It stands to reason that when it secured wide popularity in Java, it came to form an essential element in the ancestor-worship which played such a dominant part in the life of Javanese people. It may be urged in favour of this view that although ancestor-worship is a characteristic feature of all or most Indo-nesian tribes, Wajang has never been known to form a part of it outside Java. Gamelan, which is essentially bound up with Wajang may, on similar grounds, be regarded as Javanese adaptation of an Indian original. As to Batik the researches of Rouffaer and Juynboll * have established the facts that the industry is not known to any other island outside Java (except where it was directly imported from Java) and that the first reference to 1. The observations that follow are mainly based on Krom Geschiedenis, pp. 45- 52. 2. De Batik-kunst in Ned.- Indie en haar geschiedenis (1914-) SUVAKNADVlPA 35 the industry in Java belongs to a very late period, while from a much earlier period India has been a well-known centre for the mass-production and wholesale export of the commodity. Here, again, the only argument for a Javanese origin seems to be that the technical terms are Javanese. As Krom has rightly pointed out, even to-day the Javanese give indigenous names to new articles imported from America and Europe, and hence no weight should be attached to arguments based on indigenous character of the name. As to Javanese metric and system of coinage, Brandes himself puts forward the claim with a great deal of hesitation, and Krom has pointed out that there is absolutely no evidence in support of it. What Brandes claims as Javanese metre, and Javanese coins proper, make their first appearance after the Indian metre and Indian coins had remained in use for centuries. According to Brandes, these undoubtedly later phenomena are developments of old pre-Hindu state of things. We have, however, as yet had no evidence that there was any metre or coin in the pre-Hindu period. Besides, even if there were any, we are to suppose, that they absolutely went out of use during the many centuries of Hindu influence, only suddenly to come to light after an obscurity of over thousand years. Nothing but the very strongest positive evidence would induce us to believe in such an explanation, and such evidence is lacking for the present. l As regards the last two points, cultivation by means of irrigation, and developed political organisation, Brandes bases his conclusions on the use of indigenous technical terms. As has been shown above, this is by no means a satisfactory evidence. On the other hand, the irrigation system was not unknown to the other Indo-nesian tribes and might well have developed independently in Java even prior to the Hindu I. Berg points out the close connection between the Javanese metrics and Javanese phonetics, and regards it as an evidence of the high antiquity of Javanese metre (Berg-Inleiding, pp. 67-69). 36 PRE-HINDU CIVILISATION IN MALAYASIA colonisation. As regards the state-organisation, we may well conceive that there was a certain political system, however rudimentary, though it is difficult to estimate the nature and degree of the organisation, as data for such estimate are lacking. Thus, of the ten points of Brandes, by which he tried to sum up the civilisation of the Javanese before they came in contact with the Hindus, Wajang, Gamelaii and Batik may be dismissed as improbable ; two others, metrics and monetary system, are most unlikely ; while two others, irrigation and highly developed state-organisation, are, at least, doubtful. The remaining three, viz., metal industry, sea-voyage, and elementary knowledge of astronomy, may alone be accepted as undoubtedly true. Chapter IV. SUVARNADVIPA SuvarnabhGmi (gold-land) and Suvarnadvipa (gold-island), as names of over-sea countries, were familiar to the Indians from a very early period. They occur in old popular stories such as have been preserved in the Jatakas, Kathakoa and BrhatkathS, as well as in more serious literary works, mainly Buddhist. Thus, according to a Jataka story, 1 prince Mahiijanaka sailed with some merchants in a ship bound for Suvarnabhumi, in order to get great riches there. Another Jataka story 8 refers to a sea-voyage from Bharukaccha to Suvarnabhumi. The same journey is described in great detail in the Supparaka- Jataka. 8 The original Byhatkatha is lost, but its stories have been partially preserved in the KathSsarit-sagara, Brhatkatha- maSjarl and Brhatkatha-61oka-samgraha. The Byhatkatha-lloka- samgraha gives us the story of Sanudasa, who sails for SuvarnabhGmi with a gang of adventurers, and undertakes a perilous journey by land after crossing the sea.* The Kathasarit-sagara contains a few more stories of the same type. First, we have the adventurous story of the great merchant 1. Jataka Vol. VI. p. 22. 2. Jataka Vol. III. p, 124. 3. Jataka Vol. IV. p. 86. Jatakamala No. XIV. Both give practically the same details of the journey, but the latter adds that the journey was undertaken at the instance of the merchants of Suvagnabhumi who had come to Bharukaccha, It may be inferred that Suvapnabhumi was the destination of the voyage. 4. Lacote Essai sur Gugatfhya et la Brhatkatha (pp. I75 ff ) English translation by Tabard, p. 131. See below, pp. 58 ff. 38 SUVAENADVlPA SamudraSdra, who sailed in a ship for Suvarnadvipa, for purposes of trade, and ultimately reached its chief city Kalasapura. 1 Another merchant, Eudra, was shipwrecked on his way back from Suvarnadvipa. 9 It also relates the story of ISvaravarma who went to Svarnadvlpa for the purpose of trade. 8 We have also references to trading voyage to Suvarnadvipa in the romantic story of YaSahketu.* There is, again, the story of a princess of Kataha being shipwrecked near Suvarnadvipa, on her way to India. 5 The KathakoSa relates the story of Nagadatta. Being anxious to go to a foreign land, in order to acquire wealth, he went on a sea-voyage with five hundred ships. His ships fell into the hollow of the snake-circled mountain and were rescued by the efforts of Sundara, king of Suvarnadvipa, who came to know of the danger of Nagadatta from a letter fastened to the foot of a parrot 6 Among the more serious works containing references to SuvarnabhQmi, we may refer, in the first place, to Kautilya's ArthaSastra (Book II, Chap. XI) which refers to Aguru (aloe) of Suvarnabhumi. The following passage in MilindapaSha makes an interesting reference to a few centres of the over- sea trade of India : "As a ship-owner, who has become wealthy by constantly levying freight in some sea-port town, will be able to traverse the high seas and go to Takkola or Clna or Suvannabhumi or any other place where ships do congregate." 7 1. Kathasarit-sagara (Taranga 54, verses 97 ff.) (Bombay edition of 1867, p. 276). 2. Ibid, Tarahga 54, vv. 86 ff. 3. Ibid, Taranga 57, vv. 72 ff. (p. 297). 4. Ibid, Taranga 86, vv. 33, 62. 5. Ibid, Taranga 123, v. no. 6. KathSkosa Tr. by Tawney pp. 28-29. 7. Milindapafiha, p, 359, Translated in S.B.E. Vol. XXXVI, p. 269. SUVAKNADVIPA 39 The Niddesa, a canonical work, also refers to sea-voyage to Suvarnabhflmi and various other countries. 1 The Mahakarma- Vibhanga illustrates des&ntara-vipaka (calamities of foreign travel) by reference to merchants who sailed to Suvarnabhumi from Mahakosali and Tamralipti. 8 We may next refer to the Ceylonese Chronicle Mahavamsa which describes the missionary activities of Thera Uttara and Thera Sona in Suvarnabhtlmi. 3 The Mahakarma-Vibhanga attributes the conversion of Suvarnabhumi to Gavampati. The voyage of Gavampati to Suvarnabhumi is also related in the Sasanavamsa.* We learn from Tibetan sources that Dharmapala (7th cent. A. D.) and Dlpankara Atisa (llth century A. D.) visited Suvarnadvipa. 5 The name and fame of Suvarnabhumi and Suvarnadvipa travelled far beyond the boundaries of India, and we find reference to both in Greek, Latin, Arabic, and Chinese writings. Pomponius Mela was the first to refer to the island of Chryse (gold)- a literal translation of Suvarnadvipa in his 'De Chorographia', written during the reign of the emperor Claudius (41-54 A. D.). 6 The Chryse island is referred to in the Periplus of the Erythraean Sea (1st century A. D.) 7 , and is mentioned by Pliny (c. 77 A. D.) 8 , Dionysius Periegetes (2nd. cent. A. D.), 9 Solinus (3rd. cent. A. D.), 10 1. This passage is discussed below, pp. 56 ff. 2. Mahakarma-Vibhanga Edited by S. Levi p. 50 ff. 3. Geiger Mahavarftsa, p. 86. 4. Mahakarma-Vibhanga p. 62 ; Sasanavamsa, p. 36. 5. Sarat Chandra Das, Indian Pandits in the Land of Snow, p. 50 ; Kern Manual of Buddhism, p. 130. 6. Coedes Textes, p. 12. 7. Schoff's Translation, pp, 45-48. 8. Coedes Textes, p. 15. 9. Coedes, Textes, p. 71. The date of Dionysius is given as second century A.D. by Coedes. Tozer in his History of Ancient Geography (p. 282) assigns him to the first century A.D. jo. Coedes Textes, p. 86. 40 SUVAKNADVIPA Martianus Capella (5th. cent. A. D.), 1 Isidore of Seville (7th. cent. A. D), a the anonymous author of Cosmography (7th cent. A. D.), s Theodulf (8th. cent. A. D. ),* and Nicephorus (13th. cent. A. D.) 8 , in addition to several authors who reproduce the information given by Dionysius Periegctes.* Ptolemy (2nd. cent. A. D.) does not refer to the island of Chryse, but mentions, instead, Chryse Chora a literal transla- tion of Suvarnabhfimi and Chryse Chersonesus, or Golden Peninsula. 7 The Chryse Chersonesus was evidently known to Marinos of Tyre 8 (1st. cent. A. D.) and is mentioned by Marcien (5th. cent. A. D.). e The only other writers who refer to it ar'; Eustathios (12th. cent. A. D.) 10 and Etienna (6th. cent. A. D. ) X1 who quote respectively Ptolemy and Marcien. Flavius Josephus (1st. cent. A. D.) refers to Chryse as a land in India and identifies it with Sophir. 1 * The Indian tradition of Suvarnadvipa was also known to the Arabs. Albenini refers to both Suvarnadvipa and Suvarnabhumi. "The islands of the Zabaj," says he, "are called by the Hindus Suvarnadvipa i. e. the gold islands". 18 1. Ibid, p. 116. He writes the name as Chrysea. 2. Ibid, pp. 136-137. 3. Ibid, p. 149. He uses the form 'Chrisi*. 4. Ibid, p. 150. 5. Ibid, pp. 160-161. 6. Etienne (6th. cent. A.D.), Eustathios (i2th cent. A.D.) ; cf. Coedes Textes, pp. 132. 157. 159- 7. Coedes Textes, pp. 38-43, 53. 56, 60, 66. 8. Ptolemy refers to Marines' estimate of the distance between Tamala and Chryse Chersonesus (Coedes Textes, p. 38.) 9. Coedes Textes, p. 118. 10. Ibid, p. 160 11. Ibid, p. 132. 12. Ibid, pp. l7-*8. 13. Sachau's Transl. Vol. I, p. 210. 'Zabaj* is also written as Zabag. SUVARNADVIPA 41 Elsewhere he says : "The islands of the Zabaj are called the Gold Country because you obtain much gold as deposit if you wash only a little of the earth of that country". 1 Although the translator of Alberuiii has put, within brackets, Suvarnadvlpa after the expression, 'Gold Country', the phrase used by Alberuni is undoubtedly equivalent to Suvarnabhumi, rather than Suvarnadvlpa. In another place Alberuni has included Suvarnabhunii in the list of countries in the north-east, as given in Brhat-Samhita. 2 Many other Arab writers refer to Zabaj as the 'Golden land' or 'land of gold'. Among them may be mentioned Haraki (died 1138 A.D.) 3 , Yakut (1179-1229)*, Sirazi (died 1311 A.D.) 6 , and Buzurg bin Sahriyar 6 . Nuwayri (died in 1332 A.D.) calls Fansiir (Pansur or Baros on the western side of Sumatra) as the land of gold 7 . It may be noted also that Buzurg bin Sahriyar in one place calls Mankir the capital of the land of gold, though in other places he evidently follows the Arab tradition of identifying it with Zabaj 8 . The name Suvarnadvlpa was also not unknown to the Chinese. I-tsing twice mentions Kin-tcheu (gold-island) in his famous "Memoir on the pilgrimage of monks who visited the western countries in search of law", and uses it as a synonym of Che-li-fo-che or Sri-Vijaya. Having thus rapidly surveyed the wide prevalence of the knowledge of Suvarnadvipa and Suvarnabhumi in many 1. Ibid, Vol. II., p. 106. 2. Ibid, Vol. I, p. 303. 3. J. A., Vol. CCII, p. 6. 4. Ibid, p. 7. 5. Ibid, pp. 8-9. 6. Ibid, pp. 10-12. The date of this author is uncertain. Van der Lith places him in the loth century A.D., but Ferrand doubts it (Ferrand Textes Vol. II. pp. 564-5). 7. J.A., Vol. CCII., p. 9. 8. Ibid, pp, 10- 1 1. 9. I-tsing Memoire (pp. 181, 187, p. 36, f.n. 3.) 6 42 SUVAENABVlPA countries, extending over many centuries, we may now proceed to discuss in detail its precise location and antiquity of its colonisation by the Hindus. It is a striking fact that the contrast between SuvarnabhQmi and Suvarnadvipa, i.e., the Gold-land and the Gold-island, which we meet with in the Indian sources, is also faithfully reflected in the nomenclatures used by the western authors, some of them calling it an island, and the others, cither a land or a peninsula. Ptolemy, as we have seen, refers to both Chryse Chora (golden land) and Chryse Chersonesus (Golden Peninsula). He distinguishes them * as two different regions, evidently lying close to each other, as both of them adjoined Besyngeitai. We note a similar distinction even in the Periplus of the Erythraean Sea. In para 56 of that work Chryse is called an island. In para 63, however, Chryse is referred to both as a 'land' near the Ganges, and 'an island' just opposite that river. Alberuni, as we have seen above, also uses both Suvarna- dvipa and Suvarnabhumi. Gerini was perhaps the first to give serious attention to this contrast. As he has drawn very important conclusions from this, we may quote his remarks at some length. 1 "Marinos of Tyre and Ptolemy are the first to speak of the Malay Peninsula as the Golden Khersonese. The geographers that preceded them, among whom Eratosthenes, 8 Dionysius Periegetes, and Pomponius Mela may be named, all refer to it instead as Khryse or Chryse Insula : the "Golden Isle", and so does long before them the Ramayana, under the name of Suvarnadvipa, which conveys the same meaning. No stress has, so far, been laid on this wide difference in representing 1. Gerini Researches, pp, 77-/8. 2. This is misleading. Gerini himself remarks elsewhere : "Doubtless Eratosthenes had heard of them (Chryse and Argyre) although no allusion in that sense is likewise met with in the surviving fragments of his work." (Ibid, p. 670 f.n. i). SUVABNADVlPA 43 that region on the one part as an island and on the other as a peninsula. I believe, therefore, that I am the first to proclaim, after careful consideration, that both designations are probably true, each in its own respective time ; that is, that the Malay Peninsula, or rather its southern portion, has been an island before assuming its present highly pronounced peninsular character. The view I now advance is founded not only on tradition, but also upon geological evidence of no doubtful nature." Gerini then proceeds with the details of what he calls the geological evidence. Gerini's explanation, however, cannot be seriously considered. In the first place, it is to be noted that the word l dvlpa* means primarily *a land having water on two of its sides'. Thus 'dvlpa* is not identical with 'island', and includes peninsulas and sometimes Doabs also. 1 As the foreign writers got their information from Indian source, they might have taken 'dvlpa' in the sense of 'island', whereas it was really a peninsula. Further, it is a well-known fact, that ancient sailors often represent one and the same country as consisting of a number of separate lands or islands ; for, as the journey was made from one port to another by open sea, the continuity of the region was always a difficult matter to ascertain. The Arabs, even down to a late period, represented Sumatra as consisting of a number of separate islands. As to Malay Peninsula, the subject of Gerini's discussion, Chavannes has pointed out that the Chinese geographers of the T'ang period regarded it as a series of islands. 8 The real point of contrast, missed by Gerini, is the reference by one and the same author to two regions called Chryse, one of which is mainland, and the other, an island or peninsula. As we have seen above, this is the case with Ptolemy and the i. Cunningham's Ancient Geography Edited by S, N. Majumdar, Appendix I, p. 751. 3. I-tsing Memoire, (p. 36. f.n, 3), 44 SUVARNADVIPA author of the Periplus. The question, therefore, naturally arises, whether we should take Suvarnabhumi and Suvarna- dvipa as corresponding exactly to these two regions, both called Chryse by the western authors, one denoting a portion of the mainland (bhQmi), and the other, an island or a peninsula (dvipa). However tempting such a solution might appear at first, we must definitely reject it. As we have seen above, Albenmi applies the term Suvarnabhumi to the islands of Zabag which he elsewhere designates Suvarnadvlpa. Besides, the island of Sumatra, which is called Suvarnadvlpa in Chinese sources and is undoubtedly referred to by this name in later Indian literature, is designated as Suvarnabhiimi in an inscription found in the island itself. 1 It is thus quite clear, that the term 'bhiirni' in the compound 'Suvarnabhumi' should not be taken in the sense of mainland, as opposed to island or peninsula, but simply in the general sense of land or territory. This brings us to the question of the exact meaning of the term Suvarnabhumi. Pomponius Mela explains the name Chryse (gold) island by referring to an old tradition that the soil of the country is made of gold. Ho adds that cither the name is derived from this legend, or the legend is invented from the name. In any case he took Suvurimbhilmi to signify 'the country whose soil was gold'. This view was shared by a large number of ancient writers, 2 but Pliny takes a more rational view. Referring to Chryse he says : "I think the country abounds in gold mines, for I am little disposed to believe the report that the soil of it is gold/' Pliny's view is upheld by later authors, though some of them refer to the wide-spread tradition of the soil being gold. Dionysius Pcrtegetes seems 1. Een Sumatraansche inscriptic van koning Krtanagara by N. J. Krom (Vers. Med, K. Akad. Weten Lctterkunde 5* reeks deel II. 1916. pp. 306-339) reproduced in J. A. ii-XX, pp. 179-80. 2. According to Isidore of Seville the view was held by a majority of authors (Coedes Textes, p. 137). SUVAKNADVlPA 45 to explain the name as due to the strong rays of the sun which makes the soil look like gold. 1 Among the Arab writers also, Harakl and Yakut take the view that the soil is gold, while Albcruni attributes the name to the fact that the country yields a large quantity of gold. There is hardly any doubt that the old tradition of the golden soil was derived from India. For the Puranas actually refer to a country, outside Bharatavarsa, the mountain and soil of which consist of gold 3 , and Divyavadfina describes in detail the difficulties which one has to surmount in order to reach that region of the earth where the soil is gold 3 . There is equally little doubt that the origin of the name Suvarnabhumi has to be traced to this belief, though a rational explanation was substituted afterwards. The word bkUmi in Suvarnabhuim, therefore, originally stood for soil or land in general, and there was no idea of contrasting it with \lvlpa', island or peninsula. It may be noted here, that we have also reference to cities called Suvarnapura. In an illustrated Nepalese manuscript, a picture is entitled "Suvarnnapure Sri-Vijaya-pure Lokanatha" or (the image of) Lokanatha (AvalokiteSvara) in Sri-Vijaya-pura in Suvarnnapura. Srl-Vijaya is the old name of a capital city in Sumatra. So Suvarnapura should be located there, and seems to be used as a designation for a region, rather than a town*. The Kathasarit-sagara also refers to Kaficanapura, a synonym of Suvarnapura, where the merchant ISvaravarman stopped on his way to Suvarnadvlpa 5 . In Sana's Kadambarl also we get a reference to "Suvarnapura, 6 not far from the eastern ocean and the abode of the Kiratas, 1. CoedesTextes, p. 157. pp. 7i"73 2. Cf. e.g., Matsya Purana. Ch. 113, vv. 12, 42. 3. Mahantam Sauvarnabhumim prthivipradesarh (Divyavadana Cowell, p. 107). 4. J. A. ii-XX, pp. 42-43- 5. Tarahga 57, v, 76. 6. Kadamvarl-Tr. by Ridding, pp. 90-91. 46 SUVARNADVlPA Thus, in addition to the generic name Suvarnabhdmi, or gold- land, we have references to gold-island, gold-peninsula, and gold-city. It seeins to be quite clear, therefore, that Suvarna- bhumi was used primarily as a vague general designation of an extensive region, but, in course of time, different parts of it came to be designated by the additional epithets of island, peninsula or city. The original name, however, never went out of use altogether, for we definitely know that, even at a much later period, it used to denote Sumatra and portions of Burma. In order to have a general idea of the extent of the region to which the name Suvarnabhumi was applied, it is necessary to make a list of territories which we know on definite grounds to have borne that name in its primary or derivative form. The Periplus makes it certain that the territories beyond the Ganges were called Chryse. It does not give us any means to define the boundaries more precisely, beyond drawing our attention to the facts, that the region consisted both of a part of mainland as well as an island, to the east of the Ganges, and that it was the last part of the inhabited world. To the north of tliis region it places This or China. In other words, Chryse, according to this authority, has the same connotation as the Trans-Gangetic India of Ptolemy, and would include Burma, Indo-China, and Malay Archipelago, or rather such portions of this vast region as were then known to the Indians. Ptolemy's Cliryse Chcrsonesus undoubtedly indicates the Malay Peninsula, and his Chryse Chora must be a region to the north of it. Now, we have definite evidence that a portion of Burma was known in later ages as Suvarnabhumi. According to the Kalyani inscriptions (147G A.D.), RamaSSadesa was also called Suvannabhumi 1 , which would then comprise the maritime region between Cape Negrais and the mouth of the Salwin. i. Suvanijabhumi-ratta-saiiikhata Ramaflfiadesa ( Ind. Ant. Vol. 22. 1893, p. 151). SUVARNADVIPA 47 According to Po-U-Daung Inscription (1774 A.D.), 'SuvannS- paranta, a designation usually syncopated into Sunaparanta or Sonnaparanta, included the country between the Lower Iravati and Chindwin and the Arakan Yoma. Now, AparSnta means 'western end or extremity', and hence the region denoted as Suvannaparanta may be taken to denote the western end or extremity of Suvarnabhumi. Thus these two place-names would authorise us to apply the name Suvarnabhumi to a large portion of Burma, both maritime and inland, and this would also suit the location of Ptolemy's Chryse Chora 1 . There can also be hardly any doubt, in view of the statement of Arab and Chinese writers, and the inscription found in Sumatra itself, that that island was also known as Suvariia- bhumi and Suvarnadvlpa. Ferrand points out that even now Sumatra is designated by the Malays as Pulaw Ernsts or the island of gold (Suvarnadvlpa). But the Arab writers definitely imply that Suvarnadvlpa included a number of islands. Alberuni is quite clear on this point. "The islands of the Zabaj", says he, "are called by the Hindus Suvarnadvlpa, i.e., the gold islands". Ibn Said 2 ( 13th century A. D. ) definitely asserts that Zabag is an archipelago consisting of a large number of islands which produce excellent gold, and says that Sribuza (Sri-Viyaya in Sumatra) is the greatest of the islands of Zabag. The same view is implied by other Arab writers both before and after him. Thus, strictly speaking, the name Suvarnadvlpa is applied by the Arabs, on the authority of the Hindus, to 1. Gerini Researches pp. 64 ff. There does not seem to be any adequate reason for excluding the maritime region, as Gerini has done, in locating Chryse Chora. Of course we must always bear in mind that it is a fruitless task to attempt to define the exact location of Ptolemy's place-names beyond a general indication such as we have given above. 2. For the account of this and other Arab writers, cf. Ferrand J. A. ii XX, pp. 52 ff. 48 StVARNADViPA a large group of islands, roughly corresponding to Malay Archipelago of the present day. Even as late as the sixteenth century A.D., Budhagupta, a Buddhist monk, visited two islands called Suvarnadvipa in the Eastern Sea. 1 There are thus definite evidences that Burma, Malay Peninsula, and Sumatra had a common designation of Suvarnabhumi, and the name Suvarnadvlpa was certainly applied to Sumatra and other islands in the Archipelago. This does not, however, take away the possibility of other territories being designated by the one or the other name. Thus, on the whole, we shall not perhaps be far wrong, if we take Suvarna- bhumi and Suvarnadvlpa as general designations of Burma, Malay Peninsula, and Malay Archipelago, as hinted at in the Periplus. a But, keeping in view the literal meaning of the word dclpa, we should restrict the use of the name Suvarnadvlpa to the last two alone. We shall now proceed to discuss briefly some of the important localities in Suvarnadvipa which were definitely known to the Hindus in ancient times. As we have seen above, there is a reference to the island of Chryse (gold) in the Periplus. This is associated with the island of Argyre (silver) by many other classical authors, such as Pomponius Mela, Pliny, Solin, Martianus Capella, Isidore of Seville, and Theodulf. The origin of the name 'silver island' is explained in the same way as that of the 'gold island'. This close association naturally induces us to look for the two islands near each other. Now, as the name Suvarnadvipa, for the island of Sumatra, is well established, we might look 1. 1. H. Q., Vol. VII. (1931)1 pp. 698, 701. 2. The Chinese used the name 'Kouen louen' to denote Indo-China and Indo-nesia as a whole. Recently S. Le*vi has shown that in a Chinese- Sanskrit Dictionary this term is rendered by Sanskrit Dvipantara, which therefore means, not 'another island' as has been generally understood, but the 'Far East'. ( B. K. I. Vol. 88, 1931, pp. 621-627). SUVARNADVIPA 49 upon the island of Java as corresponding to Argyre, and there are several facts which speak in favour of this supposition. It is somewhat singular that Ptolemy does not refer to the large island of Sumatra, at least under any easily cognisable name. The fact seems to be, that, like the later Arab writers, he regarded it as a series of islands, which he called (1) the group of five islands, the Barousai, and (2) the group of three islands, the Sabadeibai. Next to Sabadeibai Ptolemy places "the island of labadios (or Sabadios) which means the island of barley." It is said to be of extraordinary fertility and to produce very much gold, and to have its capital called Argyre (Silver-town) in the extreme west of it/ The explanation of the name leaves no doubt that Ptolemy's labadios corresponds to Sanskrit Yavadvlpa. 1 Now, Ptolemy nowhere refers to the islands of Chryse and Argyre which figure so prominently in the writings of other western geographers both before and after him. 8 His Chryse Chersonesus may possibly represent the Chryse island of other writers, but we cannot say anything definitely on this point. The reference to an island with capital called Argyre, which is not far from his Chrysc Chersonesus, and situated quite close to Sumatra that undoubtedly bore the name Suvarna- dvipa (equivalent to Chryse island), justifies us, therefore, in identifying labadios as the Argyre island of other writers. Thus the islands of Chryse and Argyre, referred to by classical writers, would correspond to the well-known islands of Sumatra and Java or the Malay Peninsula and Java. This view is in full accord with what we find in the Periplus. The author thus describes the coastal regions of Bengal. "After these, the course turns towards the east again, and sailing with the ocean to the right and the shore remaining beyond to the left, Ganges comes into view, and near it the very last land 1, For further discussion see Chaps. VI-VII. 2. See ante. For further discussion see Chap. VI. 7 50 SUVARNADVlPA towards the east, Chryse. There is a river near it called the Ganges And just opposite this river there is an island in the ocean, the last part of the inhabited world towards the east, under the rising sun itself ; it is called Chryse." Now, although the island of Sumatra or the Malay Peninsula is at a great distance, it is undoubtedly opposite the Ganges, in the sense, that if one sails straight towards the south from the mouth of that river, he would reach the island or the peninsula direct without coming across any other land. That the author meant a somewhat remote region is indicated by the expression, 'under the rising sun itself/ It may be a vague reference to the equatorial region, but, in any case, seems to indicate a sufficiently remote locality. It is also interesting to note that Pliny and other writers locate the islands of Chryse and Argyre as simply 'beyond the mouth of the Indus river/ Thus the expression, "opposite the Ganges", should not be understood in the sense in which we would employ it to-day, but in a general way only, and Malay Peninsula or Sumatra corresponds to the position fairly well. Apart from the general and somewhat vague use of Suvarnadvlpa, we may trace in Indian literature references to various localities within that region. The earliest reference of this kind, though equally vague in character, perhaps occurs in the ninefold division of Bharatavarsa as given in the Puranas. It has been argued with great plausibility that of these nine divisions, one alone corresponds roughly to India proper, and the other eight, therefore, designate other parts of what may be called Greater India. Mr. S. N. Majumdar who propounded this idea definitely identified one of these divisions, Indradvlpa, with Burma, and suggested that another, Kaserumat, might be Malay Peninsula. 1 But the question is not free from difficulties as has been pointed out by Mr. S. B. Chaudhury.* 1. Cunningham Ancient geography of India, Edited by S. N. Majumdar, Appendix I. Sp. cf. pp. 752-754. 2. Ind. Ant, 1930, pp. 204 ft. SUVARNADVIPA 51 But when the later Puranas like Garuda and Vamana substitute Katsha and Simhala for Saiunya and Gandharva of the other Puranas 1 , we have a definite reference to a region in Malay Peninsula, for Kataha is the well-known name of the locality now represented by Keddah. 3 The name 'Kataha-dvlpa' which was thus raised to the dignified position of one of the great divisions of Bharatavarsa or Greater India may be taken as roughly denoting the same region as Suvarnadvipa, which name is entirely absent from the Puranas. In course of time, however, both the names came to be applied to particular localities. The Kataha-dvlpa figures prominently in the Kathasarit-sagara, as a rich and flourishing country, but is distinguished from Suvarnadvipa ; for, as already noted above, a story relates how the princess of Kataha-dvlpa was ship-wrecked near Suvarnadvipa on her way to India. The same story tells us that the mother of the princess was the sister of the king of Suvarnadvipa. 8 This leads to the conclusion that the Kataha-dvlpa and Suvarnadvipa were situated close to each other, but we must not count too much upon exact geographical knowledge of a story-writer. We have also the famous story of Devasmita, in which her husband, the merchant Guhasena, sails from Tamralipti to Kataha, and she follows him there after a short period. 4 The story of the foolish merchant also leads us to Kataha. 5 Geographically, the most interesting story in this connection is that of Candrasvamin who lost his son and younger sister in the wood. They were supposed to have been rescued by a merchant named Kanakavarman. Having learnt that the 1. Ibid. The verse in the two Puranas runs as follows : Nagadvipalj Katahagca Sirhhalo Varunastatha I (Garuda Purana Ch. 55-V. 5, Vamana Purana Ch. 13, V, 10.) 2. See Book II. Chap. II. 3. Tarahga 123, vv. 105 ff. 4. Tarafiga 13, vv. 70. ff. . Taranga 61, v. 3. 52 SUVARNADVlPA merchant had sailed for Narikela-dvipa, Candrasvamin embarked in a ship and went across the sea to that island. There he learnt that Kanakavarman had gone to Kataha-dvlpa. Candra- svamin followed him there, only to learn that Kanakavarman had gone to Karpura-dvlpa. In this way poor Candrasvamin visited in turn Narikela-dvlpa, Kataha-dvlpa, Karpura-dvipa, Suvarnadvipa and Simhala-dvlpa. 1 The Narikela-dvlpa is mentioned both by Chinese and Arab writers. 8 According to Hiuen Tsang the people grew no grain but lived only on cocoaimts, which evidently gave the name to the island. He places it 'thousands of li' to the south of Ceylon. It has been identified with Nicobar island. 3 The Karpdra-dvlpa is also named by Arab writers.* It is either Borneo or north (specially the north-west) side of Sumatra, where lies the port Boms from which to this day the Malays name the true camphor, Kapur Barus. Blagden considers this latter identification as more probable. 5 A similar knowledge of the islands in the Archipelago may be traced even in the Puriinas. 4"he Vflyu Parana contains a chapter describing the various dclpas to the south of India. 6 Although there is much that is imaginary or mythical, there seems to be a kernel of fact. It describes in particular a group of six islands named Anga-dvipa, Yama-dvlpa, Malaya-dvlpa, Sankha-dvlpa, Ku5a-dvipa and Vnraha-dvipa. 1. Taranga 56, VV. 54 ff- 2. Narikela-dvTpa is mentioned, among others, by Hiuen Tsang (Beal Vol. II, p. 252) and Ibn Said (i3th cent. A. D.). The latter also refers to it as a dependency of Ceylon (Ferrand-Textes Vol. II, p. 339). 3. For the identification and other details, cf. Yule- Marco Polo, Book III, Chap. XII notes. Beal identifies it with Maldive islands Beal, Vol. II, p. 252, f.n. 36. 4. Ibrahim bin Wasif Sah (c. 1000 A. D.), Ibn Ai Wardi (i4th cent,), Thousand and One Nights ; cf. Ferrand Textes, pp. 157, 422, 570-573- 5. Cf. N.M. Penzer's note in his edition of Tawney's English Translation of Kathasarit-sagara - Ocean of Stories, Vol. IV, p. 224. fn. i 6. Chapter, 48. SUVABNADVIPA 53 Among these, Malaya-dvlpa may be identified with Malay Peninsula. Malaya-dvlpa is described as producing precious stones, gold, and sandal, and this suits well with Malay Peninsula. Besides, reference is made to the city of Lanks, which may be identified with Lenkasuka (see Chap. V). The Sankha-dvipa may be identified with the island of Sankhay, frequently mentioned by Arab writers. According to them it was three days' voyage from Malaya and was included within the empire of Sri-Vijaya. It gave the name to the neighbouring sea, and there was also a town called after it. 1 The Anga-dvipa may be identified with the Angadiya of the Arab writers, which is named immediately after a place on the Siamese coast and is located in the Bay of Bengal.* In the group of three islands, named Barawa by the Arab writers, we can easily recognise the Varaha-dvipa of the Vayu Purana. These islands are placed about 100 farsangs from Fansur, i.e., Baros on the coast of Sumatra. 3 The Yama-dvipa may be the same as Yamakoti, which was regarded as being 90 to the east of Laftka.* Now, even admitting that these identifications are merely conjectural, the fact that most of these names are mentioned by Arab writers as names of islands in the Archipelago is not without importance. It certainly leads to the presumption that the Puranik writers had some real knowledge of the Malay Peninsula and Indian Archipelago, although they embellished their accounts with a great deal of mythology and fiction. Another indication of the geographical knowledge of the Hindus regarding the Far East is supplied by a passage in the Ramayana. M. Sylvain Levi has pointed out that this passage served as the basis of similar geographical accounts 1. Edrisl, (194), Ibn Said (346), Dimaski (377, 381) and Nuwayri (395). The figures within brackets refer to pages of Ferrand Textes. 2. Sidi al-Celebi (Ferrand Textes, p. 523). 3. Ferrand Textes, pp. 583-4. 4. Sachau Alberuni, Vol. I, p. 305. 54: SUVAKNADVlPA in HarivarhSa and the Buddhist Sfltra called Saddharma- SmytyupasthSna. For a critical study of the passage in all its bearings, we must refer the reader to the original article of that scholar 1 . Here we shall content ourselves with only a few points, relevant to our present study, which emerge clearly from his scholarly discussion. The most important passage runs as follows : Yatnavanto yavadviparii sapta-rajyopagobhitam I Suvarnarupyakadvipam Suvarn&karamanditam. li Unfortunately this passage appears in radically different forms not only in the different texts of Bamayana but also in HarivamSa, RSmayana-maSjar! of Ksemendra and the Saddharraa-Smrtyupasthana-Sutra which reproduce it. Thus Yavadvlpa appears only in the Bombay edition ; the Bengali edition substitutes jaladvlpam, whereas the other parallel passages omit it altogether. Similarly the first and the third words in the first line, given above on the authority of the Bombay version, are replaced respectively by 'ratnavantam' and 'phalabhojyopaSobhitam'. The reading 'Yavadvipam' is undoubtedly to be preferred, but we are less sure about the two others. Thus we cannot be quite certain if 'Yavadvlpa' was adorned with seven kingdoms as the Bombay text informs us. As to the second line, ' Siivar\ianipydka' appears as the name of a separate island in the Bengali version, but Ramayana-maiijarl and HarivamSa substitute Stivarnakitdyaka. The latter reading is supported by the Snrrtyupasthana-Siitra, as both the Chinese and the Tibetan translations of the passage render the name as 'island called 'wall of gold', an exact translation of Suvarnakudya-dvlpa. Now, Suvarnakudya is mentioned thrice as the name of a country in Kautillya ArthaSastra*, and on this ground Ldvi has preferred this reading. He takes as equivalent to this j. J.A.n-XI.,pp. 5-160. 3. Book II. Chap. XI. SUVARNADVtPA 55 name, the Chinese Kin-lin by which they designate a country, 2000 li to the west of Fou-nan (Cambodia), and situated along a bay 1 . This would locate it in the Malay Archipelago. It is to be noted here that Siivarna-rnpyaka-dvlpam is an exact equivalent of the island of Chryse (Suvarna) and Argyre (rupyaka-stiveT) of the classical writers. Further, the Smrtyupasthana-Sutra says that the soil of the island which it calls Suvarna-kudyaka is gold. This supports the reading Suvarnakaramanditam which we get in the Bombay version and Harivama, but which is replaced by Ganadvlpam, a third island, in the Bengali version. If we accept this reading, we may have here the origin of the classical tradition about the Chryse island referred to above. On the whole it seems that we have here a reference to both a gold and a gold-cum- silver island, though the two have been confused. The next important passage, which is practically identical in both the versions of Ramayana, runs as follows : AmamlnaSanaScapi kirata dvlpavasinah I antarjalacara ghora nara-vyaghra iti smytjltt II The Ramayana-MaSjarl of Ksemendra substitutes the following ; "antarjalacaran ghoran samudradvlpasamSrayan, Thus the same two adjectives arc applied, in the one case to the Kiratas, and in the other, to the people of Samudra-dvipa. As the Kiratas have already been described in the preceding passage in Ramayana, the reading of Ramayana-manjarl is preferable. In any case it presents a new name Samudra-dvipa. Now, this may mean either 'island of the sea' or the 'island called Samudra/ The first meaning is, of course, pointless, so we may take the second and find in it a reference to Samudra, which, being corrupted to 'Sumutra', has given the name Sumatra to the great island in the Archipelago. It is interesting to note in this connection that Kautilya's Arthafifistra refers to a country called 'Para-samudra', and i. fetudes Asiatiques, Vol. II, p. 36. 56 SUVARNADVlPA another, called 'Pasa/ 1 These may be taken as referring to the two neighbouring states of Samudra and Pase, in the north of Sumatra, to which frequent reference is made by mediaeval writers. 4 Further, the geographical chapter under discussion describes various barbarians in the Eastern Sea more or less in the same way as Ptolemy has done in his account of the islands in the Archipelago. The Buddhist writers show a more extensive knowledge of the countries in the East. The Milindapafiha, as we have seen above, refers to Suvannabhumi, Takkola and Cma. The Mddesa, commenting on the word "torment" in the Sutta- nipata, describes the various kinds of torments which a sailor experiences, while, overpowered by desire for wealth and enjoyment, he sails in high seas in a boat and goes to (1) Gumba, (2) Takkola, (3) Takkasila, (4) Kalamukha, (5) Maramipara, (6) Vesunga, (7) Verapatha, (8) Java, (9) Tamali, (10) Vauga, (11) Elavaddhana, (12) Suvannakuta, (13) Suvannabhumi, (14) Tambapanni, (15) Suppara, (16) Bharukaccha, (17) Surattha, (18) Anganeka, (10) Gangana, (20) Paramagangana, (21) Yona, (22) Paramayona, (23) Allasanda, (24) Marukantara, (25) Jannupatha, (26) Ajapatha, (27) Mendhapatha, (28) Saiiku- patha, (29) Chattapatha, (30) Vamsapatha, (31) Sakunapatha, (32) Masikapatha, (33) Daripatha, (34) Vettadhara (or Vettacara). This interesting passage has been the subject of a learned dissertation by M. Sylvain Levi ; and the readers are referred to his scholarly article for a detailed discussion of the various points arising out of it. 8 Its chief importance, 1. Book II. Chap. XI. 2. The name Para-Samudra is explained as Ceylon in a late commentary to which no importance should be attached. It places Suvarnakutfyaka in Assam. Dr. H. C. Raychaudhury has supported this identification by equating Para-Samudra with Palaesimundu of the Periplus. But the equation Palaesimundu Para-Samudra is not very obvious. 3. Etudes Asiatiques, Vol. II, pp. i-5$ 431. SUVARNADVlPA 5? for our purpose, is the very comprehensive view it offers of the sea-going trade in ancient India. It describes twenty-four localities (Nos. 1-24) which the merchants visited by way of sea, and ten difficult routes (Nos. 25-34) which they had to follow on land, apparently after reaching the harbour on the sea- coast. Of the twenty-four localities, Nos. 15 to 24 evidently belonged to the western side of India and do not concern us here. Suvannabhumi (No. 13), Vesunga (6), VerSpatha (7), and Takkola (2) correspond to Ptolemy's Chryse Chora, Besyngeitai, Berabai, and Takkola, the first mart in the Chryse Chersonesus. As such, Suvannabhami may be located in Burma and the same is perhaps true of Suvannabhumi, mentioned in the MilindapaSha along with Takkola. The Kalamukha (4) is mentioned as the name of a tribe both in Ramayana and Mahabharata, and the country is to be placed on the Arakan coast. Java (8) can be easily identified as the well-known island of Java. Tamali (9) is the same as Tambralinga, referred to in a Sanskrit inscription discovered at Caiya in Malay Peninsula, and has to be placed in that region. 1 Suvannakuta (12) has been equated by Levi with Suvarnakudyaka which we have already discussed above. Tambapanni (14) is, of course, Ceylon. Gumba (1), Maranapara (5), and Elavaddhana (11), are not known from any other source and cannot be identified for the present. There remain now Takkasila (3) and Vanga (10) which are both well-known places, one in the north-west, and the other, in the eastern part of India. But the usual identi- fication of Takkasila with Taxila would be somewhat incongruous, as the place is named in a list of trans-Gangetic countries to the east. While, therefore, nothing definitely can be said in this matter, Levi has drawn our attention to the river Tokosanna, mentioned by Ptolemy, near Arakan coast, and the Takkasila of the text may be located here. The identification of Vanga with Bengal seems equally i. Cf, Book II. Chap. II. 8 58 SUVAENADVlPA objectionable, particularly when we remember that it is both preceded and followed by other places in Malay Peninsula and the Indian Archipelago. Lvi has pointed out that the Manuscripts also give an alternative reading vankam, and this can be easily identified with the island of Banka to the east of Sumatra. The list of Niddesa thus practically covers a large part of the region which we have named Suvarnabhumi and Suvarnadvlpa, and of all the Indian texts available to us it shows the most detailed knowledge of the oversea centres of trade in the East. Levi has drawn attention to the points of agreement between this list and that given by Ptolemy, and has drawn the conclusion that both must belong to approximately the same period. The knowledge of the Far East possessed by Pliny and the author of the Periplus makes it highly improbable that such an extensive and detailed knowledge of the Far East, as is shown by the author of Niddesa, existed in India in the first century A. D. On the other hand, the absence of any reference to Cambodge or Champa makes it equally improbable that the list was drawn up in the third century A. D. when those countries were certainly known to India. Thus the list of Niddesa must have been drawn up between the end of the first and the beginning of the third century A. D. We shall now say a few words about the ten extraordinary routes mentioned at the end of the passage in Niddesa. The meaning of these has been made clear, partly by the commen- tary Saddhammappajotika, and partly by the occurrence of some of them in the story of the merchant Sanudasa as narrated in Brhatkatha-Slofea-samgraha. The story of Sanudasa is thus summarised by Lacote 1 . 'Sanudasa joins the gang of the adventurer Jcera, who is preparing an expedition to the land of Gold (Suvannabhflmi). They cross the sea and land at the foot of a mountain. They climb up to the top by catching hold of creepers ( Vetra). This is the "creepers' path" (Vetrapattta). On the plateau there I. Translation by Tabard, p. 131. SUVARNADV1PA 59 is a river which changes into stone everything that falls into it. They cross it by holding on to the bamboos which overhang the banks 1 . This is "the bamboos' path" (Vamspatha). Further on, they meet a narrow path between two precipices. They light a fire with wet branches ; the smoke attracts some Earatas who come and propose to sell them some goats ; the adventurers get on those goats, the only animals sure-footed enough to be able to follow the narrow edge without feeling giddy. This is "the goats' path" (Ajapatha). The adventurers do not come to the end of it without some difficulty as another gang is approaching from the opposite direction. A struggle ensues, but Accra's troops are able to pass through after having thrown their enemies into the ravines. Sanudasa begins to feel indignant at the fierceness of the gold-seekers. Acera orders his followers to slay the goats and to put on their skins with the inside out. Huge birds will mistake those men for a heap of raw meat, come and carry them away to their aerie. It is there the gold is ! Sanudasa attempts to save the goat he was riding, but his companions are pitiless. Everything takes place as Acera had foretold, but the bird which carries off Sanudasa is attacked by another bird which attempts to steal his prey. The goat's skin bursts open and Sanudasa falls in a tank which is in the heart of a luxuriant forest. The next day he comes to a river the banks of which are of golden sand ; near by, there is a hermitage from which a hermit comes out.' The story thus explains Ajapatha (26) and Vamsapatha (30), and the episode of Sanudasa being carried aloft by a huge bird evidently explains the Sakunapatha (31). Mendhapatha (27) obviously is to be explained in the same way as Ajapatha, substituting ram for a goat. The Vetrapatha is added in the story and may correspond to Vettadhara or VettacSra (No. 34). I. The bamboos on the other bank of the river are bent by strong winds, and a man catches hold of the top of one of them as soon as it is within the reach of the bank on which he is standing. Then, when the storm subsides, the bamboo reverts to its old position, and the man holding fast to it is carried along with it to the other bank, 60 SUVARNADVlPA The commentary explains Jannupatha (25) as the way where one has to crawl on knees. On Sankupatha (28) it gives a long explanatory note, describing the means by which a man could ascend a mountain. An iron hook, attached to a rope of skin, is thrown up till the hook is fixed up in the mountain. Having climbed up the rope, the man makes a hole on the hillside with a diamond-tipped iron instrument, and fixes a spear. Having caught hold of this, he detaches the hook, and throws it aloft again, till it is again fixed up in the mountain. Then he ties the rope to the spear, and having caught hold of the rope with one hand, strikes it by a hammer with the other till the spear is detached. Then he climbs up again, again fixes the spear, and repeats the process till he ascends the top of the hill. Chattapatha (29) is explained in the commentary as the way where one jumps down from a precipice with an open parasol, (chatta=chatra) made of skin, and descends slowly to the ground, on account of the resistance of the air. In other words, it involved the principle of parachute. The Masikapatha (32) and Daripatha (33) are not explained by the commentary and cannot be exactly understood. References to these extraordinary routes are not confined to the two texts mentioned above. They are met with in the VimSnavatthu, the Jatakas, Milindapanha, Vayu Purana, Matsya Purana, Kfttyayana's Vartika and Ganapatha 1 . None of these, however, mentions a large number of them, and the Puranas alone add a new one, Kharapatha, which is evidently to be explained in the same way as Ajapatha, substituting ass (khara) for goat (a/a). It is to be noted that Katy5yana associates these ways with merchants, and MilindapaSha agrees in a way, substituting cf. Etudes Asiatiques, Vol. II, pp. 45-50, for details. References ; are to Vimanavatthu LXXXIV ; Tittirajataka (Jataka III, 541), Milindapafiha (p. 280) ; Vayu Purana, Ch. 47, v. 54 ; Matsya Purana Ch. 121, v. 56 ; Patafijali's comment on Panini's Sutra V, 1,77 ; Ganapatha on Panini V. 3. ioo f SUVAKNADVlPA 61 seekers of wealth for merchants. The Vimanavatthu definitely associates them with oversea countries, agreeing in this respect with Niddcsa and Brhatkatha-Sloka-saihgraha. The Puranas also mention them in connection with countries outside India. We may now sum up the results of the preceding discussion. It is quite clear that from a very remote time the Indians possessed a vague idea of the countries in the Far East across the sea. The relation, no doubt, originated in trade, and the tradition of fabulous wealth earned by that trade gave rise to all sorts of mythical stories about the golden land. The Puranik accounts of the varsas and dvlpas, which represent this stage, were based on vague sailors' reports, but were also mingled with a great deal of fancy and imagination. The steady development of this trade is reflected in th Jatakas, Brhatkatha, Kautillya ArthaSastra and Milinda- paSha, where we have not only a more definite idea of the region, now called Suvarnabhunii, but also a knowledge of important localities within it. This intimate intercourse may be referred to the two or three centuries immediately preceding the Christian era. During the first two centuries of the Christian era, the mercantile relations led to colonisations on a fairly large scale. This is evidenced both by the popular stories as well as the by Sanskrit names applied to many localities within this region. Ptolemy and Niddesa represent this stage of development which may thus be regarded as an accomplished fact by the second century A. D. The literary evidence leaves no doubt that trade was the chief stimulus of this intercourse between India and the Far East. Missionary and political activities must have followed in the wake of trade. Indeed, if literature can be regarded as a fair reflex of popular mind, trade and commerce must have been a supreme passion in India in the centuries immediately preceding and following the Christian era, perhaps very much in the same way as it is in Europe to-day. The extraordinary routes mentioned above, together with the details of ship-wreck 62 SUVABNADVlPA and perils of the sea preserved in numerous stories, are but a faint echo of that romantic age of adventures and explorations. If the history of that wonderful epoch of new discoveries had been preserved to us, we might possibly present it as a not unworthy parallel of the similar period in modern age. We lay particular stress on this fact, as it is the background of our study of ancient Indian colonisation in the Far East. Indeed, the evidence of a commercial origin of tliis inter- course with the Far East meets us at every step. In the first place, almost all the literary references given above deal with stories of merchants or seekers of wealth. Secondly, the geo- graphical names, applied by the Indians, all refer to minerals, metals, or some industrial and agricultural products. We may note, for example, Suvarnadvipa ( and its variants Hemakdta Suvarnakdta, Suvarnakudya), Rupyakadvlpa, Tamradvlpa, Yava- dvlpa, Lankftdvlpa, Takkola, Sankha-dvipa, KarpQra-dvIpa, Narikela-dvlpa, etc. Thirdly, Kautillya ArthalSstra knows of foreign countries only in connection with their industrial products. Fourthly, we may refer to a statement of K'ang T'ai, the Chinese ambassador to Fou-Nan about the middle of the third century A. D., which runs as follows : "Formerly, during the reign of Fan-Chan, a man called Kia- Siang-li came from India to Fou-Nan for purposes of trade. He gave a short account of India to Fan-Chan who then asked him : "What is the distance of India ? How long does it take to go to that country V 9 Kia-Siang-li replied : "India is about 30,000 li from here. A journey to India and back would require, three or four years'' l This passage and another statement of K'ang T'ai" shows that the earliest intercourse between India and the Far East was through adventurous merchants, and it was well established as early as the third century A. D. i. B. E. F. E. O f Vol. Ill, pp. 277-8. 9. Etudes Asiatiques, Vol. II, pp. 249-50. SUVAENADVlPA 63 Some traditions, no doubt, represent Ksatriya adventurers from India as having conquered territories in the Far East> but they must have followed in the wake of merchants. l It is, of course, true that trade and commercial relations led to the establishment of political and cultural relations as well. But these were secondary results and not primary motives of intercourse. There is no reference in our literature to any deliberate policy of political expansion or religious propaganda across the sea, until we come to the Ce.ylonese Chronicle Mahiivaiiisa. As is well-known, it refers to tho conquest of Ceylon by Vijaya at the time of Buddha, and the despatch of a Buddhist mission to Suvamiabhnim in the time of ASoka. Whether the dates of either of those events can be accepted as true is a matter of dispute. But in any case, if true, they would constitute the only exceptions, and even then we should remember that the path had already been paved by the merchants. On the whole it can be definitely laid down, that trade and commercial activity were the first, and, for a long time, the only incentive to the perilous voyages across the sea. Tho traders spread Indian culture along with their wares, and as opportunities offered, they might have seized the political power. But it is only at a comparatively later age, that adventurous Ksatriya princes came to seek their fortune, or individual monk or bands of missionaries came to propagate their religious doctrines. We possess evidence of both, but they all belong to a later period. The subsequent history of individual colonies will show, that this peaceful penetration of the Indians resulted in the fusion of Indians with their diverse races, and the evolution of a new culture which partook of elements of both. The dominant race imposed its language, religion and social customs, but could not efface all traces of indigenous element** in respect of any of these. As years went on, and the contact with India I, These have been referred to in my work 'ChampS,' pp. XI fi" 64 SUVARNADVIPA grew less and less, the native elements again asserted themselves. All these will be illustrated by the detailed history of the Indian colonies in the Malay Peninsula, and the islands of Java, Sumatra, Borneo, and Bali to which we now proceed. Chapter V. EARLY HINDU COLONISATION IN MALAY PENINSULA. The Malay Peninsula or the Peninsula of Malacca is tho name given to that long narrow strip of territory which, projecting southwards from Judo-China, divides the Bay of Bengal from the China Sen, and forms the most southerly extremity of the mainland of Asia. It is called by the natives Tanah Malayu, the land of the Malays. It is now generally regarded as beginning at the Isthmus of Kra, in Lat. 10, but, in the widest sense, the peninsula extends from the parallel of the head of the Gulf of Siam, in Lat. 13-30'. The peninsula runs at first south, and then in a south-eastern direction, for about 800 miles. The distance from the Isthmus of Km to Cape Rumenia ( east of Singapore ), as the crow Hies, would be about 750 miles. Cape Rumenia is nearly, though not exactly, the most southerly point in the peninsula, Tanjong Bulus ( l-lt>i'N. ), a little to the west, occupying that position. The peninsula is bounded on the north by Siam, and is surrounded by the sea in all other directions ; by the China Sea and the Gulf of Siam on the cast, by the Strait of Singapore on the south, and by the Straits of Malacca arid the Bay of Bengal on the west. There are many islands along the shores of the peninsula, the most notable being Langkawi and Penang on the west, and Singapore, Batan and Biritang on the south. The islands on the eastern coast are fewer and smaller. The most characteristic physical feature of the peninsula is the long range of granite mountains which runs along its 9 66 MALAY PENINSULA whole length, descending somewhat abruptly into a wider plain on the east, and more gently into a narrower plain on the west. In addition to smaller ranges running parallel to the main chain, there are also isolated spurs and limestone buffs. The highest peak in the main range, Gunong Kerbau, has an altitude of 7,160 ft., but the highest mountain is Gunong Tahan (7, 186 ft.) on the eastern side. Almost the whole of the peninsula both alluvial plains and mountain ranges is covered by evergreen forests, mostly dense jungles, the major part of which is yet untrodden by human foot. The forests yield excellent timber, including eaglewood, camphor tree, and ebony, and also less durable, but more frequently used, materials of Malayan architecture, such as rattans, bamboos, the nibung, and the nipa palms. Guttapcrcha, rubber, oils, and resins are also obtained from the forests. The chief products of agriculture are rice, sugar- cane, coffee, cotton, sago, pepper, spices, and rubber. There are also some excellent fruit trees such as the mango-steen, durian, pomegranate, jack-fruit, custard-apple, cocoa-nut, areca- nut, sugar-pahn, and banana. The rivers are numerous, but small, and in most cases navigable for large boats only upto a short distance from the mouth. The more notable arc the Pcrak, Bornam and Muar on the west, and Patani, Talukin, Kelantan, Bcsut, Trengganu, Kuantan, Pahang and Rompin on the east. On account of the impenetrable forests, the rivers have always formed the chief highways of communication, and it is on the banks of the rivers that the main centres of civilisation have grown. The chief mineral products are tin, iron, gold, and coal. The peninsula, with the islands adjacent to it, contains by far the most extensive tin fields in the world, and supplies nearly one-third of the world's output of that metal. Gold mines exist in Pahang, Kelantan, and Perak, and they are known to have been worked even in very ancient times. Among SUVARNADVlPA 67 other mineral products may be mentioned copper, mercury, lead, silver, zinc, and coal. 1 Although it is not within the scope of the present work to deal with the existing political condition of the Malay Peninsula, a brief review of its political geography is nece- ssary for the proper understanding of the subject. The northern part of the peninsula, forming a narrow isthmus running nearly due north and south to the length of 140 miles, is inha- bited by the Siamese or a, cross between them and the Malays, known to the latter by the name of Sansam. This portion, with territories further south, is politically subject to Siam and forms an integral part of that kingdom. The Siamese dominion is confined to the northern part of the peninsula, and comprises the following states, some of which, specially those in the north, forming practically so many Siamese provinces : on the west coast, beginning from north, are Eanong, Takua Pa, Takuatung, Pukct ( or Junk Ceylon, a corruption of the Malay name of Ujong Salang), Palian and Satul ; on the east coast, Patavi, Chumpaun, Caiya, the island of Samui, Nakonsitamaraj (Nakhon Sri Tha(dha)mmarat), Patalung, Sengora, Ghana Tepa, Nongchik, Tani Jaring, Jala, Sai Ranga, Raman, and Patani. To the south of these lie the states of Perils and Keddah on the west and Kelantan and Trcngganu in the cast, over which the kingdom of Siam exercised suzerainty until recent times, but which now form the Non-Federated Malay States, protected and advised I. In spito of numerous works on the Malay Peninsula, it is not easy to get a simple and accurate description of the physical features of the land. The statements in different authorities also do not always agree, particularly as regards distance, area, and the height of mountains. The above account is based on the following books :- a. John Crawfurd A Descriptive Dictionary of the Indian Islands and Adjacent Countiies (London 1856) ; (s. v. Malay Peninsula). b. J. H. Moor Notices of the Indian Archipelago and Adjacent Countries (Singapore, 1837), pp. 241. ff. c. Major C. M. Enriquez Malaya (Hurst and Blackett, 1927). d. Encyclopaedia Britannica I4th Edition, 68 MALAY PENINSULA by the British. The southern part of the peninsula consists of states which are more directly under the British authority. The regular British territories, forming the Crown Colony of Straits Settlements, are in point of size "mere dots on the map of the Malay Peninsula. One dot is Singapore ; a little way up the coast, Malacca is another ; still following the coast, the Bindings form a third ; Penang and Province Wcllesley are two more." The other states known as Federated Malay States are not, strictly speaking, British possessions, but they are ruled all but in name by the British Resident. These are Perak, Selangor, and the group of nine states, collectively known as Negri Sern- bilan, on the west, and Pahang on the east coast. To the south of these is the important State of Johore forming the southernmost portion of the Malay Peninsula. Since 1914 it has been included among the Non-Federated States, being protected and advised by the British. 1 The Malay Peninsula ( taking it in its narrower significance, to the south of the Isthmus of Era ) has a population of about three and a half millions. This includes 1,600,000 Malays, 1,200,000 Chinese, 470,000 Indians, and about 33,000 aboriginal or primitive tribes. The racial elements among the original people of Malay Peninsula have already been discussed above. The Chinese and Indian colonists have settled there since the early centuries of the Christian era. During the last four centuries the Europeans and Americans have formed a small colony, numbering at present about 15,000, with 12,000 Eurasians. It has already been shown above that the Malay Peninsula held a very important position in respect of maritime trade in the Far East from a very early period. Indeed, its geographical j. The account of the political divisions is based mainly on "The Peoples and Politics of the Far East" by Sir Henry Norman (T. Fisher Unwin, 1907). In some respects it is corrected by 'Malaya' by Enriquez and Encyclopaedia Britannica, i^\\ Edition, SUVARNADVlPA 69 position made it the centre of carrying trade between China and the western world. It must have been known to India from very early times. As has already been mentioned above, the names of both Malaya-dvlpa and Kataha-dvlpa occur in the PurSnas, and some of the Puranas include Kataha-dvlpa among the nine dvlpas into which the known world is divided. The earliest definite reference to this region is made by Ptolemy. He calls it 'Chryse Chersonesus', an equivalent of the Indian name Suvarnadvlpa, and expressly refers to an active maritime trade between India and this region. Ptolemy has shown a fair degree of knowledge as regards the geography of Malay Peninsula. He names successively (1) Takkola, a mart; (2) a cape situated after Takkola ; (3) mouth of the river Chrysoana ; (4) Sabana, a mart ; (5) mouth of the river Palandas ; (6) cape Maleu Kolon ; (7) mouth of the river Attaba ; (8) Koli, a town ; (9) Perimula ; and (10) Bay of Perimula. In a supplementary list he refers to the inland towns, Balongka, Kokkonagara, Tharrha, and Palanda. 1 It is not possible to identify exactly any of these", but that does not take away the great importance of Ptolemy's writings. S. Levi has shown that Ptolemy's account regarding the Far East possesses a striking agreement with that given in Niddesa, a Pali canonical book. 8 This proves, in his opinion, not only the general accuracy of the Greek account, but also that the Indians had acquired a far greater amount of knowledge of the Far East since the days when neither Pliny nor the author of the Pcriplus could gain anything but a vague report of Suvarnabhiimi from his Indian informants. In other words, the century 50-150 A. D. witnessed a remarkable 1. M'Crindle Ptolemy, pp. 197-8, 226. 2. Gerini's long discourses on the identification of Ptolemy's geographical names seem to be too unscientific to be relied upon. (cf. Researches, pp. 81-115). 3. S. LeVi Etudes Asiatiques, Vol. II, pp. i if., specially cf. p. jo. 70 MALAY PENINSULA growth in the trade and maritime activity of the Indians in the Far East. This is further corroborated by the fact, that not only the general name Suvarnabhflmi, but also local place- names such as Takkola, Java, and T&mralinga, and the name- ending 'nagara' in Kokkonagara, are purely Indian. 1 It may also be noted that by the second century A. D. there was a regular intercourse between India and China, either through the Isthmus of Kra, or the Straits of Malacca. 8 This period of active intercourse must also be regarded as the terminus ante quern for the Indian colonisation in Malay Peninsula. For, Fou-nan ( ancient Kamboja ) was colonised by the Hindus in the first century A. D. s , and Champa, not later than the second century A. D. 4 It, therefore, stands to reason that the Malay Peninsula, which lies on the route to these distant countries, must have been colonised at an earlier date. This a priori reasoning is also supported by traditional accounts. The History of the Liang Dynasty describes a country called Lang-ya-su ( or Lang-ga-su ) Svhich, the people say, was established more than 400 years ago/ Now the king of this country extols the emperor of China by saying, among other things, that the precious Sanskrit was generally known in his land. This leaves 110 doubt that it was a Hindu colony. As the Chinese history, containing the account, refers to the sixth century A. D., the traditional date of the foundation of the colony would be more than four hundred years before that, or, in other words, the second century A. D. It is generally agreed that Lang-ya-su was situated in Malay Peninsula, 1. Levi (op. cit, pp. 5. ff.) was the first to point out that Takkola was a regular Sanskrit word. 2. B. E. F. E. O., Vol. Ill, p. 291. A passage in Tsien-han-Shu refers to trade between China and Huang-tche during 140-86 B.C. Huang, tche has been identified with Abyssynia, Malay Peninsula and Kaficf in South India. (T'oung Pao, 1912, p. 457 ; J. A. n-XIII (1919), p. 451 ; J. A. n-XIV, p. 4 5 ; Tijd. Aard. Gen, Vol. 45, p. 589.) 3. Ibid, p. 290. 4. Champa R. C. Majumdar, p. 21, 91 though the exact localization of this colony is somewhat difficult. The same place is referred to as Lang-kia-su by I-tsing and Kama-laAka by Hiuen Tsang, and both enumerate it in a list of countries between Sri-Ksetra (Prome ) and DvSrfivatl (Siam). On this and other grounds, Pelliot held that it must be placed either near the Isthmus of Kra, or in Tenasserim, though he preferred the latter view. 1 Pelliot further held that this Lang-ya-su is the same as Ling-ya-sseu-kia mentioned by Chau Ju-kua. M. Coedfcs has proved that this latter is the same as Lenkasuka, mentioned in the 'Keddah Annals' and Nilgara-krtagama, and is to be identified with Gimong Jerai near Keddah. Coed&s further showed that the same place is referred to, in the form Ilangafio- gam, in the Tamil inscriptions of Rajcndra Cola, as one of the vassal states of Srl-Vijaya conquered by him. Coedfcs points out that Pclliot's identification of Lang-ya-su or Lang-kia-su with Ling-ya-sseu-kia or Lenkasuka cannot be upheld, as the latter is certainly near Keddah, whereas the former is perhaps near Tenasserim, as Pelliot suggests. 9 Coedfcs, however, ignores the fact that Pelliot's identification of Lang-ya-su with Tenasserim was a very hypothetical one, based upon its resemblance with Nankasi, the old name of Tenasserim. His main point was that it should be located in Tenasserim or Malay Peninsula, because it is inserted between Sri-Ksetra (Prome) and Dvaravati (Siam). As a matter of fact he himself suggested the Isthmus of Kra as a probable location, as, according to I-tsing, the Chinese pilgrims frequently passed through Lang-kia-su on their way from China to India or back. Even, therefore, if Lenkasuka is located near Keddah, there does not seem to be uny insuperable objection in placing Lang-ya-su or Lang-kia-su also in that locality. It must be remembered that the kingdon, according to the History of the 1. B E. F. E. O., Vol. IV, pp. 406-8. The identification with Tenasserim was also proposed by Huber (Ibid, p. 475). 2. B. E. F. E. O., Vol. XVIII. No. 6, pp. 11-13. ?2 MALAY PEtf INSULA Liang Dynasty, 'was 30 days' pacing from east to west and 20 days' pacing from north to south/ It may, therefore, be regarded as having comprised the northern part of the Malay Peninsula extending as far south as Keddah. Rouffaer, however, places both Lang-kia-su and Leiikasuka in Johore in the southern part of the Malay Peninsula. 1 . On the other hand, Ferrand has traced the name in an Arabic work, in the form Lang-Saka, and has identified it with Marco Polo's Lochac. On the strength of these and fresh Chinese evidences, he has located Lang-kia-su on the eastern coast of the Malay Peninsula, in the Isthmus of Ligor. 2 Indeed the passage which Ferrand has quoted from Chavannes' 'Eeli- gieux Eminents' (pp. 78 and 100), seems to leave no doubt on the point. If, therefore, Coedfcs' identification of Leiikasuka with Gunong Jerai be accepted as definitely proved, we must hold that it was different from Lang-kia-su. Coedfcs' view, however, rests almost solely on the Hikayat Maron Mahawa&sa, a late work of no authentic character. It is also quite possible that the name of an old site was given to a newly founded city. Coedfcs has further relied upon the popular traditions about Leiikasuka or Langkasuka, noted by Blagden, and referred to hereafter. They may, however, be equally explained by the supposition that an old site of that name originally existed in the Isthmus of Ligor. On the other hand, M. Sylvain Levi's identification of MevilimbaAgam, mentioned in Rajendra Cola's inscription, with KSma-lanka 8 of Hiuen Tsang, differentiates the latter from Le&kasuka, mentioned separately as Ilangasogam in the same inscription. This would support Coedfcs' view. Thus, while it is difficult to identify definitely Lang-kia-su with Ling-ya-sseu-kia, the former may be placed in the Isthmus of Ligor. In any case we are fully justified in regarding Lang-kia-su as an old Indian colony in Malay Peninsula, dating probably 1. B. K. I., 1931, pp. 89 ff. 2. J, A. H-XII (1918), pp. I34fl. 3, J. A., Vol. CCIII, p. 44. SUVAKNADVlPA 73 from the second century A. D. Some interesting accounts of this colony are preserved in Chinese annals. 1 The manners and customs of its people, as described by the Chinese, show a strong Indian element, modified, as in other colonies, by the indigenous influence. The Chinese annals give us some information about the political condition of the country during the fifth and sixth centuries A. D. The passage is thus translated by Schlegel : "The people of this country say that their state was founded more than 400 years ago ( A. D. 100 ), but that it got weaker under its successors (sic) ; and as there was among the rela- tions of the king one who was an excellent man, the people turned towards him. When the king heard of this, he put him into prison, but his chains snapped spontaneously. On this the king thought him to be a supernatural being and dared not hurt him any more, but only drove him from his territory, whence he took refuge to India, and was married there to tho eldest daughter ( of its king ). When on a sudden the king of Lang-ga su died, the great officers called back the prince and made him king. He died more than 20 years later, and was succeeded by his son Bhagadato. In A. D. 515 he sent an envoy named Aditya with a letter to the emperor of China. "These embassies were repeated in A. D. 523 and in 531 and then seem to have been dropped."* Pelliot points out that there was a further embassy to China in A. D. 5G8. 3 In course of a highly interesting and instructive philological disquisition, M. Sylvain Levi* has demonstrated that KSma- lanka, the name given to the colony by Hiuen Tsang, also occurs i. The Chinese accounts have been translated by Groeneveldt (Notes, pp. ion), and Schlegel (ToungPao, Serie I, Vol. IX., pp. 191- 200). 2. Schlegel (op. cit, pp. 192-3). 3, B. E. F. E. O. f Vol. IV, p. 405. 4. J. A., Vol. CCIII, pp. 38$ ; translated by Bagchi in "Pre- Aryan and Pre-Dravidian in India*, pp. 104 ff. 10 74 MALAY PENINSULA in Indian literature as Karmaranga. The MaSjuSrlmfllakalpa (p. 332) "names the islands of Karmaranga with the island of Cocoanuts (Nadlkera) and Vsrusaka (Baros, Sumatra) and the islands of the Naked (Nicobar), Bali and Java as the regions where the language is indistinct, without clearness, rude, and abunding in the letter r" The same text again (p.648) mentions Karmaranga with Harikela, Kamarupa, and Kalo&a (see below). Bana, in his Haracarita, twice mentions the shield of Karma- ranga, and his commentator Sankara remarks on the excellent skins of the country. On this M. Levi remarks as follows : "The reputation of the skins of Karmaranga appears to explain Ptolemy's note on the population of the "Brigands" "Lestai" which he locates exactly in the surroundings of Karmaranga, on the southern shores of the great gulf, i.e., the Gulf of Siam (VII, 2, 6 and 21) : "It is said that the natives of the country of Brigands live like beasts, inhabit the caverns, and that they have skin almost like that of hippopotami impenetrable by arrows." The region had some centres of population and even a port of commerce. "Samara(n)de, Pagrasa, Pithonobaste which is a market, Akadra, Zabai which is the city." It can be supposed that Samara (n)de is an alteration of the name which has finally taken in Sanskrit the alternate forms Carmaranga and Karmarafiga". M. Levi further points out that India received from this country the fruit which the Europeans call carambola and which is named in Sanskrit, after the land of its origin, Karmaranga (Bengali-KSmranga). Now the Malaya name of this fruit is balimbing or belimbing, which has made its way in all parts of South India along with the Sanskrit name. This has supplied to M. Levi the key to the solution of a geographical problem. Among the countries conquered by Rajcndra Cola occurs the name Mevilimbangam which has not been hitherto identified. Referring to the Malay name of the fruit, M. Levi remarks as follows on the identity of Mevilimbangam : "Mevilimbangam should, therefore, be analysed, in the inscription of Tanjore, like Ms-Danialingam, Ma-Nakkavaram, as Me-Vilimbangam ; it SUVAKNADVlPA 75 is clear that Vilimbangam is the Indian transcription of Malaya belimbing which is the equivalent of Karmaraftga. The Indian name of the fruit, derived from the name of the country, has become in its turn the indication of the country itself ." Thus Sylvain Lvi thinks that Mevilimbaiigam is but another name of Kama-lanka=s Lang-kia-su. As pointed out above, this view of Levi would mean that Lang-kia-su was different from Langkasuka or Leiikasuka. But even if it were so, "the two countries", as Levi remarks, "are certainly very near each other". As Blagden has pointed out 1 , "Langkasuka still lives in the memory of the local Malays. It has developed into a myth, being evidently the "spirit land" referred to as Lokon Suka by the peasantry of the Patani states". L6vi has also pointed out 8 that besides Karmaranga, the Mafijusrlmulakalpa twice mentions also the name of Carma- ranga (p. 206, 233), and he considers it only a variant of the same name. Now the Brhat-Sairihita, in its catalogue of the peoples of the south-east, combines Vrsa-Nalikera-Carmadvlpa. These three names may be compared to Varusaka-Nadikera and Karmaranga (or Carmaranga) of the ManjuSrlmnlakalpa referred to above. Vrsa is possibly the same as Varusaka (Baros, Sumatra), and Carmadvlpa may be presumed to be the same as Cannaranga= Karmarafiga == Kama-laAka = modern Ligor. Carmaranga is mentioned in MafijuSrimalakalpa with Kalalavarapura (Kalafiahva p. 206 ; KalaSamukhya, p. 233). KalaSapura is referred to as a city in Suvarnadvlpa in the Kathasaritrsagara (54, 108). In the collection of Nepalese miniatures studied by M. Foucher, the representation of Bhagavat at KalaSavarapura immediately follows that of Dipankara in Yavadvipa. The New History of the T'ang Dynasty refers to a kingdom called Ko-lo-cho-f en. Apparently this kingdom is again referred i, J. R. A. S., 1906, p. 119. a. Op. cit., p, 106. 76 MALAY PENINSULA to in the same text as Kia-lo-cho-fou and Kia-lo-cho-fo. All the three forms correspond to KalaSapura. As to the location of the kingdom, the Chinese accounts place it to the north of Tou-ho-lo, which was to the north of P'an-p'an. Now Tou-ho-lo has been identified with DvarSvati, in the lower valley of the Menam river. If Kalafiapura is to be placed to the north of DvSr&vatl, it must have been an inland region far away from the sea, whereas, according to the story in the Kathasarit-sSgara, the ship-wrecked merchant SamudraSura was cast adrift at Kalafiapura, which was evidently on the sea-coast. Pelliot has shown on good grounds that the directions given in the parti- cular Chinese passage cannot be held to be quite accurate, and he, therefore, proposes to substitute Vest' for 'north', for which there is some independent authority. With this modification of the text, KalaSapura may be placed to the north-west of Siam, at the mouth of the Sittang river. On the other hand, P'an-p'an corresponds to Bandon or Ligor in Malay Peninsula, and, therefore, KalaSapura also may be placed in the northern part of it. It may bo noted that Kern amended the name KalaSapura to Kalapapura, Kalapa being the name for Batavia. This amendment, however, is unten- able in view of the forms of the name in the Chinese Text. 1 To the south-east of P'an-p'an, the Chinese locate a country called Kala or Kora. It is evidently the same as Keddah, which was the centre of trade and commerce between the cast and the west and figures so prominently in later Arab accounts. Its ambassadors visited China between 650 and 656 A.D., and the following account 8 preserved in the New History of the Tang Dynasty is apparently based on their report. 'This country is situated at the south-cast of P'an-p'an and is also called Kora Fu-sa-ra. The king's family name is Sri Pora and his personal name is Mi-si Po-ra. The walls of his city are built with stones piled upon each other, whilst the 1. B. E. F. E. O., Vol. IV, p. 360. 2. Groeneveldt Notes, p. 121. SUVARNADVIPA 77 watch-towers, the palace and other buildings are thatched with straw. The country is divided into 24 districts. "The soldiers use bows, arrows, swords, lances, and armour of leather ; their banners are adorned with peacock feathers and they fight mounted on elephants ; one division of the army consists of a hundred of these, and each elephant is surrounded by a hundred men. On the elephant's back is a cage containing four men, armed with bows, arrows and lances. "As taxes the people pay a little silver. There are no silkworms, nor hemp or flax, nothing else but cotton. For domestic animals they have numerous cows and a few ponies. "It is their custom that only functionaries are allowed to tie up their hair and to wrap a handkerchief round their heads." Another Hindu state in Malay Peninsula, of which we get some notice in the Chinese annals, is Pa-hoang (or Po-houang) which has been identified by Schlegel with Pahang. The following account is contained in the Nan-shi and the History of the First Sung Dynasty. 1 . "In A. D. 449 the king of the state of Pahang, named Sari- Pala-Varma sent envoys who presented 41 different articles of tribute. By imperial decree Emperor Wen named him "Bang of the state of Pahang". In A. D. 451 and 456 he again sent his great historian Da Napati to present a letter and offer products of his country, when H. M. gave to Napati the title of "Awe- inspiring general. "In A.D. 459 its king offered red and white parrots. In A.D. 464 and 466 he sent again envoys to offer tribute, when Ming-ti gave to his great historian Da Surawan as also to the former grand historian, the Awe-inspiring general Da Napati, the title of Dragon-horse Generals". The kingdom of Pahang with its two state historians must be regarded as a state with a high degree of civilisation. The i. T'oung Pao, Serie I, vol. X (1899), pp. 398. Pelliot, however, is doubtful about this identification of Po-houang with Pahang cf. B. E. F. E. O., Vol. IV, p. 272. 78 MALAY PENINSULA name of its king, ending in Varman, leaves no doubt that he was a Hindu. It is evident from the above account that this Hindu state in the eastern part of Malay Peninsula was in close and intimate contact with the Chinese court during the fifth century A.D. There is, perhaps, reference to another old Hindu state in Malay Peninsula, but the question is unfortunately not free from doubt and difficulties. The Chinese annals of the Liang and First Sung Dynasty refer to a kingdom called Kan-to-li or Kin-to-li situated on an island in the southern sea 1 . Neither T'ang nor later Sung annals refer to the kingdom, and it is not till we come to the History of the Ming Dynasty that we come across the name again. There it is definitely asserted that Kan-to-li was the old name of San-bo-tsai. Now, on the basis of the identification of San-bo-tsai with Sri-Vijaya and Palembang, Groeneveldt, Schlegel, and other scholars took Kan-to-li of the Liang and First Sung annals as equivalent to Palembang. This view has been strongly criticised by Gerini. Referring to the identification of Kan-to-li with San-bo-tsai by the late Ming historians, he remarks : "This late identification looks, I need not say, exceedingly suspicious, especially in view of the fact that we have more than once caught Chinese authors at fault in this sort of game ; and last, but not least, because there was and still exists a Khanthuli or Kanturi district on the east coast of the Malay Peninsula, which may very well be the old Kan-to-li of First Sung and Liang periods." The criticism of Gerini appears to be a valid one, and neither Pelliot nor Ferraiid is willing to put much faith in the identification proposed by later Chinese historians. But the identification proposed by Gerini has not found general i. For the Chinese references to Kan-to-li and discussions about its identification cf. i, Groeneveldt Notes, pp. 60-62. 2. Ferrand J. A. n-XIV (1919), pp. 238-41. 3- Gerini Researches, pp. 601-604. 4. Pelliot B. E. F. E. O., Vol. IV, pp. 401-2. 5. Schlegel T'oung Pao, Serie II. Vol. II, pp. 122-4. SUVAKNADVIPA 79 acceptance. Ferrand quotes a passage from the Hawiya of Ibn Majid (dated A.D. 1462), which shows that Kandari was a general appellation of the island of Sumatra. Ferrand suggests that the Ming historians really conveyed an authentic informa- tion, though their wordings are a little inaccurrte ; for, instead of saying that San-bo-tsai was a part of Kan-to-li, they said that San-bo-tsai was Kan-to-li. Ferrand's view does not seem to be a very probable one, and I have discussed the question in detail in Book II., Chapter I., Appendix. I hold the view that it represents ancient Kadara, a state in the Malay Peninsula. The History of the Liang Dynasty gives us the following information regarding Kan-to-li 1 . "Its customs and manners are similar to those of Cambodge and Champa. It produces clothes of variegated colour, cotton, and excellent arcca-nuts. In the reign of the emperor Hia-Wu (454-465 A.D.) fi of the Sung Dynasty, the king of this country, Che-p'o-lo-iia-lien-to (Srlvaranarendra) 3 sent a high official named Tchou-Lieou-to (Rudra, the Indian) to present valuable articles of gold and silver. In the year 502, the king K'iu-t'an-sieou-pa-to-lo (Gautama Subhadra) sent envoys to the emperor. Sometime after, the king died and his son P'i-yc-pa-mo ( Vijaya Varman or Priyavarman ?) succeeded him. In 519 the latter sent a high official called Pi-yuan-pa-mo (Vi Varman) to the emperor with a letter. 1. The translation that follows is based upon Ferrand's summary (op. cit). Groeneveldt's translation is somewhat defective. 2. The date is given as such by Cordier (La Chine, Vol. I. 335-36). Groeneveldt gives the date as 454-464 (p 60); Krom gives 452-464 (p- 81) ; while Ferrand gives 454-454