A History of Sanskrit Literature.

by Arthur A. MacDonell.

PREFACE

It is undoubtedly a surprising fact that down to the present time no history of Sanskrit literature as a whole has been written in English. For not only does that literature possess much intrinsic merit, but the light it sheds on the life and thought of the population of our Indian Empire ought to have a peculiar interest for the British nation. Owing chiefly to the lack of an adequate account of the subject, few, even of the young men who leave these shores every year to be its future rulers, possess any connected information about the literature in which the civilisation of Modern India can be traced to its sources, and without which that civilisation cannot be fully understood. It was, therefore, with the greatest pleasure that I accepted Mr. Gosse's invitation to contribute a volume to this series of Literatures of the World; for this appeared to me to be a peculiarly good opportunity for diffusing information on a subject in which more than twenty years of continuous study and teaching had instilled into me an ever-deepening interest.

Professor Max Muller's valuable History of Ancient Sanskrit Literature is limited in its scope to the Vedic period. It has long been out of print; and Vedic research has necessarily made great strides in the forty years which have elapsed since its publication.

The only book accessible to the English reader on the history of Sanskrit literature in general has hitherto been the translation of Professor Weber's Academical Lectures on Indian Literature, as delivered nearly half a century ago at Berlin. The numerous and often very lengthy notes in this work supply the results of research during the next twenty-five years; but as these notes often modify, or even cancel, the statements of the unaltered original text of 1852, the result is bewildering to the student. Much new light has been thrown on various branches of Sanskrit literature since 1878, when the last notes were added to this translation, which, moreover, is not in any way adapted to the wants of the general reader. The only work on the subject appealing to the latter is the late Sir M. Monier-Williams's Indian Wisdom. That book, however, although it furnishes, in addition to the translated specimens, some account of the chief departments of Sanskrit literature, is not a history. There is thus distinctly a twofold demand in this country for a history of Sanskrit literature. The student is in want of a guide setting forth in a clear and trustworthy manner the results of research down to the present time, and the cultivated English reader looks for a book presenting in an intelligible and attractive form information which must have a special interest to us owing to our close relations with India.

To lack of space, no less than to the scope of the present series, is due the exclusion of a full account of the technical literature of law, science, and art, which contains much that would interest even the general reader; but the brief epitome given in the Appendix will, I hope, suffice to direct the student to all the most important authorities.

As to the bibliographical notes, I trust that, though necessarily restricted in extent, they will enable the student to find all further information he may want on matters of detail; for instance, the evidence for approximate dates, which had occasionally to be summarily stated even in the text.

In writing this history of Sanskrit literature, I have dwelt more on the life and thought of Ancient India, which that literature embodies, than would perhaps have appeared necessary in the case of a European literature. This I have done partly because Sanskrit literature, as representing an independent civilisation entirely different from that of the West, requires more explanation than most others; and partly because, owing to the remarkable continuity of Indian culture, the religious and social institutions of Modern India are constantly illustrated by those of the past.

Besides the above-mentioned works of Professors Max Muller and Weber, I have made considerable use of Professor L. von Schroeder's excellent Indiens Literatur und Cultur (1887). I have further consulted in one way or another nearly all the books and monographs mentioned in the bibliographical notes. Much of what I have written is also based on my own studies of Sanskrit literature.

All the quotations which I have given by way of illustration I have myself carefully selected from the original works. Excepting the short extracts on page 333 from Cowell and Thomas's excellent translation of the Harshacharita, all the renderings of these are my own. In my versions of Rigvedic stanzas I have, however, occasionally borrowed a line or phrase from Griffith. Nearly all my renderings are as close as the use of metre permits. I have endeavoured to reproduce, as far as possible, the measures of the original, except in the quotations from the dramas, where I have always employed blank verse. I have throughout refrained from rhyme, as misrepresenting the original Sanskrit.

In the transliteration of Sanskrit words I have been guided by the desire to avoid the use of letters which might mislead those who do not know Sanskrit. I have therefore departed in a few particulars from the system on which Sanskrit scholars are now almost unanimously agreed, and which I otherwise follow myself. Hence for c and ch I have written ch and chh respectively, though in the rare cases where these two appear in combination I have retained cch (instead of chchh). I further use sh for the lingual s, and c for the palatal s, and ri for the vowel r. I have not thought it necessary to distinguish the guttural n and the palatal n by diacritical marks, simply printing, for instance, anga and pancha. The reader who is unacquainted with Sanskrit will thus pronounce all words correctly by simply treating all the consonants as in English; remembering only that the vowels should be sounded as in Italian, and that e and o are always long.

I am indebted for some suggestions to my friend Mr. F. C. S. Schiller, Fellow and Tutor of Corpus Christi College, who looked through the final proof of the chapter on Philosophy. To my pupil Mr. A. B. Keith, Boden Sanskrit scholar and Classical scholar of Balliol, who has read all the final proofs with great care, I owe not only the removal of a number of errors of the press, but also several valuable criticisms regarding matters of fact.

107 Banbury Road, Oxford, December 1, 1899.

CHAPTER I

INTRODUCTORY

Since the Renaissance there has been no event of such world-wide significance in the history of culture as the discovery of Sanskrit literature in the latter part of the eighteenth century. After Alexander's invasion, the Greeks became to some extent acquainted with the learning of the Indians; the Arabs, in the Middle Ages, introduced the knowledge of Indian science to the West; a few European missionaries, from the sixteenth century onwards, were not only aware of the existence of, but also acquired some familiarity with, the ancient language of India; and Abraham Roger even translated the Sanskrit poet Bhartrihari into Dutch as early as 1651. Nevertheless, till about a hundred and twenty years ago there was no authentic information in Europe about the existence of Sanskrit literature, but only vague surmise, finding expression in stories about the wisdom of the Indians. The enthusiasm with which Voltaire in his Essai sur les Moeurs et l'Esprit des Nations greeted the lore of the Ezour Vedam, a work brought from India and introduced to his notice in the middle of the last century, was premature. For this work was later proved to be a forgery made in the seventeenth century by a Jesuit missionary. The scepticism justified by this fabrication, and indulged in when the discovery of the genuine Sanskrit literature was announced, survived far into the present century. Thus, Dugald Stewart, the philosopher, wrote an essay in which he endeavoured to prove that not only Sanskrit literature, but also the Sanskrit language, was a forgery made by the crafty Brahmans on the model of Greek after Alexander's conquest. Indeed, this view was elaborately defended by a professor at Dublin as late as the year 1838.

The first impulse to the study of Sanskrit was given by the practical administrative needs of our Indian possessions. Warren Hastings, at that time Governor-General, clearly seeing the advantage of ruling the Hindus as far as possible according to their own laws and customs, caused a number of Brahmans to prepare a digest based on the best ancient Indian legal authorities. An English version of this Sanskrit compilation, made through the medium of a Persian translation, was published in 1776. The introduction to this work, besides giving specimens of the Sanskrit script, for the first time supplied some trustworthy information about the ancient Indian language and literature. The earliest step, however, towards making Europe acquainted with actual Sanskrit writings was taken by Charles Wilkins, who, having, at the instigation of Warren Hastings, acquired a considerable knowledge of Sanskrit at Benares, published in 1785 a translation of the Bhagavad-gita, or The Song of the Adorable One, and two years later, a version of the well-known collection of fables entitled Hitopadeca, or Friendly Advice.

Sir William Jones (1746-94) was, however, the pioneer of Sanskrit studies in the West. It was this brilliant and many-sided Orientalist who, during his too brief career of eleven years in India, first aroused a keen interest in the study of Indian antiquity by his unwearied literary activity and by the foundation of the Asiatic Society of Bengal in 1784. Having rapidly acquired an accurate knowledge of Sanskrit, he published in 1789 a translation of cakuntala, the finest Sanskrit drama, which was greeted with enthusiasm by such judges as Herder and Goethe. This was followed by a translation of the Code of Manu, the most important of the Sanskrit law-books. To Sir William Jones also belongs the credit of having been the first man who ever printed an edition of a Sanskrit text. This was a short lyrical poem entitled Ritusamhara, or Cycle of the Seasons, published in 1792.

We next come to the great name of Henry Thomas Colebrooke (1765-1837), a man of extraordinary industry, combined with rare clearness of intellect and sobriety of judgment. The first to handle the Sanskrit language and literature on scientific principles, he published many texts, translations, and essays dealing with almost every branch of Sanskrit learning, thus laying the solid foundations on which later scholars have built.

While Colebrooke was beginning his literary career in India during the opening years of the century, the romance of war led to the practical knowledge of Sanskrit being introduced on the Continent of Europe. Alexander Hamilton (1765-1824), an Englishman who had acquired a good knowledge of Sanskrit in India, happened to be passing through France on his way home in 1802. Hostilities breaking out afresh just then, a decree of Napoleon, directed against all Englishmen in the country, kept Hamilton a prisoner in Paris. During his long involuntary stay in that city he taught Sanskrit to some French scholars, and especially to the German romantic poet Friedrich Schlegel. One of the results of these studies was the publication by Schlegel of his work On the Language and Wisdom of the Indians (1808). This book produced nothing less than a revolution in the science of language by the introduction of the comparative and the historical method. It led to the foundation of the science of comparative philology by Franz Bopp in his treatise on the conjugational system of Sanskrit in comparison with that of Greek, Latin, Persian, and German (1816). Schlegel's work, moreover, aroused so much zeal for the study of Sanskrit in Germany, that the vast progress made since his day in this branch of learning has been mainly due to the labours of his countrymen.

In the early days of Sanskrit studies Europeans became acquainted only with that later phase of the ancient language of India which is familiar to the Pandits, and is commonly called Classical Sanskrit. So it came about that the literature composed in this dialect engaged the attention of scholars almost exclusively down to the middle of the century. Colebrooke had, it is true, supplied as early as 1805 valuable information about the literature of the older period in his essay On the Vedas. Nearly a quarter of a century later, F. Rosen, a German scholar, had conceived the plan of making this more ancient literature known to Europe from the rich collection of manuscripts at the East India House; and his edition of the first eighth of the Rigveda was actually brought out in 1838, shortly after his premature death. But it was not till Rudolf Roth (1821-95), the founder of Vedic philology, published his epoch-making little book On the Literature and History of the Veda in 1846, that the studies of Sanskritists received a lasting impulse in the direction of the earlier and more important literature of the Vedas. These studies have since been prosecuted with such zeal, that nearly all the most valuable works of the Vedic, as well as the later period, have within the last fifty years been made accessible in thoroughly trustworthy editions.

In judging of the magnitude of the work thus accomplished, it should be borne in mind that the workers have been far fewer in this than in other analogous fields, while the literature of the Vedas at least equals in extent what survives of the writings of ancient Greece. Thus in the course of a century the whole range of Sanskrit literature, which in quantity exceeds that of Greece and Rome put together, has been explored. The great bulk of it has been edited, and most of its valuable productions have been translated, by competent hands. There has long been at the service of scholars a Sanskrit dictionary, larger and more scientific than any either of the classical languages yet possesses. The detailed investigations in every department of Sanskrit literature are now so numerous, that a comprehensive work embodying the results of all these researches has become a necessity. An encyclopaedia covering the whole domain of Indo-Aryan antiquity has accordingly been planned on a more extensive scale than that of any similar undertaking, and is now being published at Strasburg in parts, contributed to by about thirty specialists of various nationalities. By the tragic death, in April 1898, of its eminent editor, Professor Buhler of Vienna, Sanskrit scholarship has sustained an irreparable loss. The work begun by him is being completed by another very distinguished Indianist, Professor Kielhorn of Gottingen.

Although so much of Sanskrit literature has already been published, an examination of the catalogues of Sanskrit manuscripts, of which an enormous number are preserved in European and Indian libraries, proves that there are still many minor works awaiting, and likely to repay, the labours of an editor.

The study of Sanskrit literature deserves far more attention than it has yet received in this country. For in that ancient heritage the languages, the religious and intellectual life and thought, in short, the whole civilisation of the Hindus, who form the vast majority of the inhabitants of our Indian Empire, have their roots. Among all the ancient literatures, that of India is, moreover, undoubtedly in intrinsic value and aesthetic merit second only to that of Greece. To the latter it is, as a source for the study of human evolution, even superior. Its earliest period, being much older than any product of Greek literature, presents a more primitive form of belief, and therefore gives a clearer picture of the development of religious ideas than any other literary monument of the world. Hence it came about that, just as the discovery of the Sanskrit language led to the foundation of the science of Comparative Philology, an acquaintance with the literature of the Vedas resulted in the foundation of the science of Comparative Mythology by Adalbert Kuhn and Max Muller.

Though it has touched excellence in most of its branches, Sanskrit literature has mainly achieved greatness in religion and philosophy. The Indians are the only division of the Indo-European family which has created a great national religion--Brahmanism--and a great world-religion--Buddhism; while all the rest, far from displaying originality in this sphere, have long since adopted a foreign faith. The intellectual life of the Indians has, in fact, all along been more dominated by religious thought than that of any other race. The Indians, moreover, developed independently several systems of philosophy which bear evidence of high speculative powers. The great interest, however, which these two subjects must have for us lies, not so much in the results they attained, as in the fact that every step in the evolution of religion and philosophy can be traced in Sanskrit literature.

The importance of ancient Indian literature as a whole largely consists in its originality. Naturally isolated by its gigantic mountain barrier in the north, the Indian peninsula has ever since the Aryan invasion formed a world apart, over which a unique form of Aryan civilisation rapidly spread, and has ever since prevailed. When the Greeks, towards the end of the fourth century B.C., invaded the North-West, the Indians had already fully worked out a national culture of their own, unaffected by foreign influences. And, in spite of successive waves of invasion and conquest by Persians, Greeks, Scythians, Muhammadans, the national development of the life and literature of the Indo-Aryan race remained practically unchecked and unmodified from without down to the era of British occupation. No other branch of the Indo-European stock has experienced an isolated evolution like this. No other country except China can trace back its language and literature, its religious beliefs and rites, its domestic and social customs, through an uninterrupted development of more than three thousand years.

A few examples will serve to illustrate this remarkable continuity in Indian civilisation. Sanskrit is still spoken as the tongue of the learned by thousands of Brahmans, as it was centuries before our era. Nor has it ceased to be used for literary purposes, for many books and journals written in the ancient language are still produced. The copying of Sanskrit manuscripts is still continued in hundreds of libraries in India, uninterrupted even by the introduction of printing during the present century. The Vedas are still learnt by heart as they were long before the invasion of Alexander, and could even now be restored from the lips of religious teachers if every manuscript or printed copy of them were destroyed. A Vedic stanza of immemorial antiquity, addressed to the sun-god Savitri, is still recited in the daily worship of the Hindus. The god Vishnu, adored more than 3000 years ago, has countless votaries in India at the present day. Fire is still produced for sacrificial purposes by means of two sticks, as it was in ages even more remote. The wedding ceremony of the modern Hindu, to single out but one social custom, is essentially the same as it was long before the Christian era.

The history of ancient Indian literature naturally falls into two main periods. The first is the Vedic, which beginning perhaps as early as 1500 B.C., extends in its latest phase to about 200 B.C. In the former half of the Vedic age the character of its literature was creative and poetical, while the centre of culture lay in the territory of the Indus and its tributaries, the modern Panjab; in the latter half, literature was theologically speculative in matter and prosaic in form, while the centre of intellectual life had shifted to the valley of the Ganges. Thus in the course of the Vedic age Aryan civilisation had overspread the whole of Hindustan Proper, the vast tract extending from the mouths of the Indus to those of the Ganges, bounded on the north by the Himalaya, and on the south by the Vindhya range. The second period, concurrent with the final offshoots of Vedic literature and closing with the Muhammadan conquest after 1000 A.D., is the Sanskrit period strictly speaking. In a certain sense, owing to the continued literary use of Sanskrit, mainly for the composition of commentaries, this period may be regarded as coming down to the present day. During this second epoch Brahmanic culture was introduced into and overspread the southern portion of the continent called the Dekhan or "the South." In the course of these two periods taken together, Indian literature attained noteworthy results in nearly every department. The Vedic age, which, unlike the earlier epoch of Greece, produced only religious works, reached a high standard of merit in lyric poetry, and later made some advance towards the formation of a prose style.

The Sanskrit period, embracing in general secular subjects, achieved distinction in many branches of literature, in national as well as court epic, in lyric and especially didactic poetry, in the drama, in fairy tales, fables, and romances. Everywhere we find much true poetry, the beauty of which is, however, marred by obscurity of style and the ever-increasing taint of artificiality. But this period produced few works which, regarded as a whole, are dominated by a sense of harmony and proportion. Such considerations have had little influence on the aesthetic notions of India. The tendency has been rather towards exaggeration, manifesting itself in all directions. The almost incredible development of detail in ritual observance; the extraordinary excesses of asceticism; the grotesque representations of mythology in art; the frequent employment of vast numbers in description; the immense bulk of the epics; the unparalleled conciseness of one of the forms of prose; the huge compounds habitually employed in the later style, are among the more striking manifestations of this defect of the Indian mind.

In various branches of scientific literature, in phonetics, grammar, mathematics, astronomy, medicine, and law, the Indians also achieved notable results. In some of these subjects their attainments are, indeed, far in advance of what was accomplished by the Greeks.

History is the one weak spot in Indian literature. It is, in fact, non-existent. The total lack of the historical sense is so characteristic, that the whole course of Sanskrit literature is darkened by the shadow of this defect, suffering as it does from an entire absence of exact chronology. So true is this, that the very date of Kalidasa, the greatest of Indian poets, was long a matter of controversy within the limits of a thousand years, and is even now doubtful to the extent of a century or two. Thus the dates of Sanskrit authors are in the vast majority of cases only known approximately, having been inferred from the indirect evidence of interdependence, quotation or allusion, development of language or style. As to the events of their lives, we usually know nothing at all, and only in a few cases one or two general facts. Two causes seem to have combined to bring about this remarkable result. In the first place, early India wrote no history because it never made any. The ancient Indians never went through a struggle for life, like the Greeks in the Persian and the Romans in the Punic wars, such as would have welded their tribes into a nation and developed political greatness. Secondly, the Brahmans, whose task it would naturally have been to record great deeds, had early embraced the doctrine that all action and existence are a positive evil, and could therefore have felt but little inclination to chronicle historical events.

Such being the case, definite dates do not begin to appear in Indian literary history till about 500 A.D. The chronology of the Vedic period is altogether conjectural, being based entirely on internal evidence. Three main literary strata can be clearly distinguished in it by differences in language and style, as well as in religious and social views. For the development of each of these strata a reasonable length of time must be allowed; but all we can here hope to do is to approximate to the truth by centuries. The lower limit of the second Vedic stratum cannot, however, be fixed later than 500 B.C., because its latest doctrines are presupposed by Buddhism, and the date of the death of Buddha has been with a high degree of probability calculated, from the recorded dates of the various Buddhist councils, to be 480 B.C. With regard to the commencement of the Vedic age, there seems to have been a decided tendency among Sanskrit scholars to place it too high. 2000 B.C. is commonly represented as its starting-point. Supposing this to be correct, the truly vast period of 1500 years is required to account for a development of language and thought hardly greater than that between the Homeric and the Attic age of Greece. Professor Max Muller's earlier estimate of 1200 B.C., formed forty years ago, appears to be much nearer the mark. A lapse of three centuries, say from 1300-1000 B.C., would amply account for the difference between what is oldest and newest in Vedic hymn poetry. Considering that the affinity of the oldest form of the Avestan language with the dialect of the Vedas is already so great that, by the mere application of phonetic laws, whole Avestan stanzas may be translated word for word into Vedic, so as to produce verses correct not only in form but in poetic spirit; considering further, that if we knew the Avestan language at as early a stage as we know the Vedic, the former would necessarily be almost identical with the latter, it is impossible to avoid the conclusion that the Indian branch must have separated from the Iranian only a very short time before the beginnings of Vedic literature, and can therefore have hardly entered the North-West of India even as early as 1500 B.C. All previous estimates of the antiquity of the Vedic period have been outdone by the recent theory of Professor Jacobi of Bonn, who supposes that period goes back to at least 4000 B.C. This theory is based on astronomical calculations connected with a change in the beginning of the seasons, which Professor Jacobi thinks has taken place since the time of the Rigveda. The whole estimate is, however, invalidated by the assumption of a doubtful, and even improbable, meaning in a Vedic word, which forms the very starting-point of the theory. Meanwhile we must rest content with the certainty that Vedic literature in any case is of considerably higher antiquity than that of Greece.

For the post-Vedic period we have, in addition to the results of internal evidence, a few landmarks of general chronological importance in the visits of foreigners. The earliest date of this kind is that of the invasion of India by Alexander in 326 B.C. This was followed by the sojourn in India of various Greeks, of whom the most notable was Megasthenes. He resided for some years about 300 B.C. at the court of Pataliputra (the modern Patna), and has left a valuable though fragmentary account of the contemporary state of Indian society. Many centuries later India was visited by three Chinese Buddhist pilgrims, Fa Hian (399 A.D.), Hiouen Thsang (630-645), and I Tsing (671-695). The records of their travels, which have been preserved, and are all now translated into English, shed much light on the social conditions, the religious thought, and the Buddhist antiquities of India in their day. Some general and specific facts about Indian literature also can be gathered from them. Hiouen Thsang especially supplies some important statements about contemporary Sanskrit poets. It is not till his time that we can say of any Sanskrit writer that he was alive in any particular year, excepting only the three Indian astronomers, whose exact dates in the fifth and sixth centuries have been recorded by themselves. It was only the information supplied by the two earlier Chinese writers that made possible the greatest archaeological discovery of the present century in India, that of the site of Buddha's birthplace, Kapila-vastu, identified in December 1896. At the close of our period we have the very valuable account of the country at the time of the Muhammadan conquest by the Arabic author Alberuni, who wrote his India in 1030 A.D.

It is evident from what has been said, that before 500 A.D. literary chronology, even in the Sanskrit period, is almost entirely relative, priority or posteriority being determined by such criteria as development of style or thought, the mention of earlier authors by name, stray political references as to the Greeks or to some well-known dynasty, and allusions to astronomical facts which cannot have been known before a certain epoch. Recent research, owing to increased specialisation, has made considerable progress towards greater chronological definiteness. More light will doubtless in course of time come from the political history of early India, which is being reconstructed, with great industry and ability, by various distinguished scholars from the evidence of coins, copper-plate grants, and rock or pillar inscriptions. These have been or are being published in the Corpus Inscriptionum Indicarum, the Epigraphia Indica, and various journals devoted to the study of Indian antiquities. The rise in the study of epigraphy during the last twenty years has, indeed, already yielded some direct information of importance about the literary and religious history of India, by fixing the date of some of the later poets as well as by throwing light on religious systems and whole classes of literature. Thus some metrical inscriptions of considerable length have been deciphered, which prove the existence of court poetry in Sanskrit and vernacular dialects from the first century of our era onwards. No direct evidence of this fact had previously been known.

The older inscriptions are also important in connection with Sanskrit literature as illustrating both the early history of Indian writing and the state of the language at the time. The oldest of them are the rock and pillar inscriptions, dating from the middle of the third century B.C., of the great Buddhist king Acoka, who ruled over Northern India from 259 to 222 B.C., and during whose reign was held the third Buddhist council, at which the canon of the Buddhist scriptures was probably fixed. The importance of these inscriptions can hardly be over-rated for the value of the information to be derived from them about the political, religious, and linguistic conditions of the age. Found scattered all over India, from Girnar (Giri-nagara) in Kathiawar to Dhauli in Orissa, from Kapur-di-Giri, north of the Kabul river, to Khalsi, they have been reproduced, deciphered, and translated. One of them, engraved on a pillar erected by Acoka to commemorate the actual birthplace of Buddha, was discovered only at the close of 1896.

These Acoka inscriptions are the earliest records of Indian writing. The question of the origin and age of writing in India, long involved in doubt and controversy, has been greatly cleared up by the recent palaeographical researches of Professor Buhler. That great scholar has shown, that of the two kinds of script known in ancient India, the one called Kharoshthi employed in the country of Gandhara (Eastern Afghanistan and Northern Panjab) from the fourth century B.C. to 200 A.D., was borrowed from the Aramaic type of Semitic writing in use during the fifth century B.C. It was always written from right to left, like its original. The other ancient Indian script, called Brahmi, is, as Buhler shows, the true national writing of India, because all later Indian alphabets are descended from it, however dissimilar many of them may appear at the present day. It was regularly written from left to right; but that this was not its original direction is indicated by a coin of the fourth century B.C., the inscription on which runs from right to left. Dr. Buhler has shown that this writing is based on the oldest Northern Semitic or Phoenician type, represented on Assyrian weights and on the Moabite stone, which dates from about 890 B.C. He argues, with much probability, that it was introduced about 800 B.C. into India by traders coming by way of Mesopotamia.

References to writing in ancient Indian literature are, it is true, very rare and late; in no case, perhaps, earlier than the fourth century B.C., or not very long before the date of the Acoka inscriptions. Little weight, however, can be attached to the argumentum ex silentio in this instance. For though writing has now been extensively in use for an immense period, the native learning of the modern Indian is still based on oral tradition. The sacred scriptures as well as the sciences can only be acquired from the lips of a teacher, not from a manuscript; and as only memorial knowledge is accounted of value, writing and MSS. are rarely mentioned. Even modern poets do not wish to be read, but cherish the hope that their works may be recited. This immemorial practice, indeed, shows that the beginnings of Indian poetry and science go back to a time when writing was unknown, and a system of oral tradition, such as is referred to in the Rigveda, was developed before writing was introduced. The latter could, therefore, have been in use long before it began to be mentioned. The palaeographical evidence of the Acoka inscriptions, in any case, clearly shows that writing was no recent invention in the third century B.C., for most of the letters have several, often very divergent forms, sometimes as many as nine or ten. A considerable length of time was, moreover, needed to elaborate from the twenty-two borrowed Semitic symbols the full Brahmi alphabet of forty-six letters. This complete alphabet, which was evidently worked out by learned Brahmans on phonetic principles, must have existed by 500 B.C., according to the strong arguments adduced by Professor Buhler. This is the alphabet which is recognised in Pannini's great Sanskrit grammar of about the fourth century B.C., and has remained unmodified ever since. It not only represents all the sounds of the Sanskrit language, but is arranged on a thoroughly scientific method, the simple vowels (short and long) coming first, then the diphthongs, and lastly the consonants in uniform groups according to the organs of speech with which they are pronounced. Thus the dental consonants appear together as t, th, d, dh, n, and the labials as p, ph, b, bh, m. We Europeans, on the other hand, 2500 years later, and in a scientific age, still employ an alphabet which is not only inadequate to represent all the sounds of our languages, but even preserves the random order in which vowels and consonants are jumbled up as they were in the Greek adaptation of the primitive Semitic arrangement of 3000 years ago.

In the inscriptions of the third century B.C. two types, the Northern and the Southern, may be distinguished in the Brahmi writing. From the former is descended the group of Northern scripts which gradually prevailed in all the Aryan dialects of India. The most important of them is the Nagari (also called Devanagari), in which Sanskrit MSS. are usually written, and Sanskrit as well as Marathi and Hindi books are regularly printed. It is recognisable by the characteristic horizontal line at the top of the letters. The oldest inscription engraved entirely in Nagari belongs to the eighth, and the oldest MS. written in it to the eleventh century. From the Southern variety of the Brahmi writing are descended five types of script, all in use south of the Vindhya range. Among them are the characters employed in the Canarese and the Telugu country.

Owing to the perishability of the material on which they are written, Sanskrit MSS. older than the fourteenth century A.D. are rare. The two ancient materials used in India were strips of birch bark and palm leaves. The employment of the former, beginning in the North-West of India, where extensive birch forests clothe the slopes of the Himalaya, gradually spread to Central, Eastern, and Western India. The oldest known Sanskrit MS. written on birch bark dates from the fifth century A.D., and a Pali MS. in Kharoshthi which became known in 1897, is still older, but the use of this material doubtless goes back to far earlier days. Thus we have the statement of Quintus Curtius that the Indians employed it for writing on at the time of Alexander. The testimony of classical Sanskrit authors, as well as of Alberuni, shows that leaves of birch bark (bhurja-pattra) were also regularly used for letter-writing in early mediaeval India.

The first example of a palm leaf Sanskrit MS. belongs to the sixth century A.D. It is preserved in Japan, but there is a facsimile of it in the Bodleian Library. According to the Chinese pilgrim Hiouen Thsang, the use of the palm leaf was common all over India in the seventh century; but that it was known many centuries earlier is proved by the fact that an inscribed copper-plate, dating from the first century A.D. at the latest, imitates a palm leaf in shape.

Paper was introduced by the Muhammadan conquest, and has been very extensively used since that time for the writing of MSS. The oldest known example of a paper Sanskrit MS. written in India is one from Gujarat, belonging to the early part of the thirteenth century. In Northern India, where ink was employed for writing, palm leaves went out of use after the introduction of paper. But in the South, where a stilus has always been employed for scratching in the character, palm leaves are still common for writing both MSS. and letters. The birch bark and palm leaf MSS. are held together by a cord drawn through a single hole in the middle, or through two placed some distance apart. This explains how the Sanskrit word for "knot,"

grantha, came to acquire the sense of "book."

Leather or parchment has never been utilised in India for MSS., owing to the ritual impurity of animal materials. For inscriptions copper-plates were early and frequently employed. They regularly imitate the shape of either palm leaves or strips of birch bark.

The actual use of ink (the oldest Indian name of which is mashi) is proved for the second century B.C. by an inscription from a Buddhist relic mound, and is rendered very probable for the fourth century B.C. by the statements of Nearchos and Quintus Curtius.

All the old palm leaf, birch bark, and paper Sanskrit MSS. have been written with ink and a reed pen, usually called kalama (a term borrowed from the Greek kalamos). In Southern India, on the other hand, it has always been the practice to scratch the writing on palm leaves with a stilus, the characters being subsequently blackened by soot or charcoal being rubbed into them.

Sanskrit MSS. of every kind are usually kept between thin strips of wood with cords wound round them, and wrapped up in coloured, sometimes embroidered, cloths. They have been, and still are, preserved in the libraries of temples, monasteries, colleges, the courts of princes, as well as private houses. A famous library was owned by King Bhoja of Dhar in the eleventh century. That considerable private libraries existed in fairly early times is shown by the fact that the Sanskrit author Bana (about 620 A.D.) had in his employment a reader of manuscripts. Even at the present day there are many excellent libraries of Sanskrit MSS. in the possession of Brahmans all over India.