Sanskrit Literature

by A Berriedale Keith

Sanskrit Literature A Berriedale Keith १९२०

A HISTORY OF SANSKRIT LITERATURE Oxford University Press, Amen House, London E.C.4 GLASGOW NEW YORK TORONTO MELBOURNE WELLINGTON BOMBAY CALCUTTA MADRAS KARACHI CAPE TOWN IBADAN NAIROBI ACCRA SINGAPORE FIRST EDITION 1920 Reprinted photographically in Great Britain in 1941, 1948, 1953, 1956 by LOWE & BRYDONE, PRINTERS, LTD., LONDON from sheets of the first edition A HISTORY OF

SANSKRIT LITERATURE

BY

A. BERRIEDALE KEITH, D.C.L., D.Litt.

Of the Inner Temple, Barnster-at-Law, and Advocate

Regius Professor of Sanskrit and Comparative Philology

and Lecturer on the Constitution of the British Empire in

the University of Edinburgh

OXFORD UNIVERSITY PRESS Printed in Great Britain IN MEMORIAM FRATRIS ALAN DAVIDSON KEITH

(1885-1938) PREFACE

TAKEN in conjunction with my Sanskrit Drama, published in 1924, this work covers the field of Classical Sanskrit Literature, as opposed to the Vedic Literature, the epics, and the Puranas. To bring the subject-matter within the limits of a single volume has rendered it necessary to treat the scientific literature briefly, and to avoid discussions of its subject-matter which appertain rather to the historian of grammar, philosophy, law, medicine, astronomy, or mathematics, than to the literary his- torian. This mode of treatment has rendered it possible, for the first time in any treatise in English on Sanskrit Literature, to pay due attention to the literary qualities of the Kavya. Though it was to Englishmen, such as Sir William Jones and H. T. Cole- brooke, that our earliest knowledge of Sanskrit poetry was due, no English poet shared Goethe's marvellous appreciation of the merits of works known to him only through the distorting medium of translations, and attention in England has usually been limited to the Vedic literature, as a source for comparative philology, the history of religion, or Indo-European antiquities ; to the mysticism and monism of Sanskrit philosophy ; and to the fables and fairy-tales in their relations to western parallels.

The neglect of Sanskrit Kavya is doubtless natural. The great poets of India wrote for audiences of experts ; they were masters of the learning of their day, long trained in the use of language, and they aim to please by subtlety, not simplicity of effect. They had at their disposal a singularly beautiful speech, and they commanded elaborate and most effective metres. Under these circumstances it was inevitable that their works should be diffi- cult, but of those who on that score pass them by it may fairly be said ardua dum metuunt amiitunt vera viai. It is in the great writers of Kavya alone, headed by Kalidasa, that we find depth of feeling for life and nature matched with perfection of expres- sion and rhythm. The Kavya literature includes some of the great poetry of the world, but it can never e xpect to . altainjyide- popula rity^jn— th e We st^for^it^is esse ntially untran slatable.; viii PREFACE

German poets like Riickeit can, indeed, base excellent work on Sanskrit originals, but the effects produced are achieved by wholly different means, while English efforts at verse transla- tions fall invariably below a tolerable mediocrity, their diffuse tepidity contrasting painfully with the brilliant condensation of style, the elegance of metre, and the close adaptation of sound to sense of the originals. I have, therefore, as in my Sanskrit Drama, illustrated the merits of the poets by Sanskrit extracts, adding merely a literal English version, in which no note is taken of variations of text or renderings. To save space I have in the main dealt only with works earlier than A.D. 1200, though especially in the case of the scientific literature important books . of later date are briefly noticed.

This book was sent in, completed for the press, in January 1926, but pressure of work at the University Press precluded printing until the summer of 1927, when it was deemed best, in order not to delay progress, to assign to this preface the notice of such new discoveries and theories of 1926 and 1927 as might have peimanent interest.

On the early development of the Kavya welcome light has been thrown by Professor H. Luders's edition 1 of the fragments found in Central Asia of the K alpanamanditika of Kumaralata, which is the true description of the work hitherto known to us through a Chinese translation as the Sutralamkara of Acvaghosa. That work, it is suggested, was very different in character from Kumaralata's. It may have been an exposition in verse, possibly with prose additions, of the Canon of the Sarvastivadins, and it may be represented by fragments still extant; this suggestion can be supported by Asanga's choice of title, Mahayanasiiira- lavtkara, for his exposition of Mahayana tenets. But that is still merely a conjecture, and even less proved is the view that Subandhu's famous allusion % Bauddhasamgatim ivalamkdrabhu- sitam is to such a text as that ascribed to Acvaghosa. Kumara- lata may well have been a younger contemporary of Acvaghosa, who lived after the death of Kaniska, a fact which explains an old crux, the difficulty of ascribing to Acvaghosa the references

1 Bruchstucke der Kalfanamandttikd des Kumaralata, Leipzig, 1926.

2 Below, p. 308. LeVi (Sfitralamkara. li. 15 f.) reads saihgilim very plausibly, and holds thai a woiV of Asanga is meant. * PREFACE ix

in the Sutralamkara which seemed inconsistent with the tradi- tional relation of the patriarch and that king. How the Chinese version of the Kalpanamandilika, ' that which is adorned by poetic invention ', came to bear the style Sutralamkara, remains an unexplained problem.

The fragments shed a very interesting light on the develop- ment of the style of prose mingled with verses which appears in a more elaborate foim in the Jatakamala. The narratives, eighty in number, which, with ten parables, make up the work, begin with the enunciation of some doctrine, which is then established by means of an appropriate narrative ; unlike the Jatakamala, the text does not follow a stereotyped plan of drawing out at the close of each tale the moral which it inculcates. The stanzas used are normally portions of the speeches of the dramatis personae; there is a complete breach with the tradition of the canonical texts which introduce such verses by the term bkasam bhasate ; but of course this does not mean that Kumaralata, or Arya Cura who follows this plan in the Jatakamala, is the author of all the verses used ; doubtless he often adopts or adapts current maxims. Narrative * or descriptive stanzas are rare, and they are marked out for the benefit of the reciter by the words vaksyate hi. Arya Cura, on the other hand, shows a distinct advance ; he uses descriptive or narrative stanzas to the extent of over a fifth of his total number of verses, and omits any intro- duction, inserting them freely to beautify his prose narration. The parables take a different form : in them a prose parable (drstantd) is simply followed by a prose exposition (artha). The language shows the same adherence to correct Sanskrit, with occasional lapses, as in Acvaghosa, and there is a rich variety of metres, including the earliest Aryas in Kavya so far datable with reasonable certainty ; the Cloka, Upajati, Vasantatilaka, and Cardulavikrldita are affected. Very important is the fact that Prakrit lyric written in the Prakrit of the grammarians (Middle Prakrit) is preluded in two Prakrit Aryas, written in Old Cauraseni, which already manifest that affection for long compounds which is carried to excess in the Gaildavaha.

1 Cf. below, pp. 244, 2 ?6, 332- The evidence of slow development of use of narrative stanzas is cleai. For the priority of Arya Qura to the Vessanlara Jafaka, see R. Kick, I'estgabe Jacoln, pp. 145-59. x PREFACE

Kalidasa has suffered from attempts 1 to defy style by placing him before Acvaghosa, and to ignore 2 the use of his works in Vatsabhatti by ascribing him to the period 525-75, when no great Empire existed, on the strength of his picture of India in the Raghuvanga. Much more ingenious is an effort 3 to fix his home in Kashmir, and to trace in his poetry an adumbration of the Pratyabhijnacastra of that land, with its doctrine of recogni- tion of the unity of the divine love. Kalidasa would thus be a master of suggestion, which later was definitely developed in Kashmir as the essence of poetry by the Dhvanikara, who was doubtless not Anandavardhana. Use by Kalidasa of the Padma Purana has been suggested but is not plausible. His possible relation to the Vakatakas has been investigated, and use has been made of Ksemendra's ascription to him of a Ktmte^vara- dautya, but all is mere hypothesis. 4

Discussion of the migration of fables and other literature has failed to achieve decisive results. Some stress has lately been laid on the evidence of connexions between Egypt and India contained in the Oxyrhynchus Papyri? but it is difficult to believe seriously that Isis was worshipped in India as Maia, c as asserted with complete vagueness in the Isis litany, 7 and Professor Hultzsch's effort 8 to find Kanarese explanations for certain terms in the farce regarding Charition's adventures on the coast of a country bordering the Indian Ocean, are as little plausible as those of Sir G. Grierson to discover Sanskrit. It seems prima facie absurd to suppose that any Greek farce writer would trouble to embody passages, in foreign speeches which would be utterly unintelligible to his audience. 9

1 Kshetresacnandra Cnattopadhyaya, Allahabad Univ. Stud., ii. 80 ff. ; K. G. 3an- kar, IHQ. 1. 3090. But contrast IHQ. ii. 660 for Acvaghosa 's influence on Kalidasa's grammar.

s D. R. Bhandarkar, ABI. viii. 202-4.

s Lachhmi Dhar Kalla, Delhi University Publications, no. 1.

• See POCM. 1924, p. 6.

4 In ii. no. 300 a woman Indike appears.

6 xi. no. 1380. That Maya is meant is not probable.

7 iii. no. 413. 8 JRAS. 1904, pp. 399 ff.

9 Pischel's view that mixture of language is Indian specifically is disproved by Reich, DLZ 1915, p. 591. India was known in Egypt, but there is not the slightest ground to believe that any one knew Kanarese or Sanskrit well enough to reproduce either of them in a ■farce. PREFACE xi

It is indeed probable that no assured results can be expected regarding borrowing of tales ; Sir Richard Temple's ingenious suggestions 1 as to non-Aryan origins of certain motifs, with which may be compared those of Professor Przyluski 2 regarding the influence of Austro-Asiatic peoples on early Indian thought and speech, are inconclusive, nor is it clear that, as Dr. Qaster 3 inclines to hold, we owe to India the ideas of fallen angels, genii who return to earth, or legends of asceticism carried to ludicrous extremes. Dr. Gaster, however, rightly stresses the impossibility of assuming that India gave only and did not borrow, and insists on the importance of investigating the possibility of a literary origin for many fairy tales current among the people. Moreover, parallelism should often, it appears to me, be admitted in literary development. It is instructive, for instance, to compare the scheme of development of the practice of emboxing tales within tales given below (p. 320) for India with that suggested by Schissel von Fleschenberg i for Greek literature : the simple tale passes through stages illustrated by the Milesiaka of Aristeides, the work of Antonius Diogenes, the Golden Ass of Apuleius, and the romance of Petroniiu, to the complete outcome in later romance. The many motifs found in the Kathasaritsagara, for which parallels are adduced by the learned editor 5 of a new edition of Tawney's excellent version from western literature, suggest likewise that much may be said for the doctrine of parallelism.

On Civadasa's version pi the Vetalapahcavihgatika much light has been thrown by Hertel's researches. He establishes that Civadasa used a version in verse, whence some stanzas of merit, including those cited below (p. 290), are taken ; the many verse fragments found in his prose are explained by the origin of his work. Similar features are not rare in late texts, such as

1 Ocean of Story, i. pp. xiv ff.

J For other possibilities (Sumerian connexions) cf. Przyluski, BSL. xxvii. 218-29.

3 Ocean of Story, iii. pp. ixff.

4 EntwickehingsgescJtichte lies griechischen Romans im Altertum, and Vic gricch- ische Novelle ; cf. Reich, DLZ. ioif , pp. 543 f For the parallel development of the Helen and Sita legends, see Printz, Festgabe Jacob), pp. 103 ff.

6 N. M. Penzer, Ocean oj Story, ten vols., 1924-S. For elaborate notes on motifs see references in Indexes in each volume.

6 Streitberg Festgabe, pp. 135 ff. He places him not much before A. D. 1487. xii PREFACE

Meghavijaya's Pahcakhyanoddhdra, the textus simplicior of the Qukasaptati, the Madanarekhakatha t the Kiisamasarakatha, the Aghaiakumarakatha} and that version of the Vetalapaitcavih- fatika which goes back to Ksemendra's verse rendering. This, however, does not decide the question of the original form of the Vetalapancavihcatika; the common source of Ksemendra and Somadeva may have been in prose or prose and verse ; we have not sufficient evidence to show which. Hertel proves by com- parison of texts that Civadasa was deeply influenced in vocabu- lary and syntax by Old GujaratI, and concludes that he was a man of small education, belonging to the class who did not use Sanskrit as their ' Hochsprache ', but understood it tant Men que ma/, and endeavoured to express themselves in it.

The question of the authenticity of the dramas ascribed to Bhasa by the late T. Ganapati Castrl has been frequently dis- cussed since my Sanskrit Drama appeared, but without results of value, largely because the true issues have been misunderstood and effort has been devoted to proof of the obvious. It is true that it is not a matter of much importance whether the dramas be ascribed to Bhasa or to an unknown poet, but it is important to consider whether (i) they are all by one hand, and (a) by a writer earlier than Kalidasa and the Mrcchakatiha, Both these propositions seem to me clearly established, for, though some Indian and, less excusably, some European a scholars still seem not to have weighed the evidence adduced by Dr. Morgenstierne, the English protagonist against T. Ganapati Castrl's theory recognizes that the Carudatta must be placed before the Mrccha- katika. Priority to Kalidasa seems established by evidence of use by that poet, and of greater antiquity in technique, style, diction, metre, and forms of Prakrit ; it is significant that Kali- dasa has Maharas^rl, unknown to Bhasa. Moreover, it is perfectly clear that Bhasa's Prakrits, as revealed by the manuscripts of his plays, occupy a position intermediate between the Prakrits of Acvaghosa and of Kalidasa as shown by European critical 3 editions. It is no reply to this fact to point out that manuscripts

1 Trans. Ch Krause, Ind. Erz., iv.

• Nobel, ZII. v. 141 f. He sets £udraka and the ATrcchaialika before Kalidasa.

3 Indian editions, e.g. that of the AfcaryacSdamani, have not even the value of a MS. in this connexion. PREFACE xv

dattd ; the Bhdvaprakdf. a of Caradatanaya (13th century), knew a work not merely very similar in structure, but actually con- taining a verse found in the Trivandrum text. Sagaranandin in the Natakalaksanaratnakofa ascribes to the Svapnavasavadattd a passage which undoubtedly, as T. Ganapati Castri shows, is a paraphrase of a passage at the beginning of our text, not a citation from a variant text as Professor Levi suggested. 1 I agree also with T. Ganapati Castri that the passage cited by Ramacandra and Gunacandra in the Ndtyadarpana from Bhasa's Svapnavasavadattd could easily have found a place in our text, while in any event it is clear that that play contained a scene paiallel with one in our play. The most that can be made out from these facts against the ascription to Bhasa is simply that there were probably varying recensions of the piays. That, of course, may be taken for granted ; it was the fate of every much- studied and used play, and we have it exemplified to perfection in the case of Kalidasa, 2 the variations regarding whose works seem to have been unknown to or forgotten by those who refuse to recognize .Bhasa 's authorship of these dramas. There is no evidence at all to show that any of the veisions of the Qakuntala can be credited with any greater fidelity to the original of Kali- dasa than is possessed by the Trivandrum Svapnavasavadattd in relation to Bhasa's original. Moreover, it seems too often to be forgotten that variants may be due to the dramatist himself, who can hardly be supposed to have given his dramas a single peifectly definite text. It is, of course, tempting to adopt with Hermann Weller 3 the belief that the actors of Kerala have the responsi- bility for mangling our texts, and to accept the view that Bhasa is preserved to us in a deteriorated form, and that, for example, the Pratijiiayaugandhardyana and the Svapnavasavadattd made up a single piece. But I am satisfied that to accept this view is uncritical and is to substitute our pieferences for reality; the pedestrian character of some of Bhasa's stanzas can far better be explained by the simple fact of his early date; Kalidasa exhibits the influence of increased refinement of style in his dramas, just

1 JA cciii. 193 ff., followed in the very uncritical MASI. xxvui. 11.

2 Cf. also the recensions of the Uttararamacanta, Helvnlkar, JAOS. xxxiv. 428 ff.

3 Trans, of Svapnavasavadattd, p 8. Tne same theory applies, of course, to the Qakuntala. xvi PREFACE

as in his epics he normally avoids the pedestrian traits which are easily to be found in the epics of his forerunner Acvaghosa. The dramatic defects of Bhasa need not be ascribed to actors, for Kalidasa himself in any version of even the Qakimtala is far from perfect, and Shakespeare's flaws are notorious. On the other hand, we owe a very considerable debt to Hermann Weller 1 for showing in detail, with true insight into the nature of Bhasa's poetic talent, 2 that six of the stanzas which by the anthologies are attributed to Bhasa bear remarkable resemblance to the style ol stanzas in our dramas. We may dismiss as far-fetched the suggestion that the makers of anthologies ascribed them to him because they felt in them the spirit of his poetry; it is common sense to assume that the ascriptions are correct, and that they add one more link to the chain of evidence which ascribes the dramas to Bhasa, and vindicates the suggestion of a great Indian scholar. The effort 3 to strengthen the case for dating Dandin later than Bhamaha by using the evidence of the Avantisundarikatha and its Sara is clearly a complete mistake. The Katha should never have been published from one mutilated manuscript, whose readings, even if correctly stated, have already been proved wrong by other manuscript evidence. 4 Even, however, from the muti- lated text it was clear that Bharavi was not made out to be the great-grandfather of Dandin, who is given as Damodara. But, as Dr. D6 B has pointed. out, even the most careless reader of the Katha and the Dafakitmaracarita should have been struck by the extraordinary difference of style between the two works, the Katha rivalling unsuccessfully the worst mannerisms of the Har- sacarita and the Kadambari. If a Dandin wrote the work, he was assuredly not the author of the Dagakumaracarita, and its date may be centuries later than the great Dandin, for there is no reason to accept the suggestion 6 that the writer of the Katha lived sufficiently soon after the famous Dandin to be familiar

' Festgabe Jacobi, pp. 114-25.

3 Cf. Garbe's emphatic testimony, Festgabe Jacobi, p. 1 26, in contrast with ZII ii 250; ABA. viii. 17 ff.

8 J. Nobel, ZII. v. 136-52.

4 G. Harihar Sastri, IHQ. iii. 169-71.

6 IHQ. iii. 395 ff. As Dandin wrote according to Bhoja's Qrngaraprakafa (BSOS. iii. 282) a Dvisamdhanakavya, this may be his third work (cf. below, p. 296). « Ibid., p. 403. f

PREFACE xvii

with his genealogy and to work it into his story. It may be added that the effort 1 to find in v. 17 of the Katha. an allusion to kdvyatraya of Kalidasa, thus confirming the denial to him of the Rtusamhara, is wholly impossible and has not even the authority of the editor. It is very difficult to say whether we can derive from the Katha any assurance as to Bharavi's connexion with Visnuvardhana or identify the latter with the prince who became Eastern Calukya king in A.D. 615 and was the brother of that Pulakecin, whose Aihole inscription (a. D. 634) mentions Bharavi's fame, but at least there is no flagrant anachronism, though we know already of one literary forgery 2 which ascribes to Durvi- nlta of Kongani a commentary on Kiratarjuniya xv.

Of Abhinavagupta's important commentary on the Natya Qdstra we have now the beginning of an edition, which, un- happily, is fundamentally uncritical, 3 while a new effort 4 has been made to assign their precise shares in the Kavyaprakaga to its two authors, but without any convincing result ; in cases of this sort it is probably hopeless a priori to expect to find any conclu- sive evidence ; an editor who has to fill out lacunae is certain to adapt the whole more or less to his own style and to render restoration of the original and his additions almost impossible. 5

The curious scepticism which has marked the attitude of Indian and some European scholars towards Bhasa has not been shown in recent work on the Kautillya Arthafdstra, on which I have written in the Patna memorial volume in honour of that great Indian, Sir Asutosh Mookerjee. The only ground of this differentiation of treatment appears to be the sanctity ascribed to the written word : because the work in an obviously brer appended verse assures us it was written by Visnugupta, i. e. Kau- tilya — the reading Kautalya is clearly 6 of no value — therefore it

1 ZII. v. 143.

3 Ep. Cam., iii. 107. It is noteworthy that a Durvinlta appears in the Katha.

» Gaekwad Oriental Series 36, 1926 (i-vii) ; cf. S. K. De IHQ. iii. 859-68.

1 H. R. Divekar, JRAS. 1927, pp. 505-20 ; he assigns all the commentary to Alata as well as the Karikas from that on Pankara.

•The effort of Dr. D6 to ascribe Vallabhadeva's Subh&sitavali to the 1 3th cent, has been discnssed in a note to appear in BSOS. iv. (1928). As regards Kaviraja's date (below, p. 137), Achyutacharan Chaudhuri ascribes him to the 1 ith cent, as prot^gG of a king Kamadeva of Jaintia ; see IHQ. iii. 848 f.

• Cf. P. V. Kane, ABI. vii. 89 ; Jolly, ZII. v. 216-21. Bhandarkai's theory (ABI. vii. 65-84) of a verse original known to Dandin is incapable of demonstration.

mi b xviii PREFACE

must be so, although it seems patently absurd that the minister of an Emperor should confine his work to a moderate-sized kingdom, and should not once by word or allusion betray the name of the country for which and in which he was writing. Nevertheless there is nothing too fantastic to find defenders, though it is difficult not to feel that it is a very misplaced patriotism which asks us to admire the Arthacaslra as repre- senting the fine flower of Indian political thought. It would, indeed, be melancholy if this were the best that India could show as against the Republic of Plato or the Politics of Aristotle, or even the common-sense and worldly wisdom of the author ot the tract on the constitution of Athens, formerly ascribed falsely to Xenophon. Certainly fantastic is the elaborate theory worked out by J. J. Meyer in his translation, and in his treatise Uber das Weseu der indischen Rechtsschriften und ihr Verhaltnis zn einander zmd zu Kautilya (1927). These works, prodifced in great difficulties, contain, amid much that is unsound and despite disconcerting changes of view, valuable contribution / to our under- standing of Kautilya, and throw light on many of the obscure sides of Indian life. But the main thesis of the author, who seeks to distinguish two sharply severed streams of literature, the one Brahmanical, essentially concerned with magic, the other of the people, practical and legal, is clearly based on a false foundation. The effort to regard the Brahmins as something apart in Indian life is one of those delusions which may find sympathy in the non-Brahmanical classes in India and in Europe, but which run counter to all that we know of Indian thought, which owes its life and strength to the Brahmins, not to warriors or rulers, still less to the commonalty. The efforts of the author 1 £0 establish that the Arthacaslra was used by Yajiiavalkya are certainly without weight ; the evidence tends far more to show that the borrowing was the other way. Not a single passage referred to really favours the priority of the ArtUctfastra, but in several passages the obscurities of the

W. Ruben's defence of Jacobi's date {Festgabejacobi, pp. 346 ff.) is ineffective. For Kalidasa's relation to the Artha(d!tra, cf. K. Balasubrahmanya Ayyar, POCM. 1924, pp. 2-16.

1 PP- 6 5> 6 9> 7°. 7«> 77. 'J t. I S°, t33. I5 8 ~79. l 19r9°, "3. 216, 284, 290, 294, 299, 300. PREFACE xix

Arthagastra can be readily understood by realizing that it was drawing from Yajriavalkya. Nor does Meyer attempt systemati- cally 1 to prove that Manu is later than the Artkafdstra, though on his theory of dates that text is more than a hundred years at least posterior to the Art/iafdstra. He has been as unable as the Indian supporters of Canakya's 2 authorship to explain the silence which the Artkafdstra observes regarding everything imperial and its absolute ignoring of the facts as to Pataliputra. His further effort 3 to prove the late date of the Gautama Dharmafdstra is in itself less open to objection, but his con- tentions are largely inconclusive 4 and do little more than prove, what has always been admitted, that our text of that Dharma- castra has been considerably worked over. The main principles of the development of the legal literature remain as they were formulated by Max Muller and Biihler, and further established by Oldenberg and Jolly. Indeed, Meyer's own view at present 5 — his conclusions lack admittedly any great fixity — is that Bau- dhdyana and Apastamba are pre-Buddhist, Vasistlia belongs to the fouith century B. c, and Manu may be ascribed rather nearer to 200 B. c. than to A. D. 300 ; there is, however, no tolerable proof of Vdsistka's posteriority to Apastamba, still less that Apastamba is pre-Buddhist in date. Still less convincing again are Meyer's efforts 6 to assign Narada to a period anterior to Manu and Yajriavalkya ; if we take our present texts as the basis of argument, this is certainly out of the question ; if we recon- struct originals for all three, we lose ourselves in idle conjectures which, like all guesses, merely obscure knowledge. For Ydjtia- valkya there may be noted an interesting effort 7 to reconstruct the original Smrti on the basis of comparison with parallel texts in the Agni and the Garuda Purdnas. It is very possible that

1 What is said, e. g. p. 113, is quite inconclusive; contrast IHQ. ili. 812.

2 Jacobi (IHQ. lii. 669-75) holds that Canakya and Visnugupta were distinct persons later confused wilh Kaufilya. Canikya may be original, not Canakya.

3 See references at pp. 417, 418.

' for 1 further argument as to Gautama's later date, see Bata Krishna Ghosh, IHQ. in. 607-11.

5 Altmd. RechtsschnfUn, p. vii.

6 Ibid., pp. 82-114.

T Hans Losch, Die YajHavalkyasmrti (1927). The Garuda has a version of the Nidanasthana of the Astahgahrdaya and Astahgasamhitd; Festgabe Garbe, pp. 103 ff.

3H9 b 2 xx PREFACE

the parts of the text dealing with Rajadharma and Vyavahara have been amalgamated with a text dealing with the topics of the Grhyasutras ; but it is very dubious if it is possible to recover the original form of the Smrti. It is, of course, easy to eliminate certain obviously late passages, such as those dealing with the Vinayaka- and Graha-canti and the anatomical matter in Book iii, but the more radical analysis suggested is far less satisfactorily made out.

Of auxiliary sciences architecture has at last received expert treatment from Professor Prasanna Kumar Acharya in his Dictionary of Hindu Architecture and Indian Architecture} based on a new text and rendering of the Mdnasdra, for which the period of A. D. 500-700 is suggested. Striking similarities between the prescriptions of the Mdnasdra and Vitruvius are unquestion- ably established. Unhappily, the deplorable condition of the text of the Samardnganasutradhdra' 1 of Bhoja adds to the difficulty of valuing his remarks on architecture, town-planning, engineering, and the construction of remarkable machines, pro- bably akin to the mechanical toys of the Middle Ages. 3 The Principles of Indian Silpa Sdslra, with the text of the Maya- sdstra, by Phanindra Nath Bose, is also of value. 4 Hawking figures in a Qyainikacdstra by Rudradeva.

On the early development of logic an interesting light has been thrown by Professor O. Strauss's demonstration from the Mahdbhdsya 5 that Patanjali was well acquainted with the doctrine of the causes familiar from the Sdtkkhyakdrikd* why things in themselves visible are sometimes not seen, and also had some knowledge of the theory of the syllogism — how much, is not altogether certain. The evidence, however, is useful as supporting the view that our philosophical Sutras are essentially the outcome of a long period of development, and, whatever their date as we have them, contain doctrines much earlier in point of time. The effort to distinguish strata, though energetically pursued, leads to little that is certain. For instance, while we may readily believe

1 Oxford, 1927 ff. 2 GOS. 1924-5.

3 Ocean of Story, iii. 56 ff.

4 A text and trans, of a Qilpa Qattra are in print.

' Festgabe Garbe, pp. 84-94. See also Prabhat Chandra Chakravarti, IHQ. •■ 478 ff.

Verse 7 ; cf. Caraka, Sutrasthana, ix. 8. PREFACE xxi

that the Purvamimansa and the Vedanta Sutras represent a long period of working over, it is by no means clear that we can deduce 1 from a remark of so late a writer as Surecvara that Jaimini, the author of the Purvamimansa, also wrote a more philosophical £ariraka Sutra,th& first two Sutras of which corre- spond with those of the extant Vedanta Sutra. The fact that in these two Sutras, Purvamimansa and Vedanta, references are made both to Jaimini and Badarayana is best explained, not by assuming a number of Jaiminis and Badarayanas, but simply by recognizing that each text represents a long scholastic develop- ment and that the use of the names may not represent the views of the authors in question any more accurately than do, for instance, those of the Christian Fathers or the Scholastics the doctrines of Aristotle, or those of the neo-Platonists those of Plato. Nothing, of course, conclusive can be adduced against the belief in many Jaiminis or Badarayanas, and recourse has recently been had 2 to the same device to explain the fact that Prabhakara sometimes appears in tradition as later than Kumarila, while his work as known to us shows no certain trace of such a relation. In this case the suggestion is probably needless. The much discussed question of Dignaga's place in the history of Indian logic, in special his relation to Pracastapada, has been furthered by Dr. Randle's edition of Dignaga's fragments 3 ; it appears to me that Dignaga's priority is still the more probable view, but this issue, as well as the important contributions to our knowledge of Indian philosophy by Professor M. Walleser, Th. Stcherbatsky, Louis de la Vallee Poussin, S. Radhakrishnan, Das Gupta, O. Strauss, Masson Oursel, J. W. Hauer, Ryukan Kimura, Kokileswar Sastri, Mahendranath Sircar, and others, must be reserved for discussion elsewhere. Y. Kanakura * has shown that the alleged interpolations in Cankara's Bhasya are known to Vacaspati Micra, while the date adopted by me 5 for C/afikara is supported by Jinavijaya's proof that Haribhadra, whom C/afikara

1 S. K. Belvalkar, Ftstgabe Garbe, pp. 162-70; Ind. Phil. Rev., ii. 141-54, Contra, Nilakantha Sastri, IA. 1. 172.

s Stcherbatsky, Festgabe Jacobi, p. 372. What is said in FOCM. 1924, pp. 475 ff , 523 ff. is inconclusive.

8 The Nyayapravefa is now published in GOS. 32 (vol. ii).

• Festgabe Jacobi, pp. 381-5 ; on Anandajiiana, cf. p. 382, n. 1.

6 IOC. ii. 612. xxii PREFACE

used, falls in the period A. D. 700-770. Mention, however, should be made of the controversy which has raged over the authorship of the Nyayapraveca, which is ascribed with equal confidence to Dignaga 1 and to Cafikarasvamin 2 ; a final judgement is difficult, and the matter has been dealt with by me at length in an article to appear elsewhere. 3 It should also be noted that Professor Jacobi 4 has now admitted that the Nyaya Sutra knows the Vijnanavada system, on the ground that the Sutra in iv. 2. 26 deals with a Vijnanavada tenet found in the Lankavatara ; I have already dealt with this suggestion, ,r ' and pointed out that it possesses no cogency. Professor Jacobi's further suggestion that Vatsyayana knew Vasubandhu and may be placed c. 400 accords with the results adopted by me on the score of other evidence. He criticizes the well-known attempt of S. C. Vidyabhusana to prove that Uddyotakara and Dharmaklrti were contemporaries, on the ground that (1) Uddyotakara must have flourished a generation before Bana since he was known to Subandhu, and (2) Dharmaklrti cannot have attained literary fame before Hiuen Tsang's stay in India, since he ignores him as an author of standing. These arguments are not conclusive, and it is quite possible that Subandhu, Bana, Uddyotakara, and Dharmaklrti were more or less contemporaiies ; this issue also will be dealt with elsewhere. But Professor Jacobi renders it very probable that Dignaga, perhaps even Dharmaklrti, was known to the well- known Manimekhalai in Tamil. 7

On the interesting issue of the effect of Indian philosophy on Schopenhauer and of the present importance of that philosophy for western thought reference may be made to the Fwifcehntes Jahrbuch der Schopenhauer- Gesellschaft, 1928. An eneigetic polemic against the view of early influence of Indian on Greek philosophy has been delivered by Th. Hopfner, 8 which at least

1 Vidhushekhara Bhattacharya, IHQ. iii. 152-60.

2 Tubianski, Bulletin de VAcadimie de rUSSK. 1926, pp. 975 (T.

3 IHQ. 1928. « ZII. v. 305 f. B Indian Logic and Atomism, pp. 2${.

• Ibid., pp. 27 f.

' ZII. v. 305 ; the Nyayapraveca was used in the Manimekhalai (p. 309). On "the vexed date of the Cangam literature, cf. K. G. Sankar, JRAS. 1924, pp. 664-7.

• Orient und gritckische Pkilosopkie (1925). For a probably forged reference to

Apolloiiiui of Tyana in a Sanskrit text, see M. Hiriyanna, IHQ. 11. 415. PREFACE xxiii

has the merit of showing the precaiiousncss of the assumptions of such influence. Part of the argument for Indian influence rests on the belief in early dates for the Indian schools of thought, and it is clear that there is great difficulty in arriving at definite con- clusions on this issue. Thus Professor Das Gupta 1 places the Lankavatara before Acvaghosa, but the text we have seems to know the Vijnanavada school and the barbarian inroads of c. A. D. 500. Much stress has of late been laid on the Sarhkhya philosophy, 2 as it is presented in the Sarhhita of Caraka, but it seems to be overlooked that we have not the slightest proof that this or any special part of the text is really Caraka's. 3

Some light has been thrown by the discoveries of manuscripts in East Turkestan on the Bheda Saihhita* A paper manuscript with a fragment of the text, which can be assigned to the ninth century A. D., suggests strongly that the text published from a single Telugu MS. presents a version of the Sarhhita which has suffered alteration, a chapter on raktapitta in the Nidanasthana having been replaced by one on kasa. Another manuscript frag- ment, written on leather, from South Turkestan or Northern India, dating probably from the end of the second century A. D., say a hundred years before the manuscript of the Kalpanamandi- tika and fifty years after the manuscript of Acvaghosa's plays, is of interest, as it preserves a tradition of a doctrine of eight or ten rasas as opposed to the six which Caraka and Sucruta recognize, and which are generally accepted in Indian medicine. It is possible that we here have a trace of an older medical system, which was ultimately superseded by the system of Atreya, on which the work of Caraka is based.

The vexed issue of the indebtedness of Arabia and Europe to India for the numerical system has been reconsidered by Sukumar Ranjan Das, 6 who has dealt at length with Dr. Kaye's views.

1 Hist, of Indian Phil., i. 280.

Ibid., 1. 280 f., 312 ff.

5 Cf. Hoernle, Archivf. Gesch. d. Medizin, i. 30 ff. ; Jolly, Munich Catal., p. 48. The list of Tantrayuktis in viii is, of course, by Drdhabala, who again used the Uttaratantra of Sucruta ; Ruben, Festgabe Jacobi, pp. 354-7.

1 H. Liiders, Festgabe Garbe, pp. 148 ff. ; for the doubtful character of Caraka's text, see also pp. 154 f.

5 IHQ. ii. 97-120; iii. 356-75. See also D. E. Smith, Hist, of Math., vol. ii, ch. ii. xxiv PREFACE

Some of the evidence adduced is clearly inconclusive. The Artkafdstra knows (ii. 7) an elaborate system of keeping accounts, but its date cannot be assumed as the fourth century B. c, nor does in any case the keeping of accounts imply any definite system of the use of numerals similar to that attested for the sixth century A. D. 1 References to boys learning reckoning {smhkhyanri) 2 are equally inconclusive, and the date of the Lali- tavistara is very uncertain. But the use of fiinya in the Chandas- sutra of Pingala 3 must be accorded due weight, and the Indian hypothesis has gained strength from the new investigations accorded to it. But certainty is unattainable, and it may be observed that, while the identification of Pulica with Paulus of Alexandria is merely conjectural, it is not sufficient to dispose of it by pointing out that Pulica was an authority on astronomy, Paulus on astrology, for we have nothing to show that the latter did not deal with astronomy, as would be natural enough in a professed astrologer. 4

On the question of the origin of Sanskrit no conclusive evidence has been recently adduced. Professor Hertel's conviction of the late date of the Rgveda and of Zoroaster is not likely to secure general acceptance, despite its ingenuity, 6 nor is a recent and not less ingenious effort ° to show that the Aryans lived for a time together under strong Mitanni influences and only turned definitely east, to break up into Indians and Iranians, after the Mitanni dib&cle in the middle of the fourteenth century B. c. The deductions drawn from certain terms, and from the similarity of Qiva to the Himmcls-und Wettergott of Asia Minor, whose name in Mitanni was Tesup, and of Parvati to the Great Mother of Asia Minor, Hepa in Mitanni, and from the syllabic Brahmi script, are all suggestive, biit without probative force. Very interesting and worthy of serious consideration in the field of

1 The Sumeiians {c. 3000 B. c.) and the Egyptians had elaborate systems of account- keeping ; see D. E. Smith, Hist, of Math., i. 37 ff. J Arthafastra, 1. 5 ; Lalitavistara, x. 15.

3 viii. 29). ; Weber, IS. vm. 169, 444 ff. It must be noted that this part is not probably early, and is not to be assigned to the and cent. B. C. (IHQ. hi. 374).

4 On the ketus and their influence on men's fates, see Ballalasena's Adbhutasagara (12th cent.), and J. von Negelein, JFesigabe Jacobi, pp. 440 ff. ; Festgabe Garde, pp. 47-

53-

8 Oa Zoroaster's date ci. Keith, IHQ. iii. 683-9. • W. Porzig, ZII. v. 265-80. PREFACE xxv

comparative philology are the arguments recently adduced by Professor Max Walleser 1 to refute the at present accepted theory regarding the merger in Sanskrit of the three vowels a e o into a, and to show that Sanskrit preserved as late as the seventh century A. D. the labio-velar consonants. One point is of special interest, as it confirms a view in which I differ from Professor Liebich, 2 the question of the priority of the Taittiriya Pratifakhya to Panini ; it is made most probable that the distinction between a and a as connected with the openness of the former and the closed character of the latter vowel was not noted by the Rk or Taittiriya Pratifdkhyas but by the Atharva Pratifakhya, the Vajasaneyi Pratifakhya, and Panini. Liebich's argument against the priority of the Taittiriya Pratifakhya to Panini rests merely on the identity of certain Sutras in both texts and the use of pragraha for pragrhya. The latter appears to give no possible indication of relative position in time ; it may be a local variant, which accords with other evidence as to the provenance of the text ; the former fact is most naturally explained by the certainty that Panini's work embodies much earlier material, which was made use of also by the Praticakhya, unless Panini simply is the debtor to the Praticakhya.

In an exhaustive analysis of Yaska's etymologies 3 Dr. Hannes Skold has suggested that certain of the suggested derivations are only explicable on the ground that Yaska was familiar with and used a Middle Indian (Prakrit) speech. Beside this suggestion may be placed the opinion recently expressed by Professor H. Liiders, 4 that the language of Acoka's Chancery was ' eine Art Hochsprache ', while the actually spoken speech was much further advanced and probably had reached the stage represented in the literary Prakrits, though it is candidly admitted that the latter point cannot be said yet to have been established. Nor, it may be added, are Skold's proofs regarding Yaska free from much doubt. But the more important issue is whether the matter is really to be viewed in the light suggested, of a contrast between actually spoken language and a Hochsprache. It is rather, it appears to me, a matter of class speeches ; Yaska spoke Sanskrit

1 ZII. v. 193-202 ; Zur Aussprache des Sanskrit und Tibetischen (1926).

2 Zur Einfuhrung in die indische einheimische Sprachwissenscha.fi, ii. 47. " The Nirukta, pp. 128 ff. * ZII, v. 259. ' xxvi PREFACE

much as he wrote it, and the officials of Acoka equally conversed in a speech essentially similar to that in which they wrote, while contemporaneously lower classes of the population spoke in dialects which were far further advanced in phonetic change. It is clear that the Aryan invaders succeeded in imposing their speech on many of the earlier inhabitants of the country, and there is no cogent argument to refute the natural belief that strange Prakritic forms, such as we find sporadically even in the Rgveda, when not mere later corruptions are often loan-words from class dialects with which the speakers of the more con- servative form of speech were in contact. The influence of lower speech-forms was doubtless of increasing importance, since it evoked the elaborate grammatical studies summed up in the Astddhyayi, testifying to the anxiety of the priests to preserve the Bhasa from corruption, and Patanjali's insistence 1 on the evils of barbarisms doubtless proves their occurrence. But there seems no ground for conceiving of the position as one in which the priests used a formal language only in their business, and discarded it for a true vernacular in daily life. There seems a very fair analogy with the standard English of the higher classes of society in this country; the East-end curate's true vernacular is standard English, though he ought to be able to adapt his speech to the comprehension of the dockers if he works at a mission, and a landowner's true vernacular is that which he habitually uses in his own circle, not that in which he talks familiarly to his farm workers or villagers of the old type, whose dialect often is as different from standard English as an old Prakrit from Sanskrit. The presence of many Sanskritized ver- sions of Prakrit terms, to which Zachariae 2 has suggested an interesting addition in the term protha? is a perfectly natural phenomenon where higher and lower speeches exist contem- poraneously in the same community, apart altogether from the further possibilities of speech mixture due to the development

1 So already Katjayana, Varttilca 12 on Panini, i. 3. 1. Skold's effort (IA. lv. 181 ff.) to piove l'anini older than the Rk Pratifdkhya cannot be accepted, for the reasons given by B. Liebich, Zur Einfuhi-ung in die ind. cinheim. Sprachwissenschaft, ii. 30 f,

' ZII. v. 328-31.

5 A variant for pdntham in the verse cited (from Bhasya on Panini, 1. 4. 56) below, p. 46 Kor the idea cf. Qakuntald, iv. (ed. Cappeller), p. 48. PREFACE xxvii

of local as well as class dialects. At any rate arguments used to deny vernacular character to Sanskrit are quite adequate to prove the same hypothesis of standard English, which unquestionably is a true vernacular. 1

Moreover, the fact that Sanskrit was thus regularly used in conversation by the upper classes, court circles eventually following the example of the Brahmins in this regard, helps to explain the constant influence exercised by the higher form of speech on the vernaculars which reveals itself inter alia in the constant influx of Tatsamas, words whose phonetic state runs counter to the tendencies of the vernacular. It is quite impossible to explain this phenomenon adequately by the theory of borrowing from literature only ; those who adapted the vernaculars for the purpose of writing in any form or literary composition were doubtless in constant touch with circles in which Sanskrit was actually in living use. Doubtless, important changes to the dis- advantage of Sanskrit as a spoken language resulted from the Mahomedan invasions, which culminated in the substitution of a new speech in official use at the courts of Mahomedan rulers, but for the period from A.D. 300 up to 1200, dealt with in this work, there is little evidence of any fundamental change in the extent or character of the use of Sanskrit ; the same impression is given by the Kamasutra, perhaps c. 400, the Kavyamlmahsa of Rajacekhara (c. 900), and Bilhana (c. 1100).

On the vital chronological issue of Kaniska's date certainty has not yet been achieved; a case for A.D. 128-9 as the initial year of his era 2 has been made out, while his death in Khotan is assigned to 152. 3 This places him half a centuiy after A. D. 78, and it can only be said at present that the new dating, while it has many merits, none the less leaves unexplained difficulties.

1 An interesting loan-word is suggested in kampana or kampand (below, p. 170) by B. Liebich (Festgabe Strcitberg, pp. 230-2) who sees in it a derivative of campus. Liebich has a most amuting note (ZII. v. 153-63) showing how in Paticatantra, i. 7 (below, p. 257) the original version has a bug, not a flea, but the latter was introduced by Burz5e's version. BurzOe's alleged narrative is suspected by Sir E. Denison Ross {Ocean of Story, v. pp. vff. ; BSOS. iii. 443), but the existence of a Pahlavi rendering, which alone is of importance to Indologists, is not questioned.

2 W. E. van Wijk, Acta Orientalia, v. 168 ff.

• S. Konow, IHQ. iii. 851-6. The conclusions of this article are far from

certain. xxviii PREFACE

The affairs of Harsa have recently been considered once more, 1 with the usual indecisive results.

The necessity of economy of space, no less than the meagre resources of the Library of a University perforce incurious of Oriental Letters, has necessitated the reduction of bibliographical references to a minimum, but I have, I trust, passed over nothing of permanent value ; as in my Religion and Philosophy of the Veda and Upanishads, I have omitted such work as seems to display mere ingenuity or unscientifically to revive ancient errors. Specific acknowledgements will be found in the notes ; a more general debt is due to the historians of literature and the editors of anthologies, and I tender grateful thanks to Professors Macdonell, Peterson, Thomas, Weber, Oldenberg, von Schroeder, and Winter- nitz. By devoting special attention to matters of style and literary form I have endeavoured to avoid dealing at length with issues already effectively discussed by my predecessors. In my short sketch of Classical Sanskrit Literature, written in 1923 for The Heritage of India Series, 1 have anticipated many of the views which here are set out in detail and supported by further argument.

I have to express my most sincere appreciation of the willing- ness of the Delegates of the Press to publish this work as well as my Sanskrit Drama, and of the great assistance rendered to me in preparing it by my wife.

A. BERRIEDALE KEITH.

University of Edinburgh, February 1928.

1 Nihar Ranjan Ray, IHQ. iii. 769-92. Congratulations are due to the editor, Dr. Narendra Nath Law, of this most interesting Quarterly, in which there has already appeared much useful and suggestive work on .a wide range of topics. CONTENTS

Preface ....... vii

Kumaralata and the early Kavya, Sanskrit, and Prakrit viii

Kalidasa's Date and Place of Birth . . x

Greek and Indian Fables . . . . . x

The Dramas of Bhasa . . . . xii

Dandin and the Avantisundarlkatha . . . xvi

The Authenticity of the Arthafdstra . . . xvii

The Dates of the Philosophical Systems . . . xx

Medical Fragments from Turkestan . . . xxiii

The Indian Origin of the Numerals . . . xxiii

Sanskrit as a vernacular ..... xxiv

PART I. * THE LANGUAGE

I. Sanskrit, Prakrit, and Apabhranca. <>

i . The Origin of Sanskrit .... a. The Character and Extent of the Use of Sanskrit

3. The Characteristics and Development of Sanskrit

in Literature .....

4. The Prakrits .....

5. Apabhrafica .....

PART II. BELLES-LETTRES AND POETICS

II. The Origin and Development of Kavya Literature

1. The Sources of the Kavya

2. The Testimony of the Ramayana

3. The Evidence of Patanjali and Pifigala

4. Kavya in Inscriptions

5. The Kamasutra and the Poet's Milieu

III. Acvaghosa and Early Buddhist Kavya

1. Acvaghosa's Works

2. Acvaghosa's Style and Language

3. The Avadanas

4. Arya Cura and later Poetry

IV. Kalidasa and the Guptas

1. The Guptas and the Brahmin Revival

2. Harisena and Vatsabhatti

3

8

17 26

3a

39

39

4*

45 48

5*

55 55 59 64 67

74 74 77 xxx CONTENTS

3. Kalidasa's Life

4. The Rtusamhara .

5. The Meghaduta .

6. The Kumarasambhava

7. The Raghuvanca .

8. Kalidasa's Thought

9. Kalidasa's Style and Metre

V. Bharavi, Bhatti, Kumaradasa, and Magha

1. Bharavi

2. Bhatti . .

3. Kumaradasa

4. Magha

VI. The Lesser Epic Poets

VII. Historical Kavya

1. Indian Historical Writing

2. The Beginnings of History

3. Bilhana

4. Kalhana's Life and Times

5. The Rajatarangini and its Sources

6. Kalhana as a Historian

7. Kalhana's Style .

8. Minor Historical Kavya

VIII. Bhartrhari, Amaru, Bilhana, and Jayadeva

1. Bhartrhari

2. Amaru

3. Bilhana

4. Jayadeva .

IX. Lyric Poetry and the Anthologies

1. Secular Poetry

2. Religious Poetry .

3. The Anthologies .

4. Prakrit Lyrics

X. Gnomic and Didactic Poetry .

1. Gnomic Poetry

2. Didactic Poetry .

79 82

84

87 92

98 101

109 109 116 119

124

132

144 144

H7

158 161 164 169

172

J 75

• 75

183 188 190

199 199 210

322 223

227 227 236 CONTENTS

XI. The Didactic Fable .... i. The Origin of the Fable . %. The Reconstruction of the Paficatantra and its Origin .....

3. The Subject-matter of the Paficatantra .

4. The^Style and Language of the Paficatantra

5. The Derivative Forms of the Paficatantra

6. The Hitopadeca .... XII. The Brhatkatha and its Descendants

1. Gunadhya and the Brhatkatha .

2. The Brhatkathaclokasarhgraha of Budhasvamin

3. The Kashmirian Brhatkatha

4. Ksemendra's Brhatkathamafijarl .

5. Somadeva's Kathasaritsagara

XIII. The Romantic and the Didactic Tale

1. The Romantic Tale

2. The Didactic Tale

XIV. The Great Romances

1. The Age and Works of Dandin .

2. The Dacakumaracarita

3. The Content and Style of the Dacakumaracarita

4. Subandhu .....

5. The Vasavadatta ....

6. Bana's Life and Works

7. The Harsacarita ....

8. The Kadambarl ....

9. Bana's Style .... XV. The later Romances and the Campus

1. The Romances ....

2. The Campus ....

XVI. The Aims and Achievement of Sanskrit Poetry

1. The Aims and Training of the Poet

2. The Achievement

XVII. The West and Indian Literature .

1. The Fables and Marchen of Greece and India

2. The Translations of the Paficatantra

3. The Cukasaptati ....

4. Other Cases of Contact between East and West

242 242

246 248

255 259 263

266 266

272

275 276 281 288 288

293 296 296 297

299 3°7 308

3H 316

319 326

33i 33i 33^

338 338 344

35* 352 357 359 359 xxxii CONTENTS

5. The Romance in Greece and India . 365

6. The Hexameter and Indian Metre . -37° XVIII. Theories of Poetry . .... 372

t. The Beginnings of Theory on Poetry . 37a

2. The Early Schools of Poetics . . . 375

3. The Doctrine of Dhvani . . 386

4. The Critics and Supporters of the Doctrine of

Dhvani ..... 391

PART III. SCIENTIFIC LITERATURE

XIX. The Origin and Characteristics of the Scientific

Literature ..... 403

1. The Origin of the Castras . . . 403

2. The Characteristics of the Scientific Literature 406 XX. Lexicography and Metrics . . .412

1. The Origin and Characteristics of Sanskrit

Lexicography . . . . .412

2. The Extant Lexica .... 413

3. Treatises on Metre .... 415

4. The Metres of Classical Poetry . . 417 XXI. Grammar ...... 422

1. The Beginnings of Grammatical Study . 422

2. Panini and his Followers . . . 423

3. The Later Schools . . . . 431

4. Grammars of Prakrit .... 433 XXII. Civil and Religious Law (Dharmacastra) . . 437

1. The Origin of the Dharmacastras . -437

2. The Smrti of Manu .... 439

3. The Later Smrtis .... 445

4. The Digests of Law' .... 448 XXIII. The Science of Politics and Practical Life (Artha-

castra, Nlticastra) .... 450

1. The Origin of the Arthacastra . . . 450

2. The Content and Form of the Kautiliya Artha-

castra ...... 452

3. The Authenticity of the Arthacabtra . . 458

4. Later Treatises ..... 462 5- Ancillary Sciences .... 464 CONTENTS XXIV. The Science of Love (Kamagastra)

XXV. Philosophy and Religion .

i. The Beginnings of Indian Philosophy

2. The Purvamlmansa

3. The Vedanta

(a) The Doctrine of Non-duality and

{b) Ramartuja .

(c) Other Commentators

4. Theology and Mysticism

5. Logic and Atomism

6. The Samkhya and Yoga Schools

7. Buddhism

8. Jainism ....

9. Carvakas or Lokayatas .

10. Historians of Philosophy

11. Greece and Indian Philosophy .

XXVI. Medicine ....

1. The Development of Indian Medicine

2. The Older Sarhhitas

3. The Medical Tracts in the Bower MS.

4. Later Medical Works .

5. Greece and Indian Medicine XXVII. Astronomy, Astrology, and Mathematics

1. The pre-scientific Period

2. The Period of the Siddhantas .

3. Aryabhata and later Astronomers

4. Aryabhata and later Mathematicians

5. Greece and Indian Mathematics

6. Varahamihira and early Astrologers

7. Greece and Indian Astrology

8. Varahamihira's Poetry .

9. Later Works on Astrology

English Index Sanskrit Index .

Illusion

XXXlll

467 471

47 1 472

474

475 478

479

479

48a

487 491

497 498

499 500

5o5 5°5 506

509 510

5i3 516 516

5i7 521

5*3 5*5 5*8

53° 53* 534

537

559 ABBREVIATIONS

ABA. Abhandlungen der Berliner Akademie der Wissenschaften, philol.-

histor. Klasse. ABayA. Abhandlungen der Bayerischen Akademie der Wissenschaften,

phil. Klasse. ABI. Annals of the Bhandarkar Institute.

AGGW. Abhandlungen der konigl. Gesellschaft der Wissenschaften zu

GGttingen, philol.-histor. Klasse. AKM. Abhandlungen fur die Kunde des Morgenlandes. AMG. Annales du Muse'e Guimet. AMJV. Sir Asutosh Mookerjee Silver Jubilee Volumes. AnSS. Anandacrama Sanskrit Series, Poona. ASGW. Abhandlungen der philol.-histor. Klasse der konigl. Sachs.

Gesellschaft der Wissenschaften. BB. Bibliotheca Buddhica, St. Petersburg.

BBeitr. Beitrage zur Kunde der indogermanischen Sprachen, herausgeb.

von A. Bezzenberger. BEFEO. Bulletin de l'ecole francaise d'Extreme Orient. BenSS. Benares Sanskrit Series. BI. Bibliotheca Indica, Calcutta.

BSGW. Berichte iiber die Verhandlungen der konigl. Sachs. Gesellschaft

der Wissenschaften zu Leipzig, philol.-histor. Klasse. BSL. Bulletin de la Socie'te' de Linguistique de Paris.

BSOS. Bulletin of the School of Oriental Studies, London Institution. BSS. Bombay Sanskrit Series.

ChSS. Chowkhamba. Sanskrit Series, Benares. DLZ. Deutsche Literaturzeitung.

EH I. Early History of India, by V. A. Smith, 4th ed., Oxford, 1924.

EHR. English Historical Review. EI. Epigraphia Indica.

ERE. Encyclopaedia of Religion and Ethics.

GGA Gottmger gelehrte Anzeigen.

GIL. Geschichte der indischen Litteratur, by M. Winternitz.

GN. Nachrichten von der konigl. Gesellschaft der Wissenschaften zu

Gdttingen, philol.-histor. Klasse. GSAI. Giornale della Societa Asiatics Italiana. Haeberlin. Kavyasamgraha, by J. Haeberlin, Calcutta, 1 847. Haraprasad, Report I, II. Report on the Search for Sanskrit MSS., 1895-

1900, 1901-6. HOS. Harvard Oriental Studies, ed. Charles Lanman. IA. Indian Antiquary. xxxvi ABBREVIATIONS

IF. Indogermanische Forschungen.

IHQ. Indian Historical Quarterly.

IOC. India Office Catalogue of Sanskrit Manuscripts.

IS. Indische Studien, ed. A. Weber.

IT. Indian Thought, Allahabad.

JA. Journal asiatique.

JAOS. Journal of the American Oriental Society.

JBRAS. Journal of the Bombay Branch of the Royal Asiatic Society,

JPASB. Journal and Proceedings of the Asiatic Society of Bengal.

JRAS. Journal of the Royal Asiatic Society.

KM. Kavyamala, Bombay.

KZ. Zeitschrift fur vergleichende Sprachforschung.

MASI. Memoirs of the Archaeological Survey of India.

MSL. Memoires de la Socie'te' de Linguistique de Paris.

NSP. Nirnaya Sagara Press, Bombay.

OC. Orientalistenkongresse.

POCM. Proceedings of the Third Oriental Congress, Madras, 1924.

POCP. Proceedings and Transactions of the First Oriental Congress,

Poona, 1 919.

RHR. Revue de l'histoire des religions, Paris.

RSO. Rivista degli studi orientali, Rome. w

SBA. Sitzungsberichte der Berliner Akademic der Wissenschaften.

SBayA. Sitzungsberichte der Bayerischen Akademie der Wissenschaften,

philol.-histor. Klasse.

SBE. Sacred Books of the East, Oxford.

SBH. Sacred Books of the Hindus.

SIFI. Studi Italiani di Filologia Indo-Iranica.

SWA. Sitzungsberichte der Wiener Akademie der Wissenschaften.

TSS. Trivandrum Sanskrit Series, ed. T. Ganapati £astrl.

VizSS. Vizianagram Sanskrit Series.

WZKM. Wiener Zeitschrift fur die Kunde des Morgenlandes.

ZDMG. Zeitschrift der Deutschen Morgenlandischen Gesellschaft.

ZII. Zeitschrift fur Indologie und Iranistik. PART I

THE LANGUAGE SANSKRIT, PRAKRIT, AND APABHRANCA i . The Origin of Sanskrit

SOMETIME in the course of the second millennium B.C. Indo-European tribes occupied, in varying degrees of com- pleteness, vast areas in Iran, Asia Minor, and north-west India. 1 The problems of their movements and affiliations are still far from solution, but on linguistic grounds we postulate a group conveniently styled Aryan, whose speech can be regarded as the ancestor of the speeches of India and Iran. Of these Indian speeches" our oldest evidence is the Rgveda, and the language of this great collection of hymns is obviously a hieratic and conventional one. It testifies to the cultivation of sacred poetry by rival families of priests among many distinct tribes during a considerable period of time, and in various localities. Some of the hymns were doubtless composed in the Punjab, others in the region which in the Brahmanas is recognized as the home of the Kurus and Paficalas, tribes representing the con- solidation of units familiar to us in the Rgveda. It is even claimed that Book vi is the poetry of the period before the tribes entered India proper, though the contention is still implausible. That, under these circumstances, the speech of the Rgveda should show dialectic mixture is only to be expected, and, despite the great difficulties involving the attempt to discriminate, some progress is possible towards determining the characteristics of the dialect which lies at the basis of the Rgveda. It was marked by the open pronunciation of intervocalic d/i, bit, d, and dh as h, I, and Ih ; by the change of / into r ; and by the intrusion of the pronominal instrumental plural termination ebhis into the

1 Cf. Keith, Religion and Philosophy of the Veda, Chap. I.

2 Cf. V ackernagel, Altind. Gramm., i, pp. lxff. ; H. Reichelt, Festschrift Strett- berg (1924), pp. 238 £T. ; Macdonell, Vedic Grammar (1910); Meillet, IF. xxxi. i2off. ; JA. 1910, li. 184 IT.; Mllanges Livi, p. 20; Grammont, MSL. xix. 354 ff. ; Bloch, Formation de la langue maratke (1920); S. K. Chatterji, Bengali (1926).

B 2 4 SANSKRIT, PRAKRIT, AND APABHRANCA

nominal declension. Borrowings from other dialects can here and there be confidently asserted ; in some cases the forms thus found may be regarded as of equal age with those of the Rgveda, as in the case of words in / and jajjhati, with^/Vz in lieu of ks for Aryan gzh, but in other instances we find forms 1 which are phonetically more advanced than those normal in the Rgveda, and attest loans either from tribes whose speech had undergone more rapid change, perhaps as the result of greater admixture with non-Aryan elements, or from lower classes of the population. Thus we have irregular cerebrals as in kata beside krta, kata be- side karta ; ch for ps in krchra ; jy for dy in jyotis ; i for r in githira ; busa for brga, and many other anomalous forms. To localize these dialects is in the main impossible ; the rhotacism of the Rgveda accords with its western origin, for the same phenomenon is Iranian. The use of / is Jater a sign of eastern connexion, and in one stereotyped phrase, sure du/titd, we per- haps find e (or'az, as in the eastern Prakrit.

From the language of t he R gveda_yif can_trace a steady dev^lo^rire"rTr~to~CrassTcal_ Sanskrit, through the later. Sarhhitas and the Brahmanas. The development, however, is of a special 'kind; if is hot the spontaneous growth of a popular speech un- hampered by tradition and unregulated by grammatical studies. The language of the tribes whose priests cherished the hymns of the Rgveda was subject doubtless to all the normal causes of speech change, accentuated in all likelihood by the gradual- addition to the community of non-Aryan elements as the earlier inhabitants of the north, Munda or Dravidian tribes, fell under their control. 2 But, at least in the upper classes of the population, alteration was opposed by the constant use of the sacred language and by the study devoted to it. Parallels to such restricted evolution are not hard to find ; the history of the Greek Koine, of Latin from its fixation in the first century B.C., and of modern English, attests the power of literature to stereotype. In India

1 In some cases, no doubt, forms have been altered in transmission.

2 Cf. W. Petersen, JAOS. xxxii. 414-28 ; Michelson, JAOS. xxxiii. 145-9 '< Keith, Camb. Hist. India, i. 109 if. Common sense renders Dravidian and Munda influences inevitable, though proof may be difficult ; Przyluski, MSL. xxii. 305 ff. ; BSL. xxiv. uo, 355ff., xxv. 66ff. ; Bloch, xxv. iff.; Levi, JA.ccin. 1-56. Przyluski endeavours to prove Austroasiatic influence on culture ; JA. ccv. 101 fl". ; ccviii. 1 fl". ; BSL. xxvi. 98 ff. Cf. Poussin, Indo-europiens, pp. 198 ff. ; Chatterji, i. 170 ff., 199. THE ORIGIN OF SANSKRIT 5

the process was accentuated by the remarkable achievements of her early grammarians whose analytical skill far surpassed any- thing achieved until much later in the western world. In the normal life of language a constant round of destruction and reconstruction takes place; old modes of expression disappear but new are invented ; old distinctions of declension and con- jugation are wiped out, but new differentiations emerge. In Sanskrit the grammarians accepted and carried even farther than did contemporary vernaculars the process of the removal of irregularities and the disuse of variant forms, but they sanctioned hardly any new formations, producing a form of expression well ordered and purified, worthy of the name Sanskrit which the Rdmayana first accords to it. The importance of the part played by religion in preserving accuracy of speech is shown by the existence of a special form of sacrifice, the SarasvatI, which was destined to expiate errors of speech during the sacrifice, and in the Mahabhasya of Patanjali (150 B.C.) it is recorded that there were at one time seers of great knowledge who in their ordinary speech were guilty of using the inaccurate yar vd nas tar vd nah for yad vd nas tad vd nah, but who, while sacrificing, were scrupulously exact.

The influence of the grammarians, whose results were summed up in Panini's Astddhydyi, probably in the fourth century B.C., is seen in the rigid scheme of euphonic combination of the words within the sentence or line of verse. This is clearly artificial, converting a natural speech tendency into something impossibly rigid, and, as applied to the text of the Rgveda, often ruining the metrical effect. Similar rigidity is seen in the process which sub- stitutes in many cases y and v for the iy and uv of the earlier speech. Dialectic influence may be traced in the recognition of / in many words in lieu of r, and a certain distinction between the dialect which underlies the Rgveda and- that of Panini is revealed by the absolute ignoring by the latter of the substitution of / and Ih for d and dk. x Otherwise the chief mark of progress is the gro'wth of the tendency to cerebralization, possibly under Dravidian influence.

In morphology there was elimination of double forms ; a as a variant for ena in the instrumental singular of a stems disappeared,

1 Cf. Luders, Ftstschttfl Watkernagil, pp 294 ff. 6 SANSKRIT, PRAKRIT, AND APABHRANQA

a and a yielded to au in the dual, asas to as, a to tint, ebhis to ais, dm to anam in the plural ; ni alone is permissible in the locative singular of an stems ; the effective distinction of root and derivative stems in J disappears; the intrusion of weak forms into the place of strong and vice versa is banished ; the irregular vas of the vocative of vant stems is abandoned, and by eliminating the nominative yuvam and ablative yuvat the pronominal declen- sion is harmonized with the simplicity of the three forms of the nominal. Similarly, in verbal forms the variant ntasi in the first plural active is laid aside, the e of the third singular middle yields to te, 4hva in the second plural to dhvam, and forms in r in the third plural are confined to the perfect and the root gi; in the imperative dhvat is dropped, and dhi is no longer permitted to rival hi in the second person. Far more important is the laying aside of the subjunctive, whose functions were felt to.be adequately performed by the optative, save in so far as a com- plete set of forms was made up for the imperative by utilizing the first persons. Even in the optative the wealth of forms is seriously diminished, only the present and a specialized precative being allowed. The rich variety of infinitives is steadily lessened ; the final result allows only that in turn, while of the gerunds that in tva supersedes tvl and tvdya. Against these losses can be set little more than the development of two forms of periphrasis, the future middle in take, and the perfect l composed of a nominal accusative form with the auxiliaries kr, bhu, or as, the extended use of gerundives in tavya and antya, the creation of a perfect active participle in tavant, the invention of a new third singular aorist passive as in adayisi, and the development of tertiary verbal forms.

In some of these losses Sanskrit keeps pace with popular speech, but the evidence is conclusive against ascribing too much weight to this fact. While such categories as the dual of noun and verb alike, the middle, and the past tenses, practically vanished from popular speech, Sanskrit rigidly retains them. On the other hand it rejects irregularities which popular speech permitted to survive, such as the a of the instrumental singular and nomina- tive plural neuter of a stems, the asas of the masculine plural, the

1 On changes in the use of verbal forms see L. Renou, La valtur du parfait dans Us hymnes vldiques (1925), pp. 88 ff., 188 ff. THE ORIGIN OF SANSKRIT 7

form gondiity the pronominal plurals asme and yusme, the short forms yat and tat, and verbal forms in r. Traces of the sub- junctive, the infinitive in tave, the aorist akar, the instrumental in ebhis exist in Prakrit, but are banned in Sanskrit. On the other hand, although Panini recognizes fully the Vedic accent, it can hardly be doubted that already by his time in actual speech in many regions it had yielded in part to an expiratory accent. The tendency to such a result is already visible in the Rgveda, where duhita by the testimony of the metre must at times be read dhita, comparable with Pali dhita ; ' the weakening of bh and dh to h occurs there normally after unaccented syllables, 2 and the curious mode of notation of the accent in the Qatapatha Brahmana has with some ground been ascribed to a stage of transition from the musical to the expiratory accent. 3

We must not, however, exaggerate the activity of the gram- marians to the exte nt of suggesting with some writers that Classic al Sanskrit i s an art i ficial creation, a product* of the .brahmins whe n they sought to counteract the Buddhist creation of an artistic literature "I nYali b y reca sting thek_QjyjL .Pll.kntjc_ speech with the aid~o~f~the Ve dic langua ge^ It is, in point of tact, perfectly obvious that there is a steady progress through the later Samhitas, the Brahmanas, and the Aranyakas and Upanisads, and that the Bhasa, the spoken language of Panini's grammar, is closely related to, though not identic with, the language of the Brahmanas and the older Upanisads. Nor in point of fact does Classical Sanskrit present the appearance of an artificial product ; simplified as it is in comparison with the redundant luxury of the Vedic texts, it yet presents no artificial symmetry, but rather admits exceptions in bewildering profusion, showing that the grammarians were not creators, but were en- gaged in a serious struggle to bring into handier shape a rather intractable material.

1 Luders, KZ. xlix. 236 f.

2 Wackemagel, Altind. Gramm., i. 253 f. 5 Leumann, KZ. xxxi. a 2 f.

• Hoemle and Griersou, Bihari Did., pp. 33 ff. ; Senart, JA. s£i. 8, viii. 318 fF.

Contrast Franke, B. Beitr., xvii. 86; Boxwell, Tram. Phil. Soc. 1885-7, pp. 656 ff. Ponssin (Indo-euwpiens , pp. 191 ff.) stresses the literary character of Sanskrit. 8 SANSKRIT, PRAKRIT, AND APABHRANQA

2. The Character and Extent of the Use of Sanskrit

We have seen that the Sanskrit of the grammarians is essentially a legitimate development from the Vedic speech ; it remains to consider the extent of its use, in the time of Panini and later. In examining the matter it is essential to remember the social conditions of India. In Britain to-day the varieties of English spoken and written are complex and numerous ; in India, where caste, clan, and racial distinctions were far more prominent and important, linguistic facts were far more com- plicated still. What is clear 1 is that Sanskrit represents the language of B rahma nical civilization^ and the extent oX^ihat T civilizafTorTwas ever^ncreasinp^ though thg_J|rajimanical. religion had to face competition from new faiths, in special Buddhism "and" JaTnism, from__the Jifth c entury B. C. x Th~e Buddhist texts themselves afford the most convincing evidence of all of the predominance of Brahmanism ; the Buddha is represented as attempting not to overthrow the ideal of Brahmanism, but to change its content by substituting merit in place of birth as the hall-mark of the true Brahmin. The public religious_ rites and the domestic ritual were recorde d and c ar ried out in SanskrhVand education wasTn"Brahmin hands. The Buddhist texts repeatedly conhrm trie "Brahmanical principle that instruction of the people {lokapakti) was the duty of Brahmins, and the tales of the Jatakas 2 show young men of all classes, not merely Brahmins but boys of the ruling class, Ksatriyas, and children of the people, Vaicyas, seeking instruction in the north from Brahmin teachers. Sanskrit was the. ianguage__of science, not merely grammar, prosody, astronomy, phonetics, etymology, buf Hoiibt- Iess also of more magic arts," such as ""the pTj jysjflgngmy and_ "Semonology recorded in the Buddhist texts and confirmed by the inclusion of-magiq SarpajanavKlyaTand DevajanaVrdya in the list of the subjects taught by the Brahmin to the people given in the Qatapatha Brahmana? The same text* mentions also

1 'Ihomas, JRAS. 1904, pp. 465 ff. 2 Kick, SoaaU Gliedirung, p. 131.

3 xih. 4. 3. 9 ff.

• xi 5.6.8. Cf Brhadarauyaka Upanisad,.$. 10; iv. 1. 3 ; 5. n ; Chandogya,

vii. 1. a; Faddegon, Act. Or. iv. 4 ff., 133. Yakovakya perhaps denotes the dialogues which develop into philosophy. CHARACTER AND EXTENT OF THE USE OF SANSKRIT 9 Anucasanas, Vidyas, Vakovakya, Itihasa, Purana, Gathas, and Naracansis, and the continuity of tradition is attested by the Mahabhasya l which includes under the range of Sanskrit speech the four Vedas with their Angas and Rahasyas, the Vakovakya, Itihasa, Purana, medicine. The Agvalayana Grhyasutra? pro- bably not far removed from Panini in date, repeats in the main the list of the fatapat/ia, but adds Sutras, Bhasyas, Bhdrata, Mahabharata, and the works of the Dharmacaryas. Other sciences such as those of the bow, music, architecture, and politics are recorded in the Mahabharata? and, so far as they were in the hands of the Brahmins, we need not doubt that Sanskrit here also had its place.

These facts are not in dispute, and the predominance of San- skrit in the sphere in question remained unchallenged until the Mahomedan invasions brought a new literary language into prominence. The evidence indicates clearly that Sanskrit must have been in constant use as a means of teaching and performing .religious duties arrrong the Brahmins at least. It has been denied that it was really even their vernacular in the time of Panini, and a fortiori later, but the evidence for this view is unsatisfactory. Panini has rules 4 which are meaningless for any- thing but a vernacular, apart from the fact that the term Bhasa which he applies to the speech he teaches has the natural sense of a spoken language. Thus the doubling of consonants is ex- pressly forbidden in passionate speech, as in the term of abuse putradinl applied to' a cruel mother ; he prescribes the use of prolongation in the case of calling from a distance, in greeting, question, and reply ; he gives information on the terminology of dicing and the speech of herdsmen ; he cites expressions redolent of real daily life. Indeed, it is the grammarians alone who preserve for us such usages as the repetition of the second person imperative followed by the present indicative to express intense action : khdda khadeti khadati, ' eagerly he eats ', whence we have in colloquial Marathl kha kha. khato ; other popular uses are udarapuram bhunkte, ' he eats filling his belly ' ; dandadandi £egakegi, l a. struggle in which sticks are brandished and hair is

1 i, 9. s iii. 3. 1 ; 4. 1. Cf. Utgikar, POCP. 1919, ii. 46 ff.

3 Hopkins, Great Epic, pp. 1 1 ff.

4 Wackernagel, 1, p. xliii ; Bhandarkar, JBRAS. xvi. 330. io SANSKRIT, PRAKRIT, AND APABHRAN^A

pulled ' ; atra khadatamodata vartate, 'eat and enjoy ' is the rule here ; jahistambo 'yam, ' he is one who says " strike the sheaves of corn"'. They record also the parenthetical use 1 of manye, ' I think ' ; the humorous apacasi, ' you're no cook ' ; and authorize such quaint forms 2 as yamaki, 'I go '. The elaborate rules regarding the accent reflect also actual speech.

ConGrmatory evidence can also be adduced from the references of Yaska, 3 Panini, and Katyayana to particular usages of the northerners and the eastern peoples ; Katyayana also recognizes as a matter of notoriety the existence of local variations, which Patanjali illustrates by reference to the practice of the Kambojas, Surastras, Pracyamadhyas, &c. Here too may be mentioned the references of Katyayana and Patanjali to changes in usage after Panini's time, as when the former 4 finds fault with Panini for not giving nama as well as naman as the vocative, for not mentioning that pronominal forms are permitted in the masculine as well as in the feminine singular of dvitiya and trtiya, and for allowing only the feminines upadhyayi, dryd, ksatriya, and matulani. Patanjali shows us that in his time participial phrases had superseded the second person perfects such as tera, usa,peca, a fact specially characteristic of a genuine living speech. 6

Further information of a precise character is incidentally given us by Patanjali. 8 He insists that grammar does not exist to create words, but to make clear what are correct uses; in ordinary life (loke) a man thinks of a thing and uses the appro- priate word without going to a grammar ; the words of Sanskrit are of ordinary life (laukika). We find a grammarian and a charioteer (sutd) engaged in a discussion conducted in Sanskrit, and the latter has decided opinions of his own on the etymology of his designation and on that of the term prajitr, driver. The norm of speech is that of the £istas, and these are people who speak correct Sanskrit without special tuition ; the purpose of grammar is to enable us to recognize who are Qistas, and thus to

1 As in Pali ; Franlce, ZDMG. xlvi. 311 f.

2 Keith, JRAS. 1915, pp. 502 ff.

3 tfirukta, ii 2 ; v; 5, Mahabhasya^ 1. 9 ; v. 8 on vii. 3. 4s.

4 Bhandarkar, JBRAS. xvi. 273. Cf. Macdonell, Vedic Grammar, p. 307, n. 2. 6 Bloch, MSL. xiv. 97; L.Renou, La valtur du parfait, p. 189.

8 vi. 3. 109; Bhandarkar, JBRAS. xvi. 334 ff. Grierson (JRAS. 1904, pp. 479 ff.) misunderstands the passage to mean lhat Cistas require to be taught Sanskrit. CHARACTER AND EXTENT OF THE USE OF SANSKRIT n

apply to them to find the correct form of such terms as prsodara, which do not fall under the ordinary rules of grammar. The Cistas are further defined as Brahmins of Aryavarta, the region south of the Himalayas, north of Pariyatra, east of the Adarca, west of the Kalakavana, who are not greedy, who do good dis- interestedly, and who store only so much grain as a pot can hold. Other persons may make errors ; thus they may pro- nounce sasa for gaga, palasa for palaga, manjaka for tnancaka ; or they may commit graver errors by using incorrect forms {apagabdd) such as kasi for krsi, disi for drgi, gavi, goni, gota, gopotalika for gaits, or even verbal forms such as anapayati x for ajnapayati, vattati for vartate, and vaddhati for vardhate. But from the Cistas they could acquire the accurate forms. This suggests a close parallel to modern conditions in England, where an upper educated class sets the norm to all those in lower social classes ; the speech of that class is clearly a living language, and Sanskrit was so in much the same sense. The standard com- parison of Latin in the Middle Ages is somewhat unsatisfactory ; in the earlier period of the use of Sanskrit it is clear that it was much more closely similar to the speech of the lower classes in its numerous varieties than was Latin in medieval Europe. Comparison of Sanskrit with the dialects of the inscriptions of Acoka is significant in this regard ; their differences are not essential nor such as to hinder mutual comprehension, and could easily be paralleled in English speech to-day.

Moreover, the conclusions thus attained are directly supported by the evidence of the drama, in which Brahmins and kings and other persons of high station and education use Sanskrit, while inferior characters employ some form of Prakrit. It has been attempted to argue against this view on the score that the drama was originally in Prakrit; and that Sanskiit was introduced only when it became essentially the general language of culture. But this contention ignores the fact that on one side at least the drama is closely connected with the epic in Sanskrit ; Bhasa, indeed, has one drama without Prakrit, and there is little of it in his other dramas based on the epic. Nor was the Sanskrit

1 So A^ka's Brahmagiri inscr. I ; vafoati (the usual single consonant is merely graphic ; CII. i, p. lix ; Grierson's argument (JRAS. 1925, p 228) from the writing of other conjuncts is clearly untenable) occurs in Delhi-Topra, iv. 20. 12 SANSKRIT, PRAKRIT, AND APABHRAN^A

unintelligible in early times at least to the audience, which might be one including persons of quite humble rank ; the Natyagastra expressly lays it down that the Sanskrit is to be such as is easify intelligible to every one. The denial that realism was ever aimed at in the use of language by the characters in the drama is negatived by the facts ; the Prakrits used by the dramatists show a steady advance from those of Acvaghosa through those of Bhasa to the dialects of Kalidasa, who introduced to the stage the Maharastrl which, earlier unimportant, had won fame in India as the medium of erotic lyric. 1 The evidence of Acvaghosa is of special value, for it attests the fact that about A. D. ioo the stage tradition was so firmly in favour of the use of Sanskrit by the persons of the highest rank that he adopted it in his plays despite their Buddhist theme, and despite the fact that the Buddha himself, according to tradition, had forbidden the employment of Sanskrit as the medium for preserving his sayings. 2

The extent to which Sanskrit was used or understood is further attested by the epics. It is perhaps hardly necessary now to do more than mention the implausible conjecture 3 which ascribes the writing of the epics in Sanskrit to some period after the Christian era and sees in them translations from some Prakrit. The silence of antiquity on this vast undertaking is inexplicable, and it is incredible that the translation should have taken place at a period when Buddhism was triumphant and Brahminism comparatively depressed. The language itself has a distinctive character which renders the idea of translation absurd ; * we have in Buddhist literature of the so-called Gatha type abundant evidence of the results produced by efforts to Sanskritize, and the arguments which are adduced to establish the reality of translation would suffice to prove that Vedic texts were likewise translations. Moreover, there is conclusive evidence that Panini 6 knew a Mahabharata or at least a Bharatan epic in Sanskrit, and that the bulk of the Ramayana* was composed

1 Keith, Sanskrit Drama^pp. 72 ff., 85 ff, ia,i f., 140, 155.

2 Cullavagga, v. 33. 1 ; Keith, IHQ. i. 501.

8 Grierson, IA. xxiii. 52 ; Barth, RHR. xxvii. 288.

• Jacobi, Ramayana, p. 117 ; ZDMG. xlviii. 407 ff. ; Keith, JRAS. 1 906, pp. 2 ff.

Hopkins, Great Epic, p. 385. « Keith, JRAS. 1915, pp. 318 ff. CHARACTER AND EXTENT OF THE USE OF SANSKRIT 13 long before Acoka. Now, though the Brahmins made the epics largely their own, they were not the earliest composers of this form of literature, and the fact is attested in the simpler, more careless, language which shows indifference to many of the refine- ments of Brahmanical speech. Panini ignores these deviations from his norm ; it was no part of his aim to deal with the speech current outside the hieratic circle, and in the epic speech we have doubtless the form of language used by the Ksatriyas and the better educated of the Vaicyas during the period when the poems took shape. Both the Mahabharata and the Ramayana are, it must be remembered, essentially aristocratic ; they corre- spond to the Iliad and the Odyssey, and like them became the objects of the deep interest of wider circles. In recent times, no doubt, the epics have been unintelligible to the audience, to whom interpretation has been requisite, though delight is still felt in the sound of the sacred language. But this doubtless was not the -case in older times ; we must postulate a long period when the epic was fairly easily intelligible to large sections of the people.

Doubtless, as time went on, the gulf between Sanskrit and the languages of the day became more and more marked; even between the epic language and that of the Brahmin schools there were differences to which express reference is made in the Ramayana?- and both the practice of the dramas and such passages 'as that in Kalidasa's Kumarasambhava? in which Sarasvatl addresses Qiva and his bride, the one in Sanskrit, the other in Prakrit, attest dialectic differences based on rank, sex, and locality. In a sense doubtless Sanskrit came more and more to resemble Latin in the Middle Ages, but, like Latin, its vitality as the learned speech of the educated classes was unimpaired, and it won victories even in fields which were at first hostile to it. 3 The medical textbook current under the name of Caraka tells us that Sanskrit was used in discussions in the medical schools of the day. A work of very different character, the Katnasutra of Vatsyayana, bids its man of fashion in his con-

1 v. 30. 1 7 f. ; iv. 3. 28 f. ; li. 91. 22 ; vii. 36. 44 ; Jacobi, Ramayana, p. 1 15. Cf. Hopkins, Great Epic, p. 364.

2 vii. 87.

3 Cf. Jacobi, Scuntia, xiv. 251 ff. ; Oldenberg, Das Mahabharata, pp. 129 ff. i 4 SANSKRIT, PRAKRIT, AND APABHRANQA

versation in polite society use both Sanskrit and the vernacular of his country {degabhdsa). Hiuen Tsang tells us in the seventh century that Buddhist disputants used officially Sanskrit in their debates ; in his Upamitibhavaprapaiicakatha the Jain Siddharsi (a. d. 906) gives as his reason for preferring Sanskrit for this allegory of human life that persons of culture despise any other form of speech, and claims that his Sanskrit is so simple as to be understood even by those who preferred Prakrit. The writing of Sanskrit poems which even women and children — of course of the higher classes — can understand is contemplated by Bhamaha in his treatise on poetics (c. A.D. 700). Bilhana (a. D. 1060) would have us believe that the women even of his homeland, Kashmir, were able to appreciate Sanskrit and Prakrit as well as their mother tongue (janmabhasa). The famous collection of tales known as the Pancatantra owes its origin in theory in part, according to one later version, to the importance of instructing princes in Sanskrit as well as in the conduct of affairs.

There were, of course, spheres in which Sanskrit was at first rejected, beyond all in the early literatures of Jainism and Buddhism, which were probably couched in an old form of what became known as Ardhamagadhl Piakrit. As has been shown, 1 however, the question was early raised, if we may trust the Buddhist tradition, whether Sanskrit should not serve as the medium to preserve the Master's instruction, a notice which bears emphatic testimony to the predominance of Sanskrit as a literary medium. In both cases, however, Sanskrit finally won its way, and first Buddhists, then Jains, rendered great services both to Sanskrit literature and grammar.

The Buddhist revolt against Sanskrit had, however, one important result. The edicts of Acoka, in which he impressed on his subjects throughout his vast realm the duty of practising virtue, were inevitably couched in Prakrit, not Sanskrit, and the epigraphic tradition thus established died hard. But it had to contend with facts ; inscriptions were intended to be intelligible, and in the long run it proved that Sanskrit was the speech which had the best chance of appealing to those who could read inscriptions. In the second century B. c. traces of the influence

1 Keith, IHQ. i. 501 f. CHARACTER AND EXTENT OF THE USE OF SANSKRIT 15

of Sanskrit are apparent ; in the next century on one view 2 is found the first inscription which on the whole may be called Sanskrit, and Sanskritisms are on the increased In the first century A. D. Prakiit still prevails, but, though it is prominent also in the next century, we find the great Sanskrit inscription of Rudradaman which displays clearly the existence of an elaborate Sanskrit literature. In the next century Sanskrit and Prakrit contend, in the fourth Prakrit becomes rare with the Brahmanical revival under the Gupta dynasty, and from the fifth it almost disappears in Northern India. A parallel process was going on in literature ; in such Buddhist works as the Lalitavistara and the Mahavastu we find the results of an effort to convert a Prakrit into Sanskrit, and similar results are to be found in other fields, as in the medical tieatises of the Bower manuscript. From this the Buddhists soon advanced to the stage in which Sanskrit proper was used, as in the Divyavadana, perhaps of the second century A. D. 3 The Jains showed more conservatism, but even they ultimately accepted the use of Sanskrit as legitimate. Serious competition with Sanskrit as the language of literature again arose when the Mahomedan conquests brought Persian into play, and when the vernaculars in the period shortly after A. D. 1000 began first to influence Sanskrit and then to develop into literary languages.

The true home of the Qistas is given by Patanjali as Aryavarta, but even in his time the Dekhan was a home of Sanskrit ; Katyayana himself seems to have lived there in the third century E. C. Yaska 4 (c. 500 B. C.) already mentions a southern use of the Vedic word vijamatr, and Patanjali records the love in the south for derivative formations and the use of sarasi, large pond. Even in Southern India, despite the existence of a vigorous Kanarese and Tamil literature, Sanskrit inscriptions appear from

1 On sacrificial post at Isapur, 34th year of Vasiska, 33 B. c. ace. Fleet, JRAS. 1910, pp. 1315 ff. ; Hoernle, Bower MS., p 65 ; Ann. Rep. A. S., India, 1910-11, pp. 39 ff. It is much more probably of the second century A. D. (' A. D. 102) ; an inscr. of Huviska shows almost correct Sanskrit; JRAS. 1934, pp. 400 ff.

3 Franke, Pah und Sanskrit, pp. 13, 58; Rapson, JRAS. 1904, p. 449.

3 Przyluski {La ligtnde de Vimftreur Afoka, pp. 14 ff.) ascribes much to the influence of Mathura and its Sarvastivadin school, and places its use of Sanskrit in the Acokavadana at least in the second century B. C. (cf. pp. 166 ff.).

• vi. 9. Cf. Buhler, WZKM. i. 3. For Aryavarta, see I A. xxxiv. 179 (Madhyadeca)

and Kavyamimdhsa, p. xxiv. 16 SANSKRIT, PRAKRIT, AND APABHRAN^A

the sixth century onwards, often mixed with Dravidian phrases, attesting the tendency of Sanskrit to become a Koine, 'and Sanskrit left a deep impression even on the virile Dravidian languages. Ceylon fell under its influence, and Sinhalese shows marked traces of its operation on it. It reached the Sunda Islands, Borneo, the Philippines, and in Java produced a remark- able development in the shape of the Kavi speech and literature. Adventurers of high rank founded kingdoms in Further India, where Indian names are already recorded by the geographer Ttolemy in the second century A. D. The Sanskrit inscriptions of Campa begin perhaps in that century, those of Cambodia betore a. D. 6oo, and they bear testimony to the energetic study of Sanskrit grammar and literature. Of greater importance still was the passage of Sanskrit texts to Central Asia and their influence on China, Tibet, and Japan.

It is characteristic of the status of Sanskrit as the speech of men of education thatjn one sphere of use it only slowly came to be widely employed. Coins were meant for humble practical uses, and even Western Ksatrapas, like Rudradaman, who used Sanskrit for their inscriptions, were contented with Prakrit for coin legends ; but even in this sphere Sanskrit gradually prevailed. 1

The results which we have attained are in accord with the evidence afforded by Greek renderings of Indian terms. 2 These are neither wholly based on Sanskrit forms nor on Prakrit. Derived doubtless from the speech now of the upper, now of the lower classes, they remind us of the salient fact that at any given moment in India there were in active use several forms of speech varying according to the class of society. The denial of the vernacular character of Sanskrit 3 rests largely on a failure to realize the true point at issue, on a confusion between the earlier period when Sanskrit was far more close to the speech of the lower classes and later times, or on the fallacious view that the only speech which deserves the style of a vernacular must be

1 Bloch, Mllanges Lhii, p, 16.

2 L£vi, BSL. viii, pp. vui, *, xvii j Franke, ZDMG. xlvii. 596 ft. ; Bloch, Mllangts Lhii, pp. 1 ff.

' Grierson, JRAS. 1904, p. 481. On this view standard English would not be a vernacular. CHARACTER AND EXTENT OF THE USE OF SANSKRIT 17

the language of the lower classes of the population. Still less plausible is the suggestion 1 that Sanskrit as a vernacular was preserved in Kashmir during its eclipse in India generally, a view which has no support either in tradition or in the form of the Kashmirian vernacular. What we do find is that the Buddhism which penetrated Kashmir was strongly influenced by Mathura, where the new faith had fallen into the hands of men trained in the Brahmanical schools, who applied their own language to the propagation of the faith. We have in this one more proof of the hold which Sanskrit had in Brahmanical circles, and of the obvious fact that it was far better fitted as a language of theology and philosophy than Ardhamagadhl or any similar dialect.

3. The Characteristics and Development of Sanskrit in Literature

It is a characteristic feature of Sanskrit, intimately connected with its true vitality, that, unlike Medieval Latin, it undergoes important changes in the course of its prolonged literary existence, which even to-day is far from ended. Moreover, we must note the existence of two streams of movement, the Sanskrit of the Brahmanical schools as summed up in the grammar of Panini, and the less formal language of the ruling class and the Brahmins in their entourage as shown in the epics. The works of Classical Sanskrit literature show the clearest evidence of influence in both directions; the Brahmins, to whom or to whose influence and tradition we owe most of the literature, were schooled in grammar and were anxious to avoid solecisms, but they were also under the literary influence of the epics, and in special of the Ramayana, and it was not possible for them to avoid assimilating their language in great measure to that of their model.

Hence it follows that much of what is taught by Panini and his followers has no representation in the literature. As we have seen, Katyayana and Patanjali recognize the disuse of certain verbal formS ; there disappear also many idioms, 2 such as anvaje- or upaje-kr, strengthen, nivacane-kr, be silent, matto- or kane-

1 Franke, Pali und Sanskrit, pp. 87 ff.

• Bhandarkar, JBRAS. xvi. 273 ; Speijer, Sansk. Synl., pp. 39, 45, 61 {., 65 t, 7a,

89 f., 108.

SH» C 18 SANSKRIT, PRAKRIT, AND APABHRANQA

han, fulfil one's longing, celaknopam vrstah, 'rained until the clothes were wet'; many words are no longer used, such as anvavasarga, allowing one his own way, niravasita, excom- municated, abhividhi, including, utsaiijana, throwing w^,abkresa, equitableness. The pronominal base tya disappears ; in the verb the infinitive tavai is lost, many formations such as jajanti dis- appear, and the perfect participle middle in ana is disused. The adverbial form in tra, as in devatra, and the old VQrd parut are lost. Many nominal derivatives are not exemplified, and the^use of such phrases as quklisyat disappears. Many syntactical rules are obsolete, such as the use of the accusative with adjectives in uka ; the instrumental with samjnd or samprayam ; the dative with gldgh and stha ; trnam man or fune or gvdnam man ; the ablative with words denoting far or near ; the genitive with verbs of remembering other than smr, with naih, hope, with jas and other verbs denoting injury, and impersonally with expressions of illness, caurasya rnjati; the instrumental with prasita and titsuka ; uta in simple interrogations, and many other usages.

It is, however, true that beside this feature we have the deliberate employment by poets of usages, prescribed in the grammar, but so rare as to reveal themselves as purely learned reminiscences. From Acvaghosa on, the great authors are fond of displaying their erudition; Kalidasa has anngiram, 'on the mountain', though this is given by Panini 1 merely as an optional form, and sausndtaka, ' asking if one has bathed well ', from a Varttika. 2 Magha is adept in these niceties; he has khalu with the gerund to denote prohibition ; ma jivan, ' let him not live ' ; he distinguishes vi-svan, eat noisily, and vi-svan, howl ; he affects the passive use of the perfect, revives aorist forms and gerunds in am, including vastraknopam, and uses klam as a finite verb. Qriharsa, author of the Naisadhiya, is responsible for the solitary example of the first person periphrastic future middle, dar$ayitdhe, yet cited. 3 The case is still more extreme with Bhatti, whose epic is at once a poem and an illustration of the rules of grammar and rhetoric, and who has imitators in Bhau- maka.' s RdvaHdrjunlya and Ha) ay udha' s Kavira/tasya (10th cent.). Even in writers of the folk-tale knowledge of grammar sometimes

1 v. 4. 112 (Senaka). 2 iv. 4. 1, v. 3.

3 Cf. grammatical similes ; Walter, Indica, iii 38. CHARACTERISTICS AND DEVELOPMENT IN LITERATURE 19

is exhibited quite unexpectedly in the shape of recondite forms culled from Panini or his successors. So serious a philosopher as Cankara resorts to the use of the negative with finite verbs — which originally must have been merely a comic use — and he is guilty also of the employment of the comparative of a verb, upapadye-tarant, a linguistic monstrosity of the worst kind.

The influence of the grammarians explains also the free use of the aorist in the writers of elaborate prose^ Bana and Dandin, moreover, observe the precise rule for the use of the perfect in narration prescribed by the grammarians. It has been suggested that this may be explained by the derivation of prose from a different tradition than poetry, but the suggestion appears needless. 1 Subandhu ignores the rule as to the perfect, and the simple explanation of the accuracy of the other writers is the desire to display their skill in grammar, which was naturalfy facilitated by the absence of metrical restrictions. The same liberty explains their practice in postponing the verb to the end of the sentence, unquestionably, its traditional resting-place, but one impossible to observe in verse.

Very different was the effect on Classical poetry of the influence of the epics. 2 They show, with special frequency in the case of the Mahabharata? the tendency of uncultivated speech to ignore fine distinctions and by analogical formations to simplify grammar. Thus rules of euphonic combination are not rarely ignored ; in the noun the distmction of weak and strong case-forms is here and there forgotten ; there is confusion of stems in i and in ; by analogy pusanam replaces the older pusanam ; there is confusion in the use of cases, especially in the pronoun ; in the verb primary and secondary endings are some- times confused ; active and middle are often employed for metrical reasons in place of each other ; even the passive is found with active terminations ; the delicate rules affecting the use of the intermediate i are violated at every turn ; the feminine of the present participle active is formed indifferently by antl or atl ; the

1 Speijer, Sansk. Synt., % 328 ff. ; Renou, La valeur du par/ail, pp. 86 ff.

2 For the Ramayana cf. Bohthngfc, BSGW. 1887, pp. 213 ff. ; ZDMG. xliii 53 ff.; Roussel, Mus&on, 1911, pp. 89 ff. ; I9i2,pp. 25 ff. , 201 ff. ; JA. 1910, i. 1-69; Keith, JRAS. 1910, pp. 468 ff., 1321 ff.

3 Holtzmann, Gramm.aus d. M. (1884).

C 3 20 SANSKRIT, PRAKRIT, AND APABHRANgA

middle participle of causatives and denominatives is often formed by ana, partly doubtless on grounds of metrical convenience; the rule that the gerund is formed by tva in simple, in ya in compound, verbs is constantly disregarded ; minutiae such as the substitution of dhdvati for the present of sr are habitually neglected. The tendency to prefer a bases is seen in the verb and the noun alike, giving such forms as difd and duhitd. ' It was inevitable that so distinguished models as the Maha- bhdrata and the Ramdyana should deeply affect later poets, and Patafijali, in citing an epic fragment containing the irregular term priydkhya in lieu of priydkhydya, expressly asserts that poets commit such irregularities (chandovat kavayah kurvanti). We find, therefore, occasional errors such as the confusion of antla.nA atl, of tvd and ya, of active and middle, as well as regular dis- regard of the specific sense of the past tenses as laid down by the grammarians but ignored in the epic. As in the epic, the perfect and imperfect freely interchange as tenses of simple narration without nuance of any kind. Even Kalidasa permits himself sarati and dsa for babhuva, and Cnharsa with the Ramdyana uses kavdta for the kapdta of Panini. Lesser poets, especially the poetasters who turned out inscriptions, are naturally greater sinners by far against grammatical rules, especially when they can plead metrical difficulties as excuse.

Neither the epic nor the grammarians, however, are responsible for the fundamental change which gradually besets the Kavya style, in the worst form in prose, but in varying degree even in verse. This is the change from the verbal to the nominal style, as Bhandarkar l not inaptly termed it. In the main, Vedic and epic Sanskrit show a form of speech closely akin to Greek and Latin; verbal forms are freely used, and relative clauses and clauses introduced by conjunctions are in regular employment. The essential feature of the new style is the substitution of the use of compounds for the older forms. 2 In its simplest form, of course, the practice is unobjectionable and tends to conciseness ; hataputra

1 JT3RAS. xvi. 366 ff. ; cf. Bloch, MSL. xiv. 27 IT. ; Renon, La valeur du parfait, pp. 90 ff. ; Stchoupak, MSL. xii. 1 ff. ; Jacobi, IF. xiv. 2360.

1 Jacobi {Compositum tind Ntbtnsatz, pp. 25, 91 ff.) points out that they are properly used for ornamental description, not for important qualifications, and also suggeits poetic convenience as a cause of popularity ; cf. Chap. II, § 4. See also Wackernagel, Altind. Gramm. t II. i. 35, 27, 159; Whitney, Sansk. Gramni., § 1346. CHARACTERISTICS AND DEVELOPMENT IN LITERATURE 21

is less cumbrous than ' whose sons have been slain '. But when new members are added there are soon lost the advantages of an inflective language with its due syntactical union of formed words into sentences ; brevity is attained at a fatal cost in clearness. A compound like jaldntagcandracapala, ' fickle as the moon reflected in the water ', is comparatively innocuous, but even a stylist like Kalidasa permits himself such a phrase as viciksobha- stanitavihagagrenikdhclgund, ' whose girdle-string is a row of birds loquacious through the agitation of the waves '. True, in such a case there is no real doubt as to the sense, but often this is not the case, and in point of fact it is one of the delights of the later poets to compose compounds which contain a double entendre, since they can be read in two ways ; of such monstro- sities Subandhu is a master. Moreover, the nominal forms of the verb are given a marked preference ; the expression of past time is regularly carried out by a past participle passive in form of an intransitive verb, such as gatas, he went, or if the verb is active the subject is put into the instrumental and the past participle passive is employed, as in mrgenoktam, the deer said. Or an active past participle is created by adding vant to the passive participle, fyrtavan, he did ; a distant parallel in the grammarians has been seen in the sanction by Panini of the use of such forms as ddgvdhs in lieu of a finite verb. Or the use of any save a verb of colourless kind may be avoided by substitut- ing such an expression as pakvant karoti for pacati, he cooks, or pakvo bhavati, it is cooked, for pacyate. Similarly the peri- phrastic future is preferred to the finite verb. Or the verb may wholly disappear as when for ayam indnsam bhaksayati we have mdhsabhojako 'yam, he is a meat eater. In harmony with this is the tendency to lay great stress on case relations as expressing meaning, a practice which in the later style in philosophy, exegesis, and dialectics results in the occurrence of sentences passim with no verb and practically only the nominative and ablative cases of abstract nouns. Frequent, and indeed in some forms of composition, such as the folk tale, tedious in its reiteration, is the use of gerunds in lieu of subordi- nate clauses.

We are reduced to conjecture as to the cause of this tendency. The desire for brevity is already seen in the style of the Vedic 23 SANSKRIT, PRAKRIT, AND APABHRANCA

Sutras, and the grammarians carried it to excess ; their works furnish abundant instances of insistence on using cases in a preg- nant sense and in affecting compounds ; gerunds are frequent in the ritual texts. It has been suggested that the love for partici- pial forms is partly explained by Dravidian influence ; l the periphrastic future in both Sanskrit and Dravidian uses the auxiliary verb only in the first and second persons ; the type krtavan has a parallel in geydavan ; the rule of the order of words in which the governed word precedes and the verb is placed at the end of the sentence is Dravidian. Unhappily, the arguments are inconclusive ; 2 the omission of the auxiliary in the third person is natural, for in that person in any sentence whatever it is commonly omitted as easily understood ; the order of words in Sanskrit has parallels in many other languages than Dravidian and rests on genera] rules of thought.

Beside the correct or comparatively correct Sanskrit of the poetic literature we find, especially in technical and non-Brah- manical works, abundant evidence of a popular Sanskrit or mixed Sanskrit in various forms. Generically it can be regarded as the result of men who were not wont to use Sanskrit trying to write in that language, but there are different aspects. Thus the early Buddhist writers who decided to adapt to the more learned language the Buddhist traditions probably current in Ardhama- gadhi were hampered by the desire not to depart unduly in verse at least from their models, a fact which explains the peculiar forms found especially in Gathas, but also in prose in such a text as the Mahavastu? Traces of this influence persist even in much more polished Buddhist writers such as Acvaghosa, and much of it may be seen in the Divyavadatia, though that work

1 Konow, LSI. iv. 2 79rT. ; Grieison, BSOS. I. iu. Ji; Carnoy, JAOS. xxxix. 1 1 7 ff . ; Chatlerji, i. I74ff.

2 Cf. R. Swaminatha Aiyar, POCP. 1919, i, pp. Ixxi ff., who legitimately points out that the evidence of Dravidian is very late in date, and these languages probably bor- rowed from Aryan. K. G. J>ankar (JRAS. 1924, pp. 664 ff.) points out that the Tot-kappiyam, the oldest Tamil work, must be after 400 a. D. as it refers to the Poruladhikaramsutra, horary astrology, and that the Moriyas of the Sangam are the Mauryas of the Konkana, who date after 494 A. d,

8 Cf. Senart, i, pp. iv, xrii ff. ; Wackernagel, Altind. Gramm., i, p. xxxix. Contrast F. W. Thomas, JRAS. 1904, p. 469, who regards the mixed Sanskrit as representing middle-class speech. Poussin (Indo-turoplcns t p. 305) stresses convention as stereo- typing usage. CHARACTERISTICS AND DEVELOPMENT IN LITERATURE 23

marks in part a successful attempt to adapt Sanskrit prose, as known at Mathura and elsewhere, to Buddhist use. The degree of cultivation of those who endeavoured to write in Sanskrit might vary greatly ; thus the Sanskritization of the treatises in the Bower Manuscript, perhaps of the fourth century A.D., is comparatively good in the case of those on medicine, and de- cidedly poor in those on divination and incantation. In part the deviation from Sanskrit as laid down in the grammars is purely a case of Prakritic forms intruding scarcely disguised into the texts, but in other instances popular influence reveals itself in a Sanskrit which ignores delicate distinctions and confuses forms. The distinction between Prakritisms and careless Sanskrit is not absolute, but it is convenient and legitimate.

Thus we have in the phonology of this popular Sanskrit as seen in the Bower MS. some confusion of r and ri, of n and «, of ( , s, and s ; metrical lengthening and shortening of vowels is not rare ; ml becomes mbl, and rarely a is prefixed as in alatd. In Sandhi hiatus and hyper-Sandhi, even to the extent of an elided consonant {afvibhyanumatah), are known, while a is occasionally elided when initial. In declension we find is and reversely u as feminine nominatives for i and us ; is is often replaced >y yas as the accusative feminine, and in stems are treated as /stems, as in pittinam for pittinam. In the verb we have simplification in class, as in lihet for lihyat,piset for pinsydt; and, as in the epic, very free interchange of active and middle forms ; the gerunds in tva and yd are confused. Stem formation shows frequently the mixture of bases in a, i, or u for those in as, is, or us, and, rarely, such a base as hantdra from the accusative of hantr ; there is con- fusion in feminine suffixes, as in ghna for ghni, caturthd for catur- thi, while ordinals in composition are sometimes replaced by cardinals. Very characteristic is confusion of gender, especially between masculine and neuter, more rarely between masculine and feminine or feminine and neuter. Case confusion is common, as is non-observation of rules of concord and confusion of numbers, white the interpolation of particles within compounds or sentences, absolute constructions, and very loosely compacted clauses are common.

Existing as it did side by side with Prakrit dialects, it was inevitable that there should be frequent borrowings on either 24 SANSKRIT, PRAKRIT, AND APABHRAN^A

side, 1 despite the objections raised from time to time by gram- marians and sticklers for purity in. the use of the sacred language in sacrificial matters. 2 Thus, though Classical Sanskrit lost many of the words and roots recorded in the Ganapatha and the Dhatupatha associated with Panini's grammar, it was enriched by numerous additions, some easy, others difficult, of detection. In many cases the Prakrit forms were taken over with only the neces- sary changes requisite to make them seem to have terminations allowed in Sanskrit. It appears as if even Panini a recognized this practice, since he allows eastern place-names to pass as correct though having the Prakrit e and o for the regular ai and au which his rules require. In other cases the retention of the Prakrit form was aided by the possibility of regarding the form as genuine Sanskrit ; thus the poetic technical term vicchitti^ really from viksiptif in all likelihood seemed to be derivable from w- chid ; Krsna's epithet Govinda, perhaps Prakrit fox gopendra, was felt asgo-vinda, winner of cows ; in late texts bhadanta, from the phrase of greeting bhadram te, is defended as from bhad with the suffix anta, and uttr is not recognized as from avatr through Prakrit otarati ; duruttara, hard to Overcome, really from Prakrit duttara for dustara, was felt as dur-uttara. In many cases, doubt- less, Prakrit words were correctly rendered into good Sanskrit equivalents, in which case borrowing cannot now be established. In others, however, the process is betrayed by false forms ; thus Prakrit marisa, friend, where s stands for g, was mechanically made into marisa ; guccha, for the lost grpsa, became gutsa, cluster ; masina, Sanskrit mrtsna, reappeared as masrna, soft ; rukkka, 5 for ruksa or rather vrksa, ruksa, tree ; and hettha, from adhastat, gave by reconstruction fiesta. A common formation in Jain texts is vidhyai, go out, which is based on Prakrit vijjhai, from Sanskrit viksai; similarly vikurv, produce by magic, is traced through viuvvai, viuvvae to vikr. Later there are borrowings from vernaculars such as GujaratI or MarathI or

1 Zachariae, Beitr. z. Lexikogr., pp. 53 ff.

2 See Qabarasvamin and .Rumania on Mimahsd Sutra, i. 3, 24 ff. ; Sarasvati- ianthabharana, i. 16 ; Mahabhasya, i. 5. * i. I. 75.

4 Zachariae, B. Beitr., xiii. 93 ; cf. argala (IA. xix. 59) through aggala for agralaka ; Kielhorn, GN. 1903, p. 308.

• See Hultzsch, CII. i, pp. lxxff., contra Turner, JRAS. 1925, p. 177. I agree with

Oldenberg that in RV. vi. 3. 7 ruksa is not = vrksa. CHARACTERISTICS AND DEVELOPMENT IN LITERATURE 25

Hindi. 1 Often, of course, the Sanskrit version has been ingeni- ously made to appear valid In itself, as when pabbhdra is meta- morphosed into prdgbhdra, though prahvara is its origin.

Occasionally we find the process of Sanskritization applied to what was really Sanskrit ; probably thus are to be explained prasabham, violently, from pra-sah ; Naghusa for the older proper name Nahusaj varsabhu, frog, for varsdhu.

From foreign sources borrowings also occurred naturally enough in those cases where, as in the Dekhan or Further India, Sanskrit was used side by side with a native speech. Kumarila permits the incorporation of Dravidian terms, provided that they are given SanskrifHerminations, and names especially such as Sayana were freely thus Sanskritized. The / which marks South Indian texts 2 in lieu of the d and / of the north is doubtless in part due to Dravidian influence. On the other hand, invasions from the north brought early and late Iranian words such as lipi, writing, Old Persian dipt, 3 ksatrapa, satrap, and perhaps mudrd, seal,* or divira, scribe, mihira, Mithra, bahadura, sdka, and sdhi. The Greek invasions in the north left little trace in the language, but probably later India borrowed surungd from syrinx in the technical sense of an underground passage, and a large number of terms of astrology. Many of these they ingeniously altered to seem true Sanskrit, as when for hydrochoos we find hrdroga, or jdmitra for diametron. With similar ingenuity the useful camel was metamorphosed into kramela? suggesting connexion with kram, go. The Mahomedan invasion brought with it Arabic and Turkish terms, and the European powers have contributed occasional additions to the modern Sanskrit vocabulary, testify- ing to its capacity of assimilation. The scientific literature in special has shown its willingness to appropriate the terms used by those from whom knowledge has been acquired, together with considerable skill in disguising the loan.

1 Cf. Bloomfield, Festschrift Wackemagel, pp. 320-30 ; Hertel, HOS. xii. 29 f.

' Luders, Festschrift Wackernagel, p. 295.

3 Buhler, Ind. Stud., lii. 21 ff.; Hultzsch, CIL i, p. xlii.

• Franke, ZDMG. xlvi. 731 ff. Hala has vandi, captive. Cf. Weber, Monatsber.

Bert. Ak., 1879, pp. 810 ff.

5 L4vi {De Graecis vet. Ind. Men., p. 56) doubts this, but the word is late; lopaka (awmj() is different, as lopafa is Vedic. Hala has kalama (/niXa/ios) and tnaragaa (aptipan/Sos). 26 SANSKRIT, PRAKRIT, AND APABHRANQA

As the passage of time made Sanskrit more and more a language of culture, it reveals in increasing measure a lack of delicate sensi- bility to idiomatic use of words, such as is engendered by usage in a living speech more closely in touch with ordinary life. The defect, however, is sometimes exaggerated, for it must not be forgotten that poets of all times are apt, through considerations of metre or desire for effect, 1 to adopt unusual senses of words and to strain meanings ; Pindar and Propertius illustrate a tendency which is found more or less markedly throughout classical litera- ture, while the Alexandrian Lykophron is guilty of as distinct linguistic monstrosities as any Indian poet. The tendency in their case was accentuated by the growing love for paronomasias, and the tendency to study poetic dictionaries which gave lists of synonyms, ignoring the fact that in reality two terms are practi- cally never really coextensive in sense. The grammatical know- ledge of the poets also led them into inventing terms or using terms in senses etymologically unexceptionable but not sanc- tioned by usage.

4. The Prakrits

The most widely accepted etymology of Prakrit current in India treats the name as denoting derivative, the prime source {prakrti) being Sanskrit. Another view reverses the position ; Prakrit is what comes at once from nature, what all people without special instruction can easily understand and use. 2 It is impossible to decide what was the process which led to the use of the term ; perhaps speeches other than Sanskrit received the name from being the common or vulgar speech, the language of the humble man as opposed to him of education who could talk the pure language. In the grammarians and writers on poetics the term more especially denotes a number of distinctly artificial literary dialects, which as they stand were certainly not vernaculars ; but it is customary to use the term to apply to Indian vernaculars prior to the period when the modern vernaculars became fixed. An even wider sense is given by Sir George Grierson, who classifies Prakrits in three great stages :

1 Catnllus' curious compounds in the A His illustrate this theme. 1 Pischel, Grammatik der Prakrit -Sprachcn (1900), §§ 1, 16. पृष्ठम्:Sanskrit Literature.djvu/६१ 28 SANSKRIT, PRAKRIT, AND APABHRAN?A

agree against the east in assimilating ty to cc and ks to cck, against the representation of ty as tiy and the assimilation to kkh ; the east again is marked by the use of,£ for primitive az as against o, and by its rejection of r in favour of /. This eastern dialect may fairly be regarded as a forerunner of the Ardha- magadhI of the grammatical tradition, though that language has been largely affected by western influences in its later form. An inscription in a cave on the Ramgarh hill, probably of the second century B. c, reveals to us the precursor of the later Magadhl, since it shows its characteristics, e for o, I for r, kkh for ks, and f for s.

Our next information of a definite character regarding the dialects is afforded not so much by the various inscriptions of the post-Acokan period as by the dramas of Acvaghosa, which may be regarded as good testimony for the period c. A. D. 100. Here we find dialects which may justly be styled Old ArdhamagadhI, Old CaurasenI, and Old Magadhl; of these the former may well have been the dialect in which, as tradition asserts, Mahavlra preached his doctrines and established Jainism, and in which Buddhist teachers carried on their work. 1 The early Jain scriptures, however, have admittedly perished, and the actual canon of the Cvetambaras now extant is redacted in a form strongly influenced by the later south-western speech Maharastri, while later texts are written in what has been fairly called Jain Maharastri, and the Digambaras adopted under western influence what has been styled Jain (pauraseni. The canonical language of Buddhism, on the other hand, is more ancient ; it is not, however, ArdhamagadhI, but is distinctly of a western type, perhaps more closely connected with AvantI or Kaucambl than any other region. To the group of old Prakrits belongs also the mysterious PaicacI, in which the famous Brhatkatha of Gunadhya was written ; its home is still uncertain ; it has been connected by Sir G. Grierson 2 with the north-western dialect of the Acokan inscriptions on the one side and the modern languages of the north-west, which with dubious accuracy he has styled Picaca ; against this may be set, inter alia, the fact that the north-western

1 Cf. Keith, IHQ. i. 501 ff.

• Pisaca Lang., pp. 1 ff. ; ZDMG. Ixvi. 49 ff. ; JRAS. 1921, pp. 424 ff. ; IA. xlix.

U4; AMJV. i. H9ff. THE PRAKRITS 29

dialect of Acokan times kept the three sibilants which Paicaci reduces to one, although the Gipsy dialect and the dialects of the Hindu Kush distinguish still between s and s on the one hand and g on the other. 1 The possession by Paicaci of the letters / and /, and the use of one nasal n only, have been adduced by Konow 2 as proof of location in addition to its close con- nexion with Pali, and, as these features were preserved in modern Malvi, and its hardening of soft consonants is probably due to Dravidian influence, Paicaci has been located in accord with Indian tradition in the Vindhya region. Inscriptions suggest also that south of the Narmada there was a measure of indepen- dent development, adding a south-western to the three great groups already known ; thus in the south we have duhutuya, dhua in the later Maharastrl, pointing to the source of Ardhama- gadhi dhiiyd, as opposed to the dhita of the northern inscriptions, Pali dhita, £aurasenl (beside duhida) and Magadhi dhidd, Vedic dhita. beside the normal duhita?

The characteristics of these Old Prakrits are simple.* They include the loss of the vowels r and /, and of the diphthongs at and au ; reduction in the number of sibilants and nasals ; and the assimilation of consonants. They show also the operation of the substitution of the expiratory for the musical accent, a feature which is obvious in Sanskrit during the same period. Fuither, they are subject to a most important law which reduces each syllable to the form either of a vowel, short or long, a short vowel followed by one or two consonants, or a long vowel followed by a single consonant ; the resulting changes of form are intensified by the confusion which results from substituting a long vowel with a single consonant for an originally short vowel with two consonants, or the use of a nasal vowel in lieu of

1 Reichelt, Festschrift Streitbirg, p. 245.

J ZDMG. lxiv. 95; JRAS. 1921, pp. 244 ff. ; cf. Ranganathaswami Aryavaraguru, IA. xlviii. 211 f. Przyluski {La ligende de tempereur Afoka, p. 72) holds that Pali may have had relations with Kaucambl.

s Luders, KZ. xlix. 233 f.

• ~ * LiiJers, Bruchstucke buddh. Dramtn, pp. 29 ff. ; Keith, Sanskrit Drama, pp. 72 ff.

85 ff, 121 ff. Contrast Michelson, AJP. xli. 265 ff ; Bloch, JA. 1911, ii. 167. In a Prakrit of the Western Panjab is composed the Dhammapada of the Dutreuil de Rhins MS. ; Konow, Festschrift Wmdisch, pp. 85 ff. (1st cent. A. D.) ; Luders, SBA 1914, pp. 101 ff. (3rd cent. A. D.) 30 SANSKRIT, PRAKRIT, AND APABHRANQA

a long vowel, or a short vowel and a consonant, when another consonant follows.

It is probable enough that literature of a secular character was composed in these Old Prakrits until the second century A. D., but about that date we have clear evidence of the fundamental changes which mark what may be called the Middle Prakrit of the grammarians and of most of the extant literature. This consists in the softening or disappearance of intervocalic con- sonants, carried to the furthest in Maharastri in the dominions of the Catavahanas of the south-west, but noteworthy also in the other Prakrits recognized by the grammarians, MagadhI, and Caurasenl. We see in the dramas of Bhasa, as compared with those of Acvaghosa on the one hand and of Kalidasa on the other, clear evidence of transition, the omission of intervocalic consonants, the softening of surds to sonants, the reduction of aspirates to A, the change of y into j, the substitution of n for n, the simplification of double consonants with compensatory lengthening. The evidence of inscriptions supports the view which assigns the loss of intervocalic consonants to the second century A. D., 1 in which century Maharastri lyric began its successful career, made known to us in the anthology of Hala. Once stereotyped by the grammarians at an uncertain date, the Prakrits rapidly lost in importance as they became more and more divorced from current speech, while they did not possess the traditional sanctity of Sanskrit or its clarity of structure and beauty of form.

Of the Prakrits Maharastri held pre-eminence by its use in drama, whence it was introduced perhaps by Kalidasa from lyric poetry, and by its adoption for epic poetry. Caurasenl was normally the prose Prakrit, though it appears to have been occasionally used in verse ; its employment in prose outside the drama was probably once much wider than was later the case when the Jains used a form of Maharastil for prose as well as for verse, though the presence of Caurasenl forms in prose suggests that Maharastri is here intrusive. 2 C aura senl wa s markedly more

1 Bloch, Milanges Llvi, pp. ] 2 fT. (kamdra, however, is from iarmdra). As regards lingualization cf. Turner, JRAS. 1924, pp. 555 ff., 58a ff. (danda, however, is not for dandra ; see Lklen, Stud. z. altind. und vergl. Sprachg., p 80).

3 Jacobi, Bhavisatta Kaha, pp. 88 ff. ; RSO. li. 231 ff. THE PRAKRITS 31

closely akin to Sanskrit than Maharastrl ; its place of origin was within the sphere of the strongest influence of Sanskrit, and it remained in specially close relation with it both in morphology, syntax, and vocabulary. Hence it was appropriately used-for persons of good position in the drama. MagadhI, on the other hand, was reserved for those of low rank, and, though tales 1 were composed in it, it was of comparatively minor importance. The Ndtyagdstra, perhaps in the third century A. D., enumerates other dramatic dialects (vibhasds) which are clearly of no real popular origin ; such are Daksinatya, Pracya, AvantI, and Dhakkl or Takkl, which are mere varieties of (paurasenl, while Candali and (pakarl are species of MagadhI. 2 PaicacI, though practically unknown in the extant dramas, enjoyed, it appears, a consider- able vogue in the popular tale, as a result, doubtless, of the fame of the Brhatkathd.

The comparatively late date at which Maharastrl appears to have come into fame, as indicated by its exclusion until late from the drama, suggests that some other Prakrit was employed for poetry before its rise into repute. Jacobi has found traces of such a Prakrit in the verses cited in the Ndtyagdstra ; 3 it was marked by the facultative retention or change or loss of inter- vocalic consonants, and was akin on the one hand to (paurasenl, for example in such forms as sadisa for sadrga and the gerund in iya, while it shared with Maharastrl the locative in amnti and the gerund in una; from these local indications he suggests that it had its centre in Ujjayim. It was, he holds, from this dialect that the softening of t to d passed into (paurasenl, which in Afvaghosa hardly shows any trace of it, and also in the dialect, otherwise similar to Jain Maharastrl, which on this account Pischel 4 named Jain (paurasenl. This poetic Prakrit, like (paurasenl, is essentially closely akin to Sanskrit.

1 Probably in verse, like Maharasfii and Apabhranca tales ; Dandin, i. 38; Rudrata, xvi. 36. Dandin's Gaud! Prakrit may be MagadhI ; he mentions also Latl.

a Cf. Keith, Sanskrit Drama, pp. 140 ff., 337 ; Gawronski, KZ. xliv. 247 ff. Iranian traits in Cakarl aie not proved (JRAS. 1925, pp. 237 ff.) ; the points adduced all are essentially MagadhI (cf. ibid., pp. 218 ff.).

• Bhavisatta K'aha, pp. 84 ff. He does not touch on its relation to Pali

' Op. cit., § 21. 32 SANSKRIT, PRAKRIT, AND APABHRANCA

5. Apabhranqa

Pischel 1 and Sir G. Grierson 2 have given currency to the view that the term Apabhranca denotes the true vernacu- lars as opposed to literary Prakrits, and the latter has con- structed a scheme for the derivation of modern vernaculars from the various local Apabhrancas; thus from Qaurasena (or Nagara) Apabhranca came Western Hindi, RajasthanI, and GujaratI ; from Maharastra Apabhranca MarathI; from Magadha Bengali, Biharl, Assamese,, and Oriya; from Ardha- magadha Eastern Hindi; from Vracada SindhI; and from Kaikeya Lahnda. Unfortunately this theoretical scheme will not stand investigation, for the evidence of texts and even of the literature proves clearly that Apabhranca has a different signification. 3

The essential fact regarding Apabhranca is that it is the collective term employed to denote literary languages not Sans- krit or Prakrit. Bhamaha 4 expressly gives this threefold division, and Dandin 6 expressly says that Apabhranca is the term applied to the idioms of the Abhlras, &c, when they appear in poetry. Guhasena of Valabhl, whose inscriptions have dates from A. D. 559-69, is declared to have composed poems in the three languages, Sanskrit, Prakrit, and Apabhranca. Rudrata, 6 in the ninth century, asserts that Apabhraiica is manifold through the difference of lands, doubtless in agreement with Dandin. Hema- candra also does not identify Apabhranca with the vernaculars. The vernacular {degabhasa) is a different thing; hetairai are required to be skilled in the eighteen vernaculars according to the Jain canon ; the Kdmasutra, in enumerating their sixty-four accomplishments, includes knowledge of vernaculars as well as of literary speeches {kavyakriya) ; moreover, it preserves the

1 Gramm, der Prakrit-Spracken, § 4.

8 BSOS. 1. iii. 61 ff. ; cf. IA. li. 13 ff.

• Jacobi, Bhavisatta Kaha, pp. 53 ff.; Sanatkutnaracaritam, pp. xviiiff. ; Fest-

schrift Wackernagel, pp. 1 34 ff.

< i. 16.

1 i. 31. Nobel's effort {Indian Poetry, pp. 13a, 159) to distinguish between Bharoaha's and Dandin's use of Apabhrai^a is a failure. APABHRANgA 33

interesting notice that a man of taste would mingle his vernacular with Sanskrit, as is .the way with modern vernaculars, not with Apabhranca. The identification of the vernaculars and Apabhranca is given as the opinion of some authorities by the commentator of the Prakrta Pihgala, and other late authorities adopt this view. But the oldest authority who has been cited 1 for it is the Kashmirian Ksemendra (nth cent.), and it is extremely doubtful whether he meant anything of the sort when he refers to poems in vernacular ; it is as likely as not that in Kashmir, as probably in the case of Maharastra, Apabhranca was never a literary language, vernacular poems supervening directly on Prakrit poetry.

The first actual remnants of Apabhranca preserved occur in a citation in Anandavardhana, in the Devlgataka, and in Rudrata. By preserving r and r it is clear that these verses belong to the species of Prakrit styled by the eastern school of grammarians (Kramadlcvara, Markandeya, Rama Tarkavagica) Vracata, which also is styled the speech of the Abhiras. This tribe appears to have entered India some time before 150 B.C., when it is mentioned by Patanjali. Its early home was Sindhudeca, by which is meant 2 not Sindh but the Peshawar district of the ^Rawalpindi division, where they had as eastern neighbours the Gurjaras. 3 Later both tribes spread ; the Gurjaras are found as Gujars in the United Provinces ; in the main, however, they went south and occupied Gujarat. The Abhiras are recorded in the Mahabharata as in the Panjab, later they are heard of in Kuruksetra, and their descendants, the Ahirs, range as far east as Bihar ; some went south and settled on the coast to the west of Gujarat ; they won considerable fame, and an Abhira dynasty is stated in the Visnu Purana to have succeeded the Andhrabhrtyas. Both Abhiras and Gurjaras were probably of the Dardic branch of the Indian race, to judge at least from the strong Dardic

1 Jacobi, Bhavisatta Kaha, p. 69, corrected p. 214.

2 Jacobi, Festschrift Wackernagel, p. 124, n. 2 ; cf. Faghuvahca, xv. 87, 89. See Mahabhasya, i. 3. 72, v. 6.

3 See references in EHL pp. 427 ff. ; R. C. Majumdar, The Gurjara- Pratlharas (1923). The view of them as Khazars or Huns is unproved, and their earliest date unknown, but Alexander did not find them in the Panjab. Cf, Grierson, IA. xliii. 141 ff., 159 ff.

SHS 34 SANSKRIT, PRAKRIT, AND APABHRANQA

element in Lahnda, the speech of the western Panjab. As they grew in civilization, they must have sought to create a literature ; whether they attempted it in their own dialect at first and later produced Apabhranca must remain uncertain ; what is clear is that Apabhranca originally was an effort to infuse into Prakrit a measure of their vernacular.

The effort to make Prakrit more readily intelligible to the people was not new ; in the earliest epic in Jain Maharastrl known to us, the Paumacariya x of Vimala Suri, probably not before A. D. 300, we find the free use of what the grammarians style Decicabdas, words for which no derivation from Sanskrit is obvious or normally possible ; similarly it seems that Padalipta's Tarangavati, mentioned in the Anuyogadvdra (5th cent.), though written in Prakrit, contained very many of such words. The large number of Dec! terms preserved in the Deginamamala of Hemacandra, some four thousand in all, testifies to the prevalence at one time of this practice, which, however, failed to retain favour. The reason for this may easily be conjectured ; the words takenjrom the vernaculars were a barrier to comprehension in a wide circle, and with the rapid change of the vernaculars became obscure even in the poet's own land, so that poets who desired permanence of repute and wide circles of readers pre- ferred to content themselves with those terms which had general currency. In Apabhranca, however, the effort was made to simplify Prakrit by adopting as the base of the grammar the vernacular, while using in the main the Prakrit vocabulary, and to some extent also Prakrit inflexions. There is a certain parallel with modern vernaculars which borrow freely from Sanskrit as opposed to Prakrit, but they do not use Sanskrit inflexions at all.

The Prakrit used as the base of early Apabhranca seems to have been often Maharastrl, but sometimes also Qaurasenl. But once Apabhranca had become popular, perhaps through the activity of the Abhlra and Gurjara princes, it spread beyond the west and various local Apabhrancas arose, as is recognized by Rudrata ; in these, we may assume, the special characteristics of the Vracata or Vrajada Apabhranca were refined. We find this

1 Jacobi, ERE, vii, 467. APABHRAN^A 35

confused condition reflected in the grammarians. Hemacandra, who belonged to the western school which goes back to the Vdlmiki Sutras, describes one kind of Apabhrarica, but alludes to others; in the eastern school we find a division as Viacata, Nagara, and Upanagara, in all of which r after consonants is kept while in the first r before consonc^ts also. Faint traces of the observance of this rule may be found in a few verses cited by Hemacandra ; the great poems, Bhavisattakaha and Neminaha- cariu assimilate r, and thus belong to a later type of Apabhrarica. In Bengal we find a type of Apabhrarica long in use in Buddhist texts, and a much degraded form, Avahattha, is evidenced in the Prdkrta Pihgala (14th cent.), but the basis even of this Apabhrarica is Maharastrl, not Magadhi, testifying to its ultimate western origin.

From the nature of Apabhrarica it follows naturally that in Old Gujarat! we find a considerable amount of resemblance in inflexion to Apabhrarica, as was to be expected from the fact that the vernacular is a descendant in considerable measure of that vernacular which was applied to Prakrit to form the early Apabhrarica. In other cases we could not expect to find any such important coincidences ; thus in Bengal the Apabhrarica used was not formed by applying vernacular inflexions to the local Prakrit ; at most some local colour was given to a speech which came from the west, and the same remark clearly applies in other cases. Sir G. Grietson's efforts 1 to establish a Maha- rastra Apabhrarica as a connecting link between Prakrit and MarathI are clearly unsuccessful. Nor indeed, it must be added) is there yet any adequate proof even of the relations suggested by him between the Prakrits and the vernaculars ; 2 thus traces of Magadhi in Bengali are extremely difficult to establish with any cogency. 3

There is no reason to suppose that Apabhrarica formed a necessary step towards composition in vernaculars, and in Maharastra and Kashmir Apabhrarica appears to have been

1 BSOS. f. iii. 63.

2 E. g. his view (JRAS. 1925, pp. 228 ff.) as to single consonants in the Noith-West Prakrit is clearly improbable.

3 M. Shahidullah, IHQ. i. 433 ff. Bloch {Formation de la langue marathe ; JA. '9 12 , i. 336) insists that the modern dialects presuppose a Prakrit koine.

D % 36 SANSKRIT, PRAKKIT, AND APABHRANQIA

unknown, while in the latter region vernacular poetry appears to have been practised in the eleventh century. Literary evidence of compositions in the vernaculars is fragmentary, but at least from the twelfth century there was a Hindi literature, from the thirteenth one in Marathi, and probably enough still earlier dates may be assigned to the adaptation of vernaculars to literary uses. 1

1 For Bengal see Dinesh Chandra Sen, Hist, of Bengal Lang, and Lit, (1911) and S. K. Chatterji, 1. 129 ff. PART II

BELLES-LETTRES AND POETICS II

THR ORIGIN AND DEVELOPMENT OF KAVYA LITERATURE

i . The Sources of the Kavya

INDIA produced no historian of her Sanskrit literature, and, naturally enough, the appearance of great poets of the calibre of Kalidasa, Bharavi, and Magha so eclipsed earlier efforts that their works and even their names passed into oblivion. Natural causes helped the result ; it was difficult to multiply manuscripts, difficult to preserve them, and it is not surprising that the lesser poets should have passed from recollection. On the other hand, the absence of literary remains for the centuries just before and after the Christian era, and the fact that foreign invasions, Greeks, Parthians, and Cakas, and Yueh-chi deeply affected the north-west of India, gave an appearance of reason to Max Miiller's famous suggestion J that there was a comparative cessation of literary activity in India until in the sixth century a great renaissance began with Kalidasa and his contemporaries. The theory is now wholly discredited in the form in which it was put forward, if for no other reason than that it ignored the Brah- manical revival of the Gupta empire at the beginning of the fourth century A. V>. But it lingers on in the form of the suggestion 2 that in the period up to that revival Sanskrit was little used for secular poetry, which was composed in Prakrit, until the reviving power of the Brahmins resulted in their creating the epic by translation from Prakrit originals, developed a lyric poetry to replace the simpler Prakrit songs of the people, and transformed the popular beast-fable and fairy-tale.

For this theory of a Prakrit period of Indian literature preced-

1 India (1883), pp 281 ff. Contrast Lassen, Ind. Alt., ii." 1159 a*.

2 Bhandarkar, Early Hist, of India (1920), pp. 70 ff., who admits the existence of some Sanskrit literature, but places Apvaghosa under Kamska c. A. D. 300. But as early as 185 B.C. theie was a Brahmanical revival under Pusyamilra; EHI. pp. 308 ff. ; Przyluski, La llgendt dc remfcreur Afoka, pp. 90 ff. 4 o ORIGIN AND DEVELOPMENT OF KAVYA LITERATURE

ing the Sanskrit period there is no evidence of value. The sug- gestion of the translation of the epic may be dismissed as absurd, but the case with other forms of literature is more worthy of consideration. The fairy-tale is a thing which readily circulates among the people long before it is dignified by literary treatment by the higher classes of society, and in point of fact there is a strong tradition to the effect that it was in a Prakrit dialect, though one closely allied to Sanskrit, that the great collection of such tales, which powerfully affected Sanskrit literature, as the Brhafkatha of Gunadhya, was composed. Gunadhya's work, however, is of very complex art and uncertain date, and in all probability came into being at a time when we have abundant evidence of the existence of Sanskrit literature, so that this instance is irrelevant to the contention in favour of a Prakrit period of literature. Equally little value attaches to the argu- ment for the priority of Prakrit lyric. It was founded on a wholly misleading view of the antiquity of the anthology of Hala, who was placed in the first centiuy A.D. Against this view must be set the form of Maharastil Prakrit, which shows a development in the language such as cannot be dated before the latter part of the second century A.D., if regard be paid to the evidence of the inscriptions and of the Prakrits of the dramas of Acvaghosa. 1 It is true that Vararuci's Prakrit grammar recognizes Maharastrf of the type of the anthology, but there is no evidence that Vararuci is early in date, for his identification by later tradition with the Katyayana who criticized Panini is without serious value. Jacobi, 2 on the other hand, has identified Hala with the Satava- hana under whom Jain tradition records a change in the Church calendar in A.D. 467. There is no cogent reason to accept or deny this date ; what .is clear is that so far as the evidence goes there is nothing to suggest great antiquity for Prakrit lyric. Ltiders, who finds traces of its existence about the second cen- tury B.C. in the short inscriptions of the Sltabenga and Jogl- mara caves on the Ramgarh hill, and who assigns to the same

1 Bruchstuckt buddh. Dramen, pp. 61 ff. On the SUabenga inscr. cf. Boyer, Melanges Livi, pp. lai ff. Kharavela's date is still disputed.

1 Ausg. Erzahlungtn in MAhdrdshtrt, p. xvii ; cf. Bhavisatla Kaha, p. 83. The Paumacariya of Vimala Sun, the oldest Maharasfrl epic, is not before A. D. 300 and may be much later (cf. ibid., p. 59). THE SOURCES OF THE KAVYA 41

century the Hathigumpha inscription of Kharavela of Kalifiga, which displays, though faintly, some of the characteristics of Sanskrit prose Kavya, makes no claim for the priority of Prakrit to Sanskrit in these literary uses ; on the contrary he acknow- ledges fully the coexistence of a Sanskrit literature.

Still less can be said for the priority of Prakrit in the sphere of the beast-fable. Such fables are readily current among the people, and the Mahabharata shows their popularity in the circles to whom the epic appealed. The Jataka tales of the Buddhists show likewise the skill by which they could be turned to the service of that faith, but of an early Prakrit fable literature we know little or nothing. On the other hand, the Sanskrit litera- ture is marked by the fact that it adopts the fable to a definite purpose, the teaching to young princes and their entourage the practical conduct of life, and thus constitutes a new literary genre.

The causes of the rise of Sanskrit literature are in fact obvious, and there was no need for writers in Prakrit to set an example. It would indeed have been surprising if the simplicity of the earlier epic had not gradually yielded to greater art. The Upanisads show us kings patronizing discussions between rival philosophers and rewarding richly the successful ; we need not doubt that they were no less eager to listen to panegyrics of themselves or their race and to bestow guerdon not less lavishly. We have indeed in the Vedic lists of forms of literature refer- ences to the Naracansls, encomia, 1 which candour admitted to be full of lies, and we have actually preserved a few verses from which we can guess the high praise promiscuously bestowed on their patrons by the singers. Into the Rgveda itself have been admitted hymns which contrive to flatter patrons as well as extol the gods, and added verses, styled praises of gifts (danastulis), recount the enormous rewards which a clever singer might obtain. We cannot doubt that from such contests must have sprung the desire to achieve ever-increasing perfection of literary form as compared with the more pedestrian style of the mere narrative of the epic.

In yet another sphere such heightening of style must have

1 Macdonell and Keith, Vedic Index, i. 445 f. 42 ORIGIN AND DEVELOPMENT OF KAVYA LITERATURE

been striven after. The Vedic poets, who can compare 1 the goddess Dawn to a fair dancer, to a maiden who unveils her bosom to a lover, cannot have been "incapable of producing love poetry for secular use. Nor is it doubtful that it was the early writers of the love lyric who enriched Sanskrit with a vast abun- dance of elaborate metres ; for the flow of epic narrative such metrical forms were wholly unsuited ; on the other hand, the limited theme of love demanded variety of expression if it were to be worthily developed. The gnomic utterance of which the Aitareya Brahmana has preserved some Vedic specimens natur- ally shared in the cultivation of the lyric, and the elaboration of verse doubtless reacted on prose style, inducing writers to seek to reproduce in that medium something of the elegance after which poets now habitually strove. There is, then, no justification for presuming a breach in literary continuity, and, despite the fact that so much has perished, we have indisputable proofs of the active cultivation of Sanskrit literature during the period from 200 B.C. to A.D. 200, when on one theory it had not yet come into being, and secular literature was composed in Prakrit.

2. The Testimony of the Ramayana

The validity of the Ramayana as evidence of the growth of the Kavya has been disputed on the score that the poem was, even if in large measure early in date, 2 still under constant revision, so that those features in it which foreshadow the later Kavya and justify its own claim to that title as the first of Kavyas may be dismissed as interpolations. The argument, however, is clearly unsatisfactory, and does not establish the result at which it aims. We may readily agree that some part at least of the elegancies of style 3 which mark the poem is a later addition, but there is no ground whatever to admit that these additions fall later than the second century B.C., and they may

1 Hirzel, Gltichnisse und Mttaphtrn im Rgvtda (1908). For the early, which is also the later, ideal of feminine beauty, see Qatapatha Brahmana, i. 2. 5. 16 ; iii. 5. 1. 11 ; the love charmsvof the Atharva attest the beginnings of erotic*poetry (IS. v. ai8ff.).

' Keith, JRAS. 1915, pp. 318 ft.

' Jacobi, Ramayana, pp. 1 19 ff. The Ramayana also shows the development of the (Jloka metre almost to its classic state ; cf. SIFI. VIII. ii. 38 ff. See also Krishnamachariar, Raghuvanfavimarfa (1908). THE TESTIMONY OF THE RAMAYANA 43

be earlier in date. The Ramayana in fact, as we have it, affords an illustration of the process of refinement which style was under- going, but it is essential to realize that even in its original form the poem must have shown a distinct tendency to conscious ornament. The mere theme, the blending together of two distinct legends, the couit intrigues of Ayodhya and the legend of Rama's war on Ravana for the rape of Sita — in ultimate origin a nature myth — is the work of an artist, and the same trait is revealed in the uniformity of the language and the delicate perfection of the metre, when compared with the simpler and less polished Mahabharata. Valmiki and those who improved on him, probably in the period 400-200 B.C., are clearly the legiti- mate ancestors of the court epic.

Anandavardhana ' has' not inaptly contrasted the object of the court epic with that of the legend (itihasa) ; the latter is content to narrate what has happened, the former is essentially depen- dent on form. The Ramayana occupies an intermediate place, and its formal merits are not slight. But in any case it essenti- ally anticipates the means by which the later poets seek to lend distinction and charm to their subject-matter; as they drew deeply, upon it for their themes, so they found in it the models for the ornaments of their style. If the city of Ayodhya appears in human form to the king in Kalidasa's Raghuvanga, Valmiki has set the example in his vision of Lanka in the Sundarakanda. The action in the later Kavya is all but obstructed by the wealth of the poet's descriptive powers ; Valmlki's followers have de- scribed with no less than twenty-nine similes the woes of Sita in her captivity, with sixteen the sad plight of Ayodhya bereft of Rama. 2 Descriptions of the seasons, of mountains and rivers, bulk largely in the Kavya, but Valmiki has set the example in his elaborate accounts of the rainy season and autumn, of the winter, of Mount Citrakuta, and of the river Mandakinl. 3 Meta- phors of beauty abound in the Kavya side by side with those of strained taste and pointless wit ; the Ramayana is guilty of visadanakradhynsite paritrasormimalini kitn mam na trdyase magiidm vipule gokasagare ?

1 Dhvanyaloka,?. 148. ! 11. 19 and 114.

' iv. 28 ; iii, 16 ; ii. 94, 95. There is a brilliant picture of the sound of the sea : parvasudirnavegasya sagarasyeva nifisvanah. 44 ORIGIN AND DEVELOPMENT OF KAVYA LITERATURE

' Why dost thou not save me that am sunk in a broad ocean of woe, whose coronal of waves is horror, and in which dwell the crocodiles of despondency ? '

Much happier is the famous simile :

sagaram cambaraprakhyam ambaram sagaropamam Ramaravanayor yuddkam Ramaravanayor iva.

' Ocean peer of sky, sky ocean's counterpart ; Rama and Ravana alone could match their mortal combat.' A later commonplace is foreshadowed in :

tvam krtvoparato manye rupakarta sa vigvakrt na hi rupopamd hy anya tavasti gubhadargane.

' When he had made thee, I ween, the All-maker stayed from his making of lovely forms, for there is no beauty on earth to match thine, o fair-faced one.' As later, we find as prognostications of good the wind that blows free from dust, the clear skies, the flowers that are rained down to earth, and the resonance of the drums of the gods. Indra's banner, erected and then taken down at the festival in his honour, affords material for similes ; eyes expand with joy (harsotphullanayana) ; men drink in faces with their eyes (locanabhyam pibann iva) ; breasts are like golden bowls {kucau suvarnakalagopamau) ; before men's wondering eyes the host stands as if in a picture ; the Ganges shows her white teeth as she smiles in the foam of her waves {pkenanirma- lahdsini) ; winds blow with fragrant coolness ; the clouds rumble with deep and pleasant sound (snigdhagambhiraghosd) ; the action of the fool is like that of the moth that flies into the flame ; man leaves his worn frame as the snake its old skin. The love of alliteration is already present, as in daksina daksinam tiram ; we find even an example of the figure, concise expres- sion {samasokii), in which the dawn is treated on the analogy of a loving maiden :

cancacccwdrakaraspargaharsonmilitataraka aho ragavati samdkya jahatu sv'ayam ambaram.

• Ah that the enamoured twilight should lay aside her garment of sky, now that the stars are quickened to life by the touch of the rays of the dancing moon.' The Ramdyana ifc not given to erotic descriptions ; its tone is serious and grave, but such pasTHE TESTIMONY OF THE RAMAYANA 45

sages 1 as the description of the vision by Hanumant of the sleep- ing wives of Ravana mark the beginning of a tradition which Acvaghosa handed on to his successors. Imitation in detail of the Ramayana is frequent and patent, and its language and verse technique deeply affected the whole of the history of the Kavya. The content of the Mahabharata naturally afforded to later poets an inexhaustible material for their labours, but save in its later additions the great epic suffered little elaboration of style, and affords no evidence comparable to that of the Ramayana attesting the development of the Kavya style.

3. The Evidence of Paiafijali and Pjiigala

Direct and conclusive evidence of the production of secular Sanskrit literature before 150 B.C. is afforded by the testimony of the Mahabhasya? Much earlier evidence from the point of view of grammar would be available, if we could believe the assertion 3 of Rajacekhara — perhaps the dramatist — that Panini was the author not merely of the grammar but also of the Jamba- vatlvijaya ; that epic and apparently another, the Patdlavijaya, are ascribed to him by anthologies which cite verses from them. The fact, however, that grammatical errors occur in a verse from the latter work renders the ascription implausible, even if epic excuse can be alleged, and we may reasonably accept the exis- tence of two or more Paninis, despite the rarity of the name.

The testimony of the Mahabhasya, however, is quite clear, and its value is all the greater because it is given incidentally and by accident in the discussion of disputed rules of the master. Patafi- jali, of course, knows the Bharatan epic, but he refers also to dramatic recitals of epic legends — perhaps to actual dramatic performances — and the topics mentioned include the slaying by Krsna of his wicked uncle Kaiisa and the binding of Bali by the god Visnu. We are told of rhapsodes who tell their tales until the day dawns, and stories were current which dealt with the

1 Not probably by Valmiki. For Vedic precedents in alliteration and Yamakas see Hillebrandt, Kdhddsa, pp. 161 Ff. , for the epic, Hopkins, Great Epic, pp. 200 ff.

» Cf. Weber, IS. xiii. 356 ff., 477 ff. ; Kielhoin, IA. xiv. 326 f. ; Buhler, Die indi~ schcn Inschriften, p. 72 ; Bhandarkar, IA. ili. 14.

3 See Thomas, Kavlndravacanasamuccaya, pp. 5 1 ff. 46 ORIGIN AND DEVELOPMENT OF KAVYA LITERATURE

legends of Yavakrlta, Yayati, Priyafigu, Vasavadatta, Sumanot- tara, and Bhimaratha. A Vararuca Kavya is actually mentioned, though unfortunately we know no more of it. We have, how- ever, invaluable help in appreciating the growth of Kavya in the incidental citation of stanzas clearly taken from poems of the classical type. Many are tantalizing in their brevity ; we hear of a maiden bought with a price who was dearer to her lord than his life (sd hi tasya dhanakritd prdnebhyo 'pi gariyasi). The verse varatanu sampravadanti kukktitah, ' O fair one, the cocks pro- claim together', has afforded later authors an opportunity of exhibiting skill in filling up the missing three verses {samasyd- purana)} Erotic verse is attested also by priyam mayurah pratinamrtiti, ' The peacock danceth towards his beloved ', perhaps also by a variant 'ad odakdntat priyam pdntham anuvrajet, ' Let her follow the wanderer she loveth to the end of the woods, to the end of the waters '. Epic or panegyric is found in the address prathate tvayd patimati prthivi, 'The earth with thee as lord maketh true its name as wide ' ; so also asidvitiyo 'nusara Pandavam, ' With sword as mate he attacked Pandu's son ', jqghdna Kansath kila Vasudevah, ' Vasudeva slew Kansa.' Brief as it is, there is pathos in

yasmin daga sahasrani putre jate gavaih dadau brahmanebhyah priydkhyebhyah so 'yam unchena jivati. ' On his scanty gleaning now he liveth, he for whose birth were given ten thousand kine to the Brahmins who brought the good tidings.'

Gnomic poetry is also strongly represented :

tapah grtttam cayonig cety etad brahmanakarakam tapahgrutabhydm yo hino jatibrahmana eva sah. 'Asceticism, learning, birth, these make the Brahmin; he who lacks asceticism and learning is a Brahmin by birth alone.' Or again, bubhuksitam na pratibhati kimcit, ' Nothing seems right to a hungry man.' Solomon's maxim regarding the education of children has a worthy parallel :

samrtaih panibhir ghnanti guravo na visoksitaih ladandgrayino 2 dosds tddandgrayino gundh.

1 See Chap. IX, § t.

2 Cf. the forms in Festschrift Wackernagel, p. 303. THE EVIDENCE OF PATANJALI AND PINGALA 47

' Fraught with life, not with poison, are the blows that teachers give ; vice grows by indulgence, virtue prospers by reproof.' The inevitability of death is recorded :

ahar ahar nayamano gam agvam purusam pa gum Vaivasvato na trpyati surety a iva durmadl.

' Though day by day he takes his toll in cattle, horses, men, and beasts, Vivasvant's son is sated never, as a drunkard is never wearied of brandy.' A maxim of political wisdom may be seen in

kseme subhikse krtasamcayani : purani rajnam vinayanti kopam.

' Citadels well stored in peace and abundance calm the wrath of kings.'

Noteworthy also is the fact that in the scanty number of verses there occur specimens of such ornate metres as the MalatI, the Praharsini, the Pramitaksara, and the Vasantatilaka, beside the normal C/loka and Tristubh. These new metres lead us into a different sphere from the Vedic metres, and striking light on this development is afforded by the metre of the Karikas, 1 mostly, if not all, written probably by predecessors of Patanjali, which deal with disputed points of grammar. Among these are besides the C/loka and Vaktra, Indravajra, Upajati, CalinI, Van- castha,all later usual, and the much less common metres, SamanI, consisting of four verses each of four trochees, Vidyunmala, similarly made up of spondees, the anapaestic Totaka, and the Dodhaka, in which the verse has three dactyls and a spondee. This richness and elaboration of metre, in striking contrast to the comparative freedom of Vedic and epic literature, must certainly have arisen from poetical use ; it cannot have been invented for grammatical memorial verses, for which a simple metre might better suffice. The names Totaka and Dodhaka have been sus- pected of Prakritic origin, and the latter of ultimate Greek origin, but these are unproved hypotheses without literary or other support.

In addition to the clear indications thus given of the existence of epic, lyric, and gnomic verse, we may deduce from other hints the existence of the material whence later developed the beast -

1 Cf. Kielhorn, IA. xv. 229 ff. ; Jacobi, Festschrift Wackernagel, p. 127. 48 ORIGIN AND DEVELOPMENT OF KAVYA LITERATURE

fable. We have allusions 1 to such proverbial tales as that of the goat and the razor (ajakrpaniya), of the crow and the palm fruit {kakataliyd), and to the hereditary enmity of the snake and the ichneumon, and of the crow and the owl, later famous as the theme of a book of the Pancatantra.

Corroboration of the evidence of Patanjali can be obtained from the Chandassutra of Pingala, which ranks as a Vedanga but is mainly devoted to the exposition of secular prosody. Pingala ranks as an ancient sage, being .sometimes identified with Patanjali ; the aspect of his work suggests considerable age, and many of the metres which he describes are certainly not de- rived from the Kavya literature which has come down to us. They suggest a period of transition in which the authors of the erotic lyric 2 were trying experiment after experiment in metrical effect. The names of the metres can often most plausibly be ex- plained as epithets of the beloved ; the stanzas may have been so styled because the word in question occurred in them. Thus we have the metre Kantotplda, the plague of her lovers, Kutila- gati, she of crooked gait, Cancalaksika, she of the glancing eyes, Tanumadhya, she of the slender waist, Caruhasini, the sweet- smiling one, and Vasantatilaka, the pride of spring. Other names suggest poetic observation of animal life ; thus we have Acvalalita, the gait of the horse, Kokilaka, the cry of the cuckoo, Sinhonnata, tall as a lion, Qardulavikrldita, the tiger's play. The plant world gives others as Manjarl, the cluster, Mala, the garland. That a strong school of lyric poetry existed about the Christian era and probably much earlier we cannot seriously doubt ; to its influence we may with reason ascribe the appearance and bloom of the Maharastrl lyric about A.D. 200.

4. Kavya in Inscriptions

Chance has preserved for us certain evidence in the early in- scriptions 3 which disposes definitely of the theory of the dormancy of Sanskrit during the period of foreign invasions in India. An inscription at Girnar * dated about A. D. 150-2 under the Maha-

1 Mahabkasya, li. 1. 3 ; v. 3. 106 ; IS. xiii. 4S6.

2 Jacobi, ZDMG. xxxvni. 615 f.

s Buhler, Die indischen Inschrtften und das Alter der indischen Kunstpoeste (1 890).

• EI. vin. 36 ff. ; EHI. pp. 139 f. ; IA. xlviii. 145 f. KAVYA IN INSCRIPTIONS 49

ksatrapa Rudradaman, grandson of the Ksatrapa Castana, known to Ptolemy as Tiastanes of Ozene, Ujjayinl, is written in prose {gadyam kavyam) and shows in a most interesting manner the development from the simple epic style to that of the Kavya. Grammar is obeyed, but epic licence is found ; patina, for patyd, is thus explained, and vtgaduttarani is a Prakritism for vincad-, which the epic, though not the grammar, permits ; epic again is the pleonasm in Parjanyena ekarnab/tutayam iva prihivydm krtdydm, ' when the storm had turned as it were all earth to ocean '. But in anyatra samgrdmesu, ' save in battles ', we have a pure error. From the epic style a distinct departure is made in the use of compounds ; Dandin, doubtless following earlier authority, bids them be used freely in prose, and approves of their being long. The inscription prefers compounds to simple words, and at the beginning presents us with a compound of nine words with twenty-three syllables ; the description of the king produces even a finer effort of seventeen words of forty syllables. The length of the sentences vies with that of the compounds ; one attains twenty-three "Granthas, each of thirty-two syllables. Of the figures of sound (gabdalamkaras) alliteration is freely used as in abhyastandmno Rudraddmno, sometimes with real effect. Of figures of sense (arlhalamkaras) one simile compares in the later manner the curtain wall of a reservoir to a mountain spur in the Kavya phrase parvatapratisparddhi. The description, if never of a very high order, displays some merit, especially in the vivid picture of the destruction by flooding of the dam of the reservoir. But what is far more important is that the author thinks it fit to ascribe to the king the writing of poems in both prose and verse ; flattery or not, it was obviously not absurd to ascribe to a Ksatrapa, of foreign extraction, skill in Sanskrit poetry. Moreover, the poems are qualified by a string of epithets as adorned by the qualities of simplicity, clearness, sweetness, variety, beauty, and elevation arising from the use of conventional poetic terminology (sphutalaghiHtiadhuracitrakdnta- gabdasamayoddrdlamkrta). The term almnkrta points unmis- takably to the author's acquaintance with a science of poetics prescribing the ornaments of poetry, and a comparison with the merits ascribed by Dandin 1 to the Vaidarbha style which he 1 Kavyfidarfa, i. 40 ff. See below, cbap xviii, § 2.

3H9 E 50 ORIGIN" AND DEVELOPMENT OF KAVYA LITERATURE

admires is decidedly instructive. Simplicity and clearness may well be equivalent to the arthavyakti and prasdda which he mentions ; sweetness is his mddkurya which includes richness in tasteful sound and sense (rasavai) ; variety is probably akin to the strength or force (oj'as) prescribed by Dandin, and he recog- nizes that in the view of some authorities elevation was induced by the use of the stock terms of poets such as kridasaras, a lake for sport.

The evidence of this inscription is confirmed and strengthened by that derivable from a record 1 of Siri Pulumayi at Nasik, written in Prakrit prose. There can be no doubt of the familiarity of the writer with Sanskrit ; it is even possible that he wrote his» text in that language and then, in order to comply with the usage of the- day, rendered it into Prakrit for purposes ot publication. Siri Pulumayi may be identified with Siro-Polemaios of Baithana, Pratisthana on the Godavarl, of Ptolemy and the date of the inscription is not far removed from that of the Girnar record. It begins with an enormous sentence of eight and a half lines, long compounds fill lines 2-6, then a brief rest is given by the insertion of short words, and the whole ends with a compound of sixteen words and forty-three syllables. This is deliberate art, however little we may admire it, and the same technique is found in Bana, used perhaps with greater skill. Alliteration is freely used ; the queen is mahddevl tnahardjamdta mahdrajapatdmahi. What, however, is specially interesting is the appearance of mannerisms of the later Kavya, used in a way which implies . current familiarity with the themes. Thus the king is of like strength with the mountains Himavant, Meru, and Mandara, a brief allusion to the view that the king, like the Himalaya, possesses abundant treasures, like Meru is the centre of the world and overshadows it with his might, and, like Mandara, which the gods used as their churning stick when they churned the ocean, can produce and preserve LaksmI, the for Uma regum. The king again is compared with the heroes of the epic in a manner which preludes the frequent use of this theme made by Subandhu and Bana. Finally, he is described as winning

1 EI. viii. 60 ff. ; S. Levi, Cinquanlenaire de Vicole pratique des Hautes Etudes (1921), pp. 91 ff., who holds that its hero Gotamlputa's, death in victory is described. KAVYA IN INSCRIPTIONS 51

victory in a battle in which in wondrous wise the Wind, Garuda, Siddhas, Yaksas, .J&aksasas, Vidyadharas, Bhutas, Gandharvas, Caranas, the sun, the moon, the Naksatras, and the planets take part. Thus early we find that confusion of the mortal and the supernatural which induces an alleged historian like Bilhana to allow Civa to intervene when needed in the fate of his patron. There can be no doubt from these inscriptions of the existence of Sanskrit Kavya, and doubtless also of a science of poetics among the Brahmins. 1 It is, therefore, accident only which has preserved Buddhist works like those of Acvaghosa as the earliest specimens of the Kavya. Moreover there is a simple explanation of the accident ; Acvaghosa was one of the great names of Buddhism ; no one arose to surpass his achievement in depicting the life of the Buddha, whereas the glory of earlier poets was eclipsed by that of Kalidasa. Nor is this mere theory ; we know in fact that of the predecessors in drama enumerated by Kalidasa himself the works of all save one are lost, apparently irretrievably.

5. The Kamasutra and the Poet's Milieu.

Vatsyayana's Kamasutra 2 is of uncertain date, but it is not improbably older than Kalidasa, and in any case it represents the concentrated essence of earlier treatises on the Ars Amoris. There is no question of the importance of knowledge of this topic for the writers of erotic poetry, and there is abundant proof that the Kamasutra was studied as eagerly by would-be poets as were grammar, poetics, and lexicography. To Vatsyayana we owe a vivid conception of the Indian parallel to the man about town (nagaraka) whose existence was due to the growing elaboration of Indian life, and whose interest the poet was anxious to pro- pitiate. We see him, 3 opulent, a denizen of the town which lends him his name, or, if compelled by adverse fortune to vegetate in

1 The use of compounds in ornamental epithets appears to have been much pro- moted by their convenience in eulogies of kings, places, &c, in inscriptions, just as in

Jain texts they are heaped up in stock descriptions.

2 See below, chap, xxiv ; cf. Haraprasad, Magadkan Literatttrc, chap. iv. On the arts, Kalas, sixty-four in number at least, of early India, see A. Venkatasubbiah and E Midler, JRAS. 1914, pp. 355-67.

• The comm. allows him to be of any caste.

E a 52 ORIGIN AND DEVELOPMENT OF KAVVA LITERATURE

the country, seeking, like Martial in his retreat from Rome, to find congenial society with which to continue the pleasures of his town life. His home boasts all the luxury of the age, soft couches, a summer house in a park, seats strewn with flowers, and swings to amuse the ladies who share and lend zest to his leisure moments. Much of his time is devoted to his toilet; he must bathe, be anointed, perfumed, and garlanded ; «then he can teach the cage birds which surround him to speak, or enjoy the brutal spectacle of ram or cock fights, both favourite amusements of the gilded youth of the period. Or, in the company of ladies of the demi-monde, he may visit the parks outside the town, returning home crowned with the flowers which they have plucked. There are concerts to be attended, ballets and theatrical spectacles to be visited ; he has a lute beside him so that he may make music when he will, and a book to read at leisure. Boon companions and hangers-on of various ranks, the Vitas, Pithamardas, and Vidusakas of the texts, are essential to his happiness, and drinking parties are not unknown, but the ideal forbids mere rude licence ; even in his enjoyments the man about town aims at elegance, moderation, and a measure of dignity. He con- descends to the use of the vernacular, but blends it with Sanskrit, thus indicating his fine culture. Hetairai are essential to him, but they also are not without accomplishments ; indeed the Kamasutra demands from them knowledge encyclopaedic, in- cluding poetic taste. The most famous of them achieved great riches, as we learn from the description of the palace of the heroine in the Mrcchakatika and, as in the Athens of Perikles, discussions on literature, music, and art, must often have afforded the participants a pleasure which could not be expected from their own wives, from whom they demanded children and care for their homes.

An atmosphere of this kind is unquestionably favourable, if not to the highest poetry, at least to the production of elaborate verse, and the care demanded from those who are exposed to keen criticism cannot but produce excellent results in the case of men naturally gifted, though on the other hand it leads to ex- aggerated love of style with inevitable tasteless extravagance. If under such a system Maecenases produce few Vergils, they are responsible for a plentiful crop of Valerii Flacci, and to the kings THE KAMASUTRA AND THE POET'S MILIEU 53

of India 1 we unquestionably owe most of the poets of repute; patronage by the king was at once the reward of skill in panegyric and the means of obtaining the leisure for serious composition and a measure of publicity for the works produced. It was the duty of the king to bridge the gulf between wealth and poetic talent, of the poet to save his patron from the night of oblivion which else must assuredly settle on him when his mortal life closed. At the royal courts poets vied in eager rivalry with one another ; probably in quite early times there were practised such arts as the composition of verses to complete a stanza when one verse was given, and the production of extempore poems on a given topic. The festival of SarasvatI each month afforded opportunities for displays in honour of the patroness of poetry and the arts. Fortunately, too, for the poets, kings were willing to claim renown for skill in poetry; we have seen that his panegyrist thought well to ascribe fame in this sphere to Rudradaman and we shall see that the great Gupta Emperor Samudragupta strove for renown as a man of letters. 2 Harsa not only patronized Bana, but claimed the authorship of dramas and poems, though unkind hints were prevalent that others were the true begetters of his literary offspring. 3 Four hundred years later Bhoja of Dhara was more fortunate, for we have no real knowledge to disprove his claim to polymathy exhibited in a large variety of works. In the twelfth century * the court of Laksmanasena revived the glory of Harsa's patronage, for besides the famous Jayadeva, other poets such as Umapatidhara, Dhoi, and Govardhana wrote with acceptance. The kings of Kashmir often distinguished themselves by generosity to their laureates, {kaviraja) and to such enlightened activity we owe Somadeva's

1 Rajacekhara {Kavyanumansa, p. 55) gives Vasudeva (' the Kanva or the Kusana), Satavahana, Cudraka, and Sahasfinka (?Candragupta II J Pischel, GN. 1901, pp. 485- 7) as famous patrons.

• Minor royal authors include the dramatists Mahendravikramavarman (c. 675) ;

Ya90varman, patron of Bhavabhuti (c. 735), the Kalacuri Mayuraja (c. 8oo), and Vigraharajadeva (1153). We have stanzas of a Nepalese king (8th cent.), of Amogha- varsa (815-77), of MuBja (975-95), and Arjunavannan's comm. on Amaru (13th cent.). Cf. Jackson, Priyadarhka, pp. xxxvii ff,

• Cf. Keith, Sanskrit Drama, pp. i7off.

4 Smith, Kill. pp. 419ft., 432 wishes to place this king about fifty years before the usual date, but ignores important evidence; see K. C. Majumdar, JPASB. 1931, pp. 7 ff. ; C. V. Vaidya, IHQ. i. 126 ff.; C. Chakravarti, iii. 186 ff. 54 ORIGIN AND DEVELOPMENT OF KAVYA LITERATURE

Kathasaritsagara. Yet it is worth remembering that we cannot prove any royal patron for Kalidasa, greatest of Indian poets, or even for Kalhana, the one historian of teal merit in Sanskrit litera- ture. Nor, of couise, was royal generosity confined to Sanskrit poetry; to a king, Hala or Satavahana, is ascribed the anthology of Maharastri verse, and Vakpatiraja wrote his epic, Gaiidavaha, for Yacovarman of Kanauj, thus assuring him an immortality to survive his defeat at the hands of Lalitaditya of Kashmir. So, too, if we believe tradition, it was perhaps the patronage of Kaniska which produced the first great work of the court epic preserved to us, the Buddhacarita of Acvaghosa. Ill

AgVAGHOSA AND EARLY BUDDHIST KAVYA i. Agvaghosci s Works.

THE deplorable darkness which still envelops early India renders it impossible to establish with certainty the date of Acvaghosa, famous alike as a poet and as a philosopher. Tradition unquestionably makes him a protege of the famous Kaniska, but the matter is complicated by the fact that if the Siilralamkara l is his, he tells two stories in which Kaniska's reign seems to be referred to as in the past ; this may be explained either on the theory that Kaniska died before him, which does not accord with tradition, or on the view that the stories are interpolated in whole or as regards the name, or that there was an earlier Kaniska ; again an inscription 2 held to belong to the time of Kaniska mentions an Acvaghosaraja who has been temerariously identified with the poet. Assuming the validity of the tradition despite these difficulties, the date of Acvaghosa would fall to be determined by that of Kaniska, for whom c. A.D. ioo 3 still seems a just estimate. Tradition also tells that he was originally a Brahmin, that he first adhered to the Sar- vastivada school of Buddhism, but was attracted by the doctrine of the saving grace of faith in the Buddha, and became one of the forerunners of the Mahayana school. I-tsing, who travelled in India in A.D. 671-95, refers to him as one of the great teachers of the past, and asserts that a collection of his works was still studied in his time. From the colophons of his own works we learn that his mother was named Suvarnaksl and that his home was Saketa, while he is given the style of Acarya and Bhadanta.

I Nos. 14 and 31 (Huber's trans., Paris, 1908). Cf. Levi, JA. 1896, ii. 444 ff. ; Kimura, IHQ. i. 417. Kumaralata (c. 150) is more probable.

II EI. viii. 171; S. Ch. Vidyabhusana (POCP. 1919, I. xxxiiiff.) puts Kaniska, patron of Acvaghosa, about A. D. 320.

3 Cf. Smith, EHI. pp. 272 ff. ; Foucher, VArt Grtco-Bouddhiquc, ii. 484 ff., 506 ff., who finds in the (Jaka epoch merely the beginning of the fifth century of the Maurya epoch, placing Kaniska c. A. p. 81. Cf. D. R. Sahm, JRAS. 1924, pp. 399 ff. 56 AgVAGHOSA AND EARLY BUDDHIST KAVYA

Whether the Mahaydnagraddhotpdda, a famous text-book of early Mahayana views, or the Vafrasiia, an able and bitter attack on the Brahmanical caste system, are rightly ascribed to Acvaghosa need not be discussed, and his dramas are preserved only in fragments, which reveal little of his poetic skill. 1 Of the songs for which he was renowned the Gandistotragatfia 2 displays great metrical skill and attests his comprehension of the power of music ; it is an effort to desciibe in words the religious message carried to the hearts of men by the sounds produced by beating a long strip pi wood with a short club. Of later authorship is the Sutralamkdra or Kalpanamanditikd, which unhappily is preserved only in a fragmentary condition in Sanskrit, though Huber has translated into French the Chinese version of A.D. 405. The wide culture of the writer displays itself in his allusion to the Bharatan epic 3 and the Rdmdyana, the Samkhya and Vaicesika philosophies', and Jain tenets, while in the tales he exhibits himself as a fervent believer in the doctrine of the saving power of worship of the Buddha. The collection is made up of tales, in the main already current in Literature still preserved, inculcating the Buddhist faith; many are attractive, even pathetic, but the doctrine of devotion carries the author to strange results, as in the tale of the sinner who never in his life did one good deed, but because in deadly terror of his life from attack by a tiger he uttered the salutation, 'Homage to the Buddha ', is granted entrance to the order and straightway pro- ceeds to sainthood. From the literary point of view the essential fact is that the tales are written in prose and verse, clearly of the classical type. We need not doubt that this combination was taken over by the author direct from the contemporary Jatakas current in Pali, even if no strict proof of -this view is possible.

The Sutralamkdra mentions a Buddhacarita, perhaps Acva- ghosa's work, and there is reason to suppose that that epic was later than the Saundarananda* At the close of that work Acvaghosa frankly declares the purpose which led to his adopting the Kavya

1 Cf. Keith, Butldh. Phil., pp. 252 ff. ; Sanskrit Drama, pp. 80 ff.

2 Ed. BB. 15, 1913.

s We find two verses from the Harivah$a in the Vajrasiici.

4 Ed. Haraprasad Sastri, BI. 1910. Cf. Baston, JA. 1913, i. 79 ff. ; Hnltzsch, ZDMG. lxxii-Ixxiv ; Gawronski, Studies about tht Sansk. Buddh. Lit., pp. 56 ff.. A?VAGHOSA'S WORKS 57

form ; he recognizes that men rejoice in the delight of the world and seek not salvation, and therefore he sets out the truth which leads to enlightenment in attractive garb, in the hope that men attracted by 1 it may realize the aim and extract from his work the gold alone. As he makes no allusion to an earlier poem, we may conclude that the Saundarananda was his first attempt. The topic of the poem is the legend of the conversion of the reluct- ant Nanda, his half-brother, by the Buddha, a story recounted in the Mahavagga and the Niddnakatha, but Acvaghosa deals with it in the approved manner of the later Kavya. He begins with an account of the foundation of Kapilavastu, which gives him occasion to display his knowledge of heroic tales and mythology (Canto i). There follows the description of the king, £uddho- dana, and briefly an account of the birth of Sarvarthasiddha and his half-brother Nanda. The Buddha is described in full in the next Canto (iii) ; then we hear of Sundarl's beauty and the perfection of her union with Nanda as of the night with the moon. Reluctantly Nanda leaves her (iv), and the Buddha hastens to secure his ordination as a monk, much against his inclination (v). Bitter is Sundarl's grief (vi), and Nanda himself seeks by a long list of legendary parallels to defend his desire to cling to his beloved ; kings of yore have laid aside the hermit's garb and returned to the world of joy and life (vii). In vain are the demerits of women, the flattery on their lips, the treachery in their hearts, pointed out (viii) ; in vain is he warned of the evils of pride illustrafed by the fate of heroes of the past (ix). The Buddha determines on a bolder plan ; he carries him to heaven and shows him on the way in the Himalaya a one-eyed ape of hideous form, asking him if Sundari is fairer than it. Nant'a energetically asserts his wife's loveliness, but on the sight of the heavenly Apsarases must admit that their beauty raises them as far above Sundari-as she is above the ape ; with fickle faith he resolves to win an Apsaras as bride, but is warned that he must win heaven by good works, if he is to obtain this end (x). Re- turned to earth he strives for this end, but Ananda warns him, adducing a wealth of examples, that the joys of heaven are fleeting and that, when man's merit is exhausted, he must return to earth again (xi). Nanda is thus induced to lay aside all thought of heavenly joys and to seek and obtain instruc58 AQVAGHOSA AND EARLY BUDDHIST KAVYA

tion from the Buddha ; he becomes not merely a saint, but on the Buddha's bidding determines on. the nobler course of seek- ing salvation not for himself alone', but of preaching it to others (xii-xviii).

The Buddhacarita J deals with the greater theme of the life of the Buddha, and it is a misfortune that as we have it the poem contains but seventeen Cantos and qf these only the first thirteen — with certain exceptions — are genuine, the remainder being an addition made a century ago by Amrtananda who records that he did so because he oould not find a manuscript of the rest of the text. The poem now ends with the conversions made at Benares, but the Chinese version, made between a.d. 414 and 421, and the Tibetan, have twenty-eight Cantos, and I-tsing still knew of this number. The exact source which influenced Acvaghosa in his choice of incident is unknown, for it is not proved that the Lalitavisiara existed in his time in anything like its present form. In any case the contrast between the two works is remarkable ; the LalttavislaraJ^wiitten in the».main in Sanskrit* prose of the plain type, intermingled with Ballads in mixed Sanskrit of the so-called Gatha style ; at best it is confused, at worst incoherent. Acvaghosa's poem is essentially the work of an artist ; in choice of incident and arrangement he seeks to produce the maximum effect, and, though he does not vary in . essentials the tradition, he renders vivid and affecting the scenes " which he describes.®The prince's fatal journeying forth from the palace which brings him into contact with the hateful spectacle ' of age, is preceded by the account of the fair women who crowdg to Watch his exit ,Gth_e poet again shows his skill in depicting the loving ruses by which the ladies of the harem seek to divert his mind from the desire to renounce the vanities of tha wiyWi . and in describing the famous scene when the prince gazing •dn them in their sleep resolves to abandon &e palace. Nor is he skilled in the Kamacastra alone ; he adduces the arguments by which the family pjiests, fortified by the precepts of political science, seeks to dereruie prince from his resolution to abandon

1 Ed. E. B. Cowell, Oxford, 1893 ; trans. SBE. 46 ; Formichi, Baii, 1912. See also Hullzsch, ZDMCJ. lxxii. 145 ff. ; Cappeller, ZII. ii, iff. ; Speyer, JRAS. 1914, pp. 105 ff. ; GavvroAski, Rocznik Oryentalistyczny , i. iff.; i-v ed. and trans. K. M, Joglekar, Bombsty, 191a. On Buddhist Sanskrit Literature cf. G. K. Nariman, Sanskrit Buddhistic (1923). AgVAGHOSA'S WORKS 59

secular life with its duties, and true to the rule which requires a description of a battle he provides a spirited picture of the contest of Buddha against the demon Mara and his monstrous hosts. . £&*)

. There is not the slightest doubt of one of the sources of Acvaghosa. Though Cowell was unable to find decisive proof of his knowledge of the Ramayana as opposed merely to the legend of Rama, the fact is put beyond doubt, apart from a men- tion of the poem in the Sutralainkdra, by careful study of the references in the Bitddhacarita itself 1 ; when the people of the town see that Siddhartha has not returned they weep as afore- time when the chariot of Dacaratha's son returned without him ; Quddhodana compares himself to Dacaratha, bereft of Rama, whose death he envies, and in 'these and many other passages there is clear knowledge by Acvaghosa of the wording of our present text. It was natural that the parallel should deeply affect Acvaghosa, and the broad structure of the episode of the return of Sumantra to Ayodhya without Rama and of Chan- 1 daka to Kapilavastu without Siddhartha is unmistakable ; the charioteer leaves his master, and returns to the city now sadly changed ; the eager citizens rush out to greet him, learn his news, and are filled with lamentation ; the women throrfgThe windows and then withdraw in deep depression to their inner chambers ; the charioteer enters the presence of the king. Similarly again, Yacodhara's lament for the sufferings of the prince in his new life of hardship is modelled on Sita's sorrow for her husband's sufferings in the forest. Nor does it seem reason- able to deny that the description of the aspect of the women 6f the harem in sleep is based on the portraiture of Ravana's harem. 2

2. AQvaghosd s Style and Language.

Dandin 3 draws a vital distinction between two styles as preva- lent in his day, the Gauda and the Vaidarbha, easterr? and 4S outhern, and from his account and other evidence we gather that

1 Gawronski, Studies about the Sansk. Buddh. Lit., pp. 27 ff.

• v. 9-1 1, which Wintemitz (GIL. i. 417) asserts to be based on Acvaghosa. But

see Waller, Indifa, iii. 13. ' Kavyadarca, i. 40 ff. 60 AQVAGHOSA AND EARLY BUDDHIST KAVYA

among the characteristics of the former was the love of long compounds not merely in prose, where they were accepted even by the Vaidarbha, but in verse also ; love co^literation and of

'-harsh sound effects; the use of recondite "etymalogTzing, phrase- ology, and a desire for strength resulting often in bomba'Sf* alia affectation. It has been suggested by Jacobi * that the contrast of styles has a historical basis; Sanskrit poetry was. practised, it is argued, eagerly in the east and Sanskrit poetry there had developed the evil effects of old age, before the art became current in the west and south. The' simpler style of the south^ was also on this view influenced by the freshness of the lyric of Maharastra born of close contact with the people. It is already a serious objection to such a conclusion that in the Natyaqastra we find the qualities which Dandin ascribes as characteristic of the Vaidarbha ascribed to the Kavya style in general ; this is a strong suggestion that at the time of the Natyaqastra there had not developed those characteristics of the Gauda style, and that they emerged gradually with the development of poetry at the courts of princes of Bengal. This view gains support from the fact that, though Dandin praises the Vaidarbha style, and evidently disapproves of the Gauda, in practice poets of later date often affect the Gauda manner. Acvaghosa, however,

^affords a more convincing proofj>till of the e arly character of the Vaidarbha.; his style umnTstakably is of the Vaidarbha type ; as Bana later says of the western poets, it aims at sense rather than mere ornamenT; it is his aim to narrate, to describe, to preach his curious but not unattractive philosophy of renunciation of selfish desire and universal active benevolence and effort for the good, and by the clarity, vividness, and.elegance' of nis diction to attract the minds of those to whom blunt truths and pedestrian statements would not appeal. This project left no room for mere

^elegance or for deliberate^straming after effect, and thus it results that Acvaghosa's works attain a .high measure of attractiveness, especially when we make the necessary allowance for the decidedly bad condition of the'text tradition of both epics. Simple.of course, in the sense in which it can be applied to English poetry, is an inappropriate iepuriet as regards any Sanskrit Kavya, but rela- tively to ihe later standard, even in some measure to Kalidasa, 1 Ausgeitiahlte Erzahlungen in Mdhdidshtrt, pp. xvif. ^^-"^ACVAGHOSA'S STYLE AND LANGUAGE 61

Acvlighosa's style is simple. Nor may we deny it the epithets of sensuous and passipnate ; the picture of the pleasures of love drawn by Acvaghosa is~already marked by that wealth of

by Acvaghosa is~already marked by e detail which appeals to all Indian p

^intimate detail which appeals' to all Indian poets, but proves

a grave stumbling MDlock to critics who find matter for offence even in the charming picture of the deceiving Zeus in the Iliad and reprobate in the_author of the Odyssey the' episode of the amour of Ares and Aphrodite. But still more sincere is the burning emnusi&sm of the poet for his own ideal, not the Arhat,

^contented to seek his own freedom from rebirth in this world of misery, but the Bodhisattva, the Buddha to be, who delays, how- ever, his entering into Nirvana until he has accomplished his view

of freeing all other creatures from the delusion which makes

Uhem cling throughout the ages to mortal life and its woes.' This is a new note in Sanskiit poetry; Valmlki has majesty and a calm seriousness, but he is fr&^£ rom passion like his hero, who though he experiences vicissituaes yet stands apart from them, and of whose ultimate success we never doubt. Nanda's rejection of Sundarl may seem to us heartless enough ; his transference of his fickle affection to the Apsarases has its comic side, but in the end he seeks the welfare of others, even as does tire Buddha ; Rama on the contrary in his rejection of Slta aft(*r the long agony of separation from him has no warmer motive than obedi- ence to the doctrine that Caesar's wife must be above- suspicion. As Cuddhodana reminds us— of Dacaratha, so Sundarl has traces of Slta, but with a vehemence of passion unknown to that queen, and without her dignity and steadfast courage. Nor is it in theme and character-drawing alone that Valmlki is laid under contribution ; the metaphors and similes of the Ramayana x appear in more refined form ; the king, hearing of his son's final

' resolve, falls, smitten by sorrow as Indra's banner is lowered when the festival is over {Qaclpater vrttfl ivotsave dhvajah) ; the maidens stand drinking in the prince's beauty with eyes that stay wide-open in joy {nfycalaih prltivikacaih pibantya iva loca- naik) ; they display their bosoms that are like bowls of gold {suvarnakalagaprakhyan dargayantyah payodfiaran). The epic speaks of the ocean laughing with the foam of its waves, the poet embodies the idea in the picture of a sleeping beauty of the 1 Cf. Walter, Indica, iii. 1 1 ff. / 62 AgVAGHOSA AND EARLY BUDDHIST KAVYA

harem, with a daintiness of elaboration which is far removed from the epic :

vibabhaii karalagnavenur any a: stanavisrastasitdhguka gay ana rjusatpaaapanktijustapadma : jalaphenaprahasattatd nadlva.

' And one lay resplendent, holding a flute in her hand, while her white garment slipt from her bosom, like unto a river whose banks laugh with the foam l of her waves, and in whose lotuses long rows of bees delight' Acvaghosa unquestionably is at his best in simple and elegant description by which a clear picture is presented to the eyes :

tathapi pdplyasi nirjite gate : digah praseduh prababhau

nigdkarah divo nipelur bhuvi pnspavrstayo : raraja yoseva vihalmasa

nigd. ' So when the evil one had retired worsted, the sky became calm, the moon shone forth, flowers fell in rain from heaven on the earth ; night shone clear like a maiden free from stain.' When the charioteer returns :

punafv kumdro vinivrtta ity atho : gavdksamdldh pratipedire

','tgandh vivikiaprstham ca nigamya vajinam : punar gavaksani pid-

liaya cukrugnh. ' " 'Tis the prince returned ", said the women and rushed to their windows, but, seeing the steed's back bereft of its master, closed them again and wailed aloud.' Yacodhara, who is more akin to Sita than Sundarl, laments her husband's new lot :

gucau gayitvd gayane hiranmaye : prabodhyamdno nigi titr-

yanisvanaik kathatn bdfa svapsyali so 'dya me vrati: pataikadegdntarite

mahlthle. ' How can he skep to-night, my faithful one, on one poor mat covering the banb earth, he who hath slept aforetime on a couch of gold undefileld, and whom music hath aroused from his slumbers ? ' Afvaghosa is also a master of simple pathos :

mahatyd irsmiyd duhkhair garbhendsmi yayd dhrtah ' tasyd iiisphalayatndydh kvdham mdtuh kva sd mama.

1 Cf. Meghaduta, 50. THE AV AD AN AS 65

believed wholeheartedly in the efficacy of any act of devotion to the Buddha or his followers as having the power to influence indefinitely for good the life of man; equally they held that an insult to the Buddha was certain to bear appalling fruit. Of the Avadana texts preserved the oldest may be the Avadanagataka}- which is stated to have been rendered into Chinese in the first half of the third century A.D., and which, as containing the term dlnara, can hardly belong to any period earlier than A.£>. 100. Artistically the work has scanty merit ; its arrangement in ten decades each according to subject-matter is schematic ; the tales open with set formulae, contain set formulae of description, as of the laughter of the Buddha, and of moral exhortation ; exaggera- tion and long-windedness mark the whole, and beauty of form is sacrificed to the desire to be edifying. From this point of view, indeed, the tales often reveal thoughts of some beauty ; Maitra- kanyaka, condemned for wrongs done to his mother to endure in hell the punishment of bearing on his head a, wheel of red-hot iron for 66,000 years until another who has committed a like sin comes to relieve him of his burden, resolves that rather will he for ever and ever endure the pain, and is rewarded forthwith by the disappearance of the instrument of torment. QrimatI, wife of Bimbisara, pays homage to the relics of the Buddha which the king had enclosed in a Stupa for worship by the ladies of his harem ; the parricide Ajatacatru forbids such homage on pain of death, but C/rimatl disobeys, and, slain by the king's order, is born again in the world of the gods.

Far more interesting as literature is the Divyavadana, 2 a col- lection of legends which draws, like the Avadanagataka, largely on the Vinayapitaka of the Sarva^tivaditi school of Buddhism. Its date is uncertain ; its origin is complex ; one section is definitely described as a Mahayana Sutra, while the body of the work is still of the Hinayana school. The term dlnara occurs, and one famous tale, the Cjirdulakarnavadana, was rendered into Chinese in A.D. 465. It tells how the Buddha by his skill in persuasion converted to the faith the maiden Prakrti, who had con- ceived a deep love for the beloved disciple Ananda and would have won him from his vows, had he not at the moment of his greatest

1 Ed. J. S. Speyer, BB. 3, 1902-9; trans. L. Feer, AMG 18, 1891.

2 Ed. E. B. Covvell and K. A. Neil, Cambridge, 1886. sw F 66 AÇVAGHOŞA AND EARLY BUDDHIST KĀVYA danger sought refuge in his master's strength. The gem of the collection is doubtless the pathetic legend of Kunāla, son of Açoka, whose false stepmother succeeds in poisoning his father's mind against him and in having him blinded without his per- mitting himself either hate or reproach. We find, however, also a still more gruesome and to us repellent theme in the tale of Rūpavati, who severs her own breasts in order to feed a hungry mother when on the point of eating her own child; Rūpavati is extolled as a pattern of the Bodhisattva who seeks to save the whole world, and is accorded the somewhat quaint honour of being reborn as a prince, Rūpavata. The style of the book is very uneven, as a result of the diversity of its sources. Besides ordinary simple Sanskrit prose, intermingled here and there with Gathās, we find here and there passages in elaborate metres and prose with the long compounds approved by writers on poetics. Thus Avadana xxxviii is a version in elaborate style of the story of Maitrakanyaka in the form found in the Avadānaçataka. More interesting to us is the preservation, as part of the cycle of legends of Açoka (xxvi-xxix), of the dramatic episode of the conversion of the demon Māra by the virtuous Upagupta. The idea, ingenious in itself, is carried out with spirit and imagination; Māra is converted and Upagupta, who desires to see with his eyes the Buddha long since dead, asks him to appear before him in the Buddha's form. Māra obeys, and the devotee falls down in worship before the wondrous apparition of the master he loved. We can recognize here, without question, borrowing from Açvaghosa in manner, as in substance from the Sutralam- kāra; style and metre are of the classical type which his poems display. Moreover, we can trace 2 in this section of the work clear instances of knowledge of the Buddhacarita and even of the less popular Saundarananda; thus Gupta's son is described as beautiful beyond men but yet inferior to the gods (atikranto mānuṣavarṇam asamprāptaç ca divyavarṇam), and this some- what clumsy expression can hardly be derived from any source other than Açvaghoşa's elegant atītya martyan anupetya devan. 1 The original Açokāvadāna, according to Przyluski, La légende de l'empereur Açoka (1923), was composed by a monk of Mathurã about two centuries before Kanişka (between 150-100 B.C.). 2 Gawroński, Studies about the Sansk. Buddh. Lit, PP 49 ff. THE AVADĀNAS 67 Similarly, both xxii and xxxviii contain reminiscences of the Buddhacarita both in the polish of their style and in actual verbal similarities; in the latter we have: trsnanilaiḥ çokaçikhapracandaiç: cittani dagdhani bahu- prakāram açavatām sapraṇayābhirāmair: dānāmbuṣekaiḥ çamayām- babhūva. 'The flames of desire, kindled by sorrow, in the minds of those full of longing were extinguished by the torrents of his gene- rosity, made beautiful by his courtesy.' In the less polished parts of the collection we find many curious specimens of the influence of Pāli or Prākrit on the writers. Thus we have forms like sarpi for sarpis, parval for parva, yam for yat, tāvanta for tāvant, pithi for vithi. The use of particles often deviates from Sanskrit practice: thus api... api serves as equivalent to et... et; apy eva means perhaps, prag eva often, yavat quippe; the favourite Buddhist form of denoting place, yena... tena, is common; and yataḥ, yadbhu- yasă, tatprathamataḥ, and yat khalu are common as conjunc- tions. As prepositions we find sarvänte, after, sakāmam, to please, sthāpayitva, except. Rare words and meanings abound, as apatti, sin, kola, raft, gulma, custom-house, uddhava, cheer- fulness, paribhas, abuse, niçritya, going to, pragharati, ooze forth (prakṣar-), vyatisarayati kathām, converse, anyatara, anyatama, any one, bhūyasya mātrayā, still more. 4. Ārya Çūra and later Poetry The influence of Açvaghoṣa is unquestionably to be traced in the elegant and interesting collection of lectures or sermons in the form of edifying anecdotes of the Buddha's action in former births produced by Ārya Çūra under the style of Fatakamālā.³ The mere fact that the tales appear in Sanskrit of the Kavya ¹ The Vedic ghy may be the origin of this formation, if it is not itself a Prākrit- ism; cf. Geiger, Pâli, p. 67. 2 Ed. H. Kern, HOS, i, 1891; trans. J. S. Speyer, London, 1895. Cf. Luders, GN. 1902, pp. 758 ff.; F. W. Thomas, Album Kern, pp. 405 ff.; on the Chinese ver- sion, Ivanovski, RHR. xlvii. 298 ff.; cf. E. Wohlgemuth, Über die chinesische Version von Afvaghoşa's Buddhacarita (Leipzig, 1916). F 2 68 AÇVAGHOŞA AND EARLY BUDDHIST KĀVYA type is sufficient proof of the spread of the use of that language for purposes of literature and discussion in the courtly circles in which, we may safely assume, Arya Çūra moved and lived. The material of the tales was doubtless ready to hand; nearly all of them are extant in the Pali Jataka book,¹ and twelve of them are also found in the Pali Cariyapitaka. Moreover, as in that book, the tales are told with the definite purpose of illustrating the various perfections (pāramitās) ascribed by Buddhist theory to the Buddha to be. Their chief defect to modern taste is the extravagance which refuses to recognize the Aristotelian mean. The very first tale, which is not in the Jataka book, tells of the extraordinary benevolence of the Bodhisattva who insists on sacrificing his life in order to feed a hungry tigress, whom he finds on the point of devouring the young whom she can no longer feed, and the other narratives are no less inhuman in the disproportion between the worth of the object sacrificed and that for whose sake the sacrifice is made. But these defects were deemed rather merits by contemporary and later taste. I-tsing mentions the Fatakamālā as one of the popular works among Buddhists of his day, and the frescoes of Ajaṇṭā include both pictures and verses, proving the existence then of the text. The date of this evidence, unfortunately, is not certain, but the style of writing suggests the sixth century, and with this accords the fact that a Chinese rendering of another work of Arya Çūra was made in A. D. 434. The author may then have written in the third, or more probably the fourth, century. Ārya Çūra's style is classical, showing command of the resources of his art, but restrained and saved from exaggeration by good taste. His prose and verse alike are careful and polished, and, though he is not averse to the use of fairly long compounds, especially in prose, he employs them naturally and is seldom obscure. His good taste is conspicuous in the lines put in the mouth of the son whose father in his insensate generosity has given away his wife and children; the child speaks in simple but pathetic words: naivedam me tathā duḥkham yad ayam hanti māṁ dvijaḥ napaçyam ambāṁ yat tv adya tad vidārayatīva mām ¹ GN. 1918, pp. 464 ff. ĀRYA ÇŪRA AND LATER POETRY rodisyati ciram nūnam ambă çünye tapovane putraçokena krpaṇā hataçāveva cātakī. asmadarthe samãhṛtya vanăn mülaphalam bahu bhavisyati katham nv ambă dṛṣṭvā çünyam tapovanam? ime nav açvakās tāta hastikā rathakāç ca ye ato'rdham deyam ambayai çokam tena vineşyati. 69 ""Tis not so much that the Brahmin beats me that causes me sorrow, but that I have not seen my mother to-day pierces my heart. Long will my mother weep in the penance grove, now lonely, sorrowing for the woes of her children, like a cuckoo whose young are slain. She has gathered for our sake many a fruit and root from the forest; how then will she feel when she sees the penance grove left lonely? Here, daddy, are our toy horses, our elephants, our cars; give a half to mother; thus will she assuage her grief.' But he is equally happy in more elaborate themes, as in the description of the rule of the just king : samaprabhāvā svajane jane ca: dharmānugă tasya hi danda- nitih adharmyam avṛtya janasya mărgam: sopānamāleva divo babhūva. 'Impartial to kin and stranger alike, his rule followed in the steps of righteousness; blocking the path of unrighteousness to men, it was as a ladder to raise them to the sky.' No doubt in his language there are traces here and there of Palicisms,¹ but these do not seriously detract from Ārya Çūra's claim to correct- ness of language, and his metrical skill is considerable. The form of his tales as composed of prose with verses inter- mingled, now singly, now in larger numbers, is of historical interest. It is not, of course, an invention of Arya Çūra, who followed Kumāralāta and doubtless many others in the employ- ment of this style. But its origin is disputed. Oldenberg 2 developed with his usual skill the thesis that the original form of literature in India, as perhaps elsewhere, was prose, with verses rposed at those points where the primitive mind naturally tends to give utterance to its feelings in verse form, as when ¹ He is praised in the Saduktikarṇāmṛta, ZDMG. xxxvi. 365. For his Pālicisms. see Franke, IF. v. Anz. 31. 2 GGA. 1909, pp. 66 ff.; GN. 1911, pp. 459 ff.; 1919, pp. 79 ff. Cf. Winternitz, WZKM, xxiii. 102 ff. 70 AÇVAGHOŞA AND EARLY BUDDHIST KĀVYA a god is invoked, a curse is pronounced, a benediction uttered, a prayer put up, in short at any point where emotion is let free and the pedestrian prose is inadequate as an expression of the feeling. He has found proofs of the existence of literature of this kind in the Rgveda, the Brāhmaṇas, the epic, and in Pāli texts, including the Jatakas. In principle the verses alone were preserved in fixed form, and they only received skill and care, the prose being supplied by those who told the tales. The pro- cess of development which followed was, on the one hand, the elimination of the prose by substituting verse, and it has been suggested that a remnant of the old condition is to be found in the Mahabharata, where the speakers in case of dialogue are given in prose, while in the more finished Rāmāyaṇa such devices are unknown, the poet, like the authors of the Iliad and Odyssey, working into verse the name of the spokesman. On the other hand, the step was taken of applying to the prose the artistic polish which marked the verse, and Oldenberg ¹ claims that, apart from an exceptional case like the Kuṇāla Fataka of the Pāli Jataka book, where the verses are accompanied by an ornate prose, the Fātakamālā and the Pañcatantra or Tantra- khyāyika are among the earliest examples of this form. It seems clear for reasons elsewhere adduced that the theory is not substantiated by Vedic evidence, and that it must stand or fall according as other considerations may appear to render it credible. The evidence of comparative literature is still quite inadequate to support it, and from the Indian point of view matters can much more simply be explained. The earliest form of prose with verse intermingled which we find in Indian litera- ture appears to be that in which gnomic verse is cited to illustrate ¹ Allind. Prosa, pp. 82 ff. What is true is that elaboration of prose style is later tban and based on development of verse; cf. Jacobi, Compositum und Nebensatz, p. 93, who cites the symmetrical Varnakas of the Jain canon and their long com- pounds (cf. IS. xvii. 389 ff.). 2 Keith, JRAS. 1911, pp. 979 ff.; 1912, pp. 429 ff.; HOS. xxv. 43 ff. There are cases of intermixture of prose and verse in other languages, e. g. Latin (Varro's Saturae Menippeae, Petronius, Martianus Capella (c. A. D. 400), Boethius (480-524),` and two novels, Julius Valerius (c. 300) and Historia Apolloni Tyrii; Teuffel- Schwabe, Rom Lit., §§ 28, 165, 305, 399, 452, 478, and 489); Norse; Mediaeval Irish (Windisch, Irische Texte, ii. 447 ff.); Chinese; Old Picard, Aucassin et Nico- lett; Boccaccio's L'Ameto; Sa'di's Gulistan; Basutos and Eskimos (MacCulloch, Childhood of Fiction, pp. 480 ff.); Gray, Vasavadattă, p. 32. ĀRYA ÇŪRA AND LATER POETRY 71 what is stated in the prose; this is akin to the practice of the Brāhmaṇas to adduce occasionally Yajñagāthās, verses on sacri- ficial points, in their discussions, and to the habit of the Dharma- sūtras to enforce the rules which they lay down with verse cita- tions. Here and there in the Upanişads we find similar cases, verses being cited in illustration and explanation of a doctrine stated in prose; in these cases it is made quite clear that the verses are quotations, from which, no doubt, it was an easy step to the writer composing verses of his own to enliven his theme or summarize his moral. The Kārikās found in the Mahābhāṣya prove that grammarians recognized the convenience of thus putting on record in easily remembered and accurate form their observations on disputed points. In the case of narrative the evidence seems clearly to indicate that originally in India prose and verse were used independently; if so, it is easy to understand how they could come to be combined, especially as in the other instances adduced above there already existed examples of the combination of verse and prose in one literary form. The few cases in the epic of prose and verse combined seem to be dis- tinctly instances of contamination, not remnants of an older form of composition. How far models in Pāli were available for the author of the Fatakamālā or Kumāralāta we cannot, of course, prove, for the Jataka book in Pāli as we have it presents grave problems which are yet unsolved. But the Kunāla Fataka at any rate suggests that it would be unwise to claim that the transition first took place in Sanskrit versions of Jataka tales. Other Buddhist writeis contributed much less to literature than to philosophy. The mysterious Nāgārjuna, perhaps of the latter part of the second century A.D., in his Madhyamakakārikās shows a perverse ability to develop paradoxes, while Ārya Deva (c. A.D. 250) in his Catuḥçatika¹ shows considerable power of irony in his onslaught on the Brahmanical practice of bathing in the Ganges to remove sin and acquire merit. The Çişyalekha- dharmakavya 2 of Candragomin, in which instruction is given in the form of a letter to a pupil dealing with the essential facts of the . 1 Ed. Calcutta, 1914. On his Hastavalaprakaranavytti, cf. Thomas and Ui, JRAS. 1918, pp. 267 ff. Cf. P. L. Vaidya, Études sur Aryadeva (Paris, 1923). 2 Ed. I. P. Minayeff, Zapiski, iv. 72 AÇVAGHOȘA AND EARLY BUDDHIST KĀVYA Buddhist faith, has a predecessor in the Suhṛllekha¹ of Nāgār- juna, in which he summarizes Buddhist doctrine for a king, unhappily unidentified. The Subhāṣitāvali cites a verse actually found in the letter, though omitted in the Tibetan version: visasya viṣayāṇāṁ ca dūram atyantam antaram upabhuktam visam hanti viṣayāḥ smaraṇād api. 'Vast indeed the difference between poison and objects of sense; poison slays only when tasted, but the things of sense by mere thought thereof.' The name of the author is given in the text as Candragopin, but on the whole it is improbable that he is to be distinguished from Candragomin, and we may place him in the seventh century A.D., as his grammar was used in the Kāçikā Vrtti, while he seems to have been alive as late as the time of I-tsing, though his reference is not free from doubt. As might be expected from a grammarian, the poem is written in correct and fluent Sanskrit, but without special distinction. The case is other with Çantideva, author of the laborious com- pendium of Buddhist dogmatics of the Mahāyāna school, the Çikṣāsamuccaya, in his Bodhicaryavatāra, in which he sketches the career of him who seeks to attain Buddhahood as opposed to the narrow Hinayāna ideal of saintship. Çantideva, who lived in the seventh century and whom tradition alleges to have been the son of a king who was induced by the goddess Tārā to lay aside royal state, disclaims any literary pretension; he writes for himself only and for those of nature akin to his. His poem is a strange blend of passionate devotion to the aim of aiding men to achieve freedom from the miseries of life coupled with the utter negativism of the Mahāyāna philosophy. There is nothing real, nothing can be gained or lost, none honoured or despised; joy and sorrow, love and hate, all are idle names, without reality; search as you will, nothing can be found that is. None the less Çantideva seems to be intoxicated with the nobility of the aim of seeking to be a saviour of mankind; the good we do in our efforts is a joy to the Buddhas and Bodhisattvas; we are allied with them in the struggle to attain the end. It is a delusion by 1 Trans. H. Wenzel, JPTS. 1886, pp. 1 ff.; for the king Satavahana, cf. Vidyabhu- sana, POCP. 1919, i. 125. 3īd. de la Vallée Poussin, BI. 1901 ff.; trans. Paris, 1907. 8 ĀRYA ÇŪRA AND LATER POETRY 73 which we treat our own bodies as something essentially our own; we must realise that the grief of another is our own, the joy of another not alien to us. The poetic power of the author stands out brilliantly when contrasted with the uninspired verses in which his predecessors Vasubandhu and his brother Asañga, probably in the fourth century, preached their doctrines. Of the latter we have the Mahāyānasūtrālamkāra, written in correct but undistinguished Sanskrit, utterly overloaded with technicalities, and, despite its great length and the obvious efforts of the author to express himself effectively, deplorably obscure. But the poem is of literary interest as proving how fully Buddhist teachers had adopted Sanskrit as their literary medium. IV KĀLIDĀSA AND THE GUPTAS 1. The Guptas and the Brahmin Revival UTTER obscurity attends the decline of the power of the followers of Kanişka in India, but it is certain that in A.D. 320 Candragupta founded, as a result of a matrimonial alliance with a Licchavi princess, a dynasty with head-quarters at Pāṭaliputra, which under his son, Samudragupta (c. A.D. 330-75), stood out as the paramount power in northern India, while his grandson, Candragupta II, completed its success by overthrowing the Kṣatrapas and adding Malwā, Gujarāt, and Kāṭhiāwār to the empire. His son and successor, Kumāragupta (A.D. 413-55), seems to have reigned in unbroken prosperity, and Skandagupta, his son, shortly after his reign began, won a decided success over the Hūna invaders who were advancing from the north-west and menacing India. But between A.D. 465 and 470 the Hūna advance seems to have become irresistible, and at any rate after the death of Skandagupta about 480 the greatness of the empire was irretrievably departed, though the dynasty con- tinued to rule sadly diminished dominions for several genera- tions. By 499 Toramāṇa, leader of the Hūņas, was established as ruler of Malwā, while his successor, Mihiragula, had his capital at Sialkōt in the Panjab. The expulsion of the Huns seems to have been the result about 528 of a victory won by Yaço- dharman, a ruler of central India, and the Gupta Balāditya of Magadha, but the records are curiously unsatisfactory. At any rate Mihiragula retreated to Kashmir, where he won an unenvi- able reputation,2 and shortly after 550 the Turks conquered the Hun kingdom on the Oxus. There can be no doubt that the Gupta empire signified a distinct revival of Brahmanism and a reassertion of Indian ¹ Smith, EHI. chaps. x and xi; Bhandarkar, Early History of India, pp. 47 ff. 2 To him is ascribed the ruin of Gandhara and its art; Foucher, L'Art Gréco- Bouddhique, i1. 588 ff. THE GUPTAS AND THE BRAHMIN REVIVAL 75 nationality as opposed to the somewhat cosmopolitan Kushan régime, under which Buddhism was decidedly in chief favour, though Brahmanism and Jainism must have been widespread. The art of the period is of a high order, reflecting a national spirit reacting to the impulse of Greek inspiration, although the architecture of the period has largely disappeared, owing doubt- less to the appalling destruction wrought by the Mahomedan invaders of north India. The sculpture, however, exhibits an unusual beauty of figure, dignity of pose, and restraint and refinement of treatment in detail. The coinage, often of merit, shows clear traces of intercourse with the Roman world, also attested by records of missions to Rome and Constantinople in 361 and 530. Mathematics, astronomy, and astrology flourished, taking new life under Greek influence, as is abundantly esta- blished by the Pancasiddhantikā of Varahamihira (c. 550) and by the works of Aryabhaṭa (born 476). Relations with China were maintained by visits of Buddhists from and to India. Fa-hien (401-10) gives us a most favourable picture of India under Candragupta II. There was freedom of movement throughout mid-India; justice was dispensed with mercy, fines being normally inflicted, capital punishment being disused, and mutilation restricted to rebels or brigands; the revenues of the crown were derived mainly from land, and the royal officers and servants received regular salaries. Among Buddhists at least- and they still were very numerous-the rule of refraining from animal food or taking life was widely observed, and in many places butchers' shops and distilleries were unknown. What is of special interest is that he alone records a very significant proof of the revival of Brahmanism; the Caṇḍālas or outcasts were obliged to live apart, and, when they approached a town or bazaar, to strike a piece of wood as a warning of their presence, in order that others might avoid pollution by contact with them. The emperors were clearly devotees of Vişnu and attached to the Bhagavata faith, but religious toleration was still the order of the day, and the signs of the decadence of Buddhism were con- cealed from Fa-hien's eyes. Nor is this surprising, for it is probable that Samudragupta himself was a friend of Vasubandhu when that Buddhist sage attended his father's court.2 Samudra- 1 Foucher, 11. 756 ff. 2 Cf. Vamana's evidence; Smith, EHI. pp. 346 ff. 76 KĀLIDĀSA AND THE GUPTAS gupta, however, was careful to assert his devotion to Brahmanical ideals; thus he renewed the ancient horse sacrifice as a sign of his paramount sway, and Kumāragupta appears to have followed his example. The centre of Gupta power, originally fixed at Pāṭaliputra, seems clearly to have shifted during the reign of Candragupta II to Ujjayini, doubtless in order to secure the stead- fast adherence to the empire of the newly acquired lands. That such princes should favour poetry and fine arts was inevitable. Samudragupta was proud of his skill with the lute, and a coin depicts him playing that instrument. But a more secure support for his claims is afforded by the assertions of the panegyrist Harişeṇa (c. 350), who assures us that his patron had a poetic style which was worth study and wrote poems which in- creased the poet's spiritual treasure, and again that his title of king of poets, Kavirāja, was well grounded through his composi- tion of many poems deserving imitation by others. He delighted also in the society of the earnest students of literature, was inter- ested in the explanation and defence of holy scripture, and de- voted to music. Moreover, he won fame by removing the dis- crepancy between the poet's art and riches, doubtless his chief merit in the eyes of many of his flatterers. Of his great son Candragupta we know that he adopted the title Vikramaditya, reminiscent of the legendary Vikramāditya of Ujjayini, and it is certainly plausible to suggest that the fame of Vikramāditya as the patron of poets, attested in the late and in itself worthless legend of the Nine Jewels,¹ was due to the literary distinction of Candragupta's court. The list of Jewels runs Dhanvantari, Kṣapaṇaka, Amarasinha, Çañku, Vetāla Bhaṭṭa, Ghaṭakarpara, Kālidāsa, Varahamihira, and Vararuci. Of these Dhanvantari, as the author of a medical glossary, is older than Amarasinha, who also used Kalidasa; the fourth and fifth are mere names; Varā- hamihira definitely lived in the sixth century, and the dates of Kṣapaṇaka as a lexicographer and of Vararuci are unknown. But we have a distinct corroboration of the idea of Candragupta as a patron of poets in the fact that his minister of external affairs, Virasena Kautsa Çaba, was interested in poetry. Probably the succeeding emperors manifested equal' concern in poetry. ¹ Weber, ZDMG. xxii. 708 ff.; Zachariae, Die indischen Worterbucher, pp. 18 ff.; Fleet, IA. xxx. 3 1. THE GUPTAS AND THE BRAHMIN REVIVAL Nor is there any doubt that the drama must have flourished under their patronage; indeed it has been suggested that Can- dragupta's epithet rupakṛtin denotes maker of plays, which would make the king a predecessor of Harṣa as a dramatist; the accuracy of the rendering is not, however, beyond cavil. What, however, is certain is that Sanskrit was essentially the language of the court and of learned men; even Buddhists such as Vasubandhu and Asañga resorted to it as a matter of course as the means of securing a respectful hearing for their doctrines. The disputes between the rival schools were probably friendly enough; the Samkhya philosophy as expounded in the Karikā of Içvarakṛṣṇa seems to have been the Ebject of special attack by Vasubandhu, and Samudragupta's interest in these matters may have been aroused by that teacher. 77 2. Harişena and Vatsabhatti Fortune has enabled us to obtain an interesting insight into the poetry of the Gupta epoch by the preservation of two Pra- çastis, separated by about a century in time, the panegyric of Samudragupta inscribed on a pillar at Allahabad and composed by Hariṣena, perhaps in 345,¹ and Vatsabhaṭṭi's inscription in the temple of the sun at Mandasor, written in 473-4. These inscrip- tions alone would suffice to prove abundantly the existence of a developed Kavya poetry during the whole period of the Gupta power, and in the first case we actually find a poet of distinct power, though he was foreign minister and general of the king. Harişena's poem bears expressly the title Kävya, though it consists both of prose and verse. Its structure is similar to the delineation of kings adopted in the prose romances of Subandhu and Bāņa, in which all is crowded into a single long sentence, made up of relative clauses and adjectives and appositions heaped upon one another. In this case the whole poem is one sentence, including first eight stanzas of poetry, then a long prose sentence, and finally a concluding stanza. The thought is no less complex than the form, for the poet's ingenuity has been equal to the effort to connect the pillar with the emperor's fame. That, as ¹ Cf. Gawroński, Festschrift Windisch, pp. 170 ff.; The Digvijaya of Raghu (1915); Bühler, Die indischen Inschriften (1890); Smith, EHI. pp. 298 ff. 78 KĀLIDASA AND THE GUPTAS usual in the Kavya, is personified as feminine and is regarded as having embraced the whole world so that no more room for remains on earth. It passes therefore by the way of the pillar up to the abode of the gods. There it appears as the Ganges, and, pure as that stream, it overflows on heaven, atmosphere, and earth. The metre is no less elaborate than the thought; of seven verses preserved there are four metres, Sragdharā, Çārdūlavikri- dita, Mandākrāntā, and Pṛthvī. The style is markedly and un- deniably of the Vaidarbha or southern manner; the verse eschews long compounds while the prose delights in them, one having no less than 120 syllables, though it is but fair to say that on the whole they are not difficult to understand. Of figures of sound alliteration is used, but sparingly; metaphors are most used of the figures of sense, rarely similes and double entendres as in Samudragupta's epithet sadhvasādhūdayapralayahetupuruṣasya- cintyasya, 'a hero unfathomable, the cause of the elevation of the good and the destruction of the bad (and thus a counterpart of the unfathomable absolute, which is the cause of the origin and the destruction of the world, and in which good and bad have their being)'. But Hariṣeņa spares us much of this; he shows his skill rather by new turns of ingenious thought, and by the care with which his long compounds are relieved by the inter- position of short words to give the reciter time to recover breath and the hearer to understand the sense, and by the cunning arrangement of words in the compounds themselves in order to produce the maximum of metrical effect. His choice of words and care in their arrangement are no less seen in his verses, of which one certainly has the right to be ranked as among the most perfect effects of Indian miniature word pictures, the description of the scene when before his rivals and the court Candragupta in his old age designated Samudragupta as his successor : aryo hity upaguhya bhāvapiçunair utkarṇitai romabhiḥ sabhyeşucchvasiteşṣu tulyakulajamlānānanodvīkṣitaḥ snehavyālulitena bāṣpaguruṇā tattvekṣiṇā cakṣuṣā yaḥ piträbhihito nirīkṣya nikhilām pāhy evam urvīm iti. ""He is noble", with these words he embraced him, tremors of joy betraying his emotion; he gazed on him with tear-filled eyes, following his every movement, and weighing his worth-the 79 HARIŞEŅA AND VATSABHAȚȚI courtiers sighed in relief and gloomy were the faces of his kins- folk and said to him, " Do thou protect all this earth"." Very different is the work of Vatsabhaṭṭi,' no minister of an emperor but a humble local poet, glad to earn a fee by writing for the guild of silk-weavers of a provincial town. What is inter- esting in him is his testimony to the prevalence of the Kavya in his time; the adjective purvā, above, is used as sufficient descrip- tion of his poem, the missing praçasti, eulogy, being so naturally supplied by those familiar with current verse. He asserts that his work was done with effort or care (yatnena), and there is every evidence of the truth. In obedience to the laws of poetics he inserts in his forty-four stanzas descriptions both of Lāṭa and of the town Daçapura, of the seasons, winter and spring, and shows by the use of twelve metres his skill in versification, though the effect is marred by his inability to bring off his results without free use of the weak caesura. His style is the eastern or Gauda, as is clearly proved by his love of long compounds in verse, and by the way in which in one stanza he has fitted the sound of the verses to the altering sentiment, advancing from soft harmonious sounds in describing the gentleness of his hero to discords when proclaiming him dvitdrptapakṣakṣapaṇaikadakṣaḥ, 'peerless in destroying the proud hosts of the foe'. His alliterations, similes, and metaphors all are of types abundant in the Kavya, but his skill is small, and his poem is disfigured by tautologies as in tulyopamānāni, the use of verse-fillers or needless particles as in tatas tu, or prefixes as in abhivibhāti, or words as in samudrānta, while sprçanniva for the necessary neuter and nyavasanta are offences against grammar. But his panegyric is invaluable testimony to the widespread cultivation of Sanskrit poetry and it helps definitely to aid us in determining the date of India's greatest poet. 3. Kalidasa's Life We know nothing whatever of value from later sources re- garding the life and character of Kalidasa.2 Anecdotes are told 1 Buhler, Die indischen Inschriften, pp. 31 ff. 2 On his date see Liebich, IF. xxxi. 198 ff.; Keith, Sanskrit Drama, pp. 143 ff.; Hillebrandt, Kalidāsa (1921). S. Ray (POCP. 1919, i, p. lix) held him to be Agnimitra's court poet (c. 150 B. c.), but K. G. Sankar (IHQ. i. 309 ff.) puts him between 75 and 25 B. C. KĀLIDĀSA AND THE GUPTAS asserting that he was originally extremely stupid, and won skill in poetry by the favour of Kālī, an obvious deduction from his name, slave of Kāli. He is alleged also to have shown remark- able skill in the ready manufacture of verses to order, either to describe a given situation or to complete an imperfect stanza, and a more circumstantial legend¹ tells of his murder in Ceylon while a guest of King Kumāradāsa at the hands of a greedy hetaira. There is not the slightest ground to accept the sugges- tion, still less to find in it an indication of date, Kālidāsa's visit to Ceylon on this view being due to the Hun inroads. His own poems, on the other hand, and especially the description of Raghu's conquests, prove him intimately acquainted with many Indian scenes, the sandal of Kashmir, the pearl fisheries of the Tamraparņi, the deodars of the Himalayas, the betel and coco-palms of Kalinga, the sand of the Indus, but it would be hazardous to claim for him any part in the great expedition of Samudragupta when he won his right to perform the horse sacrifice as a sign of his paramount power in India. Nonetheless it is difficult to dissociate Kalidasa from the great moments of the Gupta power. He was later than Açvaghosa and than the dramatist Bhāsa; he knew Greek terms, as his use of jamitra proves, the Präkrit of his dramas is decidedly later than Açvaghoṣa's and Bhasa's, and he cannot be put before the Gupta age. His complete acceptance of the Brahmanical system, the sense of sharing in a world of prosperity and power, the mention of the horse sacrifice in the Malavikägnimitra, Raghu's conquests in the Raghuvança, seem best explicable as the out- come of the enjoyment of the protection of a great Gupta ruler, and we must remember that Candragupta II had the style of Vikramaditya, with whose name tradition consistently connects Kalidasa. Nor is it absurd to see in the title Kumarasambhava a hint at the young Kumāragupta, the heir apparent, or even in Vikramorvaçī an allusion to the title Vikramaditya. It has been attempted to refer Kālidāsa to the sixth century by making the Vikramaditya of tradition the Yaçodharman who defeated the 80 1 Geiger, Lit. und Sprache der Singhalesen, pp. 3 ff.; Rhys-Davids, JRAS. 188⁹ pp. 148 ff.; Bendall, p. 440; Nandargıkar, Kumāradāsa, pp. v ff.; Vidyabhusana, POCP. 1919, i, p. clxxii. 2 Hoernle, JRAS. 1909, pp. 89 ff. KĀLIDĀSA'S LIFE 81 Huns, but this theory is no longer in repute. More favour ¹ has been shown to the view that Kalidasa lived under Kumāragupta and Skandagupta, mainly on the score that Mallinātha and Dakṣiṇāvartanātha ascribe to him in v. 14 of the Meghaduta a double entendre referring to Dignāga, the Buddhist logician, as a hostile critic, and that his own reference to the Hūnas and the river Vankṣū in the Raghuvança alludes to the time when these warriors were still in the Oxus valley just before their defeat by Skandagupta. The first argument is invalidated by the grave improbability of the tasteless reference in the Meghaduta and by the fact that, even if it were real, Dignāga's date need not be later than 400. The second imputes to Kalidasa a desire to achieve historic realism quite out of keeping with his poetic aim, and irreconcilable with his mention of the Greeks as on the north-west frontier as well as the Pärasīkas, Kambojas and Hūnas. That Kālidāsa lived to see the Hūņa victories is most implausible, while his evident affection for Ujjayini suggests that he spent much of his time there under Candragupta's favour. This conclusion is strongly supported by evidence culled from Vatsabhaṭṭi. Two of his verses run: calatpatākāny abalasanathany: atyarthaçuklany adhikonna- tāni tadillatăcitrasitabhrakūṭa-: tulyopamānāni grhāni yatra. Kailasatungaçikharapratimāni cānyāny: ābhānti dirghava- labhini savedikāni gandharvaçabdamukharāṇi niviṣṭacitra-: karmani lolakada- līvanaçobhitāni. 'The houses there, dazzling white and towering high, with their waving banners and tender maidens, are rivals of the cloud- pinnacles, snow-white, but stained by the lightning-creeper. Yet others match Kailasa's lofty peaks, with their long balconies and seats of stone, as they resound with music, are decked with pictures, and are adorned with groves of waving plantains.' 1 Gawroński, The Digvijaya of Raghu, pp. 1 ff.; Smith, EHI. p. 321, n. I. The term found in the epic was perhaps first used of the Hiung-nu of the 2nd cent. B. C. 3149 G 82 KĀLIDĀSA AND THE GUPTAS These stanzas can hardly be deemed other than an attempt to improve on v. 65 of the Meghadūta: vidyutvantam lalitavanitāḥ sendracāpam sacitrāḥ samgitaya prahatamurajāḥ snigdhagambhiraghoṣam antastoyam manimayabhuvas tuñgam abhraṁlihāgrāḥ prāsādās tvām tulayitum alam yatra tais tair viçeşaih. 'There the palaces can vie with thee at every point: their fair maidens rival thy lightning, their paintings thy rainbow, their drums beaten in concert thy lovely deep thundering, their jewelled floors thy water, their peaks that touch the sky thy height.' To suppose that Kālidāsa knew these clumsy verses of an obscure poetaster and turned them into the simple elegance of his verse is absurd; to hold that a local poet appropriated and tried to improve on a verse of the great poet of Ujjayini is natural and simple, and, if confirmation were needed, it is supplied by the fact that v. 31 of the inscription deals simi- larly with vv. 2 and 3 of Canto v of the Rtusamhara. Kalidāsa then lived before A. D. 472, and probably at a considerable dis- tance, so that to place him about A.D. 400 seems completely justified.² 4. The Rtusamhāra 3 The opinion of India which makes the Rtusamhāra, cycle of the seasons, a youthful work of Kālidāsa, has recently ³ been assailed on many grounds. Thus it has been complained that the poem lacks Kālidāsa's ethical quality, that it is too simple and uniform, too easy to understand. The obvious reply is that there is all the difference between the youth and the maturity of a poet, that there is as much discrepancy between the youthful work of Virgil, Ovid, Tennyson, or Goethe, and the poems of their manhood as between Kālidāsa's primitiae and the rest 1 Kielhorn, GN. 1890, pp. 251 ff. 2 On the later emperors, see R. C. Majumdar, JPASB. 1921, pp. 249 ff. 3 Walter, Indica, iii, 6 ff.; Nobel, ZDMG. lxvi. 275 ff.; JRAS. 1913, pp. 401 ff.; Hari Chand, Kalidasa, pp. 240 ff. Contra Keith, JRAS. 1912, pp. 1066 ff.; 1913, pp. 410 ff.; Hillebrandt, Kalıdäsa, pp. 66 ff. Kielhorn, Buhler, Hultzsch, Mac- donell, von Schroeder, among others, accept Kalidasa's authorship; often ed., e.g. Gajendragadkar, 1916. 83 THE RTUSAMHARA of his work. Nor is it the slightest use to argue that Sanskrit poets differed from other poets since they were essentially learned and artificial; the poets mentioned are precisely of the analogous type, men who worked steadily at their art until at their prime they could create structures which make their youthful attempts seem childish folly. In point of fact the Ṛtusamhāra is far from unworthy of Kālidāsa, and, if the poem were denied him, his reputation would suffer real loss. The contention that Mallinātha commented on the other three of his poems but not on this is met effectively by the consideration that its simplicity rendered it poor game for the very learned commentator to deal with. The fact that the writers on poetics do not cite from the poem has an obvious explanation in the same fact; these authors never exhibit the slightest trace of liking what is simple, and they could find in the later poems abundant material to use as illustration. More deplorable still are some of the æsthetical arguments adduced; complaint is made that the poet begins with the summer, whereas the spring was the usual beginning of the year, forgetting that Kālidāsa was not composing an almanac or writing a Shepheard's Calendar. Again, heat or its derivatives (tap) is found seven times in Canto i, as if this did not accord with summer, as does eagerness (samutsukatva) with the rains and longing (utkanth) with autumn. The poet is censured for asserting that the swans excel maidens in beauty of gait and the branches rob their arms of loveliness; later, he was not guilty of such discourtesy. He mixes a metaphor in speaking of clouds as having the lightning as creeper; as we have seen, Vatsabhaṭṭi borrows the phrase, and exploits two other verses of the poem, proving its antiquity and rendering most probable its authorship. It is objected that he uses here only the construction à mulataḥ, in lieu of the ablative, though equally once only in the Kumāra- sambhava he has āmekhalam; the freshness and liveliness of the seven verbal forms (ii. 19) is unparalleled and, therefore, not by Kalidāsa. Even the lack of developed use of figures of speech is adduced against him, and the use of samhara in the title has been questioned as unique. Poets happily do not feel themselves bound to be parrots.¹ ¹ His developed style is seen in his pictures of spring (Kumāras. iii ; Ragh. ix), and summer (Ragh. xvi). G 2 KĀLIDĀSA AND THE GUPTAS The poem is far from a mere description of the seasons in their outward aspect, though Kālidāsa exhibits delicate observa- tion and that loving sympathy with nature which seems innate in Indian poets. Throughout he insists on the relation of the diverse moods of the year to the loves of man and maiden or husband and wife. Though the days of summer are a burden, the nights are the more delightful, when the moon is bright and coolness refreshes the earth; at midnight the young delight in song and dance and wine; the moon in jealousy of youthful love retires in sorrow. The rainy season comes in kingly guise, the clouds the elephants which bear him, the lightning his standard, the thunder his drum. The emotion of love is awakened by the sight of the clouds which bend down to kiss the peaks of the mountains. Autumn comes like a young bride, clad in a garment of sugar cane, girdled with ripening rice, and with face of lotus blooms. Winter's cold makes all the more welcome, all the more close and tender, the embraces of lovers. In the cool season the nights are cold, the moon shines chill, the lovers close the window of their chamber, wrap themselves warmly in their garments, and enjoy every moment of the still feeble rays of the sun, or rest beside the fire. But spring brings to them and to all nature new life and joy; we see now why the poet begins with summer; it enables him to end with the season in which young love, in harmony with the birth of a new year, is made perfect.. The poem in every line reveals youth; the lack of the ethic touch¹ is in perfect accord with the outlook of the young, and though Kālidāsa was to write much finer poetry, he was also to lose that perfect lucidity which is one of the charms of the poem to modern taste, even if it did not appeal to writers on poetics. 84 . 5. The Meghaduta In distinction to the Rtusamhäāra the Meghadūta² is un- questionably a work of Kālidāsa's maturity; the mere fact that he adopts for it and maintains throughout with only occasional ¹ Stenzler, ZDMG. xliv. 33, n. 3.

• Ed. E. Hultzsch, London, 1911 (with Vallabhadeva's comm.); ed. and trans.

Pathak, Poona, 1916; ed. TSS. 54, 1919. THE MEGHADUTA 85 harshness a metre so elaborate as the Mandākrāntā is conclusive proof that he was no novice, though we may admit the possibility that he desired by this metrical tour de force to establish his capacity once and for all, and to exhibit himself as a great poet. Suggestions for the subject-matter may have been taken from the Rāmāyaṇa,' where Rāma's deep longing for his lost Sītā offers an obvious prototype for the Yakṣa's sorrow for the wife from whom he is severed, and the description of the rainy season in iv. 28 has some points of similarity. But the idea is carried out with marked originality and beauty. A Yakṣa banished for a year by Çiva his master, because of failure of duty, is reminded by the approach of the rainy season of his wife, lamenting him in their abode at Alakã, and begs a passing cloud to bear to his beloved the news of his welfare and the assurance of his devotion. From Rāmagiri, his place of exile, the cloud is bidden go, in the company of the cranes and the royal swans en route for Lake Mānasa, to the region of Māla and to mount Amrakūṭa. Thereafter it is to seek the Daçārṇa country with its city of Vidiçã, and then must drink the waters of the Vetravati before proceeding to visit Ujjayini, after crossing the Nirvindhyā and the Sindhu. The shrine of Mahākāla must be visited, the Carmaṇvati crossed, and the holy Brahma-arta after passing Daçapura; there the cloud will visit the field of Kuruksetra, the scene of Arjuna's great deeds, and drink the water of the Sarasvati, for which Balarama, who fought not for love of his kin, abandoned his beloved wine. Thence it must go to where the Ganges descends from the Himalaya near mount Kanakhala, and then to Kailasa, passing through the gap of mount Krauñca which Paraçurāma made as a path to the south. Then the water of lake Mānasa will refresh the cloud, and on the top of the mountain is Alakã where the beloved of the Yakṣa dwells. The delights of the divine city are fully depicted, and the poet then describes to the cloud the home he is to seek out; it can be seen from afar off through its archway; in the garden is a coral trec, its mistress's pet, and a flight of emerald steps leads to a well in which golden lotuses grow, and the swans, delighted, think no more even of their beloved Manasa. There is the beloved, sorrowful, and blighted by separation, emaciated, 1 There is in the Kämavilāpa Jataka (ii. 443) a very distant parallel. 86 KĀLIDĀSA AND THE GUPTAS seeking by many a device to while away the long days until her husband's return. Gently she is to be wakened from her slumber by the cloud, which is to give her a message of tender love from her husband, and an assurance of his faith and certainty of reunion.. .At first sight the effect of the poem seems to be marred by an element of unreality in the longing of the Yakṣa, whose separa- tion is but temporary and who as an attendant of Çiva cannot in truth fear either death or even injury for his beloved from his absence. The message would have read very differently had it been sent, as in Schiller's Maria Stuart, by a helpless captive awaiting in resignation or despair an ineluctable doom. But to understand the poem aright we must remember that the poet doubtless felt that it was, as later writers expressly allege, the. duty of the poet to suggest rather than to say outright; the loves of the two immortals is a symbol of human love; perhaps ¹ Kālidāsa had some experience of his own which the poem indicates, for the vivid colours in which he describes the Yakṣa's abode seem to be drawn from real life. Certainty is wholly unattainable, but in any event it is difficult to praise too highly either the brilliance of the description of the cloud's progress or the pathos of the picture of the wife sorrowful and alone.. Indian criticism has ranked it highest among Kalidasa's poems for brevity of expression, richness of content, and power to elicit sentiment, and the praise is not undeserved. 1 Popularity has had the penalty of many interpolations of the text. There is a remarkable mass of evidence available; in the eighth century Jinasena, applying the art of Samasyāpūraṇa, worked the whole of the text of 120 verses as he knew it into an account of the life of the Jaina saint Parçvanātha ;2 it exists in a Tibetan" version in the Tanjur, and in a Sinhalese rendering; many stanzas are quoted in works on poetics; it was repeatedly 4 imitated from the Pavanaduta of Dhoi in the twelfth century onwards; we have from that century and later many com- · ¹ Bhau-Dāji, Lit. Rem., pp. 50 f. 2 Pathak's ed. (1916) rests on this. A Nemiduta of Vikrama in 125 verses ends each with a line from a rather interpolated text. 3 H. Beckh, Ein Beitrag zur Textkritik des Kálidāsas Meghaduta (1907); G. Huth, SBA. 1895, pp. 268 ff., 281 ff.; date 13th cent.

• Aufrecht, ZDMG. liv. 616, mentions other imitations; cf. IHQ. iii. 273 ff. 87

THE MEGHADŪTA mentaries, including that of Vallabhadeva,' who gives 111 verses, of Dakṣiṇāvartanātha (c. 1200), who has 110, and of Mallinātha,² who has 118. Inevitably many other lyric poems were ascribed to Kālidāsa, including two of some merit, the Ghaṭakarpara and the Çrīgāra- tilaka, but there is no real probability of proving them his. 6. The Kumarasambhava High as Indian opinion ranks the Meghaduta, which won also the commendation of Goethe,³ to modern taste the Kumāra- sambhava appeals more deeply by reason of its richer variety, the brilliance of its fancy, and the greater warmth of its feeling. The Meghaduta has, with reason, been ascribed the merit of approaching more closely than any other Indian poem to the rank of an elegy; the Kumarasambhava varies from the loveliness of the spring and the delights of married love to the utter desolation induced by the death of the beloved.. The subject is unquestionably a daring one, the events which bring about the marriage of the highest god Çiva to Uma and the birth of Skanda, the war god, and Anandavardhana5 tells us that there were critics who deemed it wrong to depict the amour of two deities. Still less permissible does the subject naturally appear to modern taste, unless we realize that as in the Meghadüūta we must see the poet's power of suggestion; the wedlock of Çiva and Umā is no mere sport, no episode of light love such as that of Zeus with Danae or many another. From this union springs a power destined to perform the slaying of the demon Taraka, who menaces the world with destruction; moreover, their nup- tials and their love serve as the prototype for human marriage and human love, and sanctify with divine precedent the forces which make the home and carry on the race of men. ¹ Hultzsch places him in the 10th cent., but see Pathak's ed., pp. xiv ff. He knew Bilhana and Hemacandra, but is cited in 1140 A. D. > This famous commentator, who also explained the epics of Kalidāsa, Bhatti, and Māgha, and Vidyādhara's Ekāvali (see ed., pp. xxiv ff.) A comm. on the Nalodaya is given, Madras Catal., xx. 7923. 8 Cf. von Schroeder, Indiens Lit. und Cultur, p. 548.

• Ed. NSP. 1906; 1-viii, TSS. 1913-14; i-vii, trans. R. H. Griffith, London,

1879- 5 iii. 6, p. 137. Mammața disagrees. Bhāravi, c. 1400. KĀLIDĀSA AND THE GUPTAS The poem begins with a brilliant piece of description of the Himalaya, the abode of Çiva. Kālidāsa, unlike many a classical and even modern poet, had no hatred of mountains; his fancy makes them the dwelling of merry sprites who play in their caves, round which eddy the clouds, affording welcome screens for the maidens when they undress; the wind, wet with the drops of the streams of the Ganges as it descends from heaven, beats on the trunks of the deodars, and bends the peacock feathers, the scanty dress of the gnomes who chase the antelope. In marked contrast to this innocent frolic sits Çiva, sunk in deepest meditation, and on him with other maidens waits Umā, born of the mountain god himself, plucking flowers to offer to him, and fetching water and grass for his service. Canto ii shows us the gods in deep distress, for a demon Tāraka has arisen to menace them, and Brahman himself can afford no aid, for he has accorded him his protection, and even a poison tree cannot be cut down, if one has reared it oneself. Only Çiva can aid, Çiva who surpasses Brahman and Vişņu in glory, and, if Umā can win him, from them will spring a deliverer. Indra then seeks the aid of Kāma, god of love, to win Çiva's heart for Umā. The next Canto shows Kāma ready and willing to effect the end desired if Spring will be his comrade as well as his dear wife Rati. There follows a brilliant picture of the new life and love awakened in nature by the advent of Spring with Kāma, but the sight of Çiva seated still as a flame when no wind blows, a cloud without rain, daunts even Kāma's heart and he quails. But Umã with her friends appears, and Çiva is begged to hearken to their devotions; he feels himself strangely moved, and glan- cing sees Kāma on the point of discharging at him his deadly arrow. One fiery glance from the god's eye reduces him to ashes. Then follows (iv) a brilliant and touchingly pathetic picture of the lament of Rati for her dead husband; she will not accept the consolation urged on her by Spring; instead she bids him heap the pyre so that she may follow him in death. But her fatal purpose is stayed by a voice from on high, which assures her of reunion with her beloved when Çiva shall have relented and taken Umā to spouse. In sorrowful hope Rati con- tinues her life. 3 The first throw has failed and Umā is bitterly disappointed, 88 89 THE KUMĀRASAMBHAVA bitterly ashamed. She determines, despite all protests, to per- form asceticism until she wins her desire; in summer she exposes herself to the appalling heat and smoke of four fires, in winter lies in icy water, in the rains sleeps on the naked rock. As she is engaged in these acts a hermit appears before her and questions her; from her sighs he learns that she loves, and from her maids who that lover is. He proceeds to depict in appalling colours the god of her desire, but she fiercely and bitterly rebukes his attacks; delighted he reveals himself as Çiva incar- nate (v). All now is ripe for the wedding, but Kālidāsa detains us with a gay picture of the solemn scenes which lead up to it. The Seven Seers themselves with Arundhati come as wooers from Çiva to seek the maiden's hand; she stands, eyes downcast, counting the leaves on the lotus in her hand, at her father's side, while his eyes wander to the face of his consort, for in matters affecting their daughters householders are wont to obey their wives' desires (vi). The wedding follows, described, doubtless from the model of imperial ceremonies, with rich abundance of detail; the mother, in her excitement between joy and sorrow, cannot see to place correctly the painted mark on her daughter's forehead, and misplaces the woollen marriage thread which the nurse, more calm and practical, sets aright. With this ends the poem in many manuscripts; others add ten cantos. Of these Canto viii describes, according to the principles of the Kāmaçāstra, the joys of the wedded pair; doubtless such frankness is abhorrent to western taste, but the doubts of its genuineness which have been expressed are clearly groundless; it seems certainly¹ to have been known to Bhāravi, to Kumāradāsa, and to Māgha, and quotations from it occur in the writers on poetics. Nor in poetic skill is it in the least inferior to Kālidasa's work. The case 2 is other with the following cantos. They tell of Agni's approach, first in dove shape, then in his s proper person, to Çiva as he prolongs for centuries the joys of dalliance, begging his aid. From the seed of Çiva, cast in the ¹ See Walter, Indica, iii. 21, 25 f., ni. 6. who suggests use of viii. 63 in Vikramorvaçi, 2 Jacobi, OC.V. ii, 2. 133 ff. i-viii are used in the Çankarasamhita of the Skanda Purana, but it in ix-xvii; Weber, ZDMG. xxvii. 179 ff., 190 ff.; Pandit, ii. 19ff., 85 ff. 90 KĀLIDĀSA AND THE GUPTAS Ganges and shared by the six Kṛttikās, Pleiades, Kumāra is miraculously born, and grows up delighting his parents by his childish play. But the gods are in terror, the city of the gods is dismayed through Taraka; Indra comes to demand help; Çiva grants his prayer and assigns Kumära to the task. The great host of Taraka is described in Canto xiv, then the portents which warn him not to war (xv). Blinded by pride he refuses, bids his young opponent go back to his father and mother rather than fight, assails him with his whirlwinds and magic fire, until pierced to the heart he falls dead. The poem thus goes far beyond the birth of Kumāra as its title promises, and the inferiority of the new cantos is obvious on every ground. The metre is carelessly handled; in five cases caesura is neglected at the end of the first and third verses of the Çloka, a negligence quite foreign to Kālidāsa; the same carelessness is seen six times in Upajāti stanzas, where too weak caesuras-at the end of a compound, not of a word-are used far more often than by Kalidasa. In order to manage his metres the poet has to resort to versefillers, abhorred of really good writers; su is repeatedly thus used, as well as sadyaḥ and alam; the constant use of periphrasis is doubtless due to the same cause: the writer expends much ingenuity in coining new designations for his characters, and is so fond of the superfluous anta at the end of compounds-which we have seen in Vatsabhaṭṭi-that Jacobi has conjectured that he was a Maratha, in view of the Marathi locative amt. In the later manner is the free use of prepositional compounds and the impersonal passive with subject in the instrumental; the former use just appears in Kälidāsa, the latter is common from Bhäravi onwards. Moreover, save occasionally, as in the battle scene, the poetical value of the cantos is small, and in confirmation of the internal evidence it may be added that neither commentators nor writers on poetics cite them nor are imitations found in later poets. 1 Of Kalidasa's model for his poem we know nothing, but we can trace in it the influence of Vālmīki. In the Rāmāyaṇa ¹ we have a brilliant picture of the contrast of the beauty of spring in the Kişkindhā forest as contrasted with the ceaseless sorrow of Rāma, bereft of Sitã, nor can we doubt that this has influenced 1 iv. 1. THE KUMĀRASAMBHAVA 91 Kālidāsa to draw the wonderful picture of Spring's advent and the revival of the youth and life of the world. There is a parallel too for Rati's despair ¹; when Vālin falls Tārā addresses him with words not less sincere because they bear the stamp of the classic style: 'Why dost thus speak no more to thy beloved? Arise and share this fair couch with me; the best of men lie not, as thou, on the ground. Too dear dost thou hold, o lord, the earth even in death, since me thou dost leave alone and her hast clasped in thine embrace. Ended our days of joy together in the fair forest; sunken am I in a deep sea of sorrow, without joy, without sustenance, since thus hast departed. Hard my heart that it can see thee stretched on the ground and yet not break from sorrow. Hints too for the demon Tāraka are clearly taken from the description of Rāvaņa in the Rāmāyaṇa. There are doubtless reminders here and there of Açvaghoşa, as in the description of the actions of the women of the city on the advent of Çiva and Parvati, which has a prototype in the description in the Buddhacarita³ of the entrance of the prince, and which is taken up again in the description in the Raghuvança of the entry of Aja and Indumati. The problem why the poem was never finished by its author remains insoluble. The loss of the last pages of a solitary manu- script may be the explanation, but it is far more likely that the poet, deterred either by contemporary criticism of his treatment of the divine pair, or by the feeling that the legend of the birth with its strangeness and miracles was not a true theme for poetry, abandoned the purpose and left his work unfinished. It can hardly be claimed that death intervened, for there can be no doubt that the Raghuvança is a later work. This shows itself both in the graver tone, in the references to the Yoga philosophy and the less personal conception of the universe as compared with the magnification of Çiva in the Kumarasambhava, and in the growing pedantry seen in the use of similes derived from grammar, of which we have only modest suggestions in the Kumarasambhava. Thus Rāma's army follows him to serve ¹ iv. 23; cf. vi. 111 (of Rāvana). 2 Cf. also Rām. vi. 124. 45 with xiii. 36. 4 vii. 56-69. 6 vii. 5-16. 7 ii. 27; vii. 69; Raghuvança, xii. 58; xi. 56; i. 1; xv. 7, 9. $ Cf. Walter, Indica, iii. 11 ff. 5 ii. 13-24. 92 KĀLIDĀSA AND THE GUPTAS his purpose as the prefix adhi is followed by the root i to make the word adhyayana; Sugrīva is put in Vālin's place as king as a substitute replaces the root, and husband and wife are theme and suffix. Moreover, in the constant parallels between the two poems, as in the description of the marriage rites, the priority seems to belong to the Kumarasambhava; it is curious that Kālidāsa shows a distinct love of using the same metre for the same theme; thus in both we have the Çloka used in prayers,¹ death is described in the Viyoginī,2 a ruined state in the Upajāti.³ 7. The Raghuvança Though inferior in some slight degree to the Kumarasambhava, the Raghuvança may rightly be ranked as the finest Indian specimen of the Mahākāvya as defined by writers on poetics. Dandin lays down that the subject should be taken from old narratives or traditions, not therefore invented; the hero should be noble and clever; there should be descriptions of towns, oceans, mountains, seasons, the rising and setting of the sun and the moon, sport in parks or the sea, drinking, love-feasts, separa- tions, marriages, the production of a son, meeting of councils, embassies, campaigns, battles, and the triumph of the hero, though his rival's merits may be exalted. It should not be too compressed, and it should be replete with sentiments (rasa) and the emotions which underlie them (bhava). It should have effective transitions (sandhi), an allusion to the five stages of action recognized by the writers on drama, by which from its opening the movement advances after a halt to the central moment, pauses, and reaches the dénouement. The metres must be charming, and each Canto, which should not be too long, should end with a change of metre. The poem should begin with a prayer, paying homage or in addition invoking a blessing, or an indication of the subject-matter. It should promote the ends of Dharma, conduct, Artha, worldly success, Mokṣa, final release, and Kāma, love. ¹ Kum. ii. 4-16; Ragh. x. 16-32. 3 Kum. xili; Ragh. xvi.

• i. 23 ff.

2 Kum. iv; Ragh. vii. THE RAGHUVANÇA 93 The Raghuvança¹ is true to the type, for the central figure is Rāma, though in accord with the title the poem first sketches the history of the dynasty of the sun-born kings, descendants of the Ikṣvāku` whose name occurs in the Rgveda, and whose family is renowned in the epic and the Purāņas. This wide theme gives the poet full space to exercise his power of descrip- tion; war and the coronation of a king, the choosing of her mate by a young princess at a Svayamvara, the marriage rite, the loss of a darling life and the grief of the bereaved husband, town and country, the seasons, the incidents of a great Digvijaya, the triumphal progress of a king who seeks to conquer the earth, all form occasions for the poet's skill. The poem carries us at once into an atmosphere strange to us; Dilīpa is king but childless; he learns that by chance when returning from a visit to Indra he has failed to show reverence to his sacred cow, who has cursed him; to make amends he determines to follow in worship the movements of her daughter, Nandini, on earth; dutifully he carries out his vow, saves her from a lion by offering his own body in exchange, and Nandini accords him the wish of his heart. Soon the father gazes, with eyes as still as lotus blossoms shielded from the wind, on the lovely face of his son, his heart overflows as the sea at the sight of the moon. The young Raghu waxes fast, is given the rank of Crown Prince and bidden guard the horse that must wander for a year before his father can perform the sacred horse sacrifice; the steed disappears, but with Nandini's aid Raghu's eyes are opened until he can see where in the east Indra has taken the horse. Vainly he strives against the god, but pleased by his valour he accords him every wish save the return of the horse, and the gallant youth demands that his father shall have the full fruit of the sacrifice. The offering performed, Dilīpa gives to his son the white parasol, emblem of sovereignty, and, true to his family's rule, retires to the life of an ascetic in the forest (i-iii). Canto iv recounts the knightly adventures of Raghu as conqueror of India; he advances against the Suhmas, defeats the princes of Bengal, and erects pillars of victory on the islands of the Ganges; neither the elephants nor the arrow hail of Kalinga stay his course, Ma- ¹ Ed. S. P. Pandit, BSS. 1869-74; Nandargikar, Bombay, 1897; trans. Walter, Munich, 1914. KĀLIDĀŠA AND THE GUPTAS 94 hendra yields, the Käverī is crossed, the south invaded, the Pandyas pay tribute of pearls. Thence the hero bends his path north, through the Malaya and Dardura hills, the sea of his host covers the long slopes of the Sahya mountain, the dust of the army clings to the hair of the ladies of Kerala, the Murala river, the Trikūṭa hill witness his fame. Thence by land, as a pious king, not by the polluting sea, he advances against the Persians and the Yavanas, Greeks; the dust of the conflict hides the warring hosts whose presence is revealed by the twang of their bows alone, the bearded foemen cover thick the ground, those who escape death cast off their helms in token of submission; the victors wearied slake their thirst with wine. Next Raghu bids his steeds roll in the Indus-a variant has Oxus-sands, overthrows Hūnas and Kambojas; the winds of the Himalaya set the reeds hymning his victories. The mountain folk feel his power, fire flashes from the mountain-sides beneath the rain of spears and arrows, and the folk of the Utsavas lose for ever their joy in festivals (utsava). The Lauhityā is crossed, Prāgjyotiṣa subdued, and Kāmarūpa yields tribute of wild elephants. In this spirited and martial narrative we may justly see the reflex in the poet's mind of Samudragupta's great conquests, and with customary skill the subject changes in Canto v to a very different theme. Raghu's generosity impoverishes him; when a Brahmin Kautsa begs him to aid him to meet the vast de- mands of his teacher, he resolves to storm the treasure-house of Kubera, god of wealth, but a rain of gold saves him from impiety. The Brahmin's gratitude secures him a son, Aja, who soon equals his father. Bidden to take part in the Svayamvara, at which the sister of a kingly neighbour will choose her mate, he sets forth; on the way he boldly attacks a monstrous wild elephant, which under his stroke changes to a Gandharva, con- demned by a curse to wear this shape until released by the blow of an Ikṣvākuid's arrow, who gives him in reward a magic weapon. Canto vi presents us with a brilliant picture of the Svayamvara; the princess, with her companion Sunanda beside her, passes by prince after prince as they stand eager before her ; 1 This fact renders it far more probable that his Açvamedha is that present to Kalidasa's mind than that of Kumāragupta, of whom we have no record of great military achievements. THE RAGHUVANÇA 95 none please her, one is a dicer, therefore bad as a man; in vain Sunandā presses on her Añga's lord; he has all merits, but tastes vary. In revenge she bids Indumati pass on, when she notes that her heart is won by Aja, but the maiden lays shame aside, and accords to him the coronal which marks him as her spouse. The marriage ceremony is performed, the young pair set out home, but the shamed princes have planned revenge, and re- solved to take away by force the princess. Aja wages fierce battle with them, in the end the Gandharva's gift prevails, and he takes from his foes their honour, though he spares their lives (vii). His reign is fortunate; while Raghu as a hermit tames the senses, Aja destroys the foes of his realm, and, when Raghu dies, he pays him all the honours of a Yogin's funeral. But a fatal misfortune awaits him; a garland from the sky blown by the wind falls on Indumatī's breast and slays her, though in truth for her death means release from her mortal bondage imposed on her, in reality an Apsaras, through a curse. No consolation is this thought to Aja; in vain is he reminded of the folly of mourning. for the dead who are burnt by the tears of the living; in vain every consolation regarding the shortness of life and the duty of kings is urged on him; broken-hearted, he dies and Dasharatha reigns in his place. Of him Canto ix has no concrete facts to tell us, until after a brilliant description of spring we are told of the fatal hunt, when, after displaying equal prowess and pity, Dasharatha in pursuit of an elephant mortally wounds a Brahmin boy; he bears the dying youth to his aged parents, and hears the curse of a like doom. In Canto x we leave the realities of life to learn of the magic incarnation of Visņu in the sons born to Dasharatha; in xi Rāma's youth, his visit to Viçvāmitra's hermitage where he slays the demon Tāḍakā, his journey to Janaka's court, where he wins at the Svayamvara the hand of Sita, and his overthrow of Paraçurāma, who recognizes in him the godhead, are rapidly re- counted. The banishment of Rāma by Kaikeyi's device, the life of Rāma and Sītā in the forest, her capture by Rāvana, the search for Lankā,¹ the crossing of the ocean with the monkey horde, and the great battle between Rāma and Rāvaṇa, described in vivid colours, bring us to Canto xii in which Kālidāsa's descriptive 1 Cf. for its situation M. V. Kibe, Rawana's Lanka Discovered (1920). Hopkins (Great Epic, p. 80) appears to accept Ceylon as Lankā. 96 KĀLIDĀSA AND THE GUPTAS powers find congenial subject-matter in describing the sights of India as seen from the aerial car on which Rāma and Sītā return to Ayodhyā. Then follows a series of brilliant sketches; Rāma and Sītā visit the widows of the king, who scarce can see them for their tears, which speedily change to joy. Sitä alone weeps for the trouble her beauty has brought her husband, a foreboding of woe. For the moment all is brightness; the glorious ceremonial of the royal consecration follows. But disaster is at hand; malicious voices reproach the king whose one wife has stayed so long in Rāvana's home. Rāma places duty above love; he bids Lakṣ- mana take Sītā-now pregnant-to Vālmīki's hermitage, and there break to her the truth of her fate; overwhelmed, she de- plores her lot but utters no reproach. Rāma rules in solitude, her sculptured form his companion in his sacrifices (xiv). From his sorrow he is awakened to overthrow demon foes on the Yamunā banks, while in the hermitage Sitā bears two boys who, taught by Vālmīki the tale of their father's deed, console her sorrowing heart by reciting it. The day comes when Rāma determines to perform the horse sacrifice; he rests in a hut be- side the golden statue of his wife; he hears from the boys the song of his deeds; the people, Rāma himself recognize them for his own, Vālmīki begs reinstatement for the queen. Rāma asks only that her stainless purity be made clear; she comes before him, swears to her truth as she drinks the holy water; the earth goddess appears and takes her in her bosom to bear her to the realm below. Rāma transfers to his sons the burdens of the state, saddened by the restoration of Sītā only to be lost forth- with; in due course, followed by all the people, he goes forth from the town and is caught up in a heavenly chariot. The effective and pathetic picture of Sita's end and the return to heaven of Rāma might well have closed the poem, but Canto xvi is not without merit. Kuça, Rama's son, reigns at Kuçāvati; in a dream Ayodhyā appears to him in the guise of a woman whose husband is afar, reproaches him with her fallen condition, and bids him return. Kuça obeys, Ayodhyā once more is glorious, and a description of the delights of summer rivals, but fails to equal, that of spring in Canto ix. For the rest the poem sinks in interest, as Kālidāsa has nothing to tell us but THE RAGHUVANÇA 97 names of worthless kings whose harems supplied their sole interest in life. We cannot deny ¹ his authorship of Cantos xviii and xix; no ancient authority questions them, and they are cited, if rarely, by writers on poetics. But their brevity and the utter abruptness of the end, when the widow of Agnivarman, a worth- less debauchee, is awaiting the birth of her child, suggest that we have no more than a rough draft. Yet we would gladly assign to a poetaster meaningless puns on names of kings, as when Pāriyātra is merely said to have exceeded in height the Pāri- yātra mountains, or the incredible tastelessness of the action of a king who hangs his foot out of the window for the people to kiss. Vālmīki, of course, is the chief creditor of Kālidāsa in this poem. Here and there one certainly surpasses the other; though normally the advantage lies with the younger poet, yet there are exceptions. Fine as is Kālidāsa's picture of Rāma's meeting with the sons who know him not, it yet is still more affecting in the leisurely march of the epic, and Kālidāsa has failed to improve on the scene of Sīta's vindication. But his merit shines out in such cases as his description of the return to Ayodhyā; future poets were to imitate it, but not one to equal it. No other epic of Kalidasa has come down to us, and the rela- tion in time of his epics to his dramas is insoluble. The sugges- tion that he is responsible for the Setubandha, which relates the tale of Rāma from the advance against Rāvana and the build- ing of the bridge to Lankā down to Rāvana's death, is excluded by the style, with its innumerable plays on words, alliterations, recondite similes, exaggeration, and its enormous compounds. Its date is uncertain, as of Pravarasena of Kashmir 4 its author or patron we know nothing definite. Still more ludicrous is the suggestion that the Nalodaya is his; that rimed poem of ¹ As does Hillebrandt, Kālidāsa, pp. 42 f. They seem known to the Aihole inscr. (EI. vi. 8 f.) of Ravikirti who boasts his rivalry with Kalidasa and Bharavi. For un- evenness in great poets cf. Aeneid v as criticized by Tyrell, Latin Poetry, pp. 153 f. ² On alleged use of the Padma Purana, see H. Sarma, Calc. Or. Series, 17. Ed. and trans. S. Goldschmidt, 1880-4. Date before Bana, perhaps late 6th cent., Slein, Rajatarañgiṇī, i. 66, 84 f.

• That the Vākāṭaka Pravarasena had anything to do with the poem seems quite

unproved. Ed. and trans W. Yates, Calcutta, 1844; Bhandarkar, Report, 1883-4, p. 16; A. R. S. Ayyar, JRAS. 1925, pp. 263 ff., who ascribes Vasudeva as author also of the 3149 H KĀLIDĀSA AND THE GUPTAS intolerable affectation is perhaps not the production of Ravideva, author of the Rākṣasakāvya, of equal demerit, before the seven- teenth century, but the work of Vasudeva, protégé of Kulaçek- hara and Rama. 98 Kalidasa's Thought As Sophokles seems to have found his perfect milieu in the Athens of Perikles' happy days, so Kālidāsa appears to us as the embodiment in his poems, as in his dramas, of the Brahmanical ideal of the age of the Guptas, when oder had been restored to a troubled earth, foreigners assimilated or reduced, and prosperity broadcast.¹ Ingenuity has traced in the history of the first five of the rulers in the Raghuvança an exemplar of the exploits of the first five of the Gupta kings; granted that Kālidāsa may have known and profited by the literary activity of Harisena, which doubtless extended far beyond the one inscription which has come down to us, still we may safely doubt any such parallelism. But Kalidasa does represent, if we may judge from his poetry, the complete carrying out of the rule of life laid down for a Brahmin or a warrior or clansman. Youth, in this view, is the time for study under a teacher, then follows the period of manhood with its happy wedlock, then in stages that of the hermit whose mind is set on things eternal. The scheme is in many ways perfectly adapted to Indian life; it starves no side of man's life; four aims of existence are recognized by Kālidāsa himself, who finds them embodied in the sons of Dilipa, them- selves reflexes of Visņu himself. They are duty, governing man's whole life; the pursuit of wealth and of love, the occupa- tions of his manhood; and release, the fruit of his meditations in old age. We may not share the affection of Indian and even of a section of modern taste for the erotic scenes of the last cantos of the Raghuvança, but we must not regard them as the outpour- ings of a sensual mind. The sages of the Upaniṣads themselves deemed marriage obligatory and the Brhadaranyaka gives the Yudhisthiravijaya, Tripuradahana, and Çaurikathodaya, all rimed, to the 9th cent. The date is improbable; ZII. v. 226 f. ¹ Cf M. T. Narasimhiengar, IA. xxxix. 236 ff. with Hillebrandt, Kalidasa, PP. 137 ff. 2 A. Gawroński, The Digvijaya of Raghu (1915). KĀLIDĀSA'S THOUGHT 99 spell to obtain a male son; the saintly Çvetaketu is deemed an authority on the Kamasutra, and Kālidāsa expressly claims the divine precedent of Çiva and Umā as sanction for the most passionate married love. Statecraft again is essentially part of the material ends of life, and not only does he paint in Rāma an ideal ruler, but throughout the Raghuvança we are reminded of the duties of kings to the subjects. Let us grant that his vision was Brahmanical; he deliberately repeats the condemnation of the Rāmāyaṇa on the Çūdra who threatens the security of established order by venturing to expose himself, head down- wards, hanging from a tree to fire, in order by penance to acquire merit. This reminds us of Fa-hien's ¹ emphatic testimony of the degradation of the Caṇḍālas in the Gupta realm. 1 Youth and manhood are no time for deep philosophic views, and the Kālidāsa of the Rtusamhāra, Meghaduta, and Kumāra- sambhava remains within narrower limits. We feel, however, a growing sense of the greatness and glory of Çiva; the remote figure of the Meghadūta is definitely brought nearer to us in the Kumarasambhava. Even Brahman and Visņu are less than he, and the term Lord, Içvara, is his par excellence; moreover, despite his all-embracing majesty, he is intensely personal. Yet neither Brahman nor Vişņu is forgotten; to Brahman in the Kumarasambhava itself, to Viṣṇu in the Raghuvança two noble prayers are addressed in which in the true spirit of kathenotheism either appears as the greatest of gods, as more than the world, as beyond all comprehension. The inconsistency, however, is rather apparent than real; it is possible to ascertain with fair certainty the view Kalidasa took of the universe, and this affords a recon- ciliation of his diverse views. Both epics, but especially the Raghuvança, show that Kāli- dāsa accepted Sāṁkhya and Yoga views of the nature of the universe. The three constituents of nature, goodness, passion, and dullness, in their ethical aspect afford themes for simile; the Brahman sea as the source of the Sarayū is like the unmanifested (avyakta) whence springs intelligence. Yoga practices are recog- nized the aged king practises concentration (dharaṇā) as he sits on Kuça grass; the difficult posture known as Vīrāsana of ascetics is compared to trees standing motionless; Sītā by asceticism 1 Smith, EHI. p. 314; Foucher. L'Art Gréco-Bouddhique du Gandhāra, 11. 8. H 2 100 KĀLIDĀSA AND THE GUPTAS seeks to secure reunion in her next life with her spouse: the power to pass through closed doors may be won, and the Yogin needs not cremation, but like Raghu is buried in mother earth. But we cannot hold that the godhead envisaged by Kālidāsa is the pale Içvara of the Yoga; in Brahman we are told are united both matter and spirit as they are known in the Sāṁkhya, and this we may fairly take as indicating that to Kālidāsa, as to the author of the Katha Upanişad, over the spirits and matter stood the absolute, who to Kālidāsa takes specially the form of Çiva but who is also Brahman and Visņu, the spirit that perishes not beyond the darkness. With this absolute man is merged on death if he has attained enlightenment, for this is the sense of brahmabhūyam gatim ājagāma in the Raghuvança. If enlighten- ment is not his but good deeds, he has heaven for his share, for knowledge alone burns up man's deeds which else force him to life after life. We need have the less hesitation to accept this view in that it is essentially the standpoint of popular Vedāntism and that it afforded to a man of thought and good sense an effective means of reconciling belief in the three great gods. What is clear is that in his advancing years Kālidāsa's mind turned more and more to the conception of the all-embracing character of the godhead and of the efficacy of Yoga practices to attain union with him. From such a philosophy it would be idle to seek any solution for essential conflicts in the heart of man, or to demand any independent criticism of man's aims and fate. India knew atheists enough, but their works have all but perished, and we must rather be grateful that we have preserved in such perfection the poetic reflex of the Brahmanical ideal both in its strength and in its weakness. Nor, let us remember, does such an ideal shut out deep human feeling such as we may suspect in the longing of the Meghaduta, the lament of Aja over the dead Indumati, of Rati for Kāma slain. But it does demand resignation, and if in perfection of form Kālidāsa's poems proclaim him the Virgil of India, we may admit that he was incapable of the vision and imagery of the sixth book of the Aeneid. 4 KĀLIDĀSA'S STYLE AND METRE IOI 9. Kalidasa's Style and Metre. In Kālidāsa we have unquestionably the finest master of Indian poetic style, superior to Açvaghoṣa by the perfection and polish of his work, and all but completely free from the extrava- gances which disfigure the later great writers of Kavya. Dandin ascribes to his favourite style, the Vaidarbha, qualities which we may fairly sum up as firmness and evenness of sound, avoiding harsh transitions and preferring gentle harmonies; the use of words in their ordinary sense and clearness of meaning; the power to convey sentiment; beauty, elevation, and the employ- ment of metaphorical expressions. He assures longevity to a poem which, in addition to conforming to the rules for a Mahākāvya, is rich in ornaments (alaṁkāra), and Kālidāsa is not sparing in his use of these means of adding grace to his work. But he has the fundamental merit that he prefers suggestion to elaboration; his successors too often thought that they could only prove their capacity by showing all of what it was capable; he was content to produce a definite effect, and to leave well alone; his was the golden mean of Virgil between rustic simpli- city and clumsiness and that over-refinement which is specially fatal. Thus it results that his miniature-painting in its polished elegance often attains relative perfection. The truth of his delineation is seen in the picture of the sorrowing bride in the Meghadüta: utsange vã malinavasane saumya nikṣipya viņām madgotränkaṁ viracitapadaṁ geyam udgātukāmā 1 The critics occasionally find fault, e.g. in the Vyaktiviveka (p. 66) Raghu- vança xvi. 33 is censured for the position of taliye, but they cite him repeatedly as a master, first of Mahakavis; Dhvanyaloka, pp. 29, 207; Kavyaprakāça, p. 2. Bhamaha's assertion that a cloud is not suitable as a messenger must refer to the Meghaduta and may be put beside his attack on Bhāsa's Pratijñāyaugandharayana, proved by T. Ganapatı Sāstrī, çf. Thomas, JRAS. 1925, p 103, who (pp 100 ff.) deals effectively with the attacks on the authenticity of Bhasa's dramas. His verse (Subhāṣitāvali, 1353) is imitated in Ragh viii. 66; GIL. in 159, n. 1. 2 His improvements on Açvaghoșa are numerous and undeniable; cf. the passages in Nandargıkai, Raghuvança (ed. 3), pp. 161 ff.; Formichi, Alvaghosa, p. 350; cf. also Saund. v. 43 with Kum v. 45. The parallel Kum. vii. 56 ff. ; Ragh. vii. 5 ff. with Buddh. in. 13 ff. is conclusive and Hillebrandt's doubts (pp. 103 f.) are hyper- critical. 102 KĀLIDĀSA AND THE GUPTAS tantrīm ārdrām nayanasalilaiḥ sārayitvā kathamcit bhuyo bhūyaḥ svayam api kṛtām mūrcchanāṁ visma- ranti. 'Or perhaps, placing her lute on her lap, whose dark garment proclaims her grief, she will seek to sing a song wherein she has worked my name, but, scarce able to move the string which her tears have bedewed, she will forget the air which she herself hath made. Or, again: tvām ālikhya praṇayakupitāṁ dhāturāgaiḥ çilāyām ātmānam te caraṇapatitam jāvad icchāmi kartum asrais tavan muhur upacitair drstir alupyate me krūras tasminn api na sahate samgamam nau kṛtāntaḥ. 'When I have portrayed thee in love's anger on the rock with my colours and seek to add myself lying at thy feet, my tears well up and ever blot out my sight; cruel the fate which even thus will not permit our union.' There is a brilliant picture of Uma's confusion and of her joy when Çiva reveals himself: adya prabhṛty avanatāīgi tavāsmi dāsaḥ kritas tapobhir iti vadini candramaulau ahnaya să niyamajam klamam utsasarja kleçaḥ phalena hi punar navatāṁ vidhatte. ""From this moment, o drooping maiden, I am thy slave, bought by thy penance," so spake he whose crest is the moon, and straightway all the fatigue of her self-torment vanished, so true is it that fruitful toil is as if it had never been.' There is perfect simplicity of passionate longing in Rati's address to the dead Kāma: kṛtavan asi vipriyam na me: pratikulam na ca te maya kṛtam kim akāraṇam eva darçanam: vilapantyai rataye na diyate? 'Thou hast never displeased me; thee I never have wronged; why then, without cause, dost thou hide thyself from thy weep- ing Rati?' The timid shyness of the new-made bride and her lover's ruses are delicately drawn: vyahṛtā prativaco na samdadhe: gantum aicchad avalambi- tānçukā sevate sma çayanam paranmukhi: să tathāpi rataye pina- kinah. KĀLIDĀSA'S STYLE AND METRE 103 'Addressed she could not answer; when he touched her gown she sought to leave him; with head averted she clung to her couch; yet none the less did she delight the lord of the trident.' ātmānam ālokya ca çobhamanam: ādarçabimbe stimitāya- tākṣi Haropayane tvaritā babhuva: strinām priyalokaphalo hi veçaḥ. 'When with her long eyes fixed on her mirror she saw the reflection of her radiant loveliness, swift she hastened to seek Çiva, for the fruit of woman's raiment is the light in the lover's eyes.' Equally complete in its own effectiveness is the descrip- tion of the tragic shock received by Rati: tīvrabhişangaprabhavena vṛttim: mohena samstambhaya- tendriyāṇām ajñātabhartṛvyasanã muhurtam: kṛtopakāreva Ratir ba- bhūva. 'The bitterness of the blow cast Rati into a faint which dulled her senses and for the moment with true kindness robbed her of memory of her husband's ruin.' Aja's tears have their excuse in nature itself: vilalapa sabāṣpagadgadam: sahajām apy apahāya dhīratām abhitaptam ayo'pi mārdavam: bhajate kaiva kathā çaririşu ? 'He wailed aloud, his voice broken by sobs, forgetting the high courage that was his; iron in the fire yieldeth its strength; how much more feeble mortals?' He feels that his wife has doubted his love: dhruvam asmi çathaḥ çucismite: viditaḥ kaitavavatsalas tava paralokam asamnivṛttaye: yad anapṛcchya gatasi mām itaḥ. 'Surely, sweet smiling one, thou hast judged me traitor whose love was feigned that thou hast gone from me to the world whence there is no return and hast not bidden me even a word of fare- well. No woman could desire a more perfect eulogy: grhini sacivah sakhi mithaḥ: priyaçişya lalite kalāvidhau karunāvimukhena mṛtyunā: haratā tvāṁ vada kim na me hṛtam? 104 KĀLIDĀSA AND THE GUPTAS 'Wife, counsellor, companion, dearest disciple in every loving art; in taking thee tell me what of me hath not pitiless Death taken.' The fatal blow is depicted: kṣaṇamātrasakhiṁ sujātayoḥ: stanayos tām avalokya vihvalā nimimila narottamapriya: hy tacandrà tamaseva kaumudi. 'For a moment she gazed on the garland as it lay on her rounded breasts, then closed her eyes in unconsciousness, like the moonlight when the darkness obscures the moon.' There is humour, on the contrary, in Indumati's rejection of the Anga prince: athăngarajad avatarya cakşur: yahiti janyām avadat kumārī nâsau na kāmyo na ca veda samyag: draṣṭum na sã bhin- narucir hi lokaḥ. 'But the princess turned away from Añga's lord her gaze, and bade her maiden proceed; it was not that he had not beauty nor that she could not see it, but folk have different tastes.' This has the same graceful ease as often in the Rtusamhära: vivasvatā tīkṣṇatarāňçumālinā: sapañkatoyāt saraso 'bhită- pitaḥ utplutya bhekas trsitasya bhoginah: phanatapatrasya tale nişidati. 'As the sun's garland of rays grows ever hotter, the frog sore tormented leaps up from the muddy water of the lake only to fall into the mouth of the thirsty snake, who spreads his hood to shade him from the glare.' There is a pretty picture of girlish haste: alokamargam sahasa vrajantyä: kayacid udveṣṭanavānta- mālyaḥ baddhum na sambhāvita eva tavat: karena ruddho 'pi ca keçapāçah. 'As she rushed to the window, her garlands fell from their place, and she did not even trouble to knot the abundant hair which she caught together in her hand.' The structure of each of these cameos is simple; throughout KĀLIDĀSA'S STYLE AND METRE 105 it is normal to have each verse complete in itself, a single verb serving to support a number of adjectives and appositions, though relative clauses with verb expressed or implied are not rare. The compounds are normally restricted in length, but this is less closely observed in the Mandākrāntā metre, though even then clearness is aimed at and normally achieved. The order of words is very free, partly no doubt by reason of metrical neces- sity. Of the figures those of sound are employed not rarely but usually with skill. Beside the ordinary forms of alliteration as in nirmame nirmamo 'rtheşu, we find the more important Yamaka, in which the same syllables are repeated, in the same or inverted order, but with different sense. There is a certain liberality in the process; thus Kālidāsa is able to match bhuja- latām with jadatām, for I and d, like r and 1, b and v, are admitted as similar, and the same principle is clearly to be seen in cakāra să mattacakoranetrā: lajjāvatī lājavisargam agnau. 'She with the eyes of the intoxicated Cakora, in modesty (lajjā) made offering of fried rice (lāja) in the fire.' In Canto ix of the Raghuvança Kalidasa deliberately shows his skill in Yamakas; there is no doubt that this offends the sound rule of Ananda- vardhana that to seek deliberately such a result destroys the function of poetry which is to suggest-or express-not merely to exhibit form, and we can only conjecture that in this canto, which also is marked out by the amazing number of metres employed, Kālidāsa was seeking to prove that he could vie with any rival in these niceties. In Canto xviii also, Yamakas are superabundant. Throughout, however, we feel Kālidāsa seeking for the matching of sound and sense, to which the Indian ear was clearly more susceptible than our own. Of figures of sense Kālidāsa excels in Indian opinion in the simile, and the praise is just. The Indian love of simile appears freely in the Rgveda, and is attested by the elaborate subdivisions of Indian poetics. The width of Kalidasa's knowledge and the depth of his observation of nature and life are here shown to the highest advantage. But his world is not ours, and doubtless at ¹ As distinct from alliteration the repetition should be in corresponding parts of the verse (Jacobi, ZDMG. lxii. 303, n. 1). 106 KĀLIDĀSA AND THE GUPTAS times his figures¹ seem grotesque to our taste, as when the king comes from his bath and plays with his harem like an elephant on whose shoulder still clings a shoot of the lotus sporting with the females of his herd. But often there can be only admiration; the chariot of the prince is so covered by the arrows of his foes that only by the point of its standard can it be discerned, as the morning wrapped in mist by the feeble rays of the sun; the wound torn by the arrow is the door of death; with joyful eyes the women of the city follow the prince as the nights with the clear stars of autumn the polar star. Characteristic is the love of elaboration of a comparison; the reader is not to be contented with a mere hint, the comparison must be drawn out in full. The Pandya king is peer of the lord of mountains, for the neck- laces which hang over his shoulders are its foaming cascades, and the sandal that reddens his limbs the young sun which colours its peaks. Or again, the princes who hide their jealousy under the semblance of joy are compared to the pool in whose calm depths lurk deadly crocodiles. Or again, the ruined city, with towers broken, terraces laid down and houses destroyed, is like the evening when the sun sets behind the mountains and a mighty wind scatters the clouds. To us, no doubt, both similes and metaphors sometimes seem far-fetched; those from grammar leave us cold, but there is wit in the assertion that the wearing by Rāma of the royal dress when the ascetic's garb revealed already his fairness is equivalent to the vice of repetition (punarukta). The bowmen whose arrows strike, one another are like disputants whose words con- flict. The king seeks to subdue the Persians as an ascetic his senses through the knowledge of truth. Kalidasa is rich also in plays of fancy which present a vivid picture (utprekṣā); it is natural to him to think vividly, to attribute to the mountains, the winds, the streams the cares, sorrows, joys, and thoughts of men. He loves also the figure corroboration (arthantar anyasa); indeed, its caieless use reveals the hand of the forger of the last cantos of the Kumarasambhava. But the double entendre is rare indeed; the instances of it are very few, and they lend no 1 Cf. Hillebrandt, Kalidasa, pp. 112-20. For the Çakuntala, cf. P. K. Gode, POCP. 1919, ii. 205 ff. A very interesting comparison is afforded by Lncan's similes (Heitland in Haskins' Lucan, pp. lxxxiv ff.). KALIDASA'S STYLE AND METRE 107

credit whatever to the suggestion that v. 14 of the Meghaduta is an attempt obliquely to praise Nicula and damn DignSga. Of the former we know nothing, and it was doubtless the later love for Clesas which bade men find them in Kalidasa, where not one elaborate case even can be proved to exist. 1

Kalidasa' s metrical skill is undoubted. In the Rtnsamhara he used normally the Indravajra and Varicastha types, with Vasan- tatilaka and MalinI; one stanza only in Cardulavikrldita occurs. The Meghaduta shows the more elaborate Mandakranta used without variation ; a few slight roughnesses as regards caesura may be adduced as proof of the relatively early date of the poem, but the evidence is too slight to weigh seriously in itself. In the Kumar asambhava we find the normal rule that the canto is written in a single metre with change, as the writers on poetics require, at the close. Thus i, iii, and vii are written in the Indra- vajra; ii and vi in the Cloka, iv in the Vaitallya, and v in the Varicastha, while viii is in the Rathoddhata. The closing changes are furnished by Puspitagra, MalinI, and Vasantatilaka. The Raghuvahga follows on the whole this principle, but exhibits greater variety, suggesting later date. The Indravajra type serves for ii, v-vii, xiii, xiv, xvi, and xviii ; the Cloka for i, iv, x, xii, xv, and xvii ; the Vaitallya for viii, and the Rathoddhata for xi and xix. Canto ix is orthodox up to v. 54, being in Drutavilambita, then it deliberately displays the poet's skill in new metres, each with a verse or so, Aupacchandasika, Puspi- tagra, Praharsini, Mafijubhasinl, Mattamayura, Vasantatilaka, which is also used for 11 verses in v, Vaitallya, CalinI, and Svagata. There occur also odd verses in Totaka, Mandakranta, Mahamalika, and iii is written in Varicastha, with a concluding ' verse in Harinl. There are thus nineteen metres in all to eight in the earlier epic. Detailed efforts to find some sign of develop- ment in any of the metres in respect of caesuras &c. have failed to yield any results worthy of credence. 2

In the Cloka the rules had already been established by epic

1 In Meghaduta 10 dfdbatidha may have a double sense ; 28 rasa; Kumdrasam- bhava, viii. 22 ; Raghuvahga, xi. 20. But in v. 14 Nicula is to be a poet friend, else- where utterly unknown.

a Huth, Die Zeit des Kalidasa (1890), App. ; Hillebrandt, Kalidasa, p. 157. Cf. SIF1. VIII. u. 40 ff. 108 KALIDASA AND THE GUPTAS

practice, and Kalidasa observes them carefully. Of the four Vipula forms he uses the last once only ; the figures ' for the other three out of 1410 half-stanzas in the epics are 46, 27, and 41, or 8-15 per cent, showing that the third Vipula was Kali- dasa's favourite. It is interesting to note that in the form of the syllables preceding the first Vipula Kalidasa shows special care

to select that form (^- ) which is not allowed in the second

Vipula as against that (^-^-) which is permitted in both. The Kumarasambhava has 11 cases of the first to 3 of the second form, the Raghuvat'iga 31 to 1 ; this doubtless indicates increasing care to secure elegance, and it accords with this that in the Kumarasambhava alone is the fourth Vipula found. 2

1 l'"or the Raghuvaiifa they are 32, 18, 27 out of 1096 ; Jncobi's figures (IS. xvii. 444 f.) are corrected from SIFT. /. c. The percentage in Bharavi is 0.-6 ; Magna 27.15 ; Bilhana 8-64 ; (J!r!harsa 0-53 ; Kumaradasa 2-35.

2 X agliuvahfa, mi. 71, should perhaps be read dviliyahemaprakaram. In Kumarasambhava, vii. 11 on one reading position 15 neglected as in Qifupalavadka, x. 60, both dubious (SIFI vili. ii. 7). For the schemes of the metres see chap, xx, §4V BHĀRAVI, BHAṬṬI, KUMĀRADĀSA, AND MĀGHA 1. Bharavi F Bharavi's life we know nothing whatever, though he O second of the Kavya. External evidence proves that he was older than A.D. 634 when he is mentioned with Kalidasa in the Aihole inscription, and he is cited in the Kāçikā Vṛtti; on the other hand he manifestly is influenced by Kālidāsa, while he strongly affected Magha.¹ Bāņa ignores him, so that he can hardly have preceded him long enough for his fame to compel recognition. It is, therefore, wiser to place him c. A.D 550 than as early as A.D. 500. His Kirātārjuniya2 is based, as usual, on the epic. The Mahabharata tells us how, when the Pāṇḍavas with their wife Draupadi have retired under their vow of twelve years' banish- ment to the Dvaita forest, Draupadi, with truly feminine faith- lessness, urges the heroes to break their pledge. A council is held; Yudhisthira pleads for the bond; Bhima controverts his contentions. Vyāsa counsels retirement from the Dvaita forest, and the brothers go to the Kāmyaka wood, where Yudhisthira takes the prudent course of bidding Arjuna, as a preliminary to war, to secure from Çiva divine weapons. Arjuna obeys, prac- tises in the Himālaya severe penances, meets and struggles with a Kirāta, who proves to be Çiva himself; he grants the boon desired, to which the other gods add further largesse. This theme Bhāravi has chosen to expand and illustrate with all the resources of a refined and elaborate art. The opening shows at once the hand of the artist; in the epic the discussion of the brothers arises merely from the dreary plight in which they are ¹ Cf Jacobi, WZKM. ii. 121 ff. 2 Ed. NSP. 1907; trans. C. Cappeller, HOS. 15, 1912; i-iii, with Citrabhānu's comm., TSS. 63. 3 i. 27-4¹. no BHARAVI, BHATTJ, KUMARADASA, AND MAGHA

placed ; Bharavi begins instead with the return of a spy whom Yudhisthira has sent to report on the deeds of Suyodhana— as he is always styled ; he bears the unwelcome tidings that the king is walking in the ways of virtue and charming the hearts of the people. Hence, naturally, Draupadi, anxious for the future, taunts Yudhisthira with his inglorious plight and urges swift battle (i). Bhima adds his support ; Yudhisthira, the unready, has scruples of honour (ii), but seeks counsel from Vyasa, and the sage admits that war must be, but, since the foe is so strong, urges that Arjuna should by penance in the Himalaya win Indra's aid. He vanishes, but a Yaksa appears to lead Arjuna on his way, and they depart, cheered by the good wishes of the re- mainder of the party (iii). At this point the poet's invention displays itself in elaboration; just before, by omitting all mention of the move to the Kamyaka wood, he had shortened the narra- tive, improving greatly the effect ; now he takes the opportunity to display the poet's command of language. In Canto iv the Yaksa leads Arjuna on ; and a brilliant picture is drawn of the autumnal scene, partly in narrative, partly in the mouth of the Yaksa. Then follows (v) the description of the Himalaya itself, the Yaksa lays stress on the mystery which guards it and on its close kinship with Civa and ParvatI, and vanishes after bidding Arjuna do penance on Indraklla. The penance of Arjuna terrifies the Guhyakas, the spirits who haunt Indraklla ; they appeal to Indra to aid them, and he sends Gandharvas and Apsarases to disturb the asceticism which menaces the quiet of his mountain (vi). The heavenly host speeds through the air to Indraklla and makes there its camp ; their elephants merit special description (vii). The Apsarases now leave their palaces, just made by their magic power, and wander in the woods to pluck the flowers ; then the Ganges invites them to the bath, and the bathing scene is described with much charm and beauty (viii). Evening comes, the sun sets, the moon arises — the banal theme wins new effect through the poet's skill ; the nymphs and their lovers drink and seek the pleasures of love ; the day dawns (ix). The Apsarases now turn their minds to their task ; aided by the seasons who now appear six in number to second their efforts, they expend, but in vain, all their charms on the young ascetic (x). Seeing his minions thus foiled through ArjTlna.'s constancy, Indra appears BHARAVI in

himself in the guise of a sage, admires the fervour of the penance, but contends that to bear arms and practise asceticism are incon- sistent ; Arjuna admits the logic of the censure, but asserts that he will do all to save his family's honour. Indra is touched, reveals himself, and bids him win the favour of Civa (xi). Here ends the poet's invention, and we again find the epic as his source. Arjuna continues his penance in order that Civa may bless him ; the seers in distress appeal to the great god, who expounds to them Arjuna's divine nature as an incorporation of Nara, a part of the primeval spirit ; a demon Miika in boar form plans to slay him ; therefore Civa bids his host follow him to guard the prince (xii). The boar appears to Arjuna ; it falls pierced by his own and Civa's dart ; the prince advances to recover his arrow, but is challenged by a""Kirata who claims it in his master's name (xiii). Arjuna rejects the demand in a long speech ; the Kirata returns, and Civa launches, but in vain, his host against Arjuna, who endures unscathed the shower of their arrows (xiv). The host is rallied from flight by Skanda and Civa himself, who then begins a deadly battle of arrows with Arjuna (xv). The two then strive with magic weapons, the hero is beaten (xvi), but grasps again his bow, and with sword, mighty rocks, and the trunks of great trees assails the god, but all in vain (xvii). They box, at last they wrestle ; Civa reveals his true form, and the hero, humbled at last, praises the greatness of the god and begs him for strength arid victory ; the god and the world guardians, who come to the scene, accept his devotion and give him the magic weapons tliat he craves.

The introduction of Civa's host, of its struggles under Skanda's leadership with the hero, and the whole episode of the contest with magic weapons are the fruit of the poet's imagination. One difficulty is obvious ; it is made necessary to duplicate the episode of the force of the penance causing fear and evoking divine inter- vention, and the {Jrolongation of the conflict results in some repetition of ideas. Duplication also results from the description of the amours of the nymphs with the Gandharvas and their attempts on the prince. The poet's skill led him, we must con- fess, to exhibit it too freely, and the introduction of magic weapons leaves us cold. In this regard Valmiki has a fatal influence on Sanskrit poetry ; the mythical background of the 112 BHĀRAVI, BHAȚȚI, KUMĀRADĀSA, AND MAGHA Rāma legend produced the unreality of his combats, which every epic poet felt bound to copy. Another influence seen strongly in the first two Cantos is that of the political principles of the day, which have ample opportunity of illustration in the record of Suyodhana's rule and in the arguments by which Yudhisthira seeks to justify the keeping of their faith by his brothers. There is no doubt of the power of Bharavi in description; his style at its best has a calm dignity which is certainly attractive, while he excels also in the observation and record of the beauties of nature and of maidens. The former quality is revealed re- peatedly in the first Canto, the very first line of which strikes the true note of high policy; then follows: krtapramāṇasya mahim mahībhuje: jitāṁ sapatnena nive- dayisyataḥ na vivyathe mano na hi priyam: pravaktum icchanti mṛṣā hitaisinah. 'When he bent low in homage his mind wavered not, thougi. he had to tell the king that his realm had been won by his foe, for men who seek one's good care not to speak flattering words.' In the same strain Suyodhana is praised: na tena sajyam kvacid udyatam dhanuḥ: kṛtam na vā tena vijihmam ananam gunanurāgeṇa çirobhir uhyate: narādhipair mālyam ivāsya çāsanam. 'Never has he raised his bow to shoot, never has a frown dis- torted his face; loving his virtues the kings bear as a garland on their heads his royal orders.' The setting sun and the rising moon are happily portrayed: ançupanibhir ativa pipasuḥ: pankajam madhu bhṛçam rasayitvā klībatām va gataḥ kṣitim eşyanl: lohitam vapur uvāha patangaḥ. 'Ruddy glowed the sun as he hastened to rest, as though over- deep he had drunken with his rays, in his thirst, the sweetness of the lotus.' BHĀRAVI samvidhätum abhişekam udāse: Manmathasya lasadançu- jalaughaḥ yaminīvanitayā tatacihnaḥ: sotpalo rajatakumbha ivenduḥ. 'For Love's consecration the lady night raised aloft the moon with its shimmering sea of beams and its spots full in view, like a silver chalice decked with lotuses.' The advent of the cool season is thus greeted : 113 katipayasahakārapuşparamyas: surabhimukhahimagamāntaçańsī: raikabandhuḥ. 'Then came the cool season, Love's one friend, lovely with its mango blooms here and there, when frost is rare and but a few Sinduvāras awake from sleep, the harbinger of the end of winter and the coming of spring.' The bathing scene is rich in pretti- nesses: duvăraḥ tanutuhino 'Ipavinidrasin- samupayayau çiçiraḥ sma- tirohitäntäni nitantam akulair: apām vigahad alakaiḥ prasaribhiḥ yayur vadhūnām vadanāni tulyatam: dvirephavṛndāntari- taiḥ saroruhaiḥ. 'Hidden by their long hair in utter disorder through plunging in the water, the maidens' faces seemed like lotuses covered with swarms of bees.' priye 'para yacchati vācam unmukhī: nibaddhadṛṣṭiḥ çithi- läkuloccayā 3149 samādadhe nānçukam āhitam vṛthā: viveda puspesu na pāņi- pallavam. 'Yet another, face upturned and eyes fixed on her lover as he spoke, gathered not together her garment, though the knot slipped and fell, nor realized that her tender hand had missed the flowers it sought.' Characteristically, the same idea is varied later in the canto: 0 vihasya panau vidhṛte dhṛtāmbhasi: priyena vadhvă ma- danārdracetasaḥ sakhiva kañci payasa ghanikṛtā: babhāra vitoccayabandham ançukam. I 114 BHĀRAVI, BHATTI, KUMĀRADĀSA, AND MAGHA 'As her hand, full of water, was laughingly grasped by her lover, 'twas her kindly girdle which the water had stiffened that saved from falling the garment of the loving maiden, for the knot that held it had slipped." His play of fancy is constant and extensive; he acquired the style of parasol-Bhāravi from his comparison (v. 39) of the lotus dust driven by the winds to the goddess Lakşmi mirrored in a golden parasol. Still less attrac- tive to our taste is a simile ¹ based on the mute letter (anubandha) between ste and ending in grammar. Bharavi, however, is guilty of errors of taste from which Kāli- dāsa is free. Especially in Canto xv he sets himself to try tours de force of the most foolish kind, redolent of the excesses of the Alexandrian poets. Thus one verse has the first and third, second and fourth lines identical; in another all four are identical ; one has practically only c and r, another only the letters s, f, y, and 7; in other stanzas each line reads backwards the same way as the next, or the whole stanza read backwards gives the next; one stanza has three senses; two no labial letters; or each verse can be read backwards and forwards unchanged. One sample must serve: na nonanunno nunnono nănā nānānanā nanu nunno 'nunno nanunneno nānenā nunnanunnanut. 'No man is he who is wounded by a low man; no man is the man who wounds a low man, o ye of diverse aspect; the wounded is not wounded if his master is unwounded; not guiltless is he who wounds one sore wounded.' But at least he eschews long compounds, and, taken all in all, is not essentially obscure. Bharavi sets a bad example in his fondness for showing his skill in grammar, and he is in many ways the beginner of manner- isms in the later poets. The ridiculously frequent use of the root tan begins with him; he is fond of passive perfect forms, in- cluding the impersonal use; the adverbial use of prepositional compounds is a favourite form of his; many of Pāṇini's rules of rare type 3 are illustrated by him, as ças with double accusative, 1 x111. 19; cf. xvii. 6. Cf. Māgha, i. 47, 95, 112; x. 15; xiv. 66; xvi. 80; xix. 75. 2 Walter, Indica, ini. 34 f. 3 Cappeller, pp. 153 ff. On the perfect cf. Renon, La valeur du parfait, p. 87. BHĀRAVI 115 darçayate in the same use, anujīvisātkṛta, stanopapiḍam, the double negative as a positive, and na compounded as in nanivṛtam; it occurs also with the imperative. Most interesting in his elaborate care in the use of the narrative tenses, which Kalidāsa and the other poets treat indifferently. In Bharavi the imperfect and the aorist are not tenses of narrative use; they occur only in dealing with what the speaker has himself experienced (aparokse), and the imperfect denotes what happened in the more remote past (anadyatane), the aorist the immediate past (adyatane), exceptions being minimal; the aorist hence is extremely rare, occurring only ten times to 272 times in Mägha. The perfect is the tense of narrative, save in the case of the present perfects aha and veda. The present occurs with sma not rarely in narrative as a past; the participle in tavant is used in speeches only, that in ta in both. Both the imperative and the aorist with mã are found in interrogations beside their normal uses, and labdha is used in the passive, the periphrastic future having always its precise sense of a distant event. Errors in grammar are few, but ajaghne seems indefensible. In metrical form Bhāravi is as developed as he is in the use of the figures of speech, of which scores can be illustrated from his poem. Only once does he condescend to use a single difficult metre, the Udgatā, for a whole canto (xii), a single Praharsini terminating it. In v he uses sixteen, in xviii also sixteen different forms. The Upajäti of Indravajrā type predominates in iii, xvi, and xvii; Vançasthā in i, iv, and xiv; Vaitālīya in ii ; Drutavi- lambita in xviii; Pramitākṣarā in vi; Praharṣiṇī in vii; Svāgatā in ix; Puspitāgrā in x ; Çloka in xi and xv; Aupacchandasika in xiii. Of the other metres few save Vasantatilaka ¹ have much use; Aparavaktra, Jaloddhatagati, Jaladharamālā occur, like Candrikā, Mattamayūra, Kuțila, and Vançapattrapatita, once only. The Rathoddhatā is a good deal used in xiii; but Çālini, Mālini, Prabhā, and Çikhariņi are all rare.² In the Çloka Bhāravi conforms in general to the same rules as Kālidāsa. But he never uses the fourth Vipulā form, and in his 250 half-stanzas he uses the first three Vipulās respectively fifteen, ¹ The final syllable is in three cases in line a, in one case in line c. 2 Thus Bharavi has eleven or twelve principal metres to six of Kalidasa and sixteen of Magha. I 2 116 BHĀRAVI, BHAṬṬI, KUMĀRADĀSA, AND MAGHA eight, and two times; Kālidāsa, on the contrary, likes best the third Vipulă. 2. Bhatti Bhaṭṭi, the author of the Ravanavadha, more usually simply styled Bhaṭṭikavya, tells us that he wrote in Valabhi under Çrīdharasena. But four kings of this name are known, the last of whom died in A. D. 641, so that we remain with nothing more secure than that as a terminus ad quem. The suggestion ² that he is to be identified with Vatsabhaṭṭi of the Mandasor inscription lacks all plausibility, if only for the reason that Vatsabhaṭṭi commits errors in grammar. The name Bhaṭṭi is Prakritized from Bhartr, and it is not surprising that in tradition he has been either identified with Bhartṛhari or made a son or half-brother of that famed poet. There is, however, nothing but the name to support the suggestion. We know, however, that he was imitated by Māgha, and it is a perfectly legitimate suggestion that his work gave Magha the impetus to show his skill in grammar to the extent that he does. More important still is the plain fact that he was known to Bhãmaha. In ending his poem he boasts that it needs a comment: vyakhyagamyam idam kävyam utsavaḥ sudhiyam alam hata durmedhasaç casmin vidvatpriyatayā mayā. 'This poem can be understood only by a comment; it suffices that it is a feast for the clever and that the stupid come to grief in it as a result of my love of learning.' Bhamaha rather clumsily repeats in almost identical terms this verse. The list of Alam- kāras given by Bhaṭṭi is in a certain measure original, when com- pared with those of Dandin and Bhamaha; its source is still unknown. Bhaṭṭi's poem, a lamp in the hands of those whose eye is grammar, but a mirror in the hands of the blind for others, is esssentially intended to serve the double plan of describing Rāma's history and of illustrating the rules of grammar. In the latter aspect its twenty-two cantos fall into four sections; the first ¹ Ed. with Jayamañgala's comm., Bombay, 1887; with Mallinātha, BSS. 1898; i-iv ed, and trans. V. G. Pradhan, Poona, 1897. Cf. Hultzsch, EL. i. 92; Keith, JRAS. 1909, p. 435. 2 B. C. Mazumdar, JRAS. 1904, PP. 395-7; 1909, p. 759BHATTI 117 four cantos illustrate miscellaneous rules; v-ix the leading rules, x-xiii¹ are given to illustration of the ornaments of poetry, the names of the figures unfortunately being supplied merely in the commentary or the manuscripts, and the rest of the poem illus- trates the use of the moods and tenses. The combination of pleasure and profit is by no means ill devised, and Indian opinion gives Bhaṭṭi without hesitation rank as a Mahakavi. It is dubious if any sound taste can justify this position; what is true is that, considering the appalling nature of the obstacle set and the rather hackneyed theme adopted, Bhaṭṭi contrives to produce some fairly interesting and, at its best, both lively and effective verse. His aim in some degree helps his style, as it prevents the adoption of long compounds or too recondite allusions or ideas. His style may best be judged by a fragment of the scene where Rāvana in his need turns to Kumbhakarṇa for aid, and airs his aorists: najñāsis tvam sukhi Rāmo yad akārşīt sa rākṣasān udatārīd udanvantam puram naḥ parito 'rudhat vyajyotista rane çastrair anaişīd rākṣasān kṣayam. na prāvocam aham kimcit priyaṁ jāvad ajīvisam bandhus tvam arcitaḥ snehän mä dviso na vadhir mama. vīryam mã nà dadarças tvam mā na trāsthāḥ kṣatām puram. tavādrākṣma vayam vīryam tvam ajaisīḥ pură surān. 'Hast thou not known in thy happiness what Rāma hath done to the Rākṣasas? He hath crossed the ocean, and completelyhemmed in our city. He hath warred brilliantly and his weapons have brought death to the Rākṣasas. Never in all my life have I spoken one word of flattery; thou hast been honoured by me from love of kin; do not fail to slay my foes. Fail not to show thy might, fail not to guard our smitten town; thy might have we beheld, thou didst aforetime conquer the gods.' The flow of the narrative is, it will be seen, simple and limpid, but it lacks fire and colour, and the task of illustrating the figures of speech proves extremely wearisome to all but the commentators, whose joy the poet was. Some, no doubt, of the passages are happy enough; in one we find a proverb known from the Vikramorvaçi : ² ¹ x is on figures; xi on the quality of sweetness; xii on Bhāvika, vivid description; xiii gives verses which can be read as Sanskrit or Prakrit. 2 11. 16 (ed. Pandit). 118 BHĀRAVI, BHAȚŢI, KUMĀRADĀSA, AND MĀGHA Rāmo 'pi dārāharaṇena tapto: vayam hatair bandhubhir at- matulyaiḥ taptena taptasya yathāyaso naḥ: sandhiḥ pareṇāstu vimuñca Sītām. Rāma is aflame through Sītā's rape, we through the death of kinsfolk dear as ourselves; let us make compact with our foe as flaming iron with flaming iron; let Sītā go free.' Another ex- ample¹ describes Rāvana's advent: jalada iva taḍitvān prājyaratnaprabhābhiḥ: pratikakubham udasyan nisvanam dhiramandram çikharam iva Sumeror āsanaṁ haimam uccair: vividhama- nivicitram pronnataḥ so 'dhyatişthat. 'Like a lightning cloud through the rays sparkling from his jewels, and emitting like it on all sides a deep dull resonance, the lofty prince sat him on a high golden throne, radiant with many a gem, as the cloud clings to a pinnacle of mount Sumeru.' The use of viçala, broad, in the next example illustrates the straits into which a poet may be driven, even if he is a grammarian: ² kva strivişahyaāḥ karajāḥ kva vakşo: daityasya çailendra- çilāviçālam sampaçyatartad dyuşadam sunitam: bibheda tais tan nara- sinhamurtiḥ. What can finger-nails meet for maidens' breasts avail against the bosom of the demon, that is broad as a rock of the lord of mountains? Nay, consider this cunning scheme of the immortals; with these in his shape as man and lion (Viṣṇu) clove this bosom.' The chief metre used by Bhaṭṭi is the Çloka, which is used in Cantos iv-ix and xiv-xxii. Upajati of the Indravajrā type prevails in i-ii, xi and xii. The Giti form of Āryā prevails in xiii, and x is largely in Puspitāgā; no other metre has any currency of im- portance. Only Praharṣiṇī. Malini, Aupacchandasika, Vançasthā, and Vaitālīya occur six times or more; Açvalalita, Nandana, Prthvi, Rucirā, and Narkuṭaka occur only once each; others used are Tanumadhyā, Toṭaka, Drutavilambita, Pramitākṣarā, Praha- raṇakalikā, Mandākṛāntā, Çārdūlavikrīḍita, and Sragdhara. The ¹ xi. 47; imitated by Magha, i. 19. 2 xii. 59; Magha, i. 47 (below, § 4). पृष्ठम्:Sanskrit Literature.djvu/१५१ पृष्ठम्:Sanskrit Literature.djvu/१५२ पृष्ठम्:Sanskrit Literature.djvu/१५३ पृष्ठम्:Sanskrit Literature.djvu/१५४ पृष्ठम्:Sanskrit Literature.djvu/१५५ पृष्ठम्:Sanskrit Literature.djvu/१५६ पृष्ठम्:Sanskrit Literature.djvu/१५७ पृष्ठम्:Sanskrit Literature.djvu/१५८ पृष्ठम्:Sanskrit Literature.djvu/१५९ पृष्ठम्:Sanskrit Literature.djvu/१६० पृष्ठम्:Sanskrit Literature.djvu/१६१ पृष्ठम्:Sanskrit Literature.djvu/१६२ पृष्ठम्:Sanskrit Literature.djvu/१६३ पृष्ठम्:Sanskrit Literature.djvu/१६४ पृष्ठम्:Sanskrit Literature.djvu/१६५ पृष्ठम्:Sanskrit Literature.djvu/१६६ पृष्ठम्:Sanskrit Literature.djvu/१६७ पृष्ठम्:Sanskrit Literature.djvu/१६८ पृष्ठम्:Sanskrit Literature.djvu/१६९ पृष्ठम्:Sanskrit Literature.djvu/१७० पृष्ठम्:Sanskrit Literature.djvu/१७१ पृष्ठम्:Sanskrit Literature.djvu/१७२ पृष्ठम्:Sanskrit Literature.djvu/१७३ पृष्ठम्:Sanskrit Literature.djvu/१७४ पृष्ठम्:Sanskrit Literature.djvu/१७५ पृष्ठम्:Sanskrit Literature.djvu/१७६ पृष्ठम्:Sanskrit Literature.djvu/१७७ पृष्ठम्:Sanskrit Literature.djvu/१७८ पृष्ठम्:Sanskrit Literature.djvu/१७९ पृष्ठम्:Sanskrit Literature.djvu/१८० पृष्ठम्:Sanskrit Literature.djvu/१८१ पृष्ठम्:Sanskrit Literature.djvu/१८२ पृष्ठम्:Sanskrit Literature.djvu/१८३ पृष्ठम्:Sanskrit Literature.djvu/१८४ पृष्ठम्:Sanskrit Literature.djvu/१८५ पृष्ठम्:Sanskrit Literature.djvu/१८६ पृष्ठम्:Sanskrit Literature.djvu/१८७ पृष्ठम्:Sanskrit Literature.djvu/१८८ पृष्ठम्:Sanskrit Literature.djvu/१८९ पृष्ठम्:Sanskrit Literature.djvu/१९० पृष्ठम्:Sanskrit Literature.djvu/१९१ पृष्ठम्:Sanskrit Literature.djvu/१९२ पृष्ठम्:Sanskrit Literature.djvu/१९३ पृष्ठम्:Sanskrit Literature.djvu/१९४ पृष्ठम्:Sanskrit Literature.djvu/१९५ पृष्ठम्:Sanskrit Literature.djvu/१९६ पृष्ठम्:Sanskrit Literature.djvu/१९७ पृष्ठम्:Sanskrit Literature.djvu/१९८ पृष्ठम्:Sanskrit Literature.djvu/१९९ पृष्ठम्:Sanskrit Literature.djvu/२०० पृष्ठम्:Sanskrit Literature.djvu/२०१ पृष्ठम्:Sanskrit Literature.djvu/२०२ पृष्ठम्:Sanskrit Literature.djvu/२०३ पृष्ठम्:Sanskrit Literature.djvu/२०४ पृष्ठम्:Sanskrit Literature.djvu/२०५ पृष्ठम्:Sanskrit Literature.djvu/२०६ पृष्ठम्:Sanskrit Literature.djvu/२०७ पृष्ठम्:Sanskrit Literature.djvu/२०८ पृष्ठम्:Sanskrit Literature.djvu/२०९ पृष्ठम्:Sanskrit Literature.djvu/२१० पृष्ठम्:Sanskrit Literature.djvu/२११ पृष्ठम्:Sanskrit Literature.djvu/२१२ पृष्ठम्:Sanskrit Literature.djvu/२१३ पृष्ठम्:Sanskrit Literature.djvu/२१४ पृष्ठम्:Sanskrit Literature.djvu/२१५ पृष्ठम्:Sanskrit Literature.djvu/२१६ पृष्ठम्:Sanskrit Literature.djvu/२१७ पृष्ठम्:Sanskrit Literature.djvu/२१८ पृष्ठम्:Sanskrit Literature.djvu/२१९ पृष्ठम्:Sanskrit Literature.djvu/२२० पृष्ठम्:Sanskrit Literature.djvu/२२१ पृष्ठम्:Sanskrit Literature.djvu/२२२ पृष्ठम्:Sanskrit Literature.djvu/२२३ पृष्ठम्:Sanskrit Literature.djvu/२२४ पृष्ठम्:Sanskrit Literature.djvu/२२५ पृष्ठम्:Sanskrit Literature.djvu/२२६ पृष्ठम्:Sanskrit Literature.djvu/२२७ पृष्ठम्:Sanskrit Literature.djvu/२२८ पृष्ठम्:Sanskrit Literature.djvu/२२९ पृष्ठम्:Sanskrit Literature.djvu/२३० पृष्ठम्:Sanskrit Literature.djvu/२३१ पृष्ठम्:Sanskrit Literature.djvu/२३२ पृष्ठम्:Sanskrit Literature.djvu/२३३ पृष्ठम्:Sanskrit Literature.djvu/२३४ पृष्ठम्:Sanskrit Literature.djvu/२३५ पृष्ठम्:Sanskrit Literature.djvu/२३६ पृष्ठम्:Sanskrit Literature.djvu/२३७ पृष्ठम्:Sanskrit Literature.djvu/२३८ पृष्ठम्:Sanskrit Literature.djvu/२३९ पृष्ठम्:Sanskrit Literature.djvu/२४० पृष्ठम्:Sanskrit Literature.djvu/२४१ पृष्ठम्:Sanskrit Literature.djvu/२४२ पृष्ठम्:Sanskrit Literature.djvu/२४३ पृष्ठम्:Sanskrit Literature.djvu/२४४ पृष्ठम्:Sanskrit Literature.djvu/२४५ पृष्ठम्:Sanskrit Literature.djvu/२४६ पृष्ठम्:Sanskrit Literature.djvu/२४७ पृष्ठम्:Sanskrit Literature.djvu/२४८ पृष्ठम्:Sanskrit Literature.djvu/२४९ पृष्ठम्:Sanskrit Literature.djvu/२५० पृष्ठम्:Sanskrit Literature.djvu/२५१ पृष्ठम्:Sanskrit Literature.djvu/२५२ पृष्ठम्:Sanskrit Literature.djvu/२५३ पृष्ठम्:Sanskrit Literature.djvu/२५४ पृष्ठम्:Sanskrit Literature.djvu/२५५ पृष्ठम्:Sanskrit Literature.djvu/२५६ पृष्ठम्:Sanskrit Literature.djvu/२५७ पृष्ठम्:Sanskrit Literature.djvu/२५८ पृष्ठम्:Sanskrit Literature.djvu/२५९ पृष्ठम्:Sanskrit Literature.djvu/२६० पृष्ठम्:Sanskrit Literature.djvu/२६१ पृष्ठम्:Sanskrit Literature.djvu/२६२ पृष्ठम्:Sanskrit Literature.djvu/२६३ पृष्ठम्:Sanskrit Literature.djvu/२६४ पृष्ठम्:Sanskrit Literature.djvu/२६५ पृष्ठम्:Sanskrit Literature.djvu/२६६ पृष्ठम्:Sanskrit Literature.djvu/२६७ पृष्ठम्:Sanskrit Literature.djvu/२६८ पृष्ठम्:Sanskrit Literature.djvu/२६९ पृष्ठम्:Sanskrit Literature.djvu/२७० पृष्ठम्:Sanskrit Literature.djvu/२७१ पृष्ठम्:Sanskrit Literature.djvu/२७२ पृष्ठम्:Sanskrit Literature.djvu/२७३ पृष्ठम्:Sanskrit Literature.djvu/२७४ पृष्ठम्:Sanskrit Literature.djvu/२७५ पृष्ठम्:Sanskrit Literature.djvu/२७६ पृष्ठम्:Sanskrit Literature.djvu/२७७ पृष्ठम्:Sanskrit Literature.djvu/२७८ पृष्ठम्:Sanskrit Literature.djvu/२७९ पृष्ठम्:Sanskrit Literature.djvu/२८० पृष्ठम्:Sanskrit Literature.djvu/२८१ पृष्ठम्:Sanskrit Literature.djvu/२८२ पृष्ठम्:Sanskrit Literature.djvu/२८३ पृष्ठम्:Sanskrit Literature.djvu/२८४ पृष्ठम्:Sanskrit Literature.djvu/२८५ पृष्ठम्:Sanskrit Literature.djvu/२८६ पृष्ठम्:Sanskrit Literature.djvu/२८७ पृष्ठम्:Sanskrit Literature.djvu/२८८ पृष्ठम्:Sanskrit Literature.djvu/२८९ पृष्ठम्:Sanskrit Literature.djvu/२९० पृष्ठम्:Sanskrit Literature.djvu/२९१ पृष्ठम्:Sanskrit Literature.djvu/२९२ पृष्ठम्:Sanskrit Literature.djvu/२९३ पृष्ठम्:Sanskrit Literature.djvu/२९४ पृष्ठम्:Sanskrit Literature.djvu/२९५ पृष्ठम्:Sanskrit Literature.djvu/२९६ पृष्ठम्:Sanskrit Literature.djvu/२९७ पृष्ठम्:Sanskrit Literature.djvu/२९८ पृष्ठम्:Sanskrit Literature.djvu/२९९ पृष्ठम्:Sanskrit Literature.djvu/३०० पृष्ठम्:Sanskrit Literature.djvu/३०१ पृष्ठम्:Sanskrit Literature.djvu/३०२ पृष्ठम्:Sanskrit Literature.djvu/३०३ पृष्ठम्:Sanskrit Literature.djvu/३०४ पृष्ठम्:Sanskrit Literature.djvu/३०५ पृष्ठम्:Sanskrit Literature.djvu/३०६ पृष्ठम्:Sanskrit Literature.djvu/३०७ पृष्ठम्:Sanskrit Literature.djvu/३०८ पृष्ठम्:Sanskrit Literature.djvu/३०९ पृष्ठम्:Sanskrit Literature.djvu/३१० पृष्ठम्:Sanskrit Literature.djvu/३११ पृष्ठम्:Sanskrit Literature.djvu/३१२ पृष्ठम्:Sanskrit Literature.djvu/३१३ पृष्ठम्:Sanskrit Literature.djvu/३१४ पृष्ठम्:Sanskrit Literature.djvu/३१५ पृष्ठम्:Sanskrit Literature.djvu/३१६ पृष्ठम्:Sanskrit Literature.djvu/३१७ पृष्ठम्:Sanskrit Literature.djvu/३१८ पृष्ठम्:Sanskrit Literature.djvu/३१९ पृष्ठम्:Sanskrit Literature.djvu/३२० पृष्ठम्:Sanskrit Literature.djvu/३२१ पृष्ठम्:Sanskrit Literature.djvu/३२२ पृष्ठम्:Sanskrit Literature.djvu/३२३ पृष्ठम्:Sanskrit Literature.djvu/३२४ पृष्ठम्:Sanskrit Literature.djvu/३२५ पृष्ठम्:Sanskrit Literature.djvu/३२६ पृष्ठम्:Sanskrit Literature.djvu/३२७ पृष्ठम्:Sanskrit Literature.djvu/३२८ पृष्ठम्:Sanskrit Literature.djvu/३२९ पृष्ठम्:Sanskrit Literature.djvu/३३० पृष्ठम्:Sanskrit Literature.djvu/३३१ पृष्ठम्:Sanskrit Literature.djvu/३३२ पृष्ठम्:Sanskrit Literature.djvu/३३३ पृष्ठम्:Sanskrit Literature.djvu/३३४ पृष्ठम्:Sanskrit Literature.djvu/३३५ पृष्ठम्:Sanskrit Literature.djvu/३३६ पृष्ठम्:Sanskrit Literature.djvu/३३७ पृष्ठम्:Sanskrit Literature.djvu/३३८ पृष्ठम्:Sanskrit Literature.djvu/३३९ पृष्ठम्:Sanskrit Literature.djvu/३४० पृष्ठम्:Sanskrit Literature.djvu/३४१ पृष्ठम्:Sanskrit Literature.djvu/३४२ पृष्ठम्:Sanskrit Literature.djvu/३४३ पृष्ठम्:Sanskrit Literature.djvu/३४४ पृष्ठम्:Sanskrit Literature.djvu/३४५ पृष्ठम्:Sanskrit Literature.djvu/३४६ पृष्ठम्:Sanskrit Literature.djvu/३४७ पृष्ठम्:Sanskrit Literature.djvu/३४८ पृष्ठम्:Sanskrit Literature.djvu/३४९ पृष्ठम्:Sanskrit Literature.djvu/३५० पृष्ठम्:Sanskrit Literature.djvu/३५१ पृष्ठम्:Sanskrit Literature.djvu/३५२ पृष्ठम्:Sanskrit Literature.djvu/३५३ पृष्ठम्:Sanskrit Literature.djvu/३५४ पृष्ठम्:Sanskrit Literature.djvu/३५५ पृष्ठम्:Sanskrit Literature.djvu/३५६ पृष्ठम्:Sanskrit Literature.djvu/३५७ पृष्ठम्:Sanskrit Literature.djvu/३५८ पृष्ठम्:Sanskrit Literature.djvu/३५९ पृष्ठम्:Sanskrit Literature.djvu/३६० पृष्ठम्:Sanskrit Literature.djvu/३६१ पृष्ठम्:Sanskrit Literature.djvu/३६२ पृष्ठम्:Sanskrit Literature.djvu/३६३ पृष्ठम्:Sanskrit Literature.djvu/३६४ पृष्ठम्:Sanskrit Literature.djvu/३६५ पृष्ठम्:Sanskrit Literature.djvu/३६६ पृष्ठम्:Sanskrit Literature.djvu/३६७ पृष्ठम्:Sanskrit Literature.djvu/३६८ पृष्ठम्:Sanskrit Literature.djvu/३६९ पृष्ठम्:Sanskrit Literature.djvu/३७० पृष्ठम्:Sanskrit Literature.djvu/३७१ पृष्ठम्:Sanskrit Literature.djvu/३७२ पृष्ठम्:Sanskrit Literature.djvu/३७३ पृष्ठम्:Sanskrit Literature.djvu/३७४ पृष्ठम्:Sanskrit Literature.djvu/३७५ पृष्ठम्:Sanskrit Literature.djvu/३७६ पृष्ठम्:Sanskrit Literature.djvu/३७७ पृष्ठम्:Sanskrit Literature.djvu/३७८ पृष्ठम्:Sanskrit Literature.djvu/३७९ पृष्ठम्:Sanskrit Literature.djvu/३८० पृष्ठम्:Sanskrit Literature.djvu/३८१ पृष्ठम्:Sanskrit Literature.djvu/३८२ पृष्ठम्:Sanskrit Literature.djvu/३८३ पृष्ठम्:Sanskrit Literature.djvu/३८४ पृष्ठम्:Sanskrit Literature.djvu/३८५ पृष्ठम्:Sanskrit Literature.djvu/३८६ पृष्ठम्:Sanskrit Literature.djvu/३८७ पृष्ठम्:Sanskrit Literature.djvu/३८८ पृष्ठम्:Sanskrit Literature.djvu/३८९ पृष्ठम्:Sanskrit Literature.djvu/३९० पृष्ठम्:Sanskrit Literature.djvu/३९१ पृष्ठम्:Sanskrit Literature.djvu/३९२ OTHER CASES OF CONTACT BETWEEN EAST AND WEST 361 have usually Indian parallels ; that of the ichneumon is taken from the Pancatantra, and the others are often specimens of women's tricks to cover their infidelities, which are common in India, forming as it were a supplement of the Pancatantra. The Greek Syntipas contains various'passages which can only be read successfully by recognizing that they are merely corruptions of a Sanskrit original, and everything supports the conclusion that we have here another case of an Arabic original rendered from a Pahlavi translation of a Sanskrit text. It is natural to extend the doctrine and to find the original of the Arabian Thousand and One Nights in India, 1 and something substantial has been done in this direction by proving that the prologue and setting of the tales are a contamination of motifs which are quite well known in India. Thus we have the Jain legend of Kanakamanjari, who retains for six months the un- divided love of the king by the device of beginning a tale each night but not finishing it. Again, we have in a Chinese render- ing of a Buddhist tale (a.d. 251), in the Kathasaritsagara, and in Hemacandra, variations of the theme of the man who is utterly depressed by finding out that his wife is unfaithful, but recovers happiness because he discovers that the king himself is equally being made a mock of. The further adventure of Shahriar and Shahzeman has a parallel in the K athasaritsagara. There are other traces of Indian influence in the tales, and it is clear that it is impossible to ascribe them to borrowing from Persia ; trans- lations from Persian into Sanskrit are normally late, as in Crlvara's Kathakautuka 2 written on the theme of Yusuf and Zuleikha under Zainu-l-'Abidln in the fifteenth century. The only matter that can be in doubt is the extent of the influence ; certainly there is nothing in this case to prove the taking over of a whole cycle of stories from an Indian work, now lost. In Europe, apart from the translations enumerated, traces of real Indian origin are hard to prove. 3 A Carolingian poem of the 1 Cosquin, op. cit., pp. 365 ff. ; Przyluski, JA. ccv. 101 ff., who finds in the Svayarhvara of India a relic of the Austroasiatic festival dance at which young people were mated. Cf. Macdonald, JRAS. 1924, pp. 353 ff. a Ed. and trans. R. Schmidt (Kiel, 1898). 3 Giinter, Buddha, pp. 99 ff. The famous tale of the poison maiden in Indian literature and in the west — told of Aristotle and Alexander in the Secretum Secretcrum (cf. Hawthorne, Rappaccin? s Daughter), is discussed by Penzer, Ocean of Story, ii. 31 1 ff. पृष्ठम्:Sanskrit Literature.djvu/३९४ पृष्ठम्:Sanskrit Literature.djvu/३९५ पृष्ठम्:Sanskrit Literature.djvu/३९६ पृष्ठम्:Sanskrit Literature.djvu/३९७ पृष्ठम्:Sanskrit Literature.djvu/३९८ पृष्ठम्:Sanskrit Literature.djvu/३९९ पृष्ठम्:Sanskrit Literature.djvu/४०० पृष्ठम्:Sanskrit Literature.djvu/४०१ पृष्ठम्:Sanskrit Literature.djvu/४०२ पृष्ठम्:Sanskrit Literature.djvu/४०३ पृष्ठम्:Sanskrit Literature.djvu/४०४ पृष्ठम्:Sanskrit Literature.djvu/४०५ पृष्ठम्:Sanskrit Literature.djvu/४०६ पृष्ठम्:Sanskrit Literature.djvu/४०७ पृष्ठम्:Sanskrit Literature.djvu/४०८ पृष्ठम्:Sanskrit Literature.djvu/४०९ पृष्ठम्:Sanskrit Literature.djvu/४१० पृष्ठम्:Sanskrit Literature.djvu/४११ पृष्ठम्:Sanskrit Literature.djvu/४१२ पृष्ठम्:Sanskrit Literature.djvu/४१३ पृष्ठम्:Sanskrit Literature.djvu/४१४ पृष्ठम्:Sanskrit Literature.djvu/४१५ पृष्ठम्:Sanskrit Literature.djvu/४१६ पृष्ठम्:Sanskrit Literature.djvu/४१७ पृष्ठम्:Sanskrit Literature.djvu/४१८ पृष्ठम्:Sanskrit Literature.djvu/४१९ पृष्ठम्:Sanskrit Literature.djvu/४२० पृष्ठम्:Sanskrit Literature.djvu/४२१ पृष्ठम्:Sanskrit Literature.djvu/४२२ पृष्ठम्:Sanskrit Literature.djvu/४२३ पृष्ठम्:Sanskrit Literature.djvu/४२४ पृष्ठम्:Sanskrit Literature.djvu/४२५ पृष्ठम्:Sanskrit Literature.djvu/४२६ पृष्ठम्:Sanskrit Literature.djvu/४२७ पृष्ठम्:Sanskrit Literature.djvu/४२८ पृष्ठम्:Sanskrit Literature.djvu/४२९ पृष्ठम्:Sanskrit Literature.djvu/४३० पृष्ठम्:Sanskrit Literature.djvu/४३१ पृष्ठम्:Sanskrit Literature.djvu/४३२ पृष्ठम्:Sanskrit Literature.djvu/४३३ पृष्ठम्:Sanskrit Literature.djvu/४३४ पृष्ठम्:Sanskrit Literature.djvu/४३५ पृष्ठम्:Sanskrit Literature.djvu/४३६ पृष्ठम्:Sanskrit Literature.djvu/४३७ पृष्ठम्:Sanskrit Literature.djvu/४३८ पृष्ठम्:Sanskrit Literature.djvu/४३९ पृष्ठम्:Sanskrit Literature.djvu/४४० पृष्ठम्:Sanskrit Literature.djvu/४४१ पृष्ठम्:Sanskrit Literature.djvu/४४२ पृष्ठम्:Sanskrit Literature.djvu/४४३ पृष्ठम्:Sanskrit Literature.djvu/४४४ पृष्ठम्:Sanskrit Literature.djvu/४४५ पृष्ठम्:Sanskrit Literature.djvu/४४६ पृष्ठम्:Sanskrit Literature.djvu/४४७ पृष्ठम्:Sanskrit Literature.djvu/४४८ पृष्ठम्:Sanskrit Literature.djvu/४४९ पृष्ठम्:Sanskrit Literature.djvu/४५० पृष्ठम्:Sanskrit Literature.djvu/४५१ पृष्ठम्:Sanskrit Literature.djvu/४५२ पृष्ठम्:Sanskrit Literature.djvu/४५३ पृष्ठम्:Sanskrit Literature.djvu/४५४ पृष्ठम्:Sanskrit Literature.djvu/४५५ पृष्ठम्:Sanskrit Literature.djvu/४५६ पृष्ठम्:Sanskrit Literature.djvu/४५७ पृष्ठम्:Sanskrit Literature.djvu/४५८ पृष्ठम्:Sanskrit Literature.djvu/४५९ पृष्ठम्:Sanskrit Literature.djvu/४६० पृष्ठम्:Sanskrit Literature.djvu/४६१ पृष्ठम्:Sanskrit Literature.djvu/४६२ पृष्ठम्:Sanskrit Literature.djvu/४६३ पृष्ठम्:Sanskrit Literature.djvu/४६४ पृष्ठम्:Sanskrit Literature.djvu/४६५ पृष्ठम्:Sanskrit Literature.djvu/४६६ पृष्ठम्:Sanskrit Literature.djvu/४६७ पृष्ठम्:Sanskrit Literature.djvu/४६८ पृष्ठम्:Sanskrit Literature.djvu/४६९ पृष्ठम्:Sanskrit Literature.djvu/४७० पृष्ठम्:Sanskrit Literature.djvu/४७१ पृष्ठम्:Sanskrit Literature.djvu/४७२ पृष्ठम्:Sanskrit Literature.djvu/४७३ पृष्ठम्:Sanskrit Literature.djvu/४७४ पृष्ठम्:Sanskrit Literature.djvu/४७५ पृष्ठम्:Sanskrit Literature.djvu/४७६ पृष्ठम्:Sanskrit Literature.djvu/४७७ पृष्ठम्:Sanskrit Literature.djvu/४७८ पृष्ठम्:Sanskrit Literature.djvu/४७९ पृष्ठम्:Sanskrit Literature.djvu/४८० पृष्ठम्:Sanskrit Literature.djvu/४८१ पृष्ठम्:Sanskrit Literature.djvu/४८२ पृष्ठम्:Sanskrit Literature.djvu/४८३ पृष्ठम्:Sanskrit Literature.djvu/४८४ पृष्ठम्:Sanskrit Literature.djvu/४८५ पृष्ठम्:Sanskrit Literature.djvu/४८६ पृष्ठम्:Sanskrit Literature.djvu/४८७ पृष्ठम्:Sanskrit Literature.djvu/४८८ पृष्ठम्:Sanskrit Literature.djvu/४८९ पृष्ठम्:Sanskrit Literature.djvu/४९० पृष्ठम्:Sanskrit Literature.djvu/४९१ पृष्ठम्:Sanskrit Literature.djvu/४९२ पृष्ठम्:Sanskrit Literature.djvu/४९३ पृष्ठम्:Sanskrit Literature.djvu/४९४ पृष्ठम्:Sanskrit Literature.djvu/४९५ पृष्ठम्:Sanskrit Literature.djvu/४९६ पृष्ठम्:Sanskrit Literature.djvu/४९७ पृष्ठम्:Sanskrit Literature.djvu/४९८ पृष्ठम्:Sanskrit Literature.djvu/४९९ पृष्ठम्:Sanskrit Literature.djvu/५०० पृष्ठम्:Sanskrit Literature.djvu/५०१ पृष्ठम्:Sanskrit Literature.djvu/५०२ पृष्ठम्:Sanskrit Literature.djvu/५०३ पृष्ठम्:Sanskrit Literature.djvu/५०४ पृष्ठम्:Sanskrit Literature.djvu/५०५ पृष्ठम्:Sanskrit Literature.djvu/५०६ पृष्ठम्:Sanskrit Literature.djvu/५०७ पृष्ठम्:Sanskrit Literature.djvu/५०८ पृष्ठम्:Sanskrit Literature.djvu/५०९ पृष्ठम्:Sanskrit Literature.djvu/५१० पृष्ठम्:Sanskrit Literature.djvu/५११ पृष्ठम्:Sanskrit Literature.djvu/५१२ पृष्ठम्:Sanskrit Literature.djvu/५१३ पृष्ठम्:Sanskrit Literature.djvu/५१४ पृष्ठम्:Sanskrit Literature.djvu/५१५ पृष्ठम्:Sanskrit Literature.djvu/५१६ पृष्ठम्:Sanskrit Literature.djvu/५१७ पृष्ठम्:Sanskrit Literature.djvu/५१८ पृष्ठम्:Sanskrit Literature.djvu/५१९ पृष्ठम्:Sanskrit Literature.djvu/५२० पृष्ठम्:Sanskrit Literature.djvu/५२१ पृष्ठम्:Sanskrit Literature.djvu/५२२ पृष्ठम्:Sanskrit Literature.djvu/५२३ पृष्ठम्:Sanskrit Literature.djvu/५२४ पृष्ठम्:Sanskrit Literature.djvu/५२५ पृष्ठम्:Sanskrit Literature.djvu/५२६ पृष्ठम्:Sanskrit Literature.djvu/५२७ पृष्ठम्:Sanskrit Literature.djvu/५२८ पृष्ठम्:Sanskrit Literature.djvu/५२९ पृष्ठम्:Sanskrit Literature.djvu/५३० पृष्ठम्:Sanskrit Literature.djvu/५३१ पृष्ठम्:Sanskrit Literature.djvu/५३२ पृष्ठम्:Sanskrit Literature.djvu/५३३ पृष्ठम्:Sanskrit Literature.djvu/५३४ पृष्ठम्:Sanskrit Literature.djvu/५३५ पृष्ठम्:Sanskrit Literature.djvu/५३६ पृष्ठम्:Sanskrit Literature.djvu/५३७ पृष्ठम्:Sanskrit Literature.djvu/५३८ पृष्ठम्:Sanskrit Literature.djvu/५३९ पृष्ठम्:Sanskrit Literature.djvu/५४० पृष्ठम्:Sanskrit Literature.djvu/५४१ पृष्ठम्:Sanskrit Literature.djvu/५४२ पृष्ठम्:Sanskrit Literature.djvu/५४३ पृष्ठम्:Sanskrit Literature.djvu/५४४ पृष्ठम्:Sanskrit Literature.djvu/५४५ पृष्ठम्:Sanskrit Literature.djvu/५४६ पृष्ठम्:Sanskrit Literature.djvu/५४७ पृष्ठम्:Sanskrit Literature.djvu/५४८ पृष्ठम्:Sanskrit Literature.djvu/५४९ पृष्ठम्:Sanskrit Literature.djvu/५५० पृष्ठम्:Sanskrit Literature.djvu/५५१ पृष्ठम्:Sanskrit Literature.djvu/५५२ पृष्ठम्:Sanskrit Literature.djvu/५५३ पृष्ठम्:Sanskrit Literature.djvu/५५४ पृष्ठम्:Sanskrit Literature.djvu/५५५ पृष्ठम्:Sanskrit Literature.djvu/५५६ पृष्ठम्:Sanskrit Literature.djvu/५५७ पृष्ठम्:Sanskrit Literature.djvu/५५८ पृष्ठम्:Sanskrit Literature.djvu/५५९ पृष्ठम्:Sanskrit Literature.djvu/५६० पृष्ठम्:Sanskrit Literature.djvu/५६१ पृष्ठम्:Sanskrit Literature.djvu/५६२ पृष्ठम्:Sanskrit Literature.djvu/५६३ पृष्ठम्:Sanskrit Literature.djvu/५६४ पृष्ठम्:Sanskrit Literature.djvu/५६५ पृष्ठम्:Sanskrit Literature.djvu/५६६ पृष्ठम्:Sanskrit Literature.djvu/५६७ पृष्ठम्:Sanskrit Literature.djvu/५६८ पृष्ठम्:Sanskrit Literature.djvu/५६९ पृष्ठम्:Sanskrit Literature.djvu/५७० पृष्ठम्:Sanskrit Literature.djvu/५७१ पृष्ठम्:Sanskrit Literature.djvu/५७२ पृष्ठम्:Sanskrit Literature.djvu/५७३ पृष्ठम्:Sanskrit Literature.djvu/५७४ पृष्ठम्:Sanskrit Literature.djvu/५७५ पृष्ठम्:Sanskrit Literature.djvu/५७६ पृष्ठम्:Sanskrit Literature.djvu/५७७ पृष्ठम्:Sanskrit Literature.djvu/५७८ पृष्ठम्:Sanskrit Literature.djvu/५७९ पृष्ठम्:Sanskrit Literature.djvu/५८० पृष्ठम्:Sanskrit Literature.djvu/५८१ पृष्ठम्:Sanskrit Literature.djvu/५८२ पृष्ठम्:Sanskrit Literature.djvu/५८३ पृष्ठम्:Sanskrit Literature.djvu/५८४ पृष्ठम्:Sanskrit Literature.djvu/५८५ पृष्ठम्:Sanskrit Literature.djvu/५८६ पृष्ठम्:Sanskrit Literature.djvu/५८७ पृष्ठम्:Sanskrit Literature.djvu/५८८ पृष्ठम्:Sanskrit Literature.djvu/५८९ पृष्ठम्:Sanskrit Literature.djvu/५९० पृष्ठम्:Sanskrit Literature.djvu/५९१ पृष्ठम्:Sanskrit Literature.djvu/५९२ पृष्ठम्:Sanskrit Literature.djvu/५९३ पृष्ठम्:Sanskrit Literature.djvu/५९४ पृष्ठम्:Sanskrit Literature.djvu/५९५ पृष्ठम्:Sanskrit Literature.djvu/५९६ पृष्ठम्:Sanskrit Literature.djvu/५९७ पृष्ठम्:Sanskrit Literature.djvu/५९८ पृष्ठम्:Sanskrit Literature.djvu/५९९ पृष्ठम्:Sanskrit Literature.djvu/६०० पृष्ठम्:Sanskrit Literature.djvu/६०१ पृष्ठम्:Sanskrit Literature.djvu/६०२ पृष्ठम्:Sanskrit Literature.djvu/६०३ पृष्ठम्:Sanskrit Literature.djvu/६०४ पृष्ठम्:Sanskrit Literature.djvu/६०५