= ¢ ९१ —— LN BIBLIOTHECA INDIGA A COLLECTION OF ORIENTAL\\ VARNA-RATNAKAR? OF JYOTIRISVARA-KAVISEKHARACARYA EDITED WITH ENGLISH AND MAITHILI INTRODUCTIONS AND INDEX VERBORUM BY SUNITI KUMAR CHATTERJI, M.A., D.Lrr., F.R.A.8.B. CIE RARY ecretary of the Royal Asiatic Soctety of Bengal Khatra Professor of Indian Linguistics and Phonetics, Calcutta University, hilological F AND BABUA MISRA, JYAUTISA-TIRTHA, JYOTISACARYA Lecturer in the Departments of Ancient Indian History and Culture and odern Indian Languages, now Research Fellow, Caloutta University Work Issue Number Number 1540 ` , 262 New Series (Complete Work) CALCUTTA : Printed at the Baptist Mission Press Published by the Royal Asiatic Society of Bengal, 1, Park Street 1940 Price Rs.5/-. ` . NOTICE BIBLIOTHECA INDICA PUBLISHED BY THE ROYAL ASIATIC SOCIETY OF BENGAL Tux Bibliotheca Indica is a collection of works belonging to or treating of Oriental literatures and contains original toxt editions as well as translations into English, and also bibliographies, dic- tionarios, grammars, and studies. The publication was started in 1849, and consists of en Old and a New Series. The New Series was begun in 1860, and is still running, The issues in the series consisted originally of fascicles of 96 or 100 pages print, though occasionally numbers were issued of double, triple or larger bulk, and in a few cases even entire works were published under a single issue number. Of late years the single issues are made as much as possible to constitute complete volumes. Several different works are always simultaneously in progress. Each issue bears a consecutive issue number. The Old Series consists of 265 issues; in the New Series, till January Ist, 1940, inclusive, 1,534 issues have been published. These 1,799 issues represent 258 different works; these works again , represent the following literatures :— : Sanskrit, Prakrit. Rajasthint, Kashmiri, Hindi. Tibetan. Arabic, Persian, Several works published are partly or wholly sold out, 0४08 are still incomplete and in progress. Two price-lists concerning the Bibliotheca Indica are available and may be had on application. One describes the Indian and the other the Islamic works published in the series. These lists are periodically revised. The standard sizes of the Bibliotheca Indica are :— Royal (or large) octavo. Quarto. Single issues may be bought separately, but three years after the completion of a work no complete sets are broken for the sale of loose component parts. pi BLIOTHECA INDICA Work No, 262? VARNA-RATNAKABA EARLY MALPPHILE TEXT WIPE ENGLISH INTRODUCTION AND INDEN VERBORUM ( ८९५, 314 ५ | VARNA-RATNAK [ना <>: EDITED WITH ENGLISH AND MAITHILI INTRODUCTIONS AND INDEX VERBORUM BY | धपा KUMAR CHATTERJI, M.A., D.Lir., F.R.A.S.B. ‘hatra Professor of Indian Linguistics and Phonetics, Calcutta University, and Philological Secretary of the Royal Asiatic Socicty of Bengal “AND हि BABUA MISRA, JYAUTISA-TIRTHA, JYOTISACARY A ‘ormerly Lecturer in the Departments of Anctent Indian History and Culture and Modern Indian Languages, now Research Fellow, Calcutta University. PRINTED AT THE BAPTIST MISSION PRESS PUBLISHED BY THE KOYAL ASIATIC SOCIETY OF BENGAL CALCUTTA 1940 Printed by P. Knight, Baptist Mission Press, 41a, Lower Circular Road, Calcutta SL moe \9490 उ ४० 2. gat: भियिष्लाभाषाक दं प्राचोमतम उपलब्ध war मेयिलोक उच्नतिक रकमाच ख।भाखस विविध-विरद्‌ावलशो-विराजमान AAA महाराजाधिराज सर्‌ Stary कामेश्र सिंह बहादुर के-सौ-ष्याद्र-र, रल-णल्‌-डो, डो-लिट्‌-क करकमशमे माभावा मेथिक्लोक प्रति wae प्रेमक चेतु खाना SF खोकर एषटपोषकता करबाक Vay UAT AUS साद्र समपित। To THe Hon’BiE MAHARAJADHIRAJA SIR KAMESHWAR SINGH K.C.LE., LL.D., D.Litt. OF DARBHANGA this First Edition of the Varna-ratndékara, the Oldest Text in his Mothor-Tongue so far obtained, is respectfully dedicated by the Editors in Appreciation of his enlightened Patronage of the Cause of Maithilf and its Study in its Home Province. CONTENTS. Page Dedication... 9 4 7 6 ष Gontents ४ 8 ६ = त vii English Introduction = .. - = .. ix-lxiv 1. The Manuscript .. क ix 2. The Author—His Works, His Date and His Times .. 2111 3. The Work, its Subject-Matter, and its General Interest Xxi 4. The Language of the Varna-ratnakara .. .. XXXvi (1) Orthography, Phonetics and Phonology ,, XXXvili (11) Morphology és iu .. Xxivi (III) Syntax... ‘2 “3 = lix (1४) The Vocabulary ee ॐ = Ix Maithilf Introduction ,., = ४ ९, १.८ Maithill Text : ॥ व्देरलाकरः | 7 त .. १-७ Index Verborum: 4 शब्दस्च ॥ ६ a , , 9११९९ INTRODUCTION. 1. Tae MaANvscript. The ‘ Varna-ratnakara’ is the oldest work in the Maithili language of North Bihar so far known, and it goes back to the Ist half, perhaps to the Ist quarter, of the 14th century. It is preserved in a unique MS. on palm-leaf now in the library of the Royal Asiatic Socioty of Bengal, in its Government Collection of MSS. (No. 48/34). The MS. is written in old Maithili characters. It originally contained 77 folia, but 17 of these are now missing (folia 1 to 9 at the beginning; folia 11 and 12, 14 and 15, 17, 19, 26 and 27). The leaves measure fiom 15” long by 1एˆ to 2” broad, fairly well preserved, but some pages are slightly damaged by white ants. Generally there are five lines on each page, but there are a few pages with four lines and a few more with six. The writing is fairly legible. The work now begins with page 10a. Fortunately the last page (77b) giving the colophon hae been preserved, and we know from it that the MS. was written in the year 388 of the La Sam era which is still current in Mithila: this corresponds to 1507 A.C., the La Samm having commenced from 1119 A.C. The presont MS. has been copied from two other MSS., both of which seem to have been imperfect. or incompleto. At least one certainly was. The work is in several chapters called kallolas. The seventh kallola in the present MS. ends १६ page 69b, line 3, after which the eighth apparently commences, and it is continued to p. 70b, where, while the work is in the course of enumerating various kinds of ° ships and boats, it breaks off abruptly at line 4. A verse in Sardila- vikridita metre, with a number of errors and omissions, then follows, indicating the termination of the work: यावन्नि (=m) [र] धिनण्डिनौ मुररिपोवचःखरं arwt यावद्निजं( = भं efor x स [म] सुकं रलो विन्दति ATTRA TAT भू( = भ )वनन्युदोतयने कराः कायं त्रौ afrta = ख)रस्य दुधियां तागत्कुषौ टत्वम्‌ 1 after which comes this statoment, in corrupt Sanskrit: आद्‌ १ प्रन्दनेकं GATH (,) fetrra afavafefenaafa तक्ष्यते ॥ We have after that several other varnands or descriptions, beginning with राच्थव्णा, up fo the end (page 77b), and again the work is made to end, without the expected title of the last kallola. The verse arew Ganga-deva > Narasimha-deva, This Narasimha would be contemporaneous with the first Turki invasion of Eastern India under the lieutenants of the Slave Kings of Delhi. His date, as given in the ‘ Mithilé-Darpan’ (p. 62), which partly relies on traditional evidence, is 1149-1201 A.C. An invasion of Mithila by the Turks is in the probability of things during his rule, but the earliest Mohammadan authorities beginning with Minhaju-d-Din (second half of the 13th century) do not say anything about it, although they give enough details in their accounts of the fortunes of the Turks and of Mohammadan arms in South Bibar, in Bengal, in Assam and in Orissa. Besides, the V.R. of Jyotiriévara shows a numbor of naturalised Persian words (see infra, the section of the language of the V.R., § 60); and these from their nature could be adopted and popularised in the Maithili speech only after a century’s contact with the Persian-using Turks. From this, an earlier Narasimha- deva of whom we know the name only, with a problematic conflict with the Turks about 1200 A.C., cannot be regarded as the patron of our poet, especially when wo have the well-attested Harasimha-deva about whom we know through evidence from various quarters something fairly definite. One bit of traditional information about Harasimha-deva we find from Vidyapati’s ‘Puruga-pariksa’, which is a collection of short contes édifiantes in Sanskrit. In the section under the rubric we मोतविद्यकथा in Chapter III of the book, there is given the story of a singer from Mithila (Tirabhukti) called Kalanidhi, who went to the court of king Udayasimha of Goraksa-nagara, and the latter was highly pleased‘ with him and gave him much wealth: which made the local talenta ‘angry, and they challenged Kalanidhi to a contest in singing, ऋ the king as umpire. Kalanidhi refused this arbitration, and in -exouse said that the only mortal who was a judge of music and singing 2B INTRODUCTION xix after Siva himself was Harisimha (Harasimha), evidently of his own ‘and, Mithila; and now that Harasimha was no more, only Siva could properly act as judge: इटो वा इरि (इर) -सिंशो वा गोतिविद्याविश्ारदो। इरि (दर )-मिंङ मते wa मौतिवित्‌ केवलं इरः ॥ (I am indebted to Mr. Amaranatha Jha of the University of Allahabad and to Pandit Babua Miéra of the University of Calcutta for the story and the quotation.) This story ives a good sidelight into the accomplishments of Harasimbha: that music and singing were well patronised in his court wo can easily infer trom the fact of Jyotiriévara taking pains to vaunt his accomplish- ment in it in the ‘Pafica-séyaka’ and the ‘Dhirtta-samagama’, and from the elaborate accounts of the musicians and singers with their cortéges which we find in the V.R. Besides, Jyotiriévara is not an isolated figure in the history of literary culture in Mithila. According to a tradition current in Mithila, which has been given by Mr. Nagendranath Gupta in the Introduction to his edition of the Poems of Vidyapati (published by the Vangiya Sahitya Parigad, Caloutta, Bengali year 1316, p. vi), Jyotiriévara was a cousin of the grandfather of Vidyapati (c. 1400) ; ५0 that the former could easily have flourished in the varly part of the 14th century. He was evidently a momber of a great family of Sanskrit scholars who flourished in Mithila in the 13th and 14th centuries and - whose names are great in the smrls and other later Sanskrit litorature. Candéévara Thakkura, the greatest name in smyrli in Mithilé, was a scion of this family. He was a minister of Harasimha-deva, and conquored Nepal for him, and he was at the samo time the author or compiler of a great digest, the’ Smrti-ratnakara’ in seven sections. The family troes run as follows: Tripathin Karmaditya Thakkura Raimedvara | Devaditya Dhiréfvara (Sandhivigrahika ) | | Jyotiriévara | | Thakkura Virésvara Ganésvara Jayudatta Candéévara Ramadatta Ganapati Thakkura Vid vapati (cf. Manomohan Chakravarti in the J.A.8.B., 1915, p. 385; Nagendranaéth Gupta, Introduction to the Pooms of Vidyapati, pp. vi-vii; G. A. Griorson, ‘Modern Vernacular Literature of Hindustan e p. 9). XxX INTRODUCTION The exact relationship between Jyotiriévara and the sons of Devaditya is not known. For one thing, Jyotiriévara, it is said, has not been mentioned in the ‘ Pafijis’ or Genealogical Lists of the Mithila Brahmans and members of the other high castes, and these ‘ Pafijis’ are stated to have started under the auspices of Harasimha-deva from the Saka year 1232 (= 1311 A.C.) (cf. ‘Mithilé-Darpan’, 1, p. 206). This has induced one or two friends from Mithilé with whom I discussed this matter to think that Jyotiriévara must have flourished before the establishment of the ‘Pafijis’, and preferably during the reign of Narasimha-deva. But the evidence against that is found in the work V.R. itself in its Persian words; and the absence of ४ positive statement in the ‘Pafijis’, the authority of which can be challenged, as in the case of the genealogical treatises of an uncritical epoch in all countries, should not overweigh other evidence in favour of the first quarter of the 14th century for the date of Jyotiriévara. It is significant that the great smrtt work of Candéévara is called a ‘Ratnakara’, as also its component parts—the ‘Krtya-ratnakara’, the ‘Déana-ratnikara’, the ‘Vivadda-’, ‘Vyavahara-’, ‘Suddhi’., ‘Gphastha-’ and ‘Pijaé-ratnakaras’. The ‘Smpti-ratnaékara’ with its seven component parts seems to have heen compiled between 1315-30 A.C. (J.A.S.B., 1915, p. 386). The ‘Varna-ratnakara’ may have heen earlier than the ‘Smpti-ratnékara’; or by adopting ratndkara in the title of his vernacular work, did the author intend to pay a compliment to his great kinsman, CandéSvara, who was a statesman, scholar, jurist and possibly also a general at the same time ? The period during which Jyotiriévara flourished was the golden age of Sanskrit studies in Mithilé of post-Mohammadan times. Barring the episode of the Moslem incursion, the normal life in the land of Mithilé seems to have been a peaceful and ४ happy one, when the scholars could carry on their literary activities without any hindrance. The 14th century was a very important one in the history of empti studies in Mithilaé, under both the Karnata kings and the kings of the family of Kaméévara. If it were a period of Mohammadan conquest, sweeping away the native dynasty, we could not have expected this flourishing state of Sanskrit studies in the land. The references to the fight with the Moslem invaders in both Cand6évara and Jyotiriévara have a note of exultation, which shows that it did not permanently or seriously affect the normal life of the Hindu state, at least in its inner life. ` The vernacular of the land was not neglected by the scholars of Mithila: if the masses bad their ballads about Lorik, as at the present day, scholars seem to have found pleasure in compositions inspired by sanskrit models and guided by the rules of Sanskrit poetics. Two vnerations after Jyotiriévara came Vidyapati, the greatest lyric poet of Eastern India, with only Candidisa of Bengal as his rival, till modern times. Mithilé was the resort of Sanskrit students from Bengal for some three hundred years after the conquest of the latter provinoe by the Turks. She was the teacher and inspirer of Bengal in Sanskrit learning, in smptt and specially in nydya. Bengali scholars would come back home after finishing their studies in Mithila not only with Sanskrit learning in their heads, but also with Maithili songs on their lips—songs by Vidydpati, and also probably by his predecessors and his successors. These were adopted by the Bengali people, and they yave a new literary model and a new literary dialect, the Brajabuli, to Bengal. The Maithili lyric similarly naturalised itself in Assam and in Orissa during the 15th century. At the head of this important Maithili literature stands Jyotiriévara Thakkura, Considering the loving care shown by the scholars of early Mithila for their mother tongue, it is sad to see the neglect of this highly cultured language among its present-day speakers, both scholars and others. 3. THe Work, Its SuspsecT Matrer, AND [TS GENERAL INTEREST. Pandit Haraprasada’s note roughly indicates the subject-matter of the work. It is a sort of lexicon of vernacular and Sanskrit terms, ५ repository of literary similes and conventions dealing with the various things in the world and ideas which are usually treated in poetry. We have in it either bare lists of terms, or the similes and conventions are set in the frame-work of a number of ‘descriptions’. The work is in prose. Thore were apparently more than seven chaptors, probably there were cight. The chapters are suitably called kallolas, ‘ streams’ or ‘waves’, as the work is a ratzdkara or ‘sea’. In each kallola there aro a number of these lists of terms and conventional similes: each of these lists, or descriptions, is preceded by the formula—wu...... वदना. Kach kallola has at its end its name, together with the namo of the author und the title of the work. In these colophons to the chapters, the name of the book has been consistently given as * Varna-ratnakara , thief कविरोकराययग्रोष्योतिरोश्चरविरचितवथरल्याकरे भमरववनो नाम प्रथमः were: ॥ Pandit Haraprasida, however, bas always referred to the work ae ‘Varnana-ratnakara’, The rubric tarnand preceding each list or description certainly affords scope for this emeridation, for varna might easily be a scribal error for varnana, But it is better to keep the name given regularly at the end of each kallola ofthe work. Varna of course does not mean ‘description’. But there xxii INTRODUCTION is one sense. of it found in medieval Sanskrit which would seem to apply in this case. Among other things, varna means, according to Hemacandra, Halayudha and Mallinatha (cf. Bébtlingk and Roth’s St. Petersburg Lexicon), gita-krama, i.e., ‘the order or arrangement of asongora poem’. The work, it thus appears, is not so much an artistic composition in itself as a collection of clichés, ready-made material, to be utilised in an artistic composition. The purpose in writing the book is not to compose a descriptive poem in Maithili: the varnand or varnana of the various subjects in itself was not the aim. Rather, the aim was to furnish lists of things one must mention in describing these subjects—of things in their proper order and proper setting (cf. the meaning gita-krama for varna). The obligation to follow the accepted conventions was imposed upon the writer by the rules of rhetoric. Tho descriptions or the descriptive parts in the work are frequently very summary, and at times they are nothing but a mere string of comparisons. But the connected objects are fully enumerated, and the order of events in narrating a process is given in full. Fre- quently the author gives no description, but merely a number of names of connected objects to which reference is expected to be made in fully describing something. Examples will be found below. The utility of such a work appears to have been that of a hand- book of poetical figures and a lexicon of connected topics and objects. It is a book of ready-made patter—often highly poetic no doubt—in the approved, orthodox Sanskrit style, on a variety of topics, which would form the stock-in-trade of a Public Reciter—a Kathaka, as he is called in Bengal, or a Vydsa, as he is called in Hindustén—who would cleverly bring it all in to embellish his narration of the Ramayana or the Mahabharata, or the stories from the Puranas. An aspiring poet in Sanskrit or the vernacular might freely draw from this literary vade mecum. From a survey of the form and contents of the V.R., its author does not seem to have had any other aim than providing a book of ‘order or arrangement’ in describing things 111 ` a poem. The author was an accomplished Sanskrit scholar well-read in literature, and was a successful writer, too, in Sanskrit. If he wanted seriously to compose in Maithili, we could expect something really artistic from him. But frequently the book is nothing but an enumeration of names,—cyclopsedic in character, no doubt, but a oyclopsedia is not literature. The author’s knowledge and reading were Yharvellous. Much of his material he got ready-made from Sanskrit books, e.g., the names of the 18 Pur&nas, the 49 Winds, the 12 Adityas, the 18 Chaste Wives of Legend, the 36 Weapons of War, the various movements in dancing, and so forth; but in other cases, INTRODUCTION xxiii in dealing with topics not found in the Sanskrit, and in making lists of vernacular terms, ¢.g., in enumerating the moves in gambling and in the game of chess, the passes in shampooing, the kinds of crocodiles, of trees, of flowers, in describing boat-gear,—the author undoubtedly had to go in for a considerable amount of personal research, Such books of collections of terms and of aids in their profession for the professional reciter are not unknown in other parts of India. In the sister province of Bengal, such Kathaka's hand-books have been found: and it would be interesting to compare their method and their contents with those of the ‘Varna-ratnakara’. The late Rai Bahadur Dr. Dinesh Chandra Sen, in his ‘History of Bengali Language and Literature’ (Calcutta University, 1911, pp. 685- 588) gave a brief account of the methods of the Kuthakas of Bengal in narrating the Pauranik tales. Ho says: ‘there are formuley which every kathaka has to get by heart,—set passages describing not only Civa, Laksmi, Visnu, Krisna, and other deities, but. also describing a town, a battlefield, morning, noon and night, and many other subjects which incidentally occur in the course of the narration of a story. These set passages are composed in Sanskritic Bongali with ५ remark- able jingle of consonances the effect of which is quite oxtraordinary’. Dr. Sen mentions a book of formule, supplied to him by a Kathaka, in which there are set passages on the following subjects: 1. A city. 2. Noon-day. 3. Morning. 4. Night. 5. A cloudy day, 6. Woman’s beauty. 7. The Sage Narada. 8. Visnu. 9. Rama. 10 Laksmana. 11. Siva. 12. Kali. 13. Sarasvatl. 14. Laksmi. 15. A forest. 16. War. 17. Bhagavati. It may be just mentioned here, that, except the descriptions of the deitios, almost all these subjects have been treated in the V.R. Dr. Sen has given specimens of these set passages. They aro rather more elaborate than whut we find in the V.R., more finished and artistic. That is only natural, as the Bengali work is late, belonging apparently to the 19th century. The device has been perfected, but the spirit nevertheless is the same in both. Dr. Sen believed that the manner of delivering stories, with set formule and all, which is followed by tho present-day Kathakas, is derived from the Vaignavas. But the V.R. is unques- tionably a book of poetical conventions and of set formule going back to the 14th century, so that the pedigree of this stylo of literary narra- tion, always falling back upon set passages for aid, goes back to the times before the Vaisnava revival in Bengal. Such aid-books for professional kathakas are not unknown even at the present ‘day; we find books in Bengali, called ‘Kathakat&-éikg&’, actually to be advertised: evidently these are books of set formuls and descriptions of the type XXIV INTRODUCTION of the MS. seen by Dr. Sen, and of the type of the V.R. Consequently, the title ‘Varna-ratnakara’, being that of a work which gives the order or arrangement of a subject treated in a composition, orally delivered or written (generally orally delivered), is quite proper for it. The tradition of having set formulae and prepared descriptive passages to embellish a narrative appears to be fairly old in India, and may be said, on the evidence of the Jaina canon, to go back to the middle of the lst millennium B.C. (I am indebted to Professor Dr. P. L. Vaidya for drawing my attention to it, and to Pandit Sukhlal-j1 and Muni Sri Jinavijaya-ji for giving me some informa- tion about it.) In the Ardha-magadhi Sittras of the Jaina canon, such descriptive passages called vannaa (= varnaka) are quite common. Ordinarily these vannaa are not quoted in full: the text simply suggests that the descriptive passage is to be read at the place indicated by the rubric, jdva ot java vannaa. The reader is expected to know it by heart. In many of the texts these vannaas are given in full, e.g., in the ‘ Bhagavati’, and the * Uvavaiya-sutta’. In the Sanskrit commentaries, the explanation of the passage where the vannaa is to be introduced is given in terms like the following : sampraty asyd nagaryd varnakam aha; ‘ydvac’ chabdakarandt. . Aupapdtikagrantha-pra- siddha-varnaka-parigrahah (reterence—‘ Réyapasenaiya-sutta ’). Muni Sri Jinavijaya-jt tells me that such varnakas are found also in Pali works, and there are Sanskrit and Old Gujarati works where such set descriptions abound. It was customary with old Pandits, Purina and Nibandha reciters, to memorise such descriptive passages and to trot them out whenever the occasion was suitable. Sanskrit prose works like Bana-bhatta’s ‘K&dambari’ and ‘ Harga-carita’ and Dhanapila’s ‘ Tilaka-mafijarf’ were very popular in Gujarat as sources of these descriptive passages. The titles of the seven kallolas are as follows: (1) मगर-वषन, p. 18a; (2) arfawt-wea, p. 210; (3) qrwa-aea, p. 33a; (4) चतु -व्ंन, p. ५1९; (5) प्रपानक-बषंन, p. 55a; (6) wetfz-waa, 7. 600; and (7) खश्ान- वेन, p. 69b. The title of the eighth kallola is missing as that kallola is incomplete. Owing to the loss of the first nine leaves, very little of the let kallola has been preserved. We have merely lists of some of the lower castes and classes (¢.g., [पु] मु Ren tu mag तेग तापि तेलि ताति तिबर तुरि oa FUNSTeS Ue Ve धाक मुक Wert yfrer vigeic ote Sresrew otf wre wie wife we चन्डार चमार ats मोष्छि afin मोखार जावर wie ufe सार पञ्चकवार wefan परि wife सुष्छवारि योन्द eee mac प्रति INTRODUCTION अञ मन्दलातोय ते बास, से कदसगाइ जभं। सङ wae लोभो सवास शपटोर av नणि नेषु नरका AST HUTS रबन्विध द गलकार संयुक्ता), of some of the criminal classes (९.0. अवर कद्सन देषु । चोर Tee Nae fea wane नेखोवकार्ड yew tease नकट कमकड TAS मुष्डफोलुख मडितोशुखं निषसन निस निकार निकिषभर करिार उधखलुख अनेक ये waza खनुचिषोतो तकर अश्रय रेष), and also of various kinds of beggars and mendicants (खाखोर TEA देष। कभा योगै नमारि भरर भष्डुखा Get यतरि सुरतरोखा मौर मोरदष्या बाद परभा प्ब्टतिषे[ = ले] खनेकभिषारि ( = भिखारौ] ते" भर). The noises and sounds of the city, through playing of all kinds of musical instruments, singing of ballads and songs connected with Lorika, and shouts of people crying ‘take! give! break! raise! give again! increase |’ and all kinds of seemly and unseemly acts which would come to one’s sight in a city with its motley crowd, are mentioned. Many of the vernacular terms, referring to society and life in medieval times in Mithila, probably still exist in the country, possibly with slightly altered form and meaning. They can only be expected to have sought refuge with the pure Maithili idiom of the lower castes, unaffected by Sanskrit or Hindi. A great many are probably obsolete. It will not be possible to go through the entire lists here. Many of the vernacular terms remain obscure, and Maithili scholars whom 1 so far consulted (including my colleagues the late Pandit Khuddi Jha and Pandit Babudé Miéra) expressed their inability to explain many of the old words; and meanings suggested in some cases do not appear to be convincing. To enable students of literature, society and culture in genoral in North-Eastern India in medieval timos to utilise this work properly, explanations of these vernacular Maithili: word» will be absolutely necessary. And this work can only be done with the help of the scholars of Mithila. It may be hoped that now that the toxt as it is in the MS. is published, the work will receive the attention it मत ricbly deserves from proper quarters. I shall only give the names of the various topics ‘described’ or listed in the following kallolas, mentioning noteworthy points of interest. At the end I| shall quote tn eztenso a few passages by way of illustration of the general style and treatment of subjects in the work, Kallola 2 beging with the description of a hero, wree-wdar. The ndyaka or hero is to be an expert-in archery, skillful with the eight lesser attainments (upa-siddhis, which are enumerated), and with the eight attainments relating to the world (prdkrta-siddhi, which also are named). He knows also the eight great attainments (mahdsddhw, XXVi INTRODUCTION also named). He is practised in the use of the thirty-six kinds of wea- pons, the names of which are given. He knows the essence of the eighty- four kinds of kingly devoir and polity (rdja-nit:), beginning with the control of horses and elephants and ending in statecraft and decision at important junctures. He is endowed with mercy, charity, friendliness and all other qualities of the cultured folk. Besides, he is perfect, possessing all the thirteen qualities (names enumerated) of a lesser hero (upa-ndyaka). Then comes aifaat-auq. The personal charms of the ndyikd or heroine are detailed forth, and also her mental and spiritual attainments. Her ornaments are named. Then follow some stock comparisons and figures—afa कामदेव संसार fafa आयण, wafe पताका। sf एकर कूप SoH एन्द्र सहला रु मेरा, HUTS चतुरं ख कर्दलु । छनि रहि wifere रागि रक शुष्ण eee भर Few: ‘as if Kima-deva came after conquering the world, and she is her banner; as if Indra became thousand-eyed to 866 her beauty, and Brahm& made himself four- headed; as if for embracing her one Krsna became four-armed’. The attendant and confidential friend (sakhi) of the nayikd is then described—some four different descriptions enumerating the qualities and charms of the sakht are given—sakhis of the type (1) known as wat, चिजिषो, मोडिनो and भद्रा. (This खखौ-वषंना has been given in part below.) Not content with this beautiful description, our author gives the points in describing the smile of the heroine ( भाविका-डास्य-वकेना ). Her smile is like all white and pure things known in Sanskrit literature—¥qq, कुष्ट, कदम्ब, कास, wre (of. भाषो qre:, कालिदासो faens:), Sure, कपूर, पोयुषक कान्ति ; the smile ‘ ripples along like the waves on the ocean of milk moved by the southern breeze’ ( कौर समुद्रक दकिशानिले VIG तरङ्क Wytt ere), etc., and the effect of the smile on the heart of poor young men who seo it, is described. This finishes the second kallola. = The third kallola first gives an account of a royal court ( @Ta- war) with the various officials and other people who throng in it. The term sthdna is used to mean ‘an exalted place’: its Persian equivalent astdn, astdna means ‘threshold’, or ‘the exalted dwelling of a pir or religious devotee’. In the sense of ‘court’ or ‘palace’ we have sthana = asidna current in Malay as a loan word from India. After an audience in the darbdér, which is elaborately set forth, the . king goes to his gymnasium and bath (samara-hara) where he is seated on an elaborate throne of wood, which is adequately described, and servants, some four shampooers (maradanié) named SSndii, Gindi, Kirata and K&nhi, come with perfumed oil INTRODUCTION XXVIll and the king has his body shampooed with oil, all the approved movements like शकर, Waster, wawe, wer, मु डवल, etc..—36 of them—ere practised on him. Then follows the bath. Waters from the 12 holy rivers (all named) are placed in a ‘copper-jar of vold’ ( घौनाक तमकुष्ड ), the king sits on a sandal-wood stool, and has bath, and dries his person with a towel made of a costly cloth. He changes into dry garments, and a mirror is brought to him. Then comes @Wretfa auat—account of the temple into which the ndyaka yoes to perform his worship. We have here an enumeration of the paraphernalia used in worship (2/2). Then he goes to take his midday meal: the articles of food are described: it is a vegetarian dinner, milk and curd preparations predominating, and we have a list of all the delicacies of ancient Mithila. Jyotirigvara was a true Brabman in his long descriptions of feasts ( त्यन्ति wert विप्राः). There is another description also in this book, occurring at the end of the MS., and he gives a menu also in his ‘Dhirrta-samagama’, in which meat and fish are included: मांस माषपडोरतक्रवडिकावाखूकन्राकं वटः संजो वन्यथ मन्छयमुटुमविदशलप्रायः प्रकारोष्करः। खादिष्टं च पथो wa दधि नवं रक्भाणलं शकर संेपादिति साध्यतां सुवदने भिरा agar द्रुतम्‌ ॥ (Act 1.) After his meal, the ndyaka must have betel-leaf, like a good Indian, and this gives the author occasion to enumerate the different kinds of limes and spices used in preparing the betel-leaf. Thon wo have wan-aqar, with an elaborate account of the bedstead and the bed- clothes and the appurtenances of the bed-room. The hero goes to sleep, two expert barber-servants massaging his feet. He rises’ in the morning: and then follow the essential points to note in a series of descriptions of nature. Wo have a प्रभात वक्ना, a मध्याक-व्षना, 2 सध्या. wear, and a राजि-वष्षना, with an appendage, an अन्धकार वेना ; thon there is further a चन्डमा-वर्षना, and a मेच-वषना. With this tho third kallola finishes. The fourth kallola is called ‘the Description of the Seasons’ ( wy-3¥4), and consequently it begins with accounts of the six seasons wem, पौष्य, wl, ace, Yaw and fafac, Spring, Summer, Rains, Fall or Autumn, Early Winter, and Winter. From nature, the author passes to art, and gives us a list of the 64 kalds. Then there is a list of the 16 great gifts (rexwwiqrrt). We have further lists of gems (+ Purénas (only 15 names given), the 10 Upapuranas (these may bo noted: 16 Purdnas=few, fru, देवो, ब्रह्म, ब्रांड, wre, म्य, कुमे, बरा, 4:मन, वायु, Gay, मकष्डेय, कालिका, wifey; 10 Upapuranas: wee, नारदोय, भःगवत, भविष्य, aie, wfcaw, भविष्यो्र, विष्णधमासर--8 names only occur), tne 18 Smptis, and finally the Agamas ( ब्रामण, गौरोजामल, far, trace, शिवधमासर, विब्णधमे, विष्णधमानर, भूतडामर, कालो्तर, प्रपच््ार. प्ज्ञापारमिताच्रा, Weaiefeat, नारायणो, मस्चेवञ्ज, AAA, खतुःद्रतोय wet कम्पि aqquea, खङ्कोणश (?) सावारणो प्रण्टति quay खागम). The seventh kallola closes with this. We have next an enumeration of the Rajput clans (cTeg~.- कुल-वशन ) : 72 clans: in the list we find side by side with well- known names like Jadava, Pamara, Baisawara, Kachwwaha, Cauhana, Candela, Gohilauta, Bhatti, Padihara, ete., names of countries like Mirabhafija, Gomanta, Gandhara, Vardhana, and Khurasaina., Then tollows a list of the 36 weapons, which was already given before in the account of the ndyaka in kallola 2. Thon aftor the rubric दे -वरेना we have only 3 names, and thon the MS. takes us in the midst of a list of works on medicine which evidently formed part of a वैद्य-वपेना or description of a physician. Some leaves or lines seem not to have been copied here—either through their absence in the original, or through the scribe’s inadvertence. After the tgy-zet comes वद्धिजि वंमा, wccount of ships and river-craft: and while the book is describing boats and their gear, it breaks off, with the colophon as noted above. The portion which follows is from the Second Source MS., and it commonces with the heading खथ राश्य-वगना ; it gives an account of women from various lands, and then passes on to a doscription of a Vidydvanta again; and while in the midst of this account, there is sudden transition to the description of some ceremony. This is followed by an enumeration of the ceremonies connected with a wedding ( खथ faary-auat). The next items are: दइादश्पुज-वणन (12 different kinds of sons—list only); wearfawr (enumeration only of 8 kinds of ndyikds); afaaga-awar, which is rather in- teresting, giving a list of articles sold by baniyds—sandal and other perfumes, spices of various kinds, metals, gems, clothes; and com- mercial transactions aro mentioned here. Then we have an unex- pected चौर -वकना, description of a thief, a string of similes, telling us that the thief is like such and such in his cleverness, his resourcefulness, his greed, his cruelty, his recklessness, ctc., etc. Then comes a दुम-वकना, an elaborate account of tho surroundings of a fort. Thore is a further account of boats ( #rer-aqar), followed by an account of a physician (slightly difforent from the fragment occurring before) and an account of ships ( वोड्डित-व्ना) which agrees with the बड्िश-वकंना XXxxii INTRODUCTION occurring before. Finally, comes an interesting descriptive account uf a Jawiera—an evening meal ( विषखारो ), and with this the MS. closes. The above gives the general contents of the work. A few illustra. tive passages are quoted below. अथ सखोवदना॥ पु्िमाक चान्द खम्दत पुरश Scam gw) चेत पङ्कजां दल wat वयि असन aift) काजरक awe wren wae) zat फले ममैदाक शलाका Gre अरूखन Ne) पवराक पक्चव STEN अधर । कणिच्यराक करद्‌ खदूसन माक । Tc मोति Piss खप्सन दान । वेतक खाट खदसभ वाड | पारिजातक THT WEA इथ । WEY कोरर UCIT पयोधर, etc. (Page 18a.) wa ye faueira निष्प्शि याक gua whet देखि WW जणप्रवेष् करः निक शोभा eft इरि aq age Aue शोभा देखि चमरो पलायन करर दालक wre देवि wifes चदय tate wee अधरक wan देखि sate franc dee atra wi देषि ate wiafen wee कष्टक शोभा देखि कम्बु समुद्र प्रवे करर शनक शोभा देषि चक्रवाक sway भेर वाङंयुगलक शोभा देषि wae नार पकनिसग्र He... eae भोभा देषि कद विपरोलगति axfeo चरणक शोभा देषि रकम निकृष्न आश्रय करस । रवम्विध रलारष्ार युक्ति जिभूवनमोहिनो देष ॥ (Pages 20a, 200.) अथ प्रभालवणना। देवक आयतम GEM Ie TI दण्ड पल घलौञे प्रभा]तक्नान करा्योर० गजराजे शब्द ee बायसन्डि कोरा we मसज तिरोहित भेक GTR STH भेजा TH दोर Wafer भेर० भमर GUTH चश ° वेद कर्ने वेदध्वनि WCC कुशो सरण्न vise as ay जलाशय ्यारद्लण० बन्दोजनन्हि sy aK. wryafad खदा waatieww (Page 29b) wfaaet walquara करर मायके द्षटहेवतासारण् करण प्रुभोत्धान कर्‌ | (Page 30a) अथ भड्वण्येना॥ मार परि wat परिषशमेर खाय Sine zw चारि ufcyae ्रङनोक पाम रक मथा वन्धने सोगखढयोक wow wae देवमिरिष्ा qwewrer we Ure बन्धने, लोपि योषि afe नोकि dens परजे fewer शोडाक finirefa va डोर रो एक बाम क (650) ए बन्धने पमु AKER भाढ। संसक्त परात्‌ अवट Gufes Seat मामो es भाषाक तश्च कारो wrt qraret qt nife@t ओक विजातोपा errs उपभाषाक कुकर, पानिनि चण्ड कापदामोदर SRA AVR AI सारखत Taft बे GSW खाकर लाक पारन। धरि विश anfe मर wrafer qerqure area exe उत्पशिणो मेदिनोकर wows प्रग्लि ओ अडारदखो कोष तं arg.) ध्यनि वामन cal मदा sever cred oT प्एक्ारतिरूक सरखतोकष्डाभर णादि अमेव quyTCS विश्च । wy रभरताकर erates अन्दोविचिलि .भारतोङषल after प्रभ्डति ete दन्दोपन्य तं कुण । कादम्बो THE वायस eae ayer इरवयरित wy ere भाङभन्नौ कपुर. wu(S6a)e प्रति पुष्यं wa छलाभ्या । केवारो after साकिक Kegs fircty व Xxxiii दाता कवि खातो जे wage ते eae) oafiawien Tear मष्डशि write धरीजले भाढ Vay । तंका पा केश्यो free quae set vache ervet नारका svat धरमे० काका FW काडका wearfto कश्योगभो सुत कातो चरर sey dares तषो we dente बरवड चरि शरि Sry eee ओगण See अरूसनि afd करसे ख्योडङणक मारा एकशो परिडले० amy ख्यानक मारि से afeu forrg धारले० विरले वद्धे FS बाङ्के we HIT Gauy que नक साप qrearye कार्तिक कल्याण HTK खड, मगारि feaate ते परिवेहटित wis Say From the above account of the various subjects described or listed in this work, and from the extracts given above, the very great value of the V.R. as a compendium of life and culture in medieval India will be easily seen. The book would seem to stand worthily beside the ‘Manasollaisa’, the Sanskrit cyclopsdic miscellany compiled during the reign of King Soméévara III, Bhiloka-malla, the Calukya king of Maharastra, who ruled from 1127 to 1138 A.C. (published in part in the Gdekwad’s Oriental Series, Baroda). In tho picture it presents of court-life and its surroundings, it reminds one of the ‘*AIn-i-Akbart’, which, with its lists and its detailed accounts, was written with an entirely different purpose—being a gazetteer with a conscious scientific and historical value, which the V.R. as a literary lexicon is not. The atmosphere of the V.R. is purely Hindu, and pre-Mohammadan, although it was composed a little over a century after the establish- ment of the Mohammadan Turki power in Northern India. The pre- Mohammadan atmosphere which the work breathes is a sufficient indication of its genuineness, although the MS. was copied some two centuries later. A few Persian and Perso-Arabio words occur in it, which are given below, and which show the presence of Porsian-using Turks in the land. The all-embracing scope of the work is a noteworthy thing, and we 870 exceedingly thankful to the scholar and the poet who had the happy idea in his mind of preparing a book for the guidance of poets and narrators in the vernacular. Jyotiriévara Thakkura must have been a man with a wholesome all-round interost in life. Ho was not a simple Vedic priest whose mouth had frequently to taste the acrid soma-juice; and he was not a pedantic littérateur either. His ‘ Pafica- siyaka’ shows that he was ५ keen student of the ars amorts also. His catholic observation, like that of our Kathakas, who to drive a moral lesson home must largely draw upon their own observations in life, and must bring in verisimilitude in their narrations, did not consider any aspect of life with which he had to come in contact as too low or beneath his notice. He takes us through the city, and gives us a little . 3 . 2.4 । ~ - } - y xxx „+ , \ छः ; '. 4५ £ >» , १२४५. ¢ 1८ + . ॥ VY १ ॥1 ४ # ५ 4 ¶ * iy) : ५ > uO Df | १५ glimpse Into the ugliness that was in & medieval Indian city, cities of other ages and climes: he tells us what knayes and we meet, what low and vulgar fellows congregate and shout and jostle and move in dirt and filth; and he gives us also romantic descriptiong ` of noble heroes and beautiful heroines, perfect in their personal charms and accomplishments. He shows us round the court, and tells us who 18 who in the throng. He gives us little inside views of the intimate life of the princes and noblemen, shows us how they bathe and what they eat, and even lets us have a peep into their sleeping chamber. He is a poet by instinct, and this in addition to his powers of observa. tion. The quiet dignity of his sweet Maithili tongue gives him an additional charm. His little sketches of morning, noon, evening, and night with its darkness, of the various seasons, of the forest,— all these have the stamp of poetic genius. In spite of the conventional devices employed, the light of a broad intellectual sympathy shines through everything. His similes seem to come so naturally, although we know them to be the conventional phrases passing current as small change of literature in medieval India. And how lightly does he step among all sorts and conditions of men, and their wares and their stock-in-trade! He is our guide through a fashionable gambling house, and he bewilders us by his familiarity with the various games that are on, as well as by his knowledge of the ways of the men who gather thore; he is apparently a connoisseur knowing the various kinds of stuffs, and gems, and spices and perfumes, which the baniyds of the bazaar, the drapers and gem-cutters and druggists sell. He stands to watch the troops on the march, or a royal cavalcade going out to hunt in the jungles of the Tarai: and he knows the Rajput soldiers riding past by their clans, and he knows what weapons they wield, and what horses they ride, or what dogs are led in the leash. The Bhd{a or official bard of the court, who was often a sort of ambassador to his king, was a person of consequence, well-trained in all kinds of learning; and he was apparently a person whom our scholar-author knew very well, and admired. He also knew very well the professional singer, the Vidydvanta; for Jyotiriévara himself was by his own showing an accomplished musician and singer (in his ‘Pafica-siyaka’ and in the ‘Dhirtta-samigama’). He recognises Gndi, Sdndi, Kiratu and K&énhu, the four humble servants who come to massage their master, and he calls them by name. He had & fine sense of humour: certainly he was not a %enqqwervafe, one whose intellect had got clouded by too much study of the Veda; descriptions, like that of the old Ku#fané, and, besides, his characterisa- tions in the ‘Dhirtta-sam&gama’, with the decision of Asajj&ti-miéra 38 85 in al] beggars dat INTRODUCTION ०५ as the final dénouement, show how heartily he can laugh at the oddities and the frailties of men. He was equally at home at a wrestling match and at a nautch party, and he takes a man of the world’s delight in acquainting himself with the details in either kind of exercise of the human body. Asa Brahman of the court, he gives us ample indication of his being an adept in the gastronomic art also. In fact, he gives us an epitome of the life in a Hindu court in the early part of the [4th century. Unfortunately, through the MS. being defective, we have no means of knowing whether his survey included life in the village as well. In the ‘Dhirtta-samigama’, he has given us just the kind of a little description of the house of a prosperous farmer which tempts us to believe that he did not neglect the life in the country- side: wea, tag tag, fafec-weaere-qga-cfcq-cgec.afyel-eq-diem. TSH, षदो तदो सश्चरका-वारु-गोवनच्छ -सोडद पौवशद्रत्यशारूुषपरिक्लणन् -मन्दसचार- रमरिव्लावासपरिषर-सष्चरमा-चेडिष्धासमृषं safe awiveg बासभग्यशं विष्डोरष्यदि (Act I). As it is, his lists and his little descriptions give us a veritable ‘Bihar Court Life’ for the 14th century, and as this court was in intimate touch with the life of the commonalty, we get valuable hints as to the life of the common people also from it, if not actually of ‘Bihar Peasant Life’ in those days. Perhaps he did not feel so much attracted to the rustic folk and their ways as to the cultured people, Pandits, musicians and others, of the little provincial town of Simardima-pura (the present-day Simraon) which was the capital of his patron. The religion that he describes or hints at is the ordinary religion of the cultured Hindu of medieval times. He has an occasional mention of Buddha and of the Buddhists, for Nepal with its Buddhist Newari rulers was a neighbouring state, and Buddhism was not yet dead or transformed into the current Hindu cults on the plains of Eastern India, even in the 14th century.” The Natha or Yogi sect was strong, a sect which apparently had combined Saiva Yoga practices with some of the notions and traditions of Tantric Buddhism: and Jyotiriévara, good Brahman though he was, thought it to be in the nature of things to include the 84 Siddhas of this sect which had established itself in popular favour by the preceding century, and apparently was ingratiating itself with the orthodox by its frank and open allegiance to Siva and to the Yoga practices. ° The kaleidoscopic view of life in North-Eastern India of the 14th century as presented in the V.R. affords a valuable commentary on the epigraphic and other literary records of the contemporary and XXXvi INTRODUCTION earlier periods. The list of officers and courtiers given under दयान. वेना or description of the court, for instance, is longer than similar lists found in earlier Bengal and other North-Eastern grants on copper plate, and these are mutually complementary. The gambling saloon described by Jyotiriévara is apparently of the type known to the author of the ‘Mrcchakatika’ as well. Jyotiriévara calls a gambling house a fenta-sdra, i.e., tenta-ald, and there used to be a temple of the Devi near by; he also knew the word tentd-kardla (p. 39a), in what sense exactly we do not know, but apparently to mean a person who visited a gambling house; and more than four centuries before him, Raja-sekhara has used the word in the feminine form (fen{d-kardld) as a term of abuse, in his ‘Karpira-mafijari’. The names of the various objects of luxury, and other articles mentioned in the V.R., are explained by similar terms found in the earlier Sanskrit literature, and vice versa. All this goes to make the work a document of first- rate importance in the study of culture in early and mid-medieval times in Northern India. The fact that the work is written in a vernacular speech, with tadbhava or Prakritic forms of a great many of the Sanskrit words, renders its importance all the greater, as affording us & sure evidence of all or most of the items of this culture, material and intellectual, haying become a part of the daily life of the people even when they did not know Sanskrit. Apart from its special impor- tance for Mithilé and apart from its linguistic interest, the ‘Varna-ratnakara’ thus becomes a work valuable for the student of Hindu oulture in general. Eastern India unfortunately does not possess any considerable literature going back to Early Modern Indo- Aryan times from which one could draw inferences about the life of the period. With its Prakrit and Apabhraméa and Early Gujarati (or Old Wostern Rajasthani) literature, Gujarat and Rajputana are more fortunate in this respect. Hence the unique position of this Old Maithilf work, giving a great deal of exact and useful information. 4. Tae LANGUAGE OF THE ‘VARNA-RATNAKARA’. Vory few authentic specimens of literature in a Modern Indo- Aryan language going back beyond the 15th century are available. We have tho Old Bengali ‘Caryas’ (their languago is slightly tinged with Western Apabhraméa forms), which belong to the period 950 or 1000 A.C. to 1200 A.C., and after that, we have the ‘Srikpsna-kirttana’ of Candiddsa, preserved in a fairly old MS. dating probably from the 15th century. The Caryas have been a fortunate find for the history of Bengali. As for Hindi, the literary traditions of Western Hind! go back to the 12th century, not considering the references to INTRODUCTION अदश) earlier works, or to writers who certainly employed Saurasent Apabhraméga, ¢.g., poets like Pugya-kavi of the early part of the 8th century, and Khumana of the first half of the 9th: Saurasen! Apabhrarhéa had not as yet developed into a Modern Indo-Aryan speech, ४.९. into Western Hindi. But the works attributed to the 12th century and 13th century authors, like Kedara Kavi (c. 1160 A.C.), Ananya-deva (1148 A.C.), Canda Baradai, and Nalha (author of the ` Biséla-deva Rasau’, ९, 1216 A.C.), are either unobtainable, or are late and spurious, 80 far as their reputed authorship is concerned. Authentic specimens of Western Hindi poetry belong only to the 15th century, and even then it is questionable how far the language of the time is preserved, as there are very few old and reliable MSS. A poem ascribed to Ramananda (14th century) is preserved in the Sikh ‘Adi Granth’: but the language is, to some extent, modernised, besides showing one or two forms which do not strictly belong to Western Hindi. Eastern Hindi remains are later still, first half of the 16th century, in the ‘Padumawat’ of Malik Muhammad J&yasi. Present- day Gujaréti and Western Rajasthani (Marwari) are derived from 8 common speech which has been called Old Western Rajasthénf, and which originating in the 13th century out of the Apabhraméa dialect current in W. Rajputana and Gujarat in the llth and 12th centuries was current as a single and undivided speech up to the 16th century. Specimens of literature in this Old Western Rajasthani speech are therefore equally Old Gujarati and Old Marwari specimens, and these have been found dating from the 14th, 15th and 16th centuries, but none earlier than the 14th century. Marathi, however, possesses a genuine old text dating from the last decade of the 13th century—the ‘Jtanéévari’ of Jiana-deva, a verse commentary with translation of the ‘Bhagavad-gita’. Of the other New Indo-Aryan languages, Ofiyé is the fortunate possessor of one or two short inscriptions belonging to the second half of the 13th century; but the others—Panjibs, Lahnda, the Pahari dialects, Sindhi—cameo very late in the field to be preserved in a literature which is still extant. Outside India, there is Eu, or Old Sinhalese, which in somo of its extant remains goes back to the 10th century. The importance of the V.R. as being one of a comparatively small number of authentic works in a modern Indo-Aryan language, which goes back to the 14th century, may well be realised in discussing the development of Modern Indo-Aryan. Its position is equally important with the ‘Caryi&s’ and the ‘Srikrsna-kirttana’ of Bengali, the ‘ Jfianéé- vari’ of Marathi, and the earlier Old Western Rajasthani, Braj-bhakhé and Awadhi works. f ae अ . HO INTRODUCTION. ५ Words and forms in the V.R. have their parallels in the othe New Indo-Aryan languages, especially in Bengali, the sister-g peech 2 Maithili, and to some extent also in its cousin and neighbour Awadhj These parallel forms elucidate each other’s history. A good many Early Bengali words, for instance, which could not be explained because of the great advance in their phonetic development, became clear as to their origin and meaning when the corresponding Early Maithili forms as in the V.R. were found. I can mention two such words. Early Bengali ahutha (< ardha-caturtha = 33) and mauhari, mohari (‘a kind of wind instrument’, Skt. madhukarika). Above all, there is the supreme importance of the work in the study of Maithili philology, as the oldest document in the language, antedating Vidyapati by two generations. Its language is more archaic than anything we find in the current poems of Vidyapati, which itself is archaic enough for Modern Maithili. It 18 clear that the Maithili speech was a far simpler idiom than what it is at the present day, especially when we consider its rather complicated conjugational system. The MS. of the work is two hundred years younger than the work. Consequently we must assume that some at least of the original 14th century forms have been altered to those of the later period. This reservation is to be made first of all in considering the language ; and we shall have to see how far the older, more genuine speech of the 14th century could be modified into the language current at the time of copying,—in pronunciation, in forms, in syntax and in vocabulary. The peculiarities of the language are now discussed. [I] OrnTHOGRAPHY, PHONETIOS AND PHONOLOGY. § 1. The orthography of the MS., dating as it does from the beginning of the 16th century, cannot be expected to represent faithfully the pronunciation of the 14th. But the general indications regarding the phonetics of Early Maithill of the 14th-16th centuries are valuable. The orthography also throws some light on the medieval pronunciation of Sanskrit in Mithila. (a) The Vowels. § 2. The simple vowels were & [=a, ॐ], 9 (long and short), i (i), u'(a), € (=, ठ), ० (=©, 5), and probably also a long [2]. § 3. The pronunciation of & seems to have been as in Modern Maithill, १.९., an intermediate sound between the North Indian [a] (=v in South English hud) and [9) (७0 in South English hot), when it INTRODUCTION अड as stressed; and when unstressed, and in final positions, it had probably become [9], the so-called neutral vowel, as early as the heginning of the 14th century. This final and unstressed -% = [a] vas being dropped from pronunciation at time of the copying of the MS., if not earlier: witness orthographies like खदसन्द (खं) च as in patala afsaind duhpraweSsa, stri-k& (दम्प aisandurlakea (=aisina durlakga) (310) ; kaisanaha beside kaisan& aha (fre- quently) ; tinu beside tindhu (49a). Cf. also the use of the virama in nagal, lagal, cafical (10a), nirmmal (16b). (Is this श्‌ a device for?) The dropping of the final unaccented -& seems to have been established in spoken Maithill by the beginning of the 16th century : in Bengali, it certainly did by the middle of the 15th. § 4. ३ becomes weakened to & in compounds and in some suffixed forms when it loses its stress: ¢.g., kana ‘ear’, but kina- kata ‘ear-cutter’ (10a); chadae ‘leaves’, chadawia ‘is disjoined ’ (77a); kapala ‘cloth’, k&pala-ghara ‘tent’ (36b); raja, but rajaesa < rajadeSa (47b) ; ५६६, danta ‘ tooth’, beside d&ta-cha < danta-ksata- (60a); bajana bd&jaia ‘music was played’ (47a). Unaccented a was probably pronounced short, a8 in Modern Maithill (ef. Grierson, Maithili Grammar, § 7): cf. the spellings bidafiota and bid&dfota, semt-tatsama <~ vidyavant-; marahathi (57a) and mara- hathant (71b) <~ marahaftha-, mahdrastra-; dora, dora, dwara, awara ‘und’ < apara. §5. There is occasionally no uniformity as to the use of {, f and ४, i, especially finally: e.g., Suc ‘clean’ (76b), deg (= dekh ), beside desi (= dekhil), ‘seen’, frequently; bidirnna = vidirna (20b); kgirodaka (76b) beside kgirodaka (35b); dharani (55b); tini (16b) =tini (42a) ‘three’; nirmmikta (40b). Generally, the short forms are preferred; and considering that in Modern Maithilf, these final short -i, -u are very short (frequently unvoiced, almost indistinct) sounds, it is natural to expect this modification or weakening of them in the 16th century, and possibly even earlier. §6. eando were both long and short. They were short ospecially when they formed the second element of a diphthong. Examples: bétid ‘girl’ (76b), ka&le ‘done’, bhaé gelaha ‘became’, gdara < gopala (29४), athad ‘eight’ beside athahu, caladle ‘gone’, karadle ‘done’, etc. See Grierson, § § 11, 12, 13, 14. In the interior of words, € and 0 commonly stand for y& (yé) and wa (w6): ¢.g., kaela beside kayala ‘done’, fora beside &wara ‘and’. Cf. aéraye = 98१8९ == &éraya (20b). Conversely, ya (and wa) figure for e (and 0): ¢g., rajaputray& = rajaputraé, instrumental (22a); gosafifid = 8०७2९158 = 8०8९1९६ = *& ०७३ 1६ < gosvimi-+ -ena (13a). xl INTRODUCTION § 7. The sound of 3, long, seems also to have occurred: it was written ४, and 80 : eg., a (= 5 < au) ‘and’ (38a), also occurring as au (56a) and as u (38a); the post-position -safio (= saWo) ‘with, from’ figures also as 88 (= [83:]) at p. 75a: the present Maithili equi- valent is[s3:]. Cf. sarahara (= 8518 - < *siwara-) beside samara- hara (24b), masah@ri ‘mosquito net’ (36b) beside musari = musdri (29a); ath’ onacasa-bayu barnnand, where 019८8889 = unScaése < tinapaficagat (68a). § 8. The diphthongs were ai and au. These are written ai, al, ayi and au, ali: e.g., baisala, 0218219, bayisala ‘sat’ (the last occurring at p. 18a); gati ‘went, gone’. The disyllabic a-i (and ४ -#)} probably also occur, written ai or ayi (and all): e.g., japaite beside jarayite (63b). Other vowel combinations occurred, but it seems that they had a y or w glide in between, making two distinct syllables: thus, deite acha ‘is giving’ (29b) beside deyite (77a); hoite ‘to be’ (29b), koiri ‘a caste’ (290), dhunia (10a), dekhuaha (common), etc. The groups ae, ao, oa, 18 may be regarded as diphthongal. §9. Nasalisation of Vowels. This is denoted by the candra- bindu sign, but in the case of 1, & (or yi, Wu, Fi, fu), we find the use of fi: eg., gosafifia = gosavi¢é = gosaWié (13a), mafiusi = mfiWusi, maWusi = méatrsvasr- (4la); pau = pawl < *pa(w)uWwa < paduma < padma (20b); sarafii = sarai < éarave (77b); bhafie = bhayé ‘through fear’ (30b); safio = saWo < sama-; bidafiota = bidawWota < vidyavanta- (57a); bhafiuha = bhaWuha < bhamuha=bhrii (18a, b, 40b). There are cases of nasalisation through contact with a nasal in the word: e.g., mafiusi< 10208 (4la), bandhala = bandhala ‘bound’ (2la), kana = kéna< karna (200), mantha, j.e., matha beside matha < mastaka- (56b), kinti = kanti (77a), agafiungi, pachafiungi < agra-, pascat- + ahga- (75b). The above examples would demonstrate that the habit was to nasalise contiguous vowels if there was a nasal sound in the word, as is the way in Bengali. Spontaneous nasalisation, of Middle Indo-Aryan (Prakrit) as well as Modern Indo-Aryan (Maithilf) origin, is also found: e.g., 11108878 = 018४118 < hresana (47a); ०६1४ < *benta, betta< vétra (74b); dasafiudhi = daéa + dyudha- (22a); 7408, bah® = bahu ; ahtkari = adhikari ; ufica = ica < ucca; sfisu ‘porpoise’ (51b), cf. Pali surhsumfra, but éuéuk in Bengali. Cases of loss of nasalisation are not absent: e.g., the instrumental affix -& < -ena also occurs as -e. See below, passim. INTRODUCTION xli § 10. The Sanskrit ¢ was pronounced as ri. Thus, trparvva = triparva (75a). The ru pronunciation (ru, ur) is not found in Old Maithilf. It is now found only in the Deccan and the South. Oriya falls within the ru area now, but the ri pronunciation obtained in that language in the 13th century, as epigraphical records show. Later Maithili Umlaut is unknown. But Epenthesis seems to have come in: cf. § 52 (i) below. (8) The Consonants. 811. The sounds were the following, as in Common Modern Indo-Aryan: k kh ¢ gh (9) € ch j jh (A) tthd dh (n) tthddhn p ph b bh m yriw 8,8, ॥ ` $ 12. The nasals #, fi, m occurred only before their corresponding stops and aspirates, never singly. Intervocal fi means only the nasalisation of the contiguous vowels, accompanied by a glide y or w (866 supra, §9). mn occurs in tatsamas only; and in tadbhavas, as well as in semi-tatsamas and frequently also in tatsamas, n and n are used indiscriminately, generally as n (or its substitute 1): which shows that the cerebral n sound was lost. Thus, Canura = Canura (25a), pani = Pkt. paniya, Cauhina (220) = Cauhana, Panini = Panini (55b), Narayani = Narayani (69b), banika beside banika (748). T'adbhava words as a rule have n. In Mithila, at the present day, Pandits pronounce ए as a nasalised cerebral इ, ¥, t.e., F, in tsa. and stss. In tbh. words, however, there is only n at the present day: which shows that the cerebral sound is not proper to Maithill. The loss of the 9 sound in Maithili took place undoubtedly at the time of the copying of the MS., but we are not sure whether the absence of it characterised Maithili of the 14th century. §13. The Modern Maithili assimilation of a voiced stop or aspirato to a preceding nasal—ng(h) > 7 (1), ij(h) >A(h),gd(h) >n(h), nd(h) >n(h), mb(h) > m(h)—seems not to have been characteristic of the language of the V.R.: e.g., 81088 = 8082 - (10b), agafiudgi, pacha- hungi = agra-, pag&cat- +anga- (57b), dandia = dandgika- (29a), ०१९६११४ = danda (39b), candara = candala (10a), cinda = candra (18a), kumbhira (52a), kindha = skandha (57a), khindga = khagda ‘sugar’ (77a), sindiira (188), bandhane (5§a), 81779911 (49b), sondha < sugandha, dialimba (50a), kambala (29a), ete. The nasal in all the above instances was the ‘reduced’ nasal sound, a sort of half-way house between the full nasal of Middle IndosAryan xlii INTRODUCTION and the nasalisation of the vowel of later New Indo-Aryan. The spellings danta (beside dAta), pafica, canda (beside ५869), sompa = khompa (cf. Bengali khdp& ‘done up hair of a woman’), banhki< vakra- (55a), kanta< kanta(ka) (74b), clearly indicate the ‘reduced nasal’ pronunciation for the 14th century. See ‘Origin and Develop- ment of the Bengali Language’, I, pp. 359 ff. That the post-nasal stop (or aspirate) was fully pronounced (at least in the case of the gutturals) is evidenced from spellings like pange = panka <~ panka (52b), ahka-rakhaka = anga-raksaka (23b), Kongana = Konkana ‘a country’ (22a), and donkala = dongara ‘hill’ (530), where we have interchange of ng and nk, the latter certainly indicating the preservation of the stop sound. Cf. also andhari (58b), kandha- < skandha- (46a, 6lb, 52b), which show the preservation of dh after n. But already the assimilation in the case of the group mb was establishing itself in the language in the 16th century, and evidently it is due to the later habit that we have forms like Kamoji (44b) beside Kamboja (43a), tamauli, tama-kunda < MIA. tamba-, tamra- (25b), daliwa < *daliwa < ‘*dalima < dadimba (18b) beside dalimba (50a), in which mb has become m. § 14. Tatsama kg was pronounced as (k)kh, perhaps as (k)khy, as in Bengali and Oriyé: eg., anga-rakhaka (230), biyagkhani = bijakhkhani < vicaksana- (40b), khyara-pradipa = ksara- (76a). Cf. the sts. name Lakhima = Lakgmi at the time of Vidyapati. In tbhs., Old Indo-Aryan (Skt.) kg figures both as kh and as ch: eg., khira = ksira, dakha = drakga (50a), data-cha = 02118 - 189६9 - (60४). § 15. As in other parts of North India (excepting Bengal) § ¥ had the sound of kh, and ¥ is the letter commonly employed instead of the proper ख. This घ = kh is really a semt-tatsama pronuncia- tion, in North Indian medieval Sanskrit. The OIA. 8 had become 8 (or 8) in tho tbh. element in the Modern Indo-Aryan languages. The OIA. 8 was originally an ich-sound (i.e., the sound of ch in German ich); and g, with its tongue-tip properly curled up and touching the dome of the palate, gave rise to a hard sh sound which could be easily altered to a guttural spirant [x] (like the ch in German ach) ; and when the old articulation of § was lost in the vernacular Prakrits, but the tradition of a back, semi-guttural cerebral s remained with the Sanskrit scholars, this ¢ > [x] (= ch in German ach) could easily become altered to the familiar Indian guttural aspirate sound of kh. (See infra, § 21.) INTRODUCTION xhiii $16. j was indicated by both # | and च $: which shows that pronunciation णय = y was as in Bengali, even in pronouncing Sanskrit words, viz., the letter य was j initially, and medially also in some consonant conjuncts. The forms je, ja are frequently written ye, ya; ci. biyagkhani = bijakhkhani < vicakgana- (400), Dewajani ({sa) = Devayani ; and jantra-dhanuka = yantra- (46b). § 17. 10 in semt-tatsamas had the medieval and modern North Indian value of gy or g¥: cf. the spelling Angapala (21b) for 2948. pala = 2102 -7218. § 18. y, फ were glide sounds, 83 much as in Modern Maithill. These never occurred initially. Medially, these glides were indicated by y and w (generally by y,—w and b being confused, and the letter for w being commonly used for both w and b): thus—jarayite (63b), suyara = suwara < suara < siikara (52b), guya = guwa, beside gua,< guvaka (72b, 50a), khelawara < khelapala (38a), awara ‘and’, kadava < kardama (43b); or by fi, when accompanied by a nasalised vowel (see supra, § 9); or by € and 0 (see supra, § 6). Somo- times the glides were left unindicated in writing: eg., pahalla = pahadiya ‘hill folk’ (50a), dhunia = dhuniya ‘cotton-carder’ (10s), Sauria, a village (77b), cadoa = cidowa <~ candr&atapa- (29a), goara< gowara< gopala (29b), musa-réa< -roWa > -loman (52b). Intervocal w (=v) could become ‘nasalised, and be represented by m:e.g., Remanta = Revanta, ‘the God of Hunting, son of the Sun God’ (43a), yamanika = jaWanika < yavanika (59a). Conversely, we have denasalisation of m to w = W in daliwa = daliWa< dadimba (18b), kaidawa <~ *kadaWa < kardama (43b), Rawayana = ‘ Ramayana ’ (68b). Initially, ts. and sts. ए (ज) scems to have been pronounced as b, as now, in spite of the script of the MS. having two lotters 4 = b and दें = w. The letter for w is strictly used for the glide w sound in tbhs., and also for ts. and s/s. intervocal -v- (-w-). The anusv&ra seems to have changed -v- to -b-: ¢g., wafew ewamwidha = ewambidha for evarn-vidha (50b): cf. the Bengali spellings किन kimba, न्वा sambad, etc., for किर kim-va, नवा sam_-vada. § 19. The ¥, { sound, as a modification of earlier intervocal g, is as a rule written lin both tbhs. and tss.: e.g., Byali = Vyadi (55b), ८011 = cudi ‘bracelets’ (60a), daliwa = dadimba (18b), brila = vrig& (20a), Lala, Cola = Lada (Lata), Coda (22a), Palihara = Padgihara << Pratihara (22a), ghali = ghadi (20b), niwila = nivida (31४, 319), 1८६0219 = 27242 ‘cloth’ (318), ghola = ghoda (37b), Rath- aula = Rathauda < Rastrakiita (44a), ahela = aheda < akheta- (4b), baheli = bahedi (49b), pahalid = pahadiy& (50a), pahdla = xliv INTRODUCTION pahdda (51b), Drawili = Drawidi (55b), koli = kodi ‘scoro Z Bengali kuri (60b), nauka-ghala = -ghafa (75b), etc. d, however, ocours in one or two cases: ¢.g., padari (33b), pauli (49b) = patali. Conversely, we find d for 1 in some instances: mrnada = 7010213 (36b), khega = khela (37b), cf. khedi ‘play’ in Middle Bengali, In Modern Maithili, as in other Bihari dialects, Magadhi 1 (< r,1 of OIA.) in ébhe., Sanskrit r, 1 in tss. and stss., and original or derived d> fF, these three groups are all confused with each other, and genuine Bihari prefers only r. The significance of d > rf being written 1 in the V.R., and also in other Early Maithili texts (e.g., Vidyapati), is not clear: more so when we find r for 1, and vice versa—e.g., sinkara = $rnkhala, Bengali 81४81 (70b), padari beside padali = patali (33b), baura ‘toe-knob of wooden clogs’ = MIA. baiila-, Skt. mukula-, Bongali baula (76b), taru = talu (77a), lewari = nava-mallika, Middle Bengali neyali (29a), dotkala = dongara ‘hill’ (53a), narikera (50a), katahara = kanta-phala‘jack-fruit’ (618), sorahia (75b) adj. from solaha < sodaga (55b), ajagala (50a), etc. Con- fusion between r and | is characteristic of the Tibeto-Burman Newari —at least in Newari MSS. of works in Indian languages: and Newari 18 the northern neighbour of Maithili. This want of uniformity in spelling in the V.R. is probably to be explained as being due to the r pronunciation (OIA. -r-, -l-> Magadhi Pkt. -1- > Bihari -r-) being established for intervocal 1 in Maithili by the 14th century, but the influence of Sanskrit and old traditions in spelling kept or restored the 1 in many casos, and sometimes brought it in wrongly; and it sub- stituted -l- also for -r- to which the oarlier -r- from MIA. -d- (<< OFA. -t-, -d-) and Skt. -d- apparently had abutted, as in kapala (< *kapara < kapara< kapada< kappada < karpata), Drawili (< *Drawirit < Drawidi = Dravidi). It is not likely that the 1 for ५ > fF stands for a cerebral 1: that sound, it is not impossible, obtained in Magadh! Apabhramhéa, but probably it was altered to r in Early Maithilf. The letters for n and 1 resemble each other a great deal in the Maithili script. It is likely, judging from dialectal Maithili (and Bengali and Oriya), and also from Middle Bengali, that initial n- and 1- interchanged a great deal in Early Maithill; it was probably a case of a tendency to change initial l- to n-, There is a notable form: birani (18b) ‘pleat of hair’, for bilani? < *binani, cf. Bengali binani,; binuni (which seems to be from *varnapana+vinyasa. § 20. g occurs for kh in tbh. words as a device in spelling (see above, § 15); in ¢s. words, it was retained, but was pronounced as kh; ¢g., déga = dokha (688), manuga-ka (47a). In compound INTRODUCTION xlv consonants, it certainly had an sh, may be an s, sound, as at present: ( ¢ . Sista, baligtha, ragtra, णप 2, etc. OLA. § became 8 (= dental ? palatal?) in tbh. derivatives in Maithili: mahisa (46a), musa-rda (32b), 11088118 = 1652008 (47a), solaha = godaéa (619). § 21. 8 and 8 frequently interchange, both in tbhs. and in tse. und 585. The dental letter is more common. Thus: maréa, (63a), susruta (75b), rajaesa = rajadeSa (44b), nisina = nihsvana (t7a), 89612 = saciva (22b), 21997 < *darasia < *adarasika < adarsika, basi = varhéi- (10b), Sarh$aya (16a), siisu ‘ porpoise’ ; of. Skt. 81811818 (52a), bafsala< upa-vi&, pafsu < pra-vi§, etc. What was the pronunciation, since this interchange shows that there must have been one sound for the two letters? In Modern Maithill, the dental sound only obtains. As a descendant of Maigadhi, we would expect only the 8 sound to be inherited in Early Maithill, as it is in Bengali. In current Maithill writing, the interchange between the two letters is found commonly enough, and in the Kaith! alphabet as em- ployed in Bihar, श = 8 is the character for the sibilant sound. One sound obtained in Early Maithili undoubtedly, and it was probably a kind of sy sound, as in Oriya which is to be taken as intermediate between § and 8. The use of ch for 8 in Dhandchi = Dhanaéi, 0121881 = 0020281, ‘name of a Ragini’ (58b), is noteworthy: similar change between ch and ©, and the sibilant (= 8), is found in Bengali. § 22. h generally remains, and was in all likelihood a voiced sound as in Common Indo-Aryan. Intervocal h in the particle hu in the numerals was very unstable, and was probably dropped from pronunciation when the MS. was copied. (See infra, § 40.) A consonant preceded by r is always doubled (in (5. words). This was characteristic of Bengali also, e.g., sampirnya, aiévaryya, varnnand, dharmma, marddana, nirbbhugana, durvvacana, karppiira, etc. Some new aspirates developed in Maithili, as in other New Indo- Aryan speeches: nh, Ih, mh, rh: cf. ¢.g., Kanha = Krgna (25a); the plural affix of the noun, -nhi; kolha = kolla = ‘Kol’ (50a); unhasaité = ulhasaité < ullas- (63b). Instances of mh and rh do not occur, but doubtless they oxisted. The Phonology of the language of the V.R., १.९. the history of the development of its sounds from MIA. and OIA., agrees closely with the general lines of development of most other New Indo-Aryan speeches, which are well known. This could properly be studied in connexion with a systematic history of the origin and development of the Maithilt language. xIvi INTRODUCTION [II] MorPHoLoey. (a) Declension of Nouns. § 23. 708. The various vowel and consonant stems of OIA. were reduced to a few vowel stems in second MIA. (Prakrit); and in the late MIA. (Apabhrathéa) stage, through the further weakening of the final vowels, 3 stems only remained, -&, -i, -i. The affixes of the -& stem, 7.e., what little remained of it from the elaborate declensional system of OIA, e.g., an instrumentalin -é or -e, a genitive singular in -aha or -aha, a locative in -hi or -hi, and a genitive plural in -ma, came to be added to the nouns of the other stems also, irrespective of their origin (either tbh. or ts.) or their final vowel -i or -0 (which in the oblique forms tended to become -4). The NIA. languages inherited these few forms from the Apabhraméa, and as necessity arose, built up new post-positional inflexions to indicate case relations. In their early stage consequently there was one system of declension only, with a few survivals of the other declensions. § 24. (णा. The feminine affixes were, (i) the tbh. affix “I, -1< -ika; (ii) the tbh. -ni, -ni<*-inika, *-anika; and (iii) the ¢s. affixes -a and -ini, -Aani (in ८6. words). Grammatical gender largely obtains. Adjectives, occasionally the genitive in -kara which is properly an adjective, and the past tense of the verb (really a passive participle adjectival form) take the feminine affix -i when they qualify or refer to feminine nouns. Thus: takari pataka ‘his banner’ (18a); je athao nayika athikaha, sehao mandi hothi jakare riipé ‘those 8 Nayikas that are, even they become ugly (mandi) (by comparison) with whose beauty’ (18b); kafsani nayika (18b); BiSwakarmf@fie nirmmauli..mukha-ka 8079018 ‘beauty of face created by Viévakarman’ (20a); trayodasa-gune samyukti nayika (218) ; afsani usni dharani, aisani santapti prthwi bheli acha ‘the land became so hot, the earth so heated’ (30a); kajara-ka bhiti tele sicali aisani ratri ‘such a night as was like a wall of lamp-black moistened with oil’ (3la); Markkanda-ke sahodara jethi bahini 9188111 kufttani ‘a go-between woman like the elder uterine sister of Markandeya’ (418); kula-stri salajja bheli ‘the wife of the family became ashamed ’ (29b); awasanni nalini ‘ faded lotus’ (34a), etc. $ 25. NomBER. The old plural affixes (OLA. > MIA.) were lost: e.g., OLA. putrah—putrah > MIA. putto, putte—putta > Late MIA. (Apabhraméa) puttu, “putti, puttad—putté > Early NLA. pitu, *piti, piita—pit&. New devices had to be found out in NIA. to indicate the plural: but in the earlier stage of NIA. (as in the V.R.) INTRODUCTION xlvii ‘he distinction between plural and singular was not indicated, and ordinarily was left to the context. § 26. In the Early Maithili of the V.R., an affix -aha features commonly for the plural in the adjectives and passive particle (= past) forms: ९.4. je aneka bala ghola, se anutha; se kaisanfha— tarunaha, nonufiaha, baliaha, siiraha . . . baga baga-sdta pra- bhrti aneka a8wa-Siksa-prakara, t&ka uttirnaha ‘those numerous voung horses, they were brought; what were they like? young, gentle, strong, brave; and the numerous methods of training horses, with the reins, with tightening of the reins, etc., they were passed in all that’ (42b-43a); daSao je admanai-ka guna, té samyuktaha je dmanaik&ha, se rajadeSe hakari haluaha ‘those ten virtues of the servant (vassal), endowed with all of those the vassals (that were there), they raised a shout and marched at the order of the king’ (44b); kaisanaha betalaha = kidrSah vetalah (62a); etc. Byt this plural affix -aha is frequently omitted with reference to plural nouns, and on the other hand, it has become an honorific affix for the singular as well: e.g., nakgatra tirohita bhela, canda mlana bhelaha ‘the stars disappeared, the moon became pale’ (29b); aditya ...astacala gai apagata bhaiiaha ‘the sun went to the mountain of setting and passed away’ (300); Krsna caturbhuja bhae gelaha ‘Krsna became four-armed’ (18a); nayake paera pakhalala, suci bhae baisalaha ‘the master washed his feet, and purifying himself sat down’ (76b); udgita nigaraitaha, sama gabaitaha, aneka {81 - kumara dekhuaha ‘ many sons of 15015 were seen chanting udgithas and singing sémane’ (55a); ct. je aneka rastrakiha, se kaisandha (fom. form, rastraki+aha: 71b) raja dekhuaha ‘the king was seen’ (23a); etc. etc. This -fha seems to be the Apabhrathéa genitive singular affix = -asya of OJA.: source ?), which was extended to form tho plural. A similar use of the genitive singular for the nominative plural is not un- known in other New Indo-Aryan speeches: witness the origin of the Bengali -era, -(a)ra < -erd, -ara of the genitive singular, witness the use of the forms hamani-ka, tohani-ka = ‘we, you’, lit. ‘our, your’ in Bhojpuriyé. This -&ha affix gradually grew restricted in Modern Maithili to form the honorific of the verb intransitive, in the past tense: e.g., calal-ah(a) ‘he (they) went’: the intransitive past verb refers to the nominative, and retains its old participial and adjectival nature in this way. § 27. The instrumental plural has the affix -nhi (= 9 blend of the instrumental plural -hi of Apabhrarhéa, < -bhib of OIA., and of the genitive plural affix -na < -inim of OIA.). The -nhi affix xl viii INTRODUOTION was used for the plural oblique base, to which the genitive -ka wa added. Thus: gaja-rajé 82002 karu, bayasanhi kolahala kary ‘the royal elephant trumpeted, the crows cawed (lit. made noise)’ (29b); juwatinhi (= yuvati-) jalakeli arahu ‘the young women began their sports in water’ (30a); bhamaranhi padma tyajalg ‘the bees left the lotuses’ (30b); ulka-mukhanhi-ka udyota (62b). khadyotanhi-ka taranga, juwatinhi-ka utkantha (30b); etc. This instrumental-oblique -nhi now occurs in Modern Maithili as an honorific form referring to the object in a transitive verb: ९.0. dekhala-thi-nhi ‘they (or he, honorific) saw them (or him, honorific)’, = Magahi dekhal-thi-n. The old instrumental-nominative use of -nhi lingers in the Modern Maithili form lokanY, which is added as a noun of multitude to denote the plural. § 28. The Apabhraméa genitive plural affix -na (< -na, -namm < -&nam) became rare in Karly Maithill: a solitary example is jani kdficana-giri-k& &rhga mayiirana caraité acha ‘as if peafowl were sporting on the peak of a mountain of gold’ (212). § 29. Plural by agglutination of a noun of multitude appears to have begun. Thus woe have nayi(ka)-jana‘ youthful women, heroines’ (210), j6 brkga-samiha ‘all those trees’ (49b) ; but this sort of aggluti- nation (cf. loka, manava > mana or mana, sabha > sabha, jana, sakala, sarva > saba, otc., in other NIA. speeches) does not seem to have been popular in Maithili. Generally, we find the word aneka preceding the noun to emphasise the plural notion: instances are exceedingly common in the work. A numeral also definitely indicates the plural number. § 30. Casp. Tho following aro the affixes (post-positions are given below): " Nominative: No affix for the singular. For the plural -na seems to have been used, though it is very rare in the V.R. (see supra, § 28). Accusative: No affix. § 31. Instrumental. Singular (also extended to the plural), =€, -e (= -éna of OIA., MIA. -ena, -enath, -erh), contracted form as in stri < stri-é (7la); -hi, e.g., taru jihvahi biw&da (77a), whore it seoms to be the old locative affix -hi = OIA. *-dhi, or -smin, extended to the instrumental; -nhi for the plural (see supra, § 27). This -§ is still preserved in Maithill. § 32. dfenitive. For the genitive plural, there is the composite affix, -nhi+--ka (sce supra, § 27). Ina few instances, -8 (< -inam) has been found: ¢.g., bana-dewat&-k& dyatana ‘ in the shrines of the f INTRODUCTION xlix wood-nymphs’ (50a). or the singular (also extended to the plural), the affix -kara is found, only with the pronoun; and -ka, which is -k ‘1 Modern Maithili, found with the noun. A form -kai (rarely -ke) also features with the nominative in masculine, feminine, or neuter: this may represent a MagadhI Apabhraméa nominative *-kkal in place of the Magadhi Prakrit kade,