LITERATURE.] SANSKRIT 293 .atantra. more practical system of arrangement the principal heads under which the grammatical matter was distributed usually being rules of euphony (sandhi) ; inflexion of nouns (ndman), gene- rally including composition and secondary derivatives ; the verb (dkhydta) ; and primary (krid-anta) derivatives. In this way a number of grammatical schools l sprang up at different times, each recognizing a special set of Sutras, round which gradually gathered a more or less numerous body of commentatorial and subsidiary treatises. As regards the grammatical material itself, these later grammars supply comparatively little that is not already contained in the older works, the difference being mainly one of method, and partly of terminology, including modifications of the system of technical letters (anubandlid). Ot the grammars of this descrip- tion hitherto known the Ghdndra-vydkarana is probably the oldest, its author Chandra Acharya having flourished under King Abhimanyu of Kashmir, who is usually supposed to have lived towards the end of the 2d century, 2 and in whose reign that grammarian is stated, along with others, to have revived the study of the Malulbhashya in Kashmir. Only portions of this grammar, with a commentary by Anandadatta, have as yet been recovered. The Kdtantra, 3 or Kdldpa, is ascribed to Kumara, the god of war, whence this school is also sometimes called Kaumdra. The real author probably was Sarva-varman, who also wrote the original commentary (vritti), which was afterwards recast by Durgasimha, and again commented upon by the same writer, and subsequently by Trilochana-dasa. The date of the Katantra is unknown, but it will probably have to be assigned to about the 6th or 7th century. It is still used in many parts of India, especially in Bengal and Kashmir. Other grammars are the Sdrasvatl Prakriyd, by Anubhuti Svarupacharya ; the Sankshipta-sdra, composed by Kramadisvara, and corrected by Jumara-nandin, whence it is also called Jaumara ; the Haima-vydkarana,* by the Jaina writer Hemachandra (1088-1172, according 'to Dr Bhao Daji) ; the Mugdha-l/odha, 5 composed, in the latter part of the 13th century, by Vopadeva, the court pandit of King Mahadeva (Rfimaraja) of Devagiri (or Deoghar) ; the Siddhdnta-kaumudi, the favourite text-book of Indian students, by Bhattoji Dikshita (17th cen- tury) ; and a clever abridgment of it, the Lagliu- (Siddhdnta-) kaiimudi, 6 by Varadaraja. Several subsidiary grammatical treatises remain to be noticed. The Paribhdslids are general maxims of interpretation presupposed by the Sutras. Those handed down as applicable to Panini's system have been interpreted most ably by Nagojibhatta, in his Paribdsliendusckhara. 7 In the case of rules applying 'to whole groups of words, the complete lists (gaiia) of these words are given in the Gfanapdtha, and only referred to in the Sutras. Vardha- mana's Ganaratna-mahodadhif a comparatively modern recension of these lists (1140 A.D. ), is valuable as offering the only available commentary on the Ganas which contain many words of unknown meaning. The Dhdtupdthas are complete lists of the roots (dhdtu) of the language, with their general meanings. The lists handed down under this title, 9 as arranged by Panini himself, have been com- mented upon, amongst others, by Madhava. The Unddi-sMras are rules on the formation of irregular derivatives. The oldest work of this kind, commented upon by Ujjvaladatta, 10 is by some writers ascribed to Katyayana Vararuchi, by others even to Sakatayana. The oldest known treatise on the philosophy of grammar and syntax is the Vdkya-padiya, 11 composed in verse, by Bhartrihari (? 7th century), whence it is also called Harikdrikd. Of later works on this subject, the Vaiydkarana-bhushana, by Konda- bhatta, and the Vaiydkarana-siddhdnia-manjushti, by Nagoji- bhatta, are the most important. IV. LEXICOGRAPHY. Sanskrit dictionaries (kosha), invariably composed in verse, are either homonymous or synonymous, or partly the one and partly the other. Of those hitherto published, Sasvata's Anekdrtha-samuchchaya, or " collection of homonyms," is probably the oldest. While in the later homonymic vocabu- laries the words are usually arranged according to the alphabetical order of the final (or sometimes the initial) letter, and then accord- ing to the number of syllables, Sasvata's principle of arrangement viz., the number of meanings assignable to a word seems to be more primitive. The work probably next in time is the famous Amara-kosha 13 ("immortal treasury") by Amara-simha, one of "the nine gems" at the court of King Vikramaditya (c. 550 A.D.). This dictionary consists of a synonymous and a short homonymous part ; whilst in the former the words are distributed in sections 1 Dr Burnell, in his Aindra School, proposes to apply this term to all grammars arranged on this plan. 2 Prof. Bhandarkar, Early History of the Dekhan, p. 20. proposes to fix him about the end of the 3d century. 3 Ed., with comm., by J. Eggeling. 4 1'art ed. and transl. by R. 1'ischel. s Ed. by 0. Bohtlingk, 1847. 6 Ed. and transl. by J. R. Ballantyne. For other modern grammars see Colebrooke, Essays, ii. p. 44 ; Rftjendraiaia Mitral Descriptive Catalogue, i., Grammar. 7 Ed. and transl. by F. Kielhorn. 8 Ed. by J. Eggeling. Ed. by N. L. Westergaard. 10 Text and commentary, ed. by Th. Aufrecht. 11 In course of publication, with commentaries, at Benares. 12 Ed. by Th. Zachariae. [ 13 Edited by H. T. Colebrooke (1808), and by L. Deslongchamps (1839-45). according to subjects, as heaven and the gods, time and seasons, &c., in the latter they are arranged according to their final letter, without regard to the number of syllables. This Kosha has found many commentators, the oldest of those known being Kshira- svamin. 14 Among the works quoted by commentators as Amara's sources are the Trikdnda and Utpalini-koshas, and the glossaries of Rabhasa, Vyadi, Katyayana, and Vararuchi. A Kosha ascribed to Vararuchi, whom tradition makes one of the nine literary "gems," and hence the contemporary of Amara-simha, consisting of ninety short sections, has been printed at Benares (1865) in a collection of twelve Koshas. The Abhidhdna-ratna- mdld, 15 by Halayudha ; the Visvaprakdsa, by Mahesvara (1111) ; and the Abhidhdna-chintdmani (or Haima-kosha], by the Jaiua Hemachandra, seem all three to belong to the 12th century. Somewhat earlier than these probably is Ajaya Pala, the author of the (homonymous) Ndndrtha-sangraha, being quoted by Var- dhamana (1140 A.D.). Of more uncertain date is Purushottama Deva, who wrote the Trikdnda-sesha, a supplement to the Amarakosha, besides the Hdrdvali, a collection of uncommon words, and two other short glossaries. Of numerous other works of this class the most important is the Mcdini, a dictionary of homonyms, arranged in the first place according to the finals and the syllabic length, and then alphabetically. Two important dictionaries, compiled by native scholars of the present cen- tury, are the Sabdakalpadruma by Radhakanta Deva, and the Vdcliaspatya, by Taranatha Tarka-vachaspati. A full account of Sanskrit dictionaries is contained in the preface to the first edition of H. H. Wilson's Dictionary, reprinted in his Essays on Sanskrit Literature, vol. iii. V. PROSODY (Chhandas). The oldest treatises on prosody have PROSODY. already been referred to in the account of the technical branches of the later Vedic literature. Among more modern treatises the most important are the Mrita-sanjivani, a commentary on Pingala's Sutra, by Halayudha (perhaps identical with the author of the glossary above referred to) ; the Vritta-ratndkara, or "jewel-mine of metres," in six chapters, composed before the 13th century by Kedara Bhatta, with several commentaries ; and the Chhando-manjari, likewise in six chapters, by Gangadasa. The Srutabodha, ascribed, probably wrongly, to the great Kalidasa, is a comparatively insignificant treatise, dealing only with the more common metres, in such a way that each couplet forms a specimen of the metre it describes. The Vritta-darpana treats chiefly of Prakrit metres. Sanskrit prosody, which is probably not surpassed by any other either in variety of metre or in harmonious- ness of rhythm, recognizes two classes of metres, viz., such as con- sist of a certain number of syllables of fixed quantity, and such as are regulated by groups of breves or metrical instants, this latter class being again of two kinds, according as it is or is not bound by a fixed order of feet. A pleasant account of Sanskrit poetics is given in Colebrooke's Essays, vol. ii. ; a more complete and systematic one by Prof. Weber, Ind. Stud., vol. viii. VI. Music (Sangita). The musical art has been practised in Music. India from early times. The theoretic treatises on profane music now extant are, however, quite modern productions. The two most highly esteemed works are the Sangita-ratndkara ("jewel- mine of music "), by Sarngadeva, and the Sangtta-darpana ( mirror of music"), by Damodara. Each of these works consists of seven chapters, treating respectively of (1) sound and musical notes (svara) ; (2) melodies (rdga] ; (3) music in connexion with the human voice (prakirnaka) ; (4) musical compositions (prabandha) ; (5) time and measure (tdla) ; (6) musical instruments and instru- mental music (vddya) ; (7) dancing and acting (nritta or nritya}. The Indian octave consists like our own of seven chief notes (svara) ; but, while with us it is subdivided into twelve semi-tones, the Hindu theory distinguishes twenty-two intervals (sruti, audible sound). There is, however, some doubt as to whether these Srutis are quite equal to one another, in which case the intervals between the chief notes would be unequal, since they consist of either two or three or four Srutis, or whether, if the intervals between the chief notes be equal, the Srutis themselves vary in duration between quarter-, third-, and semi -tones. There are three scales (grdma), differing from each other in the nature of the chief intervals (either as regards actual duration, or the number of srutis or sub-tones). Indian music consists almost entirely in melody, instrumental accompaniment being performed in unison, and any attempt at harmony being confined to the continuation of the key-note. A number of papers, by various writers, have been reprinted with additional remarks on the subject, in Sourindro Mohun Tagore's Hindu Music, Calcutta, 1875. Compare also " Hindu Music," reprinted from the Hindoo Patriot, September 7, 1874. VII. RHETORIC (Alankdra-idstra). Treatises on the theory of RHET- ORIC. 14 A grammarian of this name is mentioned as the tutor of King Jayaplda of Kashmir (755-786 A.D.) ; but Kshira, the commentator on Amara is placed by Prof. Aufrecht between the llth and 12th centuries, because he quotes the 3ab<3&nusasana ascribed to Bhojaraja. i* Ed. by Th. Aufrecht (1861) Ed. by 0. Bohtlingk and C. Rieu (1847).
