________________
63
Canda's manner and since two Mss. attest them, they may be regarded genuine. Now in one of them there is a peculiar Apabhramsa adverb 'dadavada' with its meaning. Hemacandra bas it under IV 422 in his Grammar but he has given it an unnatural meaning viz. 'avaskanda.' No. (8) gives davadava, a peculiar Deśī word, not uncommon in Apbh. But piost important Apbh. is No. (11), which lays down 'jima and tima' as substitutes of 'yathā and tathā' and gives an Apbh. Dohā as illustration, For translation of the verse see Pischel, There is another Apabhramsa Dohā in Appendix A but as it is supported by one Ms. only we have to leave it out of consideration.
Most of these Sūtras, some of them verbatim, some split up into two and some with slight change, are found in Heniacandra; but in such a manner that Hemacandra can always be called a borrower. That Hemacandra was a borrower on a large scale has been proved in other literary spheres also.
About Caņda's priority to Hemacandra there cannot be much doubt." But Hoernle's claim that his grammar presents a very old phase of the Prākrit language (p. XI) cannot be accepted, as it is based on the evidence of one solitary Ms. as against three that speak to the contrary. His other claims that Canda is prior even to Vararuchi and that he must have written his work at a "some-what later time than the 3rd cent B. C." are equally untenable. Caņda evidently lived at a time when the Apabhramsa had ceased to be a mere dialect of the Abhīras and become a literary language, i.e. after the sixth century A. D. and not before.
(3) Hemacandra:-Of all Prākrit grammarians Hemacandra a Jain like Canda is the most important from Apabhramba point of view. In his Siddhahemacandra, Prākrit part, he has dealt with it more carefully and at greater length than others and what is more important, has given Apabhramsa Dobās in illustration. He has in fact done more justice to Apabhramsa than to any other Prākrit except the Mābāsāstri. He deals with it in Sūtras IV 329 to IV 448. The so-called Dbātvādesa-sūtras IV 2 to IV 259 are also many of them really A pabhramśa Sūtras as they contain roots which mostly occur in Apabhramba. So that Hemacandra may be said to have treated the Apabhramsa in about 378 Sūtras as against Saurseni in 27, Māgadhi in 16, Paisāci in 26 Sūtras. Even if we omit the Dhātvādebas, the number for Apabhramsa still remains at 120.
Cf. Kavyamīmāms a of Rajasekhara, G. 0. S., Notes pp. 8, 9, 13, 14, 15 and the table at the end. Some of these references attest to wholesale borrowing of chapters on the part of Hemacandra.
Hærnle Op. cit. p. XXII, Pischel Op. Cit. $ 84. 3 Ed. Pischel, Part I, Halle 1877.
