________________
It is curious to note that a grammarian like Hemacandra should not have noticed any dialects of the Apabhramsa, although the rhetorician Nami-sādhu a hundred years earlier, noticed them. But that Hemacapdra is mixing up at least two dialects is clear from his Sūtras and their examples. (1) Thus retaining of e. g. trņu and sukțdu along with taņu and sukiu in IV 329, grhanti under IV 34, kțdantaho under IV 370, groņa for graha IV 394 and grhneppiņu under it, Gphanti IV 341; (2) change of non-initial hard unaspirates into soft unaspirates and of hard aspirates into soft aspirates, IV 396 e. g. vicchohagaru, sugha, kadhidu, savadhu and sabhalau; compare nādho, kadham in Sauraseni, IV 267;(3) change of non-initial m to nasalised V IV 397, e.g. kavalu, Chavaru; (4) retaining r as second member junct, IV 398 prieņa, prāu IV 393; dhruvu IV 418; pemmadrahi IV 423; prassadi IV 393, dhrum, tram IV 360; drammu, dravakkau IV 422: Prayāvadi IV 404; (5) alternature forms in û, and hứ for the Present First Pers. Sing and P1, IV 385 and 396 kattau, lahahu; (6) forms in di and hi for Pres. 3rd Sing, and Pl. IV 382 (for latter); e. g. Prassadi IV 393 (indication of Saur. character) dharahi, karahî; (7) forms in i, u and e for Imper. 2nd sing. IV 387; sumari, karu (IV 330) kare etc.; (8) forms in 'sa' instead of 'ha' for Future, IV 388 hosai (indication of Sauraseni character); (9) alternative forms for Passive, IV 389 kīsu; (10) forms like jāmahı, tāmahi IV 406. Thus side by side with the Māhārāstra (?) Apabhramsa, he may be said to be treating partially of Sāuraseņa Apabhramsa.
Although he does not mention any dialects, Hemacandra's treatment of the Apabhramsa is very complete. The value of his Grammar is enhanced by the quotations in verse that he has given to almost every Sutra; sometime one single Sutra has several verses as instances. Until more literature in Apabhramsa is published, it is not possible to trace the source of them; but they appear, as Pischel has rightly remarked, to be taken from some anthology like Hāla's Sattasai. Although many of the Dohās (a few other metres) are erotic in character, there are: (1) about eighteen verses heroic in character, (2) about sixty didactic (3) about ten religious with Jaiu bent and (4) five legendary, Mythological,-one about Kệsņa and Rādhā, another about Bali and Wāmana, one about Rāma and Rāvaņa and two about Mahābbārata. Among the erotic verses there are two which refer to Munja and might have been composed during the days or immediately, after the death of that un. fortunate king of the 10th century. Two (IV 357, 2. 3 and IV 420, 5) are found in Sarasvatīkanthābharaṇa a work some hundred years older thau Hemacandra. Possibly therefore Hemacandra seems to have culled his quotations from literature that was composed in the 9th and 10th centuries.
Hemacandra's arrangement is the opposite of that of Caņda; first comes Svaravidhāna, then Vyañjapavidhāna and lastly Vibhaktividhāna, This has become the model of later writers.
