________________
56
Paramatma-prakasa
So it will be necessary and interesting to compare and contrast Hema.'s Apabh. with that of P.-prakasa and see first, what features of the dialect of P.-prakada are recorded by Hema.; secondly, what features of it are not represented in Hema.'s grammar; and lastly, what points noted by Hema. have not got their counterparts in P.-prakaša.
On the Homogeneity of Hema's Apabh.-Hemacandra does not explicitly mention the dialects of Apabh, as it is done by Markandeya and other later authors. It has been already detected, and a careful study of his remarks and rules would show that his Apabh. is not a homogeneous one and that he has mixed together different dialects. By his remark "prayðgrahandd yasyapabhranda vidişð vakşyati tasyapi kvacit prakstavat saura sēni vac ca karyam bhavati" (iv. 329) understood in the light of iv. 396 and 446 as distinguished from other features noted throughout, it is clear that he accepts two bases for his Apabh., namely, Prakrit and śaurasenia whose characteristics he has discussed in his previous sections. The illustrations on and the sutras iv. 341, 360, 372, 391, 393, 394, 398 (especially its alternative concession), 399, 414, 438, etc., show elements of an Apabhraíśa which is not in tune with the dialect described by him in other sūtras. Some of these characteristics, when studied in the light of Prakrit dialects discussed by Hema., are mutually so conflicted that they are not possible in a homogeneous dialect.
Hemacandra's Apabh. Compared and Contrasted with that of P.-prakāša-Hemacandra's sutra 'svardnan svarah prayo pabhramsē' should not be understood as a licence for violent vowel changes; but it only means that in the Apabh. literature analysed by Hema. much liberty was taken in vowel-changes which could not be canonised in short, and hence this rule. In P.-prakasa we do not find such vowel-changes as would obscure the sense. A bit of liberty is taken in some forms: parin (V. 1. pari) =paramil. 28). vatthu as the Loc. or Inst. sg. form (II. 180); at times the case termination u appears even where it is not needed as in viņu (II. 59), sahu (II. 109); and very often the quantity of vowels, short or long, is ignored as in jiū=jivah (I. 40), niccu-nicah (I. 89), vivariü-viparitan (I. 79). At times a compensatory long vowel is obtained by simplifying the duplicate remnant of a conjunct group : isaru, nisu (I. 91), būdhaü (1. 91), phasat v. 1. påsat (ll. 112); against this tendency we have kacca = kaca (II. 78), also note nibhamtu (II. 88). Hema. has noted (iv, 410) that often e and o' are to be pronounced short. In our text they are necessarily short before a conjunct with the effect that North-Indian Mss.
1 Pischel : Grammatik, etc. $28. 2 Mr. Manomohan Ghosh of the University of Calcutta in his interesting paper
'Maharastri a later form of Saurasen!' (Journal of the Department of Latters Vol. XXIII, 1933, Calcutta University) shows that Praksta means pre-eminently Saurasent, the language of the Indian Midland, of which Maharaştri is only a later phase,
Jain Education International
For Private & Personal Use Only
www.jainelibrary.org