________________
APRIL, 1903.] 1
As regards phonology, we may note the change of k to g in Jaina Mâhârâshtri, Ardha Magadhi and Magadhi; thus, Ardhamâgadht, Jaina Maharashtri savaga, Magadhi dvaga, a pupil. Compare, however, the remarks about d, f, above.
MAHARASHTRI AND MARATHI.
185
Though 8-sounds in Magadhi generally become s, we also find instances of the change of ss to h, which plays a great rôle in the formation of the future in Mâhârâshtri and Ardhamâgadhi. Compare forms such as Magadhi puttaha and puttassa, of a son.
We may further note the cerebralising of dental sounds and the change of ? to l (i. e, ) in Maharashtri, Ardhamâgadhi, and Magadhi. Compare Pischel, Grammatik der Prakrit Sprachen. Strassburg, 1900, §§ 219, 289, 238.
Such changes also occur in Saurasênî, but only sporadically.
If we turn to the inflexional system we find several indications that Magadhî is based on a dialect with the same rich variety of forms as Maharashtri and Ardhanagadhi.
Thus Magadhi has, like the two last-mentioned dialects, preserved the old dative of a-bases, at least in verses. (Pischel, § 361.) It has two different forms of the genitive of the same bases, thus, puttass and puttaha, of the son; two forms of the locative, thus, muhé, in the mouth, kúrammi, in the well. There is even a third form of this case, ending in him, thus, kulahi, in the family. The genitive plural ends in anu and dham, the vocative plural in d and dhe, etc.
Note also the form tô, therefore, in all dialects with the exception of Sauraseni.
The Atman&padam, which in Sanrasênî is only used in the first person singular, occurs also in other persons in Magadhi, not, however, so often as in Maharashtri and Ardhamâgadhi. Considering the scanty materials for our knowledge of Magadhi, this point is of importance.
Optatives such as karejjú, I may do, do not occur in Saurasêni, but are occasionally found \in Magadhî and are the common forms in Maharashtri and Ardhamigadhi. Forms such as lohé am or lahé, I may take, which are the only ones used in Saurasêni, on the other hand seldom occur in the other Prakrits.
Verbal bases ending in a short a usually form their imperative in a; thus, pira, drink. In Maharashtri, Ardhamâga dhi, and Magadhi, however, we also find forms such as picáki.
The suffix illo, which plays a great rôle in Mâhârâshtri and Ardhamâgadbi, is related to élua in Magadhi gámélua, Sanskrit gramya, and the modern dialects spoken in the old Magadha country show that an l-suffix must have been common in Magadha Apabramsa,
Such indications seem to show that Magadhi was based on a vernacular which was much more closely related to Mihârâshtri and Ardhamigadh! than to Saurasêni. It should be remembered that it is used both in prose and in verses, and there is nothing improbable in the supposition that the Magadhi of the prose passages has been largely influenced by Sauraseni, the principal prose Prakrit of the plays.
Thongh the so-called Apabhraméas will not be dealt with in these pages, it may be noted that the Saurasêna Apabhramén (as also the so-called Jaina Saurasênt) often differs from Saurasêni Prakrit and agrees with the other dialects. This may, at least partly, be explained by the fact that it is, of a somewhat different origin. The Saurasêna Apabhraméa is principally known from Hemachandra's grammar, and it is a well-known fact that in many points it agrees with old Gajarati. Now Gujarat was conquered from Mathurâ, and old Gujarati was, therefore, a Saurasena Apabhramsa. It has, however, certainly been influenced by the speech of the former inhabitants. We do not know who these were. The many points of analogy between Gujarati and Konkant, however, make it almost certain that their dialect was closely related to old Marathi.