________________
182
THE INDIAN ANTIQUARY.
(SEPTEMBER, 1914
$56. The Inflectional declension is limited to the cases following: nominative, accusative, instrumental, ablative, genitive, locative and vocative. Of these the nominative and accusative have the same termination and so have on the whole the instrumental and locative, the confusion having already taken place in the Apabhramça. Further, the ablative has lost its original case meaning and has passed into that of the locative, a change of which there are also traces in the Apabhramça. In the usual grammars of Modern Indo-Aryan Vernaculars, the instrumental and genitive cases are now termed as agentive and oblique, but I prefer to hold to the older terms, as being more correct from the point of view of historical grammar. Nouns are not all subject to inflection in the same degree. As a rule inflection is common to all nouns in the instrumental, ablative, locative and vocative cases only; in the other cases only vocal bases are inflected and consonantal remain unchanged. There are, however, some exceptions, chiefly formed by consonantal adjectives which may be inflected in all cases, consonantal nouns which are sometimes inflected in the nominativeaccusative singular, and vocal nouns in oi, ou, which are not inflected in the nominativeaccusative and genitive. In the latter three cases, bases in oi, ou may also optionally remain uninflected and bases in a are uninflected as a rule. Feminine bases in oa, o are subject to inflection only in the instrumental and locative, and feminine adjectives in oi remain generally unchanged in all cases alike. Let us now proceed to deal with each case particularly.
$57. Nominative-accusative singular.-(1) Masculine vocal bases take the termination -u, from Apabhramça - <Skt. -ah, -am. Ex.:
prahuraü Âdi 51, velaü P. 548, kucaliu Adi 77, vivekarûpiu hâthiu Çil. 1,
pau Çal. 26, râu Çal, 109, Vi. 59, Ratn. 150. Cousonantal bases and vocalic bases in a remain uninflected and so bases in oi optionally. Ex.:
vidväsa Adi. 75, bâlaka Kal. 5,
sarathi Çrâ., raja Adi. 81, Rarely consonantal bases take also -u Ex.:
Jinavaru Rs. 196, murativantu Çal. 28, bokadu Indr. 77.
In the accusative singular, masculine bases in aa form an exception in that they may optionally take the ending Raü instead of aü. This ought not to be considered as an irregularity, but rather as a survival of the Apabhramca habit of representing Sanskrit kam by oũ, instead than by ou (See Pischel, $ 352). Instances of such nasalized accusatives are chiefly met with in the declension of pronouns and adjectives. Not rarely ai is contracted to a, according to $ 11, (3). Modern Gujarati and Marwasî contract oaü into .
(2) Feminines have the nominative-accusative identical with the base. Substantival feminine bases end mostly inoa, oi, rarely in oa, oi. Adjectival feminine bases end always in o. So appears to be the termination characteristic of the feminine gender in Old Western Rajasthani. In Apabhra ça the oi feminine termination had already begun to supersede oa, not only in adjectives, but also in substantives (Cf. bâli, Pischel's Materialien zur Kenntnis des Apabhramça, XVI). Examples of the four classes of feminine bases are :
måla Dd. 5, kanya, Vi. 125, ghadi Âdi. 20, pútali Dd. 3, pida Çal. 33, tarasa P. 541, ana Çra.
säpini Kal. 35, tâņi P. 366, koti P. 391, bhamuhi P. 564, seji P. 344, vakhari Çal. 110. Observe that bhamuhi and seji in the last class are from original nouns in oá, viz. Skt. *bhruvika, çayya (Pischel, SS 206, 124). These feminine bases in i have lost the latter vowel in
