________________
326
THE INDIAN ANTIQUARY.
[NOVEMBER, 1988.
which the Vedic Literary language was deve the more common. Other documents (Vedic as loped, or at least some of them, died out at an well as Classical Sanskrit) appear to know only early date.
náth (except perhaps in vayonddha). Pali has The author next quotes Professor E. Kuhu likewise natha, Prakrit has nddha and ndha. Here (Beiträge sur Pali-Grammatik, p. 10), in support it seems probable that the PAli and Sanskrit of the view above advanced. He then considers have preserved the older form, while the Rig.v. examples of PAli, Prakrit, and modern forms prefers a form influenced by a popular dialect. which must be referred back, not to the language Again, cf. Rig-V. árbhaga with arbhaká; V tuj, of the Rig Veda, but to the popular dialecte which tuji (RV) with tuch, toka, &c.; karta (RV, developed contemporaneously with it. Examples AV) with garta (first appearing in the Brahof this kind have been given by Kuhn, but some manas). are doubtful. But, at any rate, the PAli rukkha (4) In the old literary language, a dental beappears to go back to an Indo-Germanic by-form came a cerebral only under the influence of besides the Sanskrit vriksha. So also the Påli certain special sounds. A dental explosive besuroti, sunati, appears referable to a form bruņoti, came cerebralized through the influence of a older than the Sanskrit frinoti.
preceding original sh, ah, or . The dental It is well known that already in the Vedic nasal was also cerebralized under the influence language there are appearances which remind of a preceding t-sound. In the more modern one of Praksit. A complete and systematic dialects, the cerebralization of a dental occurs examination of these is desirable, but would under & much wider range of circumstances. exceed the bounds of the essay. He, therefore, Dental explosives are cerebralized under the inproceeds only to collect a series of Vedio forms, fluence of an t-sound also, and even without any which run contrary to the sound-tendencies, apparent reason, beyond the tendency of some as hitherto known to us, of the language of the dialects towards cerebralization. Still more proHymns, and to examine whether the forme minent is the general tendency to cerebralize the comply with the sound-tendencies of later Indo. dental nasal. This kind of cerebralization of a Aryan dialects, or at least are in accordance with dental can in the cases of certain words be pointed the general direction of these tendencies. This out in the Rig V. We find kata (Prakrit form of may enable us to answer the question now under karta), katya; nala, cf. ná? (later nala; Pâli, consideration.
nala, nart, nási, ndit, ndli, beside nada): again He begins with a few isolated forms, which kévata (Rig.V.); avata (Ath.-V.), beside avata appear to be loan-words in the Vedic language. (RV); markața (V9, TS); in all of which there
(1) In Vedic and Classical Sanskrit, & in certain is possibly a change of dental (suffix ta) to a cases becomes sh. In the later dialects the cerebral. These changes are still more frequent three sibilants of Sanskțit are treated as one,- in the later language; cf. bhan beside (RV) usually the dental. Now there are Vedic forms bhan; at beside (RV) at; nat beside nart, nrit. like ribisa, busa, brisaya. These may be consi. The conjecture is not unreasonable that the forms dered as loan-words from other Indo-Aryan in which the cerebral appears have arisen from dialects. Indeed the 8, if it represents an Aryan the influence of Indo-Aryan popular dialects. $, and perhaps also the b, may point to Iranian (5) The old literary language of India, the influence. Cf. further Rig.V. prishant, prishati, Classical as well as the Vedic Sansksit, retained, besides prisni, Greek tepkvos; Ath..V. ruéant, we know, the old r vowel in its entirety. The beside Rig-V. rusl.
remaining Indo-Aryan languages agree, on the (2) Vedio and Classical Sanskrit usually pre- contrary, in endeavouring to rid themselves of serve medial consonants, while the later Indo- it, in consequence of which the r-sound usually Aryan dialects often drop intervocalic medial entirely disappears, and another short vowel, a, i, consonants; cf. titai, 'a sieve,' praiga (Rig-V.) or u, enters into its place. We find analogous
a fore yoke,' which are probably loan-words from examples in the Rig Veda; e. g. giha beside other Indo-Aryan dialects. The former the authorgriha ; edh, 'dhate, beside ardh, ridh; 50 also connects with tak, and the latter he refers to perhaps jah, beside (jrambh)jrimbh (cf.jrimbha) *prayuga.
and (RV) hesh, to be referred to *hrish. Here (3) In the later Indian dialects there is a ten- ri is represented by d, and not by a, i, or u. The dency to change an intervocalio tenuis, and later dialects, however, sometimes have &; thus tenuis aspirata, into a media and media asp. (Pali and Pråkpit) geha, beside giha, ghara; edh respectively. Cf. Rig.V. uddhin nddhamana, occurs in PAli, beside idh, hati, iddhi, Praksit uddhita, nddhas with the rarer náth in náthita, iddhi; hesh appears in Pali and Prakpit regularly andtha. In the Ath-V. náth (ndthita, natha) is as hes. If jéh in Rig Veda is to be referred to
