________________
152
THE INDIAN ANTIQUARY.
(MAY, 1892.
The groups which the grammarian expressly writes , with the dental. (cf. Sútra 289), the Msichchhakali, extending the use of the palatal é peculiar to Màgadhi, writes áll, and the verb tishtha ti, for which the spelling chishthadi is expressly enjoined by Satra 298, is written in the drama chitadi (Pischel, loc. cit.). Between the grammarian and our inscriptions there is a still wider discord : !! is no more written 8! at Girnar, than rth is written st.
The mere observation of facts such as those which exist at Girnar would be sufficient to awaken our scruples. I find it difficult to believe, as Dr. Pischel has ingeniously suggested, that the absence of the aspiration in stita and sésta, are a direct inheritance from the primitive period which existed before the birth of the secondary aspiration of Vedic Sanskrit. Should we farther conclude that the word sresta at Kapur di Giri (1st edict) is also a witness of this same period, when the sibilant sh and the other cerebrals. had not as yet developed ? As for claiming the same antiquity for the Pali form affa (equivalent to arta) for artha, the uniform use of the aspirate in all our versions is far from favouring this conjecture. In any case, the Pali spelling atta being uniformly absent from all our inscriptions cannot be relied npon as a basis for the archaic origin of the I in oțita. I therefore consider that I am right in doubting whether the popular pronunciation had really eliminated the aspiration, in a case in which, as everyone knows, as everyone can judge by a reference to Praksit orthography, the consonant is invariably aspirated, even when the aspiration is not original, 1.c., when Sanskrit does not write it as aspirated. Is it really to be believed that the people pronounced matúna (Girnar, VI, 9, 10), when the assimilated form utth dna is the only one used, even in the learned language and in its system of etymological spelling? If they really did pronounce stana, sita, can ustdna be considered as anything but a parely orthographical approximation to these words, guided and determined by the feeling of etymology P The forms anusasti (for anusasti, the only probable one) beside sashstuta, gharastáni (instead of afani), beside stita, and at Kapur di Giri, sresta (instead of bréfa) by the side of br&than (IV, 10), tistiti beside tithe and adhithana (V, 12; al.), dipista beside afha (= ashtau) are as many errors which it would be hard to explain if we considered the orthography as an actual expression of the existing pronunciation.
Now, Girnar is comparatively near the tract which furnishes us numerous inscriptions for the period following. Would it not be surprising that in none of them, not even in the most ancient, at Sabchi and at Nânághát, has a single trace of so significant a dialectic peculiarity been discovered ? What we do find is at Sáñcht (No. 160), the proper name dhamastkiri, while in all the analogous instances, sethin, &c., the assimilation is carried out. Again at Kårli (No. 22), in a text of the time of Vasithipata Sân takani, we find hitasughasth[i]tay[d, beside nithito. In this instance forms such as puttasya, sovusakasya beside budharakhitasa, wpásakasa, leave no doubt as to the nature of the spelling. We have here a text couched half in Prakrit, half in mixed Sanskrit, and we know, without any hesitation, that the spelling sthiti is a tatsama, or, which comes to the same thing, an instance of learned orthography. Does not all analogy, every probability, compel us to accept the same conclusion for Girnar ?
It is true that this mode of writing, at and <!, appears at Girnar with a certain regularity, but this should not mislead us, after the facts which we have already pointed out regarding groups which contain an . I maintain that st and are conservative methods of spelling the groups tth and th which arise in Prakrit from a dental or cerebral sibilant followed by its mute. They have been extended to groups originating from sth and shth (that is to say a dental or cerebral sibilant followed by an aspirated mute), for the very simple reason that, in the assimilation of Průksit these groups result in the same pronunciation as do st and sht. From
I do not sperk of chilathitika in the inscription of Piyadasi. It is in Magadh, and, as we shall see, ondnot be taken as an authority for the local dialect.
At Kapur di Giri, the analysis, st, which Dr. Bahler has proved for a sign hitherto generally read th, har drawn the spelling of the word sresta from ita isolation : but the inconsistencies which bave been cited above in the transcriptions of the Swokpit groups shy, shih, still remain not one whit low charaéteristic and instructive.
