________________
154
THE INDIAN ANTIQUARY.
(MAY, 1892.
spelling of our inscriptions, not being consistent with itself, does not endeavour to accurately present the pronunciation.11
Kapur di Giri in several instances poes j and y, the one for the other: ja[w] (equivalent to yal),.,11; anija, VI, 16; samaya, I, 2; kanboya, V, 12; XIII, 9; raya, V, 11; VI, 14; jx, 18; X. 21, XI, 23 ; XIII, 1, beside raja, VIII, 17, &c. Perhaps even at Girnar we find an analogous case, if we must really really read (XI), 7) sruneju, for srundju, equivalent to sruneyu. At any rate, nayásu for niyásı, is purely sporadic, contrary to analogy, and, to all appearance, an arbitrary spelling.
These exceptional spellings follow a double direction. Several bear witness to an effort to approach the etymological forms. For instance, sudvisati, against all analogy, retains its final consonant. No one can doubt that Dhauli and Jaugada represent exactly the same dialect, and the same pronunciation, and hence ékatiya at Jaugada (1, 2) and saimyapalipati (IX, 16), as compared with ekucha and summá at Dhauli, can only be taken as kinds of tatsamas. So also with forms like akusma at Dhauli. Adhigichya, equivalent to adhikriya, for adhigicha, at Bhabra, shews us an orthography which is undecided and hesitating.
In other places the writing betrays by inadvertencies that the phonetic level of the spoken language has already fallen below that which is usually marked by the ruling habits of the written one. I refer to softenings like adhigichya for adhiksitya at Bhabra, libi alongside of lipi at Dehli, 18ga, logika, lahéyu at Jaugada, or, inversely, to irregular hardenings such as kunbôcha at Dhauli, paļipátayati at Jaagada, padha at Kapur di Giri, or, again, isolated inflexions like janáó at Khálsi, mahidayo at Girnar.
It would not be impossible to increase the number of indications of this nature; but, neither the condition of the monuments, nor the accuracy of our fàcsimiles, would allow us to attain to complete statistics. I stop myself here, and proceed to sum up.
11 The translator ventures to take the liberty of appending the following note by him on this character, which originally appeared in the Acaderay for October 1890:
This character has excited considerable controversy. It is admittedly & compound of the sign for k and the sign for y, and, graphically, it representa kya. In the Khelai inscription it is substituted (but by no means uniformly) for the k which we should expect in the tormination ikd; and it also oocurs in the foreign word alikyasadale. It is also found twice in the Delhi columnar inscription. All scholars agree that no completely satisfactory explanation has been given for this form. It seems to me that the following is not unreasonable.
The spelling of Piyadasi's inscriptions prononte several instances of false analogy. M. Sopart has given strouw reasons for believing that when Piyadasi at Girnar wrote af, he meant to represent the sound th. It was a mistaken attempt to revive an old-fashioned spelling. The seribe knew that Sanskrit aht became th in Prakrit, and benea wrongly assumed that every Prakrit tth was derived from sht. Therefore, to shew his learning, whenever he came to th, he wrote it , even in Ossee when a represented not ahs, but sheh.
I think that this skya is a similar instance of false analogy. The Magadhi Pr&krit termination ikd is liable to have its ponultimate vowel lengthened, thus, fkd. Then, by a well-known role, the can again be shortened, the Consonant following being at the same time doubled in compensation, thus, ikka. Instances of this sro aot uncommon in literature; and, judging from the modern languages of India, must have been extremely common in conversation. Prákpit examploy will be found in 6 203 of Dr. Hoornle's Gardian Grammar and I need not quote them here. As tho Khálsi and the Delhi inscriptions were written in Piyadasi's Magadhi dialect, we need not be surprised if we find this doubling occurring in them too.
Now Sanskrit ikya doos become ikka in Prakrit; and I believe that the seribe, coming upon an ikka with totally different derivation of which he was ignorant, and wishing to show his learning, represented that ikka also by ikya, just as his brother at Girnar represented tth by sht, even when it had nothing whatever to do with that compound. If we assume, as sdgrosted by M. Senart, that the soribe endea youred to connect the foreign word Alikyasadala with the Sanskrit allka (an instance of a common kind of word-play in Sanskrit literature), we find an additional confirmation of my suggestion. The fin allka is long its being shortened shows that the word must have been pronounced alikker in Piyadasi's time. Accordingly, the engraver, coming upon another kk, followed his custom and wrote it kya.
It will be observed that this accounts for the want of uniformity with which bya appears in Piyadasi's inscriptions. M. Senart shows that at KhAlai ikyd ooours seventeen and id seven times. At Dolhi there are only two instances of iky4, ikd being used everywhere else. So, also, in Magadhi Prekrit both the terminations ikkd and kd appear to have been concurrently and indifferently used, just as at the present day a man of Magadha will say in the same breath, chhotakd and chhotakka, tanild and tanikku, tanud and tanukka.-G. A. G.
