________________
No. 17.)
INSCRIPTIONS ON THE MATHURA LION-CAPITAL.
137
of Plate No. 1 are based chiefly upon these inscriptions. The type may be described as intermediate between the Asoka forms and the cursive derivatives of the Dhammapada MS. and those discovered by Dr. Stein, to which approximate the vase scratchings from Manikišla and other places and the inscription of Sue-Vihar (see the Plate given by Dr. Hoernle in the Indian Antiquary, Vol. X. facing p. 325). It does not, however, appear to be quite identical with that exhibited in any of these or in the Taxila plate (see the facsimile above, Vol. IV. facing p. 56), where on account of the historical contiguity we should expect to find, and do find, a near congener. The inscriptions edited by M. Senart in the Journal Asiatique, Sér. IX. Vol. XIII. (1899), p. 526 ff. and Vol. IV. (1894), p. 504 ff. (No. 35), are also in similar characters (see the Plates).
The forms of the individual letters may be seen from Plate IV. Those of tha, sa, mu, spa, the position of medial e and the looped form of u have been described by Bühler. Attention may be called to the rda in Q. 1, the spa in A. 7, the rna (?) in I. 1, the curious form in J. 1 interpreted by Bühler as shte, the archaic ma with the original angle beneath in A. 2. A few points call for further consideration. The form of ta, which occurs many times, has in other records the value of tra (see the Plate). Although we have certainly one case (J. 2, in kürita), and perhaps two others (E". in krakdrita ?, and R. 1 in Tachhilasa or Rachhila sa ?), where the bottom curve is wanting, the character occurs so often that there can be no doubt as to what it denotes. In supposing it to be a da Bühler, who in his articles in the 2. D. M. G. Vol. XLIII. pp. 133 and 294-6 bas read a similar character as tt (60 also Johansson, Der Dialekt der ShahbasgarhiRedaktion, I. p. 126), was influenced by finding it in pratesó (M.2) and similar cases. But the regular form of da is several times instanced, and the t for d in prates and elsewhere is a dialectical, not an orthographical, feature, found in other Kharðshthi records.
The only question remaining, as concerns ta, is part of a larger matter. How do we find represented the conjunct? In a number of cases the addition is made by a stroke inclined to the left attached to the lower curve. This occurs in A. 1 (chhattrao), A. 6 (matra), A. 7 (pitrá). In these examples there was some slight danger of confusion with tô, which shows (see the Plate) a similar stroke attached to the upper curve. Perhaps for this reason another device is more frequently adopted, namely a dot in place of the stroke ; see B. 1 and 3 (chhatra), B. 2 (putra), M. 1 (chhatrao), Q. 2 (chhatrao). There remain A. 3 (where te is written), A. 5 (where I can detect no clear indication of the r-stroke or dot), G. 1 (where there is nothing attached to the ta, though the reading may be chhratava, if not chchhatava), G. 2 (chhatava), A. 9 (where we seem to have atraürêna for até). In A. 1 we have two strokes, both of which seem to be intentional; one of them, which resembles the downward stroke of the Aramaic and Pahlavi t, I have conjectured to denote a doubled t before the r. Of a cerebral (ta) I can find no trace, unless indeed it is contained in the mysterious third sign of A. 13.
With other consonants than t the, is regularly inserted where required. We may quote bhra in A, 7 (where it takes the form of a detached stroke), gra in A. 2 and N. 2, pra in A. 10 (pradhravipra'), A. 11 (prali'), M. 3 (pratesó), N. 3, dhra in A. 10 and M. 2, chhra (chchha ?) in G. 2 and I. 1 (chhratavao and chhratrava). In the last example, as also in pradhravi (A. 10, padhruvf in I. 2), it would seem to be even repeated; but this may be a feature of pronunciation rather than of writing. A really otiose r-stroke is found in maheshri, A. 2, and probably in Pispasri, A. 7.
The case of kra is peculiar. If we disregard the doubtful inscription E"., it is found four times, in bhakravató (A. 12), nákraruasa (F. 1 and N. 1), and sakrastánasa (P. 1), i.e. in places where we expect a simple ga. This can hardly be accidental; it would seem probable that in these cases the combination kra represents a spirantic pronunciation of ga, analogous to the Persian ; more especially as in the name of Seistán (Pahlavi Sayhastán) this sound no doubt prevailed at the time.
