________________
$ 26, A.]
INDIAN PALEOGRAPHY.
is a cursive substitute for the triangle. The triangle itself is a modification of the top-stroke with a semi-circle below, occasionally met with in ornamental inscriptions from Northern and Central India, as, e. 9., in Vinayakapāla's plate (letters with this peculiarity have not been giveu in plate IV, col. XXIII) and in the Candella inscription in CUNNINGHAM's Archwological Reports, Vol. 10, plate 33, No. 3. This last-mentioned form again is connected with, and gives the outlines of, the thick top-strokes, rounded off at both ends, which are not rare in ornamental MSS. like that figured by BENDALL, Catalogue of Sanskrit Baddhist MSS, from Nepal, plate ?,. Nos. 1, 2, and in the alphabet of plate VI, col. XIV (see particularly lines 5, 7, 15, 30, 84, 37, 49).
Among the abnormal single signs, not received into the modern Bengali, the following deserve special remarks :
(1) The forms of I in plate V, 3, XVIII, and VI, 3, X, are cursive developinents of the ancient I in plate IV, 3, IX, &c. But the I and I of plate V, 3, 4, XIX, appear to be southern forms; compare plate VII, 3, IV-VI.
(2) The curious ta of plate V, 20, XIX, seems to have been produced by an abnormally strong development of a “Nepalese hook" with a Serif at the end, placed above the ancient round !a, which is represented by the second lower curve on the left ; compare the ta of col. XVIII, and that of the Cambridge MS. No. 1693 (Bendall, op. cit. plate 4).
(3) The na of plate V, 29, XIX, without a connecting stroke between the loop and the vertical, is due to the strongly developed predilection for carsive forms, which is visible also in other letters of Vaidyadeva's inscription, such as A, Ā, sa and the ligatare tkr (plate V, 47, XIX).
(4) The triangular medial 14, for instance of ku (plate V, 10, XIX), which appears also in Lakşmaņasena's Tarpan-Dighi grant and other eastern inscriptions, gives the outline of the older wedge-shaped form, found, e.g., in thu (plate V, 26, XVIII) and in su (plate VI, 45, II).
(5) The Anusvåra of raih (plate V, 38, XIX) and of kan (plate VI, 15, X) has been placed on the line, as in the Old-Kanarese (see below, $ 29, C, 5) and the modern Grantha, and a Virāma stands below it.
(6) In the Om of plate V, 9, XVIII, we have the oldest example of the occurrence of the modern Anunâsika. In this case, it shows a little circle instead of the more usual dot, which is found in the Om of plate VI, 13, XI. Both forms are rather frequent in the eastern inscriptions of the 12th century, whereas in the wests they are more rare and are confined to the word Om. The Anunāsika, which I have not found in any Indian inscription older than the 11th century, probably is an intentional modification of the Annsvārs, invented because in Vedic MSS. the Ananāsika must be substituted for an Anusvāra followed by liquid consonants, sibilants and ha.
(7) [59] The Visarga of vah (plate V, 38, XVIII) carries a wedge at the top, which addition appears also in other ornamental scripts (see, e. g., plate VI, 30, XIV); in the ḥ of plate VI, 51, X (compare also VI, 41, XI, and the Gayā inscription), it has been changed cursively into a form resembling our figure 8. In the Gayā inscription (IA, 10, 342), as well as in MSS. of this period, it receives also a small tail (compare täh, plate VI, 30, XIV).
Compare the Gaya insoriptions in C.ASR. 3. pl. 87, No. 12; pl. 88, No. 13. . See the Mahoba insoription, C.ASR, 21, pl. 21.
* Compare the facsimiles of Bengali Mss. in Pal. Soo, Or. Series, PL 38, 82, 69; RAJENDRALAL MITRA, Notices of Sanskrit M88., Vol. 8, pl. 5,6; Vol. 5 and 6; and the Proto-Bengāli insoription, J.ABB. 48, 318, pl. 18.
