________________
$ 24, A.)
INDIAN PALEOGRAPHY,
(18) The modern form of ta with the vertical on the right, which occurs already in the Asoka edicte, reappears in the 8th century (plate IV, 22, XXI) and becomes the regular one in the 10th century.
(19) The modern form of tha, which has been derived from the notched one of the 7th century (plate IV, 23, XVII), is found already in the inscriptions of the same period (plate IV, 23, XVIII, &c.).
(20) (55) In the 7th century, the lower end of da is more clearly defined by a Serif (plate IV, 24, XVII, &c.), which soon after is changed into the characteristic tail of the modern letter. . (21) Already in the 7th century, the right side of na becomes occasionally a vertical, to the left of which the loop is attached (plate IV, 26, XVIII, XIX); compare also below, $ 30.
(22) On the transformation of pha by the development of a central vertical (see above, under 1), the curve of aspiration is attached first to the top of the new sign (plate IV, 28, XXII; V, 37, III, &c.). But in the 11th century it sivks lower down (plate V, 31, XII), and it occupies already in the 12th century the position which it has in the modern Devanāgari letter (plate V, 31, XX-XXIII). Retrograde archaic forms, like those in plate V, 31, II, XIV, are, however, not rare. Their occurrence has probably to be explained by the influence of the popular cursive alphabets,
(23) As ra was very generally pronounced ba, the ancient sign for ba was lost in Northern, Central and Western India, and it was replaced by ra in the inscriptions of the 7th and later centaries (plate IV, 29, XX; V, 82, II, &c.). In the MSS. the substitution occurs even earlier (plate VI, 37, V, VI). A new ba, consisting of va with a dot in the centre of the loop, occurs since the 11th century (plate V, 32, XVI), and this form is the parent of the modern Devanagari letter.
(24) The left limb of bha, mostly an inverted wedge with the point towards the right, is frequently changed into a triangle, open at the apex, from which the lower portion of the original vertical hangs down (plate IV, 80, XIX, &c.; V, 38, II, dc.). The modern Devanāgari bha appears in the 12th century (plate V, 33, XX, &c.) and seems to be derived from the form with the wedge, for which latter a Serif was substituted.
(25) Since the 8th century, ma usually has on the left a cursive loop (plate IV, 31, XX, XXI), which in the MSS. is mostly filled in with ink (plate VI, 39, XV-XVII).
(26) Both the MSS., and most inscriptions, with the exception of one from Udaypur (above, page 48, note 3) and some from Nepal (page 50), offer exclusively the looped or the bipartite ya, which latter occurs already in the inscriptions of the Kuşada period, and has been derived from the looped form. In the Nepalese inscriptions of the 7th century, wbich show the eastern $0,* we find a tripartite ya with a small circle at the top of the first upstroke (plate IV, 32, XVII); the Udaypur inscription has both the ordinary tripartite ya of the Gupta period, and also the bipartite letter.
(27) The right extremity of the wedge at the lower end of ra is often greatly elongated in the inscriptions of the 7th and later centaries (plate IV, 33, XVIII-XXI, &c.), and sometimes only the outlines of the wedge are marked. These forms are the precursors of the modern tailed ra.
(28) Since the 7th century, we find a cursive sa (plate IV, 36, XVIII, 42, XIX; V, 39, II, III, &c.; VI, 44, XV-XVII), the left half of which has been turned into a loop with a little tail on the right. i See above, $ 19, B, 12.
J.ASB. 60, 87.
3 J.ASB. 60, 85.
