1891.] A. F. JEt. Hoernle— On the date of the Bower Manuscript. 85 
as I have already remarked, was, in India proper, as early as the middle of 
the sixth century, superseded by the North-Western alphabet; but in 
Nepal it survived about three centuries longer, and there, with it, the 
old form of ya survived, at least in inscriptions, down to the middle of 
the nineth century A. D. It should be noted, however, that the old form 
of ya , in the shape in which it survived in Nepal, is somewhat different 
from the old form in its original shape, as it was once current in the 
North Indian alphabets. Its original shape is that of a sort of trident, 
of which the left-hand prong makes a curve or even a loop, thus c JX 
or oIn the Nepalese shape, the curve or loop, is replaced by a 
ringlet which is poised on the top of the left-hand prong, thus .* 
The difference is marked, and the two shapes can be very easily distin¬ 
guished from each other. 
The North-Western alphabet is the first to discard the use of the 
old form of ya. From it, as I shall presently show, the old form dis¬ 
appeared, even in inscriptions, as early as the end of the sixth century 
A. D.; and from cursive writing in that alphabet, according to the well- 
known rule, above stated, it must have disappeared much earlier. There 
is an obvious conclusion, which is suggested by these facts; it is, that 
the invention, so to speak , of the cursive form of YA took place in the North- 
West of India, somewhere within the area in which the North-Western 
alphabet was current. 
The first document from which the use of the old form has 
entirely disappeared is the long Bodhgaya inscription of Mahanaman of 
588 A. D. (Fleet, p. 274). It uses exclusively the transitional form, 
with one or two exceptions in which the modern form itself is used.f In 
another short Bodhgaya inscription of Mahanaman, of about the same 
date (Fleet, p. 278), the modern form is used exclusively. In fact, after 
600 A. D., there is no inscription known, which shows any trace of the sur¬ 
vival of the old form. In all of them the cursive form of ya is fully estab¬ 
lished in exclusive use ; thus in the Lakkha Mandal inscription of about 
600 A. D. ( Epigr. Ind., vol. I, p. 10),£ the Madhuban inscription of Harsha, 
of 631 A. D. (ibid., p. 67), the Aphsad and Shahpur inscriptions of Aditya- 
sena, of about 672 A. D. (Fleet, pp. 200, 208), the Deo Baranark inscription 
of Jivita Gupta, of (about) 725 A. D. (Fleet, p. 213), the Sarnath inscrip¬ 
tion of Prakataditya of somewhere in the seventh century (Fleet, p. 
284). To these may be added the evidence of those Nepalese inscrip- 
* See, e. g., the inscription No. 8, in the Ind. Ant., vol. IX, p. 171. 
t The transitional form is here used with a somewhat modified and more 
ornate shape. 
£ The intermediate forms occurs twice in this inscription, in yena , 11. 6 and 11, 
curiously enough, with the vowel e, on which see page 89. 
