________________
xxvi
THE INDIAN ANTIQUARY
[CHAPTER III
of m, developed, at a later time, the Nâgarl form (e), and its ringleted variety (/), by the production of the right lateral below the base linc..
The origin of the northern form of the Gupta m must be placed in the earlier half of the fourth century A.D. The starting point of the Gupta empire (Pâtaliputra) was in the East. On the coins and in the records of Samudra Gupta the older form of m with its curved sides (Fig. 9, a b) is still exclusively prevalent. But with his son Chandragupta II, who added the West to the empire, a total change takes place. All his coins and records show only the forms of m with straight sides (Fig. 9, c d). He commenced to reign about 375 A.D.; and he completed his conquest of the West about 395 A.D. His earliest known dated inscription of 407 A.D. (F.GI., No. 7, p. 36) shows the straight-sided m. Its locality Gadhwâ, Lat. 80° 38', is just within the eastern area. Another of his inscriptions, within the western area, at Mathura, Lat. 77° 43', which also shows the straightsided m (F.GI., No. 4, p. 25, Plate iii A) is mutilated and hence undated; but it may be some twenty years older. Anyhow, the fact that the straight-sided m shows no signs of a gradual origination or introduction, but with Chandragupta's western conquests, all at once, entirely supersedes the older curved-sided form of m in the records throughout the northern portion of the Gupta empire, proves that, at the time of that conquest, it must have been the established and prevailing fashion of writing m in the north-west of India. The beginning and growth of that fashion in the North-west itself, therefore, may with good reason be placed in the earlier half of the fourth century, though, of course, in calligraphic records of a particularly ornate kind, such as the Bijayagadh inscriptions of about 372 A.D. (F.GI., Nos. 58, 59, pp. 251-2, Plate xxxvi B. C.), the old form of m with its angular or curved sides, might tend to survive for some longer time. The only form of m, prevailing throughout the whole of the Bower Manuscript, in its calligraphically as well as cursively written portions, is the earlier of the two north-western forins, with its right side straight, but the left side twisted (Fig. 9, c; and Table I). So far, therefore, the graphic indications of the manuscript point to some time within the fourth century A.D. At any rate, they need not carry its date back of that century.
LU &
The northern type of the Gupta script, again, is divisible into two distinctly marked varieties, an eastern and a western. With regard to this division the most useful test letter is the character for the cerebral sibilants, as compared with the character for the dental sibilant. The original forms, in the Aśoka alphabet, of these two characters are shown in Fig. 10, a and f respectively. The form of the former Fig. 10. was soon modified, as in (b), by closing up the lower semicircle. In the East, gradually that semicircle was made to bulge out on the left, as in (c), and finally reduced to a sinall ringlet, as in (d), while in the West it was simply more or less angularized, as in (e). On the other hand, in the case of the dental s (f), its basal curve was angularized in the East, and at the same time Forms of the cerebral and dental its tail closed up to form a ringlet, as in (g), while in the sibilants. West the whole character was angularized, a triangle taking the place of the ringlet, as in (h). The final result of these modifications was, in the East, to cause the forms of the cerebral and dental sibilants, (d) and (g), to resemble each other so closely as to make them practically indistinguishable, while in the West the forms of the two sibilants remained quite distinct. It may be added that the western form of the dental sibilant occurs in
& 2
허리
*
& N N
18h i k
