________________
40
THE INDIAN ANTIQUARY.
[FEBRUARY, 1892.
marked. The cause of the peculiarity, therefore, cannot well have had any connection with the form of the vowels.
I would suggest that similar statistical enquiries should be made with reference to some other leading letters; e. g., m, sh, the sub-scribed y, the super-scribed r; also with regard to the numeral symbols. I have little doubt but that from such statistics may result some further useful land-marks for the determination of dates of writing. I hope to pursue the enquiry myself, so far as leisure from official duties will permit me.
ADDENDUM Since the publication of the foregoing paper in the Journal of the Asiatic Society of Bengal, I have drawn up a sketch map of the distribution, and classifactory lists, of the inscriptions on which my conclusions are based. As they may be found useful, I add them here.
I classify (following herein Mr. Fleet) the early Indian alphabets into the South Indian and the North-Indian, distinguished by their forms of the letter m. Inscriptions which show the form are South Indian ; those which have y are North Indian.
The North Indian alphabet I divide into the North-Eastern and North-Western varieties distinguished by their forms of the letter sh (cerebral sibilant). Inscriptions which show the form (bi-partite square) U are North-Western, while those which have the form (looped square) are North-Eastern.
In the North-Western alphabet I distinguish the earlier Gupta and the later Post-Gupta varieties, which are distinguished by their forms of the letter y. Inscriptions which show the form N are written in the Gupta, those which have I are written in the Post-Gupta alphabet.
In India proper the North-Western alphabet displaced the North-Eastern about the end of the fifth century. The year 500 A.D. may be taken as a convenient epoch of this occurrence.
Not long afterwards the Post-Gupta began to displace the Gupta variety of the NorthWestern alphabet in India proper. The year 600 A.D. may be taken as a convenient epoch of the final displacement of the Gupta variety. Any inscription showing the old Gupta form of y may be placed before that date.
In the following lists the inscriptions are arranged in chronological order. Where the exact date is not known, an average date has been assigned, 1. e., some year within the reign of the king named in the inscription. For these reigns I have used my synchronistic table, published in the Journal, As. Soc., Bengal, Vol. LVIII. The average dates may be easily recognised by their having no equivalents in the column of Indian dates. In a few cases, viz., Nos, 5 16, 19, 20, 21 in List I, No. 11 in List II., and Nos. 11, 12 in List III., there is nothing, at present, available to fix their dates, except the test-letters themselves. These inscriptions, therefore, do not help to support my argument, and I have only included them in the lists for the sake of completeness, and moreover distinguished them by italic type. At the same time, seeing that the lists without them amply sustain my argument, they add some weight to the latter post factum.
With regard to Nos. 10, 11, 12 in List III. (marked by asterisks) I should explain, that in No. 10 (Kosam inscription) no instance of the letter sh occurs. So far, therefore, the test fails. But it should bo noted, that in the North-Eastern alphabet, the dental and the cerebral sibilants are formed very nearly alike, with a looped square; see, e. 9., No. 13 (Kabîum inscription). This looped or ringleted form of the dental s occurs frequently in the Kösam inscription, Accordingly, I have classified it with the North-Eastern list. For the same reason, I have included in that list the two Nos. 11 and 12; for though these inscriptions exhibit one or two cases of the cerebral sh, these are, in the first place, not very distinct, and in the second place, they all occur in ligatures. The latter are not trustworthy tests, for the North-Eastern form of
