________________
86
EPIGRAPHIA INDICA.
[VOL. XI.
The inscription has suffered so much damage that we can form no opinion as to what its purport was after the opening part of it. Its interest lies in the point that it is a lithio record of a king, the Svāmin Virasēna, who may be identified with practical certainty with a king Virasons of whom we have coins. Professor Rapson has described and figured one of the coins in the Jour. R. As. Soc., 1900, p. 115, and notes there certain symbols on it which have some general resemblance (but nothing specific except the tree) with the symbols on this stone. And Mr. V. Smith has described fourteen of them, and figured two, in his Catalogue of Coins in the Indian Museum, Caloutta, p. 197, and conjectures there (p. 191 f.) from the provenance of them that Virasēna was a king in the Gangetic Doab about A.D. 300.
The record is dated in the thirteenth regnal year of Virasena,' and on the eighth day in the fourth fortnight of the hot season. As no year of an era is given and the date of Virasina is not known, we can only fix on palæographic grounds the period to which the record may be referred ; and the following remarks are offered towards elucidating this point. The letters that occur clearly are k, 9, t, u, P, M, y, r, u, sh and s, and they may be compared with the Brühmi forms in Tables II and III in Bühler's Indische Palæographie. The facts to be ascertained must be, not what are the earliest records with which letters found in this inscription tally, for particular forms, such for instance as those of k and , persisted with little or no modification for several centuries, but rather what was the period when any later or new forms found in this inscription came into use. The forms of certain of these letters, namely, m, y, v, and sh do not occur in Table II and are first found in Table III; hence these letters deserve most notice and attention may be confined to Table III. The form of m resembles most those in columns VI and VIII of that table; that of y those in columns I, II, IX and XIII: that of those in columns I and VIII; and that of sh the form in column XIV. The forms of v and v are therefore found in inscriptions varying in date from the 1st contury B.C. to the 2nd centary A.D.; that of m in the let and 2nd centuries A.D.; and that of sh not until the second century A.D. The most significant therefore of these letters is sh, and its form shows that the inscription cannot be earlier than the 2nd century A.D. There are two other features which point to a somewhat later period. First, the mark for the long vowel a, in the letters portrayed in Table III, is formed by a short horizontal bar drawn to the right from the head of a consonant, but here the bar always shows a slight curve upwards tendency that became more pronounood in one form of this vowel in the later Gupta alphabet. Secondly, the heads of the letters in this inscription are slightly, yet quite distinctly, wedge-shaped, and this feature also became well developed in the Gupta alphabet. It would therefore appear that this record should be assigned to a time later than the second century, that is, to the 3rd century A.D.. and very possibly to the latter part of it; and this agrees with Mr. V. Smith's conjecture regarding Virasēna's age, mentioned above.
TEXT.
1 Svamisa$ Virasēnass 2 samvatsare 10 3 gishmi3 nam päkshē 4 divasė
1 Mr. Burn read the year rightly, as 18, in his account mentioned above. Mr. V. Smith read it m 113. but the character before that for 10 is clearly rē, the last syllable of rarhateari.
This principle, however, must be applied with caution: Any particular record may easily give the first avail. able instances of types found in it, and so may carry them back to earlier times than had been previously established for them.-J. F.F.]
Read vamisa. There is a dot above the roa, which looks like an asusvara, bat may be only a flaw in the stone.
• Read pakaki. • There is a dot above the se, which looks like an audara, but may be only a flaw in the stone.