________________
162
EPIGRAPHIA INDICA.
(Vol. IX.
dagger and shield with a linga in its socket, exactly of the shape in which the Lingayats wear them, are engraved. There is a postscript to this inscription in which it is stated that the land was given by Dharana-mahadevi, who was probably the widow of Sômêsvara, as will appear further on. There can be no doubt that Narayanpål is the Narayanapura of the inscription. A temple of Narayana is still standing there. The image of Vishņa, about 2' high, canopied by a hooded snake, is exquisitely executed.
II.-Bårsûr inscription of Ganga-mahadevi, wife of 8ômesvaradeva. This inscription is now in the Nagpur Museum, and, as stated above, it has already been published. It is a slab 9'2' long, 14' broad and 3" thick, broken into two pieces, the bigger one measuring 67' and the smaller one 2' 11". It is inscribed on three sides. The inscribed portion of each flat side is about 4}', thos leaving half of the pillar buried underground. As the whole of the inscription could not be completed within the allotted space, the remaining portion has been inscribed on the third side, on which the writing runs to the length of 31". The stone is stated to have been brought from Kowtah near Sironcha, but the Tahsildar of Sironcha informs me that it was never seat from that place.
The stone is indisputably from Bargur. Happily Col. Glasfurd bas given a facsimile in his report on the Dependency of Bastar. Speaking of the Bârsor temples he says::-"In front of this temple I found & slab with an ancient Sanskrit and Telugu inscription on both sides; part of it had been broken off and was nowhere to be found. After offering & reward and causing search to be made I had the satisfaction of obtaining it. As the Telugu is of an antiquated character, I regret to say I have not succeeded in obtaining an accurate translation of the inscription. A facsimile is appended. From what I can ascertain it would appear that the temple of Mahadeva where the slab was found was built by a Raja Somêsvaradêve, & Nágayumsi Kshatriya, in the year 1130."
The inscription is in the Telaga character, and the language is also Telugu prose, the birudávali or titles of the king being in Sanskrit and corresponding with those in the Narayanpål Sanskrit inscription. It records that Ganga-mahadevi, the chief queen of Somèsvaradeva gave a village named Keramapuka or Keramarks to two temples of Siva (both of which she had built) on Sunday, the 12th tithi of the bright fortnight of Phâlguna in the Saka year 1130. The two temples referred to here still exist, having one common mandapa, and from local enquiry it appears that it was from this place that Col. Glasfurd removed the slab. Although the names of the temples Virasômesvars and Gangadhardsvars given after the royal couple as recorded in the grant, are forgotten, a tank still remains which is called Gangåsågar and retains the memory of the charitable queen Ganga-mabadêvi. If the Sômēsvara of this inscription is identical with that of Narayanpål, there has apparently been a mistake in engraving the date which should be 1030 and not 1130, and that is perhaps the reason why the week day does not correspond with the tithi given there, vis., the 12th of the bright fortnight of Phålguna, on a Sunday. According to Mr. Dikshit's calculations, Phálguns Sukla 12 of Saka-Samvat 1130 ended on Wednesday. So it was concluded that the year meant was Saka 1131 expired, in which year the tithi given in the inscription fell on a Sanday. But on calculating the week day for the same tithi in Saka 1030 expired I find that
Above, Vol. III. p. 314.
A similar error seems to have boon committed in relegating the Buddhist stone inscription of Bhavadeva (republished in J. R. A. S. 1905, p. 617, by Dr. Kielhorn) to Ratanpur, whereas from my enquiry in situ I found that the inscription was really brought from Bhandaka, and this is conOrmed by General Cunningham, Reporte, Vol. IX, p. 127.
* Report on the Dependency of Bastar, 1862, p. 62.