________________
128
EPIGRAPHIA INDICA.
(VOL. XXI.
5 Déunāyaka bhāktā[rēl] Ambada6 purē vāstavy? Padumaņasēthi7 kratēs kirttanē Bhāidēva-putra-Matagala8 msishe (?) [tõdraņam i(de)am (?) -- tā - -
B.-The Bārsi-Tākli Inscription of the reign of Hēmādridēva--Saka 1098.
The inscription, which is edited here from an impression kindly supplied to me by the Government Epigraphist for India, is at the temple of Bhavani in the village of Bārsi Tākli, twelve miles south-east of Akola in Berār. Its existence was noticed by Mr. H. Cousens, who has also described the temple in the Progress Report of the Archeological Survey of Western India for 1902. "It (i.e., the temple) consists of a shrine and a mandapa or hall, both being freely decorated upon the exterior with bands of mouldings and figures. The mandapa is curiously arranged with regard to the shrine, being attached as it were sideways to it, the open side of the mandapa with its entrance being on one side or at right angles to the doorway of the shrine. The plan of the mandapa is rectangular, while that of the shrine is star-shaped. Four decorated pillars support the central ceiling of the hall. The principal figures around the outside of the temple, excepting Ganapati, are the females Mahākäli and Mahishasuramardini, occupying important positions. The ceiling is particularly well decorated, the marginal panels being very much like those of temple No. 1 at Balsane in Khandesh ".
The inscription is engraved on the wall of the mandapa facing the entrance. As in the case of the Rāmtēk Lakshmana Temple inscription, the surface of the wall was first rendered smooth with lime plaster before the inscription was incised in it. When Mr. Cousens saw the temple it had already been much damaged. When I visited the place in December 1931, I found it in a still more deteriorated condition as the lime plaster was gradually scaling off from the wall. The whole record covers a space of about 3' 3" broad by 11" high. The average size of letters is about ". The characters are Nāgari. The language is Sanskrit and excepting the introductory Om Svasti and the particulars of the date the whole record is in verse. Originally it must have consisted at least of twenty lines; but the last six lines of the record are almost completely gone. Of the first fourteen, again, nearly half the portion on the right hand siue is destroyed owing to the peeling off of the lime plaster and though a few words and letters here and there can be read, they do not give much coherent sense. The verses are not numbered, but single and double dandas are, in some places, used to denote punctuation. The orthography calls for few remarks ; s is occasionally employed for $, e.g., in Vaisākha in l. 1; the anusvāra is incorrectly used in addition to the following nasal in many places, e.g., Nimilaṁn in 1. 3, apatanntam in l. 4, etc. The consonant t following ris occasionally doubled as in kirtlau and varttamāno in 1.5. The final dental nasal is changed to an anusvāra in frimän in l. 8, dāsvār (1.9).
As stated by Mr. Cousens, the temple is at present believed to be dedicated to Bhavāni. I found, however, no image of the deity in the garbhagriha. There were, instead, three round stones (the central one being somewhat more elongated than the rest) covered with vermilion on a finely carved pedestal. There was also a stone figure of Nandin (Siva's bull) in the mandapa facing the deity which inay easily have been brought from elsewhere. The crude symbols of divinity ure in striking contrast to the rich carving on the four central pillars and the ceiling of the
1 Read Dēunāyake wholturi, • Read krila
* Read mira-rāslarva. • The last lige is very indistinct.
