________________
SIYADONI INSCRIPTION.
163
to line 39, they are regularly and beautifully formed and skilfully engraved. The execution of lines 40-46 is somewhat inferior to the rest, and the difference in appearance is rendered more marked by the imperfect state of preservation of these concluding lines. The language of the inscription must be described as Sanskrit. Unless there was a date in any part of the concluding lines which is now illegible, the inscription itself is not dated; but it contains ten dates, some of which are historically important, while one is sufficiently particularized to ascertain from it the era employed, by calcu. lation, as will be shown below.
The inscription consists of two parts. The first and by far the larger portion extends to about the middle of line 39; it is in prose, and records a large number of donations made at different times, from the (Vikrama year 960 down to the (Vikrama) year 1025, and nearly every one of them by private individuals, in favour of various Brahmanical deities, at Siyadoni. The second part, which is almost entirely in verse, comprises the remaining portion up to the end, and records the erection of a temple of Murâri (Vishņu).
THE FIRST PART. To treat fully of the language of the first part, would require almost a separate treatise. The author or authors, though intending to write Sanskrit, bad a very meagre knowledge of the grammar of that language; they were evidently influenced by, and have freely employed words, phrases, and constructions of, their vernacular.
As regards orthography, b has throughout been denoted by the sign for v; and the dental sibilant has often been employed for the palatal.' The sign of the jihvamúliya occurs twenty-two times, almost exclusively in the phrases ufro and utfy. The sign of the upadhmaniya has been correctly employed six times (e.g., in the four, line 5, and o fuffeggf , line 13); but it has also been wrongly inserted three times (in tuzoufto, line 18, raegut, line 28, and farsdiereguft, line 29), and probably erroneously omitted twice (in fufe ufq9, line 25, and TATUETT, line 35). Of individual words. the numeral fa has throughout been spelt in JHTT, lines 24, 29, 30, and T., lines 25 and 27); FATET throughout HTC (e.9., in lines 3, 6, 8, etc.); ista throughout Afera (e.g., in lines 3, 6, 20, etc.); and similarly we have to fat in line 38, for HTTGT In line 8, we twice have wrea for wif ; throughout, frequently, Taraftar, apparently for watafa (e.g., in lines 7, 8, etc.); and similarly a short rowel has been employed instead of a long one, and vice versá, occasionally in other words. For afy we have at twice in line 17, and perhaps also in some of the compound numerals; for alaf (line 15), arrafera in lines 25 and 26; for a 177, aerct in line 12. Through the influence of the vernacular, we have a fun throughout for afur; at in line 24 for fura (lines 13 and 25); faftur in line 39 for (line 37); 941 and कलपाल in lines 9 and 19, for कल्पपाल ; etc.
The rules of samdhi have been persistently neglected; and as a specimen of an extraordinary samdhi I may point out • सुतारभि[:], for • सुता एभिः, i.e., सुता: एभिः , in line 35.
* I consider it sufficient to state this and some of the following points once for all, and sball pot consider it necessary to correct every error of this sort in the transcript of the text.
' Compare the common सन्मान for संमान.
