________________
146 EPIGRAPHIA INDICA
(VOL. XXIX except in line 29 where the lower part has a double curve instead of the usual one. Medial a has no less than three different forms. In addition to the usual sign resembling a danda put at the right of consonants and generally joined above with their top mātrā (cf. °Kailāsa in line 1), there are some cases where the stroke is a little curved towards the right and reaches only down to about half the length of the consonants (cf. krida in line 2), while in others it comes only a little down and then goes up, leaving a sharp curve below and ending in a stroke curved towards the right above the head of the consonants (cf. pank-āvila in line 1). Medial u and ů are usually of the ordinary types; but the forms of chyu (cf. chyutairo in line 2), ru (cf. dhvänt-Oruo in line 23). rū (cf. Pārüd hair in line 11) and others are interesting. Bhü has been written in two different ways. The ordinary sign for medial û has been employed in some cases (cf. abhüt in line 53); but often the ở sign is formed by joining the sign for u with a danda put at the right of bh; cf. lines 9 (bhūto), 12(bhüt), 18 ( bhūtā), 20(adbhūto for "ābhūto), 24 (Whümishu) and 25 (bhüri). Medial é is formed by lengthening the left end of the top mātrā of the consonant in a curve forming almost a loop at the end; but it is yet shorter than the developed sign of the matrā reaching the bottom of the line. The akshara 6 has not been distinguished from v. The consonant jh, which is of rather rare occurrence, is found in the conjunct ijhi in line 43. The final consonants (lines 5, 12, 54, 57, 59), n (lines 51, 53) and m (lines 4, 7, 8, 23, 25) occur many times in the inscription.
The language of the inscription is Sanskrit. It is written partly in prose and partly in verse. The prose portion exhibits the quality called ojas in a considerable degree in imitation of the style of masters like Bhatta Bana (cf. lines 33-46); but the versification is rather weak, there being several cases of metrical errors (cf. verses 8, 9, 19, 26, 27). The formal part of the charter from the beginning down to Vanamālavarmmadēvah kućali in line 48 is the same as in another record, viz., the Tezpur plates, of the king who issued the charter under discussion, except that two passages of considerable length have been omitted in the present inscription, perhaps inadvertently. As some mistakes are common to both the records (of. 'ādyasya for a word like yuktasya in verse 9 and vijagrihuḥ for jagrihuḥ in verse 21), it seems that they were prepared from the same draft which was faulty. There are altogether thirty verses (actually 29 verses and only the first half of another stanza). Of the five verses in the latter part of the document, two are the common imprecatory stanzas, while the remaining three describing the donee are written in a style which is definitely worse than that of the stanzes in the formal part of the record. Weakness in the composition even of the formal part is exhibited by the repeated use of some expressions ; cf. adhinäthatua in lines 7-8; nihdësha in lines 16-17; abesha in lines 17, 20, 26; anghri in lines 9, 14, 30: vairi-vira in lines 10, 12, anugā in lines 13, 19; vrata in lines 30, 36. The number of orthographical errors, although not many, is not inconsiderable. The conjunct ngh has been written us nih. The word pushkariņi is found in the form preshkirini. Anusvāra has been used instead of the final m in the imprecatory verses quoted at the end of the record. As regards other orthograpical features, the consonants g, ņ and dh have been reduplicated after r; but m and v have been only occasionally reduplicated under the same condition, while th and y have not been subject to reduplication. T has been reduplicated when followed by .
The inscription is not dated even in the regnal reckoning of the king who was responsible for its issue, although that was expected considering its age and locality. There is, however, no difficulty in determining the approximate date of the charter. As a record of the father of its issuer, vis., the Tezpur stone inscription of Harjaravarman, is dated in the Gupta yenr 610 corresponding to 829 A.C., the inscription under notice has to be assigned to a date about the middle of the ninth century A.C.
1 See JA8B, Vol. IX, 1840, pp. 766-67 ; Kamarüpa-sasan-avali, pp. 58-68. •Kamarido-daan avall, p. 187.
