________________
260
EPIGRAPHIA INDICA.
[Vol. xv.
inscription of the Mauryan type. The characters are thus confirmatory of the early date of the neighbouring sculptures, and prove that, though in the second century vast additions, if not almost entire reconstruction, were effected, the great Chaitya dates originally from perhaps about 200 B.C. It reads
.. Senagopasa Madakatalasa thabho...
"A pillar of General Mudakutala (Mundakuntala ?) " (p. 101). The earliest inscriptions in the Brāhmi alphabet discovered near to Amarāvati are the relic casket inscriptions of the stupa at Bhattipróla in the present Guntur district of the Madras Presidency, in which Amarăvati also is situated. Bühler points out (Indian Paleography, p. 8) that some signs of the Bhattiprolu alphabet, called by him Dråvidi, such as d, dh, bh, ch, j, sh, ?, differ from those of the edicts of Asoka. But all the signs of the ancient Brāhmi epigraphs from Amarāvati agree with the southern variety of the Asoka alphabet. The g with an angle at the top and ch, d and p are archaic in form (i.e. similar to those in the Asoka edicts).
(2) The alphabet of inscriptions 25, 29, 33, 36, 40, 42, 43, 44, of which the characteristic features are :-(a) the retention of the archaic d; (b) the equalisation of the upper verticals except in No. 43, which is somewhat older; (c) the absence of the curves at the end of a, k,, which constitute a very prominent featare of the inscriptions of Western India of the time of the Kshatrapas and later Andhras; (d) the curvilinear medial 1. These epigrapha may therefore be assigned to the first century B.C., or A.D. This variety is distinguished from the Andhra script of Nānāghāt by the presence of what is called serif, a thickening or a very short stroke at the upper end of the verticals. The alphabets of Nos. 37, 38 and 43 are withont serif and consequently older.
(3) The bulk of the Amaravati inscriptions classified by Bühler with the cave inscriptions of the Western Dekkhan and Konkan belonging to the second century A.D.
(4) The highly ornamental alphabet of inscriptions 24, 27, and 50, resembling those found in the Jaggayyapētas inscriptions of the time of the Ikshyāku king Sirivira Purisadata and provisionally assigned by Bühler to the third century A.D.
The Prākṣit used in these and other inscriptions of Amarāvati betrays close affinity with the Paisachi Prakrit of the grammarians. Thus we have k for g in Nakaya (No. 58); ch for j in pavachitaya (No. 58, and Lüders' List, No. 1270); t for d in vētikā (Nos. 29, 46, and Lüders' List, Nos. 1216 and 1269, and Vararuchi, X. 3); dental - for cerebral → in unisa and umnisa, the former occurring eight times and the latter twice in the published Amarāvati inscriptions, and in samanasa (No. 11), and tini (No. 19); but tini in No. 33 and apaņo in No. 27 (Vararuchi, X. 4). Bhāriya is invariably used for bhāryā in these records (Vararachi, X. 8). According to Kashmirian tradition Gaņādhya, who composed the Brihatkathi in Paisachi Prākpit, lived at the court of the Andhra kings. This tradition indica belief that the Paisāchi Prakṣit was cultivated in the Andhra kingdom under the Andhrabhritya kings. The language of the Amaravati inscriptions seems to support this tradition.
These inscriptions furnish as with no historical, that is to say, dynastie information with the exception perhaps of the clan-name Pākoțaka (No. 8) and the personal name Vākāțaka (No. 27). The identification of the Pākoțakas with the Vākāțakas (p softened to b, which was not always distinguished from v) is obvious. According to Mr. Vincent A. Smith the
1 Corrected to Mudukutala (Mridukuntala) (Lüders' List, 1266). · The Guntur district, which once formed part of the Kistas district, bas only recently been separated from
The site of Jaggayyapēta lies 30 miles to the north-west of Amaravati.
