________________
300
EPIGRAPHIA INDICA,
[VOL. XXVI.
of other evidence, more or less indefinite, of palæography, language and the mode of dating adopted in these records and the stratification of the level at which their coins and seals were found. As regards palæography the following peculiarities have been noticed :(1) The medial short i is shown by a small curve on the top of a letter as in the Kushāņa
inscriptions. In the Gupta records this curve is brought down much lower on the
left of the letter. (2) The medial é is generally indicated by a short horizontal stroke to the left as in the
Kushāna inscriptions. In the Gupta records the mātrā is usually placed on the
top. (3) The medial o is shown by a concave curve as in the Kushāna inscriptions. In the Gupta
records, the vowel is shown by two mātrās turned in opposite directions. (4) The right verticals of g, t and $ are of the same length as the left ones. This is a pecu
liarity of the Kushāna records. In the Gupta inscriptions the right verticals of these
letters are much elongated. (5) Gh, p and y are rectangular as in the Kushāņa period. In the Gupta period these letters
become round and cursive. (6) N has a curved base as in the Kushāņa records, but no loop except in the late inscrip
tion of Vaisravana's reign. This letter became looped in the Gupta period. (7) M and h have the later, so-called eastern, forms known from Gupta records, though on
the coins of these kings m has the archaic shape of the Kushāņa period. (8) S appears both in looped and unlooped forms. The unlooped form of it was generally
current in the Kushāna period and the looped one in the Gupta period. As regards language, these records bear affinity to the Kushāņa rather than to the Gupta records. They are written in a mixed dialect which was current in the Kushāņa age. Such Prakrit expressions as ētāya puruvaya for ētasyām pūrvāyām or satime for satatamē are used side by side with correct Sanskrit forms. The language of the Gupta records is generally good Sanskrit.
The method of dating adopted in these records is slightly different from that of the Kushana records from Mathură. The dates are recorded in years and seasons, but instead of mentioning the number of months within the seasons as in the Kushāņa records from Mathurā, they state the number of fortnights like the Sātavāhana records from the Deccan. In North India such season dates were current during the Kushāņa perin., but they seem to have gone out of use in the Gupta age as there is not a single certain date of that age recorded in seasons.
1 Lüders' List of Brahmi Inscriptions, Nos. 1122-26.
* Mr. R. D. Banerji drew attention to the date of a Mathuri Brahmi inscription which he took as indicating a compromise between solar month dates and season dates. He read the date as Vijaya-rajya sain 100 10 3 Kartika Héma*]nta mä[sa 3] disva]sa 20 (above, Vol. II, p. 210, and Ind. Ant., Vol. XXXVII, p. 46). But the facsimile shows that the reading Hemanta is extremely uncertain. Incidentally it may be pointed out that the date of the record appears to be 110, not 113. Mr. A. Ghosh mentions three season dates of the Gupta period, but they too are not certain. Of these the first one, viz., the date of the Kosam inscription of Bhimavarman (year 139) refers to the same era as those of the other Kosam inscriptions of this dynasty, as shown above. The date of the Mathurā inscription of Chandragupta II is not a season date as shown by Dr. D. R. Bhandarkar. As there was an intercalary month in the Gupta year 61, Dr. Bhandarkar's suggestion that the lost letters of the date were Ashadha-mase is quite plausible. As for the date of the Mathura inscription No. XXXVIII, (above, Vol. II, p. 210), its attribution to the Gupta period rests only on palæographic evidence which is not quite certain. See also J. R. A. 8. for 1903, p. 11. In South India the custom of using season dates lingered much longer; for we have some records of the Vakatakas and the Vishnukundins of the fifth and sixth centuries A. D., which have season datee. See above, Vol. II, p. 262; Vol. IV, p. 197; Vol. XVII, p. 339, eto.
