गुप्त अभिलेख संग्रह (जे. एफ. फ्लीट - गुप्त सम्राटों एवं परवर्ती नरेशों के ऐतिहासिक पाषाण व ताम्रलेख)

22 CORPUS INSCRIPTIONUM INDICARUM. VOL. III.
Udayagiri¹ is a well-known hill, with a small village of the same name on the eastern
side of it, about two miles to the north-west of Bhêlsâ,² the chief town of the Bhêlsâ
Tahsîl or Sub-Division of the Îsâgaḍh³ District in the dominions of Scindia (properly
Śindê) in Central India. On the eastern side of the hill, a little to the south of the village,
and almost on the level of the ground, there is a cave-temple, which, from its containing
this inscription, General Cunningham has named "the Chandragupta Cave."⁴ The in-
scription is on the upper part of a smoothed and countersunk panel, about 2' 4½" broad by
1' 6" high, over two figures,—one of the four-armed god Vishṇu, attended by his two
wives ; and one of a twelve-armed goddess, who must be some form of Lakshmî, the con-
sort of Vishṇu, rather than, as suggested by General Cunningham, Mahishâsurî, i.e.
Durgâ, the consort of Śiva,—which are sculptured on the face of the rock, outside the cave
and a few feet to the north of the entrance to it.
The writing, which covers a space of about 2' 3½" broad by 4¾" high, is in a state of
fairly good preservation ; the surface of the rock has peeled off in some places ; but no
letters are entirely destroyed, except the g of chandragupta in line 1, and, in line 2, the
first akshara of the name of the Mahârâja whose gift is recorded.—The average size of
the letters is about ⅝". The characters belong to the southern class of alphabets, and
give another specimen of the 'box-headed' variety, peculiar to Central India, which I have
commented on at page 18 f. above ; but in this inscription, again, there are no instances in
which enough remains of the square centre of the tops of the letters to shew distinctly in
the lithograph. In âshâḍha, line 1, they include an adaptation from the northern alpha-
bets ; viz. the separate sign for the lingual ḍh, which, in the early southern alphabets, was
represented, together with its unaspirated form ḍ, by the same character as the dental d.
They also include, in line 1, forms of the numerical symbols for 2 and 80.—The lan-
guage is Sanskrit ; and the inscription is in prose.—In respect of orthography, the
only point that calls for notice, is the doubling of dh in conjunction with a following y, in
anuddhyâta, line 1.
The inscription refers itself to the reign of the Early Gupta king Chandra-
gupta II. It is dated, partly in numerical symbols and partly in words, in the year⁵
¹ The 'Udaygiri and Udegiri' of maps, &c. Indian Atlas, Sheet No. 53. Lat. 23° 32' N.; Long.
77° 50' E.
² The 'Bhilsa and Bhelsa' of maps, &c.
³ The 'Isagarh' of maps.
⁴ Archæol. Surv. Ind. Vol. X. p. 49 ff. and Plates xvi. and xvii.
⁵ Whether "current" or "expired," is not specifically stated. But the locative case is used ;
and, there being no word signifying "expired" in apposition with it, the natural sense of it is "in
the year eighty-two," i. e. "while the year eighty-two is current."—It is often asserted that the ancient
Hindus always expressed their dates in expired years. And, according to this, we should understand
atîtê, 'having gone by, having passed by,' or any similar word, in apposition with saṁvatsarê
in the present passage, and translate accordingly.—No doubt, in making calculations, the Hindus
worked, as Europeans must work, with the number of expired years as a basis. But this is quite
a different question from that of their method of expressing the dates ; as, also, is the question
whether they did not sometimes by mistake quote years as current which were really expired, and
the reverse, and even occasionally quote, as current, or even as expired, years that were still to come.
And, as a very clear and pointed instance of the record of a date in current years, we have the Gwâlior
Sâsbahû temple inscription of Mahîpâla, dated Vikrama-Saṁvat 1150, which gives (Ind. Ant. Vol.
XV. p. 41 ; vv. 107, 108 ; l. 40) first, in words, the number of years expired, and then, partially