12 
Dr. Hoernle —Antiquities from Central Asia. [Extra No. 1, 
the script did not come naturally to the writer but was employed, so to 
speak, artificially as a calligraphic imitation of Indian models. This is 
particularly striking in the case of the letter tha; compare No. 9 in 
Col. 17 with Col. 6. The forms of the letters e and the initial long i 
(No. 4 in Col. 17 and No. 3 in Col. 16) point in the same direction. The 
latter occurs occasionally (in the proportion of 1: 8) as an alternative by 
the side of the more usual Indian form consisting of three ringlets (No. 3, 
in Col. 17), and is made by combining the sign of length of the f-vowel 
with the body of the a-vowel. The letter e is made by a similar com¬ 
bination. These peculiar forms of e and i are also found in the ordinary 
Tibetan script which originated in the middle of the 7th century A.D. 
For these reasons, I believe, the two later varieties to be peculiar scripts 
of Eastern Turkestan. I also believe them to belong to much the same 
period of time, and the variety, shown in Col. 18, to be the cursive script 
of the period, while the variety, shown in Cols. 16 and 17, is an artificial 
imitation of Indian models for calligraphic purposes. The same cursive 
script is also found in a series of Brahmi documents, which are described in 
Group II (p. 32). It is shown in Column 19, and possesses the same 
peculiar form of the initial i-y owel. The form of the letter ma of this 
cursive script (No. 13 in Cols. 18, 19) should be particularly noticed : it 
is quite different from the ordinary form, but closely allied to the Central 
Asian forms, shown in Cols. 20-22. In this connection it may also be 
noticed that the Central Asian Gupta script retains the three-pronged 
form of ya, and the long-limbed form of la (Nos. 14 and 16 in Cols. 16-22)* 
In the Indian Gupta these two forms began to disappear in the 6th 
century, and to be replaced by the two-pronged or boot-shaped form of ya 
and the short-limbed form of la (see Professor Buhler’s Indian Palaeo¬ 
graphy, pp. 45, 48). 
Of the slanting type of Gupta, also, some varieties can be dis¬ 
tinguished. Only one of them, however, (see Cols. 21, 22) is represented 
in the British Collection, in Pothis No. 2 of Set I, and No. 1 of Set II. 
Another, perhaps later, variety, distinguished by its form of the letter 
ma (the second form of No. 13 in Column 22), is found in a few 
fragments published by me in the Journal of the Asiatic Society of 
Bengal, Yol. LXVI (1897), pp. 213 ff., Plate I, fig. x. A variety* 
intermediate between the upright and slanting, is found in the Weber 
MSS-, Parts IY, Y, YIII, published ibidem , Yol. LXII (1893), pp. 22 ff., 
and shown in Column 20. 
The Pothis are written in two different languages: Sanskrit and 
another, which has not, as yet, been identified, 
anguage. Accordingly I have divided them, for the 
purpose of this Report, into two Sets. The first Set comprises seven 
