1901 .] 
Section III. Manuscripts. 
15 
and imported into Eastern Turkestan. All the Pothis which were written 
in Eastern Turkestan itself are written on paper; and it appears probable 
that those written in the Indian variety of the Gupta script, such as No. 1 
of Set I, were written by Natives of India who had settled in Eastern 
Turkestan, while those written in the Central Asian variety, such as No. 2 
of Set I, and No. 1 of Set II, were written by Natives of that country. 
As regards the Pothis, Nos. 3-6 of Set I, and Nos. 2-5 of Set II, I 
am disposed to ascribe them to a somewhat later age. They show the 
Indian upright Gupta script, but written in a curiously angular and 
artificial style : it is the calligraphy of epigraphical records applied to 
book-writing. The curious triangular form of tha and the wavy form of 
medial e first appear in Indian epigraphical writing in the 7th century 
(see Nos. 8 and 14 in Column 6 of Table II), and in Indian literary 
writing towards the end of the 6th century (see Nos. 8 and 14 in Col. 15). 
The peculiar composite forms of the initial i and e (No. 3 in Col. 16, and 
No. 4 in Col. 17) also belong to the 7th century. Accordingly it is to the 
7th centuiy that I am inclined to attribute the Pothis in question. The 
contemporary cursive style, I believe, appears in the fragment of Pothi 
No. 7 of Set I. It is shown in Column 16 of Table II and in figure 3 of 
Plate II. 
To a still later period I would ascribe the Pothi No. 6 of Set II. It 
exhibits a much more developed cursive form of the upright Indian Gupta, 
as established in Central Asia. Unfortunately only a fragment of one 
leaf exists, and that in a rather bad condition. Its letters are shown in 
Column 18 of Table II and fig. 5 of Plate II. Essentially the same cursive 
script, however, is found in the series of Brahmi documents which will be 
described in Group II (p. 32) ; and its letters are shown in Column 19 
and fier. 6 of Plate II. These documents seem to have been found 
together with some Chinese documents dated in the latter half of the 8th 
centuiy; and it is to this period that I would, accordingly, ascribe the 
Pothi No. 6 of Set II. 
Two points which bear on the question of the age of the Pothis 
have already been discussed in the para¬ 
graphs on the “stringhole” and on “paper.” 
The bearings on it of the course of the political history of Central Asia 
remain to be briefly considered. All the Pothis, so far as I know, are 
Buddhistic. Western Turkestan was occupied by the Muhammadans 
in the course of the 8th century. In 751 A.D. Samarkand was con¬ 
quered by the Arabs. From that time Buddhist intercommunication 
between India and Central Asia practically ceased: it had previously 
been gradually declining. U-kong, a Chinese Buddhist monk, visited 
India in 753 A.D., and Prajna, an Indian Buddhist monk, visited China 
Evidence of History. 
