1897.] 
253 
Central Asian Manuscripts. 
The following is a summary of Set II:— 
a, 
sheets 2, 
total 
2 
b, 
single leaf 1, 
double-leaves 
3, 
>» 
7 
j> >> 
>5 
2, 
jj 
5 
d , 
>5 >> 
55 
3) 
6 
j) » 2, 
V 
0, 
2 
Total 22 
Set III. Plate XXIII shows two leaves of this set. There are 
altogether 12 such single leaves. They measure about 6J x 3f inches, 
and have 6 or 7 lines to the page. The writing on them is much 
interspersed with what look like Bralimi ligatures, in the Tibetan type 
of characters. This seems to render it probable that the rest is also 
written in Bralimi characters of a very cursive type; but I have had 
no time to study it more closely. The leaves show no holes, and they 
do not appear to have ever been fastened together, though it can hardly 
be doubted that they form a connected series. 
Set IV. Plates XXIV and XXY show two double-leaves of this 
set. It consists of a thick manuscript of small sized double-leaves, of 
which some 3 or 4 have split iuto single leaves. Accordingly there 
should be 112 leaves, but actually there are only 111 leaves, and these 
measure about 5| x 3f inches each. The lower corners of the leaves 
are damaged. Each double-leaf, when folded up into two single leaves, 
makes up a so-called ‘ form,’ and these ‘ forms ’ are bound together 
into a * book ’ by means of a metal nail which is passed through the 
whole of the ‘ forms ’ of leaves near their left-hand margin. The 
‘ forms ’ are secured from falling off the nail, by a metal disk screwed 
into one of its ends and a metal knob, into the other. The ‘ book ’ 
begins and ends with a couple of blank ‘ forms,’ but whether this 
indicates that the manuscript is complete, I cannot say, though it seems 
probable. There are six or seven lines on each page, and these lines are 
distinctly partitioned off into four columns. The number of letters in 
a columnar line varies; it is usually six; but I have noticed them from 
four to seven. In this manuscript, too, ligatures of the Tibetan type 
occur on nearly every page, which would suggest a Brahmi cursive 
character for the rest of the writing. Whether the latter is the same 
as, or similar to, that occurring in Set III needs investigation. I have 
had no time for closer examination. 
Set Y. Plate XXYI shows three leaves of this set. It is a manus¬ 
cript, very similar in every respect to the preceding one. All its leaves 
are single, about 100; their exact number is uncertain, as a few of 
J. i, 33 
