254 R. Hoernle —Three further Collections of [No. 4. 
the leaves are broken in fragments, the paper being very brittle. They 
measure about 5|- x 2-| inches. They are also made up into a ‘ book, 5 
by a metal nail passed through the left-hand side of the leaves. There 
are two blank leaves at the end of the book, and the leaf preceding 
them is inscribed on one side only. There appears to have been also 
a blank leaf at the beginning of the book, but it is now broken into 
fragments. All this would seem to indicate that the manuscript is 
complete; but not being able to read it, I cannot say so for certain. 
There are five lines on each page, and these are partitioned off into four 
columns. The letters in each columnar line number eight. 56 Liga¬ 
tures of the Tibetan type occur much less frequently than in the 
preceding manuscript (see obverse of leaf I, line 3); nevertheless the 
writing may turn out to be a species of very cursive Brahnri. I have 
had no time for any closer examination. 
Set VI. This is a small manuscript of 8 leaves, measuring 5 x 2§ 
inches. It is in a very bad state of preservation : nearly the whole of 
its writing has become obliterated, and the leaves are very baked and 
brittle. From the little that is legible, it is certain that this manuscript 
was written in exactly the same characters as the preceding one (Set V), 
with the same sporadic interspersion of Tibetan-like ligatures. As the 
leaves show no hole, they do not appear to have ever been strung 
together. In this respect this manuscript is like that of Set III. 
In conclusion I would add a few remarks concerning the probable 
age of these manuscripts. They are not offered as embodying final 
results ; they are only thoughts which have forced themselves on my 
mind in the course of my investigations, and they are intended as 
suggestions to stimulate further researches by others. For my part, 
I am disposed to believe that they will eventually be found to err on 
the side of moderation rather than excess. 
For the purpose of an enquiry into their age, these manuscripts 
must clearly be divided into two distinct classes. First, there are those 
found near Kuchar, and dug out from the ruins of the ancient vihara. 
These are written in Brahmi characters, either of the Northern Indian 
or the Central Asian type, and are composed either in Sanskrit or in 
Turki. They are also written on palm-leaf, or birch-bark, or paper. 
To the second class belong those found in the sands, in the neighbourhood 
of Khotan. These are written in Chinese or Uighur or some other un¬ 
known alphabet and language; they are also inscribed on paper of 
(apparently) a quite different kind. I omit for the present the Godfrey 
MSS., because it is not certain, whether they were also found in that 
58 This, if the language were Sanskrit, would point to a work in (jlokas. 
