244 
R. Hoernlc —Three further Collections of 
[No. 4, 
tlie Brahma Sauatkumara. The name of the chapter would seem to be 
Nagaropama Ardma or ‘ the town-like park.’ On the obverse of Leaf I 
I have restored what can be concluded with much probability to be the 
missing portions. This will give an idea of the original state of the page. 
Set I, b. This set consists of 15 leaves. As a rule there are 9 
lines on a page, only exceptionally 10, as on fl. 23a. The manuscript 
is incomplete, both as regards the number and the size of the leaves. 
Its beginning and end are missing; but, so far as I can see from Dr. 
von Oldenburg’s paper in the Journal of the Imperial Russian Archaeological 
Society , no portion of it appears to have gone to St. Petersburg. All 
the leaves are mutilated at their right-hand side, and the only indi¬ 
cation of their original length lies in the well-known fact, that Central 
Asian manuscripts have their string-hole on the left side of the leaf, 
at the distance of about a quarter of the length of the full page. Hence 
it may be concluded with some probability, that about one-quarter of 
each leaf is missing. As the existing length is about 4J inches, this 
gives the full length as probably about 6 inches. 36 The breadth of 
the leaves is about 2 inches. The material is a very soft kind of 
paper of a darkish colour; it is in a very rotten and broken state. 
The writing is very slovenly done. Small and big letters frequently 
alternate without any apparent reason ; and the lines are not kept 
properly straight and apart, so that their letters occasionally run into 
one another. Also errors occur not unfrequently, syllables or sounds 
being occasionally omitted ; thus fl. 22a 2 pamca for pamcame , fl. 22a 4 
tryodagamam for trayodagamam; fl. 23a 3 rdstopadrave for rastropadrave , 
etc. All these blemishes aggravate the difficulty of reading the manu¬ 
script, and, I hope, will be accepted in extenuation of the imperfect 
state of the transliteration, given by me below. 
The characters used in this manuscript distinctly belong to the 
Northern Indian class of Brahmi, of the early Gupta period. They are 
of a rather archaic type, as I shall presently show in some detail. 
It will be seen from the excellent comparative tables, published by 
38 Professor Biihler in the Vienna Oriental Journal , Vol. VII, p. 261, points out 
that “ numerous copperplate grants with one string-hole on the left ” exist in India, 
and infers from it that manuscripts with one string-hole on the left “ were once not 
unknown in India.” There is every probability that this inference is correct. For 
as the material (birch-bark or palm-leaf) shows, some of the Central Asian manus¬ 
cripts, ( e.g ., the Bower MS.) must have been imported from India (see p. 258). In 
fact, in the case of such exported Indian manuscripts, the peculiar position of the 
string-hole is an additional proof of their great age. For no Indian manuscript, 
found in India itself, shows that position; they either show one hole in the middle, 
or one on either side. Even the Horiuzi MS., exported from India to Japan early in 
the 6th cent. A.D., already shows the double hole. 
