32 
Dr. Hoernle —Antiquities from Central Asia. [Extra No. 1, 
Script. 
the prevalent cursive style of writing among a certain class of people in 
Eastern Turkestan. 
The material on which the documents are written is exactly the 
same kind of thin, coarse, whitish, water-lined 
aper an n ing. p a p er as that 0 f the Chinese documents. It 
is also very similar to that of certain Pothis, especially Nos. 5 and 6 of 
Set II. The water-lines are, longitudinally, 13 or 14 to an inch ; 
transverse water-lines are mostly absent; only on three (Nos. 1, 3, 9) 
of the complete sheets any distinct traces of them can be seen, f to 1" 
apart. The size of the mould must have been about 16 x 12", the 
greatest dimensions of sheets, either way, actually measured being 15 
and 11^ inches. Professor Wiesner’s tests have revealed no trace of 
loading with starch. The writing always runs parallel to the shorter 
side and therefore, across the close-spaced water-lines. It is also, as a 
rule, confined to one side of the paper ; only a few fragments (Nos. 28, 
29, 48, 50) have a few letters on the reverse. A brush and China ink 
seems to have been used in writing. 
The script of the documents is a species of cursive Brahmi. Its 
affinities and date have been discussed in con¬ 
nection with the Pothis, and are illustrated 
by Table II, where the letters of the script are shown in column 19. 
Its approximate date is probably the 8th century A.D. Two varieties 
of hand can be distinguished in the document, one with rounded, the 
other with angular forms. A specimen of the latter is shown on 
plate VII., fig. 2. The former may be seen in Plates II, fig. 6, VI, and 
VII, fig. 1. In the Journal, Asiatic Society of Bengal, Vol. LXVI 
(1897), Plates xxi-xxiv, I published a Table of the alphabet of the script. 
Further investigation has shown that it is not correct in several 
particulars : the signs for ta and ma had been wrongly identified as bha 
and rja, and the existence of some special signs in the form of a hook or 
a curve had not been recognized : there are also some minor errors ; e.g. 
in identifying some forms of the vowels o and i which nearly resemble 
each other. Accordingly a revised alphabet is now given in Table III, 
which also now shows the full system of numeral figures. 
The language of the documents has not yet been identified ; but one 
point seems to be certain, that it is different 
from the unknown language of the Pothis of 
Set II; e.g., none of the characteristic conjunct consonants of the latter 
occur in the Brahmi documents. Only a few of the words or phrases 
have, as yet, been determined, but these seem to prove clearly that the 
language of the documents is an Indo-Iranian dialect, having affinities 
both with Persian and the Indian Vernaculars, in addition to pecularities 
Language. 
