Sec. iv] FIRST FINDS OF ANCIENT MANUSCRIPTS 257
left side of the leaf, as in most of the ancient manuscripts recovered from Eastern Turkestan.
Dr. Hoernle's examination (see list) shows that the leaf, numbered 8, contains a text written
in the non-Sanskritic (Proto-Tibetan ?) language, which he was the first to recognize in other
fragmentary manuscripts previously obtained from Eastern Turkestan, in part at least from
Khotan, and that the manuscript to which it belonged was probably that of a Buddhist canonical
text. The writing is attributed by Dr. Hoernle, on palaeographical grounds, to the seventh or
eighth century 1.
The find just described was made at a depth of about 5 feet from the surface and close MS. of
to the rough wooden post fixed in the floor as seen in Fig. 33. It was quickly followed by pâ âmitâ.
a series of other manuscript finds, all consisting of fragments, but varying in extent from
relatively large portions of single leaves to batches of numerous minute pieces. Among these
fragments (numbered D. III. 2-I I) the large majority could at the time be easily recognized, by
their conformity in paper, size, and handwriting as having originally formed part of a Sanskrit
manuscript treating of Buddhist canonical matter. Dr. Hoernle's careful analysis proves that
these fragments belong to the eighteenth and nineteenth chapters of a Mahayana text, apparently
some king of Prajnâ Pâramitâ, written in Gupta characters of the seventh or eighth century on
leaves of very large size, about 18 by 7 inches 2. A leaf of this work, as restored from two pieces
(D. III. 7, 8) is reproduced in Plate CV I I. The obverse contains the conclusion of the eighteenth
chapter (Bhûmiparivarta), marked also by a diagram of homocentric rings, and bears on its left
hand margin the pagination number 132. Dr. Hoernle's notes show that remnants of two other
Buddhist Sanskrit texts, one of them apparently containing the story of the Yaksa chief
Manibhadra (see Plate CI X), are represented among these fragments 3.
The position in which all these manuscript pieces were found, embedded in loose sand at Position
levels varying from 4 to 3 feet above the original flooring, proved beyond all doubt that they °emains.
could have got there only by accident. Their distribution in varying depths and places makes it
probable that they had fallen in from an upper story, while the basement was gradually filling
up with drift-sand. This assumption was borne out by small pieces of animal bones, felt, leather,
and oilcakes (kunjara), which turned up in the same layers. The earlier any manuscript remains
had reached the safe resting-place offered by the sand-covered basement, the more extensive they
might reasonably be expected to be. So I watched with growing eagerness the progress made
by my men on December 23 in clearing the sand nearer down to the original floor. The first
find, made a little to the east of the post already mentioned and about I2 feet above the floor,
was the portion of a document (D. III. 12) written on a sheet of coarse thin paper, about
10 by 8 inches, in cursive Brahmi characters of the eighth century (see Plate CX). In the text
Dr. Hoernle has recognized the same Eastern Iranian language in which a number of other
documents from Dandan-Uiliq are recorded. As in the majority of these, the writing is confined
to one side of the sheet only, since the thin unsized paper, like the modern Khotan paper, would
readily absorb the ink and let it pass through.
As the work proceeded towards the centre of the room, a massive beam of poplar wood Carved
(` Terek', Poftulus alba), 11 inches thick and in almost perfect preservation, was laid bare. Its Posts.
length, close on 19 feet, and its position showed that it had once stretched right across the room, undoubtedly supporting its ceiling. Two well-carved octagonal posts, each 4 feet 5 inches high
and exactly alike, had turned up before (see Figs. 32, 33). Their bell-shaped capitals, 4 inches high, were surmounted by single circular bands accurately reproducing the Amalaka ornament of
1 Comp. Hoernle, Report on C.-A. ant., ii. pp. 12 sq., 2 See Dr. Hoernle's Note i. in list of finds.
18 sqq. S See Notes ii. and iii. in list below.