reproduction or occasional repair of such frequently recurring details as lotus-petal borders or
flame-patterns edging vesicas. In Kha. ii. N. 0014 (Plate Xvi) we have actually a mould for a small
appliqué Buddha resembling Kha. 05. That the same method was also used for the reproduction
of curls, locks, hands, and drapery details in larger images worked in the round, is proved by the
moulds Kha. i. 0016, ii. 0076, 0077 ; ii. N. 0013 (Plate xvI).
The two main shrines must once have contained many statues in stucco, life-size or over, for Remains of
we found numerous pieces of fingers and hands (Kha. i. 0029 i. C. 00 6 ; i. W. 0014 stucco
p g ( 9~ 4 4~ ;mages.
ii. 0033-0037 ; ii. N. col), and fragments from heads (Kha. i. oo5 ; ii. 0020, 0021, 0063) or drapery
(Kha. i. E. 0040 ; ii. 0031, 0032). The complete decay of all larger remains of this statuary
must be attributed in the first place to the friable nature of the material, left exposed in all
probability for a long time without an adequate cover of sand, and then to the destructive effect
of the early quarrying operations. The existence here, too, of the practice of gilding is proved
by an abundance of fragments still retaining their gilt (see Kha. i. 16, 29 ; 005, 0023, 0024, 0033,
etc.). It is noteworthy that these fragments seem often to have owed their survival to the sup-
port given by a strong backing fabric. In the case of the small appliqué relievos, preservation
was obviously due mainly to the hardness of the fine plaster of Paris of which they were made ;
for of a general conflagration, which could have hardened small relievos, even if made of mere
friable clay, through a process of accidental firing, as observed at the Ak-terek ruin, no trace
could be found. This is fully confirmed by the analysis furnished in Appendix D by Sir
Arthur Church, who found in the specimen v from Khâdalik ordinary plaster of Paris without
any trace of the effects of a reducing process due to accidental burning such as the pieces of
plaster of Paris found at Kighillik near Ak-sipil exhibit.5
Positive evidence on this point is afforded by the many pieces of painted woodwork which Painted
were found in and near the main shrines. Unfortunately most of these, as already stated, con- woodwork.
sisted of mere parings purposely split off from the quarried posts and other architectural timber.
As a result of this treatment sometimes fragments fitting each other turned up in different places
(see Kha. i. N. of C. 007 in List). Figures of Buddhas and Bodhisattvas appear to have been the
prevalent subjects for this ornamentation (see e. g.. Kha. 005, oo6 ; i. 211. a, 311 (Plate XIv) ; ii. E.
005, 0013, etc.). The style shows closest approach to the painted work of this class which
has survived at Dandân-oilik (cf. Ancient Khotan, ii. Plate LXV, D. I. 04).
The same observation holds good of the numerous painted panels of wood found which, Painted
no doubt, had once served as votive gifts. In view of the number of these panels it is a matter panels.
of special regret that, owing probably to long exposure without a protecting cover of sand or
else to moisture reaching the floor on which they lay, the colours have faded so badly as to make
reproduction impossible. Many of them were painted on both sides. A reference to the detailed
descriptions given in the List from Mr. F. H. Andrews's pen will show how closely the subjects
represented and their pictorial treatment correspond to those in the series of painted panels which
more favourable conditions have fortunately preserved for us at the shrines of Dandân-oilik. Apart
from figures of Buddha and Bodhisattva, represented singly or in groups (e. g. Kha. ooi6 ; i. 18, 3o,
51, 194 ; i. N. of C. col, 004 ; ii. E. 004, 0013), we meet also with those legendary subjects, like the
rat-headed deity (Kha. i. C. 0o15), the horseman and bird ' (Kha. i. E. 0034 ; i. C. 0027), and the
` princess with the cocoons' (Kha. ii. N. 0015), which among the Dandân-oilik finds claimed such
special interest as illustrations of ancient Khotan folklore.6
Here I may conveniently find room also for a brief reference to the remains of artistic wood-
carving which survived among the débris of the main shrines. Apart from balusters such as