654 THE ANCIENT CEMETERIES OF ASTANA [Chap. XIX
The tomb next examined, iii. 4, situated in the north-eastern corner of the enclosure, offered compensation for the scantiness of the sepulchral deposits found in the last one. As seen in Pl. 33, its plan was unusually elaborate, a small outer room giving access first to another of cruciform shape and thence to the tomb chamber, which was provided with a kind of alcove raised 14 feet above the general floor level. The contents of the tomb had fared badly at the hands of those who had first opened and plundered it, but nevertheless proved of distinct interest. The headless body of its occupant was found in the approach trench close to the entrance where the coffin had evidently been dragged to be searched in daylight. The head was subsequently discovered within the tomb. The body was wound in a miscellaneous assortment of rags, ranging from pieces of silk, plain or figured in monochrome (iii. 4. o6), to coarse cotton and thin leather. On the sole of one foot was found a torn piece of paper with Chinese writing, which had been used for the same purpose amidst other rags. My explanation of the mass of torn fabrics of all kinds that we had found a year before mixed up with human remains in the grave-pits of the Lou-lan cemetery L.C.'a could not have been more satisfactorily confirmed.
Within the high layer of drift-sand which had invaded the anterooms we came upon scattered fragments of stucco figures similar to those of Ast. iii. 3, but all with a few exceptions badly broken. Among these were the large and well-modelled clay figure of a Bactrian camel, iii. 4.015 (Pl. XCVIII), painted a pinkish white, with its head raised and thrown back in life-like movement ; the carefully executed stucco figure of a lady, iii. 4. 064 (Pl. XCIX. A), dressed in a costume which, like her coiffure, closely resembles that of the donatrix figures of two early Chien-fo-tung paintings, liii. ooi ; Ch. xlvii. 001.8 The pair of figures in stucco and wood, iii. 4. 072. a, b (Pl. CII), also have carefully modelled heads, with lobed black caps such as appear on the heads of donors of the same Ch`ienfo-tung paintings and on those of various persons represented in others depicting scenes of Buddha's life ; the clothes in which these two figures were dressed have been lost but for a small remnant.9 We have in iii. 4. 073 (Pl. CII) a similar figure of a man, remarkable for the naturalistic treatment of the excellently modelled head. Very numerous are the fragments of painted wood, comprising miniature balustrades, bridges, arches, &c., iii. 4. 027, 035-60, &c. (Pl. XCIV) ; they appear to have belonged to an architectural model, possibly intended to represent a celestial mansion such as figures in many of the Chien-fo-tung paintings. The fragments of a wooden pedestal or stand, iii. 4. 021,028-33,062 (Pl. C), are of interest as its shape resembles that of stands in the Shôsôin, and the design of its decoration with drifting palmettes occurs there also. To a miniature model of a similar stand belonged the pieces iii. 4. 052-4 (Pl. XCIV). A similar stand was found intact in ix. 2.9a These stands may possibly have carried the pastry cakes iii. 4. 065-71, and similar offerings for the dead. To these belonged undoubtedly the strings of paper ` cash ', iii. 4. 04 (Pl. XCIII), and perhaps also the artificial leaves made of silk, iii. 4. 02.
But the object claiming most interest among the relics of this tomb is certainly the fine painting on silk iii. 4. 010 (Pl. CV, CVI), unfortunately surviving only in the form of numerous fragments. They are all extremely brittle, and only the exercise of great care made it possible to recover them safely, while clearing the sand from the floor of the principal chamber of the tomb. What position the painting had originally occupied it was impossible to determine. But the arrangement observed in the large piece, of which Pl. CV reproduces the most interesting portion, made it clear at the outset that the fragments belonged to a Makimono-like scroll which had been roughly handled and broken up when the tomb was plundered. Pl. CVI shows the most characteristic of the remaining
la Cf. above, i. p. 231.
8 See Serindia, ii. pp. 1049, 1056 ; Th. Buddhas, pp. 21 sq., Pl. X, XI ; also above, p. 653, concerning Ast. iii. 2. 022.