224 THE NIYA SITE [Chap. VI
tablet came to light, the only other find being a lignite seal discovered in one of the northern rooms (see N. xx. oor, Plate xxlx).
Ruin N. xxi. N. xxi was the ruin of a small dwelling about half-way between N. xviii and N. xx, and was
eroded down to a level of some six inches from the plastered floor. Only two rooms, about ten feet square each, were traceable on the little plateau created by wind erosion. Just under the sitting platform of the northern room lay a human skull, and bones of the same skeleton strewed the slope below. Were these the remains of another hapless ` treasure-seeker ' ? A small oblong Kharosthi tablet, the decorated neck and handle of a large hand-made pottery vessel (N. xxi. oor ; see Plate xxxvi), and the bottom of a lacquered wooden bowl, still retaining the rivets of some ancient repairs, were the only finds here.
N. xxii. N. xx and N. xxi. The rooms traceable formed a suite running north to south (see plan, Plate r 3),
The southernmost ruin of this group was N. xxii, situated about a quarter of a mile from
and their clearing was made difficult by the presence of a dune, about thirteen feet high, which adjoined and partly overlay them on the east. The excavation of them cost nearly the whole of October 23, but was rewarded by a good yield of Kharosthi documents on wood. In the small northernmost room, i, with walls of rushes and plaster, there were found seven tablets, mostly of oblong shape ; a heavy wooden comb as used by weavers (N. XXII. i. oor ; see Plate xXVIII) ; and a quaint little doll, N. xxii. i. 002, carved in wood with hinged legs and dressed up in gay silks. From the adjoining small apartment, ii, two oblong tablets were recovered. Much more abundant were the Kharosthi documents on wood which came to light in room iii, measuring twenty-one by eighteen feet and filled with sand to a height of over seven feet. Most of the twenty-three tablets were of the wedge type, three of them being complete double-wedges. On one of these, N. XXII. iii. i8, the obverse of the covering-tablet still retained in perfect preservation the clay impression of a seal, familiar to me already from my excavations of 1901, showing Pallas Athene with aegis and thunderbolt.28 Another but poor impression of the same seal appears on the covering-tablet N. xxii. iii. 16. All the tablets were found sticking to the floor and were encrusted with dirt, a sign that they had been thrown down there while the room was still inhabited.
Among the miscellaneous objects found, a large cupboard, raised on high and curiously carved legs, was, perhaps, the most interesting. It is reproduced in Plate r i from a drawing to scale made by Naik Ram Singh. It no doubt served, like the present nan-sandûk of Turkestan villagers, for the safe storage of food articles, and the peculiar carving of the legs was manifestly intended to render access impossible for small rodents. A cupboard closely resembling this in size and shape is seen in the photograph of N. xxvi (Fig. 57).27 To the north of the house a rush fence lined by a row of dead poplars mostly fallen could be traced for a distance of 105 feet up to where it disappeared under a dune. A short distance to the south the outlines of an ancient tank could still clearly be recognized within the oblong enclosure, about thirty-six by twenty-eight feet, formed by rows of large poplars (see Fig. 55). The big sand-cone, more than forty-two feet high, seen in the photograph by the side of the tank, was one of the very last in this direction still retaining living tamarisk growth.
SECTION III.—RECORDS FROM A HIDDEN ARCHIVE, N. XXIV
By the evening of October 24 my camp was shifted to a small group of ruins which in 1901 I had discovered only at the very close of my visit, too late for systematic exploration, and which had ever since made me wish for a return to the site. Its dwellings lay close together on the
R° Cf. Ancient Kholan, i. p. 354; ii. Pl. LXXI.
27 For similar cupboards found in 1901 at this site, cf. Ancient Khotan, i. pp. 377, 379.