beams there were found here a pair of well-carved double cantilevers in wood (Kha. v. oo3. a, b; Plate xviI) which once, no doubt, supported the roof. A wooden key, Kha. v. ooi, from the room, resembling in make others found at this site (Kha. ii. 0038 ; ix. oo8 ; Plate xvli) and subsequently elsewhere, soon received full explanation by the discovery of a wooden lock with bolt and tumbler block, which turned up outside on the veranda floor (Kha. v. oo6 ; Plate XVII) but which must have been worked by a different key. The system of these wooden locks and keys is fully explained by Mr. F. H. Andrews's lucid note and drawings given in the Descriptive List below (see Kha. v. oo6). Here it must suffice to mention that it is substantially identical with one still surviving from the southern shores of the Mediterranean to Yârkand, Khotan, and Ladak, and that its use can be traced back to classical times.13 There was found here also an excellently preserved broom, of a make identical with those I found in 1901 at Dandân-oilik and the Niya Site.14
More important for the student of ancient industry was the discovery of a heap of cotton seed, furnishing indisputable proof that the cultivation of cotton was established in the Khotan region by the eighth century A. D. Plentiful as have been my finds of ancient textile materials of all sorts from Khâdalik to Lop-nor, they have, as far as their expert analysis has proceeded, proved to be manufactured only of wool, hemp, or silk. Whether the rags and coils of thread found in this identical veranda contain any cotton I have so far not been able to ascertain. [Since the above was written, I have been informed by Dr. Hanausek that the specimen Kha. v. oo5, which he has been kind enough to analyse under the microscope, has proved to contain cotton threads.]
The subsequent clearing of what uneroded ground was left to the east of Kha. i showed a plastered flooring throughout, but did not reveal the presence of any structural remains. But near the southern edge of this area, Kha. vi, we came upon numerous fragments of leaves in Brâhmi script and an oblong tablet with the same writing. They had evidently been deposited near the entrance to a small shrine the walls of which, a little over ten feet square, had left their trace in shallow depressed lines marking the position of the foundation beams on the floor. Within the cella space were found the small relievo image in wood of a standing Buddha, Kha. vi. 6 (Plate xIV) ; a well-carved wooden pedestal for a statuette with lotus throne, Kha. vi. 17 (Plate XLVII) ; the fragment of a painted panel ; a well-preserved paper document with cursive Brâhmi script on both sides, Kha. vi. 14, with more Pothi fragments and inscribed tablets. About halfway between Kha. i and Kha. vi, but somewhat further north than the latter, a circular hole, seven and a half inches in diameter and two feet deep, was discovered in the floor. This was still closed with a wooden peg wrapped with a brown woollen fabric, Kha. vi. I. a, but the small receptacle was found empty.
On a piece of ground some eighty yards to the north-east of Kha. i small stucco fragments both from decorative relievos and frescoed walls had attracted my attention. The clearance here effected on September 29 showed that of the shrine which had once stood here, Kha. vii, nothing survived but the plastered floor and the sunk lines left in it by the foundation beams of walls, about twenty-five feet square. The depressions marking the position of the beams were three inches deep and six wide, with remains of timber and reeds at the bottom. Among
" My late assistant Mr. H. G. Evelyn-White supplied me in 1910 with the model of wooden locks of exactly the same construction which he had found in modern use at the Khargah oasis, Egypt. M. Arnold van Gennep, editor of the Revue d'Ethnographie el de Sociologie, who has made a special study of such lockings, after reading my reference to the Khâdalik find in Desert Cathay, i. 245, wrote
to me in a letter dated March 12, 1912 : ' Judging from your phrase, the system ... is the same as the Zanzibar one. I found it last summer, wooden in the Aurès, of bronze at Tlemcen. A friend sent me a description of the same, of iron, at Figuig, and another brought me back a wooden one from the Chaouïa in Morocco.'
" See Ancient Khotan, i. pp. 251, 336 ; I. Pl. LXXIII.