146 FROM KHOTAN TO LOP
might have traced to the east and north-east of the same line. My move disclosed no fresh ruins ; but in the close vicinity of N. xL, near the southern extremity of the site as explored in 1906, a badly eroded dwelling, N. xLV (plan, Pl. 5), built in timber and plaster was found to have escaped attention at that time. The clearing of what remained of two rooms brought to light eighteen Kharosthi documents on wood, including a double rectangular tablet, a very large wedge-shaped tablet, & c., besides miscellaneous household implements, & c., shown in the List below, among them a sandal-maker's wooden last, N. xLV. i. 03, and a mouse-trap, N. xLV. 01. The thick layer of sheep-dung, which covered the floor of the two rooms and had saved them from being completely eroded, also accounts for the brittle condition of the wooden documents. From the fact that on the route followed the open patches of wind-eroded ground were practically bare of any pottery fragments or other ` Tati ' debris, it may be concluded that the ancient settlement did not here extend westwards.
Relics found Muhammad 'Vag. 'fib's party, which rejoined my camp at N. xxiv, where our discovery of the
on eroded hidden archive was made in 1906, had failed to trace any previously unexplored ruins in the course ground.
of its search eastwards. But as the Surveyor, who was new to the ground and the work, had failed to recognize correctly our former landmarks, the actual course followed by him remained doubtful 14 Among the small relics, mostly beads, metal fragments and the like, picked up from bare patches of ground by the men with him or by others on the occasion of our renewed visit to the site (N. 01-25), special mention may be made of the small gold ear or nose ring in open filigree work, N. 03 (Pl. XXIV), the well-preserved bronze buckle, N. o6 (Pl. XXIV), and the barbed arrow-head in bronze, N. 021 (PI. XXIII). The latter agrees closely in type with one of the few previously found at this site.r5
I was able, by returning to this central portion of the site, to recover the large and well-preserved cupboard, N. xxvi. of (Pl. XV), which had been unearthed in 1906 from one of the rooms, viii, of the ancient residence N. xxvi (Fig. 103), and which difficulties of transport had then induced me to leave behind under a safe cover of sand.16 Its constructive features, and in particular the peculiar shape of the legs, make it certain that, like the other cupboards discovered at the Niya Site,17 it was intended to keep victuals safe from the attack of rodents. The discovery of pieces from the similar but more decoratively treated cupboard, L.B. 111.1,18 at the Lou-lan Site proves that we have in the Niya Site relic a practically intact specimen of a type of ancient household furniture uniformly in use throughout the Tarim basin during the early centuries of our era. From N. xxvi I now also removed the decorated wooden jambs, N. xxvi. oio, of the door through which access was gained from the central passage of the house into room v,19 as well as the pair of roughly carved wooden capitals, N. xxvi. 05-6, found detached and previously left behind 20
The dunes near the group of ruined dwellings N. xxiv—vi appeared to have shifted to some extent since my visit in 1906, and though they had left the condition of the ruins that I then explored practically unchanged, they had disclosed a good many more ancient fruit trees lying on the ground. In the same way there had now come to light traces of two more houses close to the east of N. xxvr, which being very badly eroded had before been completely hidden by drift-sand. These were now
Ancient cupboard from
14 I have reason to regret that Miân Afraz-gul, who subsequently gave ample proofs of his special aptitude and keenness for such archaeological reconnaissance work, was during our visit to the Niya Site incapacitated for it by illness.
15 See Serindia, iv. Pl. XXIX, N. xrv. oo8.
16 Cf. ibid., i. p. 235, Fig. 57.
17 See Ancient Khotan, i. pp. 377, 379 ; Serindia, i. p. 224, Fig. 57 ; iii. Pl. r r.
18 See Serindia, i. pp. 395, 443 ; iv. Pl. XLVII.
is See ibid., iii. Pl. 15.
20 Objects similarly recovered are the well-made agricultural fork of wood and the roughly ornamented wooden tray, N. xiii. or, 02, mentioned in Serindia, i. p. 217 and shown there in Fig. 52. Both were among the miscellaneous objects excavated in 1906 from the ruined dwelling N. mu and left behind at my old camp near N. xxvr after being photographed.