within on the right of Fig. 262. Its top bears brickwork manifestly of a later date. Except for a decayed mosque near the south-west corner, the interior of the circumvallation is entirely clear of structural remains above ground. But small terraces of natural clay, divided by a network of narrow sunk alleys, evidently once served as basements or foundations for dwellings. On these basement terraces . I frequently noticed round holes, 2 to 3 feet in diameter, which manifestly had been excavated for the purpose of storing grain, etc. The narrow alleys dividing these little insulae must also have been excavated from the hard clay soil. But this became fully intelligible to me only after I had seen the streets similarly cut into the ground, but to a greater depth and wider, at the site of Yar-khoto, the ancient Turfan capital (Fig. 273). Close to the north and west walls passes the stream from Ilikul which serves to irrigate the fields of Lapchuk. Beyond this stretches the area already referred to which has been used as the burial-ground of the village, evidently for a prolonged period. Close to the south wall are the homesteads of Lapchuk scattered among fields.
I did not succeed in obtaining any useful local information about the ruined town. But frôm such indications as the badly decayed state of the walls, in spite of very massive construction, and the total absence of habitations within could furnish, it appeared to me very probable that the circumvallation dated from a period preceding, though nbt, perhaps, by a long time, the first introduction of Islam. The existence of this old fortified town and of the ruined Buddhist shrines north of it, which probably belong to the Uigur period just like those of Ilikul, has a special historical interest in view of the evidence which recent researches of Professor Pelliot have brought to light
as to the mention of Lapchuk by early Chinese records under the name of Na-chih ä6Q.13 It
appears from the texts which he has discussed with much critical care and learning that ~tNa-chih, mentioned by the Tang Annals as a sub-prefecture in A.D. 63o and located by other Chinese texts, including one of the early ninth century, to the west or south-west of Hami,14 was founded in the sixth century A. n. as a colony of ` barbarians ' who had emigrated from Shan-shan, i.e. the present Lop region. M. Pelliot has further demonstrated, in what appears to me a very convincing fashion, that the Chinese Na-chih, in accordance with certain rules of early Chinese phonetics, is meant to reproduce an older form *Lop-chuk. This itself seems made up of Lop, the indigenous name of Shan-shan, the antiquity of which is proved by the Nob of my Tibetan documents from Miran and Hsiian-tsang's Na fu fto, and the well-known Turki suffix =chuk. Thus the naine Lapchuk presents itself as an appropriate designation for the old colony founded in Hami territory by Lop emigrants.
The antiquity proved for the settlement of which the ruined town marks the northern extremity justifies my adding here some notes on what I was able to observe about the ground visited on my ride further down the valley. The homesteads of Lapchuk, counted at over a hundred, lie scattered among gardens and luxuriant orchards, which stretch for a distance of about a mile below the old site. The fields belonging to the village extend for some three miles further down between the flanking gravel plateaus; but most of this large and carefully terraced area of cultivation is now sown only every third year, as the available water was declared to be insufficient for irrigating the whole. It seemed as if want of adequate labour for tilling and manuring might also have something to do with this present limitation.
13 See Pelliot, Le ` Cha Icheou lou fou ton king', J. Asial., 1916, janvier-février, pp. i 17 sqq.
l The Yüan ho chün hsien l'u chih, published between A.D. 8o6 and 814, places the sub-prefecture of Na-chih 120 li to the south-west of I-chou or Hami ; see Chavannes, T'oung pao, 1905, p. 532 note.
On the other hand, the important text No. 917 among my Ch'ien-fo-tung manuscripts, as quoted by M. Pelliot, when
describing the sub-prefecture of Na-chih and its origin, indicates its position as 320 li to the west of I-chou. The bearing here given is approximately correct, as a reference to Maps Nos. 69, 73 will show. But the distance is manifestly exaggerated, as it is less than forty miles by the high road from Hami town to Lapchuk. In the other notice quoted by M. Chavannes the distance given seems to come closer to reality, but the southwest bearing is wrong.