Sec. i] KORLA AND ITS OLD SITES 1231
habitable portion of the westernmost Kuruk-tagh has been mentioned above and is subject to no doubt.3 Against the plain evidence of this bearing and of the topographical indication furnished by
the passage of the Wei lio to be presently mentioned, no importance can be attached to the distance of 500 li which the Ch`ieua Han shu notice records between Wei-hsü and the seat of the Governor-General ; similar manifest errors of distance reckoning can be proved more than once in the ` Notes on the Western Regions '. We are told that Wei-llsü contained ` 700 families, comprising a population .of 4,900', figures which, whatever their intrinsic value, seem reasonably proportionate to the 4,000 families and 32,100 people mentioned by the same text for Yen-ch`i or Kara-shahr.
The Annals of the Later Han do not specify the position of Wei-hsü, but mention it along with the kingdoms of Yen-ch'i, Shan, and Wei-li among the territories by the punishment of which in A. D. 94 the Protector-General Pan Ch`ao completed his pacification of the Western regions.4 We receive a more definite topographical indication in a passage of the Wei lio (composed between A.D. 239-65) which describes the ` route of the centre ', discussed by us before, as passing from Lou-lan to the kingdoms of Wei-li, Wei-hsü, and Shan, ` which all depend on Yen-ch'i'.A In view
of the geographical order in which the Wei lio's description of the route proceeds we are justified in placing Wei-li Y,~ it between Lou-lan and Wei-hsü, and with this all other indications concerning Wei-li fall exactly into line. The Former Han Annals' notice of this territory clearly states that it adjoined Shan-shan, i.e. the Lop region, and Chii-mo or Charchan on the south.6 At the same time we are told that Wei-li lay 240.1i to the west of Shan, or the territory of the westernmost Kuruktagh, being thus slightly nearer to it than Wei-hsü. These bearings and that of Chii-li, to be discussed presently, which lay to the south-west of Wei-li, necessarily take us to the' tracts which stretch along the Konche-darya below Korla approximately as far down as its present junction with the Taring near the large village of Tikenlik.
My journey of 1915 along the Konche-darya showed me the extensive area of cultivable and easily irrigated lands which stretches from the left bank of the river below Konche (Map No. 45. D. 3) to the foot of the Kuruk-tagh hills, and which the Chinese have. in recent times endeavoured to colonize as the new district of Kara-kum or Konche.7 The surveys made on the same journey to the north-west of Tikenlik showed also the number of small agricultural settlements which have sprung up recently along the branching beds of the Tarim and the Inchike-darya, in spite of the difficulties caused by riverine vagaries. In view of these observations I think we can safely identify Wei-li with the large cultivable, though at present very imperfectly developed, area just described. Its natural boundary northward is likely to have been the belt of low barren terraces which juts out westwards from the foot of the Kuruk-tagh near the village of Shinalga, and which at present divides Kara-kum from Korla. In agreement with the relatively great extent of the area indicated, we find that Wei-li is credited in. the Former Han Annals' notice with a larger population than Wei-hsü or Korla, viz. 1,200 families and 9,600 persons.
cation, proposed in a Chinese dictionary published in 1766, of Wei-hsit with a place called Chagan-tungi to the north-east of Kara-shahr. M. Chavannes has already (T'oung pao, 1905, p. 552, note 5) duly emphasized the need of caution in the case of these identifications of the Hsi fit lung win chip. In this present instance the bearing of Wei-hstt relative to Shan suffices to prove the impossibility of the location proposed.
s Cf. above, pp. 333 sq.; Chavannes, T'oung pao, 1905, p. 552, note 7. Grenard, Mission Dulreuil de Rhins, ii. p. 6r, first correctly identified the position of Shan.
Cf. Chavannes, T'oung pao, 1907, pp. 210 sq.; also T'oung pao, 1906, pp. 234, 236.
S See Chavannes, 'young pao, 1905, p. 552, with notes 5-7 ; cf. also above, p. 418.
6 Cf. Wylie, J. Anlhrop. Inst., xi. p. lox ; for Chtt-mo and Shan-shan, see above, pp. 295 sq., 323 sqq.
For a brief preliminary reference to this interesting area which, given an adequate supply of suitable colonists and efficient administration, could easily be developed into a large oasis, see Third Journey of Exploration, Geogr. Journal, xlviii. p. 207.