Sec. iii] WATCH-STATIONS ALONG THE ANCIENT ROAD TO KORLA 777
There was no time for the examination of these humble resting-places, as we still had a long march before us that evening, to the newly established district head-quarters of Kara-kum south-westwards. The route followed took us first across a wide area of tamarisk and scrub-covered steppe, where shallow flood-beds and plentiful salt efflorescence indicated liability to inundation both from the drainage of the westernmost foot-hills of the Kuruk-tâgh and from the overflow of canals from the Konche-daryà. Some seven miles from Yâr-karaul we came upon the first detached patches of cultivation belonging to ` old Kara-kum ', on ground known as Gherilghan. Then over waterlogged roads, through an alternation of fields and unreclaimed land, we reached the half-abandoned market-place of Kara-kum ` Ki5na-shahr '. The official head-quarters had been shifted thence in the preceding autumn to the ` new town ' of Kara-kum, situated on the left bank of the Konchedaryâ eight miles farther south, and there we arrived long after nightfall, having covered fully thirty-one miles in the day.
SECTION IV.—THE TERRITORY OF WEI-LI AND THE MODERN KARA-KUM
I was obliged to halt two days at Kara-kum, partly in order to let men and animals have some rest after the rather trying desert journey and partly in order to save if possible one of our camels which had been severely wounded, near Sanje, by the accidental discharge of Ibrahim the hunter's gun. We had been obliged to leave the poor beast behind, with one of the men to take care of it, in the hope of bringing it in later. This hope was destined to disappointment, and the victim of our guide's want of care had to be killed by Hassan Akhûn, my camel factotum, on his return to the spot with assistance.
This halt at Kara-kum proved useful, inasmuch as it enabled me to secure reliable information about that out-of-the-way and comparatively little known Chinese district which comprises the northern portion of the Lop area and the tract between the upper Konche-daryd and Korla. I have already had occasion to discuss in Serindia the reasons which lead me to identify the territory
of Wei-li U with this modern district.' A brief summary of the Chinese historical notices
referring to it will therefore here suffice.
The notice in the Former Han Annals makes the ` kingdom of Wei-li ' join Shan-shan and Chü-mo on the south.2 This corresponds correctly enough to the position of the district now officially known as Kara-kum, which extends along the Konche-daryà from below Korla down to near the river's junction with the Tarim below Tikenlik. We have seen that, according to the Chien Han shu's notice of Shan jjj or the western Kuruk-tâgh, the chief place of this territory was situated 240 li east of Wei-li, which clearly takes us to the vicinity of the present Kara-kum.3 The distance to, and bearing of, the seat of the Governor-General, probably in the present oasis of Yangi-hissdr,
viz. 36o li to the west, are approximately correct. So also is the statement that from Cheu li
the riverine tracts between the Inchike-daryâ. and the Yarkand-daryâ from Shahyâr downwards, the road led to Wei-li.4 The population of 2,000 families recorded for Wei-li as compared with that of 700 families for the adjoining Wei-hsii or Korla suggests a fairly large territory.
The Later Han Annals do not furnish a special notice of Wei-li, but mention its ` king ', in significant conjunction with the chiefs of Yen-ch`i (Kara-shahr), Wei-hsü (Korla) and Shan, all territories closely adjacent to Kara-kum, as having been defeated and punished by Pan Chao in
1 See Serindia, iii. p. 1231.
2 Cf. Wylie, J. Anthrop. Inst. xi. p. lox.
3 Cf. above, ii. p. 763 ; Wylie, loc. cit., p. 1o5.
4 Cf. Wylie, ibid., p. loo. I may add here to the statement