564 TO GUCHEN AND ACROSS THE TIEN-SHAN [Chap. XVI
Tirthas. The distance of 8o li is easily reconciled with the approximate equivalence of about four li to the mile which results from the comparison of certain other distances recorded in itineraries of the Tang shu between definitely identified localities in Chinese Turkestân.2
From the ` spring of the Dragon ' the itinerary takes us into a mountain gorge and then through Liu-ku, ` the Valley of the Willows ', to the pass crossing the Chin-sha ling,' the Mountain of the Golden Sand '. Reference to the account given above of our route will make it clear that the mountain gorge here referred to is the canon which the present route from Shaftulluk enters below Yoghan-terek. It is equally clear that Liu-ku, ` the Valley of the Willows ', derived its designation from the thickets of willows that skirt the route for miles in the valley portion extending above Yoghan-terek towards San-shan-k`ou. By the Chin-sha ling,' the Mountain of the Golden Sand ', only the watershed range of the Tien-shan can be meant, and by the place of crossing, the pass between Hsi-yao-tzû and Pa-no-pea.
If we assume that the distance of 13o li is meant to refer to the marches from Shaftulluk to Hsi-yao-tzû, the place where travellers coming from the south are accustomed to halt before ascending the pass next morning, and the last place where water and fuel are obtainable, the estimate given by the itinerary may be considered a reasonably close approximation to the actual distance, which is about 35 miles. The above assumption seems justified in view of the fact that the distances in Tang itineraries, like those in corresponding classical texts, being derived from records made for the practical guidance of travellers, are always measured between customary halting-places and are not intended to serve for the location of natural features. For travellers of old times, whether from the east or the west, it was far more important to know the distance to the last stage below a pass than that to the watershed of the range, a place which they would be only too glad to pass and forget as quickly as possible. I am unable to suggest whence the name of Chin-sha ling,' the Mountain of the Golden Sand ', as applied to this portion of the Tien-shan range between Turfân and Guchen, was derived. But there can be little doubt that we have in the Hou Han shu a reference to a similar name for the mountain. In the biography of Pan Yung we are told that in A.D. 126 the northern shan yü of the Huns invaded with ten thousand horsemen the territory of Posterior Chü-shih and arrived in the valley of Chin-chii4 IL before a Chinese force sent by Pan Yung obliged him to retreat.3
The name of ` the Valley of the Willows ' can be traced even farther back ; for it is already found, in the form of Chii-shih Liu-ku, ` the Willow Valley of Chü-shih ', in the Notes on the Western Regions, contained in the Former Han Annals. The notices in that text relating to the petty territories around Anterior and Posterior Chü-shih state of the ` Kingdom of Hu-hu' that it is ` in the Chü-shih-liu valley '.4 The very small population that this notice attributes to this ` kingdom ', viz. 55 families numbering 264 persons, fully agrees with the supposition that the reference is to a petty chiefship, comprising the valley drained by the Yoghan-terek river, perhaps with some other adjacent valleys south of Tien-shan. Mr. Wylie in his note on this passage pointed out that the same territory is alluded to also in the itinerary of the Chinese envoy Wang Yen-tê, who in A.D. 982, after `passing through the government of Chiao-ho ' or Turfân, ` traversed the Valley of Willows, made the passage of the Chin ling mountain, and reached the Uigur capital'.5 Here we meet with what is evidently an abbreviated form of the name Chin-sha ling for the moun-
2 Cf. e. g. the distance of 120 li between Shih-ch`êng (Charkhlik) and Hsin-ch'êng (Vâsh-shahri), on the route from Sha-chou to Khotan, and of 220 Ii between Tien-shan (Toksun) and Yin-shan (Kumush), on the Turfân—Kara-shahr route, as discussed in Serindia, i. p. 306; iii. pp. 1X77 sq. ; see also below, Chap. xxiv. sec. ii, iii.
3 See Chavannes, Toung-pao, 1906, p. 253 ; Dix Inscriptions, p. 22.
4 See Wylie, J. Anthrop. Inst., xi. pp. 104 sq.
5 Quoted by Wylie, loc. cit., from Julien, Mélanges de géographie asiatique, p. 11.