Sec. iii] THE OASIS AND CITY OF YARKAND 89
political centre has shifted from Yarkand since the re-establishment of Chinese rule, the above description still holds good ; and we may well conclude that the flourishing condition of the city which Marco Polo's account also indicates, was maintained from early times independent of political predominance.
SECTION IV.—KARGHALIK IN CHINESE RECORDS
We have seen already that Hsüan-tsang's itinerary, after leaving Kashgar, took him over a distance of 500 li in a south-easterly direction, to the kingdom of Chê-chü-chia. Since the pilgrim distinctly mentions the crossing of the gird or Yarkand river en route, there can be no doubt that Chê-chü-chia must be identified with the present district of Karghalik. The latter may be roughly described as comprising the hill tracts which slope down from the snowy ranges confining the valley of the Raskam Darya or Upper Yarkand river, and at their northern foot are bordered by the lower course of that river and the sands of the Taklamakân. Where the streams draining these hills debouch into the plain, either to carry their water into the Yarkand river, or to be absorbed in irrigation, we find oases of remarkable fertility formed on the rich alluvial soil. The largest of these, with a town of considerable size and importance, is the oasis of Karghalik, watered by the Tiznaf river and a number of smaller streams ; it is joined on the north by a belt of highly cultivated ground with numerous villages, which extends unbroken for about 25 miles from Karghalik town to the right bank of the Yarkand river.
It remains doubtful whether the capital of Chê-chü-chia, which Hsüan-tsang describes as having a circumference of twelve li, and defended by high and solid walls', is to be located at the site of the present Karghalik or perhaps further north, where, not far from the Yarkand river, we find another large and flourishing settlement at Posgâm-Bazar. But it is certain that his description of the territory, the circuit of which he estimates at a thousand li, or ten marches, accurately fits the district of Karghalik. ' La population est nombreuse ; les montagnes
Hsüan- tsang's Ché- chü-ch:â.
I I follow Julien's translation, Mémoires, ii. p. 221. This seems more accurate than that of Beal, yu-ki, ii. p. 307, which makes the capital ' hemmed in by crags and mountain fastnesses.' In order to find a position approximately corresponding to this description, we should have to look for the old capital at least two or three marches further south, in the direction of Kök-yâr. But apart from the fact that the confined nature of that valley would scarcely be suited for a capital, such a location could not be reconciled with the distance of 500 li, or five marches, indicated by Hsüan-tsang between Kashgar (Ch'ia-sha) and Chê-chü-chia. We know that the pilgrim, in estimating distances from one territory to the neighbouring one, regularly records that from capital to capital. Now the estimate of five marches from Kashgar may be accepted as approximately right, if the capital of Chê-chü-chia is to be placed at Karghalik, or rather some distance further north. But in order to reach Kök-yar in five days, a distance of over r8o miles from Kâshgar, the pilgrim would have had to make forced marches of a kind with which we cannot reasonably credit him.
In this connexion we may note a point in Hsüan-tsang's itinerary from Kâshgar to Chê-chü-chia which is in need of elucidation. The Mémoires, ii. p. 220, tell us : ' En sortant
de ce pays, il fit environ cinq cents li au sud-est, passa la rivière Si-to (§itâ), franchit un grand passage de montagne couvert de sable, et arriva au royaume de Tcho-keou-kia (Chê-chü-chia): If strict topographical sequence must be assumed in the description of the route, we should have to look for the great ' passage de montagne couvert de sable' to the south of the Zarafshan, and we could in that case identify it only with the ' Dasht ' at the foot of the hills which the most direct route from Kashgar to Karghalik (passing through Yaka-arik and avoiding Yarkand city) skirts just north of Karghalik. But taking into account the general topography of the route, it appears to me far more probable that Hsüan-tsang's reference is really to the curious ridge of much-decayed hills which runs into the desert south of Ordam-Pâdshâh, and which, at the point where the pilgrim's track from the latter place crosses it, bears the
Ziârat' of Ulûgh-Nishân (see Ruins of Khotan, p. 158 ; Hedin, Reisen in Z.-A., p. 3). The old route from Kashgar to Yarkand and Karghalik is, as shown above, pp. 86 sq., likely to have followed this, the shortest line ; and the unexpected passage of an isolated ridge in the midst of a level expanse of sandy desert may have impressed the pilgrim, just as it surprises the modern traveller.