546 ACROSS THE PEI-SHAN TO BARKUL [Chap. XV
as will be seen below.38 Hence it is safe to assume, as suggested by M. Chavannes, that while one portion of the Ch`u-yüeh was in occupation of the Tien-shan slopes west of the Barkul lake, the other, represented by the Sha-t`o, held the valleys to the east of ît.37 The fact that I-p`i Tu-lu, Kagan of the Western Turks, ravaged H âmi in A. D. 642 with the assistance of the Ch`u-yüeh fully accords with this location.37a Mount Chin-so, apparently not otherwise mentioned, may with some probability be identified with the range extending all along to the north of the valleys of Tur-köl and Barkul and having its continuation westwards to about 910 longitude. The ` great stony desert ' of Sha-t`o might well be looked for on the northern slopes of this range, about which
I am unable to trace any information beyond the fact that they are visited by the herdsmen of Turköl and Nârin-kür for winter grazing.
Tribe of the The Sha-t`o are mentioned as forming one of the governments created by the Chinese out of
Sha-t` o. the vast dominion of the Western Turks after their final defeat in 658-9,38 and details about certain
of their chiefs are recorded in the notice above quoted from the Tang Annals down to the time when Tibetan aggression began to shake Chinese supremacy over these regions. It is of interest to note that late descendants of those chiefs rose to importance during the troubles following the Tang period, by founding the short-lived Posterior Tang, Chin, and Han dynasties (923-51 A. D.).39
Later For close on a thousand years after the period covered by the above Tang notices relating to
historyBarkul no specific information from Chinese sources regarding this territory appears to be accessible of Hamm and
Barkul. in translation. But there can be little doubt that, like Hâmi, it was held by Uighur chiefs during the
greater part of later mediaeval times.40 More abundant notices are available for the chequered history of Hâmi in Ming times. But they do not tell us what part the Barkul valley played in the frequent inroads to which the oases south of the easternmost Tien-shan were exposed.41 With the rest of the territories now comprised in Chinese Turkestan, H âmi had passed under the power of the Dzungars when the Chinese under the great Emperor K`ang-hsi resumed towards the close of the seventeenth century that policy of Central-Asian expansion which had been in abeyance for fully nine hundred years. K`ang-hsi's great victory over Galdan, the supreme chief of the Dzungars, won in 1696 in the Kobdo region, marks the beginning of the reconquest of China's ancient Central-Asian dominion. Significantly enough it was in the very year of that victory far away in the Mongolian north-west that Hâmi made its definite submission, and that the route leading to it from Barkul was secured by a Chinese post,42 like other routes of approach to that important base.
Hâmi, though garrisoned by Chinese troops, remained exposed to Dzungar attacks for nearly half a century longer, and we are repeatedly told of Chinese forces sent to Barkul and posts established in that territory to help in warding them off.43 But it was not until the successful operations initiated by the great Emperor Chien-lung had led, in 1759, to the complete conquest of Dzungaria and of the Tarim basin, that security was finally achieved for the high road past Hâmi which links the ` New Dominion ' with China.
Turkestan Where clearly defined geographical conditions reign supreme and are so little affected by human
retaken activity as on both sides of the Eastern Tien-shan, history must necessarily often repeat itself
Tungans. even in details. After 1863 both Dzungaria and the Tarim basin were lost to the Empire through
the rebellion of the Tungans, or Muhammadan Chinese, who formed a large portion of the Chinese