those springs at the foot of the Kuruk-tagh northward, while labourers and antiques were sent back to our depot at Miràn. 2E After replenishing our ice-supply and taking an indispensable store of fuel we left Altmish-bulak on February 25 for our respective tasks. Lai Singh was to survey the extreme north-eastern extension of the great salt-encrusted basin once filled by the Lop sea and the southernmost hill ranges of the Kuruk-tagh overlooking them. I myself wished to trace the ancient Chinese route from the eastern edge of the once habitable Lou-lan area right through to the point where it was likely to have diverged from the line still followed by the desert track from Tun-huang along the southern shore of the dried-up sea.
It was a task of special geographical and historical interest but beset also by serious
Search for ancient physical difficulties and risks ; for on the ground to be crossed no water
Chinese route from could be expected—over most of it not even fuel—before striking the
Lou-lan to Tun-huang. Tun-huang caravan track near the eastern extremity of the ancient sea bed, a mattgr of some ten days of hard marching. Apart from the serious risk of physical obstacles which would cause delay and exhaust our hard-tried camels, there was the problem of striking the line of the ancient route and of tracking it through a wilderness devoid of all resources since the dawn of historical times. I have related elsewhere how hints derived from topographical and archaeological observations, combined with fortunate finds of relics left behind by the ancient traffic of centuries, helped to guide me and to solve the problem. 25 Here the briefest explanation of the route, as now shown on the map, will suffice.
After regaining across difficult wind-eroded ground the vicinity of the terminal point d'appni above mentioned at Camp c, we moved for two long marches to Crossing of salt-encrust. the north-east until we struck the belt of salt-encrusted erosion terraces
ed Lop sea bed.
which the early Chinese accounts of this dreaded 'northern road' knew as the `White Dragon Mounds'.2° Then on a very trying march we crossed to the south-east the dried-up sea-bottom with its crumpled-up crust of hard salt, fortunately at the very point where it was narrowest. Thence continuing over easier ground to the south along the ancient sea shore, we reached three days later its extreme eastern extension in the shape of a great bay overlooked from the north by a low offshoot of the southernmost Kuruk-tagh.
For two more days we skirted this bay eastwards under the steep cliffs of its shoreline, and then crossed its salt-encrusted expanse, here still showing patches
Tun-huang caravan of actual salt bog. After a long day's march on March 6 we finally track gained.
reached the wells of Kum-kuduk, on the Tun-huang caravan track. 27 Here I found Lal Singh just arrived after having duly surveyed the wide northernmost bight of the dried-up sea and of the straggling low ranges further to the east. In conjunction with the work pluckily done a year later by Afraz-gul along the western shore our ' circumnavigation ' of the ancient Lop sea was thus successfully achieved.
After the timely arrival of our heavy baggage from Miràn a day later, we were able
toghrak, near the eastern end of the great desert valley leading down towards the Lop. sea basin, I picked up Surveyor Muhammad Yakàb who had meanwhile carried with praiseworthy perseverence a line of levels, carefully observed with a Zeiss instrument, all the way up from the bottom of the bay north of Kum-kuduk to the curious Mesa-filled basin east of
=' See Sheet No. 20. D. 3.
26 Cf. G. J., xlviii. pp. 126 sqq.; Geographical Review, ix. pp. 26 sqq.
x See Sheet No. 32. A, B. 8, for the route from Camp xcia to Camp cii. The point where the belt of
salt-coated 'Yärdangs', as distinct from Mesas, was first reached is marked by the entry referring to an important find of relics of ancient traffic, circ. 6 miles E. of C. ci.
27 See Sheet No. 32. B. 3, C. 3, 4, D. 4.