THE REGIONS SURVEYED
I have also explained the causes which render such terminal oases'
oases particularly liable to changes in position and extent during different
periods. 10 Evidence of these changes survives in the remains of those numerous ancient settlements abandoned to the desert which my explorations have enabled me to trace. The fact of the most important among them being found far to the north of the present line of 'terminal oases' furnishes definite proof of the progress of desiccation in this region within historical times. 11 Thus the detailed surveys attending those explorations may claim special interest in connection with a much-discussed geographical question.
SECTION IV.—THE TERMINAL DEPRESSION OF LOP AND THE
The last of the regions comprised within the Tarim basin is the terminal depression
of Lop. The smallest in extent and particularly well defined, it exhibits a variety of interesting geographical features. It may be appropriately described as containing the terminal course of the Tarim with its fringe
of lagoons, the marshes in which its waters are finally lost, and the great salt-encrusted bed of the dried-up Lop sea beyond them, together with the wastes of gravel, drift-sand and wind-eroded clay which surround it. In accordance with the traditional application of the name Lop, itself of very ancient origin, 1 we must include in this region also the dune-covered area to the east of the Tarim's final course, already referred to as an outlier of the Takla- makan, as well as the area, mostly scrubby desert or gravel 'Sai', which extends southwards of the last sections of the Charchan river and the Tarim to the foot of the mountains.
Here the streams of Viish-shahri, Charkhlik and Miran have in recent years rendered
it possible for a few small settlements to resume cultivation near ancient sites abandoned for centuries. 2 Apart from the people in these tiny oases and the survivors of the scanty nomadic population of Lopliks
('Lop people') fishing and hunting along the terminal Tarim, the whole region is now wholly uninhabited. The same applies also to the surrounding areas : in the west the Taklamakan; in the north the barren hills of the Kuruk-tagh; in the east the terminal basin of-the Su-loho, and in the south the arid ranges of the Altin-tagh, an eastern extension of the K'un-lun.
In spite of its desolate character, considerable interest, historical and geographical,
attaches to the Lop region. This explains the special efforts devoted to
Prehistoric sea bed. its survey both on my second and third expeditions, notwithstanding
the great physical difficulties besetting topographical work in a region which for the most part is devoid of drinkable water. In the great salt-encrusted bed, proved by our surveys to extend for fully 170 miles from south-west to north-east with a maximum width of some 80 miles, we have a visible remnant of that prehistoric salt sea which was fed by the drainage of the Tarim basin during periods when the climate of Central Asia was moister.
A variety of observations justify the assumption that this dried-up terminal basin, still showing salt bogs in places, even now periodically receives water
Marshes of Lop-abr. at its south-west end, either by inundation or percolation, from the
adjoining Kara-koshun . marshes of the Tarim. 3 The fact that these marshes of 'Lop-ncir'—to use the Mongol designation which from modern Chinese maps and
Extent of Lop
Cf.•.ancient Shelton, i. pp. 95 sq., 285 sq., 383 eq., 419 sq.; Serindia, i. pp. 202 sqq.
I' See Sheet No. 14. C. 1, 2 for the sites of Dan_ diin-oilik, Uzan-tati, etc.; Nos. 18. B. 4; 19. B. 1 for thé 1•iya site.
Regarding the history and early application of
the local name Lop, cf. Serindla, i. pp. 318 sqq. In its present form it is first recorded by Marco Polo, see Yule, The Book of Ser Marco Polo, (third edition) i. pp. 194 sqq.
? See Sheets Nos. 26. C. 3 ; 30. A, B. 2.
3 See Sheet No. 30. C. 1, 2.