Sec. iii] A HILL RANGE IN THE TAKLAMAKAN 83
of the Mazar-tagh, whence it irrigates the fields of Châr-bagh. Less than a mile from where we
camped by the river bank we crossed a canal, known as Kodai-daryd, which during the flood season
of the river carries water to the Chaghan-köl marsh SW. of Tumshuk ; this in turn serving as
a natural reservoir for the irrigation canals of villages farther east.
On the morning of October 26th we crossed the river at a ford where the single bed actually Ford across
holding water was about 55 yards wide with a maximum depth of 4 feet, the velocity of the current Yarkand
being only about 17 foot per second. The flood bed (Fig. 8i) was nearly a mile wide, and some idea of the huge volume of water that passes down at the height of the summer floods could be formed from the statement of our guide, a local hunter called Aziz (Fig. 86), that during about a month the water then overflows the steep clay bank, some i6 feet high, which bounds this flood bed on the south. All the day's march led through rich grazing, mostly of reed-beds, and luxuriant young jungle. The autumn tints of the belt of vegetation contrasted vividly with the dark brown and red of the desolate hill chains of the Mazar-tagh to the north and the much lower Kum-tagh to the south. The latter, which appeared nowhere to rise more than about a thousand feet above the riverine flat, was seen to have its south-west and north slopes overlaid by huge accumulations of dunes (Fig. 82) very similar to those that we had observed on the Bel-tagh.s
That the Kum-tagh, as it is appropriately known in consequence, represents a direct continua- Defile
tion of the Mazâr-tâgh is shown by the map (No. 8. B. i) ; and this is confirmed by the prevalence between
in the rocks of the Mazar-tagh of the same calcareous sandstone containing feldspar and permeated and Mazar-
by thin porphyrite strata that constitute the main formation of the Kum-tagh, as indicated by tagh.
Dr. Hedin's specimens and my own.6 The defile through which the Yarkand-daryà passes between
the two hill chains proved to be only about five miles wide. It presents a very striking feature in
the river's course, which everywhere else, from the point where it debouches above Yarkand right
down to its termination in the Lop-nör marshes, passes solely across level plains of alluvium and
drift-sand. I shall have occasion below to touch upon the question of the genetic relation which
may be surmised to exist between the river course at this point and the survival by its side of the
highest among the island-like hills that form part of the ancient diagonal range.
We camped near fine groves of Toghraks growing amidst reed-beds and low dunes to the Lake be-
north of the Kum-tagh. This is faced on the east by a bolder and higher hill range, running in tween Kum-
a parallel direction, approximately NW.-S E., and extending from the Yarkand river for a distance Chok-tagh. of over twelve miles. It is known to the Dolàns of Maral-bàshi by the name of Chok-tagh. The depression between the two hill chains, about four miles across at its widest, is filled for the greater part by a far-stretching lake, filled annually by the inundation of the river and known as Chöl-köl, the ` desert lake '. Its southernmost end was to serve as the starting-point for our desert venture, just at it had for that of Dr. Hedin. To reach it we, like him, found it best to proceed by the level and open stretch of ground that divides the west shore from the foot of the Kum-tagh. After crossing, not without some trouble, a series of depressions in which water remained from the last summer flood, we skirted the sandy glacis of an extreme north-eastern offshoot of the Kum-tagh. It was of interest here to note that the surface consisted of small flakes of completely decayed rock debris, dark red and of slaty appearance, forming a thin layer over fairly fine sand. The steady deflation and corrosion that is grinding away these remnants of the ancient mountain range could not find a better illustration.
Ever since leaving the river we had been following a well-marked cart track, and after skirting Salt-digging
the reed-beds that fringe the lake shore we came to a low but steep ridge jutting out from the above W.
foot of the Kum-tâgh. It was found to consist of detritus of a reddish rock, a hard silicious lake.
b See above, p. 75. 6 Cf. Hedin, Reisen in Z.-A., pp. 220 sq., 241.