国立情報学研究所 - ディジタル・シルクロード・プロジェクト
『東洋文庫所蔵』貴重書デジタルアーカイブ

642   THE END OF THE CHINESE LIMES   [Chap. XVII

Thus the ridges furnished excellent natural bases for the watch-towers, and the Chinese constructors of the Limes, with their unfailing eye for topography, did not fail to make the most of them. On this account these towers could always be sighted easily from afar. I had already recognized T. v and T. vt. a as towers on my first visit to T. iii, though it is fully fifteen miles away from them in a straight line. It was curious to note, as my survey progressed to the south-west, that the towers T. v, vi. a, c, d were seen to fall into an almost straight line from T. ttt as the starting-point, as if special care had been taken to fix their positions by sighting with a diopter.

Physical   The special quasi-geological interest belonging to certain physical features of this strange

features of inland coast-line have been so fully discussed elsewhere 3 that it will suffice very briefly to refer to ground.

two of them which have a direct bearing on the positions chosen for the flanking line of watch-stations. The clay ridges mentioned above which form the headlands jutting out into the marshy basin, and the long-drawn bays and inlets which intervene between them, invariably lie in a general direction from south-east to north-west. It is clearly seen from the Map (Nos. 74, 75) that this is due to the erosive action of the water once carried down from the foot of the distant mountains and across the gravel glacis, now utterly dry and barren. The deep-cut beds, which descend to the heads of the inlets and in places are traceable along the steep foot of the ridges, were also surveyed afterwards in their upper ravine-like courses cut across the glacis. There were signs that these beds are swept by exceptional floods even at the present time, though the intervals of absolute dryness may last many years. In any case, there is conclusive evidence of subsoil drainage in the springs which, as I found, rise within the bays of the basin. The water of those higher up was fairly drinkable, but further down towards the edge of the marsh area it rapidly grew salt. To this consideration we may reasonably attribute the fact that the watch-stations T. v and T. vt. a were placed, not at the far end, but about the middle of the clay ridges which they respectively occupy.

Clay   The same consideration must have also made itself felt as regards the position of the last two

terraces con- stations, T. vi. c and T. vt. d, which are found on isolated clay terraces not too far from the ` coast-

tinuing line

of ridges.   line '. These terraces themselves afford striking evidence of the great effect which another and,

throughout the historical period, certainly far more powerful agent has had upon the present configuration of the ground. It is wind-erosion which is at work here, and has been for ages. There can be no doubt that to its action are due the strings of isolated clay terraces, or Mesas, found within the wider bays and running parallel to the bordering plateau ridges, as well as those which, forming a continuation of these ridges, project into the marshy expanse of the basin and still maintain the same bearing. Both groups of Mesas are extremely numerous, though for obvious reasons only a small proportion could be marked on the map.

Ridges   The narrow ridges formed out of the alluvial plateau by the action of running water were

broken

by wind up bound to be broken up into terraces by the cutting and grinding force of the wind, as they stretched

erosion.   more or less at right angles across its prevailing directions from the north-east and east. On the

crest of the still continuous ridges the sapping and scooping action of the wind could be observed in progress, marked by wind-eroded trenches usually starting from the north-east. Such a trench, of considerable size, is seen in the foreground of Fig. 169.4 It is the interaction of water and wind-erosion, here so clearly illustrated, which helps best to explain the maze of detached clay terraces encountered along the north-eastern shores of the ancient Lop Sea bed, in the dried-up terminal basin of the Su-lo Ho east of Besh-toghrak, and again around Khara-nôr.5

3 Cf. Desert Cathay, ii. pp. 139 sqq.

' On the left of Fig. 17o is a similar but completed trench by which a clay terrace has been entirely separated from the end of the ridge on which the tower T. iv. c stands. Cuttings

of the same type, due to wind-erosion, are illustrated by Figs. 187-90, which show ground towards Khara-nôr.

6 Cf. above, pp. 341, 423 sq., 552, 575, 589 ; Third Tournai of Exploration, Geogr. Journal, xlviii, pp. 127 sqq.