CH. LXXI WIND-ERODED TOWN WALLS 239
to the prevailing wind direction have escaped with relatively little damage. What better illustration than this could I have desired to show why the east and west walls of the ancient Chinese station north of Lop-nor had completely disappeared while I could still trace those to the north and south ? Curiously enough the semicircular bastion in front of the east gate of old An-hsi has nowhere been breached, though its foot has been under-cut by driven sand in a few places. Evidently the sand not meeting here with absolute obstruction is diverted sidewise.
When I subsequently examined the walls of the present An-hsi town I found that the eastern face was in danger of falling from exactly the same cause, the onset of that relentless foe, the famous ` wind of An-hsi.' In order to prevent breaching on the top, this particular wall face had been protected, apparently since the reconquest after the Tungan rising, with a solid stone parapet. This had to some extent averted the attack above, though even thus I found that a dune fifteen to twenty feet high had formed
on the inside. But the drift sand baffled in its usual line
of assault was now scouring the clay rampart below, and in many places the under-cut wall looked as if destined to tumble before long. Yet at the same time I found scarcely any traces around An-hsi that wind erosion, with so much of its ready instrument, driven sand, close at hand, had produced those characteristic Yardang trenches of the Lop - nor desert, or lowered the ground level generally as at sites along the southern edge of the Taklamakan. No doubt the cover of vegetation, scanty as it looked in places, was sufficient to afford protection to the soil. That all trees and shrubs about An-hsi had a marked western bend in their growth is a point that scarcely needs emphasizing.
Our stay there was enlivened by a succession of dust-storms. Yet in spite of them the heat continued oppressive. So it was a great relief when during the night of June 22nd brisk showers descended, followed by drizzling rain at intervals during great part of the next day. It was the first real rain I had seen since those trying days of last August high up in the Karanghu-tagh mountains, and it seemed
IIIFカラー高解像度
