National Institute of Informatics - Digital Silk Road Project
Digital Archive of Toyo Bunko Rare Books

STEIN had passed that way twice, as had also HÖRNER and CHEN. The short stretch I had covered consisted partly of fine dust in which the camels left deep tracks, and partly of sand and queerly shaped clay terraces. Farther west HÖRNER had found country that he described as »very loose alluvial gravel ». Moreover, this desert contains very little water.

The mountain-route, on the other hand, had certain advantages. We could expect to find hard, firm ground among the low crests and ridges of the Pei-shan, and also water. And the mountain-route was attractive because it would take us into absolutely unknown country, where no European had set foot. To the west of the three trails between Tun-huang and Hami, marked on STIEI,ER'S map, the only known route, was the one I had mapped in February, igoi, when I went north with my camels from Toghraq-quduq into the Pei-shan as far as lat. 41° 29' N, thence turning off west, to Altmish-bulaq. Although thirty-three years had passed since then, I remembered clearly that we had not been held up by sand-dunes, and that the ground had been comparatively hard and good going for the camels.

I freely admit that the old, insidious desiderium terra incognita, that so often before had lured me into reckless adventures, was a factor in my decision in favour of the mountain-route. And as we were searching for a motor-road suitable for traffic, and had well-founded reasons for mistrusting the desert road, it was a clear case. So we did not turn off west for the time being, but held on our course towards the foot of the Pei-shan.

We were soon in among the mountains and following an ancient but clearly defined road, marked by numerous cairns and a watch-tower on a low hill.

At the foot of the mountain on our left were a little stone wall and a watchtower. Shih-pan-ch'üan, »Rocky Spring », at the base of a ridge, was an open basin of clear spring water 2 X 1.5 m in size and 0.7 m deep. We poured away the water in our receptacles and re-filled them with the beautiful new spring water.

The road, still marked with small cairns, led on N. N. W. between low mountains and small isolated peaks. The ground was hard and the gradient quite insignificant. Here and there grew an occasional tussock. We crossed a minor ridge and encamped on a gravel plain 1,50o m above sea-level.

The next day's journey took us northward between low dark ridges. We followed a broad valley with hard ground. The guide was alarmed at our plans, and declared that the next pass was impossible for cars. A cairn on our left marked a spring that carried water after occasional rain, though it was now dry. The »impossible » pass was low and easily negotiable, at an altitude of about 1,77o m. Extensive views opened up on the farther side. We passed through a defile only 10 m wide, coming out into an arena-like valley. Here the thorny tussocks were more numerous. Old cart-tracks indicated that this road, which doubtless led to Hami, was practicable for vehicles. There were cairns everywhere; once we counted

246