| || || || || || || || || || || |
284 THE SEARCH FOR THE ANCIENT CHINESE HIGH ROAD [Chap. VIII
surprise to me when subsequently following that ancient route beyond Camp C. ci I encountered a belt of exactly corresponding salt-coated Yardangs and, as we shall see farther on, just where conclusive antiquarian evidence allows us safely to identify them with the ` White Dragon Mounds of the Chinese itinerary.6
The Yardangs gradually decreased in height, and as they became more and more clear of salt-incrustation resumed their steep wind-carved slopes. It was significant that where this change had become marked, at a point about fourteen and a half miles' marching distance, we met with the first mollusc (Limnaea) shells and, as a further proof of fresh water having once been present, with dead tamarisk wood on low Yardangs. The furrows or trenches between these now showed coarse sand at the bottom. We had already passed isolated Mesas and sighted others, at intervals,
to the left and right of our line of march. But now we were brought up by an extensive Mesa ridge rising like a wall to over fifty feet, which forced us to make a considerable detour to the east before
we could find a passage practicable for the camels. Beyond it we again crossed ground covered for a mile and a half with Yardangs only ten to fifteen feet in height and were able at one point to collect a little fuel from the top of one of them ; a badly eroded dead tamarisk-cone was also in sight at this point. Then again a regular line of Mesas up to sixty feet in height was encountered. A breach in their wall-like array allowed us to get the camels through without too great trouble. But they had all been severely tried by the ground that we had met since leaving the edge of the Sai, and regard for their tired condition and the approach of darkness obliged us to pitch camp after having covered altogether about seventeen miles. Scanty debris of dead tamarisks allowed us to economize fuel.
The general impression left upon my mind by the day's march was that during the period when the easternmost portion of the Kuruk-daryâ. delta was still receiving some water and was habitable for semi-nomadic herdsmen and the like, a belt of utterly barren ground, wind-eroded and partly salt-encrusted, separated the foot of the gravel Sai from the riverine belt of vegetation. No vestige of the presence of man had been met with anywhere on this ground. Having regard to the position of the previously explored remains at L.E. and L.F., it was clear that the ancient route would have to be searched for farther south.
On the morning of February 26th we rose in the dark and managed to get the camels loaded and started by daybreak. A belt of Mesas sixty to eighty feet high was crossed to the south within less than a mile from Camp C. xcix. It presented clay terraces, fantastically eroded and 'often curiously suggesting ruined strongholds, watch-towers, or Stûpas (Fig. 177). The biting wind, which on the previous day had for once been blowing from the south-west, had died away during the night, and the atmosphere had in consequence become clearer, So, after clambering to the top of a towering Mesa, we were soon able to recognize far away to the south the long isolated ridge bearing the remains of L. F. I was thus assured of having steered correctly towards the eastern extremity of the previously surveyed area. The ground that we entered beyond the belt of Mesas was easy, furrowed by Yârdangs only four to ten feet high. But after about two miles' progress the soil turned to hard salt-impregnated clay, with a billowy surface. This caused fresh trouble to the camels, most of them already footsore from the previous day's march. Just before we came upon this ground dead tamarisks were found in patches, and near one of them we picked up a Wu-chu coin, evidence of the passage of man.
Having covered about four and a half miles from Camp xcix, we found ourselves about half-way between the Mesa L.Q. to the west, on which Afrâz-gul's first reconnaissance had shown the presence of some graves, and the Mesa L.I. to the east, which he had visited on February i6th
6 See below, pp. 297 sq., 309 sqq.