The night was made miserable by a violent north-east gale which nearly blew my tent down and left us all half-benumbed at the start next morning for the tramp which was to bring us at last to the ruins of ` Lou-lan '.
The hardships to which my labourers from Charkhlik were exposed and their consequent weariness rendered it a matter of serious concern for me that we should reach the eastern group of ruins, the main site for the proposed excavations, that day, December 17. I estimated the direct distance still separating us from it at only about eight miles to the north-east, if the position of Camp 123 on our plane-table and that indicated for the ruined site in Dr. Hedin's map were right. But I had not been able to sight any of the ` towers ' which his description mentioned and which, I hoped, might serve us as landmarks. So I considered it safest to continue steering the due northern course which we had followed so far and which, I hoped, would allow us to strike the far-stretched line of ruins somewhere near its centre. The strain of anxious expectancy was lightened for me when, after marching for about three and a half miles over Yârdangs and low dunes, we picked up three Wu-chu coins of the Han type in quick succession. They were a definite proof that we were approaching a site of the historical period. About seven miles from our start we came upon a broad and well-marked bed fringed by rows of dead Toghraks and decayed tamarisk-cones and running west to east with a slight north-easterly bend. Mullah, with his sure topographical memory, at once recognized it as the ancient river-course he had seen in 1901, south of the ruins visited with Dr. Hedin.
Beyond the riverine belt the eye ranged wider across the flat expanse of wind-eroded clay barrenness, with the outermost low range of the Kuruk-tâgh fully in view. The men now pushed on eagerly in the hope of earning the reward in silver which I had promised for the ruin first sighted. We had only gone a mile from the dry river-bed when one of them ahead, having climbed the top of a plateau-like )(di-clang, shouted that he could see a ` P`ao-t'ai '. My glasses showed that the tiny knob, rising far away on the horizon eastwards, was really a ruined mound, manifestly that of a Stûpa. So the course was promptly changed, and the direction now favouring our progress along the Yârdang ridges, the five miles separating us from the mound were covered in two hours. It proved, as expected, the ruin of a Stupa built of sun-dried bricks (see Fig. 97), and was the same near which Dr. Hedin had first camped on his return in 1901. Chinese coins of the Wu-chu and uninscribed types were picked up in numbers around it, and finds of bronze arrow-heads and of other small objects in metal had by now become frequent.
From this point three more mounds were in view, and among them Mullah without hesitation recognized the largest, which lay to the south-east, as marking the main group of ruins. The three miles' march to it led over frightfully eroded ground, and the succession of precipitous clay ridges and sharply-cut trenches between them, down to twenty feet and more in depth, had to be crossed at right angles. Darkness came on just as I reached the foot of the ruined Stûpa which stands out in this weirdly desolate expanse as the landmark of the ruined station of ` Lou-lan '. It was two hours before the convoy of our much-tried camels safely joined us. But as I sat by the big bonfire we had lit to guide them, the fatigues and anxieties of the trying desert march were forgotten, and only elation. at having safely reached the goal in good time remained. I felt grateful, too, for Dr. Hedin's excellent mapping, which, notwithstanding the difference of our route lines and the total absence of guiding features, had enabled me to strike the ruins without a day's loss. When subsequently the results of our own plane-table survey for these parts, checked by astronomical observations and triangulation as far as the mountains south-west of Charchan, came to be computed, I was much gratified to find that Dr. Hedin's position for the site l' differs from ours (see Map
" Cf. Central Asia, vol. iii. Pl. 51.