| || |
532 ACROSS THE PEI-SHAN TO BARKUL [Chap. XV
be the way across the range that we had so eagerly desired. It was a hope which proved well founded in the end. But its fulfilment was delayed. When we had covered some eighteen miles and found ourselves at what seemed the very head of the valley, our ` guide ', so confident before, declared that he had mistaken his course and that the true route had been left behind in a side valley leading northward near the point where we first had found water. Uncertain whether we should find water in the highest portion of the valley still before us or beyond it, we felt obliged to turn back for that night to the spring that we had passed near the rough shelters.
At this Camp 2 r 7 I decided to sacrifice a day to tracing, if possible, the regular caravan track which our ` guide ' felt quite sure we could strike again by moving a short distance south-eastwards. There is no need to describe here at length the details of the Odyssey that ensued, enlivened as it was by a succession of incidents caused by the queer and in the end almost frantic behaviour of our poor ` ta-lu-ti'. It will be enough to record that after first moving down the valley and then searching a plateau to the east, intersected by branches of the same drainage bed, we found indeed abundance of grazing for camels and marks of its having been visited, but no trace of a route leading across the range. Our hapless ` guide ', now inflamed with a kind of desperate obstinacy, continued to drag us south-eastwards for another seven miles across a much-broken plateau, previously hidden by a low spur. Then the position reached relatively to the alleged Ta-hsi-k`ou' made it quite certain that no track leading from the latter point northward could possibly have escaped us. Satisfied in this respect I turned back to the valley where we had been lucky enough to find water, glad to let our ponies enjoy the fine grass that it offered near Camp 218.
The fate of our ` guide ' now gave rise to anxiety. In his vain search for his la-lu, he had continued to rush ahead, and the two men sent after him had failed to catch him up before dark. However, after daybreak he rejoined us, with the sullen despair of a man who knew that the line I decided to follow north-westwards was bound to take us into impossible ground and end in our destruction in waterless ` Gobi '. A variety of considerations led me subsequently to conclude that the ` high road ' which the unfortunate man so desperately strove to bring us back to lay indeed much farther to the east. But we had probably lost it already at the spring of Chin-êrh-ch`üan. One of my Turki followers subsequently reported that while looking after the ponies there, he had noticed a broad track leading off to the north. It must be left for some future traveller to ascertain whether this information gives the clue to the true line followed by the caravan route towards Bai.
On September 25th we once more ascended the valley to its head. Before our start I took care to make sure that our two iron water-tanks were properly filled, and in addition to this regular provision, water was also taken in the available goatskins. As we again approached the head of the valley, I noticed on our right a big wall of almost vertical cliffs, over 8o feet high. The rocks, red in colour and suggesting coarse sandstone, showed clear stratification and a strike from east to west. After having covered nine miles from Camp 218 we reached a flat saddle at a height of a little under 6,000 feet, offering an open view towards the Karlik-tâgh proper and the southern slopes of its eastern continuation. A cap of clouds was hiding the snow peaks from view. The divide on the saddle lay between the valley we had followed and another draining south-westwards. So my relief was great when the track, such as it was, turned to the north-west, keeping along a succession of small plateaus separated by low rocky ridges. It was skirting the steep slopes of a massif that rose, on the east, to over 9,200 feet in height. Here, too, the stone seemed the same coarse red sandstone striking west to east. Then, after having thus covered some three miles, we reached what manifestly was the true watershed on the axis of the range at an elevation of 6,000 feet (Map No. 37. c. 3). When the valley that we entered on the other side was found to descend north-north-west, hope rose high that it would offer us the desired access to Bai.