loads of fodder (dry lucerne grass) we had brought so far. It was clear that we could not count further on their help. But more disquieting still was the fact that in the course of a ten hours' tramp extending over eleven miles as measured by the cyclometer we had not succeeded in covering more than a little over seven miles of direct distance. The point where we came to a halt that evening, on the top of a broad high dune, could not be appreciably less than a hundred miles in a straight line to the Khotan river where the Mazar-tagh strikes it. Assuming that our camels would be able to maintain this day's rate of progress, about which both Hassan Akhiln and myself felt grave doubt, it would have cost us another fortnight's struggle before we could reach water and grazing. In view of Dr. Hedin's experience farther north and of what Kasim Akhi-ln reported about the sand formations around the Khotan Mazar-tagh, there was little hope of gaining easier ground until we reached that hill range itself. And worst of all, there was no assurance that, however carefully our bearings might be taken, the route followed would actually allow us to sight that conjectured north-western extension of the low hill range which I was anxious to trace in the Taklamakân ; for previous experience had taught me only too well the impossibility of steering an exact course amidst high sands by the compass.l"
So there followed for me a night of anxious consideration. I realized that the effort to force our way across the forbidding obstacles created by Dawàns and dunes could not be persisted in without facing heavy losses in animals and equipment. The risk of serious delay as a result of such losses weighed even more heavily in the scale, while the chance of securing in return fresh topographical observations of interest was far too problematic to balance it. Reluctant however to abandon a cherished plan, .I left my decision till next morning. Ascending then the highest dune near our camp and carefully scanning the horizon eastwards with my glasses I saw nothing but the same expanse of formidable sand ridges, which resembled the huge waves of an angry ocean suddenly arrested in movement (Fig. 9o). Closely packed dunes stretched over and between them without anywhere a patch of eroded ground or easier sand. The deceptive appearance of hills that refraction gave for a brief time to distant sand ridges vanished as the sun rose well above the horizon.
There was a strange allurement in this vista, which suggested Nature suddenly caught in a death torpor, and I found it hard to resist it. We men might have safely struggled through in the end, but only at the cost of sacrificing some or most of our brave camels, the mainstay of our transport for the difficult explorations of the winter, and of hampering our work by loss of time and equipment. To make our way due east to the Khotan river might have meant a reduction of difficulties and distance, as we should then have kept parallel to the Dawâns, at least for a portion of the way. But such a route would have led over ground already seen by Dr. Hedin and could offer no appreciable advantage. So there remained no choice but to turn and reach the Khotan Mazar-tagh by way of the Yârkand river. It was a hard decision to take, and the knowledge that the little band of hardy men with me would have willingly shared what risks and adventures lay ahead did not help to lighten it. But experience proved the wisdom of having bowed to necessity in time ; for the next day but one there sprang up a violent sandstorm, the first of the autumn, most trying by its bitter cold even where fuel was abundant. Its icy blasts continued for days, and if met with amidst the high sands would greatly have impeded our movement and caused serious suffering ; a single camel load of fuel was all we had been able to take along.
In view of the geographical interests indicated, I still feel convinced that an effort should be made to trace the ancient wind-eroded hill range right through the sea of sand that at present separates the Chok-tagh from the Mazar-tagh on the Khotan river. But I now realize more than
15 Cf. e. g. Desert Cathay, i. p. 425, regarding the crossing from the Lou-lan Site to the Tarim.