THE KURUK-TAGH AND JING-PEN. 41
in question from two different directions. As the reader will readily see, the two descriptions agree even to the minutest details. In both cases there is a circular wall of almost exactly the same dimensions, pierced by four gates facing the four quarters of the compass. North-east of each circular wall there exists a little burial-ground, with ruins of a tower and of walls, backed on the north by a terrace, the extreme outlying portion of the detritus scree. On the south there exists in both cases alike a kamisch marsh, with saline water, and poplars growing on its margins; and in both cases there is, again, a Chinese station-house on the north side. Finally, in both cases the marsh is entered by erosion 'gullies coming from the Budschentubulak.
Nevertheless upon comparing my materials with those of Kosloff, I was tempted to think that the two places I visited were not identical. In so far as the natural characteristics of the surface are concerned, it ought to occasion no surprise to find there is such a close resemblance between the two, for the old river-bed extends east from the Kontsche-darja, and everything seems to indicate that for a considerable part of its course, at all events from Jing-pen onwards, it exhibits the same character, being bordered by the same poplar-planted terraces, and having the same kamisch-grown marsh in its bottom — a marsh that is fed by springs which bubble up in what is probably the wide-reaching subterranean delta of the Budschentu-bulak. Nor should it occasion any surprise that the terraced edge of the detritus scree advances in both cases all the way to the river-bed; for in the sequel we shall find that it extends an incredibly long way east beside the Kuruk-darja. It was however a curious coincidence to find in both cases the same evidences, both new and old, of human agency. But, assuming that Kosloff's map was put together with the necessary degree of accuracy, it is easy from a comparison of his map with mine to ascertain that his Empen and my Ju-jing-pen lie at a considerable distance apart. From Ju-jing-pen it is barely 20 km. in a direction S. 35° W. to Turfan-karaul; Kosloff however has 40 '/2 km. in a direction S. 60° W. to the same point. Hence both Kosloff's map in Lop-nor and mine in Petermanns Mitteilungen seem to be in agreement in making the distance between Ju-jing-pen and Jing-pen to be 22 km. That is to say, we seemed to have crossed the »wedge»of desert along a different route. Where I crossed it, the belt of sand was only 2, or at the most 3, km. broad; where Kosloff crossed it, it was ro versts. The distance between the two stations, 22 km., is apparenity, it is true, but a short day's march, and it would seem to be unnecessary to put up station-houses so close together; but the reason for doing so may have been, that just here alongside the old river-bed there existed fuel, grazing, and water in greater abundance than at any point farther on in the mountainous region of the Kuruk-tagh. Thus I almost felt obliged to believe in the existence of two routes between Turfan-karaul and Jing-pen: one along the northern bank of the old river-bed viâ the station-house of Ju jing-pen, the other diagonally across the desert. The latter would be the one that was used by Kosloff, and has been entered by Dr. Hassenstein on my map in Petermann Mitteilungen to the east of the route I followed in 1896. In fact Abdu Rehim told me, that along a more easterly route there are nischans, or »sign-posts», as well as three gumbes, for the purpose of indicating the road during the sandstorms.
Hedin, ,journey in Central Asia. II. 6
IIIFカラー高解像度
