SECTION I.-FROM TURFAN TO SINGER
ON February 16th, to my great satisfaction, I was at last free to leave my suburban quarters in the house of the obliging Russian Ak-sakâl of Turfân in order to cross the Kuruk-tâgh into the Lop basin. After my return from Yâr-khoto I had been able to settle with the attentive and now almost apologetic district magistrate upon a suitable diplomatic reply to the injunctions received through him from head-quarters. It was calculated to leave my hands free in the matter of excavations elsewhere and to offer no excuse for interference with the convoy of antiquities now on their way to Kâshgar. We parted on terms of mutual consideration.
My immediate programme was to move straight to Singer, the only permanently inhabited spot in that whole vast desert region of the Kuruk-tâgh, and thence, after picking up a guide in the person of Abdurrahim's youngest brother, to visit two localities, P`o-ch`êng-tzi'i and Shindi, where the former had told me that remains of old occupation were to be found. With these visits I wished to combine a plane-table survey of ground that had remained outside the explorations carried out by L5.1 Singh, both in 1907 and on this expedition. Then a descent past the salt spring of Yârdang-bulak was to bring me to two cemetery sites near the Kuruk-daryâ that Lâ.l Singh had noticed on his march a year before from Tikenlik to Lou-lan, and to a portion of the course of the ` Dry River ' that had been left by him unsurveyed on that occasion. Thence the Ying-p`an site, near the point where the Kuruk-daryâ bed branches off from the present course of the Konchedaryâ, could be gained with a view to eventual excavations.
My journey to Singer in the central portion of the western Kuruk-tâgh had to be made along the most direct of the three routes that connect it with the Turfân basin. All three had already been followed by Lal Singh ; therefore I naturally chose the shortest, leading due south from Turfân town across the deepest portion of the depression (Map No. 28. c. 3). The first two marches were short ; but as they lay across the lowest belt of K5.1éz-irrigated oases and then past the western extremity of the terminal salt marsh of the basin, they afforded opportunities for interesting observations on its physical geography. A record of them, however, must be left for the paper previously referred to.1 I must confine myself here to recording that the ground on either side of the several wide ice sheets in which the river of Toksun was then pushing its terminal course towards the salt marsh known as Aidin-köl, looked, with its bulging cakes of cracked salt crust and patches of soft shör, as if marking a stage in the formation of such a bed of hard salt as extends over the now dried-up area of the ancient Lop Sea. A rapid reconnaissance subsequently made from our camp at Béjân-tura towards the westernmost portion of the marsh showed ground which closely resembled the salt bog crossed by us ten months earlier on the way to the line of the Limes north-east of Tun-huang.2 Ice brought from the terminal branches of the Toksun river saved us from having to
1 See above, i. p. 567. 2 Cf. above, i. p. 362 sq.