746 EXPLORATIONS IN THE KURUK-TAGH [Chap. XX
dying Tarim, and that the goal he had steered for was near. The exhausted condition of one of his camels necessitated short marches. But moving over ground recognizable from the mapping here done by us in 1906 and 1913 he finally brought his little party in safety to the northern edge of the Chainut-köl on the morning of March 1st.
Regard for the condition of the camels, all severely tried by the preceding marches and their long fast, obliged Afraz-gul to halt at Chainut-köl for the next four days. He found the ice in the small pond near which our Camp lxxxix of February 3rd, 1914, had stood,ls practically all melted. But the large lake basin south of it, then completely dry, was being rapidly filled with fresh water from the newly arrived spring flood of the Tarim. On March 3rd and 4th he carefully surveyed the actual shores of the Chainut-köl (Map No. 3o. c. 1). The network of channels in which water was pouring into the lake could not be crossed. But in the main channel the surveyor measured a volume of not less than 700 cubic feet .per second. On the following day he proceeded to SW., and following those channels upwards ascertained that water overflowing from them had already filled the large lagoon of Yaghizmak-köl which had been found dry and crossed by us in February, 1914•
To the NE. of Chainut-köl, too, depressions then dry were being rapidly reached by flood water spreading in narrow channels, as Afraz-gul found when he started on March 6th for the desert crossing which was to bring him to our rendezvous at Yardang-bulak. In accordance with his instructions he proceeded by our former route to Camp xc and thence struck due north for the site of L.M. explored the year before. On the way to the latter big ridges of drift-sand up to ioo feet in height had to be crossed, as Map No. 29. c. 4 shows. Among relics in stone, pottery, and bronze picked up between L.L. and L.M. and described in the List below, the large jade celt C. ccxlv. a. 02 (Pl. XXII) deserves special mention.
When dealing in Chapter vi with the remains explored at L.M., it has already been recorded that the extensive search which Afraz-gul thence made on March 8th to the east and north-east did not lead to the discovery of more ruins. But when he resumed his march in the originally indicated direction to the NW. he came, after covering two miles, upon the remains of three detached dwellings occupying the top of wind-eroded terraces and marked by L.R. on the map. They were built like those of the I,.M. site of Toghrak timber and wattle. Of the one marked i only timber debris survived on the top, about 15 feet in diameter, of a terrace 12 feet high. No finds were made here. Some ioo yards to the NE. of this there survived portions of the walls of another structure, ii, on a terrace c. 6 feet high. One room traceable to the east apparently measured 21 by 27 feet, and held much sand. Another smaller one adjoining westwards was completely filled by it. In a third structure, iii, found about 200 yards to the NE. and on a similar wind-eroded mound, the surveyor was able to distinguish 5 to 6 rooms within a total area of about 55 by 51 feet. Here, too, the sand lay too high, on the west up to 10 feet, to permit of any serious clearing being attempted by the surveyor and his few companions ; only very limited time could safely be spared in view of the difficult desert journey still before them. The clearing of a refuse deposit outside ii yielded no finds. So Afraz-gul had to content himself with the small objects in bronze, iron, stone, and glass which were picked up on the eroded slopes of L.R. ii and iii. They have been described in the List following the account rendered above of L.M.16 They suffice to prove that the ruined dwellings were occupied about the same period as those of the sites L.K.—L.M. to the south.
On March 9th resumed progress to the NW. brought Afraz-gul within a mile from L.R. to a well-marked river-bed about ioo yards wide with a winding course apparently from the NE. Much of the bed, 5o feet deep in places, was overrun by big dunes. This old bed was last seen