27o REMAINS OF ANCIENT LOU-LAN [Chap. VII
camels for any service we might neéd and ready to set out at once. He had duly received the order from Kara-kum, the district head-quarters, some three months before, and fortunately nothing was subsequently done to revoke it ; yet the prohibition against us issued from Urumchi had reached Kara-kum, as I learned a year later. Whether the failure to act on these instructions was due to mere negligence, or to the benevolent closing of official eyes on the part of the Tungan district magistrate, luckily an acquaintance of Sir George Macartney and a well-meaning person, may be left an open question.
The local knowledge and life-long desert experience of Abdurrahim proved from the start of no less help than the excellent transport provided by his five magnificent camels, bred and reared in the Kuruk-tagh. He had more than once visited Altmish-bulak on his hunting expeditions after wild camels, besides the occasion in 1901 when he guided Dr. Hedin there. Thus Lal Singh had been able without serious difficulty to carry out the main task for which I had detached him, by mapping the course of the ancient dried-up river from Ying-p`an, on the Tikenlik-Singer route (Map No. 25. c. 3), to the vicinity of the Lou-lan sites. While carrying out this survey, he covered, as recorded on Maps No. 25. D. 3 ; 29. A—D. 3, a marching distance of close on a hundred and fifty miles, all through waterless desert except at one point, Yaka-Yardang-bulak, and was able to determine the approximate position of the head of the ancient Kuruk-darya delta near his Camp 78 (Map No. 29. B, C. 3).
In accordance with my instructions, he kept a careful look-out for any ancient remains and thus succeeded in tracing four cemetery sites on the edge of the gravel Sai overlooking the ancient riverine belt. He also came upon two small ` Tatis ' in the latter (Map No. 29. c. 3), showing coarse pottery debris and probably marking indigenous occupation, perhaps partly in prehistoric times.2 Three out of the four cemetery sites which Lal Singh's survey indicated I was subsequently able to visit and explore.3 The route followed by him had brought the little party to within about seventeen miles of the Lou-lan station L.A. But uncertain as to whether the determination of his position on the plane-table, after a circuit of some four hundred miles from Miran, was sufficiently exact to permit of his steering correctly towards a particular point in the wind-eroded desert which neither he nor Abdurrahim had visited before, he decided to turn off to the north-east and seek the spring of Astin-bulak before venturing farther. He had been there only a day when Tokhta Akhûn and the men in charge of my camels arrived from the south, and contact was duly established. Neither party was prepared for the meeting, and in consequence one of Abdurrahim's camels, while straying a little from the salt spring, narrowly escaped being shot by Tokhta Akhûn for a wild one.
The greater portion of the ground seen by Lal Singh on his seven marches from Ying-p`an to Astin-bulak and again between this and L.A. had been visited already by Dr. Hedin in I9oo—I,4 and the latter's detailed record of it made me feel less regret at the surveyor's inability to give
2 At and near these ` Tati ' areas Lal Singh's party picked up most of the miscellaneous small objects—mainly potsherds, but including also two glass beads and two stone ` blades '—which are shown in the Descriptive List below under the ` site-mark ' Lal S.
The well-made bronze arrow-head, Lal S. 015 (Pl. XXIII), is of interest as it closely resembles others of the same type found to the NW. of L.A. ; see C. xcvi. 013, 016 and one from the Niya Site (see Ser. iv. Pl. XXIX, N. xiv. 008) which approximately date from the first centuries of our era.
3 See below, pp. 276 sq. ; Chap. xx. sec. iii.
4 Dr. Hedin's route of March 1900 had lain, as appears from the sheets recording his prismatic compass survey from ` Yangi-köl to the Lop Desert ', approximately over the same line as that of Lai Singh as far as the latter's Camp 75 (Map No. 29. A. 3) ; then again between Yaka-Yârdang-bulak, C. 76 and C. 78 (Nogustu ; Map No. 29. c. 3). From the last-named point his route towards Altmish-bulak followed a somewhat more northerly line, away from the Kurukdaryâ. His routes from Altmish-bulak in 1900 and 1901 to the Lou-lan site do not diverge much from that followed by Lal Singh between his Camps 81 and 83, L.A. (Map No. 29. D. 3).