国立情報学研究所 - ディジタル・シルクロード・プロジェクト
『東洋文庫所蔵』貴重書デジタルアーカイブ

406   FROM SU-CHOU TO THE LIMES OF MAO-MEI   [Chap. XII

pletely dry, as were the heads of six canals which take off at this point. Three other canals, which we crossed a mile beyond on scrub-covered ground,2 together carried only a volume of some sixty cubic feet of water per second. This clearly showed how scanty is the supply of surface water left for cultivation below Su-chou before the summer flood from the melting snows of the Central Nan-shan comes down towards the middle of June.

Our march of May 12th to the north of the pleasant little town of Chin-tea (Map No. 42. B. 4), surrounded by shady orchards and farms, took us over ground which I had already seen in September, 1907, on a rapid reconnaissance.3 The avenues of fine elms leading first through rich cultivation and then through grass lands came to an end after we had proceeded about seven miles. Beyond this, cultivation was restricted to patches of varying size, separated by stretches of scrub-covered sandy waste. My impression was that since 1907 a good deal more of this waste, which had once probably been tilled, had been reclaimed by the industrious settlers of the scattered hamlets, and in places it was possible to observe this process actually extending. Yet it will evidently be long before the effects of the devastation wrought by the Tungan rebellion in this region is completely effaced. As we were nearing the large watch-tower of T`ou-tun towards the end of the day's march, the strange intermingling of green fields, tamarisk-cones and low scrub, and reed-covered dunes vividly recalled to my mind the scenes to the north and north-east of the Domoko oasis between Chira and Keriya, where areas abandoned at different periods were again being brought under cultivation.4 There these periodic changes had proved to be connected mainly with variations of physical conditions affecting the supply, from subterranean drainage, of water for irrigation. Here there could be little doubt that the change I saw proceeding resulted solely from economic recovery following a great political cataclysm of recent occurrence.

During the day's march a piece of very welcome, if vague, information was secured by our Mongol interpreter, gruff old ` M alum '. An aged Chinese villager whom he met on the way had told him that when going once some thirty li' to the north-east of T`ou-tun, he had come upon remains of a chiu ch`iang or old wall ' built with brushwood at the foot of the low hills running above the Pei-ta-ho valley. The information, though it probably related to the Han Limes, sounded vague, and further inquiries at T`ou-tun resulted only in assertions of complete ignorance. In order not to lose time over an uncertain search, I accordingly let the baggage with Surveyor Muhammad Yaqûb proceed by the cart-track to Mao-mei, while I myself set out with the rest of the mounted men towards the foot of the hills overlooking the wide riverine plain. After passing a dry river-bed about a mile and a half from camp we skirted an area where new fields had been opened amidst tamarisk-cones. There were signs that the jungle here was undergoing reconquest, having in places known cultivation before the Tungan devastation.

Two miles farther we passed a second dry bed about sixty yards wide, and after crossing a clayey belt covered with shôr and scanty scrub reached the foot of the outermost Pei-shan chain.

Low decayed ridges of rock striking from ESE. to WNW. rose here to heights of 40 or So feet above the gentle slope of detritus. From a plane-table fixing made here a watch-tower was visible

far away on the top of a hill northward, but I looked in vain for signs of the Limes. Yet scarcely had we moved a hundred yards or so to the east when a curiously straight double line, which had looked like the outcrop of a decayed rocky ledge, revealed itself clearly as marking the direction of the almost wholly effaced Limes agger. As we proceeded along it, the mound gradually rose to nine feet in height and its mode of construction became evident. Rough slabs of stone had been

2 The small detached area to the south-west of Chin   3 See Serindia, iii. p. 1134.

t'a-ch`êng (Maps Nos. 42. B. 4 ; 43. B. I) ought to have been   4 See ibid., i. pp. 202 sqq.

printed in yellow instead of green.

Chin-t'a oasis revisited.

Information about ` old wall '.

Agger of Han Limes.