National Institute of Informatics - Digital Silk Road Project
Digital Archive of Toyo Bunko Rare Books

Sec. viii]   THE LAKE SECTION OF THE LIMES, T. xix-T. xxiii   719

was roughly built of hard salt-permeated clods of clay, such as might still be obtained from the shör-covered marsh edge close by. Layers of Toghrak twigs were inserted to strengthen this coarse masonry, which still rose to a height of some 13 feet. No remains of quarters survived, apart from a small underground room measuring i i by 9 feet, cut into the clay to a depth of 5 feet a little to the east of the tower. From the refuse outside it came the complete inscribed slip T. XXII. a. i. i, Doc., No. 302 (Plate Ix), and the fragment of another, No. 303, was found in a large heap of dung and ashes some 20 yards from the tower.

After skirting the edge of the marsh belt north-eastwards, a stretch of firm ground was reached, covered with erosion terraces from 20 to 3o feet in height and having elsewhere a surface of fine gravel or drift-sand. It was easy to see, as Map No. 78. A. 3 also shows, that this stretch of ground marks the northern end of a well-marked tongue-like plateau, which runs out from the gravel `Sai' in the south. The plateau, together with a corresponding low spur meeting it from the foot of the Kuruk-tagh, forms the natural barrage which holds up the Su-lo Ho waters in the Khara-nor basin. This barrage is about 2 miles across, and in its lowest part northward probably liable to occasional flooding ; for when on May 13, 1907, I searched for and found the actual outflow of the river in the north-west corner of the lake, the deep-cut bed of the former, only 20 yards wide but quite unfordable and carrying a volume of water of at least I,5oo cubic feet per second, was filled almost to the brink. At other seasons a crossing could, no doubt, be effected here without difficulty owing to the firmness of the banks, and this explains why the line of the Limes wall had been carried across the barrage from the west shore of the Khara-nôr to the wide marshy belt on the other side.

The line of the wall, marked by a low straight mound, could be made out only where it crossed bare patches of coarse sand between the towers T. xxii. b and T. xxii. c. Elsewhere it had disappeared completely on finer soil overgrown with scrub. The line, as marked by the towers, had been drawn about i z miles to the south of the Su-lo Ho outflow. T. XXII. b was a badly decayed tower built with bricks on the top of a small terrace about 20 feet high and 3o to 4o feet across. The masonry stood only to a height of about i 3 feet, and was too much broken to allow exact measurement of the plan, which seems to have included small quarters adjoining on the south and west. From the refuse which had been thrown down the cliff southward sixteen inscribed wooden records were recovered. One of these, T. XXII. b. 9, Doc., No. 272 (Plate Ix), is dated A. D. I 2. Another, Doc. No. 274 (Plate Ix), is of interest as it refers to a list of government arms belonging to the Ch'ing-tui company of Ping-wang. The recurrence of the latter name also in T. xxii. b. io, Doc., No. 275, and in T. XXII. c. 22, Doc., No. 271, conclusively proves that the Ping-wang section of the Limes extended as far east as these stations close to the Khara-nôr. Local interest attaches also to the square tablet, Doc. No. 278 (Plate ix), which in big characters

names the ` fire signal of the Ta-wei   kg barrier ', and evidently was intended to be affixed to
a wall. No. 273 describes itself as a notification concerning troop movements, ` to be hung up in the [several] Ping of the barrier ', but does not give their names.

The tower T. XXII. c occupied the highest point of a small isolated ridge of clay rising at a distance of about a furlong from the actual shore of the Khara-nor. An earlier shore-line, 4 or 5 feet higher, approached the position of the watch-station even closer. The ruined tower, built of stamped clay with intervening layers of reeds, was badly decayed, and was only to or i i feet high. No quarters could be traced apart from the remains of a small room partially cut into the clay of the terrace. But at the south foot of the latter, where a small ravine descends the steep cliff, I discovered a considerable accumulation of refuse, about 16 feet in diameter and up to 4 feet in height. From masses of reed-straw and stable refuse there emerged here over two dozen records on wood and a large number of fragmentary fabrics, mainly silk, in great variety

Outflow from Kharanor.

Limes wall between

T. xxii. b, c.

Documents found at T. XXII. b.

Watch- tower

T. xxu. c.