Ruined enclosure at Lai-su.
Position of Lai-su post.
798 FROM KORLA TO KUCHA [Chap. X
well-tilled tract of Chompak, which boasts a Bâzâr of its own. Then, having covered about four miles more across a bare Sai of stone and gravel, I left the cart-road near the edge of the small scrub-covered river-bed which receives occasional flood water from the small stream of Lai-su and is called by the same name, and turned to the south-west to visit the ruins known as Lai-su-iura.
They comprise, as seen in the sketch-plan (Pl. 39), a massive watch-tower and by its side remains of a walled enclosure, both of manifest antiquity and situated about half a mile to the south of the road, on ground furrowed by some small shallow flood channels. The watch-tower K. III is remarkably well built of solid masonry, and shows constructive features which clearly suggest that it belongs to the same period as the watch-towers on the Ying-p`an—Korla route. It measures 48 feet square at its foot, where it has a plinth or base about 2 feet wide projecting up to about 8 feet from the ground. The tower still rises to a height of 47 feet, and its top could be gained only with difficulty. The bricks measure i 5" x 8" x 3" and are thus identical in size with those used in the towers above Ying-p`an. Many of them bear marks of imperfect burning. The masonry shows thin intervening layers of reeds and is reinforced by a framework of massive posts and rafters exactly as observed in the towers Y. iii and Y. iv. A peculiar feature is the presence of small triangular holes on the southern and western faces, at irregular intervals ; they can have served neither for loop-holes nor for the insertion of beams, and their purpose is puzzling.
At a distance of about II o yards to the north, lies the south-eastern corner of a much-decayed rectangular enclosure of which the side facing east appears to have measured about 192 yards. On the northern face the enclosing wall is traceable only for about 38 feet and on the southern for about 118 feet. But from the position of a tower now detached which seems to have occupied the southwestern corner it appears likely that the original shape of the small fort was square. The enclosing wall on the west has completely disappeared, evidently through the erosive action of a flood channel which passes the ruin on that side. Others have cut through the walls on the north and south. The wall, about 10 feet thick, was built of bricks of the same size as the watch-tower, but does not now rise anywhere more than 4 or 5 feet above the bare clay of the ground. That wind-erosion has helped here to accelerate destruction is shown by small Yardangs, only 1 to 2 feet in height, which it has carved out over a portion of the enclosed area. The tower, which, as previously stated, probably occupied the inner south-western corner, is built with stamped clay and measures 39 feet square at its foot. It rises to a height of about 26 feet and may have been occupied on its top by a small guard-room. Remains of a brick wall built against its northern foot, and of a fire-place near this, prove that quarters had once adjoined the tower. Stable refuse and remains of fuel lay in heaps against the foot of the tower ; some beads and small bronze fragments were also picked up here.
The position of the ruins and such details of construction as can still be made out strongly support the belief that these remains belong to a fortified post erected at à point which in Han times served as a stage on the road to Kuchà. The present road from Bugur to Yaka-ârik, on the eastern border of the cultivated area of Kuchà, lies in a practically straight line along the foot of the glacis that descends from the outermost hills ; nor could a more direct route have been followed in ancient times between these two points. The conclusion thus arrived at as to the early date of the Lai-su ruins was fully borne out by the other remains traced farther on along the present road line. That water might still be found at this point is rendered probable by the fact that I noticed marks of abandoned cultivation on a small patch within half a mile of K. 111, and that some fields farther up, near the cart-road which follows a more northerly line to avoid a deep-cut ` Yâr ', were stated to have been actually cultivated for the last few years.
Returning to the road and proceeding along it over bare wind-eroded clay for about two