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

Sec. v]   SURVEY OF LIMES LINE TOWARDS AN-HSI, T. xxxi-xxxv   605

T. xxxII, did not stand to the same height, but the bastion-like projection of the wall in front of it was particularly well marked here. The distance of the wall northward was twenty-six feet, to east and west fifty-six and thirty-three feet respectively. A rubbish deposit, probably from some quarters once àdjoining the tower but now completely eroded, only yielded a few wooden relics, among them a rough pen, T. xxxii. 002, a hemp string, and the like. The bronze arrow-head T. xxxii. 009 was picked up on the surface at the north foot of the tower.

Close on the south of the tower T. xxxiii was an enclosure about thirty-five feet square, built with lumps of clay and layers of tamarisk brushwood. Though salt-impregnation had made its wall almost as hard as concrete, erosion had reduced its height to a maximum of about two feet, and in places had completely effaced •it. Apart from the droppings of horses nothing was found within or near the enclosure. T. xxxiv proved to be a tower much injured by wind-erosion, which had reduced its solid clay masonry to about sixteen feet from north to south and about twelve feet across. The base had been undercut by erosion, and a big fissure ran down through the clay, which nevertheless still rose to a height of seventeen feet. A low heap of débris about twenty feet to the east contained the foundation of a clay wall about fifteen inches thick, together with broken bricks and bundles of reeds, possibly from a roofing. Here was found the well-preserved wooden label, No. 705 (Doc., Plate xIx), which apparently mentions the silk string for a particular type of crossbow as belonging to a certain company raised, as its name shows, under the Han dynasty. Besides some other small wooden relics this heap yielded the large fire-stick (` female'), T. xxxiv. ooi (Plate LII), and the elaborately cut block, T. xxxiv. 003 (Plate LIV),. the use of which has not been determined. In the débris, and about two feet above the natural soil, there turned up a copper coin of the Wu-chu type attributed to the first-second century A. D. Of T. xxxv nothing remained but a low mound which only contained disintegrated red clay and ashes, evidence of some structure destroyed by fire. That this had been a watch-tower is made practically certain by the bastion-like semi-lunar projection which the line of the wall makes north of it, just as in the case of the towers previously examined.

Half a mile to the east of this last tower the wall became lost amidst dunes of drift-sand that rose up to fifteen feet in height and were evidently the offshoot of a sand belt encroaching from the south. Further on, it again emerged in patches. After having thus tracked it for about two miles from T. xxxv, I reached a broad belt of gravel, fringed by dunes both on the north and on the south, and found there a remarkably preserved stretch of wall, quite unbroken for 256 yards and rising in places to a height of fully seven feet (Fig. r 58). It seems probable that its preservation in this fair state had once been aided largely by the protection of a high cover of sand, though now the drift heaped up against it lay only three to five feet high above its foundation. The uniform distribution of the sand on either side showed that the direction of the wall, east-north-east to west-south-west as before, was also that of the strongest among the prevailing winds. In fact, without this direction the wall could not have survived at all on the bare, level ground fully exposed to the erosive action of the winds.

In the centre of this stretch the wall had a remarkably solid appearance. Its sides showed scarcely any trace of erosion, except that they had lost the revetment of fascines laid horizontally in the direction of the wall which they are likely once to have possessed. Otherwise the particular method of construction could be studied with ease. As shown quite clearly by Fig. 157, layers of fascines, about six inches thick, made up of mixed tamarisk twigs and reeds, alternated with strata, three to four inches thick, of coarse clay and gravel from the soil on the spot. Where the photograph of Fig. 157 was taken, I counted eight double layers of fascines and stamped clay, making up a total height of a little over seven feet. I noticed that, while the fascines were mainly made up of tamarisk

Ruins of T. xxxIIlxxxv.

Well-preserved stretch of wall.

Solid construction of wall.