| || || || || || || || |
196 REMAINS OF AN ANCIENT DELTA [Chap. VI
appeared to have been of larger size, but had suffered very badly through erosion. One of the two rooms of which the walls of timber and wattle could still be made out measured 4o feet by 28, and the other may have been equally large. Even where the timber wall foundations still retained their position, the ground was swept completely clear by the wind. The force of its erosive action was attested by the fact that the soil to the south of the ruin was scooped out to a depth of 22 feet below the original ground level as marked by the floor of the building. Among the big Toghrak beams and other timber scattered over the slopes, two badly splintered capitals with voluted ends could just be distinguished. There, too, a number of pieces of iron household utensils (L.M. iv. 01-5, Pl. XXIV, XXVI) were picked up, together with a few bronze fragments. The only other structure of the site, L.M. v, situated about 33o yards to the north of L.M. iv, had also, as Fig. 135 shows, undergone far-advanced erosion. Among the much-withered timber pieces some double-bracket capitals and pillar bases of the usual shape were barely recognizable.
The extent of the ground, fully a mile from north-west to south-west, over which the ruined dwellings just described were found scattered, appears to me in itself a sufficient indication that the settlement to which they belonged was one of some local importance. It was only by reason of the uniformly massive construction of their timber framework that their remains had survived the destructive effects of wind-erosion. Just as at the Niya Site and at other ruined settlements explored in the south of the Taklamakâ n, it may be assumed that the extant ruins are those only of the more substantial buildings, while the mud-built dwellings serving the needs of the great majority of the inhabitants have been completely effaced by the same powerful agent. This conclusion is fully supported by the great number of small miscellaneous objects of the ` Tati ' type that could be picked up at the site practically within a single day and without such systematic search as a longer stay would have permitted. Among such finds six Chinese coins may be mentioned in the first place. They are all of the inscribed and large Wu-chu type, which belongs to Han times and the immediately succeeding period. The numismatic evidence agrees therefore with the evidence furnished by the manuscript finds in Kharosthi and Sogdian, in its bearing on the chronology of this Site.
To the period of occupation contemporary with that of the Lou-lan station may also be ascribed the very numerous fragments of glass vessels and glass beads (L.M. 025, 027, 071—IOI, 0134 ; L.K.—L.M. 01-13) in a variety of colours or gilded, some cut or showing raised patterns (Pl. XXIV). The spout in fine blue glass, L.M. 070 (Pl. XXIV), deserves special mention. The fragment of glass slag, L.M. 034, might suggest local manufacture. Among ceramic remains, the fragments of green-glazed pottery, L.M. 042, oi22, and of faceted grey stoneware, L.M. 054, may claim interest. Remains of small objects in paste, L.M. 063-9, 0133, are also represented. The numerous bronze relics include the finely modelled anthemion ornament, L.M. 0119 (Pl. XXIV) ; a cat-bell', L.M. 0131 (Pl. XXIV), similar to that found at the L.A. station ; the bezel of a jewel, L.M. 0129 (Pl. XXIV) ; the neatly made ear-pick, L.M. 0150 (Pl. XXVI), &c. Iron implements are also represented, L.M. 051, 0145-7 (Pl. XXIV, XXVI). By the side of these remains manifestly belonging to the period of early intercourse with China and the West we find, just as at other Lou-lan sites, far older stages of civilization illustrated by miscellaneous relics of the stone age which wind-erosion has laid bare. The well-made stone arrow-heads, L. M. oio, oiS5, are probably neolithic, while the numerous `knife-blades', L.M. 012-24, 0156, and jade celts, L.M. 04, 043, 055 (Pl. XXII), manifestly date back to an earlier epoch. The stone cores, L.L.—L.M. 1. of (Pl. XXII), L.M. 06-9, point to local manufacture.
The presence of stone age remains at and around L.M. must obviously be considered in connexion with the corresponding finds made on the ground extending south-eastwards to L.K. as well as with those which were so plentiful on our march 0f December 15, 1906, to the north of