| || || || || || || || |
400 THE LOU-LAN SITE [Chap. XI
scouring by -wind-driven sand had here done its work with greater uniformity. There were not the same close hachures of Yardangs as at L.A., and in places the hollows showed accumulations of drift-sand. A quarter of a mile to the east an old river-bed, winding between lines of dead Toghraks and tamarisk-cones, could clearly be traced, and about half a mile to the south I carne upon it again. There it measured over i 70 yards across, and, with its bottom only about six feet below the foot of the dead tamarisk-cones, seemed distinctly shallow.
The other ruined building, L.B. v, about thirty yards to the east, had, by the almost complete erosion of the ground it once occupied, been reduced to a heap of much-splintered timber, as seen in Fig. 112, with only one of the massive foundation beams still in situ. Some portion of the larger building, L.B. iv, had also suffered badly through the same cause, especially on the north-east side, as shown by the heavy foundation beams and other timber débris which are seen strewing the eroded slopes on the right and in the centre of the panoramic view reproduced in Fig. 74. Fortunately other parts of the terrace occupied by the large dwelling-house—for as such I easily recognized it at my first inspection—had not been much affected by erosion, even though the covering layer of sand was, nowhere more than three feet high. As excavation proceeded, I soon realized that the protection enjoyed by this piece of ground was due mainly to the thick and well-consolidated layer of sheep-dung which had accumulated within and around the rooms of the ancient residence. Evidently, for a considerable number of years after its abandonment by the last proper occupants, the building had served as a place of shelter for the flocks of shepherds. I observed this also at certain of the southernmost ruins of the Niya Site,2 and later shall have to mention the same thing as seen at Mirân.3
Excavation was begun with the small room L.B. iv. i at the south-west corner of the building. Its walls, like those of the rest of the building, were constructed of horizontal reed wattle, fixed to a framework of posts of the usual type. A wooden partition, adjoining a boarded sitting platform, projected from the south-west wall across the greater part of the width of the room and left only a narrow passage, about 3' 6" across, giving access on the south-east to the adjoining hall iv. The small apartment may well have served as an anteroom for attendants. Sheep-dung had accumulated in it above the height of the sitting platform, about one and a half feet from the floor level. Here, close to the surface, several well-carved fragments of open-work wooden panels, L.B. iv. i. ooi, 002 (Plate xxxIv), were found showing a pattern of wheels strung on interlacing bands. Another fragment, LB. iv. i. 005 (Plate xxxiv), exhibits a gracefully designed branch with leaves and berries. As pieces from the same open-work panels were afterwards found in rooms vii and viii at the opposite end of the ruin (L.B. iv. vii. ooi, 003, 004 ; viii. 00i ; see Plate xxxiv), we must assume that the position of some of these fine wood-carvings was changed after the panels had been broken. Whether this dispersal was caused in ancient times, or possibly by the burrowing of Dr. Hediri s men, I have no means to decide. But what is certain is that the floral ornament shows in all of them, as in other fine wood-carvings found in this ruin, a very strong classical influence.
Among the other finds made in this small room were a rectangular under-tablet and a covering-tablet, L.B. iv. i. 6, 7 (Plate XxxviIi), both excellently preserved and showing their neat Kharosthi script in perfect clearness. It is worth noticing that these two tablets seem to be made of the wood of the cultivated poplar, whereas all the Kharosthi documents found at L.A. are of Toghrak wood, from which we may infer that the latter material was used locally for stationery. Both tablets were discovered at the foot of the platform, and had been protected there by the thick layer of dung. The architectural use of the long wooden pin found here, L.B. iv. i. 2, has already