numbers. It was a room 26 ft. square, with a raised platform of plaster running round three of its sides, while the remains of eight posts arranged in a square indicated a central area which probably had a raised roof to admit light and air, after the fashion still observed in the halls of large Turkestan houses. As the protecting layer of sand was here only 2 ft. deep little more was left of the walls than rows of broken posts. The first inscribed tablets, too, which turned up in the sand close to the surface, had suffered greatly, their warped and split wood showing plainly the effect of the terrible summer heat to which they must have been exposed since the winds had carried away most of the sand that originally protected them.
All the more delighted was I when I found that even the light remaining layer of sand had sufficed to preserve in a more or less legible condition the threescore of tablets that were found covering the platform along the southern side of the room. In some places, particularly near the centre of the wall, they were rising in small closely packed heaps above the plaster flooring, evidently just as left by the last occupants. But a considerable number of other inscribed tablets showed plainly by their position that they had been disturbed at some early period, apparently not long after the building had been deserted. For, in addition to some twenty tablets which were found scattered about in the loose sand covering the floor just in front of the south and east platforms, I unearthed over two dozen more from the southern part of the central area of the room marked by the posts already referred to.
As the layer of these tablets was being removed it was seen that they had rested on a square piece of strong matting which, supported by some light rafters also recovered, must have once formed a roof over the central area. The matting was found lying about a foot above the floor, thus showing the depth to which the invading sand had accumulated before the roof fell. The tablets found above the matting could only have got there subsequently. They may well have been thrown there when the abandoned dwelling was visited by some one anxious to search its remaining contents after
IIIF Color HighRes
