632 THE RUINS OF LOU-LAN.
Between the cross-walls are three stalls, which at the first glance suggested a stable or a temple shrine for images of Buddha. The walls were 1.95 m. high at the most, but on the top were so irregular that in places they actually appeared to have been filed away by the wind. They were o.85 m. thick, and extended 5.6 m. from the principal wall. It was in the : stall that lay farthest towards the north-east, and which had a breadth of i .40 m., that we discovered all the fragments of written paper and tablets of wood which Mr. Himly and Prof. Conrady have deciphered and described. They were buried under an accumulated layer of dust and sand, not more than i . r m. deep, the spot being almost completely sheltered from the wind. The only side on which the stall was open was the south-south-east, and . from that direction it seems practically never to blow in the Desert of Lop, at all events the wind from that quarter is only slight. The sand and dust having been whirled into the corner of this . stall were unable to get out again, and it is to this circumstance that we must ascribe the fact that the fragments of paper showed not the slightest signs of wear .or scouring by the sand, but were as well preserved as if they had lain there only a few days. Nor did the paper show any but the slightest signs of the »tooth of time», a .circumstance that could only happen in a country where there is no rainfall, and where the atmospheric moisture is at a minimum. The fact of these scraps of paper lying at a depth of about one meter would make them hardly sensible of the 'alternations of temperature between day and night. Between winter and summer there is a greater range of temperature, but the changes are distributed over longer periods and are never abrupt. Thus it was the shape of the apartment, its orientation, its still surviving walls which, in conjunction with the protecting layer of sand and dust, had preserved these precious, though alas too few, documents from destruction, although this fate has overtaken all other perishable objects that may chance to have lain in the wooden house.
The layer of sand from which we unearthed these documents bore a striking resemblance to a rubbish-heap. I need hardly say that the contents of the stall were sifted to the very last grain of sand, until we came in fact to the hard bare ground underneath, and had literally cleared out the whole of the interior. Amongst the material thus removed we found the skeleton of a rat, with its skin scarcely injured; a great quantity of fish-bones — which witness just as unambiguously as the mollusc-shells do to the former presence of fresh water; a whip, with a loop in the end of its thong, and having the shin-bone of a sheep for the shaft; a pig's tooth; parts of the skeletons of sheep, cattle, horses, and camels; buttons; copper coins; rags of clothing; the sole . of a shoe made of horse-hair; old iron, e. g. a rusty chain, a couple of holders for. Chinese pencils; a wooden spoon; and finally a large quantity of kamisch-stalks. In a word it is little likely that all this worthless rubbish was brought there accidentally. The stall was indeed too small to have served as a dwelling. Possibly it may actually have been used as a receptacle for rubbish and sweepings of refuse. This is confirmed by the fact that most of the papers were fragments and had been 'deliberately torn to pieces. A large number of thin wooden tablets with Chinese script bore evidences- of having been used twice, the face of the tablet with the first writing having been shaved off with a knife, after which the tablet was used a second time. The other particulars connected with this discovery are
IIIFカラー高解像度
