Sec. iii] DISCOVERIES IN AN ANCIENT RUBBISH-HEAP, N. xv. 341
pieces of leather, and other promiscuous rubbish. The earth which enveloped all this litter looked exactly like the ordinary loess, and manifestly owed its origin to the consolidation of the dust deposits which the rubbish-heap must have received in the course of its gradual formation. The dust which had been bound together here, was, of course, of the same character as that which has helped to form the loess soil. It was evident that the consistency which this accretion of varied refuse intermingled with dust had acquired in the course of long centuries, had more than anything else helped to protect it against the erosive action of the winds. It was, in fact, the compactness of these deposits, filling the whole interior of N. xv. to a height of about 4 ft., which had saved the walls of the room (seen in Fig. 45) from being reduced to mere rows of posts like almost all the other walls in this ruin. The combined conditions of cohesion and dryness which the weight of this ancient dustbin and the absorbent character of its sub-aerial dust deposits assured, account for the remarkable state of preservation shown by the great mass of the precious records embedded in it. At the same time we have a clear indication of the more unsavoury associations of a dustbin in the salt impregnation of many of the wooden tablets from N. xv., mainly due to the presence of ammoniac. It makes them now particularly sensitive to atmospheric moisture, and liable to develop stains even under the protection of the well-fitting glass panes of the British Museum.
I had ample opportunities to familiarize myself with all the manifold contents of this remarkable refuse-heap during the three laborious days which its clearing cost me. As soon as I had realized the peculiar origin of its deposits, it became a matter of importance to keep as accurate a record as the conditions of the work and the available time would permit of the relative position in which each written document or other object of interest turned up. This might assist in tracing chronological order and in some cases possibly also internal connexion among the various documents. Accordingly, every inscribed piece had to be removed by myself and briefly tabulated after a preliminary cleaning. This was no easy task with fingers half-numbed by cold in the fresh north-east wind which was blowing during the greatest part of the time, and in the dust it raised from the dug-up refuse-heap. Its odours were still distinctly pungent and trying after so many centuries ; but there was all the fascination of a great antiquarian haul to make me forget such discomforts.
The four wooden posts which rose above the rubbish in the centre of the room, and which in all probability had supported an elevated portion of the roof, after the fashion of a modern Aiwdn, afforded a convenient means for demarcating the several portions of the room successively cleared. The space to the north of the central area (marked N in the plan) which was first excavated and searched on February 6, in the direction from east to west, yielded the inscribed pieces N. xv. i-79, including therein those found on my prospecting visit amidst the layer of rubbish that had slid down near the eroded north wall. The finds made in the space which extended between the east wall and the central area of the room (shown as E in the plan), and the rubbish contents of which proved relatively richest in documents, were marked with the numbers N. xv. 80-149. Here, as well as in the remaining two sections, the order of working was from north to south. The section between the west wall and the central area (shown as W), which was cleared on February 7 and 8, furnished the pieces N. xv. 150-207, while from the space between-the four posts and the one immediately to the south of it (C) came the series numbered N. xv. 300-3632. The mud platform running