National Institute of Informatics - Digital Silk Road Project
Digital Archive of Toyo Bunko Rare Books

Sec. ii]   EXPLORATION OF RUINED DWELLINGS, L.A. II—vi   38 t

Limes to Farhâd Bag-yailaki, furnish as to this primitive method-of fire-production having prevailed,

during the early centuries of our era, all along the line of Chinese advance westwards.17

Immediately to the south-west of L.A. v could be traced a fence made of tamarisk rushes Fenced

marking a rectangular enclosure about forty feet across. Beyond it were several smaller fenced enclosures.

spaces, probably meant for horses or cattle, and adjoining the remains of some modest quarters,

L.A. vi (Plate 25), roughly built with rush walls. In the small rooms, i, the wind had scooped the

ground clear in the centre ; yet light sand lay in the corners, and in the one to the south-west there

survived a solitary rectangular covering-tablet, bleached and splintered. The mud plaster of the

walls had completely disappeared, and little was left of their wooden framework.

SECTION III.—DISCOVERIES IN AN ANCIENT REFUSE-HEAP, L.A. vi. ii.

Far more interesting than the small dwellings last described was the rich mine which we struck in the big rubbish-heap, L.A. vi. ii, extending eastwards from L.A. vi to the extreme western portion of the ` Ya-mên ', where it joined the previously searched area L.A. in. i. It measured over a hundred feet across, with a width of about fifty feet (see Plate 25). Fig. ioo shows it in the course of clearing. On the south, to a height of from four to five feet gradually diminishing northward, lay a mass of consolidated rubbish consisting mainly of reed straw and stable refuse including abundant droppings of horses, donkeys, and camels, and to a smaller extent also of sheep. As soon as excavation of this unsavoury quarry had commenced from the south, Chinese records on wood and paper, together with other small relics, cropped up in numbers. Most of them were recovered from layers two or three feet above the ground level and thus close to the present surface. Evidently the sweepings from the quarters and offices close by had been thrown down here on earlier accumulations of stable litter. Does the comparative rarity of records in the latter indicate a period when the ruined station served as a halting-place on the ancient trade route, but had ceased for a time to be a place of much official activity ? Our present knowledge of local history in this region is too imperfect to permit of any definite answer. But it is certainly curious that the dated records on wood from this deposit of refuse (see Doc. Nds. 721-28, 731, 733, 735-36) are contained within the narrow chronological range of A.D. 264-70, to which two fragmentary documents on paper (Doc. Nos. 910, 912) add the date A. D. 312.

The careful clearing of these accumulations of dirt kept us busy for the best part of two days. The odours were still pungent and, combined with the icy north-east wind which was constantly driving fine particles impregnated with ammonia into eyes and throat, made the task very trying. It was impossible to remove the refuse altogether, and after being turned over and searched, it had

to be left in situ.   In this process the rubbish-heap had naturally lost its quasi-consolidated
condition. It was hence of special interest to me to find it, on my second visit in February, 1914, practically undiminished and scarcely disturbed on the surface, in spite of all the gales and sand which had driven across it during the seven intervening years. This strikingly illustrates the wind-resisting capacity of such refuse, especially when containing a thick admixture of straw, which has protected from erosion, at so many of the ancient sites that I explored,' valuable antiquarian relics.

In proportion to the extent of this refuse-heap was the abundance of interesting finds which rewarded its clearing. Among them the most numerous class are the Chinese records on wood and paper. Of the former I counted on the spot about 170, of which 121 are published in M. Chavannes' Documents, the rest being small fragments or undecipherable for one reason or

" See above, pp. 233 sq.   1 Cf., e.g., Ancieni Kholan, i. pp. 343 sq., 479 ; above, pp. 205, 276.

Rich mine of Chinese records.

Refuse resists wind-erosion.

Condition of wooden Chinese slips'.