国立情報学研究所 - ディジタル・シルクロード・プロジェクト
『東洋文庫所蔵』貴重書デジタルアーカイブ

Sec. iii] FIRST DISCOVERY OF DATED HAN RECORDS AT T. xxvit   597

of a type represented elsewhere, T. XXVII. 007-008, two roughly-cut wooden pens, made of tamarisk twigs, T. xxvII. 0018-0019, throw some light on the occupations of those who were stationed here. Passing mention may be made of two wooden spoons, T. XXVII. 0015, 0017, and of a broom and a well-made piece of fibre string matting, T. XXVII. 0021, 0023 (Plate LIV). There is in T. xxv11. 0010 a specimen of those curious wooden pegs showing a roughly-drawn grotesque face, which turned up in numbers at different watch-stations, and of which the true purpose still remains to be determined (cf. Plate III for other examples). The wooden fire-stick, T. xxVII. 0011, ' female' portion, is of interest as closely agreeing, in shape and arrangement of the holes, with the pieces found at the sites of Niya, Endere, and Lou-lan.e

SECTION IV.—SEARCH OF RUINED LIMES STATIONS T. XXVIII-xxx

The tower T. xxvlll was found to be situated about one and a half miles to the south-west of that last described, and, like it, on a low clay ridge rising above eroded ground with scanty tamarisk growth. It proved to be very badly decayed, as seen in Fig. 156, and no exact measurement of its original ground-plan could be obtained. It is likely to have been, as usual, about twenty feet square at the base ; the actual height was about thirteen feet and the material stamped clay, with thin layers of tamarisk brushwood at intervals of three to four inches. No other structural remains could be traced ; but at first sight my eye caught thick layers of refuse covering the south slope of the ridge from about ten feet down to about twenty feet below the foundation of the tower. Their total width was close on thirty feet, and at the foot of the slope, where the refuse rested in horizontal strata on the natural soil of soft alluvial loess, it lay quite three to four feet high. Its position showed clearly that, at the time when the tower was built, the clay ridge must already have risen some twenty feet above the immediately surrounding ground and thus offered itself as a good look-out place to watch the riverine depression northward. This is likely to have been covered in ancient times, even more thickly, perhaps, than it now is, with tamarisk-cones and other growth of scrub, and thus to have needed close guarding. To this special feature of the ground I would ascribe the relatively short distances at which the watch-stations of this eastern section of the Tun-huang Limes were built, nowhere more than about one and a half miles, and in places only about one.'

The chief ingredients of the refuse were straw of reeds, twigs and bark of tamarisks, and dung of horses—obviously, in the main, stable refuse thrown down here from some watch-station forming a link in the long chain of posts which ran along the wall. But from the very edge of the top of the heap protruded the small but quite legible fragment of a Chinese slip (T. xxviir. 1 ; Doc., No. 647, Plate XVVII), and as soon as clearing was started, ancient records on wood cropped up in striking abundance. It proved, indeed, a precious rubbish-heap, and in the end the day's work which it cost us was rewarded with the discovery of over seventy pieces, of course mostly fragmentary. Out of the total M. Chavannes found forty-seven sufficiently legible for publication in his Documents.2

The variety of the miscellaneous correspondence and office ` papers '—for as such I could through Chiang Ssû-yeh's help safely recognize them—found here and at T. XXVII was sufficiently great to familiarize me on the spot with the main external aspects of this ancient Chinese stationery on wood. The most usual form was certainly the thin wooden slip measuring, as already stated, from 9 to 92 inches in length and from quarter to half an inch in width. The fact that there were complete slips containing up to thirty characters and more in a single vertical line (see e. g. T. xxvlll. 54 ;

Refuse- heaps at watch- tower

T. XXVIII.

Clearing of refuse layers.

Ancient stationery on wood.

6 Cf. above, p. 233.   T. xxvli, see below, p. 604.

As to the apparent exception between T. xxvi and   2 See Documents, Slrie 7. xxviti, pp. 135-41.