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. iv]   T. xiv. A AND THE REFUSE-HEAPS OF T. xv. A   699

station—for as such it could safely be recognized notwithstanding the almost total disappearance of structural remains—must have been occupied for a great number of years ; so extensive, and in places deep, were the rubbish accumulations. To the west of the débris first discovered the scraping of the slope from about io to 20 feet below the top of the terrace revealed refuse lying to a maximum depth of 5-6 feet. At this place, marked T. xv. a. ii, most of it consisted of brick débris, ashes, and reed straw. But, just as at i, there were found plentiful Chinese records on wood mixed up with it. Almost as abundant were the documents extracted from another refuse layer, T. xv. a. iii, covering the slope about 20 yards north of ii over an area of about 6o by 3o feet. On lower ground westwards, and at about 5o yards distance from ii and iii respectively, were traced two smaller refuse-heaps, v and vi, while yet another, iv, was discovered near the southern edge of the terrace some 25 yards from i. Further to the east, on the level top of the terrace, there survived two stacks of fascines of the usual type to a height of 2 or 3 feet. The one further south (see Plate 40) was in part reduced to a slag-heap still showing calcined reeds. Some 4o yards off eastwards there rose a third stack.

As three at least of these rubbish deposits can be proved by documents to date from successive and well-defined periods, it will be convenient to deal first with the written records they yielded and subsequently with the miscellaneous and presumably contemporary relics found with them. The refuse-heap of T. xv. a. iii manifestly represents the earliest deposit ; for among the thirty-six documents from it included in M. Chavannes' publication all those which are datable belong to the Former Han period. In Doc. No. 446 the date 53 B.c. is quite certain ; in No. 447 that of 6i B.C. is probable. Of No. 449 M. Chavannes has demonstrated that the names by which the various commands of westernmost Kan-su from Lan-chou to Tun-huang are mentioned point distinctly to the closing years of the Former Han dynasty. Three records, Nos. 463, 464. 470, refer to the I-ch`iu g # company, and, as we find it named in five more documents from T. xv. a (Nos. 482, 486, 535, 536, 541) and nowhere else, it can be safely concluded that this company provided the garrison of the station for most, if not the whole, of the time during which it was occupied. In certain documents from T. xv. a. ii (Nos. 482, 484, 485) the names of three other companies occur in connexion with payments made by individual men, etc. But none of these companies are mentioned

again elsewhere, except that of Chu-chüeh j   , which, both in T. xv. a. ii. 9. Doc., No. 484, and in
No. 693 (T. xix. i. 6), is described as belonging to Ping-wang. We have seen that this section of the Limes probably extended from near the Khara-nôr to T. xiv. a.3

The reference made to Yii-mên in Nos. 451, 458, 459 would be sufficiently accounted for by the vicinity of that important frontier headquarters at T. xiv. But a record from T. xv. a. i, Doc., No. 536 (Plate XVI), containing a circular order dated in A.D. 137, distinctly implies that the station T. xv. a and the I-ch`iu company were without doubt at that time subordinate to Yii-mên. We have a record of intermediate date pointing to the same conclusion in T. xv. a. ii. 22, Doc., No. 483 (Plate xiv), which conveys an order from the commandant of the fortress of Yii-ma' and is dated A.D. 43. Another and otherwise unknown locality, ` the fortified post of Yu-ch`ang-lo', is referred to in T. xv. a. iii. 43, Doc., No. 452 (Plate XIII), and deserves notice here in view of a topographical surmise to be discussed below. The document authorizes a certain person to proceed to Chien-tu, taking the camel provided for the official use of the wes-shih, a subordinate officer, at Yü-ch`ang-lo, and drawing three daily food-rations for the journey. Chien-tu, or Ta-chien-tu as it is usually designated, has already been located at the extreme western end of the Limes wall.4 The distance to that point, i.e. T. iv, from T. xv. a is only about 32 miles, easily covered by two daily marches, and T. xi offers a convenient intermediate halting-place. The

See above, pp. 691, 698.   Cf. above, pp. 636, 648.

4U2

Refuse deposits of T. xv. a.

Dated

records from T. xv. a. iii

Document
referring to
Jade Gate.

Post of Yi1-ch`ang10.