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

740   HISTORY AND RECORDS OF THE TUN-HUANG LIMES [Chap. XX

as that traversed by the Great Wall beyond Tun-huang must have been all through historical times. This geographical fact once properly recognized forces upon us two conclusions which have their special antiquarian bearing. In the first place, we must realize that on such ground the constant maintenance of considerable detachments or bodies of troops, along a line which for great stretches was devoid even of water, would have presented most serious difficulties about supplies and transport. This makes it obvious that the regular pickets stationed at the watch-towers along the Limes wall in ordinary times must have been small, probably limited to the minimum compatible with the maintenance of an efficient signal and patrol service and with the protection of the line from such incursions as a few raiders might attempt at a time. The larger bodies, which would have been needed to repel any possible inroad in force across the deser north and west, could be kept available only within, or quite close to, the habitable areas of the Nan-hu and Tun-huang oases. In the second place, it appears to me very improbable that on such ground, for the most part absolutely bare gravel desert and almost everywhere so open that even mere débris heaps of completely decayed towers, less than a dozen feet high, could be sighted with ease miles away, structural remains of any consequence, other than those actually explored, could have escaped the notice of myself and my assistants, while we repeatedly crossed and recrossed the area within the westernmost Limes for months. We shall see further on how this observation warrants us in drawing certain inferences from purely negative evidence and in using them for the interpretation of statements contained in our documents from the Limes.

Barrenness of ground traversed by Limes.

Pickets limited by difficulties of supply.

Imperial order to governor of

Chits-cNilan.

Edict refers to extension of Limes beyond Tun-huang.

SECTION IV.—MILITARY AGRICULTURAL COLONIES

In our review of the antiquarian information to be found in the documents from the Limes it will be convenient to start with those records which can throw some light on the general organization of the Tun-huang border, and subsequently to proceed to the far more numerous class furnishing details about those who kept watch there and about the varied aspects of their duties and daily life. Among the former records by far the most important is the complete and neatly written, but in some places effaced, slip T. vi. b. i. 289 (Doc., No. 60, Plate III). It reproduces an ` imperial order given to the governor of [the command of] Chiu-ch`iian' and manifestly relating to the establishment of a military colony on the border. The order first mentions ` two thousand from among the garrison soldiers of the command of Tun-huang', who apparently together with others ` from the command of Chiu-ch'üan' were to be raised for the enterprise. ` The ssu ma and his subordinates, together with the generals, soldiers, and [indigenous] functionaries, will proceed to occupy a locality in order to establish there an agricultural colony. It will be the duty of the governor to examine the configuration of the places. By utilizing natural obstacles a rampart will be constructed in order to exercise control far away.' The order closes with a formal injunction often found at the end of other documents from the Limes : ' Let there be no negligence of any kind, and be the orders conformed to.'

If we carefully consider the context of the edict reproduced in the document and the place where the latter was found, it appears to me difficult to doubt that the agricultural colony of soldiers to be established must have been connected with the extension of the Limes beyond Tun-huang. General as the terms of the edict are, it yet refers clearly to the construction of a defensive line which was to guard the outlying ground to a distance ; it was to be built with due regard to the ' configuration of the ground' and with whatever advantages could be derived from the use of natural obstacles. After our detailed survey of the westernmost portion of the Limes it is needless to emphasize how closely the directions here given are reflected by its actual features. The