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

594   THE TUN-HUANG OASIS AND ITS NORTHERN LIMES [Chap. XV

felt quite sure. Unfortunately this proved to occur repeatedly, in the sixth as well as in the third century A. D., and further back I scarcely ventured to look. In the other Nien-hao Chien it ... , shown by T. xxv11. 2, 3, my learned Secretary was at a loss how to identify the second character, though written clearly enough in both slips. No Sinologue familiar with the intricacies of Chinese palaeography will wonder at his doubts, nor have anything but praise for the scholarly frankness with which my excellent literatus-friend was wont at all times to express his difficulties.

In vain I searched near the Yung Ping periods previously mentioned for any Nien-hao which might give Chiang Ssû-yeh a clue to Chien .... At last I had the courage to look back a few centuries : there was a Yung Ping period beginning in A. D. 58, and, separated from it only by one Nien-hao covering an interval of two years, there stood the regnal title of Chien Wu it te. When I showed it to Chiang, he at once recognized in it the second character which had puzzled him so far. Chien Wu was the first regnal title adopted by the Emperor Kuang Wu-ti, who in A. D. 25 founded the Later Han dynasty, and the twenty-sixth year of it, which is the date recorded in both the documents T. xxvir. 2, 3, thus proved to correspond to A. D. 5o. There could be no doubt any longer that the old frontier wall, the ruins of which I had determined to explore, went at least as far back as the first century of our era, and that the wooden documents which I had in my hands as proof were the oldest original Chinese records in writing, as distinct from inscriptions, so far brought to light. It was a discovery both gratifying and inspiriting : all along archaeological evidence had caused me to put faith in the antiquity of this Limes, and now I felt fresh confidence in its successful exploration.

On the morning of March 28, with an icy gale blowing from the north, I directed our camp to be moved to the ruined tower T. xxvIII, which I had sighted on the previous evening on the southwest of, and next to, the one then so successfully prospected.3 The position was convenient for following up the exploration of the other ruined towers sighted to the west. Moving ahead with every available man, I started the same morning the clearing of the thick layers of refuse which were very soon discovered on the south slope of the clay ridge occupied by T. xxv111. Then, leaving the work to be continued under my assistants' supervision, I proceeded to reconnoitre the ruined towers just mentioned. But instead of describing the tasks as they were taken up and carried out in succession of time, it will be more convenient, for the reader as well as myself, if I record the results in the topographical order of the ruins. I propose to follow the same plan also in regard to other sections of the Limes which I was subsequently able to explore in detail. Such a treatment will make it easier to consider topographical features bearing on the Limes in conjunction with the archaeological facts revealed by its exploration.

The complete clearing of the ruined watch-station T. xxvii did not disappoint the hopes which the first visit had raised. The débris lying by the tower on the west (Plate 34 ; Fig. 155) proved to belong to a room measuring fourteen feet from east to west and probably quite as much or more across. The north wall had completely disappeared through erosion. Those on the west and east were, at least in part, cut from the live clay of the narrow ridge occupied by the tower ; but on the east side a brick-facing, ten inches thick, had been given. The south wall was three feet ten inches in thickness and entirely built of bricks which measured thirteen by seven inches, with a thickness of five inches. Through it led the entrance, three feet wide within, and narrowing to two feet three inches outside, evidently for the sake of greater protection. The floor of the room, when completely cleared of the refuse filling it to a height of over three feet, was found to occupy the same level as the lowest masonry course of the tower. Obviously the top of the ridge

3 In Map No. 81. A. 3 the symbol to mark the position of Camp 164 at T. xxvin has by an oversight been omitted.

Dated Chinese record of A.D. 50.

Excavation started at T. xxvnl.

Topographical order in description of ruins.

Ruined quarters of T. xxv11.