thorough examination, as Naik Shamsuddin's digging party which followed behind me arrived at this point too late on the evening of March 22nd. When next morning I reached the Limes line again at T. XXIII. 1 with my camels and baggage, the uncertainty as to where water might be found farther on for camping obliged us to move on along the line, and no time could be spared for a return to those posts.
T. xxiii. I proved a brick-built tower of the same outer dimensions and appearance as those last described. But the guard-room, i, here 6 feet square and entered by a narrow passage from the south (see plan in Pl. 14) at a level of about five feet from the ground, was found choked with refuse to a height of over four feet. Mixed up with reed-straw, chips of wood, broken bricks and the like, there emerged here over two dozen wooden slips bearing Chinese records, most of them in fair preservation, together with a number of others which are blank and evidently had been kept ready for use as stationery. It appeared very likely that this refuse had been allowed steadily to accumulate while the small room continued to be occupied as a clerk's office ; for layers of ashes as well as reddened brickwork were found at different levels where fires had been kept burning against the western wall. [Among the inscribed slips recovered here there are found fragments of private letters (T. XXIII. 1. 2, 21, 23, o8) as well as the fragment of a literary work, T. XXIII. 1. i. 7. T. XXIII. 1. i. 3 is of interest as it appears to contain instructions as to fire-signals to be lit from a car by soldiers sent out on reconnaissance. In T. xxiii. 1. i. 8, 12 the ` signal posts' of Wei-hu j j}J and Chih-k`ou j F A are mentioned. The former post recurs in T. xxiii. 1. i. 18, which is curious as it records the meeting on a certain day of patrols sent out from Wei-hu and from another post, Hsiian-wu, to the west of it. The tablet has been cut into two halves to serve as a tally with a number of notches marked uniformly in both. The document is described as a ch`iian 4, this being the proper term applied to ` an instrument in writing, originally on a tablet of wood, which was cut in two, each party having half '.] 28
On searching the ground outside, a small refuse heap (ii) was discovered just beyond where the mound marking the Limes wall passed in a semicircle to the north of the tower. Here under only a few inches of gravel over three dozen more records on wooden slips were brought to light, unfortunately most of them completely effaced by moisture. They had evidently been thrown out together on the occasion of some clearing out of ` waste papers ', like those far more numerous packets of ` slips ', & c., found in 1907 at T. vi. b, T. xv. a, &c. Their decayed state seemed to indicate that we could not here hope for such favourable conditions as regards absolute dryness of air and soil as prevailed along the westernmost portion of the Limes. [Of the slips still partially
decipherable T. xxiii. 1. ii. 09 mentions the ` signal post ' P`o-lu j. In T. xxiii. 1. ii. 020 the
fire-signals of Tien-ho f and I-ho g AC are referred to ; the latter name occurs also in
Chavannes, Doc. No. 637, found at T. xxviir, a Limes station to the north of Tun-huang.] Among the miscellaneous objects extracted from the refuse surviving at this station and described in the List below, I may mention the wooden bowls, T. XXIII. 1. i. 03-5 (PI. XLVI), round or boat-shaped ; the fragment of an iron weapon or implement, 1. i. 06, and the wooden tally-stick 1. ii. 03 (Pl. XLVII).
The next two posts by the wall, T. xxiii. m, n, were reached at intervals in each case of three-quarters of a mile. The general direction still continued to the south-east, but the position of the individual towers, here as from T. xxiii. h onwards, diverged from the straight line. This perhaps was arranged with a view to making it easier to distinguish the fire-signals of successive stations. Both these posts were marked only by completely decayed mounds of small size, and beyond them the line of the wall ceased to be traceable on the marshy ground covered with thick reed-beds.
28 Cf. Giles, Chinese-English Dictionary, p. 3159.