National Institute of Informatics - Digital Silk Road Project
Digital Archive of Toyo Bunko Rare Books

Sec. y]   THE WATCH-STATIONS T. xi' AND T. x1II   679

in the Chinese records brought to light by my excavations I knew already from some which Chiang

Ssû-yeh had been able to decipher on the spot. But it needed M. Chavannes' translations, as

embodied in his Documents, to show me how frequent these references to fire-signals are in the

records recovered from the different stations.2

The subject is one which will best be discussed below in the review of the general antiquarian Fascines for

information furnished by the documents from the Limes. Here it will suffice to point out two lighting

signal-fires.

minor observations which support this interpretation. No doubt, such signals would ordinarily be lit on the top of the towers, whence fires, even if small, could be quickly sighted by the men on guard at neighbouring stations. The reddish burnt appearance of the clay on the top of several watch-towers where it still remained and was accessible bore direct testimony to this practice. But there might be circumstances, as on occasion of a particularly big fire needed to penetrate a murky night or to light up the foreground in expectation of an immediate attack, when it would be necessary to set a whole stack on fire. The fact that the remains of burnt stacks were usually found, as shown by Plate 39 in the case of T. xII. a, at points such as the south-east corner of the group, where the risk of igniting others was less, thus receives its proper explanation. The greatly varying height of the stacks, from 7 feet down to i foot only, at the same watch-station can best be accounted for by the successive use made of the stored materials for signal-fires kindled in the usual way on the tower itself.

After proceeding for about three-quarters of a mile to the south-east along the narrow Position of

plateau on which the watch-station T. xi'. a and the adjoining segment of the wall are built, watch-tower

xii

the tower T. xi' (Fig. 181) is reached. It occupies the southern end of that gravel ridge . . where it drops down into the wide marshy depression which I have already had occasion to mention in connexion with my return to the Limes from Nan-hu.3 The bogs and small lakes found in the Nullahs on either side of T. xii. a fill northward offshoots of the same depression. The latter is traversed by the caravan track to Lop close to the south of T. xii over ground slightly higher than the rest, and it is this topographical fact which explains why we find here the ruin of a watch-tower well removed from the line of the Limes wall. It is clear that the ancient route to Lou-lan must also have crossed the depression at this point, the ground further south being quite impassable owing to the large salt morass. A reference to Plate J3 or Map No. 74. D. 3 shows that a post maintained at T. x11 was excellently placed for guarding the ancient route and watching the traffic passing along it. But we see also that it could not be intended to strengthen the defensive line of the Limes itself, as it is well behind it and near a segment of it which was adequately protected by the natural obstacles of impassable marshes.

I am therefore strongly inclined to believe that the purpose of T. x11 was to serve as a road- Watch-post

side post for what I may call the police control of the border as distinct from its military defence. for police

control of

Guards placed here could make sure that westward-bound travellers, traders, etc., had been duly road. authorized to proceed extra muros by those in charge of the ` Jade Gate ', the main frontier station, located at T. xiv. From all that Chinese and foreign records show us of the administration of the kuan, or ` barrier ', on the empire's western border during successive periods, we can feel quite sure that this function of the police cordon maintained there was always taken as seriously as it used to be until quite recent years at the Chia yü kuan of Su-chou, the modern representative of the ancient Yü-mên kuan.4 In the same way a preliminary watch could be kept here upon travellers, etc.,

z Cf. below, pp. 752 sqq. Here I may conveniently-   been intended to be used either for kindling signal-fires or

note the curious fact that we have a record of the collection   for repairing the wall.

of such fascines in one of the slips found at T. xII. a. ii, Doc.,   3 See above, p. 629.

No. 609. It mentions the respectable total of 42,390 faggots.   ' For Chia-yü kuan, the modern equivalent of Ytt-mên,

M. Chavannes has correctly recognized that these must have   see Desert Cathay, ii. pp. 274 sqq.; below, chap. xxvii. sec. ii.