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

Sec. iii]   HUA-HAI-TZÛ AND ITS LIMES REMAINS   393

Tun-huang Limes (Chavannes, Doc., Nos. 415, 485). In 023 the commandant of a ` signal post' (sui) is referred to as a native of the canton of Fu-chaang r h of Yü-mên. 019 furnishes an inventory of various textiles and articles of clothing ; 029 and 03o are fragmentary private letters.]

The well-made string shoe, T. XLIII. h. 04 ; the wooden peg, 013, with a roughly drawn grotesque face, and the iron hoe-blade, 012, have their exact counterparts among objects from Limes posts west of Tun-huang, occupied during Han times. In view of the chronological indication afforded by these and by the coin previously mentioned, the find of a torn piece of paper, o6, uninscribed, has its special interest. The distance from the tower T. XLIII. g in the west to T. XLIII. i, the next tower eastwards, proved little more than a mile. It has occurred to me that the interposition of T. XLIII. h so close to each of these posts might perhaps best be explained by the supposition that its remains mark the site not of an ordinary watch-tower but of a point d'appui or sectional head-quarters, such as T. v1. b and T. xv. a on the Tun-huang Limes.' The comparatively large number of documents found here would well agree with such an assumption.

Neither here nor near T. XLIII. g was it possible to trace the actual Limes wall. But eastwards the belt of tamarisk-cones receded, and close to the point where the line of the wall reappears on gravel soil amid a scanty growth of scrub, we came upon the much-decayed remains of a brick-built tower, T. XLIII. i, occupying a small natural knoll to the south of the wall. It was interesting to see this knoll now completely overtopped by a tamarisk-cone close by, fully 3o feet high, which must have grown up since the watch-post was here built.8 From a refuse heap on the slope of the hillock nine Chinese records on wood were recovered, including a rectangular cover-tablet [with the name of the addressee] and a seal-case. Among the miscellaneous objects found here may be mentioned the piece of a wooden saddle-tree, T. XLIII. i. 02 ; the fragment of an iron implement, 07 (Pl. XLVII), wedge-shaped, of uncertain use, and a small quantity of oats, of which o8 is a specimen.

From here onwards the wall was found in very fair preservation, running eastwards, with an average height of 5 to 6 feet, but rising in places even a couple of feet higher (Fig. 22o). At a point about one mile distant from the post last described and a little to the south of the wall, a mound was found covered with the debris of a small brick-built structure, T. XLIII. j. A room, of which only the east wall remained clearly traceable, had its floor covered with refuse. The excavation of this yielded no less than twenty-four Chinese records on wood, including not only fragments of ` slips', but also a complete trilateral piece of the size and shape which, as is shown by corresponding finds on the Tun-huang Limes, was used for literary productions.9 [The trilateral piece, T. XLIII. j. 014,

has proved to contain the beginning of the 14th paragraph of the Chi chiu chang   vi , a
famous lexicographical text, the popularity of which during the later Han period is attested also by a number of fragments recovered at different points of the Tun-huang Limes on my second expedition.10 T. XLIII. j. 013 contains a writing exercise in which a date corresponding to the year 4o B.c. is copied. A private letter from an exile is found in 05-6. Numerous pieces, 07-12, 015-2 I, &c., seem to belong to an astrological calendar.]

Among the few miscellaneous finds are the wooden heel-piece of a shoe last, 02 (Pl. XLVI) and a piece of well-made basket-work, 04 (Pl. XLVI). But more interesting is a small carefully carved wooden peg, of (Pl. XLVII), supporting at its upper end a carved head with the details of the face neatly cut in or painted in black. The little object is remarkable on account of the resemblance it shows, as pointed out by Mr. Andrews, in the treatment of the coiffure to certain Han

Relics found at T. XLIII. i.

Limes

wall

traced to T. XLIII. j.

Miscell. finds at T. XLIII. j.

7 See Serindia, ii. pp. 644, 698 sqq.

8 For a corresponding observation made at the ancient fortified post south of the Endere site, cf. Serindia, i. pp. 283 sqq.

9 Cf. Chavannes, Documents, pp. 6 sqq., Pl. I, II ; Serindia, ii. p. 763, n. 92.

10 Cf. Chavannes, Documents, pp. r sqq. ; Serindia, ii. P• 763.

3E