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

square, a natural clay bank projecting beneath the broken west face of the masonry having been
wrongly included before in the measurement. Remains of brick-built walls rising to about two feet
from the ground, and enclosing small apartments which must have served as quarters for the
detachment guarding the watch-station, were brought to light close to the east and north faces of
tower. From the little passage iii a flight of stairs, each nine inches high and very narrow, had
once led up to the top of the tower; two of them were still intact.

Documents
found at
quarters of
T. iv. b. The clearing of these modest quarters yielded about a dozen Chinese documents, which with
one exception were on wood, several of them being in fair preservation.⁶ It is curious to note that
a complete and particularly clear wooden 'slip' (Doc., No. 432, Plate XIII),⁷ which contains a general
order enjoining 'a perfect look-out and the immediate lighting of a fire-signal on receipt of one',
was found just in the passage iii giving access to the top of the tower. The text itself describes it as
'an order to be posted up in a visible place of the [quarters of the] company of the watch-post', and
the place where it was found proves that this direction had been duly complied with. Among the
other documents, which all refer to military matters, receipts for letters, rations, and the like, two
others deserve mention here as having a special archaeological bearing. The large tablet (T. iv.
b. ii. 1 Doc., No. 430, Plate XII) is of interest as emanating from 'an adjunct of Ta-chien-tu' and
bearing a date which M. Chavannes hypothetically accepts as T'ai-shih 3, in agreement with Chiang

Ta-chien-tu,
name of
Limes
section. Ssŭ-yeh's original reading. As regards Ta-chien-tu 大 煎 都, it appears to me very probable that this
local designation, which is met with in nine records found at the stations T. iv. b, v, vi. b, and only
in one found elsewhere, T. xiv,⁸ refers to that section of the Limes which covers the westernmost
watch-towers of the wall and the detached ones guarding its south-west flank.

Dated
records from
T. iv. b. With regard to the date it must be observed that the characters read above as T'ai-shih, and
hence as indicating the year 94 B.C., could be read according to M. Chavannes also as Yüan-shih,
which would make the date correspond to A.D. 3. Considering that all the numerous dated
documents found at the watch-stations on the south-west flank belong to the first century B.C., and that
[Ta]-chien-tu is mentioned in a document of the year 96 B.C. (Doc., No. 304, T. xiv. iii. 67), I should
be inclined to prefer the earlier dating. But it must be noted that another record from this station,
T. iv. b. i. 10 (Doc., No. 434), written on a label of red cloth, shows the name of the sub-prefecture
Lo-yang written with the character 雒 which, according to M. Chavannes, necessitates its
attribution to the period of the Later Han Dynasty. Accepting this evidence, we must assume
that even the westernmost posts along the wall continued to be garrisoned at least until the
beginning of this period, even though the detached posts on the flank may have been abandoned
somewhat earlier. The discovery at the same station of records separated in date by upwards of
a century can be paralleled elsewhere along the Limes.⁹ In this case it must also be pointed out
that the narrow passage ii may have been turned into a rubbish-heap and abandoned for a long time
while the room i, where the cloth label was found, still continued to be tenanted. Among the
refuse found at T. iv. b (see List) were a bronze arrow-head and numerous fragments of silk
fabrics.

Traces
suggesting
entrenched
camp. The position occupied by the tower T. iv. b commanded so open a view of all the low ground
in the neighbourhood that, as I stood there and looked about, I could not help asking myself why