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

The tower is known to the people of the oasis by the name of T'ou-tun 頭 墩, 'the chief Modern
(or first) tower'. This application of a special name, like the continued occupation of the post designation
down to later times, suggests that this section of the Limes, otherwise wholly deserted, is crossed T. XLIII. l.
somewhere in the vicinity by a track still occasionally frequented. And in fact, one of our men
from Ying-p'an averred knowledge of a route leading northward to a place vaguely described as
'a hundred li off in the Pei-shan'. But for what purpose the local people visited it we did not
succeed in eliciting.

The examination of the wall to the north-west of T. XLIII. l. and of the ground immediately Remains
beyond it revealed two curious features. At a distance of about 40 yards from the tower survived near tower
what looked like the reed fascine foundations of four small structures built within, and immediately T. XLIII. l.
against, the Limes wall. They measured about 13 feet from north to south and were separated
from each other by intervals of about 18 feet. Could these mark the position of shelters erected
at a point of the wall which seems to have required special watch and guard? This need may have
been occasioned by the proximity of a succession of swelling gravel ridges, which here stretch from
the north to within fifty yards or so of the wall and threatened its security. It is evident that they
greatly facilitated unobserved attack. On the most westerly of these ridges were found two massive Stacks of
stacks of tamarisk brushwood, at points about 50 and 60 yards to the north-west of the previously brushwood
mentioned position on the wall, as shown by the plan (Pl. 15). They measured 13 feet square and reeds.
at the base and still rose to over 7 feet in height. On a second ridge to the east and at about the
same distance, we found a somewhat smaller stack of reeds, containing also horse-dung. It
occurred to me, as I looked at these stacks, that by setting them on fire, the foreground would be
well lit up at night and risks from sudden attack at this point much reduced. But, of course, the
possibility of other explanations must be kept in view, e.g. the intention to use them for signal fires.

Beyond T. XLIII. l the wall could be followed eastwards for half a mile. Then it became Line of
completely lost in a wide depression of soft eroded clay covered with closely packed tamarisk-cones. wall lost
We continued our search over a distance of about three miles beyond the tower without coming in depres-
upon traces of the wall. Nor had Lāl Singh been able to find any when he pushed an independent sion to
reconnaissance north-eastwards, to within sight of the Sai marking the foot of the Pei-shan. This East.
and my previous experience of the very deceptive nature of such ground decided me to seek a fresh
base to the east from which we might endeavour to locate the continuation of the line of the Limes.
Practical considerations as regards water, doubly important now with the advance of the season,
and the necessity of husbanding time in our move towards Su-chou and the Etsin-gol, indicated
as a suitable place for this purpose the next halting-place with a well on the route to Su-chou.
Before, however, recording the work carried out from that point, some remarks may be offered
here concerning the geographical position of the Hua-hai-tzŭ basin and its bearing upon the line
followed by the Han Limes.

The geographical interest of the Hua-hai-tzŭ basin, as it may conveniently be called from the Position of
designation of its single oasis, is due to the fact that it forms the terminal depression of a small Hua-hai-
but independent and typical drainageless area, intermediate between the far larger depressions tzŭ basin.
in which the waters gathered by the Su-lo-ho and the Etsin-gol finally disappear. The northern
limit of this area is marked by the southernmost hill range of the Pei-shan, the southern by that
portion of the outermost Nan-shan which lying approximately between the 97th and 98th degrees
of longitude separates the catchment area of the Su-lo-ho from that of the Pei-ta-ho or Su-chou
river (Map No. 41. c, D. 1, 2).

On the west and east the limits are not so sharply defined. Towards the Su-lo-ho the watershed
is formed by an outlier of the Nan-shan which descends west of Ch'ih-chin-sê and flattens out in
3 E 2