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

Chinese ta-lu-ti, or 'men knowing the high road (sic)', spoke of two routes by which to reach the
coal-pits of Mou-wo (Map No. 42. B. 3), some four or five marches ahead. But on September 2nd,
the day of our start from our airy temple quarters outside Mao-mei, we had still to keep together ;
for the crossing of the flooded river to the west of the town proved very troublesome and took nearly
half a day to accomplish. Though the summer flood had already greatly subsided, there was water
still flowing over most of the bed, fully a mile wide. In its western portion it was so deep and the
current so strong that the camels had to ford it unladen, and high carts had to be used for taking
the loads across, as well as most of the men. We then separated, after a short night's rest at a temple
near the western edge of cultivation on the left bank, Lāl Singh moving off for Mou-wo through
the barren low hills north-westwards.

The alternative route which I was to follow with Muḥammad Yāqūb and Afrāz-gul proved Return to
in the end to be a track leading to Mou-wo from the extreme north of the far stretching Chin-t'a Limes near
oasis. In order to reach it, our 'guide' took us along the foot of the low hillocks overlooking the Pei-ta-ho.
left bank of the Pei-ta-ho, almost as dry now as when we had seen it in May. Thus our first march
brought us back once more to the line of the Limes west of Mao-mei and allowed us to trace it for
some distance beyond the point, near T. XLVI. a, where we had first struck it from the side of Chin-t'a
(Map No. 42. c. 4). The line for about four miles west of this point could be followed quite clearly,
in the form of an agger constructed of rough stones, a material supplied in abundance by the low
decomposed rocky ridges on which the wall stood. In some places where large rough slabs had been
used, this ancient border wall still rose to a height of 7 or 8 feet. Two completely decayed watch-
towers, T. XLV. a, b, which we were able to examine before darkness came on, were found to be
built of cultivation clay, with thin layers of tamarisk brushwood interposed. Ancient pottery debris
was picked up near them.

The track by which our guide took us next morning passed through a belt of luxuriant scrub Last patch
along the river-bed ; it enabled us, however, as we proceeded, to see other mounds marking decayed of cultiva-
watch-towers, T. XLV. c, d, e, on the line where the Limes agger skirted the foot of the low hills. tion below
I was unable to visit them ; for increased pain in my left leg, the result of the severe strain to which Chin-t'a.
I had subjected it by doing the long marches between Kan-chou and Mao-mei on horseback, had
obliged me to abandon all attempts at riding and to try being carried, instead, on a kind of bed lashed
to the back of a camel. This mode of progress necessarily tied me to the slowly moving baggage
train, and soon proved so trying that I was glad in the end when the track taken by our timorous
'guide', instead of leading us, as I wished, to the north-west, where Mou-wo lay, brought us to
the small outlying patch of cultivation of Chiu-hsi-tun belonging to Chin-t'a. It was the last
chance available of securing wood for improvising a pony litter. There, through the kindness
of a friendly villager who sacrificed some pieces of timber from his roofing, I managed by that
evening to have a conveyance constructed which, in spite of frequent break-downs, carried me
during the next two months, safely and in comparative comfort, across the Pei-shan and along the
T'ien-shan.

On the morning of September 5th we set out at last for the barren low hills to the north, our Passage
'guide' having apparently reinforced his courage for the task by locally gathered information. of last
After a march of about two miles across a bare flat of clay, partly wind-eroded, we reached the foot Limes. remains of
of a stony Sai and here passed for the last time through the line of the Han Limes. It took the
form here of a low and badly decayed mound, with remains of two ruined towers, T. XLV. f, g, visible
to the east and another about four miles off to the west. There could no longer be any doubt that
the agger continued all the way across the waterless desert to where we had traced it among high
dunes north of Ko-ta-ch'üan-tzŭ (Map No. 42. A. 4). Curiously enough those of the people of
3 x 2