国立情報学研究所 - ディジタル・シルクロード・プロジェクト
『東洋文庫所蔵』貴重書デジタルアーカイブ

could be made out without any difficulty. Apart from the straightness of the line itself, there are clear marks of its direction provided by the watch-towers built along it at average distances of about North-
1¾ miles. The two that I was able to examine here measured 25 to 26 feet square at the base and east wall
rose to a height of 20 to 22 feet in their present condition. They were built of stamped clay, with watch-
layers of the same thickness as the wall which they were meant to guard. But where later repairs towers.
had been effected at the top, the material consisted of sun-dried bricks set in vertical courses,
after a fashion characteristic in these parts of late work.⁸ Cultivation extended right up to the foot
of this wall and in places beyond it, affording clear evidence, along with the big gaps already men-
tioned, that the far advanced decay of this defensive line dated back for a considerable period. At
the same time man's destructive vicinity and the moisture brought by irrigation are likely to have
accelerated the process.

I may conveniently record here some facts which I was able to ascertain as regards the con- Northern
tinuation of this wall further east. They will help, together with what the map shows of the sections angle of
actually surveyed towards Su-chou and beyond, to explain its character and origin. From the side wall at Yeh-
of Chia-yü kuan we were able to lay down the line of this wall quite clearly on the plane-table as far as ma-wan.
Yeh-ma-wan (Map No. 88. B. 1), where a large fort-like structure could be sighted rising at a sharp bend
of the wall. This was said to guard a gate through which passes a direct route coming from Hāmi.
In 1914, when making my way to Su-chou from the Su-lo Ho by a route which lies north of the
desert range flanking the high road, I was able to verify this statement and found it quite
correct. At the same time I ascertained that cultivation, made possible by canals from the
Pei-ta Ho, extends in considerable patches as far as Yeh-ma-wan and even slightly beyond.
This fact plainly accounts for the curious great bend which the protective line of wall makes
here to the north.

From Yeh-ma-wan, where I found that the wall was of the same construction and in the same Continua-
state of decay as near Chia-yü kuan, its line turns to the south-east. Skirting a great expanse of tion of wall
marshy grazing, it continues to a point about 7⅛ miles due north of Su-chou city (Map No. 88. c. 1). to Su-chou
Here, not far from the village of Hsin-ch'êng-tzŭ, I actually examined the wall again on a recon- and Kan-
naissance specially made for this purpose from Su-chou on July 26, 1907. I found its remains there chou.
exactly of the same type and condition as seen near Chia-yü kuan. Its line takes thence an easterly
direction running roughly along the northern limits of Su-chou cultivation. Subsequently, on my
return from Kan-chou in September, 1907, I ascertained that the same wall runs on to the left bank
of the Pei-ta Ho, or Su-chou River, near the north-eastern end of the great Su-chou oasis. Starting
afresh from the right bank near the hamlet of Ai-mén (Map No. 88. D. 1), it is continued north-eastward
to the low desert range along the southern foot of which runs the high road to Kan-chou (Map No. 91.
A. 1). Its further course thence to the south-east keeps more or less parallel to this great line of
communication and along the right bank of the Kan-chou River. I traced it in 1907 and 1914 to
the vicinity of Kan-chou itself; but there is no need to give details here. It will suffice to emphasize
the main fact that this much-decayed clay wall, originally, too, neither solid nor imposing, represents
the 'Great Wall' of Kan-su as known to present local tradition.

The questions to be considered here concern firstly the purpose and origin of this wall, and Wall dis-
secondly its relation to the 'chiusa' of Chia-yü kuan as it exists at present. If as regards the former tinct from
question I am able to indicate here what appears to me a safe and definite answer, it is largely due Han Limes.
to the results which my explorations of 1914 in this region have yielded. They have proved that
the line of the ancient Han Limes did not pass from the great bend of the Su-lo Ho south-east