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

Section IV.—THE POSITION OF THE 'YANG BARRIER'

Identifica-
tion of Nan-
hu with
'Yang
barrier'.
After this survey of the old remains actually traced at Nan-hu it remains for us to consider
whether the ancient frontier station west of Tun-huang, which under the name of Yang kuan 陽 關,
the 'Yang barrier', repeatedly figures in the Han Annals side by side with the more famous 'barrier of
the Jade Gate' (Yü-mên kuan), is really to be located at this site. I found the claim to this proud
identification put forward in a modern stone inscription which some learned Tun-huang Mandarin of
antiquarian tastes had set up by a side of a small shrine, between the south face of the ruined
shrine and the artificial lake feeding the Nan-hu canals. Topographical and antiquarian observa-
tions, gathered in the course of my explorations on the Tun-huang Limes, made this location of the
'Yang barrier' appear to me distinctly probable at the time, and the examination of all Chinese
records at present accessible to me in translation has since confirmed me in this belief. The
references to the Yang barrier contained in the Former Han Annals are so closely bound up with
those made there to the 'Jade Gate' that their evidence can be properly utilized only if we take
into full account the results which my explorations along the westernmost Limes, by combined
archaeological observations and documentary finds, have established as to the true position of the
Yü-mên barrier. The discussion of these results must necessarily be left for a subsequent chapter.¹
It will, therefore, be more convenient here to start, in our inquiry as to the position of the Yang
barrier, from what later Chinese records can tell us.

Yang
barrier at
Shou-ch'ang
hsien or
Lung-lo.
These records are scanty enough; but there is fortunately among them one precise and
distinctly helpful. According to a passage of the Old T'ang Annals quoted by M. Chavannes,¹ᵃ the
Yang barrier was situated 6 li to the west of the sub-prefecture of Shou-ch'ang 壽 昌, whereas the
Yü-mên barrier was 118 li to the north-west of the same place. We learn from the passage also
that the Shou-ch'ang sub-prefecture was the one known as Lung-lo 龍 勒 under the Han which lay
roughly south-west of Tun-huang.² That Shou-ch'ang is represented by the present oasis of
Nan-hu is made quite certain by a passage of the Tun-huang Lu, a short geographical treatise on
the Tun-huang region composed towards the close of the T'ang period, which Dr. L. Giles has
translated from a manuscript found among the great collection of Chinese texts I secured at the
'Thousand Buddhas' of Tun-huang, and to which I shall have repeated occasion to refer hereafter.³
In this passage we find the river of Tun-huang, here called Kan-ch'üan ('sweet spring'), the present
Tang Ho, spoken of as entering the Tun-huang district through the territory of Shou-ch'ang hsien in
the south-west, and a glance at the map shows that by this hsien, or district town, must be meant the
present Nan-hu. There is no other tract near the course of the Tang Ho which during historical
times could have supported a settled population such as the establishment of a hsien presupposes.

Nan-hu
ruins mark
Shou-ch'ang
hsien.
The location of Shou-ch'ang at Nan-hu is fully confirmed by Dr. Giles from other Chinese
records:⁴ 'Shou-ch'ang hsien, called after the Shou-ch'ang, a lake south of the town, was founded
in [A.D.] 521 on the site of the ancient Lung-lo, but a few years later was incorporated in Ming-sha
hsien [a designation of the Tun-huang district under the Later Chou dynasty, A.D. 557–81]. In
619 it was again established, and after a chequered existence definitely disappeared before the close
of the T'ang dynasty.' The reference here to the lake south of the town is of particular interest, as
it proves that Shou-ch'ang hsien must have occupied the same position as the ruined town of Nan-hu.