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

624   THE OASIS OF NAN-HU AND THE YANG BARRIER [Chap. XVI

established, and, further, that it is mainly a collection of local folk-lore stories bearing on the mirabilia of the district, interesting in various ways, but not a text to be accepted as a source of reliable historical information.

As regards the origin here given for the name Yang, Dr. Giles himself has rightly observed that it ` does not seem a very probable derivation '. The author of the only other Chinese text, the T`u shu chi ch`êng, in which Dr. Giles has been able to trace a reflex of this story, seems, in fact, to have entertained a similar critical misgiving.19 The suspicion that we may be dealing here with a local ' popular etymology ' of the quasi-learned variety so common in all regions is strengthened by the fact that the Tung Tien, a Chinese text, of which Dr. Giles quotes a passage from a source not specified, gives quite a different derivation : ' The Jade Gate is in the north of the hsien (Lung-lo), and the Yang Barrier is south of the Jade Gate ; that is why it is called Yang (the quarter of light and warmth, i. e. south).' 20 This etymology, too, seems to me to smack of a learned origin, if a non-Sinologist may be allowed to express an opinion on the subject.

There still remain for our consideration two statements of the Tun Huang Lu : one which places the Yang barrier west of Tun-huang city, and the other which declares it to be ` the same as the ancient Yü-mên Barrier '. As regards the first, the fact that no distance is stated makes it impossible for us to determine with certainty whether popular tradition at Tun-huang, towards the close of the ninth century A. D., still located the Yang barrier at Nan-hu, or at some point closer to Tunhuang town. In any case, by that time the quondam frontier-station must have long lost its original significance. Even at the commencement of the Tang period, as we know from the Life of Hsüantsang, the western gate station of the empire, the Yii-men kuan of those times, was established north of Kua-chou and not far from the present An-hsi,2' and the Tun-huang tract had thus passed kuan wai-1`ou, or ' outside the Wall '. We find this transfer also duly noted in the concluding remark of the above-quoted passage of the Tun Huang Lu.

The second of the statements I have singled out from this text for attention seems to imply that, at the time when it was written, popular local opinion at Tun-huang identified the ` Yang barrier ' with the Jade Gate. It is impossible to discuss this statement of the Tun Huang-Lu without going also into the question of the successive positions occupied by the Jade Gate. Hence its consideration may be left until a subsequent chapter, where I shall have occasion to examine the earliest traceable site of the Jade Gate in the light of the archaeological evidence furnished by my explorations along the westernmost Limes. Here it must suffice to mention that, in view of what combined geographical and archaeological facts conclusively prove as to the quite distinct original purposes and positions of the two ` barriers' of Yang and Yü-mên, I am unable to attach to this statement of the Tun Huang Lu the special historical value which Dr. Giles is inclined to assume for it in his otherwise very helpful comments.22

SECTION V.—ABANDONED VILLAGE SITES NORTH OF NAN-HU

On April i i I left Nan-hu in order to regain the Limes line through the desert northward. The collection of ten additional labourers, the maximum contingent which the little oasis could spare, had cost so much time that the start was delayed until noon. The route we followed, under the guidance

Other derivation of name Yang.

Statements about Yang kuan in

Tun Huang Lu.

19 Cf. Giles, J.R.A.S., 1914, p. 717. The name of the Han general who is supposed to have fled through this gate after his defeat is there given as Yang Hsing.

2° See Giles, J.R.A.S., 1914, p. 716.

41 Cf. Julien, Vie, pp. 17, 21 ; Chavannes, Dix inscrip lions, p. 67, note 2. [See now my paper, The Desert Crossing of Hsilan-Tsang, Geogr. Journal, 1919, liv. pp. 27o sq.]

22 Cf. Giles, J.R.A.S., 1914, p. 7r5: ' This is a most interesting statement. Even if made at random or without full appreciation of what it involves, it furnishes, I venture to think, a valuable clue to the mystery which has hitherto surrounded the relation between these two famous frontier gates.'

Yang and Yii-mên barriers wrongly identified.

Start from Nan-hu northward.