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

Desert vegetation scarce on An-hsiHâmi route.

710   THE JADE GATE BARRIER   [Chap. XIX

itself keenly felt at any of the watch-stations which we may assume to have been established on the ` new route ' leading through the barren wastes of the Kuruk-tâgh northward. It is this very difficulty, together with the almost total want of grazing, which is brought home to every traveller as he passes along the present Chinese ` high road ' from An-hsi to Hâmi and has to pay the somewhat exorbitant charges made for fuel and reed straw at all the halting-places in that desert of gravel and stone.'*.

All the information that we have about this region, now known to the Chinese by the vague but convenient designation of the ` Pei-shan ' Gobi, proves that the scarcity of vegetation, just like that of water, increases the further west one crosses this desert. Hence it appears to me certain that, even though desiccation is likely to have made conditions worse since Han times, the difficulty about fuel must have been felt quite as much then on the ` new route of the north ' as it now is on the An-hsi-Hâmi route which has taken its place.' Though the latter runs on a line that is on the average some 6o miles further east than the one likely to have been followed by the ancient ` new route ', I can well imagine that any attempt to maintain a system of fire-signals over the An-hsi-Hâmi road, before the telegraph wire reached it late in the last century, would have met with exactly the same difficulty which our document from T. xv. a. v records.

Whether a systematic search of the desert to the north-west might yet reveal remains of the modest hovels likely to have once stood at the successive- stages of the ` new route ', that is, of their refuse-heaps, is a question which I cannot answer, but which I should like to recommend to the attention of future explorers. I may, however, point out that we have, perhaps, a reference to one of the nearest of these conjectured roadside stations of the ` new route ' in the previously quoted record T. xv. a. iii. 43, Doc., No. 452 (Plate XIII).16 It directs the issue of food-rations for three marches to a person dispatched from ` the fortified post of Yii-cla`ang-lo' to Ta-chien-tu, i. e. the western terminus of the Limes wall. The distance to that point from T. xv. a accounts for two marches only, thus proving that the starting-place lay one march beyond. This fact and the mention of a service camel being provided as a mount for the journey seem to justify a presumption that Yu-ch`ang-lo was a post in the desert where, owing to scarcity of water and grazing, it was preferable to station riding camels instead of horses for official use. Finally, it may also be mentioned that it would be easier to account for the silk strip, T. xv. a. iii. 57, with its Brâhmi inscription, having been left behind here by some trader from the West, if we may assume that the place where it was discovered was the `gate' station for an important new trade route, and not a mere watch-post amongst others guarding the line of the Limes.

Yii-ch`anglo located on ` new route ' (?).

SECTION VII.—THE GREAT MAGAZINE OF THE LIMES

The series of watch-towers T. xv-T. xvii and the continuous stretch of wall connecting them form a link, as it were, between the two well-defined sections of the Limes which I have distinguished above : the first running across a succession of spring-fed marshes and the second skirting lakes and swamps filled by the Su-lo Ho waters.1 My remarks on T. xv-T. xvii can be brief. I may, therefore, conveniently prefix them to the account I shall have to give of the most notable ruin on the second of these sections.

" For a brief account of this desert route, cf. Desert Cathay, ii. pp. 239 sqq., and below, chap. xxvitt. sec. i ; see also Maps Nos. 73, 76, 77, 8o, 81.

le We cannot suppose that the watch-station referred to may have been situated along the Lou-lan route beyond the

western end of the Limes; for fuel is plentiful along it all the way down the depression of Besh-toghrak until the dried-up bed of the ancient salt sea is reached beyond Kum-kuduk.

1" See above, pp. 699 sq.

1 Cf. above, p. 663.