Explanation of double walls in Hao-spän k ou defile.
Early notices of Chia-yü kuan Gate.
Karaul passed by Shah Rukh's embassy.
1124 THE NORTH-WEST FRONTIER OF KAN-SU [Chap. XXVII
It is only _left for me to recur to the curious detail previously noted about Hao-shan-k`ou, the two walls facing each other across the middle of the gorge. In the lower and later wall, with its parapet facing south-west, we obviously have an outwork of the Chia-yü kuan ' barrier ' wall, rendering the portion outside Huang-ts'ao-ying more secure. As regards the upper and older wall facing to the north-east, the most likely explanation appears to me to be that it was built for the purpose of protecting the high road to the west of the Chia-yü-kuan Gate from attacks of robbers and the like. It is very improbable that the Chinese, even during the period of seclusion in Ming times, could have left the great route westwards without such protection as advanced stations would assure. The account of Shah Rukh's embassy distinctly mentions its reception by Chinese officials 13 at several points ' in the desert ' before reaching the Chinese frontier at Chia-yii kuan. On the other hand, we see from Benedict Goes' narrative how much need there was for such protection in view of the general insecurity prevailing in his time on the route from Hami to Su-Chou.'} In addition there is the exact analogy of the advanced watch-towers which were maintained in Han times beyond the point where the Lou-lan route left the Tun-huang Limes," and of those which Hsüan-tsang had to pass after going outside the Yü-mên of early Tang times.' If the Hao-shan-leou gorge was thus closed during Ming times by a wall preventing access to the high road, the latter was rendered safe, at any rate against attacks from the north, for some distance beyond the Chia-yü kuan gate ; for the very precipitous range on the north is crossed by no practicable track until we come to the gorge through which the Po-yang Ho has cut its way towards the small basin of Hu-hai-tzti (Map No. 86. D. I).
Having now completed our survey of Chia-yii kuan and cleared up the origin of its walls, we may turn to the notices which early travellers from the West have left us of this famous gate into China. That Marco Polo, who must have passed the defile of Chia-yü kuan on his way from Shachou to Su-chou, makes no reference to the site is easily understood in the light of the preceding explanations. At the time of his passage the walls closing Chia-yü kuan and bending round the westernmost border of Ho-hsi were not yet built. It is true that the Chinese mission to Khotan of A.D. 938 mentions a defile called the ' Gate of Heaven ' in a position a hundred li to the west of Su-chou, which manifestly corresponds to that of Chia-yii kuan." But even if a watch-station was maintained here after the Mongol conquest of China, it could not have been of any special importance in. the days when the great Kublai Khan's reign kept China open to intercourse with the West.
Conditions had completely changed when the embassy sent by Shah Rukh passed here in A.D. 1420.16 The envoys, accompanied by many merchants, spent twenty-five days on their way through the Great Desert from Hami. On the last few marches ' near the frontier of China Proper' they were fêted in great style by Chinese officers who had come to meet them. ' On their arrival at a strong castle called Karaul, in a mountain defile, through the middle of which the road passed,19 the whole party was counted and their names registered before they were allowed to proceed. 'l'hey
" See Yule-Cordier, Cathay, i. p. 273.
" Cf. ibid., iv. p. 240.
16 See above, pp. 638 sqq., regarding the towers T. 1, n.
16 Cf. Julien, Vie, pp. 17, 24 sqq. ; below, PP. 1144 sq.
" Cf. Rémusat, Ville de Khotan, p. 77. A ' passage named the Gate of the Jade Stone' is mentioned in the same report a hundred li to the west of the Gate of Heaven: There is a temptation to take this as referring to Yü-mênhsien, but the distance in that case is greatly under-estimated.
14 See Yule-Cordier, Cathay, i. pp. 273 sq., for an abstract
translation from ' Abd-ur-Razzaq's ' narrative of the mission ; also the translation by E. Rehatsek in Indian Antiquary, ii. (1873) pp. 75 sqq.
19 Rehatsek's translation, Indian Antiquary, ii. p. 76, describes the position more closely : ' Qarawul is a very strong fort among the mountains, and can be entered only on one side by a road which also leads out of it on the other.
Cf. also a translation of the same narrative which Ritter, Asien, ii. p. 21 I, quotes from Thévenot, Relation de Voyage, Paris, 1696, t. n. fo1.,3 (not accessible to me).