This dating from Ming times is also supported by whatever archaeological observations I was
able to gather about the wall at the different points where I touched its line. Its state of decay, generally far advanced, and its total disappearance in places, -when compared with the fair condition of the wall closing the Chia-yü kuan defile, seemed to me to preclude the assumption that its age reached back only to the Manchu period, i.e. the last two or three centuries. On the other hand,
it appears to me distinctly improbable that a wall of stamped clay, of no great thickness and lacking the special protection of those reed or brushwood fascines which so effectively reinforced the ancient Han wall of the Limes, could in localities exposed to wind-erosion or else to moisture have anywhere remained standing to such a height, as e.g. the segment of it seen in Fig. 254 shows, if it had been built many centuries ago. Taking into account the climatic conditions also, which grow distinctly less arid as we follow the line of the wall south-eastwards to Kan-chou and Liang-chou, a terminus a quo is thus indicated which fits well the assumption that the wall- was first constructed under the Ming dynasty. From the fact that Shah Rukh's embassy in A.D. 142o, as we shall presently see, found the police control of the western frontier already fully established at Chia-yu kuan, we may conclude further that the wall, too, was in existence by then. The fifty odd years intervening since the accession of the Ming dynasty allowed ample time for the erection of the wall and the perfection of the system for which it was needed.
There still remains the question of the relation between the wall that we have examined and the line closing the Chia-yü kuan passage, as it now runs from the left bank of the Pei-ta Ho in the south to the Hao-shah-k`ou defile in the north. That this wall in its present condition is distinctly more recent than the former, and also different in construction, I have already shown. It appears to me easiest to account for the relation between the two in the following way. Some time after the accession of the Manchus, perhaps under the great Emperor K`ang-hsi (1662-1723), when the ` forward policy ' leading ultimately to the conquest of the ` New Dominions ' had begun, the long police border of Ming times, then probably already largely in ruins, was replaced by a fortified line at Chia-yii kuan. This, by effectively closing the great highway from the west, could serve a good military purpose in case of need. Though frontier garrisons of importance had already been pushed forward to Bulungir and Tun-huang before 1697,1° the danger of attack from the powerful Dzungar or Eleuth tribes in the north-west was not finally removed until imperial authority was
established over the whole of Eastern Turkestan by their final destruction in 1757.11 •
Under these conditions the construction of the extant closing line of Chia-p.:1 kuan was of distinct military advantage.12 Su-chou must have then, just as during the operations leading to the reconquest of Chinese Turkestan in 1877 after the last great Tungan rebellion, served as the main base of supplies, etc., for the imperial forces. By the defence of the Chia-pi kuan ` chiusa ' Su-chou was protected from any serious direct attack. About that period, I suppose then, the old wall already in existence from the foot of the detached rugged ridge overlooking Chia-yü kuan to the left bank of the Pei-ta Ho was repaired and put into its present state. At the same time this section of the wall, running approximately from south to north, was prolonged to the north-west so as to close also the debouchure of the Hao-shan-k`ou gorge, through which the Chia-yü kuan position could easily have been turned. I must leave it to others who have access to Chinese historical works to ascertain whether the conjectural explanation here offered finds support in original records, as far as such may be available. From the antiquarian and topographical points of view it affords a very likely solution of what otherwise would appear a very puzzling relation.
10 Cf. Ritter, Asien, ii. pp. 37o sq. p. 21 i; Chia-yii kuan was the gathering-place and starting-
" See ibid., ii. pp. 459 sq. point for the army which the Emperor Chien-lung in 1754
12 According to a notice quoted by Ritter, Asicn, ii. sent forth for the conquest of the Dzungar power.
7 D 2