510 TO KAN-CHOU AND THE CENTRAL NAN-SHAN [Chap. XIV
crossed before approaching the wide inundation bed of the Kan-chou river, has been fully described before.3 This huge bed, close on four miles wide here, presented an appearance strikingly different from that seen at the close of August, 1907. Whereas then the summer floods made several of the channels quite unfordable for laden animals and covered most of the bed with water, leaving exposed only island-like strips of ground, none of the four or five channels now actually filled by the river were over 5o to 6o yards wide and none carried water more than 2 feet in depth. It was striking evidence of the lateness of the season at which the rapid melting of the glaciers and high snow-beds of the Central Nan-shan takes place, and explained at the same time the difficulties which in consequence beset cultivation where it is dependent on irrigation from the lower course of the river. I f the northern ranges in this portion of the Nan-shan were assured of a more adequate rain and snow-fall during winter and spring, conditions would necessarily be very different.
I was obliged to stay ten days at Kan-chou in order to make the necessary arrangements for the further surveys I had planned in the Central Nan-shan. My object was to extend the mapping which we had carried out in 1907 in the mountains near the sources of the Su-lo-ho and Su-chou rivers, by careful surveys of those high ranges and valleys farther to the south-east which contain the head-waters of the Kan-chou river. They were intended, in conjunction with our recent labours in the Etsin-gol region, to complete our topographical record of that large north-western portion of Kan-su which, inasmuch as it sends all its waters into drainageless basins, may well, in respect of its hydrography and general physical conditions, be attributed to Central Asia rather than to China. The need for hired transport and for some measure of local support and guidance involved dependence on the help of the Chinese authorities. I had already had experience of the reluctance shown by Chinese settlers in the submontane oases to venture far into those mountains, and of the official apprehension of danger from Hei-fan-tzû, i. e. Tangut robbers, &c.A I was therefore not surprised at the objections raised at the outset to my plans by the Kan-chou district officers, civil as well as military, and at the difficulties that I encountered in consequence in connexion with transport.
This official attitude was evidently prompted by the desire to avoid all responsibility for my safety, and considering the strength of the motive I should have had to be prepared for prolonged obstruction but for a piece of good fortune ; for the reported approach of an old Chinese friend in the person of the worthy General Ch`ai Hung-shan Aar, just then raised to the military command of Kan-chou as ` T`i-t`ai', gave hope of assistance. I well remembered all the kindness that this genial old soldier had shown me during my visits to Su-chou in 1907,5 and my reliance on his help was not disappointed ; for a stay of a few days after his arrival from Lan-chou-fu on July 2nd, made pleasant by repeated meetings with my old friend, sufficed to clear the way for our departure for the mountains. Meanwhile the peaceful retreat that I had again secured in my old quarters in the temple outside the south-western corner of the city walls (Fig. 267) had enabled me to dispose of much writing work, including accounts, Serindia proofs, &c. I was also able to make use of the facilities that this halt on a newly opened postal line offered for safe and comparatively rapid communication both with Europe and India. At the same time the kind help of the Belgian Mission head-quarters, established at Kan-chou under the Rev. Father Van Eecke, permitted me to gather useful information both about local conditions in the neighbouring cultivated tracts and about the route through the mountains towards Hsi-ning.
I have already recorded in Serindia such observations as my two stays at Kan-chou enabled
3 Cf. Serindia, iii. pp. 1132 sq. In Map No. 46. B. 2 the 4 Cf. Desert Cathay, ii. pp. 290 sq.
name Hei-shui-kuo has been wrongly spelt as Hei-slii-kou 5 See ibid., ii. p. 288.
by a draughtsman's error.