Yol. 57.] CHA]!fGES IN NORTHERN AND CENTRAL ASIA. 249 



the area from below sea-level the rainfall for a considerable period 

 must have been so great, that all these basins overflowed their 

 watersheds until they became freshened ; while, since the evapora- 

 tion has exceeded the precipitation, time enough has not elapsed 

 for salt to accumulate to any great extent. It is difficaU to 

 explain this upon any theory, except that of a recent and gradual 

 elevation of the region, which left under water a sufficient extent of 

 the low-lying portions to secure the increased precipitation implied. 

 This points to a deposition of the loess by water rather than by 

 wind, for the seolian hypothesis implies a dry condition where 

 certainly there was just the opposite. 



At Trebizond, on the Black Sea, we found the clearest direct 

 evidence of this great continental subsidence. Behind that city 

 a prominent mass of volcanic rock rises precipitously to a height 

 of 850 feet. Clinging to the sides of this, at an elevation of 

 650 feet, is a deposit of fresh-looking beach-gravel, about 100 

 f 3et thick, and extending for nearly half a mile. Scattered patches 

 of gravel extend up to 750 feet, but none could be found higher. 

 This deposit was subsequent to the entire rock-erosion of the region, 

 and certainly belongs to very recent geological times. A subsi- 

 dence to that extent would submerge all Southern Russia, and 

 help to account for the extensive loess-plains which form the most 

 striking characteristic of that region. 



Another piece of direct evidence of such a subsidence in Southern 

 Russia was found in the lower part of the Dariel Pass, on the north 

 side of the Caucasus Mountains. A careful examination of this 

 pass shows that it was never occupied by a glacier. Yet for several 

 miles below the narrowest part of the pass, where the elevation is 

 3000 or 4000 feet above the sea, the valley, after the rock-erosion 

 had proceeded to its present extent, was re- filled with clay, sand, 

 and gravel to a depth of 300 or 400 feet. It is noteworthy that 

 these deposits are finest near the bottom. The facts would not 

 necessitate a general depression to sea-level, but only a differential 

 subsidence, increasing towards the axis of the range. But, as 

 there is free drainage to the north, and no chance for local ob- 

 struction of the channel, extensive recent changes of level must 

 have taken place in that region. 



We found that the geologists in St. Petersburg agreed with the 

 conclusion to which we had unwillingly come, that there was no 

 general glaciation of Siberia south of lat, 56^ N., and that however 

 much the wind had to do with the loess-deposits of China, it was not 

 adequate to account for the Icess of Turkestan and Russia. This 

 gives special significance to the recent discovery of Palaeolithic im- 

 plements beneath the loess at Kiev, a discovery described by Prof. 

 Armashevsky in a pamphlet prepared for the Vllth International 

 Geological Congress in 1897. The whole situation was explained 

 to us, in November 1 900, by him on the spot. The general level of 



Q. J. G. S.No. 226. s 



