CHINA, MONGOLIA, AND JAPAN. 9 



Were near enough to the hmestone to be cemented mto a hard mass by the waters 

 flowing from it. 



The bed of the Yangtse must have been cut to about its present depth, when a 

 diminution of its average fall took place, permitting the formation of these terrace 

 deposits. Subsequently another change, by increasing the fall, caused the river to 

 scour out, again, the greater part of the valley. As with the river so Avith the 

 Tungting lake ; this large sheet of water, which then occupied all the plain of 

 Hupeh and Hunan, must have been filled up with the terrace deposit, the remains 

 of which now form its shores. With the returning increase of fall, the lake was 

 scoured out by the rivers Yangtse, Han, Siang, and Yuen. Since this erosion, it 

 would seem probable that the velocity of the current has slightly diminished, as 

 the material brought do^vn by these rivers has converted nearly nine-tenths of the 

 former lake into dry land. A large part of this lake-plain is said, by ancient 

 Chinese writers, to have been an immense marsh where it is now cultivated land. 



We have, at present, no observations to show whether the oscillations of Central 

 China, which are thus recorded in the Yangtse Valley, were contemporaneous with 

 the raising of the western edge of the delta-plain ; but whether they were or not, 

 the cause which was exerted across the whole breadth of China, must be looked for 

 in a vertical movement, either in the Tibetan highland or along the eastern coast. 



A remarkable instance of the formation of a deposit of fine material, in the 

 swiftest part of the river, is observable in the first rapids, just above the Ichang 

 gorge. Granite rocks rising to the surface, near the shore, form an obstruction to 

 the current, which is here from fifteen to eighteen miles an hour, causing eddies in 

 their lee, in which a constant x^i'ecipitation of sand takes place. Banks of quick- 

 sands are thus formed, their tops almost even with the surface of the river. Their 

 sides, too steep to remain at rest, are constantly being washed away, and as con- 

 stantly replaced by the freshly precipitated material. At low water these banks 

 line the shores, and, during the high water season of 1863, 1 noticed one more than 

 half a mile long, and twenty-five or thirty feet above the river; the result of some 

 previous very high freshet. 



