108 RESEARCH IN CHINA. 



its level; and simultaneously the ocean basins have so deepened that sea- 

 level has been lowered 5,000 feet, 1,500 meters. Or, there have been both 

 positive and negative movements not sinking only, but rising also and 

 consequently readjustments of sea-level and of altitudes of various areas. 



The element of time is important in the hypothesis which ascribes 

 all epeirogenic changes of level to subsidence in obedience to gravity; but 

 the movements in Asia are so recent that time fails. The physiographic 

 history of the mountains of Shan-si is plain and consistent. The young 

 slopes and scarps and canyons are obvious facts. However we may 

 differ in a reasonable estimate of their possible age, we can not assign to 

 them an antiquity such that during their development sea-level might 

 have sunk 5,000 feet, 1,500 meters. To do so would not only contradict 

 the direct evidence of comparison with topographic features elsewhere, 

 which are definitely dated by relation to late Tertiary sediments, but 

 would require explanation of the absence of a marine record where there 

 is an erosion record instead. This question enters into the study of the 

 highlands of Asia as a fundamental distinction between two very differ- 

 ent interpretations of the orogeny and epeirogeny of the continent. My 

 views are more fully stated in the following chapter on the hypotheses of 

 continental structure. It suffices here to state that I hold it to be true 

 that when some masses sink notably, whether in continental or oceanic 

 regions, other masses rise notably; while some parts of Asia have certainly 

 subsided during the latest movements, other parts have risen, and the 

 upward movements, which are measured by many thousand feet, are greatly 

 in excess of the downward movements within the continent. 



The mountainous region of northwestern China may be said to consist 

 of the Khingan range north of latitude 40, where that height is the margin 

 of the Mongolian plateau simply, and of the various ranges of Chi-li and 

 Shan-si, which as a group are distinct from the Mongolian plateau. Among 

 the latter are the Nan-k'ou range northwest of Peking, the Wu-t'ai-shan, 

 and the plateau of Shan-si with the Ho-shan. The physiographic history 

 of these mountains and the reasons for believing them to have been elevated 

 during the Fon-ho epoch of the Quaternary have been sufficiently set forth. 



Northwest of the mountain system of Chi-li and Shan-si is a depressed 

 zone characterized by the basins of Ta-tung-fu, latitude 40; of Hin-ch6u, 

 latitude 38 30'; of T'ai-yiian-fu, latitude 37 30'; of P'ing-yang-fu, latitude 

 36; and of the Wei valley, latitude 34 30'. The last stretches far to the 

 west, north of the Ts'in-ling-shan, and is related to the depression of Kan- 

 chou and Sii-chou, which lies along the northeastern border of the Nan- 

 shan. Von Richthofen describes the basin of Ta-tung-fu.* We refer to 



* China, vol. n, p. 359. 



