388 W. M. DAVIS OBSERVATIONS IN SOUTH AFRICA 



whereby the southern, southeastern, and eastern coastline was determined, 

 as will be more fully set forth further on. 



But the most serious objection that I feel against Schwarz's explana- 

 tion of the great antiquity and long persistence of the straight "main 

 watershed" is that such an explanation entirely overlooks the changes 

 that must reasonably be expected to have taken place by the interaction 

 of the rivers themselves during the long periods of erosion to which the 

 region has been subjected since middle Mesozoic time. The great diver- 

 sity of structure and the well proved changes of level in the region dur- 

 ing its continental existence, probably involving two important cycles of 

 erosion at least, must have contributed effectively to the changes of drain- 

 age area that the competing rivers would have themselves brought about. 

 Such changes are today in active progress at the head of the Vaal river 

 and its branches, where the shorter and steeper east-flowing streams are 

 gaining area at the ex])ense of the longer west-flowing rivers, as will be 

 more fully set forth in the section concerning the eastern escarpment. To 

 take no account of all these possible changes, and therefore to regard the 

 existing watershed as having persisted through geological ages, involves 

 probable errors of the same order as those which were introduced in the 

 interpretation of our Cordilleran physiography thirty years ago by the 

 wholesale suggestion of an antecedent origin for many rivers, small as 

 well as large, in Utah and Arizona. 



The extensive rearrangement of initial drainage lines in a long 

 eroded region, partly through the development of subsequent rivers 

 by headward erosion along belts of weak strata, partly through the en- 

 couragement of headward erosion given to rivers of all classes by favor- 

 ing crust al deformation, is now too well established a procedure in the 

 general natural history of rivers to be set aside, unless by the strongest 

 positive evidence in particular cases. In the absence of such evidence, it 

 may be safely concluded that the existing transverse drainage of the Cape 

 Colony ranges, like the comparable transverse drainage of the Appa- 

 lachians, has come into existence at a date much more modern than that 

 of the original elevation of the region above sealevel. Just how the trans- 

 verse drainage came to be established it is difficult to determine with any 

 certainty on existing information, but there is some reason for thinking 

 that, in South Africa as in Pennsylvania and Virginia, it was not devel- 

 oped until during and after the general peneplanation of the Veld and 

 the upfolded ranges, and that whatever drainage lines — already much 

 changed from the initial lines — had come to exist on the peneplaned sur- 

 face, the weak old streams of that time were profoundly affected by the 

 broad warping and the general changes of level which resulted in the 



