414 W. M. DAVIS — OBSERVATIONS IN SOUTH AFRICA 



yet reported in connection with the northern part of the Dwyka tillite in 

 the Transvaal. In the peripheral parts of other glaciated areas terminal 

 moraines and overwashed, gravels are found bordering the till sheets, 

 but no such accumulations are yet reported from the southern part of the 

 Dwyka in Cape Colony. It is possible that a search with these matters 

 in mind might aid in their discovery. 



The climatic conditions under which the Dwyka ice-sheet was formed 

 offer a most interesting subject for inquiry. The chief factors to be con- 

 sidered in this connection are : Altitude of the Dwyka land surface above 

 the sealevel of its time ; distance from the continental margin, as the con- 

 tinent was then shaped ; form of the Dwyka area and its relation to neigh- 

 boring mountains and highlands of Permian time; various atmospheric 

 factors, such as composition, temperature, and circulation; certain astro- 

 nomical factors; and the latitude of the glaciated area, as latitude was 

 then arranged. 



THE TOPOGRAPHY OF SOUTH AFRICA IN DWYKA TIME 



Evidence has already been given to the effect that the floor on which 

 the Dwyka rests had been extensively eroded and reduced to moderate or 

 small relief in pre-Dwyka time. If the region then drained into the sea, 

 and if it were not farther from the coast than it is today, it must have 

 been a lowland, for only as a lowland could it have been eroded to small 

 relief. If the Dwyka region had been of interior drainage in pre-Dwyka 

 time it might have stood as high or higher than it does now; but in that 

 case it must have been better inclosed from the ocean than it is at present. 

 Such inclosure would have required an extension of the continent, partic- 

 ularly on the east and south, where the Dwyka now reaches or approaches 

 the present coastline, and as a consequence of such continental extension 

 the Dwyka area would have probably been drier than it is today. 



As between the two suppositions of a lowland draining to the sea and 

 an interior plateau, the occurrence of marine (Devonian) fossils in the 

 Bokkeveld series, 2,500 feet or more below the basal members of the 

 Dwyka in the Karroo district, is in favor of the former; for the Bokke- 

 veld, Witteberg, and Dwyka series are conformable there, and the transi- 

 tion from the marine conditions of the first to the glacial conditions of 

 the last, while all three formations still lay horizontal, and while deposi- 

 tion continued without interruption, can hardly have involved great 

 changes of continental altitude. 



There do not appear to have been any mountains during Permian time 

 in the parts of South Africa here under consideration. The many east 

 and west ranges along the southern border of Cape Colony had not then 



