PRESIDENTIAI. .\1)|)RI<:.SS SECTION P.. 37 



River, and perhaps elsewhere, towards Rhodesia. A geographi- 

 cal change like this. Dana thought, would certainly bring about 

 a refrigeration of the climates of the continent, and thus affect 

 a very large part of the earth's surface. Ice sheets resulted, 

 and from these snow-clad regions glaciers brought down the 

 boulders and mud to form the Dwyka conglomerate. These 

 gradually filled the deepest depression in the south-west of the 

 Cape Colony to a thickness of i.ooo feet of glacial debris. 



Glacial striae show that this came from the great areas of 

 pre-Karroo rocks in the north, and it is possible that high 

 plateaux, long raised from sea-level, existed in this direction, 

 and would be planed down by the action of the ice. How 

 far north this snow-sheet extended is hard to determine in the 

 absence of evidence of glaciation, and the answer may be foiuifl 

 in North-Western Rhodesia. 



The present configuration of that province suggests some 

 such planing action in early Karroo times. The landscape that 

 resulted was much as we now find it there — wide rolling plains, 

 dotted by isolated hills of 200 feet in height, flat distant horizons, 

 and sluggish rivers. There is good reason for believing that 

 much of that country was afterwards covered by some one or 

 other member of the Karroo deposits, that have since and com- 

 paratively lately been removed by denudation. 



The growth of Gondwanaland commenced with an accumu- 

 lation of great boulders of the contiguous rock masses. Glacial 

 origin is assigned to these deposits in the Cape Colony and Trans- 

 vaal (Dwyka), and to those in lower Talchir stage in India, 

 the conglomerate in Queensland, Tasmania, West Australia, and 

 at the base of the Muree Series in New South Wales, the 

 Bacchus Marsh boulder beds in Victoria, the Orleans con- 

 glomerate of the corresponding Santa Catherine system in 

 Brazil, and the Falkland Islands, all showing how widespread 

 were the causes of ice movement. Boulder beds occur in a 

 corresponding position in the Karroo beds in North-Western 

 Rhodesia and North-Eastern Rhodesia, among which there is 

 so far no definite evidence of glacification, but they point to an 

 origin in storm deposits, screes and surface litter among hilly 

 regions. 



The conditions, which followed, led to the accumulation of 

 the Beaufort beds, of coarse sandstone, as storm deposits, then 

 the Stormberg series of fine sandstone and shale. The highest 

 series is the Volcanic, forming the flat tops of the mountains, 

 and reaching as great a thickness as 4,500 feet in Basutoland, 

 sometimes interbedded with sandstones and tuffs. 



Here again is a coincidence with other Provinces of the 

 continent, for volcanic rocks now form the summit of the sys- 

 tem in Brazil, Rhodesia and India — tlie Deccan trap. 



It is surmised from an examination of the grains of the 

 forest sandstones in Rhodesia that desert or sterile conditions 

 occurred towards the close of the age. Some of these. beds are 

 remarkably fine, and resemble marls, though they are highly 



