THE KARROO SYSTEM 179 



the Witteberg quartzites and the Dwyka boulder-bed, 

 lying conformably to both. There can be hardly any 

 doubt, therefore, that the lower shales are a definite 

 group of beds present at the base of the Dwyka serie.s 

 wherever it lies conformably upon the Witteberg beds. 1 



It seems probable that the Southern Dwyka, which is 

 thicker and more uniform in character over wide areas 

 than that in the north* represents the silt and boulders 

 carried southwards by water and floating ice and de- 

 posited in water. It includes some thick bands of shales 

 with few pebbles and boulders, traceable for considerable 

 distances. The Northern deposits are, .on the other 

 hand, extremely variable both in their nature and thick- 

 ness, but generally speaking show two fairly distinct 

 phases. In one of these the tillite forms a bed seldom 

 more than thirty feet in thickness and is succeeded rather 

 abruptly by soft shales. In the other the deposit is 

 much thicker and more or less regularly bedded, consist- 

 ing of alternations of conglomerate and shaly material, 

 shale, and sometimes limestone, while the shales above 

 may enclose occasional boulders. 



The second type is especially characteristic of the valley 

 of the Harts River and the portion of the Vaal Valley 

 that forms its south-westerly extension ; it fills a wide 

 depression which existed as a valley in Pre-Dwyka times 

 and which is now in process of re-excavation. This 

 type of deposit is most probably in great part or wholly 

 of fluvio-glacial origin, thus resembling the southern 

 Dwyka, while the first type may well have been formed 

 on land or in very shallow water at the end of glaciers 



1 Schwarz, G. C., viii., pp. 88, 89. 



