350 
Transactions of the Boyal Society of South Africa. 
striae. Even where the older rocks are hidden by the entire Dwyka series 
there are certain cases in which the directions of the major valleys in them 
can still be made out. 
Obviously the undulating surface thus determined is that of the con- 
tinent at the close of the Dwyka glaciation. In making a restoration of 
the surface as it appeared prior to glacial times the effects of glacial action 
in modifying the topographical features have necessarily to be taken into 
account. That a considerable thickness of rock was removed from the 
continent is evident both from the vast area occupied by the southern 
Dwyka, from its great thickness, and, in the case of the northern Dwyka^ 
from the high proportion of material in it, obviously of local origin. 
Differential erosion is indicated by the occurrence of rock basins, and the 
lower portion of the Kaap Valley has probably been much deepened^ 
firstly owing to the direction of glaciation almost coinciding with that of 
the valley, and secondly to the fact that the strata forming its floor were 
softer and easier removed than the beds composing the high ground to the 
north-west and south-east. Apart from this there is evidence to show 
that, except for the rounding off of ridges and the widening of valley 
bottoms, the intensity of the relief of the ground could not have been 
altered to any appreciable extent. It seems, therefore, not too much to 
presume that the main pre-Glacial drainage lines are represented with but 
slight modification in the major valleys of the glaciated Karroo floor. 
The PalcBozoic Drainage. — There is no portion of the area, indeed, 
where the palaeozoic drainage lines can be more clearly seen than that 
immediately south of Vryburg, for the present river system is practically 
a replica of that of the past. At Brussel's siding the valley is about a 
mile wide and at least 300 feet deep, but at Taungs it broadens and the 
depth must have been over 800 feet. The main artery extended down the 
Kaap Valley, receiving a small tributary from the north-west at Boetsap 
and a larger one near Bead's Drift. The channel probably continued 
to near Prieska, and then turned southwards in opposition to the course 
of the Brak Eiver, passing between the T'Kuip hills and the ridges north 
of Beer Vlei ; further into the Karroo its exact direction cannot be traced. 
The evidence regarding the area north-west of Prieska is not quite con- 
clusive, but so far as it goes does not support the assumption that the 
valley of the Orange Eiver from Prieska to Kheis is cut along a pre- 
Karroo valley. The drainage from the southern part of Hay and from 
the Doornberg Eange in all likelihood entered the main valley at Prieska. 
Another important branch came down from the Transvaal ; it crosses 
the present Vaal Eiver, first above Warrenton, again at Windsorton, 
receiving two tributaries from the east, and leaves it at Pniel, striking in a 
south-westerly direction to Douglas, where it joins the main stream in tha 
Kaap Valley. 
