The Evohition of the Biver System of Griqualand West. 349 
smooth outlines stretching south-eastwards parallel to the river to a point 
a little north of Omdraai's Vlei. The altitude of the Doornberg exceeds 
4,200 feet at several points. 
Towards Kenhardt is another tract of ^elevated country trending, like 
the Doornberg, north-west and south-east, but not quite so high, and 
composed chiefly of Kheis quartzite and mica-schists. South-east of 
the areas of high ground referred to is a wide undulating tract formed 
principally by the Dwyka series, the basal portion of which, resting 
unconformably upon the underlying formations, is the celebrated Dwyka 
conglomerate or " tillite " of glacial origin. In this area, composed of 
material of late Palaeozoic age, are numerous inliers of older rocks, prin- 
cipally the volcanics of the Ventersdorp system. These inliers form, with 
rare exceptions, ground higher than the surrounding tillite, and obviously 
represent the ridges of the pre-Karroo land surface, the troughs and 
valleys in most cases remaining buried beneath the glacial deposits. 
Such inliers are found west and south of the Doornberg, and form a 
regular chain extending from Omdraai's Vlei to the Orange River near 
Hopetown. The older rocks form high ground in the angle between the 
Orange and Vaal Rivers, and again in the neighbourhood of Schmidt's 
Drift. Between the Harts and Vaal Rivers they constitute a belt 
broadening to the north-east and covering a v^ide area in Vryburg and 
south-western Transvaal. 
Towards the south and south-east of the area the general level of the 
country rises, the covering of Karroo sediments and intrusive dolerite 
becomes thicker in that direction, and the scenery is that typical of the 
Karroo. 
A critical examination of the available data shows that the extremely 
slight southerly inclination of the Karroo beds over the greater portion of 
the area in question is most probably an original feature due to conditions 
of deposition. In the extreme south and south-east the dip of the strata 
becomes appreciable, the fall of the Karroo floor is rapid, and it is likely 
that there has been a slight sinking of the crust in this direction. Omit- 
ting this tract, however, it is clear that over almost the whole of the area 
in question the elevation subsequent to the deposition of the Karroo 
sediments was of uniform amount, and consequently that there was no 
warping either of the Karroo beds or, obviously, of the pre-Karroo surface 
underlying them. 
Where denudation has been insufficient to lay bare this surface entirely, 
remnants of the Karroo formation still occupy the bottoms of the depres- 
sions in it, and thus the ridges and valleys of the palaeozoic land mass can 
be determined. Where the tillite has but recently been removed the 
surface of older rock possesses in a remarkable degree the characteristics 
of a glaciated region, while in many places the rocks still retain glacial 
23 
