Verneuk Fan, 
81 
the rivers in the district, where they flow from soft, nearly horizontally 
bedded rocks on to the hard granite " of Kenhardt affords part of the 
explanation. The valleys are extremely wide and usually without a 
definite river-bed as long as they lie on the Upper Dwyka shales ; but 
when they reach the tillite below, a bed sometimes appears, and the sides 
of the valley close in ; when they reach the granite there is always a well- 
defined narrow bed with steep banks, and the valley sides become cliffs in 
places, as, for instance, at Modder Gat on the Hartebeest Eiver. Measur- 
ing the fall on the contoured 4-inch-to-the-mile maps made by the 
Topographical branch of the War Office, the grade of the rivers increases 
from about 1 foot to the mile on the Dwyka formation to 8 feet to the 
mile on the granite. The granite forms a hard bar, and the rivers cannot 
widen their valleys in the granite so long as the Orange Eiver cuts down 
its bed at a faster rate than they can keep pace with, which it does 
because it drains a large area with a rainfall greater than that of the 
catchments of the Hartebeest and its tributaries. 
Above the granite barrier, which is always moving up-stream and, on 
the average, falling in level, the rivers have cut their beds to the gentlest 
possible grade for a considerable distance, and their present work is, over 
that distance, confined to cutting plains laterally in the soft rocks behind 
the barrier and distributing thin sheets of fine silt over these plains. 
On the small streams tributary to the Zak and Olif ant's Ylei Elvers, 
in the lower parts of their courses, a similar process goes on behind each 
dyke sheet of dolerite or other hard rock which lies athwart their paths, 
but in these cases the soil becomes very brack, probably from lack of 
sufi&cient drainage when the valleys have passed a certain slope, and 
vegetation will no longer thrive on the mud flats in these valleys as it 
does on the flats of the Zak and Olifant's Vlei Elvers, which are 
periodically thoroughly washed by heavy floods. When there is a dearth 
of vegetation the wind attacks the mud more readily than before, and lays 
bare the underlying shales in places. The wind perhaps does a certain 
amount of actual rock-cutting with the help of the sand grains picked up 
from the floor of the pan, but a careful inspection of many outcrops and 
loose fragments on Verneuk Pan failed to reveal any noticeably sand-cut 
or polished surfaces, probably because the shales and thin fibrous lime- 
stones break up too rapidly under the influence of the weather to allow 
the retention of the evidence. In any case the wind removes material 
from the outcrops, as well as from the dry mud of the pan, as soon as the 
fragments are small enough. 
* For the purpose of this description the ancient rocks of Kenhardt are called granites 
throughout, but they are in fact not often true granites, for they consist largely of gneiss 
and various schists of sedimentary or volcanic origin. Their important character here is 
their comparatively great hardness. 
