THE OPEN VELD 423 



swells to the streams and along the streams to the rivers. Not only the 

 stream courses, but the general surface of the plain, is and has long been 

 reduced to grade. The delicacy of drainage organization here involved 

 is worthy of close attention; it can be explained only by the long undis- 

 turbed and continued action of erosive agencies, and its attainment may 

 therefore be taken, along with other features above mentioned, as evi- 

 dence that the region of the Veld has really been undisturbed by crustal 

 movement through a long period of time, and that it is actually a plain 

 of long continued erosion. It is distinctly not a dissected peneplain in 

 the areas here considered; the streams do not flow in narrow valleys that 

 are eroded beneath the level of the peneplain, as is so often the case in 

 other parts of the world, but flow in mere channels through the very shal- 

 low and very wide Open valleys appropriate to old age. 



The absence of rill channels on the slopes of the faint swells is a char- 

 acteristic peculiarity of the Veld very appropriate to old age. This 

 feature is probably to be explained, in so far as a peneplain in a normal 

 climate is concerned, by the failure of the gently sloping surface of the 

 residual swells to intersect the ground-water surface, and hence by the 

 failure to develop springs and small streams, as already suggested. In the 

 case of the Veld, the absence of rills is also favored by the dryness of the 

 climate. If contour lines were drawn on a good map of the district, they 

 would, in the absence of rill channels, sweep around the swells in large 

 curves ; the many reentrant angles by which contour lines indicate the re- 

 peated divarication of streams and rills in a surface of stronger relief 

 would here be wanting. The very broad convexity of the faintly arched 

 swells is another matter of interest in showing how far the sharp ridged 

 divides, appropriate to the action of streams in an early stage of the cycle, 

 are now replaced by flat and broadly rounded divides on which normal 

 stream action is replaced by some other process — probably wet weather 

 wash rather than soil creep — which becomes dominant in the late stages 

 of the cycle. It should be noted that the flat divides swell from 30 to 100 

 feet above the pale valleys, and from this it may be fairly inferred that 

 peneplanation here can not be due to the lateral swinging of streams, but 

 simply to the general wasting and washing processes of subaerial degra- 

 dation. The few "pans" or undrained depressions that I saw seemed to 

 be distinct departures from the rule of normal degradation that obtains 

 so widely in the Veld, as will be more fully stated below. 



EFFECTS OF THE DRY CLIMATE 



Some of the more special features of the Veld, considered as a pene- 

 plain, may now be mentioned, and, first, certain features associated with 



