90 Correspondence. 



or three feet in breadth, will take its course in curves, bearing some- 

 thing like its relation to its breadth of the sketch, — 

 the stream being perhaps 10 or 12 feet deep in the 

 alluvium of its valley, — and its course through the 

 alluvium must be continually changing, as the silt is 

 deposited in the receding angle, and the bank worn 

 away on the salient one. The Kromme and Diep Rivers, 

 larger tributaries of the Kat, take the same course, as their names in- 

 dicate, only with wider sinuations. These latter seem to have some 

 definite relation to the steepness of the course and the volume of the 

 rivers. With allowance for the different degree of erosion exerted on 

 the various rocks, this character seems to be common to all the rivers 

 of the coantry, and makes it unpossible to believe that they originated 

 in channels formed by oceanic currents, or in cracks by earthquakes, 

 or the upheaval of land. As the rocks of the " Lacustrine," or, as 

 they are locally called, from a genus of Eeptile common in them, 

 "Dicynodon formation," are suj)posed by Professor Owen to be 

 referable to about the Triassic Age, it would seem that the time 

 which has elapsed since the desiccation of the lakes is siifficient to 

 account for the vast denudation that I have shown to exist. Some 

 years ago I took advantage of a three daj^s' detention by a freshet of 

 the Sunday's River to collect evidence with a view to the calculation 

 of the approximate amount of denudation affected by it. The data, 

 particularly with the imperfect surveys of the country at our 

 disposal, which are extremely unsatisfactory, led me to a result of 

 about "8000 inch in a century over the area drained by the river, I 

 have little doubt that the estimate is much too great, for it would 

 require little more than 100,000,000 years to effect the amount of 

 our denudation. 



A better guide to the denuding power of our rivers is offered by 

 the Bushman's River. Its circumstances are similar to those I have 

 described of the Orange and Caledon, in that it conveys in its bed a 

 quantity of pebbles of agates (from a layer of amygdaloid in the 

 Zuurberg) over a formation in which they do not occur. After 

 passing through the Zuurbergen, its course to the sea lies through a 

 basin of Devonian Schists and Sandstones, on which rest 300 feet of 

 Sub-Cretaceous Sandstone, Claj^ s, and Marls, and these are surmounted 

 by 50 to 80 or 100 feet of Limestone, which is called " Tertiary " by 

 Mr. Bain, but which, so far as I am aware, does not contain any shells 

 different from those now living in the Indian Ocean,^ — certainly, if 

 any of its species are extinct, they are but few, though the Ostrece, 

 Pectunculi, and Pectines are much larger than any I have seen living. 

 These strata of the three formations I have mentioned have been cut 

 through by the river to the depth of 500 feet, and a width of many 

 miles ; and all this since the deposition of the Tertiary or recent 

 limestone ; and the proof that this has been done by the river is that 

 the agates, evidently rolled by its action, are found all the way up 

 its banks to nearly the level of the plain. 



I was struck with this fact of the occurrence of the evidence 

 of the existence of the river-beds in all parts of the Bushman's and 



