president's address. 13 



for holding water underground. The date at which the exposure 

 of the Dolomite began is not known, but the Kuruman and 

 Molopo still traverse Karroo beds for many miles above their 

 confluence,^' and it is possible that in early Tertiary times their 

 valleys were mainly in that formation.^* The Kuruman Eiver 

 has no tributaries between the confluence of the Mashowing and 

 that of the Molopo, a distance of over 100 miles, and water 

 flows between these points only at intervals of years. 



The present state of the Molopo is sometimes regarded as 

 conclusive proof of decreased rainfall in the southern Kalahari, 

 even during the human occupation of the country. It is doubtful 

 whether native tradition is strong enough to substantiate this, 

 and the meaning of the word Molopo (a common spelling) is, 

 according to Mr. Stigand^^ " an intermittent creek or back- 

 water." The deeply-cut trenches and the gravels observed ai. 

 certain parts of its course prove that more water ran down the 

 valley at some time in the past than now,^" but the development 

 of the sandveld, which greatly diminished the run-off from the 

 region, and the replacement of an impervious floor by dolomite 

 in those parts of the catchment with the higher rainfall, which 

 must have had a similar effect, may well account for the observed 

 result without calling in a period of increased rainfall of long 

 duration. 



There have been no discoveries of fossiliferous beds which 

 date the commencement of the accumulation of the sand in the 

 Southern Kalahari. The remains of mollusca belonging to living 

 ■species at Witkop in Gordonia"*' and at a few other places are 

 more recent than that event. The pink and red marls underlying 

 the silcretes in parts of Bechuanaland appear to be the oldest 

 known post-Karroo rocks in the region, but they have as yet 

 yielded neither bones nor shells.^' 



The Kalahari sand which covers so large a region in the 

 Kalahari itself and has outlying representatives of great area in 

 the Waterberg Flats of the Transvaal and in Bushmanland, and 

 of smaller area in Kenhardt, Prieska, Carnarvon and Kimberley, 

 is no doubt derived in part from rivers and in part directly from 

 the parent rocks by insolation and wind transport.'*^ Wind must 

 have been the chief agent in giving it the wide distribution it 

 now has. It is known to be 130 feet thick in places^^ south of 

 Morokwen. 



An instance of a partially sand-fiUed valley far from the 

 Kalahari is the Zand Leegte in Clanwilliam.^* This is a well- 

 marked valley cut in the Table Mountain, Sandstone of the coast 

 belt which is gradually being filled in, and down which water has 

 not been known to flow continviously since the occupation of the 

 farm, over 100 years ago. The country between the Berg River 

 and the Olifants west of the mountains flanking the left bank of 

 the latter is a sand veld, an old land surface buried to a con- 

 siderable depth under sand through which the tops of hills project ; 

 40 feet was considered a low estimate of the average depth of 

 sand. The presence of a raised beach containing shells of living 



