26 puesidext's addukss. 



(39) A. (J. Stigaiid, " Notes on Nganiilaiid," Geographical Jdunial, 

 1912, pp. 376-379. 



(40) A. L. flu Toit, Ana. Rep. Geol. Com. for 1905, p. 255; for 1907, 

 p. 96; P. A. Wagner, Report of the 16th Annual Meeting of the S.A.A.A.S., 

 1919, p. 187. The Hygap traverses a region where rounded boulders are 

 set free from a rather soft matrix of Dwyka tillite, and therefore gravels 

 are apt to be better developed along it than would otherwise be the case. 



(41) L. Schultze, " Aus Namaland und Kalahari," Jena, 1907, pp. 706-7, 

 and Ann. Rep. Geol. Com. for 1907, p. 107. 



(42) Ann. Rep. Geol. Com. for 1907, pp. 106, 151, 156. 



(43) Passarge, who had seen wide areas of this sand summed up his 

 views in " Suedafrika," 1908, p. 60, in this way: "The Kalahari sand is 

 the characteristic formation (amongst the younger deposits of the interior). 

 Thick masses of red and white sand cover the greater part of the Kalahari 

 steppes and condition their physical nature. They may well he mainly 

 desert deposits of great age, but they were in part redistributed by rivers 

 in the Pluvial, period." This seems to be a better-founded opinion than 

 the one expressed on p. 373 of " Die Kalahari," 1904 : •" Thus we always 

 return to the notion that the Kalahari sand owes its distribution in the 

 first place to water, and fast flowing water. The period of the Kalahari 

 sand must have been a time of huge precipitation, which covered wide 

 stretches of the land with raging waters. I freely admit that I have armed 

 myself to the utmost against such a view, and that I am even now at a 

 loss when I try to picture to myself the appearance of the country at 

 that time. But the most varied phenomena in different regions of the 

 Kalahari speak only too plainly for such an abundance of water." There 

 is a note of exaggeration in this, but, provided that "period*' may mean 

 a few days or hours, the picture, though difficult to raise, is fully justified. 

 In 1894 there was very heavy rain in Bechuanaland, and Mr. I^. G. Mayers 

 informed me that he was on the road between Grootfontein and Vryburg 

 at the time, and that for hours, over a distance of fully 60 miles, he 

 travelled in water inches deep with the bush projecting from it. He told 

 me this in 1906, and in answer to a recent letter of mine in which I had 

 asked for confirmation or correction of my recollection of his account, he 

 writes : "I was only able to distinguish the road by the flow of the water 

 in it, which resembled a river, the rest of the country being like a lake 

 out of which the ' aars ' (slight ridges of surface limestone marking the 

 courses of the dolerite dykes — A. W. R.) stood forth as the only recognis- 

 al)le features in the landscape. The year before last (1920) much the 

 same thing occurred again, the whole country flooded, houses falling down, 

 horses drowned in harness on the road, and an enormous loss of stock." 

 It was in 1904 that Abiquas Puts was converted into a lake for several 

 months and contained large fish. Such a flood must have moved large 

 quantities of sand in a short time. An instance of a flood causing the 

 removal and deposit of sand is found in the Matlabas River during 1909. 

 Before 1909 the river, as Mr. Heyzak, of Welgevonden, informed me, had 

 pools at intervals in the dry season, but heavy rain in the catchment 

 brought down sand during a storm in 1909, filling the pools, and water 

 is now only seen during the winter months where outcrops occur ; elsewhere 

 it has to be dug for. This is a good instance of the silting up of a river 

 bed without any permanent change of climatt! in the region. Whether 

 there were special circumstances, such as extensive burning of grass in 

 tlie catchment, depriving the sand of part of its usual protection, I do 

 not know. In September, 1920 (i.e., at the end of the dry season), the 

 Matlabas was the only one of the four northward flowing rivers (the otliers 

 are the Pongola,, Palala, and Magalakwin) seen by me that was dry, tlie 

 other three had flowing water. 



Though there is no doubt that great quantities of sand may be trans- 

 ported by water within a few hours in the usually dry country of the 

 interior, it is probalde that wind is by far the more important agent, 

 because on the rare opportunities one has of seeing vertical sections 

 through the sand bedding planes and current bedding are not seen (see 

 the description of a cutting through 20 feet of sand at Put Pan, \'ryburg. 



