50 



Til AHA ROSTGO SAND DUNES. 



The agent which has brought about the removal is, in the 

 writer's opinion, chiefly the wind, since rain or running water 

 has no access, at any rate immediately under the overhanging 

 slabs. 



The wind, charged with sand, would also act in the same 

 way as a sand blast, causing further abrasion of the rock surface. 



This sand-laden wind, then, rising above the top of the 

 plateau, forms eddies and drops at least some of the sand, which 

 has accumulated and formed dunes in suitable spots. 



Fig. 2. 



Figure 2 illustrates what is assumed to occur, and the 

 lowest photograph, Plate 2, shows the end of a dune in the posi- 

 tion shown in Fig. 2. 



In high winds sand will undoubtedly be blown off the 

 dunes and off the mountain also, thus diminishing their size, 

 while at other times the wind velocity will be such that sand 

 accumulates and their size increases. Further, with a change 

 of direction, their position tends to alter, and this, of course, is 

 known to take place. 



That their size and position can vary with great rapidity 

 under no other action than that of the wind is borne out by 

 the writer's knowledge of the behaviour of sand residue dumps 

 on the Witwatersrand, which afford an excellent example of the 

 transporting power of wind. 



The top surface is frequently loose to a depth of a couple 

 of feet or so, and this appears to move from side to side of 

 the dump according to the direction of the wind, that side on 

 which the wind is blowing being often quite bare and relatively 

 hard* through the loose sand shifting over to the lee side. 



Tn one instance in a single night nearly three feet of sand 

 lias been blown over on to a place on the lee side, a newlv-erected 

 water tank affording means of measurement, and a few hours' 

 high wind in a fresh direction is sufficient to cover with loose 

 t and a previously bare surface. 



The '-and moves along partly in a series of shifting ripples, 

 and partly in actual suspension in the air, the latter being only 

 too obvious to anyone working on the dump, but the former 



* The sand particles arc in these cases loosely cemented together by 

 slime .-ind salts formed by oxidation of pyrites. 



