28 GENESIS AND DEVELOPMENT 



alternately from opposite quarters, but does not hold in one quarter 

 suflBcienth' long to reverse the work of preceding winds. Cornish'^) 

 remarks that in speaking of amplitude instead of height of dunes, 

 one avoids the common confusion which results from the fact that the 

 vertical distance from the bottom of the trough to the summit may 

 increase even by raising of the crest or by deepening of the trough. 



A dune sufficiently large is stationary, and it is an established law 

 that as the amplitude increases, the velocity of the dune wave decreases. 

 The velocity of the dune is the rate of advance of the crest which takes 

 place by accumulation of sand upon the lee slope. There is also a group 

 velocitj' of dunes to be recognized, that is, the rate of travel of the sand. 

 Increase of wind will increase formation of new dunes to leeward 

 rather than the rate of travel of the individual dune, and is usually 

 accompanied by a considerable lowering of the general level, especially 

 in the case of simultaneous diminution in the supply of sand. 



The sorting action of wind already mentioned is supported by rain- 

 water ^hich washes the finer particles down into the trough, and con- 

 sequently we find the summit of dunes to consist of coarser material. 

 But on the other hand the lower part of the eddy is gouging out the 

 trough and the finer material is carried away through the combined 

 action of the eddy and the wind. The sand is therefore finer in the 

 dunes generally than in the hollows between them. On a large sandy 

 surface the particles are fijier at the extremity towards which the wind 

 blows. 



Through this winnowing process the dust which consists of frialjle 

 matter, having been reduced to the size of powder by grinding between 

 the sands is carried away from the dune district and deposited beyond 

 its limit. It is especially in desert regions, where aridity excludes 

 vegetation and allows the wind to play with full force upon the finer 

 particles of the soil, that we notice the development of sandy deserts 

 covered with quartz sand, j'et surrounded by grassy steppes consisting 

 of clay dust. This remarkable distribution of the products of rock 

 disintregration by wind and its effect on the physiograplw of Northern 

 Africa has been eminently shown by JValthei:^) Already Buvay^) 

 described the transition between the cultivated coast lands and the desert 

 of Africa, which must be called a steppe, and the genetical relation of 

 these formations is now a generally admitted fact. 



1) 1. c. p. 287. 



2) Die Denudation In der Wiiste. 



3) Zeitschr. fur Allgemelne Erkunde. Berlin, 1857, p. 290. 



