G2 GENERAL CONCLUSIOXS. 



appear as a distinct layer in the deposit. But this prob- 

 ably never occurs except in the drifting dunes, and near 

 them. In dune sand the most perfect lamination is often 

 to be seen, even when the actual difference in the coarse- 

 ness of the separate seams is very small (Nos. 47. 48). 

 The deposits which accumulate nearest in the lee of 

 drifting" tracts may also sometimes become more or less 

 stratified, Avhen coarse layers from exceptionally heavy 

 storms are thick enough to remain separate. This does 

 not always happen, for rains and growing plants are 

 effective agents in mingling successive laminae, Avhen not 

 too thick, into a homogeneous unstratifled mass. 



But the lulls which occur even in the strongest winds 

 soon cause the coarser particles of their load to fall out, 

 and after a Avhile only the finer ones remain suspended. 

 This is plainly indicated by the composition of the sam- 

 ples of sand, which were collected in front of dunes (Nos. 

 64, 75). As the wind travels away from the place of 

 loading, its many convection currents, turns, and wind- 

 ings cause it to disperse vertically and horizontally, and 

 the load is pari passu dispersed and thinned. From such 

 an atmosphere sedimentation is very slow. From each 

 transient current, marked off by cyclonic, diurnal, or 

 shorter irregular periods, deposits are laid down, which 

 no doubt are different from each other in mechanical 

 composition, but the quantity from each is never suffi- 

 cienth' great to form a separate lamina. Each deposit is 

 thoroughly mingled with that which has settled before, 

 either by the settling of the particles of the latest deposit 

 in the interstices of that laid down before, as this is not 

 thick enough to completely cover the ground surface, or 



