DISTRIBUTION FROM A GROWTH OF THE SANDHILLS. 147 



the obliquity being due not to a diversion of the direction of the wind, 

 as assumed by v. Middendorff, 1 but to the fact that the upward diver- 

 sion of the wind currents takes place more and more down wind than 

 in the centre of the sandhill. 



The growth of the horns on either side of the notch is seldom, if 

 ever, symmetrical, one is longer than the other and occasionally one 

 end of the sandhill gives origin to a long ridge of sand extending down 

 wind and practically in the same direction as it. These long ridges 

 seem to have much in common with the longitudinal sandhills, so 

 conspicuously developed in the desert of Sind and Rajputana, but the 

 scale of the two is so different that their origin is probably also differ- 

 ent. Where one of the horns of the sandhill is extended in this manner 

 the side slopes are steeper than those of the horns of normal propor- 

 tion and usually markedly steeper on the inner side. 



The manner in which the sandhills change their form under a 

 change of wind is worthy of note. The principal winds seem to blow 

 from west-south-west and east-north-east, the former prevailing 

 for the greater part of the year, and to them the principal features 

 of the sandhills are due, while the latter blow during the winter 

 months, and cause a reverse slope and bank of sand to be formed, 

 seen near the summit of the long gentle slope which faces the west- 

 south-west winds. The formation of this reverse slope is accompanied 

 by a good deal of scour of the original steep leeward slope, though in 

 no case that I saw was this sufficient to cause a complete reversal of 

 the shape of the sandhill. This shape, as will be remembered, is 

 typically crescentic, with the horns pointing down wind ; on the rever- 

 sal of the wind these horns come to point up wind, and the principal 

 vertical displacement of the wind-currents takes place further down 

 wind than the minor vertical displacements on either side. This again 

 appears to be an unstable condition, and a modification soon occurs. 

 I am not able to say for certain whether the reverse concavity of the 

 lee side of the new sand ridge cuts off part of the original concavity 

 and a hollow is so left, or, as seems more probable, the hollow is 



1 Loc, cit. 



( '5 ) 



