148 OLDHAM: SANDHILLS OF CLIFTON NEAR KARACHI. 



scoured out by a wind eddy ; but whichever be the origin, there is 

 usually a hollow formed on the crest of the sandhill, to leeward of 

 the new lee slope, a hollow which grows by removal of sand by the 

 eddy set up in it, and corresponds on a small scale to the fuljes of 

 the Arabian desert. 



This hollow is very well shown on the crest of the sandhill repre- 

 sented in Plate I, and this photograph also shows the manner in 

 which a change of form of the sandhill takes place, when it is attacked 

 by a wind from a fresh direction. The sandhill in question was first 

 shaped by west-south-west winds, then a period of east-north-east 

 winds caused a partial modification of form, heaping up the sand from 

 that side and forming the steep slope facing to left of the picture. 

 These winds ceased for a while, and the sandhill was attacked by a 

 south-west wind which has commenced to re-shape it, but this re-shap- 

 ing does not in the first instance take place by a general and uniform 

 removal of sand, but by the formation of notches in the cre'st, in which 

 the wind is concentrated and increased in force, setting up a violent 

 scour and excavating deep pits to leeward. In the photograph can be 

 seen the ribbing of the sides of these notches, caused by the sand 

 slipping down the sides as the notch is deepened by the scour, and it 

 will be noticed that the furthest of these notches has been nearly cut 

 down to the foot of the steep slope ; in it downward scour is nearly at 

 an end, the effect of the wind will be to widen the notch, and, as it and 

 its fellows increase in size, the violence of the rush of air through them 

 will diminish, its effect become more regular, and instead of the greatest 

 scour being in the notches it will be on the pinnacles left between 

 them, which will be gradually lowered till the crest is reduced to a 

 smoothly-rounded outline. 



4. — Composition of the Sand. 



The laboratory of the Geological Survey is not equipped with 

 apparatus for the complete physical analysis of soils, so a substitute 

 was used, in the shape of a series of sieves of 10, 24, 50, 70, and 100 

 meshes to the inch, through which samples of sands from Clifton, and 

 dredgings from the harbour, were passed. 

 ( 16 ) 



