102 W. T. Blanford — On the Physical Geography [No. 2, 



ting sandy plain. Over a breadth of about a quarter of a mile, and rigbt 

 and left, in the direction of the wind, as far as the eye could distinguish, the 

 surface was covered with sand in small newly formed hillocks, mostly of 

 crescentic form and about 5 to 10 feet high. The direction of the wind 

 was shewn to be from S. 35 E., this being at right angles to the lines of 

 ripples, and to the chords of the crescentic arcs formed by the hillocks ; and 

 the general direction of the sand-ridges immediately to the westward was 

 the same. Many of these sand-ridges were so regular that it was difficult 

 to conceive that they could have been formed otherwise than in long lines. 

 But I do not quite understand how the wind can thus form them. There 

 is a great difference between forming a line of hillocks and uniting them 

 into one continuous ridge. 



The third suggestion is that the tract of country along the edge of the 

 Indus alluvium was originally covered at least as deep as the height of the 

 present sand-hills by sand arranged more or less in ridges at right angles 

 to the prevailing south-west wind, and that the valleys between the present 

 sand-hills are the result of wind-denudation, their contents having been 

 swept away and the intervening ridges left. The abrupt terminations of 

 the ridges mark the former leeward slope of the sand-hills. On the whole, 

 I think this last theory is the most probable of the three. 



The sand, it is true, accumulates in long ridges behind any obstacle, and, 

 when a ridge is once found, it will tend to be prolonged to leeward. But 

 no obstacle exists of sufficient size to account for the commencement of a 

 ridge 100 to 200 feet high. 



§ 12. Source of the Sand. — There is yet one point which demands notice 

 and that is the source of the sand. Rounded sand-grains are rarely pro- 

 duced in any quantity by simple subaerial disintegration, except in the case 

 of the degradation of a sandstone, and in the present instance there is no 

 sandstone area to windward. All the sand may safely be assumed to be 

 derived from river- channels or the sea coast. 



Part of the sand may be derived from the bed of the Indus, and pro- 

 bably a large portion of the sand-hills of Rohri are supplied from this 

 source. But it is difficult to conceive that all the sand-hills of Thar and 

 Parkar, Mallani, Jaysalmir, &c, can have derived their sand from the Indus, 

 to say nothing of those of Jodhpur, Bikanir, &c. 



Some of the sand also may be derived from the present coast-line. But 

 all the sand-hills are at a distance from the coast, and it is difficult to con- 

 ceive that all the sand has been blown across the delta of the Indus and 

 the Ran of Kachh to reach the region where it so greatly abounds. Had 

 all the sand which is spread over the plains of western Rajpiitana been 

 blown across the Ran, the latter would surely have been converted into a 

 sandy desert long since. 



