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



sides, but they often end in a higher point at their north-eastern extremities. 

 The scattered hills east of Balmir are always highest to the north-east and 

 slope away very gradually to the south-west. The sand-hills as a rule are 

 evidently of very great antiquity ; they often shew evidence of denudation 

 from the action of rain, and sometimes they are worn into ravines several 

 feet in depth. When it is considered how small the desert rainfall is (11*8 

 inches in the year at Umarkot, 18 at Nagar Parkar, but much less 

 in the central portion of the desert, and especially towards Jaysalmir), 

 it is evident that a long series of years must be required for ravines 

 even a foot in depth to be cut in the sand, since it is only in exceptionally 

 heavy showers that any water can run off so porous a surface. At the 

 north-eastern termination of these sand-hills, however, there is frequently 

 found a quantity of sand which is shewn to be newly deposited by its sur- 

 face being ripple-marked, by the absence of holes made by burrowing ani- 

 mals, and by the stems of bushes being partially buried. Lastly, from the 

 north-east corner of most of the high hills near Balmir a long ridge of sand 

 runs out, evidently deposited by the wind under the lee of the hill. 



Now there is one point to which it is necessary to advert before going 

 further. I must apologize for mentioning a fact doubtless familiar to most 

 of my readers, but although familiar with it on a small scale, I did not 

 clearly understand its application when I first went into the desert, and in 

 consequence I was for some time greatly puzzled by the phenomena presented 

 by the sand-hills ; I think, too, that it has been overlooked by Sir Bartle 

 Frere, and that this accounts in part for his doubting the efficacy of the 

 wind in producing the sand-hills of the Thar. On the possibility that it 

 may not be universally familiar I will venture to call attention to it. 



When the wind blows over any surface composed of particles which can 

 be moved independently, it forms waves or ridges more or less at right an- 

 gles to its direction, with a long low slope to windward and a steeper slope to 

 leeward. Something similar is seen in sand-banks formed by rivers and must 

 be well-known to all who have navigated any Indian river in the dry season. 

 In descending the stream the depth of water every here and there will be 

 found to diminish gradually up to a certain point, which is part of a shoal 

 stretching more or less across the channel ; below this the water becomes 

 suddenly deep. Here again the long gradual slope is on the side from 

 which the ciu'rent runs, the steep slope in the direction towards which the 

 river is flowing. 



This phenomenon on the small scale must be familiar to every one, 

 as it may be seen on sand or dust wherever the wind blows over it. 

 The long slope to windward is variable, the steeper slope to leeward is 

 that assumed naturally by a talus of the material forming the ripples. 

 The sand is driven up the longer windward slope by the wind and 



