1849.] On the Embankments of Rivers. 697 



range. To the northward of the Indus, or on a line running from 

 Garo towards Yarkund, I dare say that undulating ground or moderate 

 slopes, rather than deep ravines with steep sides, may perhaps be found. 

 These downs or steppes, or at least tracts, afford pasturage to the 

 best description of shawl- wool goats, and Lieut. Strachey is right in his 

 opinion that, elevated although they be, they are as free from snow 

 during summer as the plains of India. What he supposes of the Kailas 

 or Gagri of the Manasarawar lake, viz. that the height of its (northern) 

 snow line may be 19,500 feet, would also be fully verified on any moun- 

 tains which may break the sameness of these steppes, and not be so far 

 to the north as to be much affected by the latitude. 



On the Embankments of Rivers, and on the Nature of Overflowing 

 Rivers in Diluvial Plains. By Capt. J. D. Cunningham, Engi- 

 neers. 



A short time ago I addressed the Editor of the " Englishman'' about 

 the embankments of the Damooda and other streams, and partly be- 

 cause my propositions were well spoken of by that paper (see issue of 

 the 31st May, 1849), I am induced to write to you more at length on 

 the general question of such dikes, and also to make my views more 

 plain by a few illustrative sketches. 



Proposed Scheme of Embankments. — My scheme is founded on the 

 fact that rivers, how capricious soever they may seem to be in any one 

 neighbourhood, do nevertheless on the whole usually wind or vary within 

 given limits, or will ordinarily be found to flow between lines parallel, 

 at an ascertainable distance — although in their deltas and in the upper 

 portion of their courses, they do also, after long periods, occasionally 

 quit such belts of petty variation, and strike out new channels for them- 

 selves at considerable distances to the right or left. Hence, disregard- 

 ing these latter changes as not giving cause for yearly care, I recom- 

 mend that there should be two sets of embankments, one primary, and 

 of a large section, following generally the limits of the belt of variation, 

 and the other secondary and smaller, surrounding islands, or covering 

 detached portions of land within the belt, which can be enclosed without 

 materially impeding the flow of an ordinary inundation. By keeping 



4 x 



