PT. nr. OR POISONED BY VEGETABLE GROWTH? 165 



rain, the waste by tlie direct action of rivers may be 

 reckoned as nothing ; and even this waste by the 

 direct action of rivers takes place, I might say, entirely 

 when they are flooded by rain. The real main geo- 

 Icgical work of rivers is indirect ; that is, the carrying 

 off the traffic brought to them by the wash of rain : 

 and they carry this mighty traffic for the entire ter- 

 restrial surface of the globe ; at least, their channels 

 do. 



And the channels of most rivers would exist 

 whether the rivers existed or not, as in the south 

 you constantly see river beds dry, or almost dry, ex- 

 cept when filled by the superficial I'un of rain, or the 

 thawing of mountain snow ; and the size of the chan- 

 nels of all torrents and rivers (except in alluvial parts, 

 for a reason which will be given) is in proportion 

 to the heavy floods of rain which occasionally rush 

 through them, not to the comparatively small volume 

 of spring w^ater which always flows down them. 



Lyell, quoting Mr. Everest, calculates that in the 

 rainy season (four months) the Ganges discharges into 

 the sea a weight of earth equal to fifty- six great Pyra- 

 mids ; and in the other eight months only the weight 

 of four Pyramids. Now, if Professor Sedgwick will 

 only grant us these four Pyramids for the snows of 

 the Himalaya, we shall have an annual superficial wash 

 of soil amounting in weight to sixty Pyramids. This 

 is the ordinary and annual work resulting from the 

 operation of ordinary and annual rains on one river 



