252 SCIENCE PROGRESS 



overlooked. Alluvial land, irrigated and manured, is, clearly and 

 obviously, subject to rapid denudation. The ground is porous. 

 The roots of trees and crops are continually loosening new rock. 

 The ground is soaked by the percolation of water. Passage 

 is made in winter for the water to enter the rock below, to 

 freeze and break it up. The humus acids formed by the rotting 

 of manure are not without their effect. It is, indeed, not 

 unlikely that much of the observed erosion is due to human- 

 kind. We must not forget the influence of man as a geologic 

 agent. 



The data at our disposal are too scattered for us to form 

 definite conclusions, but it is an interesting fact that all rivers 

 with a high, or an abnormal, discharge of sediment are situated 

 in densely populated and highly cultivated districts. Those 

 with a calculated rate of erosion greater than a foot in 2,000 

 years are the Ganges, the Irrawady, the Hoang-Ho, the Po, 

 the Rhone, the basins of all of which have been highly cultivated 

 for generations. Those with a moderate rate of erosion (more 

 than a foot in 7,000 years) are the Potomac, the Mississippi, the 

 Danube, all of which drain districts of considerable cultivation. 

 The rivers with an abnormally low rate of discharge of sediment 

 are the Uraguay, the Rio Grande, the Nile. The Nile is ex- 

 ceptional owing to the absence of rainfall in the lower part of 

 its basin and to the fact that a proportion of the sediment from 

 the upper reaches is deposited in the rainless district during 

 the annual river overflow. The Uraguay and the Rio Grande 

 are situated in districts of comparatively sparse cultivation. 

 These facts are striking. The conclusion may or may not be 

 that here suggested, namely that the discharge of sediment 

 does not represent true geologic erosion, but merely the effect 

 of cultivation, but, at least, the coincidence shows that the 

 problem of the rate of erosion under diverse conditions requires 

 further investigation. 



Wider data are needed to avoid possible sources of error. 

 We should endeavour to find river basins under conditions 

 similar to those which existed before the earth was trodden by 

 the foot of man. If we could obtain, for example, reliable 

 experimental data for the Amazon (a tropical and sparsely 

 populated district), the Colorado, and Murray (districts of 

 scanty rainfall), the Mackenzie (a district under glacial condi- 

 tions), and one or two miscellaneous results (such as the 



