SOME ASPECTS OF GEOLOGIC TIME 253 



Zambesi) for other districts where the population was sparse, 

 we should throw some light on our problem. Measurements 

 for the upper reaches of rivers would be helpful. Also we must 

 note that, though the rate of discharge of sediment is, as yet, 

 our best guide to the rate of erosion, it is not impossible that 

 others may be discovered. For the present, however, we must 

 clearly realise that such information as we do possess is scanty 

 and uncertain. There is one other point of importance. It is 

 highly probable that, in all normal cases, there must be some 

 relation between true geologic erosion and the soluble content 

 of the river. The relation would not be strictly proportionate 

 because of the solubility of carbonate of lime, but there would, 

 as a rule, be some relation. Now it is a suggestive fact that so 

 many rivers which pass through districts of sparse population 

 have a comparatively small soluble content. If we mark out 

 those like the Colorado and the Kansas, draining " bad lands," 

 impregnated with large quantities of saline deposits, the soluble 

 content is unusually small. The Amazon, for example, has a 

 soluble content of less than 50 parts per million. The rivers 

 of Northern Sweden are remarkably pure. Other instances 

 can be given. Though rough and inaccurate, the suggestion 

 is one on which I lay some stress. It has been shown that 

 the process of weathering is, largely, a chemical change, in 

 which a portion of the substance is carried away in solution, 

 and, by that change, the remainder is loosened and comes away 

 in the form of sediment. Erosion and solvent denudation must 

 always be interrelated. 



Other circumstances that point to the conclusion that the 

 rate of erosion has probably been overestimated are the long 

 periods, in all climates (except the neighbourhood of large 

 manufacturing towns), during which inscriptions will remain 

 legible. Some, not deeply cut, will last for many thousands 

 of years. Once again, it is well known that we can still see, 

 on the rocks in mountainous regions, striae which date back 

 to the last glacial epoch. If this occurred (say) 30,000 years 

 ago, several feet of strata must, according to current theories, 

 have been removed in the meantime. How anything of the 

 kind could happen and leave the striae as we now find them 

 requires some explanation. It thus seems probable that the 

 rapidity of land erosion may be smaller than our data would 

 tend to show. This suggestion I put forward for what it is 



