EHYTHMS IN DENUDATION" 761 



erosion cycle, would be preserved as fossil remnants of an older belt of 

 weathering beneath the surface of a peneplain, though no longer serving 

 as zones of underground flow. Facts which appear to require such excep- 

 tional interpretations are known, but as long as they are isolated examples 

 they can not be used as arguments against the generality of the principle. 



Thus a very flat emerged plain in time would come to have its uncon- 

 solidated sediments removed down to baselevel. Below that level cemen- 

 tation rather than solution would go forward, and close below the sur- 

 face would exist conditions for the indefinite preservation of limestone, 

 fossils, and other soluble materials. Thus on peneplains not only has the 

 mechanical erosion become reduced, but solvent denudation from ground- 

 water action is also reduced to a minimum. 



To sum up this discussion, it is seen that no simple law can be given 

 which expresses the relation between rate of denudation and slope. It 

 varies with the absolute magnitude of uplift, and probably, for the average 

 rock and pluvial climate, the exponential factor lies between the first and 

 second power of the angle of the slopes. Thus a region whose surface 

 bas an avei'age relief of five degrees probably suffers denudation between 

 two and four times as fast as it will when its average relief is reduced to 

 two and one-half degrees. 



Upon the first recognition of the existence of uplifted and dissected 

 peneplains, the date of the older was generally placed as Cretaceous. More 

 recent investigation shows that in regions of moderately soft rocks the 

 oldest recognizable peneplains are hardly older than Tertiary, and may 

 even be late Tertiary, in age. Stages of partial peneplanation of soft 

 rocks and of mature valleys in hard rocks are even younger, belonging 

 to the late Pliocene or early Pleistocene. Tlie Cenozoic is now regarded 

 as an era of general continental elevation; yet from the evidenccNof the 

 long duration of these partial cycles of erosion the average slope of most 

 parts of the earth's surface due to erosion might be placed as perhaps 

 twice as steep now as during the mean of Tertiary time. The present 

 rate of denudation may therefore be several times the mean rate for the 

 entire Cenozoic. 



SIGNIFICANCE OF PRESENT VALLEY FORMS 



Over the greater parts of the continents the rivers are now trenched 

 in inner valleys. Rock walls or bluffs more or less narrowly hedge them 

 in. Above the inner valley lies a higher and wider outer valley whose 

 floor represents the level of the stream previous to the last acceleration of 

 erosion. Terraces on this older valley margin or hills with accordant tops 

 give traces of still higher and older levels. The highest level commonly 



