282 N. S. SHALER — BROAD VALLEYS OF THE CORDILLERAS 



the Madison and the Rub}'^ rivers, at the base of the mountain, have cut 

 down 3,000 feet to the present level of these streams, and also for the 

 unknown depth of the troughs that lie below the detritus that partly 

 fills the valleys. The total erosion can not we|l be less than 6,000 feet 

 and may very much exceed that estimate, 3^et it has apparentlj^^ been 

 accomplished in geologically modern times, certainl}^ since the Cre- 

 taceous period. This instance of high l3^ing, apparentl}'' recent gravels 

 is but one of many that could be cited from the region about the head- 

 waters of the Missouri. They occur at a number of points about the 

 Madison river and elsewhere. 



Although we have to estimate the original depth of the valle3^s as sur- 

 prisingly great, the estimate does not have to go so far as it would if 

 we were to neglect the broad shoulders or bed-rock slo})es which, we have 

 reason, as at Butte, to suppose, commonly lie beneath the lateral parts 

 of the detrital accumulations. If we should suppose the declivities of 

 one of the broader of these troughs to be carried downward as contin- 

 uous slopes to the center of the excavation, we would have to reckon an 

 almost impossibly strong topography. On this basis we should have 

 valleys equal to, if not exceeding, any now existing on this sphere. As 

 it is, we cannot well escape the conclusion that some of these vales have 

 the sites of their original streams below the level of the sea ; that if the 

 debris that now fills thenl were removed the}' would have more than 

 twice their existing depth. 



, An evident objection to the supposition that these valleys once had the 

 great depth which is here assigned to them is found in the fact that the 

 streams which drain them sometimes pass over bed rocks at a level 

 which is not far from the present beds of the streams that traverse 

 them. Tlius the Missouri at Great Falls flows on the bed rock, and like 

 conditions maj' be found elsewhere. In answer to this objection it 

 may be said that these valleys are evidently very ancient, and that ex- 

 tensiv^e local warpings have clearly taken place since they were formed. 

 Moreover, in such instances as that of the falls of the Missouri it is 

 possible that the stream has been so far displaced by the infilling of its 

 original channel as to flow at some distance from the path it followed 

 when the channels were excavated. 



Method of Infilling of the broad Valleys 



We have now to consider the conditions which have led to the depo- 

 sition of the beds that have so far occluded the greater part of the Cor- 

 dilleran valleys within the limits of the United States. Incidentally we 

 have noted that these accumulations were not to any considerable ex- 



