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



neatli the alluvial deposits of the Silver Bow river at Butte, Montana. 

 They have wide bottoms, declining gently from the base of the border- 

 ing ranges to near the center of the trough, and then a sharp canyon- 

 like gorge descending to the ancient level of the stream. A fair though 

 small example of these structures is seen in canyon Diablo, on the Santa 

 Fe railway. The simplest explanation of these vallej^'s would be to sup- 

 pose that after they were broadened out in adjustment to one baselevel 

 of erosion the region in which they lie had been elevated or tilted so as 

 to increase the cutting power of the streams. There are, however, several 

 reasons why this view is not satisfactor3\ In the first place, as just above 

 noted, the sha}-)e of these vales is identical with what we find in those of 

 the more arid district which has been filled with debris. In the second 

 place, these troughs lie too far above the sea to have had the regimen of 

 their streams altered by any such change in its level as has recently 

 occurred. The central canyon is by no means continuous down to the 

 coast line, as should be the case if its existence was due to recently 

 augmented erosion brought al)out by a change of level. As for the sug- 

 gestion that there has been an increase of erosion due to tilting, evidence 

 to show such movement is quite lacking; there is no such contrast in 

 the conditions of streams flowing in diverse directions as normally, in- 

 deed necessarily, marks a drainage system which has been thus perturbed. 



Any alteration of the slopes of this region sufficient to have produced 

 the reexcavation of these valleys would be indicated by a general dis- 

 turbance of drainage such as has evidently not occurred. 



Along the eastern face of the Rocky mountains there are several fea- 

 tures which deserve notice, for the reason that they appear to throw 

 some doubt upon the hypothesis that the long continuance of an arid 

 climate will account for the refilling of valleys excavated during humid 

 ])eriods. First of these is the deep canyon valley of the upper Arkansas 

 river. This gorge is on some accounts more puzzling than that of the 

 Colorado, for the reason that while the U})per part of the trough is of 

 normal width and has been extensively refilled, the canyon section is 

 relatively narrower than almost any other formed by the Cordilleran 

 streams. I have been unable to determine the reason for this peculiar 

 feature. It may be due to some recent uplifting of the section in which 

 the can)''on lies, or possibly, though not probably, to a diff'erence in the 

 rate of decay of the rock in the different parts of the valley. It should 

 be noted that this stream,. like the Colorado, is fed from the melting 

 enows of a high country where the rainfall is and doubtless has long 

 been considerable. The fact that the valley above the gorge was the seat 

 of extensive glaciation during the last ice epoch clearly indicates that 

 the annual precipitation has recently been much greater than it is at 

 present. 



