INFILLING OF THE VALLEYS 285 



plete that the stream is forced mto other channels. When the ejections 

 are as ash the embarrassment of the drainage may be more extended and 

 nearly as complete, for the reason that the dust, though spread over a 

 wide area, is likely to be rapidly swept away to the low ground by the 

 rain in such quantities that when cemented in the form of tuff the rivers 

 can not bear it away. Where a region is the seat of extensive and long- 

 continued volcanic activity the original topography may become en- 

 tirely hidden, as is the case in the great area of the Columbia River ash 

 and lava fields, so that a new drainage system has to be developed. 



Although in certain parts of the Cordilleran district volcanic ejections 

 of ash or lava may have had a considerable share in forming the ac- 

 cumulations of the ordinary broad valleys, the evidence goes to show 

 that they have not been the principal agent in this work. The fact that 

 the ash deposits at Alder gulch lie in a channel cut through the slopes 

 by a mountain torrent, and that they do not extend under the ordinary 

 detritus of those slopes, indicates that the only accumulation of fragment- 

 ary volcanic matter of which I have been able to trace the history came 

 too late to have Ijad any decided effect on the filling process. Moreover, 

 very many sections, though none of them except those near Butte are 

 of any great depth, go to show that the mass of the detrital slopes are of 

 ordinary water-worn detritus. We may therefore consider volcanic waste 

 as an element other than controlling in the process of deposition, and 

 we are forced to seek the principal factor in changes in the quantity 

 and distribution of rainfall. 



So far as I have been able to find the normal mountain valleys of the 

 world, those in which the streams which are still excavating their beds 

 appear to be generally limited to those areas in which the rainfall is 

 considerable and rather uniformly distributed, while the broad type, 

 such as we are considering, is common in regions where the precipita- 

 tion is scanty and irregular, where it takes on the character common in 

 partly arid countries of torrential rains so local and violent as to receive 

 the name of '' cloud-bursts." There are many reasons why the differ- 

 ence in the quantity and order of the precipitation should produce a wide 

 diversity in the amount of debris brought down by mountain torrents 

 into a valley. As this matter has not, so far as I have learned, been set 

 forth, I shall give it a somewhat extended presentation. 



First let us note that the degradation of a mountainous country is 

 likely to be hindered by a thick coating of vegetation, for such a cover- 

 ing tends to prevent the access of frost. Moreover, the joints and other 

 interstices of the rocks are kept continually full of water, and thus in 

 part protect it from the access of air, while in arid conditions they are 

 filled by the occasional rains, and as the water-table sinks, as it may do 



