292 N. S. SHALER BROAD VALLEYS OF THE CORDILLERAS 



have usually gentle and completely adjusted slopes. Moreover, in all the 

 many instances I have inspected the torrent is now engaged in construct- 

 ing a detrital fan within the valley which a while ago it was excavating. 

 Not onl}^ are there valleys where the permanent streams intersect the 

 benches, but at many points we find such valleys where there is at pres- 

 ent no effective stream at work at an}'' season. These might be taken 

 as the result of torrential rains but for the fact that they all appear to be 

 very ancient. Moreover, they seem to have been excavated by moderate, 

 continuously acting streams and not by a catacl3^sm. A part of the evi- 

 dence on tliis point is afforded b}'' the fact that these ancient loadys are 

 in all the regions abounding in alluvial gold, the seat of local placers, 

 the metal being in winnable quantity because it is a concentration from 

 the general mass of the talus deposits. The distribution of this gold in 

 the loady^ where it occurs in benches of rearranged gravel, shows that 

 tliese now dr}^ valle3^s have had an ordinary stream liistory. 



Tlie duration of this last period of more abundant rainfall must have 

 been considerable, for the erosive work done on the taluses was evidently 

 great. The gorges formed by the main torrents were, in many instances, 

 a mile or more wide and from 1 to 300 feet in de[)tb, though the}^ have 

 since been to a great extent refilled during the later now existing stage 

 of accumulation. They generally retnain very distinct features. At the 

 present rate of growth of the detrital fans, it would require a geologically 

 long time for the replacement of the debris which was removed from 

 these cross-valleys. It appears likely that this channeling of the valley 

 deposits which is shown where the torrents cross the slopes continued 

 down to the centers of the troughs of the rivers, so that the drainage 

 flowed in canyon-like gorges of considerable depth. Although these 

 canyons cut in the debris are in process of refilling and are sometimes 

 almost occluded, they can often be traced by the slight escarpments 

 which show where their margins lay. Such sections as have been dis- 

 closed of the beds which have been laid down in these ancient water- 

 ways show relatively fine, worn debris differing in texture from that 

 which makes up the slopes on either side of the river. 



So far as the evidence goes, it is to the effect that this period of in- 

 creased rainfall and consequent reinstitution of erosion in part coincided 

 with the last develoi)ment of glaciers in this region. It appears to have 

 continued after the retreat of the local ice-streams from the valle3'S, as 

 is shown by the fact that the morainal fields are incised by channels 

 which are no longer the seat of cutting streams. It may also have some- 

 what preceded the development of those icefields, as is occasionally 

 shown by the semblance of such dead valle3's partly covered by drift 

 materials. It is not easy to make observations of value as to this ques- 



