rRE-LACUSTRlNE RELIEF OE BONNEVILLE REGION 359 



a])le ill that it Las ])oen able to maintain itself in spite of the fact that 

 three great mountain l)arriers have snccessively arisen across its course. 

 The means b}^ which it has been able to do this is also a clue to the rea- 

 sons wh}^ other streams of the region were not permitted to accomplisli 

 the same thing. The former existence of such a stream in Bonneville 

 basin is borne out b_y many facts. 



A large river occupying tlie Bonneville Valley before tlie lake period 

 naturally would be the twin stream of the Green Eiver. Its main gath- 

 ering basin, like that of its sister stream, would be most likely in the 

 Eockies of the Yellowstone Park district. Its course would follow tlie 

 deep and relative naiTOw valley ])etween the Port Neuf and Bannock 

 Ranges, enter the broader Bonneville basin, which it would traverse its 

 entire length, and finally reach the Colorado River at the point where 

 that stream makes its great sharp bend from a westward direction to a 

 southern one. The recent visits to the region had for one of their main 

 objects the gathering of facts l)earing on the support or disproval of such 

 a working hypothesis. AYith the former the observations seemed to agree. 



HEADWATERS OF OLD VIRGEN RIVER 



The complex of head streams which now gather together to make up 

 the Snake River of Idaho presents strong evidences of being a newly 

 adjusted system. Traces of relatively recent diversion are indicated in 

 many places. One of the principal trdnitaries, the Port Neuf River, 

 after flowing directly south for many miles, swings sliarply around to the 

 northwest and empties into the Snake River near Pocatello town. An- 

 other stream, the Bear River, likewise turns sharply on itself, its upper 

 half running parallel to the lower half, but in a diametrically opposite 

 direction. Other streams of the region have equally erratic courses. 



Near Pocatello the Port XeuE Creek crosses an old stream channel of 

 large size which is now filled high by lava, for the ancient drainage-ways 

 and valleys of southern Id alio seem to have often flowed with liquid rock 

 as well as water. This old channel extends apparently far to the north- 

 ward. Southvvard it enters Red Rock Pass and coincides with the old 

 river bed which Gilbert interpreted to represent the outlet of Xjake Bonne- 

 ville. That this old channel belonged to a south-flowing and not a north- 

 running river of considerable size seems clearly indicated by disposition 

 of the old stream grav'cls. The pebbles composing these gravels are not 

 chiefly derived from local rock ledges, but from Precamlirian crystallines, 

 such as characterize the Yellowstone Park region. They are numerous, 

 of nearly uniform measurement, and of ^a size that suggests their removal 

 a distance of al)out oik; hundred miles from their parent ledges. 



