52 



only a terrace of sand. There is but one agency that could pro- 

 duce such effects, a sea sufficiently deep to float off the glacial ice 

 in bergs and filling the whole Basin and thus preventing the for- 

 mation of moraines, as must have occurred otherwise. This unex- 

 pected discovery solved all problems of plant distribution in the 

 region as well, and entirely changed the conception of Shoshone 

 Lake, a name given to the arm of the great Columbia Lake which 

 filled the Snake river valley from near Pocatello to Huntington 

 Oregon. The Flathead and the Bitterroot also were arms. With 

 this new viewpoint I was so fortunate as to rctravorse the whole 

 region, nearly, in the fall. I traced the lake beaches of the Co- 

 lumbia Lake from near Chopaka peak to Spokane and the Flat- 

 head arm beyond Missoula, a distance of over 1000 miles. I saw 

 abundant evidence. There were river terraces but most of the 

 large ones so conspicuous on the Okanogan were lake beaches as 

 was shown by their filling all the coves where river action would 

 be absent. This fresh water lake was wholly distinct from the Ter- 

 tiary lake and did not drain away till after the close of the glacial 

 period as is shown by the absence of moraines far up toward the 

 head even of Chelan canon which was the longest-lived glacier. 

 The existence of this great glacial lake explains why there is no 

 terminal moraine at Thompson Falls, only a delta, and why there 

 is no more erosion there than in most river (not glacial) valleys. 

 It does away with the violent assumptions to account for the Bit- 

 terroot Lake at Missoula, for the boulder-speckled surface of the 

 Basin and, best of all, accounts for the similarity of the flora of 

 the Flathead, Blue Mts. and Cascades. It also explains the char- 

 ter of the soil especially that overlying the Tertiary lava flows which 

 for the most part overlie the Tertiary lake beds. It also does away 

 with the assumption that the Tertiary lava flows near the Wallo- 

 was of eastern Oregon caused Shoshone Lake. This Columbia 

 Lake, as I have named it, also covered the Salmon watershed and 



all the strings of lakes in southern British Columbia to the Fraser 

 river. 



f 



