The Great Terrace of the Columbia. — Russell. 369 
glacier could not have divided so as to send a branching ice- 
distributary through it. The conditions as they existed at the 
northern end of the coulee during the maximum extension of 
the Chelan glacier, are paralleled, I think, in every essential 
particular at the present day, on the east border of Hayden 
glacier, near Yakutat bay, Alaska. At the locality referred 
to, a deep, narrow, transverse channel, a remnant of an ancient 
stream channel, cuts across a mountain spur. This channel is 
at right angles tO' the east border of the Hayden glacier, which 
forms a wall of ice from 150 to 200 feet high across its upper 
extremity but does not extend into it. A small stream, fed 
by the melting ice, courses down the lateral channel, but this 
is all the direct influence that the glacier has upon it. The 
Chelan glacier rose about 200 feet above the lowest point in 
the entrance to Knapp's coulee, but did not enter it. The 
melting of the ice very likely fed a small stream which found 
its way southward down the coulee, but of this I have no good 
evidence. 
There are several other portions of Mr. Dawson's inter- 
pretation of the topographical history of the region drained 
by the Columbia, which dilifer widely from the conclusions 
based on my own observations, but it does not seem as if 
farther discussion at this time would be profitable. It is to be 
remembered that no detailed geological studies have been 
carried on in the portion of the country under consideration, 
at least my own work there is entirely of the nature of a 
reconnoissance, and the radical dififerences between Mr. Daw- 
son's conclusions and my own, can, apparently, only be ad- 
justed after more extended field -study. 
