CHANNELED SCABLANDS OF THE COLUMBIA PLATEAU 645 



Quincy Basin, like the Hartline Valley, does not have a scabland 

 floor. Both lay too low to be eroded. But both belong to the 

 glacial drainage plexus. Quincy Basin probably contains more 

 gravel than all the scablands of the plateau together. It was an 

 enormous settling basin for the glacial rivers from Grand Coulee 

 and Crab Creek. The flood of waters entering it was so great that 

 at first three discharge ways were simultaneously in operation.^ 

 The southern and larger one obtained all the discharge later, and 

 by deep notching of the basin's rim, caused the glacial waters which 

 traversed the fill to incise their deposits. Tv/o great channels and 

 one smaller one were thus formed. The two large channels are 

 each about 3 miles wide. Each was eroded about 100 feet deep 

 during the later part of the episode. The one which contains 

 Rocky Ford Creek also carried the later Wisconsin discharge and 

 was further modified then.^ 



Erratic bowlders, some of them striated, are widely distributed 

 at all altitudes on the basalt plain and the scablands. They also 

 occur in valleys of the mature topography which open northward 

 on to the basalt plain, and in some which open on to scabland tracts. 

 The size, angularity, and striated surfaces indicate that these 

 erratic bowlders were not rolled to their positions by running water. 

 In the scablands, they must have been carried by berg ice on the 

 great rivers. In their peculiar and limited distribution in the 

 valleys in loess is evidence of small glacial lakes, in which the drift- 

 bearing bergs floated.^ 



' Frenchman Springs, The Potholes, and Drumheller Channels. 



2 This interpretation is a modification of that published earlier by the writer. 

 Further study of Grand Coulee, Quincy Basin, and Drumheller Channels has led to a 

 magnification of the work of the earlier flood, and a minimizing of the results accom- 

 plished during the Wisconsin epoch. Grand Falls is now considered to be a pre- 

 Wisconsin affair, none of the distributary canyons of Grand Coulee, except Dry 

 Coulee, are thought to have functioned during the second flooding, the Adrian terrace 

 is considered to be a part of the original fill and not of Wisconsin age, and all the deep 

 canyons of Drumheller Channels are thought to date back to the earlier episode. 



3 Below an altitude of about 1,250, erratic bowlders occur on every formation and 

 type of topography on the plateau. But these are a younger deposit (see Journal of 

 Geology, Vol. XXVII [1919], pp. 489-506) and do not much overlap the scablands. 

 Where overlap does occur, however, it is impossible to distinguish bowlders of the 

 two categories by any difference in the amount of decay. 



