6i8 / HARLEN BRETZ 



ington, north of Snake River. These scablands have a history 

 which is beHeved to be unique. The prevaihng feature of their 

 topography is indicated in the term here used : channeled scablands.^ 

 They are scored by thousands of channels eroded in the underlying 

 rock. The plateau in Washington, north of Snake River, has a 

 total area of about 12,750 square miles, of which at least 2,000 

 square miles is channeled scabland. The scabland is widely dis- 

 tributed over the region in linear tracts among maturely dissected 

 hills which bear the loessial soil (wheat lands) of the plateau. 



PHYSIOGRAPHIC RELATIONS OF THE CHANNELED SCABLANDS 



The following features and relations of the scablands exist in all 

 tracts. They must form the basis of any interpretation for the 

 origin of channeled scabland. The map should be examined as this 

 list is read. 



1. Scabland tracts are developed invariably on or in the Colum- 

 bia basalt formation. 



2. Scabland tracts are invariably lower than the immediately 

 adjacent soil-covered areas. 



3. Scabland tracts are invariably elongate. 



4. The elongation of scabland tracts is with the dip slope of the 

 underlying basalt flows. There are eight known exceptions to this 

 rule,^ all minor affairs so far as length is concerned. 



5. Scabland tracts, considered as units, invariably have con- 

 tinuous gradients. 



6. Scabland tracts are invariably bounded by maturely eroded 

 topography. 



' An earlier paper on this subject was published by the writer in the Bulletin of the 

 Geological Society of America, Vol. XXXIV (1923), pp. 573-608. The study on which it 

 was based involved about a 1,000-mile traverse of the plateau. Since then, the writer 

 has studied the plateau more thoroughly, having added more than 2,000 miles to the 

 previous total traverse. Much more detailed information and several modifications 

 of the earlier interpretations justify the appearance of a second paper on the subject. 

 The accompanying map (Plate III) is based on a field examination of every scabland 

 there indicated. In a few places the boundaries are inferred (dashed lines) but future 

 work will hardly do more than make minor changes or additions. 



^ A part of Othello Channels, a part of Drumheller Channels, at Palisades and near 

 Spencer in Moses Coulee, at Soap Lake and near Bacon in Grand Coulee, 6 miles south 

 of Almira on Wilson Creek, and at Long Lake in Spring Coulee. 



