THE SPOKANE GLACIATION 



599 



The lava flow which causes the lower rock terrace is much more resist- 

 ant to plucking and sapping than is the upper terrace. This is sho^\^l 

 both in the Columbia Valle}'' and in "The Potholes." The current which 

 emerged from the Potholes themselves spread considerably over this ter- 

 race and spilled over its edge in a broad sheet which later became some- 

 what concentrated in four or five different places, so that minor notching 

 of the edge of the terrace resulted; but none of these notches was cut 

 back more than a quarter of a mile. 



Above and east of the upper falls is a scabland tract extending two 

 miles farther east across the low rim of the structural basin and very 

 much diversified by ramifying channels and their separating hills. 

 Eock-basins are common, some of them being 40 feet deep. This chan- 



FiGURE 10. — One of ''The Potholes'' 



The gravel bar (terrace on left) is 200 feet thick and the clifE back of it is 200 feet 

 high. View is taken looking toward the ancient cataract. 



neled tract was an island-studded rapids descending to the brink of "The 

 Potholes" cataract. At the beginning of the cataract this channeled 

 area extended to the original edge of the upper terrace. As the twin 

 falls, narrower than the channel group, receded eastward, some of these 

 channels were left along the edge of tlie gorge. 



The talus accumulations of "The Potholes" are somewhat irregular 

 in height, because of unusually marked differences among the flows in 

 the cliffs, but the large majority constitute three-fourths or more of the 

 total height of the cliffs (see figure 7). "The Potholes" cataract was 

 formed at the time of the Spokane glaciation by discharge from the 

 Quincy basin. The Columbia Valley here was nearly or quite as deep 

 at that time as it is today. The structural basin was aggraded to the 

 level of this western rim and gravel was carried completely across from 



