THE SPOKANE GLACIATIOX 601 



of anastomosing channels with rock-basins and cataracts, leaving isolated 

 knobs and buttes irregularly disposed in a perfect maze (see figure -1). 

 Bare rock or rocky talus covers the area, whereas both canyoned portions 

 have no rock-floor exposed. There is a total descent of 600 feet along 

 this tract. Most of the channels have noteworthy tains accumulations, 

 amounting to three-fourths or four-fifths of the total depths. Only the 

 lower and central channels bear talus comparable to that in Grand Coulee 

 and J3rumheller channels. Erratics of granite, quartzites, etcetera, lie 

 here and there, even on the highest of these eroded surfaces. 



In the upper canyon, north of the moraine built by the Okanogan lobe 

 during the Wisconsin glaciation, talus is but halfway up the clifi's. Be- 

 low the moraine some talus piles seem to be of Spokane age, some of 

 Wisconsin, and there is other evidence that Wisconsin waters only par- 

 tially cleaned away the preexisting cliff M^aste. In the lower canyon the 

 talus dates back to the Spokane epoch. If there is any record of Wis- 

 consin waters, it is in the gravel fill, more than 200 feet deep, 3 miles 

 from the head of this canyon, and possibly in the existence of a promi- 

 nent rock terrace, not more than 100 feet above the aggraded floor of 

 the canyon, a few miles south of Palisades. If trenching by Wisconsin 

 waters was performed anywhere in the lower canyon, it was in the 

 production of this terrace. 



Moses Coulee, as a drainage line, antedates the Spokane glaciation. 

 Tributary valleys, well developed in the basalt, are recognizable as far 

 north as Mansfield, and the upper canyon itself extends 6 or 8 miles 

 north of the Wisconsin terminal moraine ; but the best evidence is in the 

 lower coulee. The cliffs here are deeply notched by wide-open Y-shaped 

 tributary valleys. Many notches are two-thirds or more as deep as the 

 main canyon. These notches give the cliffs a striking resemblance to a 

 series of great rounded gables in alignment (figure 11). The slopes of 

 these tributary gorges are graded and covered w^ith sage and grass. Eock 

 ledges in them are rare. They clearly are the relics of a pre-Spokane 

 drainage line, the trunk valley of which w^as entered and greatly enlarged 

 by the Spokane waters. Both widening and deepening in the basalt oc- 

 curred and the tributaries were left hanging. They have since attained 

 topogTaphic adjustment by building large alluvial fans out on the canyon 

 floor. Furthermore, Moses Coulee crosses the Badger Mountain fold, as 

 already noted. Like the crossing of the Soap Lake anticline by Grand 

 Coulee, this records an antecedent course determined long before the 

 Spokane glaciation. 



Spokane waters could not have entered Moses Coulee if the parent ice- 

 sheet had not pushed across or at least well up on the divide south of the 



