fracture.[1] At the northern end of the cañon, however, granitic rocks form a portion of its walls, and stand as isolated towerlike masses within it. Some of these towers are capped with horizontal lava sheets. When the lava was poured out it surrounded a granite ridge having the position of the Grande Coulée, but probably not extending as far south as the depression since formed. The weathering and removal of the granite gave origin to a trenchlike depression with vertical walls, composed of basalt above and granite below. The more rapid crumbling of the granite led to the breaking away of the jointed basalt resting on it, and the widening of the depression in the manner already noticed.
From the brief and inadequate description I have given of certain of the more striking features of Washington and Oregon it will be seen that landslides have modified the topography of the region occupied by the Columbia lava in several ways. There are yet other changes in the geography of that most interesting and instructive land due to the same causes. Chief among them are the obstructions to the streams formed by landslides, and the production of lakes and rapids. At several localities in the upper Columbia masses of rock which have fallen from the cliffs bordering the stream obstruct its course. There are now no lakes along the course of the river due to this cause, but terraces above rocky rapids, show where such water bodies previously existed.
Perhaps the most interesting fact brought out by the study of landslide topography is that certain broad, nearly level areas, now covered with deep, rich soil, and in the autumn golden with the sheen of ripened grain, owe the minor features in their relief to ancient landslides. The hills, with broadly rounded summits, and the shallow undrained basins between, in such regions are an inheritance from a time when long, precipitous escarpments, by their slow recession, left the land covered with a rugged, confused mass of fallen blocks. A review of the facts concerning the minor features in the relief of the broad wheat lands of southeastern Washington,[2] in the light of the conclusions here presented, leads to the suggestion that some of the ridges and basins of that region may be due to the recession of cliffs produced by stream erosion, [any portions of the deeply decayed surface of the basaltic plateau of southeastern Washington resemble closely the old landslide topography in the valley to the northwest of Lookout Mountain, shown in the accompanying illustration.
- ↑ A Geological Reconnoissance in Central Washington. By Israel C. Russell. United States Geological Survey Bulletin, No. 108, 1893, pp. 90-92.
- ↑ A Reconnoissance in Southeastern Washington. By Israel C. Russell. Water Supply and Irrigation Papers of the United States Geological Survey, No. 4, 1887, pp. 58-69.
