340 PROCEEDINGS OF THE CALIFORNIA 



clear that there has been a change in the level of the water since the forests 

 grew. Either the land has sunk or the water has been raised : the latter appears 

 to have been the fact. 



The river at the Cascades, just below the submerged forests, plunges over 

 great masses of a hard volcanic conglomerate, which forms the base of the cliffs 

 on each side. This conglomerate, which is 150 to 200 feet thick, rests upon a 

 stratum of sandy clay. This stratum is much softer than the conglomerate, and 

 yields more rapidly to the action of running water. It may be seen when the 

 water is low. at the foot of the Cascades, with the hard conglomerate overhang- 

 ing it in large masses. 



From all these facts, it appears that the river, in cutting its way downwards 

 through the- Cascade Range, reached this soft substratum, and for a long time 

 flowed in a comparatively unobstructed channel at a much lower level than now, 

 thus permitting the forests to grow along its banks. The extensive undermin- 

 ing of the conglomerate caused it at length to fall into the stream, and this, 

 together with the sliding in of the banks upon this soft foundation, I regard as 

 forming the obstruction which dammed the waters and caused the overflow of 

 the forests above. 



The mountains rise on each side to a height of 2,500 or 3,000 feet, and are 

 composed of nearly horizontal beds of lava. One of these mountains on the 

 right bank, or Washington Territory side, presents vertical cliffs towards the 

 Cascades, and these cliffs have a, freshly broken appearance, as if a large part of 

 the mountain had broken off at no very remote period. The surface of the 

 country between this cliff and the Cascades is very much broken, and the rail- 

 road which traverses it, exposes enormous masses of the conglomerate, piled con- 

 fusedly together as if they had been hurled down by a land-slide. Mr. Brazee, 

 the engineer of the Oregon Steam Navigation Company, informs me that the 

 ground is in constant motion toward the river, and it has necessitated the relay- 

 ing of the track within the past year. The same phenomena have been obsi rved 

 on the left bank, or Oregou side. The bank is in constant motion there, and at 

 low-water a fine blue clay may be seen rising in the channel, as if crowded out 

 by the pressure of the rocks above. As there has not been any perceptible 

 change of level in the stream for years past, we may conclude that the erosive 

 action of the current is fully equal to the encroachment of the banks. 



The Indians of the Columbia have a tradition of a great convulsion at the 

 Cascades. They assert tha$ the Chinook canoes formerly ascended the river as 

 far as a water-fall at the Dalles, passing, at the Cascades, under a bridge of rock. 

 This bridge, or arch of rock, they say, fell in at the time of a quarrel between 

 the two mountains, Mt. Hood and St. Helen's, and at the same time the water- 

 fall at the Dalles was destroyed, so that salmon could ascend to the Upper 

 Columbia. Before that the fall was so high that salmon could not get up, and 

 all the upper country Indians were obliged to go to" the Dalles for their supply. 



The general accuracy of this tradition seems highly probable. The Dalles 

 are now a succession of rapids and low falls, in a narrow channel, between 

 vertical walls of basaltic lava. There is very little fall or current in the river 

 below the Dalles to the Cascades, and the elevation of the water by an obstruc- 



