the margin of the plateau have been very much disturbed, which would easily have obliterated all traces of the original cañon. Entering the body of the plateau, the cañon suddenly assumes a greater breadth. As it increases in depth, it exposes a series of several hundred feet of stratified rocks, below which the granite comes to the surface. At the crest of the plateau the lower two thirds of the walls are of granite, the upper 1,000 feet being sedimentaries. From its northeastern end the cañon gradually widens from a bed-breadth of a hundred yards to one of at least a mile; at the crest, the cañon suddenly narrows to a mere stream-way, with rugged, vertical walls.
It is purely a cañon of erosion. There is no sign whatever of any local disturbance which could account for its existence. The strata on the two sides are perfectly conformable.
To attribute this cañon to the streams now occupying it is manifestly absurd. Not only are they utterly insignificant in comparison with the amount of erosion which has taken place, but no similar streams, of any magnitude whatever, could have cut the cañon down at the divide; neither could the western one have cut back into the plateau to any such extent. There is no lateral slope toward this cañon to determine the drainage of any considerable area of the plateau in this direction. It is plainly the scene of the defeat of a large stream, in its struggle to maintain its ancient course—of a victory of volcanic over aqueous forces. Another thing is equally apparent—that the stream here diverted came from the eastward and not from the westward, and that it was the Grand River. This is shown by the following considerations: (1) The general slope of the country, disregarding local accidents of topography, is from the northeast toward the southwest; (2) the direction of the course of Grand River above this cañon, which is almost precisely in line with (3) the character of its course, which shows that it antedated all other uplifts, and, as the Uncompahgre Plateau, from its trend and association, must have been coexistent with the rest, it must have antedated this also; (4) the features of the cañon itself afford the strongest possible evidence of a stream flowing southwest through it. The profile, with the summit east of the crest of the plateau, the slow descent west of the summit to the crest of the plateau, and the rapid descent beyond the crest, point unmistakably to this conclusion. The plan of the cañon is no less clear in its indications. It is well known that a rapid stream erodes its bed downward; a sluggish one, on the other hand, erodes laterally, thus broadening its bed. Here we have precisely these phenomena. Beyond the crest of the plateau, where the slope is great, the cañon is very narrow; while east of the crest, where the velocity of the stream must have been very much lessened, it widens rapidly, and then gradually diminishes in width.
At what stage in the rise of the Uncompahgre Plateau the river abandoned the unequal contest and took its present course around the
