After dinner we pass through a region of the wildest desolation. The canyon is very tortuous, the river very rapid, and many lateral canyons enter on either side. These usually have their branches, so that the region is cut into a wilderness of gray and brown cliffs. In several places these lateral canyons are separated from one another only by narrow walls, often hundreds of feet high,--so narrow in places that where softer rocks are found below they have crumbled away and left holes in the wall, forming passages from one canyon into another. These we often call natural bridges; but they were never intended to span streams. They would better, perhaps, be called side doors between canyon chambers. Piles of broken rock lie against these walls; crags and tower-shaped peaks are seen everywhere, and away above them, long lines of broken cliffs; and above and beyond the cliffs are pine forests, of which we obtain occasional glimpses as we look up through a vista of rocks. The walls are almost without vegetation; a few dwarf bushes are seen here and there clinging to the rocks, and cedars grow from the crevices--not like the cedars of a land refreshed with rains, great cones bedecked with spray, but ugly clumps, like war clubs beset with spines. We are minded to call this the Canyon of Desolation. The wind annoys us much to-day. The water, rough by reason of the rapids, is made more so by head gales. Wherever a great face of rocks has a southern exposure, the rarefied air rises and the wind rushes in below, either up or down the canyon, or both, causing local currents. Just at sunset we run a bad rapid and camp at its foot. July 9.--Our run to-day is through a canyon with ragged, broken walls, many lateral gulches or canyons entering on either side. The river is rough, and occasionally it becomes necessary to use lines in passing rocky places. During the afternoon we come to a rather open canyon valley, stretching up toward the west, its farther end lost in the mountains. From a point to which we climb we obtain a good view of its course, until its angular walls are lost in the vista. July 10.--Sumner, who is a fine mechanic, is learning to take observations for time with the sextant. To-day he remains in camp to practice. Howland and I determine to climb out, and start up a lateral canyon, taking a barometer with us for the purpose of measuring the thickness 192 CANYONS OF THE COLORADO. of the strata over which we pass. The readings of the barometer below are recorded every half hour and our observations must be simultaneous. Where the beds which we desire to measure are very thick, we must climb with the utmost speed to reach their summits in time; where the beds are thinner, we must wait for the moment to arrive; and so, by hard and easy stages, we make our way to the top of the canyon wall and reach the plateau above about two o' clock.