the valley, and expending the down-thrusting power with which they were endowed by virtue of the declivity of their channels, the trunk flowed up out of the valley without yielding much. compliance to the crooked and comparatively small river cañon extending in a general westerly direction from the foot of the main valley. In effecting its exit a considerable ascent was made, traces of which are to be seen in the upward slope of the worn, rounded extremities of the valley walls. Down this glacier-constructed grade descend both the Coulterville and Mariposa trails; and we might further observe in this connection that, because the ice-sheet near the period of transition to distinct glaciers flowed southwesterly, the south lips of all Yosemites trending east and west, other conditions being equal, are more heavily eroded, making the construction of trails on that side easier. The first trail, therefore, that was made into Yosemite, was of course made down over the south lip. The only trail entering the Tuolumne Yosemite descends the south lip, and so also does the only trail leading into the Kings River Yosemite. A large majority of deer and bear and Indian trails likewise descend the south lips of Yosemites. So extensively are the movements of men and animals controlled by the previous movements of certain snow-crystals combined as glaciers.

The direction pursued by the Yosemite trunk, after escaping from the valley, is unmistakably indicated by its immense lateral moraines extending from its lips in a west-southwesterly direction. The right moraine was disturbed by the large tributary of Cascade Creek, and is extremely complicated in structure. The left is simple until it comes under the influence of tributaries from the southeast, and both are further obscured by forests which flourish upon their mixed soil, and by the washing of rains and melting snows, and the weathering of their boulders, making a smooth, sandy, unmorainelike surface. It is, therefore, the less to be wondered at that the nature of these moraines, which represent so important a part of the chips hewn from the valley in the course of its formation, should not have been sooner recognized. Similarly situated moraines extend from the lips of every Yosemite wherever the ground admits of their deposition and retention. In Hetch-Hetchy and other smaller and younger Yosemites of the upper Merced, the as

cending strie which measure the angle of ascent made by the bottom of their glaciers in their outflow are still clearly visible. Fig. I is the horizontal sec

tion of the end of a Yosemite valley, showing the ordinary boat-shaped edge, and lateral moraines (MM) extending from the lips. The moraines and arrows indicate the course pursued by the outflowing ice. Fig. 2 represents the right lip of Yosemite, situated on the upper Merced below the confluence of Cathedral tributary. The whole lip is polished and stri

M

FIG. 1.

M

ated. The arrows indicate the direction of the stria, which measure the angle of ascent made by the outflowing ice.

டாப

FIG. 2.

In the presentation of these studies, we have proceeded thus far with the assumption that all the valleys of the region are valleys of erosion, and that glaciers were the principal eroding agents; because the intelligible discussion of these propositions requires some knowledge of the physiognomy and general configuration of the region, as well as of the history of its ancient glaciers. Our space is here available only for very brief outlines of a portion of the argument, which will be gradually developed in subsequent articles.

That fossils were created as they occur in the rocks, is an ancient doctrine, now so little believed that geologists are spared the pains of proving that nature ever deals in fragmentary creations of any sort. All of our valleys are clearly fragmentary in some degree. Fig. 3 is a section across Yosemite Valley from

Indian Cañon, which displays the stumps of slabs and columns of which the granite is here composed. Now, the complements of these broken rocks must have occupied all, or part, or more than all of the two portions of the valley, A C D and B E F. The bottom, A B, is covered with drift, but we may assume that if it were laid bare we would find it made up of the ends of slabs and columns like the sides, which filled the space A CEB; because in all valleys where the bottom is naked, the broken stumps do appear, showing that this valley was not formed by a fold in the mountain surface, or by a splitting asunder, or by

B
A=

FIG. 4.

subsidence, but by a breaking up and translation of rocks which occupied its place; or, in other words, by erosion.

Fig. 4 is a section across the lower portion of the valley of Illilouette south of Mount Starr King. In this case the bottom is naked, and the dotted reconstructed portions of the huge granite folds A B C D have evidently been eroded.* Even the smoothly curved trough of two rock-waves which afford sections like Fig. 5 can not be regarded as a valley originating in a

fold of the surface, for we have shown in the first paper of this series that domes or extended waves, with a concentric structure like A C, may exist as concretionary or crystalline masses beneath the surface of granite possessing an entirely different structure or no determinate structure whatever, as in B.

The chief valley-eroding agents are water and ice. Each has been vaguely considered the more influential by different observers, although the phenomena to which they give rise are immensely different. These workmen are known by their chips, and only glacier chips form moraines which correspond in kind and quantity to the size of the valleys and condition of their surfaces. Also their structure unfolds the secret of their origin. The constant and inseparable relations of trend, size, and form which these Sierra valleys sustain to the ice-fountains in which they all head, as well as their grooved and broken sides, proclaim the eroding force to be ice. We have shown in the second paper† that the trend of Yosemite valleys is always a direct resultant of the forces of their ancient glaciers, modified

Water never erodes a wide U-shaped valley in granite, but always a narrow gorge like E F, in Fig. 4.

† Reprinted in SIERRA CLUB BULLETIN, Vol. X, No. 1, January 1916.

by obvious peculiarities of physical structure of their rocks. The same is true of all valleys in this region. We give one example, the upper Tuolumne Valley, which is about eight miles long, and from 2000 to 3000 feet deep, and trends in a generally northerly direction. If we go to its head on the base of Mount Lyell, and follow it down, we find that after trending steadily about two miles it makes a bend of a few degrees to the left (A, Fig. 6). Looking for the cause, we perceive a depression on the

B

FIG. 6.-ILLUstrating BenD OF UPPER TUOLUMNE VALLEY

opposite or right wall; ascending to it, we find the depression to be the mouth of a tributary valley which leads to a cratershaped ice-fountain (B) which gave rise to the tributary glacier that, in thrusting itself into the valley trunk, caused the bend we are studying. After maintaining the new trend thus acquired for a distance of about a mile and a half, the huge valley swerves lithely to the right at C. Looking for the cause, we find another tributary ice-grooved valley coming in on the left, which like the first conducts back to an ice-womb (D) which gave birth to a glacier that in uniting with the trunk pushed it aside as far as its force, modified by the direction, smoothness, and declivity of its channel, enabled it to do. Below this, the noble valley is again pushed round in a curve to the left by a series of small tributaries which, of course, enter on the right, and with each change in trend there is always a corresponding change in width or depth, or in both. No valley changes its direction without becoming larger. On nearing the Big Meadows it is swept entirely round to the west by huge glaciers, rep