resented by the large arrows, which descended from the flanks of Mounts Dana, Gibbs, Ord, and others to the south. For thirty miles farther, we find everywhere displayed the same delicate yielding to glacial law, showing that, throughout the whole period of its formation, the huge granite valley was lithe as a serpent, and winced tenderly to the touch of every tributary. So simple and sublime is the dynamics of the ancient glaciers.

Every valley in the region gives understandable evidence of having been equally obedient and sensitive to glacial force, and to no other. The erosive energy of ice is almost universally underrated, because we know so little about it. Water is our constant companion, but we cannot dwell with ice. Water is far more human than ice, and also far more outspoken. If glaciers, like roaring torrents, were endowed with voices commensurate with their strength, we would be slow to question any ascription of power that has yet been bestowed upon them. With reference to size, we have seen that the greater the icefountains the greater the resulting valleys; but no such direct and simple proportion exists between areas drained by water streams and the valleys in which they flow. Thus, the basin of Tenaya is not one-fourth the size of the South Lyell, although its cañon is much larger. Indeed, many cañons have no streams at all, whose topographical circumstances are also such as demonstrate the impossibility of their ever having had any. This state of things could not exist if the water streams which succeeded the glaciers could follow in their tracks, but the mode and extent of the compliance which glaciers yield to the topography of a mountainside, is very different from that yielded by water streams; both follow the lines of greatest declivity, but the former in a far more general way. Thus, the greater portion of the ice-current which eroded Tenaya Cañon flowed over the divide from the Tuolumne region, making an ascent of over 500 feet. Water streams, of course, could not follow; hence the dry channels, and the disparity, to which we have called attention, between Tenaya Cañon and its basin.

Anyone who has attentively observed the habits and gestures of the upper Sierra streams, could not fail to perceive that they are young, and but little acquainted with the mountains; rushing wildly down steep inclines, whirling in pools, sleeping in

lakes, often halting with an embarrassed air and turning back, groping their way as best they can, moving most lightly just where the glaciers bore down most heavily. With glaciers as a key the secrets of every valley are unlocked. Streams of ice explain all the phenomena; streams of water do not explain any; neither do subsidences, fissures, or pressure plications.

We have shown in the previous paper that post-glacial streams have not eroded the 500,000th part of the upper Merced cañons. The deepest water gorges with which we are acquainted are between the upper and lower Yosemite falls, and in the Tenaya Cañon about four miles above Mirror Lake. These are from twenty to a hundred feet deep, and are easily distinguished from ice-eroded gorges by their narrowness and the ruggedness of their washed and pot-holed sides.

The gorge of Niagara River, below the falls, is perhaps the grandest known example of a valley eroded by water in compact rock; yet, comparing equal lengths, the glacier-eroded valley of Yosemite is a hundred times as large, reckoning the average width of the former 900 feet, and depth 200. But the erosion of Yosemite Valley, besides being a hundred times greater, was accomplished in hard granite, while the Niagara was in shales and limestones. Moreover, Niagara cañon, as it now exists, expresses nearly the whole amount of erosion effected by the river; but the present Yosemite is by no means an adequate expression of the whole quantity of glacial erosion effected there since the beginning of the glacial epoch, or even from that point in the period when its principal features began to be developed, because the walls were being cut down on the top simultaneously with the deepening of its bottom. We may fairly ascribe the formation of the Niagara gorge to its river, because we find it at the upper end engaged in the work of its further extension toward Lake Erie; and for the same reason we may regard glaciers as the workmen that excavated Yosemite, for at the heads of some of its branches we find small glaciers engaged in the same kind of excavation. Merced cañons may be compared to mortises in the ends of which we still find the chisels that cut them, though now rusted and worn out. If Niagara River should vanish, or be represented only by a small brook, the evidence of the erosion of its gorge would still re

main in a thousand water-worn monuments upon its walls. Nor, since Yosemite glaciers have been burned off by the sun, is the proof less conclusive that in their greater extension they excavated Yosemite, for, both in shape and sculpture, every Yosemite rock is a glacial monument.

When we walk the pathways of Yosemite glaciers and contemplate their separate works-the mountains they have shaped, the cañons they have furrowed, the rocks they have worn, and broken, and scattered in moraines-on reaching Yosemite, instead of being overwhelmed as at first with its uncompared magnitude, we ask, Is this all? wondering that so mighty a concentration of energy did not find yet grander expression.