The beds of glaciers suffer denudation.

the avalanche and the glacier bring down débris with them to be disintegrated below. Indeed, glaciers bring down their huge quota, ready ground for exportation by the torrents which they generate. According to Saussure, four of these main channels, or trunk lines of denudation, are continuous from the top of Mont Blanc, and from these the débris of the monarch of mountains' are traced for great distances.

6

That the beds of the valleys of glaciers suffer denudation like those of other valleys, I think we may infer for two reasons. First, the grooves or lines scratched on the sides of the valleys by the passage of the glaciers are discovered far above their present level. Next, if this were not so, there would be an abrupt break or step from the glacier valley to the valley in prolongation of it.

Besides the moraines or heaps of undigested detritus deposited by glaciers at their lowest ends, the streams which issue from them are always loaded with mud. These streams are flooded in the spring and summer, when the snows melt, to such a degree that, Saussure says, they raise the level of the Lake of Geneva six feet in that There is a common belief that the waters of the Rhone pass through the lake without mingling with it. The river, however, enters the

season.

lake as muddy as possible, and leaves it as pellucid as possible. Perfectly clear water reflects the colour of the sky. And the waters of the lake and of the sea are as blue as the sky above them in clear weather; but in cloudy weather they are lead-colour.

The same argument will prove that the parts

of mountains covered with perpetual snow suffer aqueous denudation precisely the same as all other parts of the earth. For if not, along the lower line of snow we should find terraces the reverse of those of Glen Roy, and similar to some which I shall mention in the next chapter; that is, the terraces would be above the denuded parts instead of below them. Summer rains, and the melting of snow by summer suns, percolate the snow, and what does not enter the ground runs off it, and denudes it the same as if there were no snow. But it is a fallacy to suppose that if a 'glacial epoch' had ever existed, a greater denudation would have gone on than now goes on by rain and rivers.

The cause of the descent of glaciers is simply the pressure of snow from above their upper ends. Mr. Hopkins (Philosophical Magazine,' October 1859), is as unfortunate here with his glacial mathematics as he is in the Weald with his igneous mathematics. He talks of the tendency of such

6

Land covered by eternal

snow suffers

denudation.

an enormous weight of ice to descend down a plane even of very small inclination;' and he mentions slopes at inclinations sometimes not exceeding 2° or 3°,' and he assures us twice over, that his conclusions were as certain as that of the parallelogram of forces, and no theory which contradicted them could possibly be true.' Very much so.

6

But then, no doubt, it equally as true is

that 'an enormous weight of ice' can no more slide down a slight incline than a small weight of ice can. Or if it can, perhaps the mathematician will tell us why it can. And on slopes of two or three degrees, and on all the slopes of the lower ends of glaciers, solids large or small (and among them ice) would rest motionless. If a glacier could be extracted from its valley, the valley would be instantly filled again by the fall of the snow from above into it, which snow was before supported by the upper end of the glacier. It is only the pressure of this snow from above, which pushes the glacier down; abstract this pressure from above, that is, let all the snow be abstracted from Mont Blanc, and let no more snow fall, every glacier would cease instantly to move downward. They would retreat upward by melting, and would finally disappear; even above the former snowline, they would melt in summer suns. But as snow constantly does fall on Mont Blanc, and as it

cannot melt there in the same proportion as it falls, it would constantly accumulate there unless it was constantly shoving the glaciers down towards the sea, as rain is constantly shoving the whole surface of the earth down towards the sea. Owing to the friction of the bed of the valley, the central ice yields most easily to the pressure from above, and travels most quickly. This principle obtains also in rivers, though water is neither viscous nor brittle. But Professor Forbes is unfortunate in calling ice (one of the most brittle things we know), viscous. A glacier travels at different rates in different parts, or alters its form to the form of the valley on the same principle that holds in the ice-house, that is, by being pounded to atoms and freezing solid again. This results from extreme brittleness, the very reverse of viscousness.

Though ice and snow perpetually surround That ice and the two poles, it is not the same ice and snow,

any more than the snow on the mountain top or the water in the river continues the same. And the idea that ice and snow must increase at the poles is fallacious. Every year thousands of enormous mountains, we might say continents, of ice (all bearing mud, earth, and stones) float off, and are melted in more temperate latitudes. The causes of increase or decrease of snow which may

snow must in

crease at the

poles is falla

cious.

Mountains while rising may be decreasing in height.

occur at the poles, are shown by a magnificent theory of the illustrious Lyell.

It does not follow that while mountains are rising they are increasing in height. They may be decreasing in height. Suppose the Alps to have been rising six inches in a century for myriads of years, if their denudation has been seven inches in a century, they have been decreasing in height.

It is true that the direct action in waste and denudation of torrents and rivers is on lines only; and were it not for the atmospheric disintegration and the lateral wash of rain this their direct action would only cut ravines and channels to the sea, and the sides of all valleys, instead of sloping, would be cliffs; that is, where a spring issues high up the mountain side, it will cut a deep ravine with precipitous cliffs, and the deeper it cuts, the more springs it will lay open. But what widens this ravine into a broad valley with gently sloping sides? The lateral wash of rain into the longitudinal valley. And what forms the broad valley, even where there is no river at the bottom? or within many miles? The longitudinal scooping power of the concentrated wash of rain, which in no respect differs from that of the torrent, except in its being a hundred-fold more powerful than the torrent. It is indeed intermittent; so is