the friction-markings run in a general uniform direction, which is that of the motion of the glacier. Such striated surfaces could only be produced by some agent with rigidity

enough to hold the sand-grains and stones in position, and press them steadily onward upon the rocks. A river polishes the rocks of its channel by driving shingle and sand across them; but the currents are perpetually tossing these materials now to one side, now to another, so that smoothed and polished surfaces are produced, but with nothing at all resembling striation. A glacier, however, by keeping its grinding materials fixed in the bottom of the ice, engraves its characteristic parallel striæ and groovings, as it slowly creeps down the valley. All the surfaces of rock within reach of the ice are smoothed, polished, and striated. Such surfaces present the most unmistakable evidence of glacieraction, for they can be produced by no other known natural agency. Hence, where they occur in glacier valleys, far above and beyond the present limits of the ice, they prove how greatly the ice has sunk. In regions also where there are now no glaciers, these rock-markings remain as almost imperishable witnesses that glaciers once existed. By means of their evidence, for example, we can trace the march of great ice-sheets which once enveloped the whole of Scandinavia and lay deep upon nearly the whole of Britain.

The river that escapes from the end of a glacier is always muddy. The fine sand and mud that discolour the water are not supplied by the thawing of the clear ice, nor by the sparkling brooks that gush out of the mountain-slopes, nor by the melting of the snows among the peaks that rise on either side. This material can only come from the rocky floor or the glacier itself. It is the fine sediment ground away from the rocks and loose stones by their mutual friction under the pressure of the overlying ice. It is thus a kind of index or measure of the amount of material worn off the rocky bed by the grinding action of the glacier. We can readily see that as this erosion and transport are continually in progress, the amount of material removed in the course of time must be very great. It has been estimated, for example, that the Justedal glacier in Norway removes annually from its bed 2,427,000 cubic feet of sediment. At this rate the amount removed in a century would be enough to fill up a valley or ravine 10 miles long, 100 feet broad, and 40 feet deep.

In arctic and antarctic latitudes, where the land is buried under a vast ice-sheet, which is continually creeping seaward and breaking off into huge masses that float away as icebergs, there must be a constant erosion of the terrestrial surface. Were the ice to retire from these regions, the ground would be found to wear what is called a glaciated surface; that is to say, all the bare rocks would present a characteristic ice-worn aspect, rising into smooth rounded bosses like dolphins' backs (roches moutonnées), and sinking into hollows that would become lake-basins. Everywhere these bare rocks would show the striæ and groovings graven upon them by the ice, radiating generally from the central high grounds, and thus indicating the direction of flow of the main streams of the ice-sheet. Piles of earth, ice-polished stones, and blocks of rock would be found strewn over the country, especially in the valleys and over the plains. These materials would still further illustrate the movements of the ice, for they would be found to be singularly local in character, each district having supplied its own contribution of detritus. Thus in a region of red sandstone, the rubbish would be red and sandy; in one of black slate, it would be black and clayey (see chapter xxvi.)

Summary. In this chapter we have seen that ice in various ways affects the surface of the land and leaves its mark there. Frost pulverises soil, disintegrates exposed surfaces of stone, and splits open bare rocks along their lines of natural joint. On rivers and lakes, the disrupted ice wears down banks and pushes up mounds of sand, gravel, and boulders along the shores. In the condition of avalanches, it causes large quantities of earth, soil, and blocks of rock to be removed from the mountain-slopes and piled up on the valleys. In the form of glaciers, it transports the debris of the mountains to lower levels, bearing along and sometimes stranding masses of rock as large as cottages, which no other known natural agent could transport. Moving down a valley, it wears away the rocks, giving them a peculiar smoothed and striated surface which is thoroughly characteristic. By this grinding action, it erodes its bed and produces a large amount of fine sediment, which is carried away by the river that escapes at the end of the glacier. Land-ice thus leaves thoroughly distinctive and enduring memorials of its presence in polished and grooved rocks, in masses of earth, clay, or gravel, with striated stones, and in the dispersal of erratic blocks from principal masses of high ground. These memorials may remain for ages after the ice itself has vanished. By their evidence we know that the present glaciers of the Alps are only a shrunk remnant of the great ice-fields which once covered that region; that the Scandinavian glaciers swept across what is now the bed of the North Sea as far as the mouth of the Thames; and that Scotland, Ireland, Wales, and the greater part of England were buried under great sheets of ice which crept downwards into the North Sea on the one side, and into the Atlantic on the other.

CHAPTER VII.

L.

THE MEMORIALS OF THE PRESENCE OF THE SEA.

We have now to inquire how the work of the sea is registered in geological history. This work is broadly of two kinds. In the first place, the sea is engaged in wearing away the edges of the land, and in the second place, being the great receptacle into which all the materials, worn away from the land, are transported, it arranges these materials over its floor, ready to be raised again into land at some future time.

I. Demolition of the Land. - In its work of destruction along the coasts of the land, the sea acts to some extent (though we do not yet know how far) by chemically dissolving the rocks and sediments which it covers. Cast-iron bars, for example, are so corroded by sea-water as to lose nearly half their strength in fifty years. Doubtless many minerals and rocks are liable to similar attacks.

But it is by its mechanical effects that the sea accomplishes most of its erosion. The mere weight with which ocean-waves fall upon exposed coasts breaks off fragments of rock from cliffs. Masses, 13 tons in weight, have been known to be quarried out of the solid rock by the force of the breakers in Shetland, at a height of 70 feet above