indicates that the action of the ice has not been continued long enough to obliterate all previous inequalities. The débris must have been deposited after the moulins had ceased to act, and in all probability during the retreat of the glacier. These "kettles" were found, unless appearances are misleading, within a very few feet of the surface, and the overlying material is not at all characteristically morainic; moreover, a rather careful search in all accessible places in the neighbourhood failed to show more than slight traces of any such deposit.

In short, at the time of the maximum extension of the ice-sheet, almost the whole of the Swiss lowland was buried, and the ice welled up against the flanks of the Jura to a height of 3000 feet above the Lake of Neuchâtel, whence it extended northward to the neighbourhood of Soleure. Along the present course of the Rhone it sent out a huge lobe far beyond the frontier of Switzerland, for erratics and other glacial deposits have been traced to within a few miles of Lyons. It is estimated that altogether the ancient glacier of the Rhone was not less than 270 miles in length.

The glaciers on the southern sides of the Alps were hardly less extensive than on the northern. They descended the valleys, they passed over the sites of the Italian lakes and on to the plains of Piedmont and Lombardy. Round one group of moraines the tide of battle ebbed and flowed in the struggle of Solferino;

another is a memorial of a glacier which passed over the site of Como, up the slopes and over the sandstone ridge enclosing that arm of the lake to the lowland beyond; a third helps to form the lake of Varese; others, farther west, rise like hills from

FIG. 7.-Map showing the lines of débris extending from the Alps into the plains of the Po (after Lyell). A, Crest of the Alpine watershed; B, Névé fields of the ancient glaciers; C, Moraines of ancient glaciers.

the plain of Piedmont; the most remarkable and extensive being those opposite to the opening of the Val d'Aoste. This moraine forms a curved chain of hills, with a frontage of over fifty miles, which rise sometimes to a height of more than 1500 feet.

In short, the glaciers of the Alps-and the statement holds good, as will be seen presently, of other mountain regions are but the dwindled representatives of gigantic predecessors, to which, at any rate on the Swiss lowland, we can hardly refuse the name of ice-sheets.

CHAPTER II

ARCTIC AND ANTARCTIC ICE-SHEETS

LAND-ICE has left its marks on the Alpine region, but in Greenland it is still in possession. The Glacial Epoch belongs to the past in the one, to the present in the other. Every process which has sculptured the surface and formed the glacial deposits of lands wherein a milder climate now prevails, should be found at work in Greenland and the adjacent parts of the Arctic region. Here the results which, in more southern countries, are subjects for conjecture, should be actual matters of fact. Of late years much has been learned about Greenland itself, not only from the members of sundry expeditions to circumpolar regions, but also from the special investigations of Steenstrup, Nordenskiold, Nansen, Peary, and others. From the accounts which have been published we shall endeavour to select those facts which seem likely to throw light on the behaviour and work of ice during a Glacial Epoch.

Greenland is the only very important land mass which, in the Northern Hemisphere, extends beyond the seventy-ninth parallel of latitude. Over almost