But it appears that such erosion could be performed by heavy glaciers with the lands at their present level. A glacier three thousand feet thick would fill and press heavily on the bottom of a gorge fifteen hundred feet in depth. Therefore, should the bottom of a fiord sink hundreds of feet below the sea-level, a glacier several thousand feet thick flowing through and over it into a sea of much greater depth, the erosion at the bottom of the sunken channel would be greater than on the land above the sea, where the ice possessed less weight. Therefore, it is not necessary that lands pierced by deep fiords should have acquired a higher level during the ice age than they now maintain. And it is probable that on the antarctic continent ice erosion may be going on at much greater depths below the sea-level than the deepest channels in the high northern latitudes. For it is likely that the temperature of a glacier is so low in such frigid regions that it holds firmly in its freezing grasp such bowlders as may become detached from the rocks, thus giving it great erosive power. But this great eroding ability could not be maintained by glaciers in the lower latitudes, where a higher temperature would largely deprive the ice of its abrading properties except on the steep slopes of mountainous lands. There are deposits of ice on the North American coast bordering the arctic shores, and also on Northern Siberia, that are supposed to have existed since the last frigid period, and are likely to be preserved into a future cold age, which now appears to have made considerable progress on Greenland and other ice-clad arctic shores on account of the independent circulation of the Arctic Ocean waters, which largely excludes the Gulf Stream from the polar seas; and it is for this reason that the glaciers on the elevated lands of Iceland are being enlarged and rapidly advancing. Yet, notwithstanding the gathering of ice and increasing coldness of lands largely removed from the warm Gulf currents, there are still mountain regions where glaciers may have been preserved through post-glacial times, although directly to the leeward and under the influence of the Gulf Stream and Japanese currents. These glaciers are situated in the Alpine districts of Europe and on the mountain ranges of Alaska. It would appear that, were the climate growing gradually colder in the northern temperate zones, such glaciers should be increasing in size. Yet it is said that such is not always the case. This is probably owing to their being subject to the genial influence of the tropical currents. For, although the climate of Europe and Alaska may have been slowly growing colder for centuries, still the slow shrinkage of these once immense glaciers may still be going on, although at a much slower rate than formerly, even if the tender plants of these latitudes, because of the growing coldness, have gradually moved southward.