And it is to be remarked that the glacier, though composed of solid ice, behaves exactly like a river, or rather like a river of some viscous fluid like pitch or treacle. Its size depends on the magnitude of the reservoir or area drained by it; it conforms to the configuration of the valley by which it descends and the obstacles which it encounters; it flows rapidly, and with a broken current, through narrow gorges and down steep inclines; slowly and tranquilly over wide and level areas; its velocity is greatest at the surface and in the middle where friction is least, slowest at the bottom and sides where it is greatest. In short a glacier is simply a solid and slowly-flowing river, discharging an excess of solid ice to the lower level from which it came, just as a liquid river does with the rainfall of warmer regions. The cause of this tendency of solid and brittle ice to flow like a viscous fluid is not quite understood, though recent researches, especially those of Tyndall, have thrown a good deal of light upon it; but all glacialists are agreed on the _fact_ that it does so, and we can argue from it with great confidence as to the conditions under which glaciation has acted in the past and is now acting. Thus even if Namsen had never crossed Greenland, or Ross had never discovered Mounts Erebus and Terror, we might have inferred with certainty the existence of enormous ice-caps, implying continental masses of elevated land, in both the Arctic and Antarctic circles, from the number and size of the icebergs floated off into the Northern Atlantic and Southern Pacific Oceans. Icebergs are frequently met with in the latter down to 50° south latitude, or even lower, of a mile in length and 500 feet high above the sea; and in some instances icebergs three miles long and 1000 feet high have been recorded. As upwards of eight feet of ice must be under water for every foot that floats above it, some of these icebergs must be considerably over a mile in thickness, which implies that there must be land ice towards the south pole so thick that it is, in places, over 5000 feet in thickness at its outer margin. It has been estimated from the great size and abundance of these icebergs, that in the interior of the great Antarctic continent the ice may be twenty miles or more thick, and in Greenland the great interior ice-cap rises in a dome to at least 9000 or 10,000 feet above the sea-level, a great part of which is solid ice, while during the great glacial period it was certainly very much thicker.