In such a condition its denser materials, such as the heavier metals, would settle toward the centre, and the surface would consist of lighter material composed of the less dense and more oxidizable substances combined with oxygen, and similar in character and appearance to the slag which forms on the surface of some ores in the process of smelting. Of this slaggy material there might, however, be different layers more or less dense in proceeding from the interior to the surface. This molten surface would, of course, radiate heat into space; and as it would naturally consist of the least fusible matters, these would begin to form a solid crust. We may imagine this crust at first to be smooth and unbroken, though such a condition could scarcely exist for any length of time, as the hardened crust would certainly be disturbed by ascending currents from within, and by tidal movements without. Still, it might remain for ages as a spheroidal crust, presenting little difference of elevation or depression in comparison with its extent. When it became sufficiently thick and cool to allow water to lie on its surface, new changes would begin. The water so condensed would be charged with acid substances which would begin to corrode the rocky surface. Penetrating into crevices and flashing into steam as it reached the heated interior, it would blow up masses and fragments of stone, and would perhaps force out and cause to flow over the surface beds of molten material from below the crust, and differing somewhat from it in their composition. All this aqueous work would accelerate the cooling and thickening of the crust, and at length a universal or almost universal heated ocean would envelope the globe, and so far as its surface was concerned, the reign of water would replace that of fire. We may pause here to consider the probable nature of the earth's crust in this condition.