remark that the crust abounds most in the oxides of those metals which have the strongest affinity for oxygen, as the alkalies and alkaline earths; while in the peridotic and lower zones the proportion of these elements is much less, and that of the earths and metals is much greater. The minerals composing the superficial crystalline rocks, as well as water, are generally absent from the meteorites. This is especially noticed in respect to the mineral quartz or silica, so common at the surface. According to these views, the granites must once have been in a melted condition, and the excess of silica present in them have assumed the amorphous form. Many geologists have supposed the silica ought to have crystallized first, if the rock cooled from fusion. It may be that our ideas of the intense heat have been exaggerated; yet the Labrador granites of New Hampshire have recently been shown by us to be situated in sheets over a plain, precisely like the erupted lava of the present day.
We have dwelt upon the present concentric structure of the earth, because it was probably the same with that existing in the igneous period, at that time fused, but now largely solid. The order of the alternations has always been the same. It corresponds also with that observed in furnaces, where the metal sinks to the bottom, and is overlaid by one or more successive layers of slag.
This complex sphere, when molten, with its fiery billows and igneous currents, being situated in a fearfully cold region, could not fail to radiate heat; and, like other melted bodies, become covered with a congealed crust. A pot of melted iron taken out of the fire loses heat, and a crust speedily forms over it, shrinking as it cools; and, if the exterior be broken, the red liquid may be poured out. The same thing may be seen on the dumping-heaps connected with melting-works. Masses of slag, with their entire surface congealed, are placed upon the car and wheeled to the end of the pile; but, when thrown down the slope, they are fractured, and the liquid interior flows out like water. When a stream of lava flows down a slope, the surface and sides of the molten river are soon covered by a thick crust, the result of cooling. This will become so firm that men may walk upon it, as upon ice over lakes in the winter. During one of the eruptions from Vesuvius, when lava covered the town of Resina—the old Herculaneum—some of the inhabitants, driven to the tops of the houses, escaped by walking over the stiffened crust, before the flow had ceased. Whenever the lateral walls of the stream are broken, the lava will flow out and change its course. In this way, a current threatening to engulf a village may be averted and directed elsewhere. This is a practical matter, and has been turned to account in Sicily, in warding off from Catania the threatened calamity rolling down the slopes of Etna.
Our entire experience, therefore, of analogous phenomena, leads us to believe that a crust will be formed, and that the several zones will cool in natural order in later periods. Not till the last melted layer
