As regards the earth-crust, we know from observations in bore-holes made in different parts of the world that the temperature increases rather rapidly with the depth, on an average by about thirty degrees Cent. per kilometre (about 1.6° F. per 100 feet). It must be remarked, however, that the depth of our deepest bore-holes hardly exceeds 2 km. (Paruchowitz, in Silesia, 2003 m., or 6570 ft.; Schladebach, near Merseburg, Prussian Saxony, 1720 m.). If the temperature should go on increasing at the rate of 30 degrees Cent. for each further kilometre, the temperature at a depth of 40 kilometres should attain degrees at which all the common rocks would melt. But the melting-point certainly rises at the same time as the pressure. The importance of this circumstance was, however, much exaggerated when it was believed that for this reason the interior of the earth might possibly be solid. Tammann has shown by direct experiments that the temperature of fusion only rises up to a certain pressure, and that it begins to decrease again on a further increase of pressure. The depths indicated above are therefore not quite correct. If we assume, however, that other kinds of rock behave like diabase—the melting-point of which, according to the determinations of Barus, rises by 1° Cent. for each 40 atmospheres of pressure corresponding to a depth of 155 m.—we should conclude that the solid crust of the earth could not have a greater thickness than 50 or 60 km. (40 miles). At greater depths we should therefore penetrate into the fused mass. On account of its smaller density the silicic acid will be concentrated in the upper strata of the molten mass, while the basic portions of the magma, which are richer in iron oxide, will collect in the lower strata, owing to their greater density. This magma we have to picture to ourselves as an extremely viscid liquid resembling asphalt. The experiments of Day and Allen show that rods, supported at their ends, of 30 × 2 × 1 mm. of different minerals, like the feldspars microcline and albite, could retain their shape for three hours without curving noticeably, although their temperature was about a hundred degrees above their melting-point, and although they appeared completely fused, or, more correctly, completely vitrified when taken out of the furnace. These molten silicates behave very differently from other liquids like water and mercury, with which we are more accustomed to deal.