At any point of the surface there are variations of temperature, depending on external causes. But these variations are found to extend only a little way below the surface,--never more than a hundred feet. At greater depths, it is found that the temperature invariably increases with the depth. Deep mines have always a temperature above the mean annual temperature at the surface. The water obtained by deep boring is always tepid when it comes to the surface. The thermal springs, so abundant in this country and in Europe, are so situated as to justify the impression that their waters come from great depths. To make these general observations of any value, we must determine the law by which the temperature increases. The result of all the observations yet made, in mines and upon wells and springs, is that, below the first hundred feet, the temperature increases by one degree of Fahrenheit's scale for every forty-five feet. Regarding this law of increment as applicable to all depths, at ten miles below the surface we should have a temperature above that produced by the combustion of wood; and at twenty-five miles, a temperature of three thousand degrees, by which nearly all mineral substances would be reduced to a state of fusion. The general conclusion of a temperature sufficient to melt the mineral substances of which rocks are composed, at no considerable distance below the surface, is confirmed by the fact that portions of the interior of the earth--at least, at the volcanic centres--are in a melted state. The intimate connection between some volcanoes situated a hundred miles or more apart, so that they are alternately in a state of activity and rest, indicates that these centres are connected,--that subterranean melted lava extends from one to the other, so that when one is active, the elastic force is relieved at the other. These deep-seated lakes of lava must therefore underlie large areas. We are justified, then, in concluding that the mass of the earth, with the exception of a comparatively thin superficial layer, has a very high temperature. By way of accounting for this temperature, it is now generally assumed that the earth was originally in a state of fusion; that it was a mass of liquid lava (if, indeed, it had not a temperature sufficient to reduce it to the aëriform state). Starting with this assumption, there must necessarily be a gradual reduction of temperature by radiation, and a time must arrive when the surface would be crusted over with solidified lava; and this crust would increase in thickness as the cooling advanced, the interior still retaining its heat and liquidity. The present condition of the crust of the earth, its form, that of an oblate spheroid, with the exact difference of the equatorial and polar diameters which is found to exist, as well as the phenomena of volcanic eruptions, will all admit of explanation on this hypothesis.