Had the earth, at its formation, been a solid globe, four times as hard as hammered iron at the centre, and gradually lessening in density towards the surface, we must admit that it would still be solid matter. Governing ourselves by these principles, how can we imagine that such immense caverns, filled with combustible matter, as would be necessary to supply those volcanoes from time immemorial, could have existed? However, that they do exist is certain, which I think is in no way more easily accounted for, than on the plan of a _mid-plane-space_, or of spaces, filled with a certain hydrogenous gas, which being much lighter than atmospheric air, if there should be any small aperture or crevice extending from the surface to the space beneath, the gravity of the outer part of the sphere pressing on it would occasion a portion of this gas to escape through the aperture; and as it comes in contact with the oxygen of the atmosphere would take fire and occasion those tremendous explosions which we know do sometimes take place and cause those mountains to burn for years, until the cavity which supplied the volcanic matter, becomes exhausted; or until some shock or convulsion consequent on the burning, may have loosened rocks or earth of the denser part of the sphere, which falling into the aperture, choke it up. Hence the gas ceasing to escape, the volcano would cease to burn, until some shock or accident should again open the aperture. The elastic fluid, with which the _mid-plane_ cavities are filled, being forced out into the common atmosphere, the greater degree of gravity would condense and set free its latent heat or caloric, and be resolved into its original base, somewhat as coal-gas, out of the tube of a gas-light apparatus, yields up its latent heat by condensation. Hence steam burns when mixed with coal-gas.