In offering these remarks on the causes of volcanic action, I take it as proved that volcanoes are in communication with a reservoir of heat, occupying the centre of the earth: which heat, however, the earth is gradually, though very slowly, losing; partly by volcanic eruptions, partly by hot springs, and partly by the slow conduction of heat towards the surface through the strata.

I also take it as proved, that the earth is not fluid but solid throughout; or, at least, that its solid crust is of such thickness that the earth is to be regarded for all purposes as a solid body. Consequently the production of lava—that is to say, the liquefaction from heat of a small portion of the solid mass of the earth—is a strictly local occurrence; this is sufficiently proved by the well-known fact, that the lava stands at different levels in different volcanic craters, even when they are in eruption at the same time, and very near each other: as in the case of Mauna-Loa and Kilanea, in Hawaii (Scrope on Volcanoes, p. 262), which would be impossible if there were any free hydraulic communication between the different reservoirs of lava.