In the infancy of Chemistry and Natural Philosophy, it was imagined that earthquakes might be easily explained; in proportion as these sciences have become more correct and more profound, this confidence has decreased. But by a propensity, for which the character of the human mind sufficiently accounts, all the new physical agents which have been successively discovered, such as electricity, magnetism, the inflammation of gases, the decomposition and recomposition of water, have been maintained in theories as the causes of the great phenomena of nature. Now all these conjectures seem to be insufficient to explain convulsions so extensive, produced at the same time over such large portions of the earth, as those which take place during earthquakes. The most probable opinion, the only one which seems to us to reconcile, in a certain degree, the energy, the extent of these phenomena, and often their frightful correspondence in the most distant countries of the globe, would be to suppose, conformably to many other physical indications, that the solid surface on which we live is but of inconsiderable thickness in comparison with the semi-diameter of the terrestrial globe; is in some measure only a recent shell, covering a liquid nucleus, perhaps still in a state of ignition, in which great chemical or physical phenomena operating at intervals cause those agitations which are transmitted to us. The countries where the superficial crust is less thick or less strong, or more recently or more imperfectly consolidated, would agreeably to this hypothesis, be those the most liable to be convulsed and broken by the violence of these internal explosions. Now if we compare together the experiments on the length of the pendulum, which have been made for some years past with great accuracy, from the north of Scotland to the south of Spain, we readily perceive that the intensity of gravitation decreases on this space, as we go from the Pole towards the Equator, more rapidly than it ought to do upon an ellipsoid, the concentric and similar strata of which should have equal densities at equal depths; and the deviation is especially sensible about the middle of France, where too there has been observed a striking irregularity in the length of the degrees of the earth. This local decrease of gravity in these countries should seem to indicate, with some probability, that the strata near the surface must be less dense there than elsewhere, and perhaps have in their interior immense cavities. This would account for the existence of the numerous volcanos of which these strata show the traces, and explain why they are even now, at intervals, the focus of subterraneous convulsions. [32] Cook's Geography, v. 2, p. 250――Also Rees' Cyclopedia, article Lake. [33] Rees' Cyclopedia, article Lake Geneva. [34] Rees' Cyclopedia, article Lake. [35] Ross' Voyage, v. 1, p. 225. [36] Ibid, v. 1, p. 144. [37] Humboldt. CHAPTER VII.