precession and nutation, provided the ellipticities of the interior and exterior surface of the supposed solid crust were equal, and the density of the crust and fluid equal and uniform. But if these limitations were not observed; if the solid shell and interior fluid were heterogeneous, and the ellipticities of the interior and exterior surface of the crust different, then the amount of the precession and nutation would depend on the difference between the ellipticities of the interior and exterior surface of the crust, and on its thickness: or on this latter quantity alone, if the solidity of the shell resulted from refrigeration. And the result of the whole inquiry appears to be, that the thickness of the solid crust cannot be less than th or th of the radius of its external surface.[1] This conclusion is probably decisive against any universal ocean of molten rock below us, at depths accessible to the disturbing agents which generate earthquakes and volcanos: but it seems not to preclude the admission of limited fluid masses, at various and far smaller depths than 1000 or 800 miles. In harmony with this view is the opinion of Mr. C. Darwin, who, from considering the circumstances which accompany volcanos and earthquakes in the Cordilleras of the Andes, proposes, as a fundamental point of reasoning, the recognition of the existence of a vast internal sea of melted rock below a large part of South America.[2]
This conclusion appears liable to so little objection; it is, besides, so perfectly in harmony with the fact historically proved of the perpetual readiness of volcanos for action, and with the geological inference of the perhaps unlimited extent below our feet of rocks once fused; that we shall venture to adopt it as a datum sufficiently established, and applicable to the whole series of volcanic phenomena, in every country, and during all past periods of time.
But this ocean of melted rock may sleep, and does remain at rest, beneath enormous areas, for centuries, or much longer periods, till some particular causes concur to "change (as Mr. Darwin expresses it) the form of
