During this excursion, Mr. Banks had an excellent opportunity to examine the rocks, which were almost every where naked, for minerals; but he found not the least appearance of any. The stones every where, like those of Madeira, showed manifest tokens of having been burnt; nor is there a single specimen of any stone, among all those that were collected in the island, upon which there are not manifest and indubitable marks of fire; except, perhaps, some small pieces of the hatchet-stone, and even of that, other fragments were collected which were burnt almost to a pumice. Traces of fire are also manifest in the very clay upon the hills; and it may, therefore, not unreasonably be supposed, that this and the neighbouring islands are either shattered remains of a continent, which some have supposed to be necessary in this part of the globe, to preserve an equilibrium of its parts, which were left behind when the rest sunk by the mining of a subterraneous fire, so as to give a passage to the sea over it; or were torn from rocks, which, from the creation of the world, had been the bed of the sea, and thrown up in heaps, to a height which the waters never reach. One or other of these suppositions will perhaps be thought the more probable, as the water does not gradually grow shallow as the shore is approached, and the islands are almost every where surrounded by reefs, which appear to be rude and broken, as some violent concussion would naturally leave the solid substance of the earth. It may also be remarked upon this occasion, that the most probable cause of earthquakes seems to be the sudden rushing in of water upon some vast mass of subterraneous fire, by the instantaneous rarefaction of which into vapour the mine is sprung, and various substances, in all stages of vitrification, with shells, and other marine productions, that are now found fossil, and the strata that covered the furnace, are thrown up; while those parts of the land which were supported upon the broken shell give way, and sink into the gulf. With this theory the phænomena of all earthquakes seem to agree: pools of water are frequently left where land has subsided; and various substances, which manifestly appear to have suffered by the action of fire, are thrown up. It is indeed true, that fire cannot subsist without air; but this cannot be urged against there being fire below that part of the earth which forms the bed of the sea; because there may be innumerable fissures by which a communication between those parts and the external air may be kept up, even upon the highest mountains, and at the greatest distance from the sea-shore.