Humboldt states that, according to tolerably coincident experiments in artesian wells, it has been shown that the heat increases on an average about 1° for every 54·5 feet. If this increase can be reduced to arithmetical relations, it will follow that a stratum of granite would be in a state of fusion at a depth of nearly twenty-one geographical miles, or between four and five times the elevation of the highest summit of the Himalaya. The following is the opinion of Professor Silliman: That the whole interior portion of the earth, or at least a great part of it, is an ocean of melted rock, agitated by violent winds, though I dare not affirm it, is still rendered highly probable by the phenomena of volcanoes. The facts connected with their eruption have been ascertained and placed beyond a doubt. How, then, are they to be accounted for? The theory prevalent some years since, that they are caused by the combustion of immense coal-beds, is puerile and now entirely abandoned. All the coal in the world could not afford fuel enough for one of the tremendous eruptions of Vesuvius. This observed increase of temperature in descending beneath the earth’s surface suggested the notion of a central incandescent nucleus still remaining in a state of fluidity from its elevated temperature. Hence the theory that the whole mass of the earth was formerly a molten fluid mass, the exterior portion of which, to some unknown depth, has assumed its present solidity by the radiation of heat into surrounding space, and its consequent refrigeration. The mathematical solution of this problem of Central Heat, assuming such heat to exist, tells us that though the central portion of the earth may consist of a mass of molten matter, the temperature of its surface is not thereby increased by more than the small fraction of a degree. Poisson has calculated that it would require _a thousand millions of centuries_ to reduce this fraction to a degree by half its present amount, supposing always the external conditions to remain unaltered. In such cases, the superficial temperature of the earth may, in fact, be considered to have approximated so near to its ultimate limit that it can be subject to no further sensible change. DISAPPEARANCE OF VOLCANIC ISLANDS. Many of the Volcanic Islands thrown up above the sea-level soon disappear, because the lavas and conglomerates of which they are formed spread over flatter surfaces, through the weight of the incumbent fluid; and the constant levelling process goes on below the sea by the action of tides and currents. Such islands as have effectually resisted this action are found to possess a solid framework of lava, supporting or defending the loose fragmentary materials.