By the laws of conduction, the daily impressions of heat which the earth receives, follow each other into the interior of the mass, like the waves which start from the edge of a canal;[6] and like them, become more and more faint as they proceed, till they melt into the general level of the internal temperature. The heat thus transmitted is accumulated in the interior of the earth, as in a reservoir, and flows from one part to another of this reservoir. The parts of the earth near the equator are more heated by the sun than other parts, and on this account there is a perpetual internal conduction of heat from the equatorial to other parts of the sphere. And as all parts of the surface throw off heat by radiation, in the polar regions, where the surface receives little in return from the sun, a constant waste is produced. There is thus from the polar parts a perpetual dispersion of heat in the surrounding space, which is supplied by a perpetual internal flow from the equator towards each pole. Here, then, is a kind of circulation of heat; and the quantity and rapidity of this circulation, determine the quantity of heat in the solid part of the earth, and in each portion of it; and through this, the _mean_ temperature belonging to each point on its surface. If the earth _conducted_ heat more rapidly than it does, the inequalities of temperature would be more quickly balanced, and the temperature of the ground (below the reach of annual and diurnal variations) would differ less than it does. If the surface _radiated_ more rapidly than it does, the flow of heat from the polar regions would increase, and the temperature of the interior of the globe would find a lower level; the differences of temperature in different latitudes would increase, but the mean temperature of the globe would diminish. There is nothing which, so far as we can perceive, determines necessarily, either the conducting or the radiating power of the earth to its present value. The measures of such powers, in different substances, differ very widely. If the earth were a globe of pure iron, it would conduct heat, probably, twenty times as well as it does; if its surface were polished iron, it would only radiate one-sixth as much as it does. Changes in the amount of the conduction and radiation far less than these, would, probably, subvert the whole _thermal_ constitution of the earth, and make it uninhabitable by any of its present vegetable, or animal tenants.