It may likewise be said, "you have calculated the duration of the refrigeration of the planets, not only by the inverted ratio of their diameters, but also by the inverted ratio of their density; this might be well founded if we could imagine that in fact there exists matter whose density is as different from that of our globe: but does it exist? What, for example, will be the matter of which Saturn is composed, since its density is more than five times less than that of the earth? To this I answer, that it would be very easy to find, in the vegetable class, matters five or six times less dense than a mass of iron, marble, hard calcareous stone, &c. of which we know that the earth is principally composed; but without quiting the mineral kingdom, and considering the density of these five matters, we have 21-10/72 for iron, 8-25/72 for white marble, for gres 7-24/72, for common marble and calcareous stone 7-20/72; taking the mean term of the densities of these five matters, of which the terrestrial globe is principally composed, we find its density to be 10-5/18. It is therefore required to find a matter whose density is in the relation of 189 to 1000 density, which is the same as that between Saturn and the Earth. Now this matter might be a kind of pumice stone, somewhat less dense than common pumice stone, whose relative density is here 1-66/70; whence it appears that Saturn is principally composed of a light matter similar to pumice stone. So likewise the density of the Earth being to that of Jupiter as 1000 to 292, we must suppose that Jupiter is composed of a more dense matter than pumice stone, but much less dense than chalk. The density of the Earth being to that of the Moon as 1000 to 702, this secondary planet appears composed of a matter whose density is not quite so great as that of hard calcareous stone, but more so than soft. The density of the Earth being to that of Mars as 1000 to 730, this planet must be composed of a matter somewhat more dense than that of gres, and not so great as that of white marble. But the density of the Earth being to that of Venus as 1000 to 12700, it may be supposed that this planet is chiefly composed of a more dense matter than emery, and less dense than zinc. Finally, the density of the Earth being to that of Mercury :: 1000 : 2040, or :: 10-5/18 : 20,(966-2/3)/1000, it must be thought that this planet is composed of a matter less dense than iron but more so than tin.