It is supposed that Jupiter and Saturn, as well as Uranus and Neptune, retain a considerable amount of internal heat, but certainly not sufficient to keep the metallic and other elements of which the sun and earth consist in a state of vapour, for if so they would be planetary stars and would shine by their own light. And if any considerable portion of their bulk consisted of these elements, whether in a solid or a liquid state, their densities would necessarily be much greater than that of the earth instead of very much less--Jupiter is under one-fourth the density of the earth, Saturn under an eighth, while Uranus and Neptune are of intermediate densities, though much less in bulk even than Saturn. It thus appears that the solar system consists of two groups of planets which differ widely from each other. The outer group of four very large planets are almost wholly gaseous, and probably consist of the permanent gases--those which can only be liquefied or solidified at a very low temperature. In no other way can their small density combined with enormous bulk be accounted for. The inner group also of four planets are totally unlike the preceding. They are all of small size, the earth being the largest. They are all of a density roughly proportionate to their bulk. The earth is both the largest and the densest of the group; not only is it situated at that distance from the sun which, through solar heat alone, allows water to remain in the liquid state over almost the whole of its surface, but it possesses numerous characteristics which secure a very equable temperature, and which have secured to it very nearly the same temperature during those enormous geological periods in which terrestrial life has existed. We have already shown that no other planet possesses these characteristics now, and it is almost equally certain that they never have possessed them in the past, and never will possess them in the future. A LAST ARGUMENT FOR HABITABILITY OF THE PLANETS Although it has been admitted by the late Mr. Proctor and some other astronomers that most of the planets are not _now_ habitable, yet, it is often urged, they may have been so in the past or may become so in the future. Some are now too hot, others are now too cold; some have now no water, others have too much; but all go through their appointed series of stages, and during some of these stages life may be or may have been possible. This argument, although vague, will appeal to some readers, and it may, therefore, be necessary to reply to it. This is the more necessary as it is still made use of by astronomers. In a criticism of my article in _The Fortnightly Review_, M. Camille Flammarion, of the Paris Observatory, dramatically remarks: 'Yes, life is universal, and eternal, for time is one of its factors. Yesterday the moon, to-day the earth, to-morrow Jupiter. In space there are both cradles and tombs.'[19]