Again: that the different bodies of the universe are composed of different elements is inferred by our author from the following among other considerations: "In our solar system Mercury is sixty or eighty times more dense than one of the satellites of Jupiter, and probably in a much greater proportion denser than the satellites of Saturn. This indicates a wide difference between the nature of their elements." This statement is again repeated in a subsequent page.[22] "The densities of the planets and their satellites prove that they are composed of very different elements. Mercury is more than sixty times, and our earth about fifty times, more dense than the inner moon of Jupiter. Saturn is only about one-ninth as dense as the earth; it would float buoyantly on water. There is a high probability that the satellites of Saturn and Uranus are far lighter than those of Jupiter. Between the two extremes of the attendants of the sun, there is probably a greater difference in density than one hundred to one; and from one extreme to the other there are regular gradations of small amount. "The difference in constitution between the earth and the moon is seen in their densities: that of the moon being about half that of the earth. The nitrogen of our globe is found only in the atmosphere, and such substances as derive it from the atmosphere. The moon has no appreciable atmosphere, and therefore, in a high probability, no nitrogen." The statements here quoted were designed to show that the physical constitution of the sun and planets is widely different from that of the earth, and that the combustion of _some_ of the elements in this indefinite variety may account for the origin of solar heat. Let us examine the facts. According to Laplace the mass of Jupiter's first satellite is 0·000017328, that of Jupiter being 1. The diameter is 2436 miles. Hence the corresponding density is a little more than _one-fifth_ of the mean density of the earth. In other words, it is somewhat greater than the density of water, and very nearly equal to that of Jupiter himself. Professor Ennis' value is therefore erroneous.[23] In regard to the densities of the Saturnian and Uranian satellites nothing is known, and conjecture is useless. In short, Saturn has the least mean density of all the planets, primary or secondary, so far as known. This may be owing to the great extent of his atmospheric envelope. The density of the moon is but three-fifths that of the earth: it is to be borne in mind, however, that the _mass_ and _pressure_ are also much less.