But if we turn back to the Table, we see that Jupiter at its mean distance from the Sun is 5·2 times that of the Earth; that is to say, it receives only 1/27th of the light and heat that we receive. But in Chapter VIII, we learnt from Mars that as this receives only 3/7ths of the Earth's light and heat, its mean temperature would sink to -30°C.; the Earth's being 16°C. Mars is therefore almost always a frozen planet; frozen except on its mere surface when this is exposed to the full rays of the Sun. No sea there would ever be melted to a depth of more than a few inches, even at noonday in midsummer. And yet Mars has at least ten times the advantages of Jupiter. Jupiter, then, must be a frozen planet through and through; no liquid of any sort can exist on its surface; no vapour of any substance can exist in its atmosphere. It must be icebound even at its summer noonday. Yet, from the description given by Prof. Keeler, it is manifestly not so; and another item in the Table emphasizes that it cannot be so. The density of the Sun is 1·4 that of water, Jupiter's is 1·33, showing that but a very small proportion (if any) of its bulk can be solid; the rest must be vaporous, or at least fluid. How then can we reconcile these inconsistencies? It is in the dimensions of Jupiter that we find the answer. The mass of the planet is 317 times that of the Earth; it is indeed nearly three times as great as that of all the other planets put together. But the aggregation of so vast an amount of material is of itself a source of heat; the chief source at the present time of the enormous output of heat from the Sun is ascribed to its gradual contraction; the slow falling of its substance, if we may so express it, a little nearer to its centre. The great mass of Jupiter points to its inherent store of heat being much greater than that of any other planet. And of two bodies equally hot, the larger must cool more slowly than the smaller. If, therefore, all the members of the solar system had at one and the same moment possessed the same surface temperature, that equality would have ceased directly they began to radiate their heat into space; the temperature of the smaller bodies falling more rapidly than those of the larger. This is another example of the principle that has already been noted, that the properties of a small world are not those of a large world divided by a constant factor. It is not possible to conceive a model of the solar system in which all the significant factors should be true to the same scale. If the diameters and distances were all made on a one-tenth scale, the surfaces would be one-hundredth of reality, the volumes one-thousandth. But a radiating body radiates from its surface, while the store of heat from which that radiation is kept up is supplied by its volume. It follows, therefore, that a large and heavy world must differ from a small light world, not merely in scale, but also in kind.