18 The Saturnian System. (January, 
circumstance. Of what materials should such a globe be 
composed? Are there any materials whatever, of such 
small density, which can be supposed to exist in such rela- 
tively enormous quantities as to constitute nearly the whole 
of the mass of Jupiter or Saturn? It would be absurd to 
regard as a reasonable hypothesis, now, either the theory of 
Whewell, that Jupiter consists mainly of water, or the 
alternative suggestion of Brewster, that the substance of 
the giant planets may be of the nature of pumice-stone. 
Such theories as these are the really startling theories 
of science; they are really fanciful, because based on no 
known physical fa¢éts. They were unscientific even when 
they were propounded; but they are doubly unscientific 
now that spectroscopic analysis has rendered it probable 
that the elements constituting the great orbs of the universe 
are in the main the same that we are familiar with on earth, 
and proportioned similarly as to relative quantity. If our 
sun were to solidify, there would be in his globe our familiar 
iron, gold, copper, calcium, sodium, and other metallic 
elements, combined with those other elements which we 
recognise as the chief constituents of the earth’s globe. 
We cannot refuse to admit the probability that what is true 
of the small earth and the gigantic sun, as well as of his 
giant brothers, the stars, can scarcely fail to be the case 
with the intermediate order of bodies to which Saturn and 
Jupiter belong. But we cannot admit this inference, pro- 
bable theugh it is, if we adhere to our assumption that the 
conditions of temperature in Saturn and Jupiter resemble 
those in our earth. We should certainly find these planets, 
in that case, as dense as our earth, or rather (considering 
the much larger pressures to which the greater part of their 
mass would be subjected) we should find their mean density 
very much greater. Since, on the contrary, they are of 
very small mean density compared with the earth, we are 
driven from the assumption that their physical condition 
resembles that of the earth. 
It is natural, under these circumstances, to inquire 
whether, since the earth gives us no satisfactory explana- 
tion of the condition of Jupiter and Saturn, we may not 
find in the sun some suggestions towards an explanation. 
It would have been natural, indeed, to have turned first to 
the sun, because of the much greater similarity of condition 
already mentioned as existing between the giant planets and 
the sun. But I preferred to consider first the less obvious 
and probable line of argument, partly to dispose of it, and 
partly because, unlikely and even unreasonable as it is, it is 
