There is, in fact, no escape from the conclusion that the interior portions of a planet like Saturn or Jupiter, nay, even of a body like our earth or the moon, must be subject to tremendous pressure, a pressure exceeding many hundred-fold the greatest which we can obtain experimentally, and that under that enormous pressure the density of the materials composing those central parts must be increased. How is it then that Saturn is of much smaller density than the earth? I can imagine no other explanation at once so natural and so complete as this, that an intense heat pervading the entire frame of the planet enables it to resist the tremendous pressure due to mere weight. The planet's mass is expanded by the heat; large portions which at ordinary temperatures would be solid are liquified or even vaporised; matters which are liquid on our earth are vaporised; and, in fine, the planet assumes (as seen from our distant station) the appearance of being very much larger than it really is. We measure not the true globe, which, for aught that is known, may be exceedingly dense, but the dimensions of cloud-layers floating high in the planet's atmosphere. In describing Jupiter, I considered the changes which have been noticed in that planet's outline, and observed that it is impossible adequately to explain the evidence, without assuming that the changes of outline are real. The outline is not that of a solid globe, however, but of cloud-layers surrounding such a globe, and probably at a great distance from its surface. [Illustration: Fig. 31.--Saturn's square-shouldered aspect.]