_Density._--So, too, the small mean density of the planet, only 1.3 times that of water and actually less than the density of the sun, suggests that the larger part of the planet's bulk may be made of gases and clouds, with very little solid matter even at the center; but here we get into a difficulty from which there seems but one escape. The force of gravity at the visible surface of Jupiter may be found from its mass and dimensions to be 2.6 times as great as at the surface of the earth, and the pressure exerted upon its atmosphere by this force ought to compress the lower strata into something more dense than we find in the planet. Some idea of this compression may be obtained from Fig. 88, where the line marked _E_ shows approximately how the density of the air increases as we move from its upper strata down toward the surface of the earth through a distance of 16 miles, the density at any level being proportional to the distance of the curved line from the straight one near it. The line marked _J_ in the same figure shows how the density would increase if the force of gravity were as great here as it is in Jupiter, and indicates a much greater rate of increase. Starting from the upper surface of the cloud in Jupiter's atmosphere, if we descend, not 16 miles, but 1,600 or 16,000, what must the density of the atmosphere become and how is this to be reconciled with what we know to be the very small mean density of the planet? We are here in a dilemma between density on the one hand and the effects of gravity on the other, and the only escape from it lies in the assumption that the interior of Jupiter is tremendously hot, and that this heat expands the substance of the planet in spite of the pressure to which it is subject, making a large planet with a low density, possibly gaseous at the very center, but in its outer part surrounded by a shell of clouds condensed from the gases by radiating their heat into the cold of outer space. [Illustration: FIG. 88.--Increase of density in the atmospheres of Jupiter and the earth.] This is essentially the same physical condition that we found for the sun, and we may add, as further points of resemblance between it and Jupiter, that there seems to be a circulation of matter from the hot interior of the planet to its cooler surface that is more pronounced in the southern hemisphere than in the northern, and that has its periods of maximum and minimum activity, which, curiously enough, seem to coincide with periods of maximum and minimum sun-spot development. Of this, however, we can not be entirely sure, since it is only in recent years that it has been studied with sufficient care, and further observations are required to show whether the agreement is something more than an accidental and short-lived coincidence.