turn formed, by breaking and concentrating, the satellites or moons
which attend the earth, as they do all the planets which lie farther
away from the sun than our sphere.

[Illustration: Fig. 1.--Saturn, Jan. 26, 1889 (Antoniadi).]

As if to prove that the planets and moons of the solar system were
formed somewhat in the manner in which we have described it, one of
these spheres, Saturn, retains a ring, or rather a band which appears
to be divided obscurely into several rings which lie between its group
of satellites and the main sphere. How this ring has been preserved
when all the others have disappeared, and what is the exact
constitution of the mass, is not yet well ascertained. It seems clear,
however, that it can not be composed of solid matter. It is either in
the form of dust or of small spheres, which are free to move on each
other; otherwise, as computation shows, the strains due to the
attraction which Saturn itself and its moons exercise upon it would
serve to break it in pieces. Although this ring theory of the
formation of the planets and satellites is not completely proved, the
occurrence of such a structure as that which girdles Saturn affords
presumptive evidence that it is true. Taken in connection with what we
know of the nebulæ, the proof of Laplace's nebular hypothesis may
fairly be regarded as complete.

It should be said that some of the fixed stars are not isolated suns
like our own, but are composed of two great spheres revolving about
one another; hence they are termed double stars. The motions of these
bodies are very peculiar, and their conditions show us that it is not
well to suppose that the solar system in which we dwell is the only
type of order which prevails in the celestial families; there may,