The answer to the question of what may be the constitution of these remarkable features may now be given with a moderate approach to certainty. It has been shown successively that the rings could not be solid, or liquid, and in 1857 Clerk-Maxwell demonstrated that the only possible constitution for such a body is that of an infinite number of small satellites. The rings of Saturn thus presumably consist of myriads of tiny moonlets, each pursuing its own individual orbit in its individual period, and all drawn to their present form of aggregation by the attraction of Saturn's bulging equator. The appearances presented by the rings are explicable on this theory, and on no other. Thus the brightness of the two rings A and B would arise from the closer grouping of the satellites within these zones; while the semi-darkness of the Crape Ring arises from the sparser sprinkling of the moonlets, which allows the dark sky to be seen between them. Cassini's division corresponds to a zone which has been deprived of satellites; and as it has been shown that this vacant zone occupies a position where a revolving body would be subject to disturbance from four of Saturn's satellites, the force which cleared this gap in the ring is obvious. It has been urged as an objection to the satellite theory that while the thin spreading of the moonlets would account for the comparative darkness of the Crape Ring when seen against the sky, it by no means accounts for the fact that this ring is seen as a dark stripe upon the body of the planet. Seeliger's explanation of this is both satisfactory and obvious, when once suggested--namely, that the darkness of the Crape Ring against the planet is due to the fact that what we see is not the actual transits of the satellites themselves, but the perpetual flitting of their shadows across the ball. The final and conclusive argument in favour of this theory of the constitution of the rings was supplied by the late Professor Keeler by means of the spectroscope. It is evident that if the rings were solid, the speed of rotation should increase from their inner to their outer margin--_i.e._, the outer margin must move faster, in miles per second, than the inner does. If, on the contrary, the rings are composed of a great number of satellites, the relation will be exactly reversed, and, owing to the superior attractive force exercised upon them by the planet through their greater nearness to him, the inner satellites will revolve faster than the outer ones. Now, this point is capable of settlement by spectroscopic methods involving the application of the well-known Doppler's principle, that the speed of a body's motion produces definite and regular effects upon the pitch of the light emitted or reflected by it. The measurements were of extreme delicacy, but the result was to give a rate of motion of 12-1/2 miles per second for the inner edge of ring B, and of 10 miles for the outer