The next great difficulty in the rigid theory is, that the rate of rotation in different parts of the breadth of the ring does not conform to the conditions of stability. The inner edge of the ring should rotate much more rapidly than the outer—as the combined breadth of the two bright rings is 28,000 miles. They should rotate as satellites would do at corresponding distances. The nearer a satellite is to the body of a planet, the more rapidly must it revolve, in order that it may not fall upon the surface. If you swing a stone at the end of a string round your head, you must, in order to keep it up, increase the number of revolutions as you shorten the string; and, for a similar reason, the nearer a satellite is, the greater must be the number of its revolutions in a given time. It has been surmised, that there is a small satellite or large meteoric stone revolving round the earth at about the distance of 5000 miles from its surface; if this be so, it must, in order that it may not fall to the earth, swing round seven times a day. The moon is about twenty-six times further off, and, to maintain its position, it has only to swing round once in twenty-eight days. Let us suppose that the moon and the little satellite are yoked together by being made parts of a rigid ring whose breadth extends from the one to the other. It is plain that the tendency will be to rend the system in pieces. The one is a slow horse, obstinately keeping back the vehicle 3 the other is a fiery steed, that will break
