speed be irregular or uniform, in comparing the speed of one star with the speed of another star. I think that the best criterion which I can give is by a piece of mechanism which has been contrived, and applied to this purpose. (See Figure 2.) The best Equatoreals are furnished with a racked wheel attached to the axis, in which works an endless screw or worm, as at E, Figure 1. By turning it, the whole instrument is made to revolve. This worm, or screw, is turned by an apparatus which is constructed expressly for uniform movement. Various contrivances have been used for making this motion as uniform as possible. The one usually adopted, with some modifications (as represented in Figure 2), depends on the use of centrifugal balls AB, similar to those which are used to regulate the motions of steam engines. Everybody knows well that whirling these balls round by the rotation of the axis CD, to which they are attached, causes them to spread out. When the speed has reached a certain limit, the spreading out of these brings the moving parts, as at E and F, into contact with the fixed parts GH, and produces a degree of friction which prevents further acceleration; and thus a uniform speed is produced, with very great nicety. This contrivance is in constant use on my Equatoreal at the present time. You will observe it is essential to have a machine moving uniformly. In the motion of a common clock, though the movement from day to day, from hour to hour, and from minute to minute, is uniform, yet it is not so with the smaller divisions of seconds: the clock works with jerks, and does not go uniformly. Now the machine here is going on without any jerks with a smoothness and uniformity scarcely to be obtained by any other apparatus. In all the best Observatories
