points, and is not at all affected by their relative velocity. The inversion of the poles being effected instantaneously, we should thus have a velocity infinitely accelerated. Now a system moving round an axis, and capable of a continuous circular motion, is that which is alone susceptible of such a velocity. It cannot become uniform, unless some resisting element, or some other action depending on the velocity, is introduced into the system. Putting aside the application to practical use which has to be made of such an apparatus, the obstacles to be overcome, inseparable from the system, consist only in the friction of the pivots in the sockets, and in the resistance of the air. As to the former, repeated experiments have proved that the friction is independent of the rapidity of rotation, at least within the limits of experiment; this resistance, therefore, can in no way contribute to render the accelerated motion uniform. It is in fact the resistance of the air which will act to produce this effect. Although it might be reduced at pleasure, principally by giving a suitable form to the rotatory system, it would not be entirely annihilated. But it will be allowed, that we should have reason to be well satisfied with the mechanical effect of magnetism, if this were the only cause which tended to reduce the accelerated movement to a uniform movement. The limits of such a uniform velocity must be very distant. I do not speak, of the great simplicity of a magnetic machine with a continuous circular motion, of the advantages of construction which are gained by being able to transform with ease this motion to any other which the working machine may require. These considerations had strongly impressed my mind, even whilst the means of execution were still unknown to me, but I always kept in sight the practical application, and the object appeared too important for me to exhaust my powers in the construction of see-saw toys, which might claim the honour of being placed in the rank with the electric chime relatively to their effect, and still more relatively to the tinkling with which they are accompanied.
4.
Fig. 1. of the annexed plate represents the magnetic apparatus of eight bars, arranged symmetrically upon a disc moveable round the axis , and of eight fixed bars similarly arranged upon a fixed platform. The arrangement of the bars admits of the greatest variety, provided it be exactly symmetrical, and that it allow the poles to approach each other as nearly as possible. To prevent the action being too oblique,—since the centre of magnetic gravity is probably at some distance from the extremity, as in the ordinary magnetic bars,—it is preferable to make this arrangement so that the axes of the cylindrical bars shall be situated rectangularly, and not parallel, as in the figure. It must be further observed, that there will be some difficulty in forging bars of considerable dimensions into the horseshoe form, so that the axes of the branches be situated exactly
