direction of the current is shown in fig. 3 by little arrows; it is reversed each time the poles meet, provided the commutator be so placed that the edges of the levers shall quit one of the divisions in order to pass to the other. This inversion acts, as is seen, instantaneously, and quite independently of the velocity of rotation. The object is too simple, and sufficiently explained by the figures, to render it necessary to enter more into its details. I may add, further, that this same system of the inversion of the poles is applicable to any number of bars, provided that the sections of the discs are equal to them in number. I have constructed, for magneto-electric experiments, a double commutator of eight discs, with seventy-two divisions. In this apparatus there are also four levers similar to the former, which rest upon the cylinders () that unite the discs in pairs. The other extremities of these levers are likewise immersed in jars of mercury, intended to receive the ends of a connecting wire, which is to be traversed by the voltaic or magneto-electric currents, sometimes in one direction sometimes in the other. The instrument is put in motion by a handle, which can be easily turned twice in a second, and effect in the same time 144 double inversions. It will be easy to change or completely interrupt the electric current 1000 or more times in a second. The nature of this current or of the magnetism will of course be better understood by decomposing it into a rapid succession of pulsations: I am persuaded, for instance, that we should succeed by this means in charging a Leyden jar, or in effecting any chemical decompositions by the thermo-electric current of a single pair of elements.
8.
The magnetic power is produced and maintained, as is well known, by the action of the voltaic apparatus. By using zinc as a positive metal, copper as a negative metal, and water acidulated with sulphuric acid as the conducting liquid, it is the transformation of the metallic zinc into sulphate of zinc which here constitutes the cost of keeping the apparatus in action. It is a matter of the greatest importance to reduce as much as possible this cost. Let us examine what is the relation between the magnetism of the connecting wire and the action of the voltaic apparatus. Since the discovery of electro-magnetism this object has engaged the attention of distinguished scientific men, but it presents so many difficulties and such a complication of circumstances, that we cannot be surprised that the theories and formulæ which they have endeavoured to deduce from experiments differ considerably.
This is not the place to enter into the criticism of these theories; but it appears to me that the theory established by M. Ohm, in a little work entitled "Die galvanische Kette, mathemutisch bearbeitet von Dr. G. S. Ohm (1827)," and developed more fully in various memoirs printed in the German Journals, presents so much simplicity, and agrees so well with all the phænomena of the voltaic pile, that I have not hesitated to
