The commutator consists of a small boxwood cylinder, carrying around its cylindrical surface two rows of eight holes, one above the other, in which are fitted sixteen contact pieces of brass which slightly project above the surface of the wood, the positions of those in the upper circle alternating or "breaking joint" with those in the lower, and each contact piece is in metallic connection with its corresponding conducting wire, and, therefore, with the junction of two of the helices on the armature. Against the edge of the commutator are pressed by means of adjustable levers two small brass contact rollers, k k, which are respectively connected with the positive and negative poles of the voltaic battery (either through or independent of the coils of a fixed electro-magnet, to which we shall presently refer), and the magnetic axis of the ring will lie in the same plane as the line joining the points of contact of the battery and rotating helix, this axis remaining nearly fixed notwithstanding the rotation of the iron ring in which the magnetism is induced.

In the apparatus figured in Figs. 3 and 4, the armature rotates between the two vertical limbs, A B, of a fixed electro-magnet furnished with extended pole pieces, A A, B B (Fig. 4), each of which embraces about six of the armature coils. The fixed electro-magnet is constructed of two vertical iron cylindrical bars, A and B, united at their lower extremities by a horizontal iron bar, F F, the one being rigidly and permanently attached to it, while the other is fastened to it by a screw, G, passing through a slot so that the distance of the pole pieces from one another and from the armature ring is capable of adjustment.

The connections of the machine, which are shown in Fig. 3, are made as follows: The positive current, entering by the attachment screw, h, passes by a wire to the right hand commutator screw, l, to the right-hand roller, k, through the commutator to the ring, around which it traverses to the left-hand roller, k¹, and screw, l¹, to the magnet coil, A, and thence through the coil of the magnet, B, to the terminal screw, h, on the right hand of the figure. This method of coupling up is of very great historical interest, for it is the first instance on record of the magnet coils and armature of a machine being included in one circuit, giving to it the principle of construction of a dynamo-electric machine, and antedating in publication, by two years, the interesting machines of Siemens, Wheatstone, and Varley, and preceding them in construction by a still longer period.

With this apparatus Dr. Pacinotti made the following interesting experiments with the object of determining the amount of mechanical work produced by the machine (when worked as an electro-magnetic engine), and the corresponding consumption of the elements of the battery: Attached to the spindle of the machine was a small pulley, Q Q (Fig. 3), for the purpose of driving, by means of a cord, another pulley on a horizontal spindle carrying a drum on which was wound a cord carrying a weight, and on the same spindle was also a brake and brake-wheel, the lever of which was loaded so as just to prevent the weight setting into motion the whole system, consisting of the two machines, when no current was flowing. In this condition, when the machine was set in motion by connecting the battery, the mechanical work expended in overcoming the friction of the brake was equal to that required to raise the weight; and, in order to obtain the total work done, all that was necessary was to multiply the weight lifted by the distance through which it was raised. The consumption of the battery was estimated at the same time by interposing in the circuit a sulphate of copper voltameter, of which the copper plate was weighed before and after the experiment. The following are some of the results obtained by Dr. Pacinotti in experimenting after the manner just described. With the current from a battery of four small Bunsen elements, the machine raised a weight of 3.2812 kilos to a height of 8.66 m. (allowing for friction), so that the mechanical work was represented by 28.45 m. During the experiment the positive plate of the voltameter lost in weight 0.224 gramme, the negative gaining 0.235 gramme, giving an average of chemical work performed in the voltameter of 0.229 gramme, and multiplying this figure by the ratio between the equivalent of zinc to that of copper, and by the number of the elements of the battery, the weight of zinc consumed in the battery was computed at 0.951 gramme, so that to produce one kilogrammeter of mechanical work 33 milligrammes of zinc would be consumed in the battery. In another experiment, made with five elements, the consumption of zinc was found to be 36 milligrammes for every kilogrammeter of mechanical work performed. In recording these experiments, Dr. Pacinotti points out that although these results do not show any special advantage in his machine over those of other construction, still they are very encouraging, when it is considered that the apparatus with which the experiments were made were full of defects of workmanship, the commutator, being eccentric to the axis, causing the contacts between it and the rollers to be very imperfect and unequal.

In his communication to the Nuovo Cimento, Dr. Pacinotti states that the reasons which induced him to construct the apparatus on the principle which we have just described, were: (1) That according to this system the electric current is continuously traversing the coils of the armature, and the machine is kept in motion not by a series of intermittent impulses succeeding one another with greater or less rapidity, but by a constantly acting force producing a more uniform effect. (2) The annular form of the revolving armature contributes (together with the preceding method of continuous magnetization) to give regularity to its motion and at the same time reduces the loss of motive power, through mechanical shocks and friction, to a minimum. (3) In the annular system no attempt is made suddenly to magnetize and demagnetize the iron core of the rotating armature, as such changes of magnetization would be retarded by the setting up of extra currents, and also by the permanent residual magnetism which cannot be entirely eliminated from the iron; and with this annular construction such charges are not required, all that is necessary being that each portion of the iron of the ring should pass, in its rotation, through the various degrees of magnetization in succession, being subjected thereby to the influence of the electro-dynamic forces by which its motion is produced. (4) The polar extension pieces of the fixed electro-magnet, by embracing a sufficiently large number of the iron projecting pieces on the armature ring, continue to exercise an influence upon them almost up to the point at which their magnetization ceases when passing the neutral axis. (5) By the method of construction adopted, sparks, while being increased in number, are diminished in intensity, there being no powerful extra currents produced at the breaking of the circuit, and Dr. Pacinotti points out that when the machine is in rotation a continuous current is induced in the circuit which is opposed to that of the battery; and this leads to what, looked at by the light of the present state of electric science, is by far the most interesting part of Dr. Pacinotti's paper, published, as it was, more than seventeen years ago.