With the year 1831 begins the period of the celebrated “Experimental Researches in Electricity and Magnetism.” During the years which had elapsed since his discovery of the electromagnetic rotations in 1823, Faraday, though occupied, as we have seen, with other matters, had not ceased to ponder the relation between the magnet and the electric current. The great discoveries of Oersted, Ampère, and Arago had culminated in England in two results: in Faraday’s discovery that the wire which carries an electric current tends to revolve around the pole of a neighbouring magnet; and in Sturgeon’s invention of the soft-iron electromagnet, a core of iron surrounded by a coil of copper wire, capable of acting as a magnet at will when the electric current is transmitted to the coil and so caused to circulate around the iron core.

This production of magnetism from electricity, at will, and at a distance, by the simple device of sending the electricity to circulate as a current around the central core of iron was then, as now, a cause of much speculation. The iron core which is to be made temporarily into a magnet stands alone, isolated. Though surrounded outwardly by the magnetising coil of copper wire, it does not touch it; nay, must be screened from contact with it by appropriate insulation. The electric current entering the copper coil at one end is confined from leaving the copper wire by any lateral path: it must circulate around each and every convolution, nor be permitted to flow back by the return-wire until it has performed the required amount of circulation. That the mere external circulation of electric current around a totally disconnected interior core of iron should magnetise that core; that the magnetisation should be maintained so long as the circulation of electricity is maintained; and that the magnetising forces should cease so soon as the current is stopped, are facts, familiar enough to every beginner in the science, but mysterious enough from the abstract point of view. Faraday was firmly persuaded that, great as had been these discoveries of the production of magnetism and magnetic motions from electricity, there remained other relations of no less importance to be discovered. Again and again his mind recurred to the subject. If it were possible to use electricity to produce magnetism, why should not the converse be true? In 1822 his notebook suggestion was, as we have seen, “Convert magnetism into electricity.” Yes, but how?

He possessed an intuitive bent of mind to inquire about the relations of facts to one another. Convinced by sheer converse with nature in the laboratory, of the correlation of forces and of the conservation of energy long before either of those doctrines had received distinct enunciation as principles of natural philosophy, he seems never to have viewed an action without thinking of the necessary and appropriate reaction; never to have deemed any physical relation complete in which discovery had not been made of the converse relations for which instinctively he sought. So in December, 1824, we find him experimenting on the passage of a bar magnet through a helix of copper wire (see Quarterly Journal for July, 1825), but without result. In November, 1825, he sought for evidence that might prove an electric current in a wire to exercise an influence upon a neighbouring wire connected to a galvanometer. But again, and yet again in December of the same year, the entry stands “No result.” A third failure did not convince him that the search was hopeless: it showed him that he had not yet found the right method of experimenting. It is narrated of him how at this period he used to carry in his waistcoat pocket a small model of an electromagnetic circuit—a straight iron core about an inch long, surrounded by a few spiral turns of copper wire—which model he at spare moments would take out and contemplate, using it thus objectively to concentrate his thoughts upon the problem to be solved. A copper coil, an iron core. Given that electricity was flowing through the one, it evoked magnetism in the other. What was the converse? At first sight it might seem simple enough. Put magnetism from some external source into the iron core, and then try whether on connecting the copper coil to a galvanometer there was any indication of an electric current. But this was exactly what was found not to result.

And not Faraday alone, but others, too, were foiled in the hope of observing the expected converse. Not all who tried were as wise or as frank as Faraday in confessing failure. Fresnel, in the height of the fever of Oersted’s discovery, had announced to the Academy of Sciences at Paris, on the 6th of November, 1820, that he had decomposed water by means of a magnet which was laid motionless within a spiral of wire. Emboldened by this announcement, Ampère remarked that he too had noticed something in the way of production of currents from a magnet. But before the end of the year both these statements were withdrawn by their authors. Again, in the year 1822, Ampère, being at Geneva, showed to Professor A. de la Rive in his laboratory a number of electromagnetic experiments from his classical researches; and amongst them one20 which has been almost forgotten, but which, had it been followed up, would assuredly have led Ampère to the discovery of the induction of currents. In the experiment in question a thin copper ring, made of a narrow strip folded into a circle, was hung inside a circular coil of wire, traversed by a current. To this apparatus a powerful horse-shoe magnet was presented; and De la Rive states that, when the magnet was brought up, the suspended ring was observed sometimes to move between the two limbs of the magnet, and sometimes to be repelled from between them according to the sense of the current in the surrounding coil. He and Ampère both attributed the effect to temporary magnetism conferred upon the copper ring. Ampère himself was at the time disposed to attribute it to the possible presence of a little iron as an impurity in the copper. There are, however, some discrepancies in the three published versions of the story. According to Becquerel, Ampère had by 1825 satisfied himself of the non-existence of induction currents.

Quite independently, the question of the possibility of creating currents by magnets was raised by another discovery, that of the so-called “magnetism of rotation.” In 1824 Arago had observed that a fine magnetic compass constructed for him by Gambey, having the needle suspended in a cell, the base of which was a plate of pure copper, was thereby damped in its oscillations, and instead of making two or three hundred vibrations before it came to rest, as would be the case in the open air, executed only three or four of rapidly decreasing amplitude.21 In vain did Dumas at the request of Arago analyse the copper, in the supposition that iron might be present. Inquiry compelled the conclusion that some other explanation must be sought. And, reasoning from the apparent action of stationary copper in bringing a moving magnetic needle to rest, he conjectured that a moving mass of copper might produce motion in a stationary magnetic needle. Accordingly he set into revolution, beneath a compass needle, a flat disc of copper, and found that, even when a sheet of card or glass was interposed to cut off all air-currents, the needle tended to follow the moving copper disc, turning as if dragged by some invisible influence. To the suggestion that mere rotation conferred upon copper a sort of temporary magnetism Arago listened with some impatience. All theories proposed to account for the phenomenon he discredited, even though emanating from the great mathematician Poisson. He held his judgment in absolute suspense. Babbage and Herschel measured the amount of retarding force exerted on the needle by different materials, and found the most effective to be silver and copper (which are the two best conductors of electricity), after them gold and zinc, whilst lead, mercury, and bismuth were inferior in power. The next year the same experimenters announced the successful inversion of Arago’s experiment; for by spinning the magnet underneath a pivoted copper disc they caused the latter to rotate briskly. They also made the notable observation that if slits are cut radially in the copper disc they diminish its tendency to be dragged by the spinning magnet. Sturgeon showed that the damping effect of a moving copper disc was diminished by the presence of a second magnet pole of contrary kind placed beside the first. All these things were most suggestive of the real explanation. It clearly had something to do with the electric conductivity of the metal disc, and therefore with electric currents. Sturgeon five years later came very near to the explanation: after repeating the experiments he concluded that the effect was an electric disturbance in the copper disc, “a kind of reaction to that which takes place in electromagnetism.”

Faraday knew of all the discussions which had arisen respecting Arago’s rotations. They may have been the cause of his unsuccessful attempts of 1824 and 1825. In April, 1828, for the fourth time he tried to discover the currents which he was convinced must be producible by the magnet, and for the fourth time without result. The cause of failure was that both magnet and coil were at rest.

The summer of 1831 witnessed him for the fifth time making the attack on the problem thus persistently before him. In his laboratory note-book he heads the research “Experiments on the production of electricity from magnetism.” The following excellent summary of the laboratory notes is taken from Bence Jones’s “Life and Letters”:—

In the seventeenth paragraph, written on the 30th of August, he says, “May not these transient effects be connected with causes of difference between power of metals at rest and in motion in Arago’s experiments?” After this he prepared fresh apparatus.

As was his manner, he wrote off to one of his friends a letter telling what he was at work upon. On this occasion the recipient of his confidences was his friend Phillips:—

September 24 was the third day of his experiments. He began (paragraph 21) by trying to find the effect of one helix of wire, carrying the voltaic current of ten pairs of plates, upon another wire connected with a galvanometer. “No induction sensible.” Longer and different metallic helices (paragraph 22) showed no effect; so he gave up those experiments for that day, and tried the effects of bar magnets instead of the ring magnet he had used on the first day.

In paragraph 33 he says:—

The fourth day of work was October 1. Paragraphs 36, 37, and 38 describe the discovery of induced voltaic currents:—

The fifth day of experiment was October 17. Paragraph 57 describes the discovery of the production of electricity by the approximation of a magnet to a wire:—

The cause of all the earlier failures was, then, that both magnet and coil were at rest. The magnet might lie in or near the coil for a century and cause no effect. But while moving towards the coil, or from it, or by spinning near it, electric currents were at once induced.

The ninth day of his experiments was October 28, and this day he “made a copper disc turn round between the poles of the great horse-shoe magnet of the Royal Society. The axis and edge of the disc were connected with a galvanometer. The needle moved as the disc turned.” The next day that he made experiments, November 4, he found “that a copper wire one-eighth of an inch drawn between the poles and conductors produced the effect.” In his paper, when describing the experiment, he speaks of the metal “cutting” the magnetic curves, and in a note to his paper he says, “By magnetic curves I mean lines of magnetic forces which would be depicted by iron filings.”

We here come upon those “lines of force” which played so important a part in these and many of Faraday’s later investigations. They were known before Faraday’s time—had, in fact, been known for two hundred years. Descartes had seen in them evidence for his hypothetical vortices. Musschenbroek had mapped them. But it was reserved to Faraday to point out their true significance. To the very end of his life he continued to speculate and experiment upon them.

All this splendid work had occupied but a brief ten days. Then he rearranged the facts which he had thus harvested, and wrote them out in corrected form as the first series of his “Experimental Researches in Electricity.” The memoir was read to the Royal Society on November 24, 1831, though it did not appear in printed form until January, 1832—a delay which gave rise to serious misunderstandings. The paper having been read, he went away to Brighton to take a holiday, and in the exuberance of his heart penned the following letter23 to Phillips:—

The second section shows that Faraday had discovered the cause of all the previous failures to evoke electric currents in wires by means of a magnet: it required relative motion. What the magnet at rest fails to do, the magnet in motion accomplishes. This crucial point is admirably commemorated in the following impromptu given by Mr. Herbert Mayo to Sir Charles Wheatstone:—

Faraday’s holiday was brief; by December 5 he was again at work on his researches. He re-observed the directions of the induced currents about which, as the slip in his letter to Phillips shows, his mind was in some doubt. Then on December 14th comes the entry:—“Tried the effects of terrestrial magnetism in evolving electricity. Obtained beautiful results.”

“The helix had the soft iron cylinder (freed from magnetism by a full red heat and cooling slowly) put into it, and it was then connected with the galvanometer by wires eight foot long; then inverted the bar and helix, and immediately the needle moved; inverted it again, the needle moved back; and, by repeating the motion with the oscillations of the needle, made the latter vibrate 180°, or more.”

The same day he “made Arago’s experiment with the earth magnet, only no magnet used, but the plate put horizontal and rotated. The effect at the needle was slight but very distinct.... Hence Arago’s plate a new electrical machine.”

When we compare these manuscript notes, recording the experiments in the order in which they were made with the published account of them in the “Experimental Researches,” we find many of them transcribed almost verbatim. But there is a difference in the order of their arrangement. In point of time the experiments on the evolution of electricity from magnetism, beginning with the ring (p. 108), preceded those on the induction of a current by another current. In the printed “Researches” the experiments on the induction of currents are put first, with an introductory paragraph on the general phenomenon of induction.28 Faraday’s habit of working up an experiment—whether successful or unsuccessful—by increasing the power to the maximum available is illustrated in the course of the experiments on the iron ring. At first he used a battery of ten pairs of plates four inches square. Then, having been eminently successful in producing deflexions of his galvanometer, he increased the battery to one hundred pairs of plates, with the result that when contact was completed or broken in the primary circuit the impulse on the galvanometer in the secondary circuit was so great as to make the needle spin round rapidly four or five times before its motion was reduced to a mere oscillation. Then he removed the galvanometer and fixed small pencils of charcoal to the ends of the secondary helix; and to his great joy perceived a minute spark between the lightly touching charcoal points whenever the contact of the battery to the primary helix was completed. This was the first transformer, for the first time set—on a small scale—to produce a tiny electric light. The spark he regarded as a precious indication that what he was producing really was an electric current. Using the great compound steel magnet of the Royal Society (constructed by Dr. Gowin Knight) at Christie’s house at Woolwich he had, as narrated above, also obtained a spark from the induced current. For some time he failed to obtain either physiological or chemical effects. But upon repeating the experiments more at leisure at the Royal Institution, with Daniell’s armed loadstone capable of lifting thirty pounds, a frog was found to be convulsed very strongly each time magnetic contact between the magnet and the iron core of the experimental coil was made or broken.

The absence of evidence as to chemical action seemed still to disquiet him. He wanted to be sure that his induced currents would do everything that ordinary voltaic currents would do. Failing the final proof from chemical action, he rested the case on the other identical properties. “But an agent,” he says, “which is conducted along metallic wires in the manner described; which, whilst so passing, possesses the peculiar magnetic actions and force of a current of electricity; which can agitate and convulse the limbs of a frog; and which, finally, can produce a spark by its discharge through charcoal, can only be electricity. As all the effects can be produced by ferruginous electro-magnets, there is no doubt that arrangements like the magnets of Professors Moll, Henry, Ten Eyke, and others, in which as many as two thousand pounds have been lifted, may be used for these experiments; in which case not only a brighter spark may be obtained, but wires also ignited, and as the currents can pass liquids, chemical action be produced. These effects are still more likely to be obtained when the magneto-electric arrangements, to be explained in the fourth section, are excited by the powers of such apparatus.” The apparatus described in the fourth section comprised several forms of magneto-electric machines, that is to say, primitive kinds of dynamos. Having in his mind the phenomenon discovered by Arago, and the experiments of Babbage and Herschel on the so-called magnetism of rotation, he followed up the idea that these effects might be due to induced currents eddying round in the copper disc. No sooner had he obtained electricity from magnets than he attempted to make Arago’s experiment a new source of electricity, and, as he himself says, “did not despair” “of being able to construct a new electrical machine.”

The “new electrical machine” was an exceedingly simple contrivance. A disc of copper, twelve inches in diameter (Fig. 6), and about one-fifth of an inch in thickness, fixed upon a brass axle, was mounted in frames, so as to allow of revolution, its edge being at the same time introduced between the magnetic poles of a large compound permanent magnet, the poles being about half an inch apart.29 The magnet first used was the historical magnet of Gowin Knight. The edge of the plate was well amalgamated, for the purpose of obtaining a good but movable contact, and a part round the axle was also prepared in a similar manner. Conducting strips of copper and lead, to serve as electric collectors, were prepared, so as to be placed in contact with the edge of the copper disc; one of these was held by hand to touch the edge of the disc between the magnet poles. The wires from a galvanometer were connected, the one to the collecting-strip, the other to the brass axle; then on revolving the disc a deflexion of the galvanometer was obtained, which was reversed in direction when the direction of the rotation was reversed. “Here, therefore, was demonstrated the production of a permanent current of electricity by ordinary magnets.” These effects were also obtained from the poles of electro-magnets, and from copper helices without iron cores. Several other forms of magneto-electric machines were tried by Faraday.

In one,30 a flat ring of twelve inches’ external diameter, and one inch broad, was cut from a thick copper plate, and mounted to revolve between the poles of the magnet, two conductors being applied to make rubbing contact at the inner and outer edge at the part which passed between the magnetic poles. In another,31 a disc of copper, one-fifth of an inch thick and only 1½ inch in diameter (Fig. 7), was amalgamated at the edge, and mounted on a copper axle. A square piece of sheet metal had a circular hole cut in it, into which the disc fitted loosely; a little mercury completed communication between the disc and its surrounding ring. The latter was connected by wire to a galvanometer; the other wire being connected from the instrument to the end of the axle. Upon rotating the disc in a horizontal plane, currents were obtained, though the earth was the only magnet employed.

Faraday also proposed a multiple machine32 having several discs, metallically connected alternately at the edges and centres by means of mercury, which were then to be revolved alternately in opposite directions, In another apparatus,33 a copper cylinder (Fig. 8), closed at one extremity, was put over a magnet, one half of which it enclosed like a cap, and to which it was attached without making metallic contact. The arrangement was then floated upright in a narrow jar of mercury, so that the lower edge of the copper cap touched the fluid. On rotating the magnet and its attached cap, a current was sent through wires from the mercury to the top of the copper cap. In another apparatus,34 still preserved at the Royal Institution, a cylindrical bar magnet, half immersed in mercury, was made to rotate, and generated a current, its own metal serving as a conductor. In another form,35 the cylindrical magnet was rotated horizontally about its own axis, and was found to generate currents which flowed from the middle to the ends, or vice versâ, according to the rotation. The description of these new electrical machines is concluded with the following pregnant words:—

In yet another machine (Fig. 9), constructed by Faraday some time later,36 a simple rectangle of copper wire w, attached to a frame, was rotated about a horizontal axis placed east and west, and generated alternate currents, which could be collected by a simple commutator c.

Within a few months machines on the principle of magneto-induction had been devised by Dal Negro, and by Pixii. In the latter’s apparatus a steel horseshoe magnet, with its poles upwards, was caused to rotate about a vertical shaft, inducing alternating currents in a pair of bobbins fixed above it, and provided with a horseshoe core of soft iron. Later, in 1832, Pixii produced, at the suggestion of Ampère,37 a second machine, provided with mercury cup connections to rectify the alternations of the current. One of these machines was shown at the British Association meeting at Oxford in the same year (p. 64).

The idea developed in the third part of this research was intensely original and suggestive. Faraday’s own statement is as follows:—

Faraday further supposed that the formation of this state in the neighbourhood of a coil would exert a reaction upon the original current, giving rise to a retardation of it; but he was unable at the time to ascertain experimentally whether this was so. He even looked—though also unsuccessfully—for a self-induced return current from a conductor of copper through which a strong current was led and then suddenly interrupted, the expected current of reaction being “due to the discharge of its supposed electrotonic state.”

If we would understand the rather obscure language in which this idea of an electrotonic state is couched, we must try to put ourselves back to the epoch when it was written. At that date the only ideas which had been formulated to explain magnetic and electric attractions and repulsions were founded upon the notion of action at a distance. Michell had propounded the view that the electric and magnetic forces vary, like gravity, according to a law of the inverse squares of the distances. Coulomb, in a series of experiments requiring extraordinary patience as well as delicacy of manipulation, had shown—by an application of Michell’s torsion balance—that in particular cases where the electric charges are concentrated on small spheres, or where the magnetic poles are small, so as to act as mere points, this law—which is essentially a geometric law of point-action—is approximately fulfilled. The mathematicians, Laplace and Poisson at their head, had seized on this demonstration and had elaborated their mathematical theories. Before them, though the research lay for a century unpublished, Cavendish had shown that the only law of force as between one element of an electric charge and another compatible with a charge being in equilibrium was the law of inverse squares. But in all these mathematical reasonings one thing had been quite left out of sight—namely, the possible properties of the intervening medium. Faraday, to whom the idea of mere action at a distance was abhorrent, if not unthinkable, conceived of all these forces of attraction and repulsion as effects taking place by something going on in the intervening medium, as effects propagated from point to point continuously through space. In his earlier work on the electromagnetic rotations he had grown to regard the space around the conducting wire as being affected by the so-called current; and the space about the poles of a magnet he knew to be traversed by curved magnetic lines, invisible indeed, but real, needing only the simplest of expedients—the sprinkling of iron filings—to reveal their existence and trend. When therefore he found that these new effects of the induction of one electric current by another could likewise cross an intervening space, whether empty or filled with material bodies, he instinctively sought to ascribe this propagation of the effect to a property or state of the medium. And finding that state to be different from any state previously known, different from the state existing between two magnets at rest or between two stationary electric charges, he followed the entirely philosophical course of exploring its properties and of denoting it by a name which he deemed appropriate. As we shall see, this idea of an electrotonic state recurred in his later researches with new and important connotations.