14. Ampere’s theory has, in a high degree, the usual fault of substituting one mystery for another; but, on the other hand, it has, in an equally high extent, the only merit to which any theory can make an indisputable claim: I mean that of associating facts so as to make them more easy to comprehend and to remember, enabling us, by analogy, to foresee results, and thus affording a clue in our investigations. Evidently, the author of this theory was guided by it in his highly interesting and instructive contrivances; and Professor Henry ascribes his success in improving the electro-magnet to the theoretic clue which he had received from Ampere.
15. Nevertheless, the postulates on which this Amperean hypothesis is founded appear to me unreasonable. They require us to concede that about every atom of a permanent magnet a process is going on analogous to that generally admitted to exist in a galvanic circuit, where two fluids pass each other in a common channel by a series of decompositions and recompositions, ([7].) In the galvanic circuit this process is sustained by chemical reaction; but without any coenduring cause, how is it to be sustained permanently in a magnet? Is it reasonable to assume that the heterogeneous constituents of an imaginary tertium quid are perpetually separating only to reunite? ([8].)[62]
16. In cases of complex affinity, where four particles, A B C D are united into two compounds A B, C D, it is easy to conceive that, in obedience to a stronger affinity, A shall combine with C, and B with D: but, without any extraneous agency, wherefore, in any one compound, should a particle A quit a particle B, in order to unite with another particle of the same kind; or wherefore should any one B quit one A, in order to combine with another A?
17. That such a process should take place in consequence of the inductive agency of a similar process already established in a magnet or galvanized wire were difficult to believe; but it would seem utterly incredible that the most transient influence of such induction should be productive of such permanent electrolytic gyration as has been above specified. Moreover, it is inconceivable that the particles of any matter should, as required by this hypothesis, merely by being put into motion, acquire a power of reciprocal repulsion or attraction of which it were otherwise destitute.
18. The vortices being assumed to take place about each atom, cannot severally occupy an area of greater diameter than can exist between the centres of any two atoms. Of course, the gyratory force exercised about the surface of a magnet by the aggregate movements of the vortices cannot extend beyond the surface more than half the diameter of one of the minute areas of gyration alluded to. Wherefore, then, do these gyrations, when similar in direction, from their concurrence approach each other; when dissimilar in direction, from contrariety move away, even when situated comparatively at a great distance?
19. I should consider Ampere’s theory as more reasonable were it founded upon the existence of one fluid; since, in that case, vortices might be imagined without the necessity of supposing an endless and unaccountable separation and reunion of two sets of particles; not only devoid of any property capable of sustaining their alleged opposite gyrations, but actually endowed with an intense reciprocal attraction which must render such gyrations impossible. But even if grounded on the idea of one fluid, this celebrated hypothesis does not seem to me to account for the phenomena which it was intended to explain. If distinct portions of any fluid do not attract or repel each other when at rest, wherefore should they either attract or repel each other when in motion? Evidently mere motion can generate neither attraction nor repulsion. Bodies projected horizontally gravitate with the same intensity, and consequently, in any given time, fall to the earth through the same perpendicular distance, whether moving with the celerity of a cannon ball, or undergoing no impulse excepting those arising from their own unresisted weights.
20. The objections which are thus shown to be applicable in the case of liquids, of which the neighbouring particles are destitute of the reaction requisite to produce the phenomena requiring explanation, must operate with still greater force where ethereal fluids are in question, of which the properties are positively irreconcilable with the phenomena. According both to Franklin and Dufay, bodies, when similarly electrified, should repel each other; yet in point of fact, collateral wires, when subjected to similar voltaic discharges, and of course similarly electrified, become reciprocally attractive, while such wires, when dissimilarly electrified by currents which are not analogous, become reciprocally repulsive.
21. Agreeably to Ampere, an iron bar, situated within a coil of wire subjected to a galvanic current, is magnetized, because the current in the wire is productive of an electrical whirlpool about every particle of the metal. When the iron is soft, the magnetism, and of course the gyrations of which its magnetism consists by the premises, cease for the most part as soon as the circuit through the coil is broken; but when the iron is in the more rigid state of hardened steel, the gyrations continue for any length of time after the exciting cause has ceased.
22. This theory does not explain wherefore the hardening of the steel should cause the gyration to be more difficult to induce, yet more lasting when its induction is effected. Evidently the metallic particles must take some part in the process; since it is dependent for its existence and endurance upon their nature and their state. Yet no function is assigned to these particles. In fact, it is inconceivable either that they can participate in, or contribute to, the supposed gyration.
23. The electrical fluid in an iron bar cannot form a vortex about each particle, all the vortices turning in one direction, without a conflict between those which are contiguous. In order not to conflict with each other, the alternate vortices would have to turn in different directions, like interlocking cog-wheels in machinery. But in that case, if magnetism be due to currents, the magneto-inductive influence of one set would neutralize that of the other. Again, how can a current, excited by a battery in one circuitous conductor, cause, by dynamic induction, a current in the opposite direction, through another conductor parallel to the first, but insulated therefrom? How can a current of quantity in a ribbon coil[63] give rise to one of intensity in a coil of fine wire, rushing of course with a velocity commensurate with the intensity thus imparted?