disturbing impulses, proceeding from the poles of the disturbing apparatus, neutralize each other intermediately. Supposing the ponderable as well as the imponderable matter in a perfect conductor, to be susceptible of the polar derangement, of which an electrified state is thus represented to consist, non-conductors to be insusceptible of such polar derangement, imperfect conductors may have a constitution intermediate between metals and electrics. When an electrical discharge is made through any space devoid of air or other matter, it must then find its way solely by the polarization of rare imponderable matter existing therein; and consequently its corruscations should be proportionably more diffuse, which is actually found to be true; but when gaseous ethereo-ponderable atoms intervene, they enable competent waves to exist within a narrower channel, and to attain a greater intensity. I consider all bodies as insulators which cause discharges through them to be more difficult than through a vacuum, and which, consequently, cannot be made to convey the waves from one body to another when the transfer would not take place in their absence. This furnishes a good mean of discrimination between insulators and conductors, the criterion being that a discharge ensues more readily as there is more of the one and less of the other in the way. Even when a bell wire has been dissipated by lightning, it has been found to facilitate and determine the path of the discharge. Both in the case of disruptive discharge through air, producing a spark, or of a deflagrating discharge through wire, causing its explosion, there is a dispersion of intervening ponderable particles; and yet there is this manifest discordancy, that in the one case the undulatory process of transfer is assisted, in the other resisted. The waves follow the metallic filament with an intense attraction, while they strive to get out of the way of those formed by the aëriform matter, as if repelled. Hence the term diruptive, from dirumpo, to break through, was happily employed by Faraday to designate spark discharges. The zigzag form of the diruptive spark, shows that there is a tendency in the aëriform particles to turn the waves out of that straight course, which, if unresisted or facili tated, they would naturally pursue. On the one hand the aërial filaments being unsuitable for the conveyanee of the electric waves, they are forced by them out of the normal path, first in one direction, then in another; while, on the other hand, the finest metallic filament furnishes a channel for the electric waves, so favorable that this channel is pursued, although the conse quent polarization of the conducting particles be so intense as to make them fly asunder with explosive violence.

70. When a small image, of which the scalp has been

abundantly furnished with long hair, is electrified, the hairy filaments extend themselves and move apart, as if actuated by a repulsive power: also when iron filings are so managed as to obey the influence of the poles of a powerful magnet, they arrange themselves in a manner resembling that of the electrified hair. There is, moreover, this additional analogy, that there is an attraction between two portions of hair differently electrified, like that which arises between filings differently magnetized.Yet the properties of the electrified hair and magnetized filings are, in some respects, utterly dissimilar. A conducting communication between differently electrified portions of hair would entirely neutralize the respective electrical states; so that all the electrical phenomena displayed by them would cease. Yet such a communication made between the poles, exciting the filings, by any non-magnetic conductor, does not in the slightest degree lessen their polar affections and consequent power of reciprocal influence. Upon the electrified hair, the proximity or the contact of a steel magnet has no more effect than would result under like circumstances from any other metallic mass similarly employed; but by the approximation, and still more, the contact, of such a magnet, the affection of the filings may be enhanced, lessened, or nullified, according to the mode of its employment. In the case of the hair, the affection is superficial, and the requisite charging power must be in proportion to the extent of surface. In the case of the magnetized ferruginous particles, it is the mass which is affected, and, cæteris paribus, the more metal, the greater the capacity for magnetic power. In the instance of the electrified hair, as in every other of statical excitement, the electrical power resides in imponderable ethereo-electric atmospheres which adhere superficially to the masses, being liable to be unequally distributed upon them in opposite states of polarity, consequent to a superficial polarization of the exciting or excited ponderable masses; but in the instance of bodies permanently magnetic, or those rendered transiently magnetic by galvanic influence, the ethereo-electric matter and the ponderable atoms are inferred to be in a state of combination, forming ethereo-ponderable atoms; so that both may become parties to the movements and affections of which the positive and negative waves consist. Since the electro-polarization, when carried to a certain degree of intensity, produces ignition, at other degrees deflagration, decomposition, or dispersion of conductors, it may be conceived, that in discharges too feeble to produce these consequences, there is still proportionably an affection of the ponderable matter.

71. I infer that the difference between the phenomena of statical electricity and those of galvanism does not arise from a

difference in quantity and intensity, but from the greater or less extent in which the polarity of the ponderable atoms of the masses affected are deranged. All magneto-polar charges, I infer to be attended by an affection of ponderable particles; and the reason why the most intense statical charge does not affect a galvanometer, is that it is only when oppositely excited bodies are neutralized by the interposition of a conductor, as during a discharge, that ethereo-ponderable particles are sufficiently polarized to enable them to act upon others in their vicinity, so as to produce a polar affection the opposite of their own. In this way dynamic induction is consistently explained, by supposing that the waves of polarization, in passing along one conductor, produce, pari passu, the opposite polarization in the proximate part of any neighboring conductor suitably constituted, situated and arranged to allow it to form a part of a circuit. A magnetized bar, so contrived as to rotate easily upon its axis, turns to the right or to the left, when made the channel of two opposite galvanic discharges, accordingly as the connections with the battery may be made. This may be explained by supposing that the opposite undulations (consistently with Oersted's idea) moving spirally, one to the right the other to the left, react with the polarized ethereo-ponderable particles of the magnetized bar, so as to produce revolutionary impulses.

72. It is only during the state of the incessant generation and destruction of what has been called the two electricities, that the circuit, which is the channel for the passage of the polarizing waves, is endowed with electro-magnetic powers. It was no doubt, in obedience to a perception of this fact, that Oersted ascribed the magnetism of a galvanised wire to a conflict of the electricities. Undoubtedly that state of a conductor in which, by being a part of an electrical circuit, it becomes enabled to display electro-magnetic powers, is so far a conflict of the two elec tricities, as the affections of matter which are denominated electrical consist of two opposite polar forces, proceeding, agreeably to the language of Faraday, in opposite directions from each side of the source, and conflicting with each other so as to be productive of reciprocal annihilation.

73. That a corpuscular change in conductors is concomitant with their subjection to, or emancipation from, a galvanic current, is proved by an experiment of Henry's, which he afforded me an opportunity, on one occasion, of witnessing. I allude to the fact that sound is produced whenever the circuit is suddenly made or suddenly ruptured. By I. P. Marrian it has been observed, that a similar result takes place during the magnetization or demagnetization of iron rods, by the alternate establishment or arresta tion of galvanic discharges through wires coiled about them so as to convert each into an electro-magnet. Mr. Marrian represents

1

the sound as resembling that produced by striking a rod upon one of its ends.* †

74. Thus it appears that there is an analogy between the state of matter which involves permanent magnetism and that which constitutes a galvanic current, so far as this; that either by one or the other, during either its access or cessation, an affection of the ponderable particles concerned ensues, sufficient to produce sound.

75. Although waves originally purely ethereal, as when generated by a discharge from a conductor or Leyden jar, are prone to pass superficially; yet as the surface of the conducting wire becomes less, the reaction between the ethereal and ethereo-ponderable matter becomes greater; so that finally deflagration may be produced. Hence the inferiority of statical discharges in producing electro-magnetic power is explained by the fact, that this power being in the compound ratio of the quantity of matter affected and the intensity of the affection, neither condition can be augmented without a diminution of the other. In order to produce the intensity requisite to the polarization of the ponderable particles, we must lessen the number of these subjected to the discharge, and, of course, the numerical force which can be brought into the electro-magnetic battlefield.

Summary.

From the facts and reasoning which have been above stated, it is presumed that the following deductions may be considered as highly probable, if not altogether susceptible of demonstration. The theories of Franklin, Dufay and Ampere, are irreconcilable with facts on all sides admitted.

A charge of statical electricity, or that species of electric excitement which is produced by friction, is not due to any accumulation, nor any deficiency either of one or of two fluids, but to the opposite polarities induced in imponderable ethereal matter existing throughout space however otherwise void, and likewise condensed more or less within ponderable bodies, so as to enter into combination with their particles, forming atoms which may be designated as ethereo-ponderable.

Statical charges of electricity seek the surfaces of bodies to which they may be imparted, without sensibly affecting the ethereo-ponderable matter of which they consist.

When surfaces thus oppositely charged, or, in other words, having about them oppositely polarized ethereal atmospheres,

* Agreeably to recent experiments of Faraday, the particles of a glass prism may be so influenced by an electro-magnet as to affect the passage of polarized light.

† L. and E. Phil. Mag. and Jour., vol. 45, p. 383, 1844.

are made to communicate, no current takes place, nor any transfer of the polarized matter: yet any conductor touching both atmospheres, furnishes a channel through which the opposite polarities are reciprocally neutralized by being communicated wave-like to an intermediate point.

Galvano-electric discharges are likewise effected by waves of opposite polarization, without any flow of matter meriting to be called a current.

But such waves are not propagated superficially through the purely ethereal medium; they occur in masses formed both of the ethereal and ponderable matter. If the generation of statical electricity, capable of influencing the gold leaf electrometer, indicate that there are some purely ethereal waves induced by the galvano-electric reaction, such waves arise from the inductive influence of those created in the ethereo-ponderable matter.

Magnetism, when stationary, as in magnetic needles and other permanent magnets, appears to be owing to an enduring polarization of the ethereo-ponderable atoms.

The magnetism transiently exhibited by a galvanized wire, is due to oppositely polarizing impulses, severally proceeding wavelike to an intermediate part of the circuit where reciprocal neutralization ensues.

When magnetism is produced by a statical discharge operating upon a conducting wire, it must be deemed a secondary effect, arising from the polarizing influence of the ethereal waves upon the ethereo-ponderable atoms of the wire.

Such waves pass superficially in preference; but when the wire is comparatively small, the reaction between the waves and ethereo-ponderable atoms becomes sufficiently powerful to polarize them, and thus render them competent, for an extremely minute period of time, to produce all the affections of a galvanoelectric current, whether of ignition, of electrolysis, or magnetization. Thus, as the ethereo-ponderable waves produce such as are purely ethereal, so purely ethereal waves may produce such as are ethereo-ponderable.

Polarization of hair upon electrified scalps is supposed to be due to purely ethereal polarity, while that of iron filings, by a magnet or galvanized wire, is conceived to be owing to ethereoponderable polarity.