turned to what it had been before it was heated, when the glass was cold.

These different facts, therefore, prove the relation of the phenomena discovered by Faraday with the molecular condition of the bodies submitted to experiment. But there is a wide difference between this and admitting that this condition is the cause of it. In fact, with the exception of a few traces of temper, and these somewhat dubious, observed by M. Mathiessen, nothing hitherto has shown directly that the electric and magnetic forces may influence, by their exterior action, the molecular constitution of solid non-magnetic bodies. Further liquids, which all manifest, though in different degrees, it is true, the rotatory property under magnetic action do not owe it to a modification impressed by this action upon their molecular constitution; for we may agitate them, may make them be traversed by electric currents in all directions, without the property being in the slightest degree altered. Moreover, direct observations, made with great care, do not seem to indicate that the exterior influence of magnetism or electricity exercises any effect upon the physical constitution of liquids, upon their volume, their fluidity, &c.

If it is not to an alteration in the arrangement of their particles that bodies owe the rotatory power that they acquire under the influence of magnetism, we must endeavour to seek the origin in some other modification which they experience beneath this influence. In fact, this phenomenon does not arise from a direct action exercised by magnetism upon light; the body is a necessary intermediate; for a polarized ray travelling in vacuo, or even in a gas, does not undergo any action on the part of a powerful electro-magnet. This has been proved by Mr. Faraday and several other philosophers: the presence of material molecules, or of molecules tolerably near together, such as those constituting a solid or a liquid, seems therefore to be a necessary condition. On the other hand, the action not being exercised upon the particles, so as in any way to modify their relative position, it is necessary to admit that it is upon the particles in their relation with the ether by which they are enveloped that it occurs; but, in

order that the action which the particles exercise naturally upon the ether may be modified by the magnetic force to which they are subjected, it is necessary that this ether be in that particular state that follows from the close approximation of the particles. Now this particular state consists in that it is more dense and more elastic in solid and liquid media than in the gases; which, as we know, is the cause of the high refracting power of the first two classes of bodies.

Thus the magnetic force would not act upon the ether by the intervention of the particles, except when it is in a certain state of density, arising from the action exercised naturally upon it by the particles, and would act the more powerfully as this density should be more considerable. As it does not depend wholly upon that of the body, that is, upon the nearness of the particles of which it is constituted, but rather upon the nature of these particles, it is not always the densest bodies that are the most refracting, and which consequently ought to experience the greatest amount of circular magnetic polarisation. Experiment entirely confirms this mode of viewing the subject; and if we cast our eyes over the table, which as yet is very limited, it is true, and very imperfect in the co-efficients of magnetic polarisation, we are struck with the fact that the substances follow each other in this table, almost in the same order as in the table of their refracting powers. Further researches are necessary, in order to establish on still more solid bases the analogy that I have been pointing out; and especially to determine the nature of the modification experienced by the action of the particles upon the ether under magnetic influence; a modification, the essence of which is to break the uniformity of its mode of action around the particle; to substitute for it another, occurring only according to a certain direction, and moreover in a contrary way at the two opposite extremities of this direction; a mode of action that the word polarity characterises very well.

However, not attributing the production of the rotatory power for magnetism to a molecular derangement produced by this agent, we do not deny that the arrangement of the

particles of a body influences its optical properties. Thus every arrangement that disturbs its uniformity of structure, such as is determined by nature in crystals, and which is produced artificially in glass, by compression, for example, gives rise to phenomena of double refraction and of polarisation, which can only be explained by admitting that this alteration of molecular constitution causes, as its consequence, the ether not to have the same elasticity in all directions equally.

But if the molecular arrangement, be it natural or artificial, developes optical properties in certain bodies, I do not at present know that it has determined in them circular polarisation, which appears to me essentially due, as M. Biot also remarked, to the intimate nature of the molecules rather than to their state of aggregation. Now, what is it that characterises the intimate nature of the particle if it is not, with its weight, its action upon the ether by which it is enveloped? It is this action, therefore, that the presence of the magnetic force would modify, by giving to it, as we have said, a particular direction and a polarity. I know that we might suppose that the particles in their natural state exercise this kind of action upon the ether, that is, that they are naturally polar. In this case certain substances would have of themselves, and others would acquire under the influence of magnetism, the property of manifesting this molecular polarity, which would be in all ordinary cases neutralised by the mutual action of the particles upon each other; a neutralisation that would constitute a state of equilibrium. But it would be necessary in this explanation to suppose a change of position of the particles in respect to each other; a change that does not appear very probable; and which, at all events, it is not easy to admit, until it shall have been demonstrated directly.

To sum up in the ideas at present received on the constitution of matter, we think that the phenomena discovered by Faraday ought to be attributed to an action of magnets and electric currents, exercised neither on the particles alone, nor on the ether alone, but on the manner of the existence of the particles in respect to the ether.

Action of Magnetism upon Polarised calorific Rays.

Radiating heat is polarised the same as light, as several philosophers have succeeded in showing. It is electricity that furnishes, in the thermo-electric pile, as we shall see further on, the fittest instruments for proving this property.

M. Wartmann, a short time after Faraday's discovery, announced that a piece of rock salt, placed in the route of polarised rays of heat, determines the rotation of the plane of polarisation, if a powerful electro-magnet is made to act upon it. Rock salt, in the experiment, plays for radiant heat the same part that glass plays for light; namely, it transmits it without its experiencing a sensible diminution in intensity; and like glass, when under the action of magnetism, rock salt, in like manner, makes the plane of polarisation of the rays that are traversing it deviate. MM. de la Prevostaye and Desains have confirmed the result that M. Wartmann had at first obtained, by employing slightly different means, and, in particular, by making use of solar heat. They even succeeded in measuring the deviation of the plane of polarisation, a determination that is very difficult, and which requires the employment of very delicate processes and apparatus. We shall return to this subject when we shall have made known these processes and apparatus.