horse-shoe magnet is substituted for the electro-magnet. But in all these cases it is essential that the glass prism be so placed that the forces emanating from the poles, and which are characterised in their direction by the magnetic curves, should pass through the glass in the direction parallel, or nearly so, to that of the ray. Thus no effect is observed when the glass prism is situated equatorially, that is to say, perpendicularly, to the line that joins the poles of the electro-magnet.

The direction of the rotation of the plane of polarisation, as we have said, is connected with the position of the magnetic poles in respect to the direction along which the polarised ray travels; and this direction is such, that if the north pole of the electro-magnet is turned on the side of the observer who is receiving the ray, and, consequently, the south pole on the side whence the ray comes, the rotation occurs for the observer from left to right. If the place of the poles is changed by changing the direction of the current that is circulating around the electro-magnet, the rotation occurs from right to left. We may reduce to a very simple law the relation that exists between the direction of the rotation of the plane of polarisation and that of the magnetism, by which the rotation is produced; for this, we have merely to suppose a piece of soft iron put in place of the transparent substance, and to represent the direction of the currents that in Ampère's theory circulate around this piece of iron in consequence of the magnetisation it undergoes. Now, the law is that the rotation occurs in the same direction as that according to which this current travels. A more direct mode of making this law manifest, is to put in place of the electro-magnet a bobbin, in the axis of which the glass prism is placed, so that it is enveloped from one end to the other by a girdle of electric currents. The effect of these currents is to produce, in the same manner as the magnetic poles, the rotation of the plane of polarisation; and the direction of this rotation is such, that we may say that when an electric current circulates around a transparent substance that is transmitting a ray of polarised light in a direction perpendicular to that of the current, it causes the plane of polarisation of this ray to rotate in a

direction similar to that according to which it is itself travelling.

When we employ bobbins or helices to produce the rotation of the plane of polarisation, we notably increase the effect by lengthening them, because it enables us to give greater length to the substance traversed by the ray, such, for example, as a column of water; but it is useless to give it a greater length than that of the helix in the interior of which it is placed. It is also indifferent, when the transparent body has a less diameter than that of the helix, to place it in the axis or out of the axis, provided it is in the interior of the helix; for out of the helix the effect is null. Finally, when the rotatory power is produced by means of electric currents, we must take the necessary precautions to prevent those currents, which are generally very intense, from heating the substance that is placed in the interior of the helix.

Determination of the magnetic rotatory Power of different transparent Substances.

Faraday, in his first researches, had already operated upon a large number of substances, both solid and liquid; he had not been able to discover any trace of rotatory power either in air or in other gases, although he submitted them to the action of a powerful electro-magnet, and to that of a long helix, traversed by an energetic current. With regard to liquids, he had tried a very great number, and had found them all capable of acquiring in different degrees the rotatory power under the influence of magnetism; even those which, like essence of turpentine, already possess it naturally: and this rotatory power, superinduced by magnetic action, is altogether independent of that which the substance possesses of itself; so that if the two rotations occur in the same direction they are added to each other, and if they occur in contrary directions, the more feeble is subtracted from the more powerful, in order to obtain the definitive result.

This law has been verified upon many liquids, such as essences, solutions of sugar, tartaric and tartrate acid, which

make the plane of polarisation turn from the left to the right; as well as upon others, which make it turn from the right to the left. With regard to solids, Faraday constantly found that the effect is at its maximum in boro-silicate of lead, and subsequently in glasses containing lead; but he was not able to succeed in producing it in crystals endowed with double refraction, no matter in what way they were cut. Rock salt, on the contrary, acquired, under magnetic influence, a rotatory power almost equal to that of flint, but which was scarcely the third of that of the heavy glass.

M. Edmond Becquerel, in order to increase the action of the electro-magnet upon substances submitted to experiment, conceived the happy idea of placing upon the polar surface armatures, each pierced with a horizontal cylindrical opening at the very spot where the magnetic pole is situate, that is to say, at the point whence the magnetic forces seem to emanate. The two armatures and the transparent substances are SO arranged, that the polarised ray that traverses it the direction of its greatest length may pass through the two poles, on the outside of which are placed, on one side, the polarising prism, and on the other, the analysing prism. By means of this apparatus, M. Becquerel successfully repeated Faraday's experiments. In order to measure the rotation of the plane of polarisation, he made use of the tint of passage pointed out by M. Biot, and he obtained, with heavy glass, a deviation of 16°.

He recognised that among liquids the chlorides, especially that of zinc, possess a considerable rotatory power. A stratum of chloride of zinc, only 0.39 in. in thickness, made the plane of polarisation deviate 6°. Finally, he was able to obtain a slight rotation by submitting to the magnetic action certain crystals endowed with double refraction. With this view he took two plates of quartz, of the same thickness, and of opposite rotation, so that on placing them perpendicularly in the route of the polarised ray, their effects neutralised each other. Each of the plates was 196 in. in thickness. The electro-magnet was put in action, and the effect was manifested, sometimes in one direction, sometimes in the other,

according to the direction of the magnetisation, as upon a plate of glass, but much more feebly. The magnetic rotation was 5° with a specimen of beryl 039 in. in thickness, placed perpendicularly to the axis. These are two examples, which prove that bi-refracting crystals present the phenomenon of circular magnetic polarisation, although in a much more feeble degree than other substances.

M. Mathiessen, in a series of experiments, out of a hundred crystals that he examined, found only rock salt which was sensible to magnetism; the rotation that was produced by a thickness of 102 in., which is that at which the maximum effect occurs, is little inferior to that of boro-silicate of lead; for it is 6°, the other being 9°. Furthermore, M. Mathiessen found several vitreous combinations that are more heavy than Faraday's glass; in particular, pure silicates of lead, which give an effect more than double that of the boro-silicate. Unfortunately they tarnish rapidly in the air. On subjecting to experiment two hundred and forty species of glass, the same philosopher was able to determine very exactly the influence of the chemical composition over the rotatory power of the glass; he found that the silicates and the chlorides hold the first rank in respect to sensibility; that oxide of lead is the base that acts most energetically, then bismuth, antimony, zinc, mercury, silver. The rotation is manifested in glasses with magnetic bases: perhaps they would produce more effect than all others, if their deep colours did not cause our being compelled to employ them only in the state of very thin plates. A very remarkable fact is that, when a glass contains iron, cobalt, or nickel in so small a quantity that it preserves a sufficient transparency, the rotation gradually increases with the thickness of the glass, up to 3 inches, which is the greatest separation that can be given to the poles in the apparatus that was employed; whilst glasses without a magnetic metal, and at the same time without boracic acid, without soda, and without potassa, have their maximum of effect at a much less thickness, which varies between half an inch and an inch. An electro-magnet, that can sustain 55lbs., gives, with silicate of lead, 20° of rotation, for a thickness of 59 in., and gave no more

for a thickness of 1.57 inches. With Faraday's heavy glass, on the other hand, under the same magnetic influence, 4° were obtained with 59 in., and 9° with 1.57 inches of thickness.

Faraday had observed, that magnetic action does not instantaneously develope in a substance the maximum of rotatory power, but that this power increases gradually for a few seconds, whilst it ceases immediately with magnetisation. M. Mathiessen remarked that, for certain slightly annealed glasses, the augmentation of rotatory power with the duration of the magnetic action is very sensible; but that this augmentation especially occurs, if we change the poles of the electromagnet between which the glass is placed. This happened to a silicate of lead, which at first gave 18°, and which afterwards gave 20°, after three or four changes of the poles: a greater number of subsequent changes made it fall back to 18°, and afterwards to 15°. After a certain time of rest, the same series of operations may be recommenced, and with the same results. It would seem to follow from this that the sudden interruption and re-establishment of the magnetism diminish the temper, and consequently increase the rotation of the plane of polarisation; but that if they are repeated too great a number of times consecutively, they restore a new temper, which diminishes the rotatory power.

Moreover, experiment shows that it is indeed in the very interior of the body and not at its surfaces, whose effect would rather consist in diminishing the result, that the phenomenon of circular magnetic polarisation occurs. Six plates of the same glass superposed, the assemblage of which constitutes a thickness equal to that of another single piece of the same glass, gave a lesser rotation almost in the relation of 11 to 13: these six plates, cemented together with Canada balsam, regain almost the force of the single piece.

All these results clearly indicate a relation between the rotatory powers and a peculiar molecular arrangement, that is determined by the influence of magnetism in transparent bodies: but, before entering upon the study of the causes of the rotation, we have yet to expose the more complete ex