placed across (Fig. 80.). These pieces of iron, both in this case and the preceding one, become magnets; they react

Fig. 80.

upon the bars, and maintain in them the separation of the magnetisms which, but for this precaution, would finish gradually by combining in a great degree, and so recomposing the neutral fluid. In compass needles, terrestrial magnetism performs the office of an armature by maintaining by its power the separation of the two magnetisms.

The armatures of horse-shoe magnets generally carry a scale-pan in which weights are placed, care being taken never to exceed the limit of what the armature can carry without being detached. These weights, which may be gradually increased, retain the power of the magnet, and even tend to increase it, providing that the access of weight never becomes such that the armature is detached. In this case, the magnet suddenly loses a great portion of its power, and it cannot recover it except by a fresh, slow, and gradual increase of the weight that its keeper can carry.

An interesting question, which was first studied by Coulomb, and afterwards by Nobili, is to know what the magnetic condition is of each of the plates, or of each of the wires, whose union forms what is termed a magnetic bundle. This determination seems to be able to lead to the knowledge of the magnetic state of the interior strata of a magnetised needle or bar. Coulomb found, by means of the balance, that the plates which formed the two exterior surfaces had a much greater magnetic force than the others. He also found some in the interior that had their poles inverted. M. Nobili, by employing the method of oscillations, succeeded in determining very exactly the magnetic state of each of the very fine needles, of the same length, which, to the number of fifty, formed a bundle that had been powerfully mag

netised. He found, after he had undone the parcel, that all the needles were magnetised in the same direction: he then made up the bundle again, maintaining the contact of the needles as perfect as possible.

Having untied the packet two hours afterwards, he found that several of the needles had acquired a contrary magnetism. Having made the same experiment with another packet, and having undone it half an hour after magnetising it, he found that a certain number of the needles had lost all their magnetism. It follows from these facts that the strongest needles demagnetise the more feeble, and even impart to them a contrary magnetism; and that if, in the outset, they had received the same degree of magnetisation, the magnetic force would be very speedily extinguished in the whole system.

These experiments led M. Nobili to conclude that we must not consider, as we have done hitherto, a magnetic bar as formed of a bundle of filaments of the same length, magnetised to the same degree, and all in the same direction; for then the whole system would be very speedily demagnetised. He supposes that the interior of the bar is divided into concentric layers, the magnetism of which diminishes rapidly from without inwards; and that the conservative condition of magnetisation does not depend upon the coercitive force, such as we understand it, but on the mode itself of the distribution of the magnetism in magnets.

In this way of viewing it, tempering acts by determining in the mass such a state, that the exterior molecules, being cooled more rapidly than those of the interior, approach more closely than the latter are able to do. It follows from this, that tempered steel possesses a crust, the density and the other properties of which differ from those of the internal strata. In particular, the magnetism is better preserved in it; and on that account it is that magnetised steel remains more strongly magnetic. It is for the same reason that soft iron, when it has been beaten under the hammer, or has been passed through the draw-plate, acquires the property of preserving a small amount of magnetism. The exterior parts

having been rendered more compact than those of the interior, there arises an unequal distribution of the magnetism, which is found in greater quantity on the exterior than on the interior. This also causes small bars to take proportionally more magnetism than large ones, their surface, in proportion to their volume, being more considerable. The following new experiment by Mr. Nobili is entirely in support of his opinion. This philosopher constructed, with the same steel, two cylinders of the same length and the same diameters; the one solid, which weighed 4324 grs. Troy, the other hollow, which weighed only 247 grs. Troy. These two cylinders were tempered in the same manner, and each magnetised to saturation. Placed at the same distance from a compass needle, the solid one gave a deviation of 91°, and the hollow one a deviation of 19°. The difference is very great in favour of the hollow cylinder, although its mass was almost one half less than the solid one. This arises from the hollow cylinder being tempered without and within, and then being covered on both sides with this crust that preserves the magnetism: whilst the solid one possesses it only on its exterior surface.

All the facts relating to magnetisation are still enveloped, as we perceive, in very great obscurity. One very evident principle, however, is derived from them all, and which we have already pointed out; it is the connection which they establish between magnetism and the molecular properties of bodies. As we have said, we shall see, in the sequel, when we are considering magnetisation by electric currents, some fresh proofs in favour of this principle; we may possibly then be able to determine, in a more precise manner, what the nature is of the relation in question.

We shall say nothing here upon the experiments of certain philosophers, and especially those of Mr. Barlow, on the magnetisation of bodies of various forms, such as rings and spheres; and of the action that they exercise upon the magnetised needle, when they are of soft iron and not magnetised previously, or merely subjected to the magnetising action of the terrestrial globe. These effects obey laws which are remarkable for

their simplicity and their regularity. So also have they been made easily the subjects of calculation, and they have been employed by M. Poisson in his theoretical researches upon magnetism. We shall confine ourselves to pointing out a single fact, which, from its connection with those that have gone before, appears to us very important. It is, that the magnetic power exercised by soft iron spheres resides entirely upon their surface, and is completely independent of their mass; so that the effect exercised by cannon-balls upon the magnetised needle is exactly the same, whether they are solid or hollow. However, this law has limits. Mr. Barlow has recognised that the metal envelope must have at least a thickness of in., in order to act as if the sphere were solid.

20

The processes of magnetisation have lately been still further improved in a remarkable manner, so that permanent magnets can be obtained of extraordinary power; but the processes that are employed are not known. We shall have occasion, further on, when occupied with the magnetism produced by dynamic electricity, to speak of the principle upon which probably they depend.

CHAP. II.

MUTUAL ACTION OF MAGNETISM AND DYNAMIC ELECTRICITY; AND OF ELECTRIC CURRENTS UPON EACH OTHER.

Mutual Action of Magnetism and Electric Currents.

FOR a long time philosophers were struck with the analogy that seemed to exist between electric and magnetic phenomena. Two magnetisms, as there are two electricities; attraction and repulsion exercised between the contrary magnetisms as between the electricities, according to similar laws; these are indeed points of resemblance between the two classes of phenomena. However, it was in vain that attempts were made to establish experimentally a more intimate relation between them. In 1805, MM. Hachette and Desormes had endeavoured, without success, by means of terrestrial magnetism, to give direction to an insulated voltaic pile, having consequently its two poles equally strong, and freely suspended: their attempts were fruitless.

It was not until 1820 that a Danish philosopher, M. Oersted, succeeded in discovering the relation, that had so long been sought after, between magnetism and electricity ; but it was not where it had been constantly thought to exist that he discovered it. Electricity acts upon a magnet; and a magnet in its turn acts upon electricity; but only when the electricity is in motion, that is to say, in the condition that we have termed dynamic: there is no action when the electricity is in the static or tension state.

The following is Oersted's fundamental experiment: —

The two poles of a pile are united by a metal wire, called a conjunctive wire. This wire is placed either above or below a magnetised needle, freely suspended, and parallel to its direction. The needle is immediately seen to suffer a