With a view of examining whether the force of the magnetism of a discharge would be modified by a resistance which, in the middle of each wire would tend to hinder the conducting wire from dilating itself, or is opposed to small transverse movements, I filled a crystal tube, whose exterior diameter was nearly four millimetres, and the interior diameter four-tenths of a millimetre, with mercury. The ends were closed by two brass wires, fixed by a little wax, and in contact with the mercury. With regard to the pressure exercised during the discharge, the effect has been to separate the tube into two, precisely in the middle of its length, to break it very near the extremities, and to furrow it symmetrically from that part of the middle to the quarters and a little beyond, by oblique fissures, such as are presented by the fillets of two helices turning in contrary directions. The other two quarters of the tube up to the extremities were whole. The pressure had even been a little stronger at the quarters, for at these two points, the mercury had escaped in very fine drops, showing that it had been driven out more abundantly at the two ends and in the middle. It is necessary to avoid the small bubbles of air or water interrupting the metallic fillet this will suffice to determine the rupture at the two points of separation.

Under the relation of magnetism, the series presents a notable irregularity, Opposite the point of rupture at the middle of the tube a needle was found very feebly magnetized between two needles that were strongly magnetized. This experiment was not repeated.

When a very fine brass or platinum wire, of such a length that the magnetism produced by a strong discharge will present changes in the sign, and the wire is broken by the current, the magnetic effects, except perhaps at a very small distance from the point of rupture, remain the same as if the wire still remained extended.

After having passed a very strong discharge through a wire conductor, we no longer obtained on this wire precisely the same series of magnetic intensities for quantities of electricity that are sensibly equal. This alteration, which is very feeble, depends perhaps sometimes on a superficial oxidation, and ought to agree in general with the degree of annealing which the metal undergoes during a discharge at a high temperature.

Hitherto I have supposed all the needles of the same diameter, of a millimetre, and of the same degree of temper, the rolled temper. These two circumstances, temper and the diameter, have the greatest influence on the direction and the intensity of the magnetism which a discharge may produce and a conductor give, whilst as we have seen, the length of the needles has scarcely any influence.

As an example of the manner in which these maxima and the changes of sign in the magnetism are displaced by the degree of temper, I transcribe the following series, obtained on on a brass wire of of a millimetre in diameter, and by a single discharge, the

needles were all 15 millimetres long and 0 mil. 30 in diameter. They had all been cut from the same piece of steel.*

First Series. Needles tempered hard.

The needles tempered hard presented two changes in the sign; the untempered needles only presented at some distance from the wire a maximum of intensity. The maximum of the hard tempered needles is very much further from the wire, and at a higher value than that of the flexible needles.

In general the magnetism in the hard tempered state is a state of equilibrium between the greatest forces and resistances: a feeble exterior cause alters it with greater difficulty. In this state the needles acquire a more elevated maximum either in one direction or the contrary one, attaining it by a less discharge, and only commence

⚫ I have sought to find whether, as M. De la Rive and M. Marianini have remarked on wires which have served to establish the communication between the poles of a pile, a platinum wire which has transmitted a strong discharge would give, when its extremities are plunged into a liquid conductor, any traces of an electric current. I could not observe any: but the apparatus I made use of was not sufficiently sensible, and this experiment merits a repetition.

losing it in a sensible manner with stronger discharges, and suddenly changing the sign The mode of magnetizing due to M. Arago, thus offers the means of comparing, with exactness, the different degrees of the coercive force, which augments nearly in proportion as the maximum of magnetic intensity, the state of saturation, is itself increased, but which may still vary considerably when this maximum is very little changed.

The influence of the diameter of the needles on the magnetism which they receive cannot be disengaged from other causes, which, such as the degree of temper, make the results vary; for the interior particles of a thick needle cannot have taken in the sudden cooling the same disposition as the exterior ones. Be it as it may, I extract from a series made with needles of three different thicknesses, and of 15 mill. in length, on a brass wire of 0 mill. 125 (mill.), the following values of magnetic intensities at equal heights :

In place of the changes of sign presented by the thin needles, the mean needles only presented a minimum in contact with the wire even, and the largest needles a continued decrease of intensity in proportion as they were further removed from this wire.

The series presented by the mean needles and the thickest ones, are those which were obtained with the small needles with discharges more and more weak; with discharges more and more strong the thick needles themselves exhibited changes in the sign.

The phenomenon of the reversement of the poles produced in the needles of the compass by thunder may already be explained by known facts; since the fluid, according as it passed on one side or the other of these needles, all things beside being equal, ought to magnetise them in opposite directions. But this phenomenon still admits another explanation, founded on the preceding facts; for the fluid always passing on the same side of the needle, the direction of the magnetism still depends on the distance and the intensity of the discharge. We may even remark that compass needles are found in circumstances which facilitate the production of alternations of the contrary magnetisms. They are of very small substance, which places them almost in the state of isolated wires of a small diameter, and receiving the hard temper.

On the Action of Discharges Transmitted by Wire Conductors Wound as Helices.

The needles I have employed in the following experiments were fifteen millimetres long and quarter of a millimetre in diameter : they were tempered hard.

I rolled into a helix, on a hollow cylinder of dry wood of nine cent. long, and of about 0TM 5 in diameter, a brass wire of Omil., 180 in thickness. The step or pitch of the helix was about 3 mil. in height. With a length of wire of Omet., 80, the same needles, placed successively in the middle of the cylinder, in the direction of its axis, were magnetized by discharges, gradually increased. I here give for each needle after being magnetized, and in the order of the increasing intensity of the discharges, the duration of sixty oscillations: +25",6; + 56′′,8; 38",2; 25",5; +28′′,9; + 27′′,1; 57",6; +27",8; +23′′,0; +34′′,6;

42",0; 33",1;

1",15; +31",3. The sign indicates that the needle was maguetized in the direction in which the voltaic current was transmitted, and which I have hitherto called positively magnetized; the sign indicates a contrary direction of magnetism.

The second

This series presents to us six changes in the sign. discharge, which corresponds to the number+ 56",8, was equivalent for the quantity of electric fluid to that of an ordinary Leyden jar. I could with difficulty perceive the luminous point drawn from the battery.

We should perhaps obtain with a less length of wire a greater number of changes in the direction in the magnetism. When on the contrary the wire is lengthened, the helically wound part remaining the same, not only does it require a greater force to obtain the first reversing of the poles, but in place of the changes of sign following, we only find some variations of magnetic intensity: thus for discharges equal to those of the preceding series, and a similar wire, but of double the length (1, 6), the corresponding numbers to the 3rd, 5th, 7th, 9th, 11th, 12th, and 13th observations of that series are +25",0; —31′′,9; +31′′,0; +25′′,6; +51′′,6; +54′′,6; +46",0. Beyond that the magnetism continues to augment in the same direction.

The period which is formed by the four latter values is that which becomes by the effect produced by lengthening the wire, the period comprised between the sixth and tenth observations in the first series. We see how it differs from it. However, these two periods would become identical, if the effects due to the lengthening of the wire could be compensated for by an increase in the intensity of the discharges. This compensation, therefore, cannot take place. We must not forget that in the two cases that part of the wire which acts immediately on the needles is exactly the same. A similar conclusion has already been presented in the examination of the action of rectilinear conductors.

I shall not here report the different series of magnetic intensities obtained by employing lengths of the same wire, gradually increased,

whilst with the wire met., 80 long, the first reversing of the poles takes place with a quantity of fluid which does not much surpass the charge of a simple Leyden jar. It is only with a very strong discharge of a battery of twenty-two feet surface that we obtain the same effect with a wire of equal diameter, and of eleven metres in length, the helically wound part being the same in both cases. But this wire has given needles magnetized to saturation, making sixty oscillations in a little less than 23", as well in one direction as the other.

The pitch of a helix always remaining the same, a brass wire of 0 mil., 09 in diameter, and 6 mil., 5 long, no longer gave a reversion of the poles. Nor could I obtain it with a silver wire of of a millimetre in diameter, and 0, 25 millimetres long, from the most feeble discharges up to those which reduced it completely into smoke. The maximum of magnetic intensity was with wires in a state of saturation.

I brought into communication by one of their extremities the helix of this same silver wire and a similar helix formed of a copper wire considerably larger. A discharge passing from one helix to the other produced changes in the direction of the magnetism in the second, when even the very fine wire was completely vapourized, and had so much power that it constantly magnetized in the same direction the wires submitted to its direct action. Also in the case where the diameters of the wires composing a circuit are very different, where at least one of these diameters is of extreme tenuity and rather fine, when during the discharge one of these wires changes its state the action of different parts of the circuit cannot be equal.

In general, for the same wire on a similar helix, the first maximum augments in value when the wire is made longer, and diminishes when its diameter is increased. In the first case, it requires a greater force to obtain the first reversion of the poles or the period of variations of intensity which replace it; in the second it requires a less force. Any maximum whatever is as much more elevated as the limits of the period to which it belongs are more scattered.

I pass, however, now to the case where leaving the diameter and the length of the wire invariable, we successively change the length of the diameter and the pitch of the helix.

The length of the helices, when it is equal to seven or eight times their diameter, and two or three times greater than the length of the needles, has scarcely any appreciable influence on the length, or the intensity of the magnetism.

The experiments of M. Arago have shown, as I have already said at the commencement of this memoir, that similar needles disposed in any manner whatever in the interior of a large helix, at least at some distance from its extremities, all receive parallel to its axis the same degree of magnetization ;* and likewise

• This is also true, is whatever way the needles are magnetized.