This double result was obtained, for example, with twelve wires of about in. in diameter, and a cylinder of grey smelted cast iron. Furthermore, the influence of the nature of the iron is itself probably due to some difference in the molecular constitution of the mass. Thus, grey rough cast iron is of all kinds of iron the one that approaches the nearest to bundles of wires with regard to the effect of induction, its physiological action being relatively greater than might be expected from the intensity of the current developed in the galvanometer. This result would seem to indicate that, in this cast iron, the part of the iron susceptible of magnetisation does not form a continuous mass, which accords with M. Karsten's chemical researches.

The experiments made by means of the magnetisation of steel, and by comparing the vividness of the sparks, have confirmed the preceding results, and have also demonstrated, in a general manner, that in an inductive current, developed by employing a bundle of wires, the same quantity of electricity moves in less time than when it is the result of induction produced by a solid cylinder. Tubes, substituted for the iron, act like forged cylinders, and completely nullify the effect of wires placed in their interior, except in the case where they are traversed by a longitudinal slit; for, in that case, the introduction of a certain number of iron wires increases the physiological effect, but in no way modifies the action upon the galvanometer, which in consequence appears to arise only from the iron envelope. In the experiments in question, the tubes were hollow cylinders of the dimensions of a gun-barrel the results are not the same when tubes of thin sheet iron are employed. The wires act through these tubes, quite closed as they are, so as to increase the galvanometric effect, and even proportionately more than the physiological effect. But if the thin tube is slit longitudinally, the result is the inverse, and the introduction of the wires causes a greater increase in the physiological effect than if they were placed in a closed tube.

:

Closed or unclosed conducting envelopes, such as an insulated wire, wound to a helix around a tube of card in

which the iron is placed, and the two ends of which are connected or insulated, such again as a tube of brass closed or slit lengthwise, produce effects that are perfectly in accordance with those that we obtained when employing iron wires or solid cylinders. The closed helix and the continuous brass tube diminish the action of a bundle of iron wires which they surround, so as to render this action similar to that of a solid iron cylinder, and even inferior to it, although it was very far superior before the wires were thus enveloped. The tube split lengthwise produces thus a less reduction of force than the closed tube, but a little more considerable than a helix, the extremities of which are not united. By closing the circuit of the split tube by means of the wire of a galvanometer, we may prove the existence of the induced current. These experiments show that the difference we found to exist between the action of a bundle of iron wires and that of a solid cylinder is essentially due to the currents of induction being enabled to be set up on the surface of the cylinder.

From all the facts that we have described, it follows that the diversity of effects manifested by the currents of induction is due to a difference in their duration, and not to a difference in their energy. The metal envelope that surrounds a bundle of wires or the continuous surface of a solid iron cylinder does not enfeeble the action of induction, but retards it. This retardation has no influence on the galvanometer needle, upon which the effects of the current are accumulated, an operation for which duration is of no importance; but it greatly diminishes the physiological effect and the magnetising action. It is to the same cause we must attribute the fact, that the sparks and shocks produced by the extra current, namely, by the current induced upon itself in a helix, the axis of which is occupied by a bundle of iron wires or by pieces of solid iron, disappear almost entirely when the wires or pieces of iron are placed in a closed brass tube. But it is not the same if the brass tube is opened. The end of a gun-barrel produces the same results, when closed and open, as does the brass tube; only with the closed gun-barrel there is still a feeble shock.

A very important remark made by M. Dove, and the accuracy of which I have often had the opportunity of proving, is, that the effects are greatly increased, all other circumstances remaining the same, by reversing the polarity of the iron, whatever it may be that is in the interior of the helices, by means of a change in the direction of the current. This result is due to the fact, that iron, even the softest, always preserves a certain amount of magnetic polarity, even when the magnetisation has ceased, if at least it has been prolonged for a short time. It follows that the inductive current, produced by the rupture of the circuit, is less powerful than if the magnetisation at this moment ceased entirely. But, by reversing the polarity, we necessarily obtain this result. In general, the most powerful effect of induction belongs to the metal whose magnetic state experiences the greatest change. The indications of the galvanometer are, in this case, analogous to those resulting from sensation. Many examples show that, if we did not take into account this principle of the increase of energy by the reversal of polarity, it would be impossible to compare together different sorts of iron. Thus soft steel and tempered steel, when their polarity is reversed, surpass all kinds of cast iron; and these in their turn, when their polarity is reversed, have all a greater energy than soft steel and tempered steel. It is the same with the different species of cast iron in respect to each other, and with bundles of iron wire.

After having employed electric currents, M. Dove employed discharges for producing induction, with a view of determining the influences that may be exercised upon induced discharges by the introduction of all kinds of metals. The apparatus that he employed, and which he termed the differential inductor, consisted of two helices made of two thick copper wires, and wound upon two glass tubes 13 in. long by 1 in. in diameter. The spirals of these wires were insulated with great care by means of a thick coat of gum-lac. Two other larger helices, but perfectly similar to each other, wound in the same direction around a tube of card, and consisting of the same number of convolutions, received in their interior

the first-named helices. The latter transmitted the discharge of a battery of Leyden jars, which traversed successively the two wires. There resulted from this an induced discharge in the two enveloping helices. Of the four ends of these helices, two, belonging each to a different helix, were placed in communication by a conductor, and the two others served to transmit the induced discharge through the human body, the wire of a galvanometer, a magnetic spiral, or a voltameter. By the arrangement of the apparatus, we could at pleasure introduce into the interior of the helices solid cylinders or bundles of wire. This apparatus, although essentially intended for induction produced by discharges of common elec

The helices of the induced circuit may be united, so that the two induced discharges travel in the same direction, or so that they travel in the contrary direction. In the latter case, they completely neutralise each other, so that if the introduction of any substance in the interior of one of the helices modifies in any way the induction brought about in that helix, it is immediately perceived, because the neutralisation no longer occurs, and one of the induced discharges surpasses the other. Unfortunately, the galvanometer, the chemical voltameter, and the magnetisation of soft iron, are in no degree influenced by the instantaneous current of the induced discharge, so that we must have recourse to other means for estimating it, namely, to the shock, and to the magnetisation of steel needles. M. Dove made use of the physiological shock, which, by its degree of intensity, indicates by how much one of the instantaneous currents is superior to the other; but, in order to obtain the definitive direction of the current, and to discover which of the two helices gives a superior current to the other, he employed M. Riess's method, namely, a condenser and figures traced by the two electricities upon cakes of resin, placed between the two extremities of the induced wire. By operating thus, he found that the physiological effect of the inductive current, developed by the discharge of the jar, is enfeebled by the introduction into the bobbin of non-magnetic metals, and the more so as they are better conductors. This diminution, therefore, is much less for antimony, bismuth, and lead, than it is for copper. The effect of these cylinders is the same as that of helices whose two ends are united; it is then the result of the reaction exercised upon the secondary circuit, by the instantaneous currents developed upon their surface. Thus, tubes cleft longitudinally, and bundles of wires, diminish much less the effects of induction than do continuous cylinders. Forged iron, tempered and untempered steel, grey or white cast iron, equally diminish the physiological effect by their introduction: but it is not the same with bundles of insulated iron wire; there is then, on the contrary, an augmentation. The condenser in this case indicates that the instantaneous