precautions are not indispensable; they are not really necessary, except in certain cases in which we are engaged upon particular effects, as we shall see further on.

One of the most important parts of soft iron magnetoelectric machines is the rheotome. The most simple is a toothed wheel, moved by the hand by means of a small handle, and communicating by its axis with one of the poles of the pile, whilst the other pole is in communication, by means of the inducing wire, with an elastic plate resting against the teeth of the wheel. Every time the plate, in consequence of the rotatory movement, leaps from one tooth to another, there is an interruption and a completion of the circuit in the inducing wire, and consequently a production of two induced currents successively in opposite directions in the induced wire. When the two ends of the induced wire are not united by a conductor, the currents give rise to a spark that escapes between the plate and the tooth, and which arises from the current induced in the inducing wire itself, which is only manifested so long as the former is not manifested.

The rheotome that gives rise to the most powerful current of induction is the mercury rheotome that is described (Fig. 130.). We require only two needles, which close the circuit when they are plunged into the mercury, and interrupt it when they come out: they are fixed transversely to an axis that is itself moved by a clock movement, the velocity of which is regulated by means of fans, to which is added the effect of the resistance of the mercury. The only inconvenience presented by this system is that the mercury is projected, and is very rapidly oxidised. It is then covered with small black globules, that interfere with the perfect contact of the needles with the liquid metal, and, as they rise above its surface, frequently make the two compartments communicate together in a permanent manner; and they ought not to have any other communication than what is established, at intervals, by the two needles that plunge into it. The apparatus annexed (Fig. 147.), constructed by M. Bonijol, represents an induction machine, in which the two rheotomes may be in

differently employed. The two cups, A and B, are intended to receive two poles of the pile (generally a single pair).

From the cup B there proceed two conductors, one of which is attached to the toothed wheel, the other to one of the compartments of the mercury rheotome. From the other compartment of this rheotome, as well as from the elastic plate that rests against the toothed wheel, comes a conducting wire; and these two wires arrive at the other extremity of the inducing wire. If we wish to employ the mercury rheotome, we set the clock movement in motion, and the needles immediately execute their immersion and their emersion: we take care at the same time to raise the elastic plate, so that it shall have no point of contact with the toothed wheel, and that the current can only be established by the mercury rheotome. If we desire to employ the toothed wheel, we make it rotate by means of the handle, after having put the elastic plate again in contact with it; we at the same time take care that the needles of the other rheotome, which are no

longer in motion, are arranged in their state of rest, so as not to plunge into the mercury, and consequently not to permit the current to circulate by this path.

The two knobs C and D represent the two extremities of the induced wire; any conductor by which they are united is traversed by the succession of induced currents. To the two other knobs, E and F, arrive metal wires, each proceeding from one of the extremities of the inducing wire: these knobs are intended for transmitting through every conductor, by which they are united, the currents induced in the inducing wire itself, in such a manner that, by using alternately the knobs C and D, and the knobs E and F, we can obtain either the one or the others; but we cannot make them act at the same time, the former currents existing only when the latter are not liberated, as we have already remarked. It is, indeed, a very remarkable thing to see this mutual dependence of the currents induced in the inducing wire itself, and the currents induced in another parallel wire. This circumstance is a good proof that the two species of currents are due to the same cause, and that they differ merely as to the place in which they are propagated.

The chemical and calorific effects that may be produced by means of these induced currents, as well by means of one as of the others, are very energetic. We can ignite a platinum wire, and can even obtain a small luminous arc between two points of coke. With regard to the chemical effects, they are very decided at the first instants: but the two gasses liberated alternately at the two wires of the voltameter very soon recombine, and the chemical power seems in appearance diminished.

A fifth knob, G, placed after E and F, communicates with a wire which comes to the cup B. It is employed for the following purpose; or, what amounts to the same thing, the following is what occurs, when it is united by a conductor to the knob E. The current induced in the inducing wire, which is liberated at the moment of the breaking of the circuit, is found by this combination compelled to traverse a circuit formed of the inducing wire, of the conductor interposed

between E and G, of the wire that goes from G to the cup B, of the voltaic apparatus that connects the cup B with the cup A, and of the wire that goes from the cup A to the inducing wire. The induced current is moving in the same direction as the inducing current, namely, as the current of the voltaic apparatus, so that the conductor placed between G and E is traversed at once by two currents, namely, the induced current and the voltaic current. There results from this an effect which is more than the sum of the effects of the two currents when separate; for it appears that the passage of the induced current through the voltaic apparatus notably increases the powers of this apparatus. This augmentation is especially sensible, when the voltaic apparatus consists of a single pair, and the conductor interposed between G and E is a voltameter. We then see the current of one pair, incapable of itself to decompose water, produce, with the addition of its own induced current, a very energetic chemical decomposition. This reinforcement of the current by itself, which I discovered in 1843, led me to give the name of electrochemical condenser to the apparatus, by means of which I succeeded in obtaining it.

This apparatus (Fig. 148.) presents a peculiarity of some

interest; it is, that in place of a toothed wheel, or a clockmovement with needles and mercury, it contains, for rendering the current discontinuous, a simple elastic metal stem, provided with a small piece of iron, placed very near to the

soft iron of the bobbin. When this soft iron is magnetised, the small piece of iron is attracted, and the elastic stem to which it is connected is raised; this occasions the breaking of the inducing circuit, which, in order to arrive at the wire of the bobbin, must traverse the stem, and pass into a small metal cup filled with mercury, or simply amalgamated, on the bottom on which the extremity of the stem rested. This rupture of the circuit destroys the magnetisation of the soft iron of the bobbin; the small piece being no longer attracted, the elastic stem immediately falls back, so that its extremity rests again on the bottom of the cup, and the circuit is made again. It follows from this that the current is interrupted a greater number of times, according as it is more intense; the rapidity with which the interruptions succeed each other depending on the power of the magnet.

Instead of having a distinct apparatus, we may adapt this kind of rheotome to the same case that already possesses the two others, as may be seen by Fig. 147. A small capsule, placed beneath G, and communicating with one of the ends of the inducing wire, receives the bended extremity of an elastic stem, that arrives at the cup A, and, passing near the soft iron of the bobbin, is furnished, in the part nearest to this iron, with the small piece, which is also of iron, and which, by the attraction that it obeys, raises the stem, and so interrupts the circuit.

Another more simple apparatus, also constructed by M. Bonijol (Fig. 149.), presents, under another form, the same system of rheotome. We interpose in the circuit of the inducing wire, a bent but fixed stem, A B, and an elastic plate, CD. The plate, by virtue of its elasticity, naturally places itself in contact with the extremity of the stem: the two surfaces in contact are covered with a platinum disc, in order to avoid oxidation and the rapid destruction of the metal. It is therefore between the two platinum discs that the contact occurs, and from which the spark darts, if there is one. The elastic plate is traversed in E by a stem that may be raised or lowered, so as to bring its extremity, which is a horizontal iron disc, at a proper distance from the soft iron of the bobbin. It follows from this, that the attraction exercised