3. If, instead of advancing or withdrawing the magnet by means of a single movement, it is advanced in jerks, a succession of currents in the same direction is produced, of which the existence can be ascertained by the galvanometer when there is a sufficient interval between the movements, but when the intervals are very slight the currents are interfused; and since inverse effects take place when the magnet is moved in a contrary direction, the needle of the galvanometer follows the movements of the magnet, and to a certain extent stereotypes them. 4. If, instead of reacting on a simple closed circuit, the magnet exerts its force on a considerable number of circumvolutions of this circuit, that is, on a bobbin of coiled wire, the effects will be considerably increased, and they will be still greater if there be a magnetic core within the bobbin, since the inducing action will then be more effectually exerted throughout the bobbin. As the magnetic core, when it is magnetised and demagnetised under the influence of its approach to or withdrawal from the inducing magnet, is subject to the reaction from all the fluctuations which occur in the movements of the magnet, the induced currents which ensue are perfectly defined. 5. If, instead of a movable magnet, we suppose it to be fixed in the centre of the coil, the induced currents of which we have spoken may then be determined by modifying its force. In order to do so, it is enough that an iron armature should react upon its poles. When this armature is brought close to one of the poles, or to both at once, it acquires force, and produces an inverse current, that is, a current in the direction which would have corresponded to an approach of the magnet to the closed circuit. On its withdrawal the inverse effect is produced; but in both cases the induced currents correspond with the extent and direction of the movements accomplished by the armature, and consequently they may reproduce its movements by their effects. If this armature is an iron plate, which vibrates under the influence of any sound in this disposition of the electro-magnetic system, the alternate movements of the plate will be transformed into the induced currents, and these will be stronger or weaker, more or less definite, according to the range and complexity of the vibrations: they will, however, be undulatory, since they will always result from successive and continuous movements, and will consequently be in the conditions which, as we have seen, are required for the transmission of speech.