As for the action produced upon the receiver, that is, on the instrument for reproducing speech, it is somewhat complex, and we shall have occasion to speak of it presently; but we can get a general impression of it, if we consider that the effects produced by the induced currents of variable intensity, which traverse the coil of the electro-magnetic system, must determine, by the magnetisations and demagnetisations which ensue, the vibrations of the armature disk; these vibrations, more or less amplified and defined, exactly represent those of the disk before which the speaker stands, and can only be obtained from them. The effects are, however, in reality more complex, although they are produced under analogous conditions, and we shall have more to say about them when we come to speak of the experiments made with the telephone. It must meanwhile be observed that, for the reproduction of speech, it is not necessary that the magnetic core should be of soft iron, since the vibratory effects may follow from differential as well as from direct magnetisation. ORDINARY ARRANGEMENT OF THE BELL TELEPHONE. The arrangement most generally adopted for the telephone is the one represented in fig. 21. It consists of a kind of circular wooden box, fitted to the extremity of a handle M, which is also of wood, and contains the magnetic bar N S. This bar is fixed by means of a screw _t_, and is so arranged as to be moved forward and backward by tightening or loosening the screw, a condition necessary in order to regulate the instrument. At the free extremity of the bar the magnetic coil B is fixed; this must, according to MM. Pollard and Garnier, be made of wire No. 42, so as to present a considerable number of spirals. The ends of this coil generally terminate at the lower end of the handle in two copper rods _f_, _f_, which traverse its length, and are fastened to two binding-screws I, I′, where the line wires C, C are fixed. In the instruments made by M. Bréguet there are, however, no binding-screws, but a little twist, made of two flexible wires covered with gutta-percha and silk, is fastened to the two rods. A wooden cap is screwed to the end of the handle, and the twist passes through a hole made in this cap, so that there is no inconvenience in working the instrument. By laying hold of the ends of the wire twist with pliers it is possible to join them to the circuit. This instrument is represented in fig. 22. [Illustration: FIG. 21.] By another arrangement, the wires of the coil end immediately in the binding-screws which are placed below the wooden box, but this arrangement is inconvenient.