Page:Collected Physical Papers.djvu/300

280

If the wire had been carefully annealed, the molecular conditions of its different points are approximately the same. The wire will therefore be practically iso-electric throughout its length. If the wire be now held near the middle by the clamp, and a vibration through an amplitude of, say, 90° be given to the end A, an upward deflection will be produced; an equal and opposite deflection will be produced by similar vibration of B (fig. 71, c). If both the ends are simultaneously vibrated, the electromotive variation at the two ends will continuously balance each other, and the galvanometer spot will remain quiescent (fig. 71, d). The clamp is next removed, and the wire vibrated as a whole; the stimulation of A and B being the same, there is no resultant deflection. Having found the balancing point for the clamp (which is at or near the middle), if the clamp be now shifted to the left, on simultaneous vibration of A and B, the A effect will relatively be the stronger (inasmuch as the torsional vibration of A is increased and that of B decreased), and there is produced a resultant upward deflection. Thus keeping the rest of the circuit untouched, by merely moving the clamp from the left, past the balancing position to the right we get either a positive or zero or a negative resultant effect. This can be repeated any number of times. The experiment shows further that when the amplitude of vibration is kept constant, the intensity of electromotive effect is increased by shortening the wire. The normal direction of the current of response in the wire is in the majority of metals, from the relatively less to the relatively more excited point.