_Considerations on the theory._--Although the explanation we have given will make the effects of the phonograph intelligible, it leads to a curious question which has greatly interested physicists--namely, how it is that the tracing made on so yielding a surface as tin can, when retraced by the stylus, of which the rigidity is relatively great, produce a vibratory movement without being completely destroyed. To this we reply that the retracing is effected with such extreme rapidity that the effects of active force which are developed only manifest themselves locally, and that under these conditions the mechanical effects exerted are as energetic in soft as in hard substances. The curious experiment, related in so many books on physics, must be remembered, of a plank pierced when a candle serves as the projectile of a gun. The various accidents caused by the discharge of paper waddings must also be remembered. Under such conditions the motion imparted to the molecules which receive the shock has not time to be transmitted to the whole mass of the substance to which they belong, and these molecules are compelled to separate from it, or at any rate to produce, when the substance is capable of vibration, a centre of vibration which diffuses waves throughout its surface, and produces sounds. Several scientific men, among others Messrs. Preece and Mayer, have carefully studied the form of the tracing left by the voice on the tinfoil of the phonograph, and they observe that it greatly resembles the outline of the singing flames so well shown by Herr Koenig’s instruments. Mr. Mayer wrote on this subject in the ‘Popular Science Monthly Review’ of April 1878. He said that he had been successful in reproducing a splendid tracing on smoked glass, which gave in profile the outline of the vibrations of sound registered on the tinfoil with their varying curves. For this purpose he fastened to the spring support of the tracing point of the phonograph a slender rod, terminating in a point, which pressed obliquely against the plate of smoked glass, and, since the latter was in a vertical position, a movement imparted to the rod enabled it to produce a sinusoidal tracing. By this arrangement, when the phonograph was at work, two systems of tracings were produced at the same moment, of which one was the profile of the other. Mr. Mayer had not, at the time he wrote, been long enough in possession of the instrument to make many experiments with it, but from a study of some of its curves it appeared to him that the registered outlines bore a strong resemblance to those of Koenig’s singing flames. [Illustration: FIG. 71.]