These two series of experiments show that sounds may be obtained from different parts of the telephone without any very appreciable vibratory movements. But Signor Luvini wished for a further assurance of the fact, by ascertaining whether the magnetisation of any magnetic substance, followed by its demagnetisation, would involve a variation in the form and dimensions of this substance. He consequently caused a large tubular electro-magnet to be made, which he filled with a quantity of water, so that, when its two ends were corked, the liquid should rise in a capillary tube fitted to one of the corks. In this way the slightest variations in the capacity of the hollow part of the electro-magnet were revealed by the ascent or descent of the liquid column. He next sent an electric current of varying intensity through the electro-magnet, but he was never able to detect any change in the level of the water in the tube; although by this arrangement he could measure a change of volume of 1/30 of a cubic millimètre. It appears from this experiment that the vibrations produced in a magnetic substance under the influence of successive magnetisations and demagnetisations, are wholly molecular. Yet other experiments made by M. Canestrelli seem to show that these vibrations are so far sensible as to produce sounds which can be detected by the microphone. He writes as follows on the subject: ‘When the broken currents of an induction coil are discharged into a coil placed on a sounding-box, it is possible to hear at a little distance the sounds produced by the induced currents thus generated. On approaching the magnet to the opening of the coil, these sounds are intensified, and the vibrations of the magnet become sensible to the touch; this vibration might even be made visible by suspending the magnet inserted into the coil to a metallic wire, which is fitted to a membrane stretched on a drum, and the latter will then reproduce sounds. When the same magnet is suspended to a microphone, it is possible, with the aid of a telephone, to ascertain the same effects, which are then increased.’ We shall presently consider how these different deductions are to be interpreted, so as to render the true theory of the telephone intelligible; but, before doing so, we will mention some other experiments which are not without interest.