These boxes may be suspended on a cross-bar (fig. 47) by the two communicating wires, so as to be completely insulated. In this case the ticking of a watch placed upon the board, friction, and external shocks are hardly heard, but on the other hand the sound vibrations of the air alone are transmitted, and they acquire great distinctness. We have often repeated these experiments, and have always found that the tones of the voice were perfectly preserved. The model represented fig. 48 is still more simple, and appears to be the latest development of this kind of instrument. It consists of a stand and a disk united by a central rod. The upper disk moves round the central rod, and permits the vertical carbon to assume any inclination which is desired. It is evident that the instrument will become less sensitive when the carbon is more oblique. [Illustration: FIG. 47.] We must also mention a very successful microphone devised by M. Lippens. It is a slightly made box, like that of M. Varey, and on its opposite faces there are applied, on two frames left empty for the purpose, two thin plates of hardened caoutchouc, in the centre of which inside the box, two carbons are fastened, and on their outer surface a half-sphere is hollowed. [Illustration: FIG. 48.] The interval between the two carbons hardly amounts to two millimètres, and a carbon ball is inserted into the two cavities which form its spherical case. This ball is supported by a spiral spring which can be extended more or less by means of a wire wound on a windlass which is fixed above the instrument, like the spring of an electric telegraph instrument. By means of this spring, the pressure of the carbon ball against the sides of the cavity which contains it can be regulated at pleasure, and the sensitiveness of the instrument and its capacity for transmitting speech can be adjusted. Under these conditions, the vibrations of the caoutchouc plates directly affect the microphone, and the currents of air have no influence on it, so that the effects are more distinct. It is so sensitive that it is best for the speaker to place himself at the distance of at least 50 centimètres from the instrument. M. Lippens’ instrument is a pretty one, mounted on a wooden stand, which is neatly turned.