The microphone may also be made of morsels of carbon pressed into a box between two metallic electrodes, or enclosed in a tube with two electrodes represented by two elongated fragments of carbon. In the latter case the carbons ought to be as cylindrical as possible, and those made by M. Carré for the Jablochkoff candles are very suitable. Fig. 43 represents an instrument of this kind which M. Gaiffe arranged for me, and which, as we shall see, serves as a thermoscope (fig. 44). It is composed of a quill filled with morsels of carbon, and those at the two ends are tipped with metal. One of these metal tips ends in a large-headed screw which, by means of its supports A B, is able to press more or less on the morsels of carbon in the tube, and consequently to establish a more or less intimate contact between them. When the instrument is properly regulated, speech can be reproduced by speaking above the tube. It is therefore a microphone as well as a thermoscope. Mr. Hughes has remarked one curious fact, namely, that if the different letters of the alphabet are pronounced separately before this sort of microphone, some of them are much more distinctly heard than others, and it is precisely those which correspond to the breathings of the voice. A microphone of this kind may be made by substituting for the carbon powders of more or less conductivity, or even metal filings. I have shown in my paper on the action of substances of moderate conductivity, that such power varies considerably with the pressure and the temperature; and as the microphone is based on the differences of conducting power which result from differences of pressure, we can understand that these powders may be used as a means of telephonic transmission. In a recent arrangement of this system Mr. Hughes has made the powder adhere together with a sort of gum, and has thus made a cylindrical pencil which, when connected with two electrodes which are good conductors, can produce effects analogous to those we have just described. As I have said, it is possible to use metal filings, but Mr. Hughes prefers powdered charcoal. Mr. Blyth states that a flat box, about 15 inches by 9, filled with coke, and with two tin electrodes fixed to the two ends, is one of the best arrangements for a microphone. He says that three of these instruments, hung like pictures against the wall of a room, would suffice, when influenced by a single Leclanché cell, to make all the sounds produced in a telephone audible, and especially vocal airs. Mr. Blyth even asserts that a microphone capable of transmitting speech can be made with a simple piece of coke, connected with the circuit by its two ends, but it must be coke: a retort carbon, with electrodes, will not act.