gas presents several inconveniences. It is difficult to prepare; it necessitates the use of gas-holders, whose size may he considerable. Besides, there is some danger in its use. I have therefore given up using hydrogen gas, and for a year I have experimented on the means of applying common illuminating gas to the pyrophone, which it is always easy to procure. In the first experiments which I attempted with two flames, with illuminating gas, in a glass tube, I was unable to obtain any sound, which proved unmistakably the presence of carbon in the flames. While the sound was produced in a very clear manner with the pure hydrogen gas, that is to say, without there being any solid foreign matter in the flames, it was impossible to make the tube with illuminating gas vibrate, when placing the flames in an identical condition. It was necessary, then, by some means or other to eliminate the carbon, a result at which I arrived by dint of the following method:
"When the flame of ordinary gas is examined, and this is introduced into a tube made of glass, or of any other material (metal, oil-cloth, card-board, etc.), this flame is either illuminating or sounding.
"When this flame is only illuminating, that is to say, when the air contained in the tube does not vibrate, it presents a lengthened form, and is pointed at the top. Besides, it swells toward the middle, and flickers on the least current of air. On the contrary, when the flame is sounding, that is to say, when the necessary vibrations for the production of sound are produced in the tube, its form is narrow, and large at the top. While the air of the tube vibrates, the flame is very steady. The carbon in a great measure is eliminated as if by some mechanical process.
"Sounding-flames proceeding from lighting gas are in effect enveloped in a photosphere which does not exist when the flame is merely luminous. In the latter case the carbon is burnt within the flame, and contributes in a great degree to its illuminating power.
"But, when the flames are sounding, the photosphere which surrounds each of them contains an exploding mixture of hydrogen and oxygen which determines the vibrations in the air of the tube.
"To produce the sound in all its intensity, it is necessary and sufficient that the whole of the explosion produced by the particles of oxygen and hydrogen in a given time should be in agreement with the number of vibrations which correspond to the sound produced by the tube.
"To put these two quantities in harmony, I have thought of increasing the number of flames so as to increase also the number of the explosions from the mixture of oxygen and hydrogen in the photospheres, and thus determine the vibration of the air of the tube. Instead of two flames of pure hydrogen, I put four, five, six, etc., jets of lighting gas in the same tube.
"I have besides observed that the higher a flame is, the more carbon it contains.
"I have then immediately been obliged to diminish the height of the flames, and consequently to increase the number so that the united surface of all the photospheres may suffice to produce the vibration of the air in the tube.
"The amount of carbon contained in the whole of the small flames will always be much less than the quantity of carbon corresponding to the two large flames necessary to produce the same sound. In this manner I have been able with separated flames to obtain sounds whose tones are as clear as those produced by hydrogen gas. When these, flames, or rather when the photospheres which correspond to these flames, are put in contact, the sound instantly ceases.
