120 Proceedings of the Boyal Society 
a glass tube containing a singing flame, and connected it in circuit 
with a battery of six Grove’s cells and an ordinary telephone, or 
still better a Gower Bell telephone. The note produced by the 
singing flame was faintly but distinctly heard in the telephone 
placed in a distant room. It will be readily seen, that, in this 
experiment, we take advantage both of the heat and light effect 
upon the selenium, and that the action is increased by the close 
proximity of the cell to the flame. 
If the indiarubber tube carrying the gas to the singing flame be 
rapidly pinched and let free, the increase and decrease of the flame 
is instantly indicated by the telephone ; and if the pinches are made 
rhythmically, a corresponding rhythm is heard in the telephone. 
The above experiment can also be performed very well by 
attaching the singing flame jet by an indiarubber tube to one of 
Koenig’s manometric capsules screwed on to an organ pipe. The 
note of the pipe is then heard in the receiving telephone. 
Having obtained a cell which was sensitive to a singing flame, it 
was an easy step to try if it would be equally sensitive to the varia- 
tions of the flame of a Koenig’s capsule produced by singing and 
speaking. To test this, I placed the cylindrical cell over the flame, 
so that the flame was in the axis of the cylinder and exactly opposite 
the selenium. When the cell was now joined up in circuit with 
fifteen Grove’s cells and a telephone, not only singing, but speaking 
was distinctly reproduced in the telephone. 
This experiment is very remarkable, as it shows that the resist- 
ance of the selenium must vary with very great rapidity ; otherwise 
it could not transmit articulation. 
Flat cells of the comb form are very easily constructed by simply 
screwing the combs upon a flat piece of wood, taking care that the 
teeth do not touch each other. The whole is then heated up to the 
melting point of selenium, which is then rubbed carefully between 
the teeth. The annealing is a matter of great importance. It can 
be done very well in an air-bath, whose temperature can be regu- 
lated by a thermometer. During the time the cell should be joined 
up in circuit with a Wheatstone’s bridge, and its resistance carefully 
tested all through the annealing process. It will be found that 
the resistance of the cell gradually diminishes with a rise of tempera- 
ture, and finally attains a minimum, after which it begins to rise 
