732 
Proceedings of the Royal Society 
platinum wire, the result was almost as good as that obtained from 
the fine iron wire itself. This is rather a puzzling result to explain, 
as it cannot arise from magnetisation of the metal, as is probably the 
case, in part a least, with the iron wire. 
Fine wires of copper and aluminum transmitted no sound what- 
ever, although they were treated precisely in the same way as the 
iron and plantinum wires. 
A short strip of cobalt inserted in place of the fine wire at A gave 
a very distinct result, although the sound was not so loud as in the 
case of the iron wire. This, however, may be due to smallness of 
the vibrations arising from the stiffness of the strip. 
From all these experiments it is obvious that, some how or other, 
rapid variations of the current strength are produced in the circuit, 
and the problem is to explain how these variations are produced. 
Now the current strength can only be varied in two ways — either by 
varying the electro-motive force or by altering the resistance of the 
circuit ; and no doubt, in this case, both ways come into play to a 
certain extent. The former comes into play, inasmuch as the electro- 
motive force will be varied by the motion of the fine wire in the 
magnetic field caused by itself and by the earth’s magnetism. The 
latter will also come into play, inasmuch as change of resistance will 
arise from at least three distinct sources. These are (1) varying 
magnetisation in the wire, especially the iron wire, produced by 
strain ; (2) variations in the temperature of the wire caused by the 
cooling effect of the air as the wire vibrates ; (3) alterations of the 
resistance caused by the varying strain to which a vibrating wire is 
exposed. It is possible to arrive at something like a numerical 
estimate of this last cause of alteration for any particular note 
transmitted ; for, knowing the mass of the vibrating wire, its initial 
tension, its number of vibrations per second, we can find the varia- 
tions in the strain to wdiich it is subject ; and if we can know how 
varying strain is connected with change of resistance, as may be 
got from Sir William Thomson’s recent paper on that subject, we 
have all the elements necessary for making the calculation. 
It is right that I should mention that, in making the experiments, 
I had the valuable aid of Professor Chrystal in the twofold way of 
helping me to make exact observations, and of suggesting various 
changes of experiment to bring out or eliminate particular effects. 
