702 
Proceedings of the Royal Society 
of its heat as suddenly as it did of its electric charge. A wire takes 
a sensible time to warm up to the balance of internal gain and 
external loss of heat and also to lose the heat it has acquired. Now, 
unless the increments and decrements of length are in the strictest 
sense momentary, they cannot affect a telephone disc. The clear 
click that a wire emits when a current begins or ceases in it, both 
exactly equal in loudness, seemed to me too sharp for the compara- 
tively sluggish course of expansion and contraction by heat and cold. 
If Professor Chrystal’s theory be correct, it can only be the initial 
increment and decrement of heat that count in this phenomenon, 
and not subsequent gradual gain or loss. Eapidity of alternation is 
by no means necessary, for one close or break in a second is as clearly 
rendered as 500. The extension theory thus appears to me to sub- 
stitute a thermal for a magnetic onset, and to make it as likely that a 
sounding clash of molecules accompanies it. 
It struck me on considering Professor Chrystal’s reasoning on the 
experiment he exhibited, that if the sound be really in the wire, the 
skin disc, if placed in the middle instead of at the end of it, would 
still sound. We should have thus two telephonic receivers, one on 
each side of the disc. If the sound was due to the push and pull 
of the wire on getting longer and shorter, the two wires would 
eliminate the effect the one of the other, but if to internal commo- 
tion, little change would be observed whether we had a sounder on 
one side or on both. So far as I have been able to ascertain the 
latter is the case. I arranged an experiment in the following- 
way : — wrefghrw is a composite thread consisting of very fine 
platinum wire and fine cotton thread — ef and gh are the platinum 
parts. The thread is kept stretched by two equal weights w,w. 
The middle of the thread is attached to the skin or paper disc of an 
ordinary mechanical telephone dd. The thread is symmetrically 
made up on each side of the disc, ef and gh being exactly equal in 
every way. At the ends of the platinum wires, wires of copper are 
soldered which dip into the mercury cups Mm and n , cz is a 
Bunsen battery of six cells. I is a current interruptor or micro- 
phone. The one used on this occasion was the mercury break I 
used for my last communication, vibrating some six times per 
second. I do not suppose that this deliberate beat, as compared 
with Professor Chrystal’s microphone, which vibrated from 200 to 
