282 MR JAMES BLYTH ON TRANSMISSION OF 
- 
at A was then distinctly heard at B, even when the can was at some distance 
from the ear. The same sound was heard when the file at A was laid against 
any part of the can, but most loudly when it happened to be against a corner or 
other sharp edge. It was remarkable also, that the sound was heard distinctly 
even when the file did not touch the can at all, but was merely laid against the 
wire attached to it, so as to complete the electric circuit without including the 
can in it. It would seem from this that some mechanical tremor is set up at 
the loose contact of the file with the wire which is transmitted along the wire 
to the can. As a variety of this experiment, I removed the can from the wire, 
and substituted in its place a poker, having the circuit wire firmly attached to 
its point. When the other end of the poker was put to the ear, and the file 
applied to the poker at any point, the sound of the distant rasping was distinctly 
heard. The same was the case when a long brass tube was substituted for the 
poker, all which very strongly suggests the idea of a mechanical tremor trans- 
mitted through the metal from the point of loose contact. 
Experiment 2.—In this experiment a common automatic make-and-break, 
consisting of a vibrating spring worked by a small electro-magnet, was intro- 
duced into the circuit at A, and a similar spring, only without the electro-magnet, 
at B. At B the sound of the vibrations of the springs at A was so distinctly 
heard, as to at once suggest the idea that the spring at B was itself vibrating. 
However, I was unable to detect any such vibration, either with the aid of a 
microscope, or by attaching a small polished bead to the spring and observing 
in it the reflection of a light. Still it would be rash, I think, to assert that such 
vibrations were not present, and it is possible that, by more refined experimental 
means, they may yet be made manifest. It was very noticeable in this experi- 
ment that the sound at B got less and less loud as the pressure on the vibrating 
spring was increased, until it ceased altogether when the contact was made 
perfectly tight. 
Experiment 3.—The sound from the poker in Experiment 1 was so like 
that produced by the TREVELYAN rocker, that it immediately suggested the em- 
ployment of that apparatus as the loose contact at B. For this purpose the 
current was led through the lead block, the rocker, and a brass plate, on which 
the ball at the end of the rocker rested. When this was done, and the make- 
and-break set agoing at A, a distinct sound was heard at B, suggesting very 
strongly the idea that the rocker was in actual vibratory motion. To test this 
in some measure, I heated the rocker and laid it on the lead block, when two 
sounds were distinctly heard, one due to the make-and-break, and the other to 
the heat effect. The one did not seem in the least to interfere with the other. 
Still farther to test the idea of actual vibration, it occurred to me to try if one 
rocker could not be made to act as the make-and-break to agitate the other. — 
For this purpose two precisely similar rockers were taken, consisting of two 
