276 Scientific Proceedings, Royal Dublin Society. 
wire, and the pair included in the circuit of the telephone, Fig. 3. 
Upon heating the junction of the iron and copper wire at A to in- 
candescence, and then allowing it to cool, the following phenomena 
were observed:—With the wire white-hot, the humming of the 
contact breaker was heard loudly, indicating a powerful current. 
At ared-heat, the sound was feebler, and at a dull red-heat, no 
sound at all was heard, indicating the “neutral point.” Allow- 
ing the wire to cool still further, an inverse current is set up, and 
accordingly the sound was heard again, growing quite loud at a 
still lower temperature, and then fading away till the junction 
was at the temperature of the air. When this had occurred the 
reverse action was, of course, noticed on heating instead of cool- 
ing the wire. Tinned copper wire and iron gave similar results. 
T failed, however, in the telephone, to detect the neutral point, 
when cobalt or nickel was substituted for the iron. With the 
copper and iron pair, remarkably easy and useful class experi- 
ments can be shown. 
