OP ARTS AND SCIENCES. 261 



Striking, were sufficiently encouraging to lead us to hope for greater 

 ultimate success. The thermal telephone, with a platinum wire 4i- 

 in. long and 0.004 in. in diameter, and an iron diaphragm, was placed 

 in circuit with a battery of 10 Grenet cells and an ordinary hand tele- 

 phone receiver. The current was sufficient to heat the wire almost to 

 redness. All the joints of the wire were soldered to prevent the pos- 

 sibility of microphonic action, which at first caused some annoyance, 

 but which was readily recognizable by the scratching sound produced 

 by it. With the joints thus soldered, and every precaution taken to 

 avoid such microphonic disturbances, we succeeded in hearing and un- 

 derstanding at the receiver the word " Hallo ! " and some brief familiar 

 sentences. The sound was low, but distinct and clear, and in quality 

 quite unlike the sound produced by mici'ophonic action. There seems 

 to be no doubt that electrical undulations were produced in the circuit 

 by variations in the stress to which the heated wire was subjected. 

 Similar results were obtained with a smaller number of cells, even 

 down to one ; but the sound was then exceedingly faint and scarcely 

 audible. Speech was also transmitted by a thermal transmitter in 

 which German-silver wire was substituted for the platinum. It may 

 be that the more marked results on heating the wire to a higher tem- 

 perature by the use of a stronger batter}' indicate that at such a 

 temperature the electrical effect of changes in stress in varying the 

 resistance of a wire is increased. 



Rogers Laboratory of Physics. 



