Trouton — On T/iermo-JEkctric Current in Single Conductors. 175 



doing this is considerable, especially in silver, for long below its 

 melting-point it appears to lose tenacity almost completely. 

 Platinum was examined at temperatures approaching its melting 

 point with the oxyhydrogen flame ; but the currents obtained were 

 very small, and were due to irregularities in the structure of the 

 wire. As the flame was being carried along in one direction the 

 needle kept to one side or the other, according to the part of the 

 wire the flame was at. A thin rod of carbon examined in the oxi- 

 hydrogen flame gave a small but regular current as the flame was 

 moved along when water was applied to cool the carbon behind the 

 flame ; without this the carbon does not cool quickly enough to give 

 a current. 



A difference in the potential along a wire of the nature sup- 

 posed above, that is due to a temporary change in structure from 

 temperature, could not be discovered by means of a galvanometer, 

 owing to the symmetry on either side, except the temperature 

 alter more rapidly than the structure. For otherwise it would 

 always be equivalent to introducing another metal, and keeping 

 the two junctions at the same temperature. 



To state shortly the conclusions finally arrived at, there is, 

 first, a permanent alteration effected in the structure of a wire 

 when it has been once heated. So that, if one of the points 

 between the altered and unaltered metal be warmer than the 

 other, a current flows similar to what would happen if a second 

 metal were introduced into the circuit instead of the altered part. 

 Secondly, that there is, at least in some metals, a temporary 

 alteration of a somewhat similar nature to the permanent one, 

 which lasts while the wire is at a high temperature ; and that it is 

 possible to obtain currents from this, is solely due to the fact that, 

 both in appearing and in disappearing, the alteration may take 

 place more slowly than a change in temperature, which ultimately 

 effects the alteration. 



What this alteration is, whether stresses similar to what Sir 

 William Thomson found could produce thermo-electric hete- 

 rogeneousness in a single metal, or whether of the nature of 

 molecular rearrangement of the nature of annealing, may be 

 doubtful. 



