of Iron when under Stress or Strain. 51 



shifted suddenly and slightly towards C; almost immediately 

 afterwards the light darted off the scale of the galvanometer, 

 indicating a strong current from A to D through C. The 

 portion D was again heated above a bright red temperature, 

 and the burner then was shifted suddenly towards D ; the 

 deflection of the galvanometer needle was now strong in the 

 opposite direction, showing a current from B to C through D. 



Experiment III. furnishes the clue to the explanation of 

 the phenomenon described in Experiment IV. The portion 

 C D, when at a bright red, is thermoelectrically negative to 

 the rest of the iron wire, and is in fact like a different metal. 

 So long, however, as the flame of the burner is kept in one 

 position the temperature at C will be the same as at D, and 

 there will be no current ; but directly we shift the burner 

 towards C we make C hotter than before, whilst I) cools more 

 or less, and we get a current from A to I) through C. 



We also see why a slight draught of air blowing the flame 

 of the burner towards C or D may suddenly generate a strong 

 current in one direction or the other and may completely 

 mask the phenomena recounted in Experiments I. and II. 



Experiment IV. may be varied in several ways, thus : — If 

 we keep the flame steadily burning so as to maintain C D at 

 a high temperature, and allow a small stream of water to flow 

 down on C, we get a continuous current from B to C through 

 D, and if we shift the stream to D we reverse the current. 

 Or, instead of the stream of water, we may blow with a small 

 pair of bellows at C or D. 



What connexion there may be between the above-men- 

 tioned phenomena and the discovery of Professor P. Gr. 

 Tait that the thermoelectrical diagram of iron cuts that of 

 iridium-platinum at least three times below a low white heat*, 

 that is to say that an iron and iridium-platinum circuit has at 

 least three neutral points, does not quite appear. 



In conclusion, it may be said that the investigation serves still 

 further to confirm the author in his opinion that a sudden 

 and profound change takes place in the molecular arrange- 

 ment of iron at a temperature of bright red. 



* Heat, by P. G. Tait, § 199. Also Trans. R. S. E. 1873, for similar 

 peculiarities in nickel at lower temperatures. The author has failed to 

 observe in nickel wire any such phenomena as those mentioned in 

 Experiments I. and II. 



E 2 



