THERMO-ELECTRIC EFFECTS OF LONG. TENSION IN D1FF. METALS. Q\ 



tions to vary so much as it might have done otherwise. Some 

 regularities are discernible from the curves, but their interpretation 

 seems not to be single. There may be an E. M. F. due to structural 

 difference and this may vary at different temperatures ; or else, there 

 may be an independent source of E. M. F. and this may vary not 

 only at different temperatures, but also from time to time. Again, 

 both these variations can quite possibly exist side by side. The 

 wires being left undisturbed throughout the night and almost the 

 whole of the next day, so that it could be assumed that all parts of 

 my arrangements had attained the atmospheric temperature/ I closed 

 the circuit of the experimental and leading wires through t e galvano- 

 meter and found a current which corresponded to an existence some- 

 where of an E. M. F. amounting to —1*7 micro-volts. Closing the 

 leading wires alone through the galvanometer an E. M. F. of nearly 

 — 0*1 micro- volts was found to exist. 



Where the seat of this E. M. F. is, and how it comes into exis- 

 tence, I can only guess at, and have no experimental data to base an 

 answer on. At all events, it is certain that the E. M. F.'s measured 

 and given in the table of § 31 are not pure inspite of the care which 

 has been bestowed upon them, and cannot be referred to as such 

 unconditionally. 



Still another thing which cannot be overlooked with respect to 

 the gradual shifting of the thermo-electric curve as a whole towards 

 the negative side, is the magnetisation of the iron wires. From Prof. 

 Ewing's experiments made on the relation, on the one hand, between 

 the magnetic qualities of iron when under stresses and when not, and, 

 on the other hand, the relations of the thermo-electric E. M. F. 

 between them, and also from a few of my own experiments, I have 



(I) The temperature of the cellar, where all of the experiments were carried out, was. 

 24.°5— 25.°5. 



