64 
Proceedings of the Royal Society of Edinburgh. [Sess. 
VII. — The Axial Inclination of Curves of Thermoelectric Force : a 
Case from the Thermoelectrics of Strained Wires. By John 
M‘Whan, M.A., Ph.D., Lecturer in Mathematics in the University of 
Glasgow. Communicated by Professor Andrew Gray, LL.D., F.R.S. 
(MS. received October 17, 1913. Read February 16, 1914.) 
In a communication to the Royal Society of Edinburgh,* Mr J. D. Hamil- 
ton Dickson has examined with great care the valuable results of Professors 
Dewar and Fleming on the thermoelectromotive forces of various couples, 
and has come to the conclusion that the curve representing the thermo- 
E.M.F. is in every case a parabola whose axis is, not vertical as had always 
been assumed, but inclined a definite though very small angle to the 
E.M.F.-axis. 
This remarkable result has led me to go back to some experiments 
which I made a few years ago on the thermoelectric properties of longitu- 
dinally strained metal wires, to see if by any chance the same phenomenon 
might be detected there, and in one instance (only) I have been able to 
establish its existence unmistakably. The experiments in question, which 
I have described elsewhere, j* were made on couples consisting each entirely 
of one and the same pure metal ; but one wire of the couple might be subjected 
to any desired longitudinal tension while the other remained unstrained. 
The temperatures of the junctions were the same in all the experiments, 
one junction being steam-heated, the other water-cooled. 
On reference to the curves showing the relation of the tension in the 
strained wire to the thermo-E.M.F. of the couple, only one was found to be of 
parabolic shape, namely, that for nickel. The E.M.F.’s in non-magnetic 
metals nearly all obey a straight-line law up to the point where overstrain 
sets in : the only other magnetic metal tested, bismuth, gave no simple 
relation between E.M.F. and strain. A closer examination of the nickel- 
curve showed that, while it afforded good grounds for suspecting an in- 
clination of the axis, the observations had been neither numerous enough 
nor of the high order of accuracy necessary for certainty, and it was 
accordingly decided to repeat the experiment. This repetition proved 
extremely laborious, and only after some five or six attempts (each necessitat- 
* Trans. R.8.E . , xlvii. 737-791, 1910-11. 
t Diss., Gottingen, 1911. 
