653 
1907-8.] Young’s Modulus under an Electric Current. 
Again, in my first paper, the wire was put through only one cycle; 
but this investigation has been carried out in more detail by putting each 
wire through a sufficient number of cycles to bring it to the cyclically 
steady state. The iron and steel had to be carried through several cycles 
before this was accomplished, the platinum required two cycles, while the 
copper reached it at the end of the first. 
The investigation has also been extended in another direction, on 
account of the totally different results obtained by Miss Noyes* in a 
second paper, where another series of experiments on various wires is 
described. In her first paper the results, generally speaking, were similar 
to those I described in my first paper, viz. an increase of the modulus to a 
maximum, and then a decrease. In her second paper, however, the graphs 
are straight lines when the wire was heated by a current through it, as well 
as when it was heated by a helix and by a non-inductive current. Now, the 
only difference in the conditions was that in the second case the load was 
much greater than in the other. It became necessary, then, to examine 
this, and experiments were performed on all the wires with much greater 
loads than in the previous experiments. My results quite confirm those 
of Miss Noyes, and show that, under a load approaching the elastic limit, 
the decrease in Young’s modulus is uniform. 
In those experiments in which the temperature was determined by 
measuring the resistance of the wire, the method is perhaps open to 
criticism, and may seem to stand in need of justification. The temperature 
coefficient of resistance had been determined in the usual way in an oil- 
bath when the wire was unstretched, whereas in the experiments the 
resistance was determined under tension. Therefore the assumption is 
that no appreciable difference is produced in the resistance of the wire by 
the load. The experiment with which I am most familiar is that described 
by Kelvin,]- where he discusses the electrical resistance of a wire under 
tension. No exact quantitative results are given, but the effect is small, 
and, as the curves are wide apart and cut at a large angle, the assumption 
seems a legitimate one to make. 
The results are correct to a unit in the fourth significant figure, that 
is, the deviation of any individual reading of a set from the mean does not 
exceed a unit on either the one side or the other. 
When the wire was bare there was radiation, and consequently a tempera- 
ture gradient in the wire. To see if any change was produced when there 
* Phys. Rev., vol. iii., p. 452. 
t Math, and Phys. Papers , vol. ii. , p. 298. The fourth paper of the series on “The 
Electro-Dynamic Qualities of Metals.” 
