655 
1907-8.] Young’s Modulus under an Electric Current. 
As I have already stated, in consequence of the results obtained by 
Miss Noyes, it was deemed necessary to examine the behaviour of the wire 
when it was subjected to a greater load. In the first experiment the load 
was small, viz. *8 kilo, which was equal to nearly 1462 kilos per sq. mm. 
The load was then increased to 2 kilos, that is, 36'54 kilos per sq. mm. In 
this case my results were quite in agreement with those of Miss Noyes, 
and these results are shown in the graph. 
The wire was next heated in the ordinary way, and readings taken at 
various temperatures. These results are shown on the graph, and it will 
be seen that by this method of heating the modulus undergoes a uniform 
decrease. Usually, when Young’s modulus has been determined at different 
temperatures, readings were taken at the temperature of the room and 
then at about 100° C., no intermediate temperatures being used. This was 
the case in the experiments described in Shakespear’s * paper on Young’s 
modulus, in which the extension of the wire is measured by the method 
of interference. The same temperatures were employed by Gray, Blyth, 
and Dunlop in their paper already referred to. The only investigations 
with which I am familiar in which intermediate temperatures are 
employed are those of Miss Noyes in her two papers. In her experiments 
the wire was heated both by a magnetising coil and by a non-inductive 
coil, and the graph of the results was always a straight line. My results 
are the same as hers, the only difference being in the value of the co- 
efficient. That, however, is a minor point, for it varies widely in different 
specimens. In three determinations of the temperature coefficient for soft 
iron wires in Gray, Blyth, and Dunlop’s paper there was a difference of 
33 per cent, between the extreme values. A still greater difference was 
found by them in copper, as the coefficient for electro hard-drawn copper 
was fully three times larger than that for commercial copper. It seems, 
then, to be beyond doubt that, when a wire is heated in the ordinary way, 
Young’s modulus undergoes a uniform decrease. 
When the wire is heated by the current and carries the greater load, 
the modulus is lower than when it is heated in the ordinary way, but the 
coefficient is smaller, so that the two lines converge, and if produced they 
would intersect at about 215° C. Now, the two graphs may not be 
absolutely straight lines, so that a slight alteration in their rates of fall 
may produce such an effect as to make them after meeting coincide with 
one another. Further, the graph for electric heating with a load of '8 kilo 
slopes down to these two lines in such a way as would seem to make them 
* Phil. Mag., 1899, p. 539. 
