660 
Proceedings of the Royal Society of Edinburgh. [Sess. 
Steel Wire. 
With this wire the method of procedure was the same as with the soft 
iron. The modulus was first determined at the temperature of the room 
before any current had been passed through the wire, and the value was 
found to be 21 ’60 x 10 11 . On passing a weak current the modulus was 
lower than before any current was passed, but on increasing the current 
slightly the modulus rose in value. With a further increase in the current 
it continued also to rise until a maximum was attained at about 45° C., and 
then fell regularly as the temperature rose. 
When the current was diminished the modulus increased, and that, too, 
more rapidly than it had fallen with the increasing current. The maximum 
was reached at about 62° C., and this was the highest value throughout all 
the experiments on this wire. As the temperature continued to fall the 
modulus diminished, and came finally to a lower value than it had when 
the current was started. On allowing the wire to cool to the temperature 
of the room, the value was a little higher than it was initially. 
Since the modulus towards the end of the cycle did not have the same 
values as it had at these temperatures at the beginning, the cycle of opera- 
tions was repeated, and the values were found to be a little higher than in 
the first. This increase obtained all through, and there was also a slight 
increase in the value after the wire had cooled to the temperature of the 
room, compared with that at the corresponding stage of the first cycle. In 
all, four cycles had to be completed before the cyclically steady state was 
reached. When allowed to cool to the temperature of the room, the 
modulus had a higher value than in any of the previous determinations 
without a current. The effect, then, is to produce a permanent increase. 
In accordance with the plan on which the experiments were carried out, 
the wire was next loaded with 2‘2 kilos, that is, about 46'6 kilos per sq. mm., 
and heated by passing a current through it, with the results as shown 
in the graph. With this load on the wire the decrease is uniform, for the 
graph is a straight line. 
The same wire was then heated in the ordinary way in the double- 
walled tube, the graph being again a straight line. 
We see that the results for steel are very much the same as those for 
soft iron, the most important respect in which they differ being that with 
the decreasing current the soft iron has a lower modulus than with the in- 
creasing, whereas in steel it is at first higher with the decreasing current. 
Again, as with the soft iron, the temperature coefficient for ordinary heating 
