192 Proceedings of the Royal Society of Edinburgh. [Sess. 
With another specimen, giving similar original values, and subjected to 
fifteen minutes’ fatigue, values were got : 
A. a = 1 B. a = 30 
n = B80 n—' 55 
6 = 225 6 = 96. 
This value of b seems to be distinctly nearer the original value. This 
was confirmed by an experiment where five minutes’ fatigue only was given. 
Here again all the constants were similar except the b of the first line, 
whose value was now 327. The value of b (169), obtained after ninety 
minutes’ torsion, was found to tend to regain the normal value after the 
wire had been allowed to stand for a few hours to recover from fatigue. 
After one hour’s recovery no appreciable difference was observed, but after 
one day’s recovery the values were for the first part of the line, 
a = 0 
n= 1-90 
6 = 266; 
and after two days’ recovery, 
a = 0 
w= T90 
6 = 371. 
With fatigue, then, the first value of b is the only one of the constants 
to be markedly changed. In the following experiments five minutes’ 
torsion was in each case given for ageing, and the change again produced 
in the first value of b is too great to be accounted for by the effect of such 
fatigue. In each case the length of the wire used was far greater than 
that of the portion actually tested. This portion was cut from the middle 
of the original length, and a suspension of one minute in the coil was in 
each case given. 
Again no change was found in the values of a or n, but the value of 
6 for large oscillations was found to vary considerably. The results were : 
Current. 
b of Line A. 
T Amp. 
588 
•2 Amp. 
728 
'3 Amp. 
946 
•4 Amp. 
519 
•5 Amp. 
748 
1*0 Amp. 
1200. 
Whilst no uniformity of value is got, a large increase on the whole is 
observed. 
