044 Prof. Trouton and Mr, Rankine: Stretching and 
very accurately by the continual removal of weight. It was 
also necessary to have some means of removing weight with- 
out disturbing the mirror of the optical lever. The removal 
of weights by hand was found to make the image of the 
scale quite unreadable. To avoid this a tin pot containing 
water was used as the weight and the latter removed as 
required by means of a syphon. This worked very well, as 
there was no visible disturbance of the image when it was 
used. 
In order to eliminate temperature variations the optical 
lever was itself supported by similar lead wires hanging from 
the same beam. The legs of the lever were removed and an 
axle put through its centre, and it was suspended as shown in 
figs.3 and 4(pp.542,543). This arrangement also had the effect 
of eliminating any yielding there may have been in the beam 
from which the wires hung. The weight of the optical lever 
was only 60 grams, and this was supported by two wires, so 
that even supposing the rate of stretching was proportional 
to the weight to such a low limit, it would in any case be 
very small compared with the rate of stretching of one wire 
under a stress of 700 grams. 
Fig. 5.—Decrease of Stress when kept at Constant Length. 
— 
(RGITANS WEGAE) 
BPFLIED FORCE 
ie eae] 
ae 
Q 2 TIME (in hoursy 
The curve obtained is shown in fig. 5, and here, even more 
closely than in the first case, an equation of the form 
W=<a+hk log (¢+c) 
fits. The actual equation is 
W=713—159'1 log (¢+1), 
