' 
. 
Torsion of Lead Wire beyond the Elastic Limit. 533 
a fulerum U. At the other end of the lever is a pen V, 
which draws the curve on the drum W which revolves once 
in 24 hours. The length WU is about three times UM. 
On the surface of the water both in B and T, a layer of oil 
is placed to prevent evaporation. 
The curves drawn by the pen were found to be smooth 
during the first 24 hours or so, but after that, when the pull 
on the wire became very small, the test of accuracy was too 
severe and the curve is not to be depended upon further than 
to give the character of the phenomenon. Curves VIII. and 
IX. (Pl. XV.) are traced from the actual drawings of the pen. 
In the first case the original weight was 500 grams, and in the 
second 250 grams. The first experiment proceeded five days, 
when all but about 18 grams had been removed and more 
was still being removed. In the case of the initial weight of 
250 grams, after ten days all but 5 grams had been removed 
and the process had stopped. It is possible therefore that 
the whole is not removed within a finite time, and the 
equation before given fails when the time is greater than a 
few hours. One would expect the curve to be. asymptotic to 
the final weight removed. A curve of the type 
VY a log (pt +1), 
where W is the weight removed, fits VIII. very well during 
the first few hours, ann there is an increasing departure from 
it as the time increases. The same applies to curve IX. 
But if the equation be slightly altered to the form 
Spit ly 
W=a log eked 
where 4 is so small that it has little effect during the first few 
hours but increasing effect with time, then it fits through the 
whole range and satisfies the end condition, namely, that at 
an infinite time the slope shall be zero and the value of W 
somewhere near 500 grams. 
The equation 
6dt+ 1 
0:0237¢+1 
fits curve VIII. reasonably well. 
When t=0, W=0. 
Whent=x, W=205 log ‘ ae 1 ‘ 
==5 0) 
W =20 log 
