433 
1910 - 11 .] Energy in Torsionally Oscillating Wires. 
show no change from those for wire heated to 535° C., and quickly withdrawn 
from the furnace. On that occasion the values got were a = 4, n = 2*30 and 
2*60. An experiment when the wire was annealed from 485° C. also showed 
no change. 
Aluminium Wire. 
In the experiments, wire of diameter *95 mm. and of length 12 inches 
was used. In observing the readings on oscillating the original wire, it 
1*3 1-4 1-5 1*6 1-7 1*8 1*9 0 *1 '2 ‘3 *4 *5 *6 *7 *8 *9 
was noticed that after about twenty oscillations the relative decrease 
per swing in scale reading became larger, and after this point the oscilla- 
tions seemed to die down in a different manner. This was seen to be the 
case on plotting log (x-\ -a) against log y. No value of a could be got to 
cause all the points plotted to lie on a single straight line; with a = 0 a 
straight line was got for the first ten points (see diagram 5), but the others 
lay along a curve concave to the origin. The break seemed to be quite 
sudden. Higher and higher values of a were taken, but it was not until a 
vol. xxxi. 28 
