806 
MR. H. TOMLINSON ON THE INFLUENCE OF STRESS AND 
Three thermometers, one near the top of the long box, one in the middle of the 
latter, and one in the box S S, served to determine the mean temperature of the wire; 
these thermometers will be designated as the thermometers T 1} T 2 , and T 3 respectively. 
The thermometer T 3 was divided so as to read to one-tenth of a degree Centigrade, 
and had been tested recently at Kew; the thermometers T, and T 3 were carefully 
compared by myself with T 3 degree by degree, and were graduated in degrees 
Centigrade. 
A delicate aneroid barometer, placed on a level with the middle of the box S S, 
served to determine the pressure of the air in the box; the aneroid had been recently 
compared with a standard mercury barometer, and was graduated to one-hundredth of 
an inch. Readings of the three thermometers T l5 T 3 , and T 3 , and of the aneroid, were 
taken at regular intervals during the experiments. 
When all the adjustments had been completed, the wire, whether in the annealed 
or hard-drawn condition, was allowed to rest for a period of time varying, with the 
nature of the metal, from one day to several weeks, until there was no shifting of the 
position of torsional equilibrium and no appreciable “ running down” of the wire 
either before or after vibration. It was necessary to adopt this precaution inasmuch 
as, though great care had been used in adjusting the wires to avoid imparting 
permanent or rather sub-permanent torsion, such was frequently found to exist, and 
manifested itself by a gradual shifting with time of the position of torsional equili¬ 
brium. Now t , when a wire is either possessed of sub-permanent torsion or is “running 
down,” or even on the point of doing so, the loss of energy which occurs during 
torsional vibration is appreciably greater than vdien there are no such things. Even 
when there may be no apparent permanent torsion, or “running down,” it is necessary 
to allow the wire to rest for some time, and to cause it to undergo frequent vibrations 
before the internal friction assumes constancy. 
When it was necessary to excite vibrations a side-door of the box S S was opened, 
and slight impulses were given by means of a worsted thread to the bar V Y until 
the amplitude of the oscillations was of the required magnitude ; the door was then 
shut, and another door* at the bottom of the long box A A was opened, so that any 
accidental pendulous oscillations might be checked : finally this door was closed, and 
when the amplitude of the oscillations had become reduced to the requisite extent 
the observer commenced to write down the extreme limits of several consecutive 
vibrations, as shown by the numbers on the scale. At intervals the time of transit of 
the light across the middle point of the scale was noted. When the amplitudes 
decreased rapidly the observations were continued throughout the experiment; but 
when the decrement was small the observer frequently left the room till the ampli¬ 
tude was sufficiently reduced, when he again noted down the extreme limits of the 
same number of consecutive vibrations as before, and so on. In all cases the experi- 
