826 
MR. H. TOMLINSON ON THE INFLUENCE OF STRESS AND 
Experiment XIV. 
The same piano-steel wire as before. The temperature ranged from 9 o, 30 C. to 
I2°-38 C. 
Yibration-period in 
seconds. 
Total load on the 
wire in grammes. 
Couple in dyne 
centims. required to 
twist the wire 
through one radian, j 
4-303 
164 
5496 
16-658 
1208 
5713 
18-106 
5208 
5738 
It will be noticed that there is a comparatively rapid rise in the torsional elasticity 
as the load increases from 164 grammes to 1208 grammes, and a much less propor¬ 
tionate increase as the load rises still further. Now we find this phenomenon to be 
closely associated with the permanent torsion, which is, to a greater or less extent, 
always imparted during the process of wire-drawing. Permanent torsion produces 
decrease of torsional elasticity, but, as I have already shown,* * * § unannealed piano-steel, 
if permanently twisted by hand, begins to temporarily untwist t when loaded, the 
amount of untwist being, for small loads, very rapid, and consequently we might 
perhaps expect the elasticity to be increased by loading, rapidly at first, and after¬ 
wards much more slowly as the loads were increased in amount. Similarly piano- 
steel behaves when permanently twisted by wire-drawing; this is apparent from the 
next experiment. | 
Experiment XV. 
The unannealed piano-steel used in the previous experiments, and which had been, 
during a period of several weeks, frequently loaded and unloaded with weights 
reaching to upwards of 20 kilos., was allowed to rest for some time with only the bar 
of the vibrator § attached ; a pan whose mass was 388'8 grammes was then suspended 
* ‘ Phil. Trans.,’ Part I., 1883, p. 21. 
f This expression needs a little explanation. By the words “permanently twisted” is meant that 
the torsion will remain in the wire so long as we do not alter the load under the influence of which the 
wire was when the torsion was originally imparted, or, if we do alter the load, but afterwards restore it 
to the original amount, the twist originally imparted will be restored. 
X I have since examined the matter more closely, and find that, with all very hard-drawn wires, there 
is a slight difference between the torsional elasticity, as inferred from torsional vibrations, wheu different 
light loads are used. This is also the case with annealed metals which, after annealing, have suffered 
considerable permanent torsion. Whether the torsional elasticity increases or diminishes with the load 
depends apparently upon whether loading causes the wire to temporarily untwist or twist more. 
§ The mass of this bar was 164 grammes. 
