STRAIN ON THE PHYSICAL PROPERTIES OF MATTER. 
831 
There was no perceptible permanent elongation produced by the loading, so that the 
increase of torsional elasticity, which is very apparent, must be due to molecular and 
not molar strain. Part of the strain is only sub-permanent and can be shaken out; 
when this is done there is an immediate loss of elasticity, which is regained if the wire 
be again loaded to the same extent as before, and subjected to the same previous 
treatment. 
Effect of Wire-drawing on the Internal Friction. 
We have seen that both permanent extension and permanent torsion are attended 
with permanent increase of internal friction. Now, in the process of wire-drawing 
there occurs both permanent extension and torsion, combined with lateral pressure. 
As a consequence, wire-drawing increases the internal friction in an astonishing 
manner. I shall have occasion to dwell more fully on the effects of annealing in a 
memoir which I hope will shortly follow this one, but the following Table will bear 
witness to the truth of the above statement:— 
Table. 
Name of metal. 
Logarithmic decrement 
due to internal friction 
in the hard-drawn 
metal. 
A. 
Logarithmic decrement 
due to internal friction 
in the annealed 
metal. 
B. 
Ratio A : B. 
Silver. 
•001540 
•0002970 
5-2 
Copper . 
•000526 
•0002622 
2-0 
Platinum .... 
•002239 
•0000766 
29-3 
Platinum-silver . 
•001792 
•0001166 
15-3 
Zinc. 
•018530 
•0032070 
5-8 
We thus see that annealing the hard-drawn metals reduced the internal friction to 
an amount which varied from one-half to one-thirtieth of the original. 
The Logarithmic Decrement due to Internal Friction is independent of the Length 
and Diameter of the Wire. 
Experiment XXI. 
The annealed copper wire of Experiment IV. was drawn to a diameter which was 
nearly one-half of the original and afterwards re-annealed. With the same length of 
wire in both cases, and with nearly the same preliminary treatment, the logarithmic 
decrement due to internal friction only was for the thicker wire '0002622 and for the 
thinner one '0002695.* The temperature of the wire was nearly the same in each case, 
* Great care must be used in an experiment of this kind to ensure uniformity of temperature during 
the period of experimenting, otherwise the thinner wire will show a greater internal friction than the 
thicker one (see Experiment VII.). 
