STRAIN ON THE PHYSICAL PROPERTIES OF MATTER. 
803 
Gustav Wiedemann, at the end of a very interesting paper on “ Torsion,”* discusses 
the question of the loss of energy experienced by a wire vibrating torsionally. He 
had by previous statical experiments proved—- 
(a.) That permanent torsion occurs even after the slightest temporary torsion. 
( b .) That when a wire is repeatedly twisted by a given torsional stress, first one way 
and then the other, the distance between the permanent positions of equilibrium on 
each side gradually diminishes to a minimum ; the period during which this diminu¬ 
tion is taking place is called “ the accommodation period.” 
(c.) That when the wire has been “accommodated” by stresses repeatedly 
acting in opposite directions, till the temporary t and permanent torsions dtT?« and 
respectively are constant, if the wire be again twisted by increasing stress in the 
direction of the last torsion, T p — P /; increases nearly in proportion to W p , where T p , 
P ; „ and W. p represent any temporary torsion, permanent torsion, and torsional stress 
respectively between 0 and T m , P,„, and W m . 
Wiedemann, from these and other results, feels himself justified in concluding that 
“ the hypothesis according to which the diminution of the amplitude is to be referred 
to an internal friction which for the entire course of the oscillations is a function of 
the velocity can no longer be maintained.” Neither does he consider that the elastic 
after-action will alone account for what is observed. The chief cause of the diminu¬ 
tion of the amplitude lies rather, according to him, in the shifting of the position of 
permanent torsion of the molecules at the end of each oscillation; what takes place at 
the alternate oscillations may be well described in his own words as follows “ In the 
absolute first position of rest of the wire, before any torsion in its molecules, we will 
draw axes parallel with the axis of the wire. Let the wire be next ‘ accommodated ’ 
by frequent vibrations to and fro, and at last be temporarily twisted in the positive 
direction, describing an angle of +«, while the molecules may be rotated so that the 
lower ends of their axes, looked at from the axis of the wire, describe an angle, -fa 
say, to the left. If the wire be slowly brought back into the permanent torsion- 
position -fb, the axes of the molecules will retain a portion +/5 of their rotation to 
the left. If the wire now receive an impulse in the positive direction, which again 
sends it to -fa, it will, according to the laws of perfect elasticity, swing back to -fb. 
If it now swing beyond this position further to the right, and if the molecules in their 
rotation had no friction to overcome, it would arrive at the position — a, while the 
axes of the molecules would be rotated to —a. Again, with perfect elasticity the 
wire would go back to the position —h, in which the molecules would be rotated — /3, 
and so forth. The to-and-fro motions of the wire between Ta and are perfectly 
elastic; therefore the performances of work in the swingings backward and forward 
* ‘Wiedemann’s Annalen,’ vol. 6, 1879, pp. 485-520; ‘ Phil. Mag.,’ Jan. and Feb., 1880. 
t Wiedemann employs the term “ temporary torsion ” so as to include not only the strain which 
disappears on the removal of the stress, but also the permanent torsion. I, on the contrary, will always 
denote the torsion which disappears on the removal of the torsional stress as the temporary torsion. 
5 k 2 
