SOME PECULIARITIES IN THE ELASTIC PROPERTIES OP CERTAIN SUB- 
STANCES. 
BY K. E. GUTHE. 
All solids deviate more or less from the theoretical perfectly elastic body 
and though many experiments have been undertaken in order to explain these 
deviations, we can hardly claim that any of the numerous theories proposed 
by the different investigators has yet cleared up the intricate relations involved. 
The best known of these deviations is the so-called elastic after effect. It 
was first discovered in silk fibers in 1835 by W. Weber and consists ip a lagging 
of the establishment or disappearance of a strain behind a distorting force. Thus 
if we twist a wire at one end while the other end is kept fixed, then upon re- 
leasing the wire there will be at first a rapid return towards the original shape, 
followed by a slower change in the same sense, extending sometimes over con- 
siderable time. The longer the original deformation has lasted the stronger will 
be the after effect. Of course the deformation is supposed to remain always 
below the value at which a permanent set occurs, so that after a sufficiently 
long time the return to the original condition is complete. Many experimental 
and theoretical investigations have led scientists to believe that elastic after 
effect may be considered as a kind of viscosity. It must, however, be remem- 
bered that it differs in its nature from the viscosity of fiuids because in the latter 
case there is no tendency to return to the original shape after the stress is re- 
moved. 
The elastic after effect is especially apparent in soft metal wires, while 
the same wires hardened show it to a much smaller degree. Quartz fibres, car- 
bon filaments and wires of hardened steel, phosphor bronze and platinum-irid- 
ium have vefy small elastic after effects. 
Another deviation from the demands of the theory has received the name 
“elastic fatigue.” Lord Kelvin^ observed that a wire which had been kept under 
torsional vibrations for some time has a larger period of vibration and a larger 
logarithmic decrement than a wire which has been at rest. J. O. Thompson® 
has later demonstrated that as long as the amplitude and temperature remain 
•constant there is no such change, even if the wire is kept vibrating for days. 
It is, however, a well known fact that when the amplitudes decrease from larger 
values to smaller the period, as well as the logarithmic decrement, decrease 
somewhat. So far only cases were known in which ttie variations of these 
quantities are quite small; in the course of some experiments with platinum- 
iridium wires which show almost no elastic after effect I found such remarkable 
1. Kelvin, Math, and Phys. Papers, III, p. 22. 
2. J. O. Thompson, Phys. Rev. VIII, p. 141 ^ 
( 147 ) 
