STRAIN ON THE PHYSICAL PROPERTIES OF MATTER. 
833 
platinum, silver, steel, iron, brass, copper, aluminium, zinc, platinum-silver, and 
German-silver, with arcs from rest to rest ranging up to 700 scale-divisions, and lead 
and tin, for arcs ranging up to 200 scale-divisions, showed no sign of the phenomenon. 
With nickel, however, the case was different, and 1 met with most marked signs of 
“ fatigue of elasticity,” even for such small molecular displacements as attended 
arcs ranging from 300 to 700 scale-divisions. 
Experiment XXIII. 
A nickel wire, unannealed, 602 centims., and ‘018622 square centim. in section, 
after having been tested for several days with three different vibrators, was finally 
tried with the most massive of the three.* 
Logarithmic decrement 
due to internal friction. 
Vibration-period in 
seconds. 
Remarks. 
•007058 
6-010 
Immediately after adjustment. 
After ten minxxtes’ rest. 
•006623 
•006225 
After 14 hours’ further rest. 
•005964 
After 2 hours’ further rest. 
•005585 
After 3f hours’ further rest. 
•005522 
5-960 
After 16 hours’ further rest. 
•005644 
. . 
Started again immediately. 
•005689 
Ditto. 
•006023 
An arc of 700 scale-divisions 
maintained for 2 minutes, and 
theu allowed to subside. 
•006427 
An arc of 700 scale-divisions 
maintained for 30 minutes, and 
then allowed to subside. 
In all the above trials the initial arc from which the logarithmic decrements were 
computed was about 600 scale-divisions, and the final arc about 150 scale-divisions. 
I next give the wdiole of the observations which were taken in the last trial from 
the initial to the final arc. 
Range of arc. 
Logarithmic decre¬ 
ment due to internal 
friction. 
605-2-516-0 
'00674 
516'0^40-8 
•00665 
440-8-372-0 
■00718 
372-0-317-2 
•00673 
317-2-270-0 
•00680 
270-0-230-8 
•00663 
230-8-201’5 
•00570 
201-5-174-0 
•00619 
174-0-152-5 
•00553 
152-5-135-5 
•00495 
MDCCOLXXXVI. 
* The total mass of this vibrator was 3666 grammes. 
5 o 
