445 
1910-11.] An Apparatus for inducing Fatigue in Wires. 
and the pendulum set in motion, the end of the wire at A is twisted through 
an angle of 360° at each half swing, and thus fatigue is imparted to any 
desired degree in the wire by noting the time of oscillation. In the steel 
spindle on which the pendulum rests, and at the other side from the toothed 
wheels R, is a small groove into which the thin iron bar H rests, and this 
bar is movable about a fulcrum I. As the oscillator moves to its maximum 
outward swing from the vertical position of rest, the end H is raised, and 
consequently C lowered, and in the inward swing of the pendulum this is 
repeated, so that a wire clamped between C and D receives two pulls 
during a complete oscillation. To measure the power applied to the end of 
the wire per pull the weight hung on the end H and the ratio of the arms 
of the lever are noted, whilst the total number of pulls is found by 
doubling the number of oscillations in the given time. The wire between 
C and D can be kept fully stretched by means of screws above and below 
the projecting arm to which D is clamped. 
The pendulum having been set oscillating, and the necessary resistance 
introduced by means of the rheostat, it was generally found that the range 
of the oscillation kept quite uniform, and the apparatus could be left for 
any desired time. On one occasion it was left going overnight, but this 
had worn down the teeth of the brass wheels so far that unsteady 
oscillation of A was being produced. 
Results to Date. 
Brass wire, similar to that used in the temperature experiments, is the 
only wire yet tested, and experiments included the finding of the effects on 
the constants n, a, and b of the equation 
y n (x + a) = b, 
after subjecting the wire (1) to rotational strain in the original, unheated 
state ; (2) to rotational strain after the wire had been raised to a red heat 
and cooled, i.e., after the points obtained by plotting log (a? + a) against log y 
could not be fitted into one straight line. It was seen, however, that these 
points could, with the proper value of a , lie along two straight lines; 
(3) to extensional strain in the unheated state; and (4) to extensional 
strain after raising to a red heat and then allowing to cool. It was 
shown in a paper accompanying this, that, since n log y + log (x + a) = log b, 
then if log y be plotted against log (cc + a), the points will, if the empirical 
law hold, fall into a straight line, provided the proper value of a be inserted, 
and the tangent of the angle this line makes with the log y axis will then 
be = n ; b can then be calculated. 
