154 
IOWA ACADEMY OP SCIENCE 
Having thought at first that the property of iridium to absorb oxygen might 
be the reason for the observed anomalies, I experimented with the same wire 
under different pressures. The changes in the period and decrement were as 
large as before, even when the pressure was only 2.5 cm. A further decrease in 
pressure could not be obtained because the large brass case was not perfectly 
airtight. 
The observations were repeated with a different moment of inertia suspended 
from the wire. The time of vibration was now only 7.274 sec. for the larger 
amplitude and decreased to 7.105 with the smaller amplitude. But most sur- 
prising of all was the fact that values of the logarithmic decrement when 
plotted as a function of the amplitude fell closely upon the same curve which 
was obtained with the larger period, and seems, therefore, to be independent of 
the velocity with which the system moves. 
It made no difference whatever with how large an amplitude I started, 
whether or not the wire had been kept swinging for some time, and if it was 
swinging under atmospheric or decreased pressure; for the same amplitude al- 
ways the same large logarithmic decrement was found. 
The phenomenon which this wire exhibits in such an exceptional degree has 
usually been considered as closely connected with elastic after effect, but the 
fact that the wire shows almost no after effect when tested by the usual method 
seems to indicate that there is little or no connection between the two. 
My first thought was that the wire did not follow Hooke’s law and I therefore 
connected one of its ends rigidly to a steel wire of equal length and at the con- 
necting point a small mirror was sealed to the wires. The lower end of the steel 
w'ire was securely fastened and the upper end of the platinum-iridium wire 
clamped in a torsion head. It was found that equal deflections of the mirror 
were produced by equal angular displacements of the torsion head. This shows 
that the wire followed Hooke’s law within the limits of the experiment. 
I am unable to explain at the present time the results obtained with the 
platinum-iridium wire, but hope to continue the investigation with wires of 
different length and diameter and possibly also with alloys of platinum with 
other rare metals. 
While the elastic after effect may have some influence upon the rate of dis- 
sipation of energy in a wire in torsional vibrations, there can be no simple re- 
lation between the after effect and the logarithmic decrement, even in the case 
of ordinary wires, because frequently wires with smallest after effect show very 
large values for the decrement. I have shown this to be true in all cases. 
Drawn wires have a considerably larger logarithmic decrement than softer wires. 
I have made a large number of experiments on this point, using first hard 
drawn wires and repeating the experiments after the wires had been an- 
nealed by drawing them through the flame of a Bunsen burner. In every case 
the soft wires show the smaller decrement and the larger moment of torsion. 
Thus copper wire, hard drawn, had a period of 5.780 sec. and a logarithmic 
decrement of 0.0057. After annealing its time of vibration with the same mo- 
ment of inertia and at the same temperature was 5.702, decreasing with ampli- 
tude to 5.692, and the decrement had decreased to 0.0022, more than one-half. 
New light has recently been thrown upon the physical process taking place 
during the drawing of a wire. Beilby® arrives by very interesting microscopic 
3. Beilby, Nature, 76, p. 572, 1907. 
