148 
IOWA ACADEMY OP SCIENCE 
deviations from tiie behaviour of ordinary wires that I decided to investigate 
the matter further. 
The wires which carried at their lower end cylinders of Tobin bronze were 
suspended from the top and inside of a large hollow brass cylinder having a 
glass window at its lower end. The whole brass case could be rotated around a 
central pivot and thus the suspended cylinders set in vibrations. The glass 
window allowed the observation of a small mirror attached to the lower end of 
the vibrating system. The period was determined by chronographic records 
and the amplitudes by a telescope and scale placed at a distance of 100 centi- 
meters from the mirror. 
A platinum-iridium wire containing 25 per cent of iridium, when hardened 
by drawing a period of 6.7766 sec. with an amplitude of 5°, which decreased to 
6.7745 sec. with an amplitude of 0.®5, while its logarithmic decrement decreased 
between the same amplitudes from 0.0012 to 0.00091. A still more surprising 
behaviour was found in a 40 per cent platinium-iridium wire whose period of 
vibration changed through the same range of amplitude from 12.060 sec. to 
11.890 sec., while its decrement decreased from the enormous value of 0.012 to 
0.0030. * 
The period, as well as the logarithmic decrement, are very nearly proportional 
to the amplitude, though the increase is slightly smaller than that of the ampli- 
tude. The decrease of amplitude between successive swings is very nearly pro- 
portional to the square of the preceding amplitude. Since such enormous varia- 
tions have never before been observed in other wires, a great many observations 
were made, but in every case the same effect was found. In the following 
tables I have picked out at random a few of them and the values are plotted in 
figures 1 to 3. 
