324 
IOWA ACADEMY OE SCIENCE 
effect of the drawing is clearly brought out. It is of especial 
interest to note that the wire that showed the greatest deviation 
from the assumed law of elasticity, No. 2, is of the size .114 
mm., most usually employed in the manufacture of galvanometer 
suspensions. The apparatus femployed will not be described, as 
both it and the method have been fully detailed in the former 
volumes of these Proceedings, already referred to. 
A New Ejfect .- — The curves in figure 23 represent the results 
that were obtained from these wires by using them just as they 
came from the draw plates, of the makers. After numerous ex- 
periments were made on the points above discussed,_ ff^ de- 
termined to see what effects would arise from various treatment 
of the wires. The two treatments that were employed were heat- 
ing almost to redness, and long continued vibrations. After 
these preliminary treatments were employed, an entirely new 
effect was discovered. This effect is to be noted in figure 24. 
Here as before the curve (marked with crosses) indicates the 
relation betwen period and amplitude. It will be noted that 
the period decreases with amplitude just as in the previous ex- 
periments, with the important exception that at an amplitude 
of about 4° per cm. of length, the curve takes a sudden rise, 
and from this amplitude downward it continues to rise. This 
seems to be an entirely new effect. The other curve in figure 
24, represented by the circles, is for the same wire except that 
in this case the period is platted against vibration number. It 
is only another way of illustrating the same point. 
To make sure that this new effect did not come about as a 
result of differences in the original amplitude of the displace- 
ment of the supported weight, it was determined to let the wire 
rest for some days, and then to start it in vibration with a small 
amplitude, carrying out the experiment as usual, and then in- 
creasing the initial amplitude in successive experiments. The 
results of these experiments are graphically shown in figure 25. 
The curves would be so near alike that only the observed points 
are represented, the curves not being drawn. It is evident that 
the form of curve is practically the same, regardless of the in- 
itial amplitude of the vibration, at least within the limits of our 
experiments. The curve is perhaps clear enough to entail no 
additional explanation. 
