228 



IOWA ACADEMY OF SCIENCE 



large the fall in the range of amplitudes common to the two ex- 

 periments is more rapid than when the initial amplitude is 

 small. 



Figure 38 shows this for wire No. 4 when in state II. The 

 initial amplitudes were 1, 4, 7 and 10 degrees per unit length 

 respectively, for the curves 1, 2, 3 and 4. Theoretically the 





curves should be parallel but careful measurements prove that 

 they are not. There is a progressive change in the slopes of 

 the period-amplitude curves, the curves tending to become 

 steeper with the larger initial amplitudes. This is not a new 

 point but has been observed in other wires by Kelvin 8 , and by 

 Sieg 9 . We have reasons to suspect that the previous history 

 plays a large part in this effect. 



States II mid HI. The reasons for the peculiar conditions 

 of phosphor-bronze wires now became the object of research. 

 As stated above, curves II and III were the rule while those of 

 type! I were the exceptions. The very first tests showed all 



"Kelvin, Math, and Phys. papers, p. 22. 

 9 Loc. cit, p. 6. 



