[coKER & m'kergow] THERMAL CHANGE TO TENSION 



TABLE IL 



11 



The relation between strain and thermal change is apparently not 

 a linear one, for the galvanometer readings are not proportional to the 

 load; this is seen very clearly in Fig. 2, vàiere the stress strain curve I 

 is very approximately linear up to a stress of 23,000 pounds applied 

 uniformly in 210 seconds. If, however, the curve II, showing the 

 thermal change, be corrected for radiation and conduction, the result- 

 ing curve III gives a linear relation for approximately the same range 

 of stress. In the actual observations a marked heating effect was 

 noticed about two seconds before the yield point was reached, corre- 

 sponding to a small difference of about 200 pounds of stress. It was 

 therefore inferred that the thermal change and strain are proportional 

 to the stress through about the same range. Other tension experiments 

 confirmed this inference. 



