Manchester Memoirs, Vol. xliii. (1899) No. 10. 5 



If the formula /= Ce"" represents the variation of actual 

 stress with strain, then 



\ +e • 



will give the corresponding stress per unit of original 

 area. 



By differentiation, the maximum value of this stress 



will occur when 



n 



e = ' 



\ - n 



For annealed copper this equals I'loi when- ;« is "524, 



or 110%, whilst the corresponding value for iron and steel 



(^ = •25) is 331 %• 



The maximum stress for annealed copper should 

 therefore be i8'98 tons per square inch of original area. 



On testing a bar of the annealed copper cut from the 

 same rod as the previous ones, the stress at yield point 

 was found to be 2'23 tons per square inch, whilst the 

 maximum stress was only I5"5 tons per square inch, with 

 an elongation (by the formula) of 30*3 %. 



In order to discover if the speed of the test could 

 account for this difference between the theoretical and 

 actual maximum stresses, a second bar, which was annealed 

 in the Laboratory, was tested very slowly, when it was 

 found that the maximum stress only amounted to 1379 

 tons per square inch, whilst the elongation had risen to 

 49 % (measured). 



This is only in accordance with the view expressed by 

 Professor Ewing,* that metals such as copper show to 

 greater advantage as regards breaking strength and less 

 advantage as regards elongation the greater the speed of 

 the test. 



Further, as the first of these two tests was made at 

 approximately the same speed as those from which the 



* Proc. R. S., vol. XXX., p. 510. 



