8 Wilson, Stress and Strain in Copper Bars. 



Each wire, after fixing, was stretched by loading the 

 hanger until permanent set took place, and then carefully 

 annealed whilst in position by means of a blowpipe. 



This stretching and annealing was repeated until the 

 wire was apparently free from kinks and bends, after 

 which the length and diameter were measured and the 

 test commenced. 



The increment of load varied from Jib. to 2lbs., and, as 

 each successive load was applied to the hanger, the wire 

 extended, and when the extension had apparently ceased, 

 the reading was taken. As the tests each occupied some 

 hours, any change in the temperature of the laboratory 

 was noted and allowed for in reducing the results. 



Again, as it was impossible to measure directly the 

 extension of the wire due to the weight of the hanger, 

 this extension was estimated by considering the elastic 

 portion of the wire and exterpolating accordingly. 



After the test was concluded the wire was re-annealed 

 and stretched several times, and then re-tested in the above 

 manner. One wire was tested five times in this way, 

 altogether corresponding to 8 stretchings. 



The results shown in the form of the logarithmic 

 stress-strain curve in Fig. j are those of the last wire 

 tested, and they may be taken as characteristic of the 

 other wires which were examined. 



The curves there shown have the strains for abscissae 

 and the quantity P{i+e) for ordinates. 



To obtain the actual stress it is necessary to divide 

 by the original area of the cross section of the wire. To 

 measure this area for each additional load would be 

 difficult, hence the assumption of constant density was 

 again made, and the areas required were calculated from 

 the data afforded by the original volume and the change 

 in length. In order to show that this assumption was 



