Manchester Metnoirs, Vol. xliii. (1899), iV^. 10. 3 



points of the extensometer screws were placed, thus 

 ensuring a correct attachment to the specimen. 



This extensometer, which is shewn in Fig. i, con- 

 tained no multiplying arrangement, but simply afforded a 

 means of reading easily the extensions after yield point, 

 whilst the load was increasing. 



It consisted of two tubes, the inner one of brass, coated 

 with white paper upon which was drawn longitudinally a 

 scale of inches and tenths, and an outer tube of glass, of 

 such a diameter as to slide freely over the inner tube 

 without too much side-play, and upon the inner surface of 

 the glass a vernier was attached, which enabled the 

 extensions to be read to the hundredth of an inch. 



After the diameter of the bar had been carefully 

 measured by a micrometer-gauge, the extensometer was 

 attached and the whole placed in the testing machine. 

 Upon the pressure being applied, readings of the extension 

 were taken when certain pre-arranged loads were reached, 

 and both load and extension were noted. 



Six bars were tested in this manner, the second how- 

 ever being rejected owing to the extensometer becoming 

 loose; with the remaining bars 24 points were obtained, 

 for which the logarithms of stress and strain have been 

 plotted, and are reproduced in Fig. 2. 



. The third bar of the series was carefully weighed 

 before testing and its specific gravity obtained. This 

 was found to be 8*9 17. After testing, a length was cut 

 out of the centre, the specific gravity of which was 8'905. 

 For the purposes of these experiments the density of the 

 bars was therefore assumed to remain constant. 



If P is the load, e the strain due to P, and ^„ the original 



area, then assuming constant density, 



p 

 stress on reduced section =— -(r + e)- 



