MECHANICAL TESTING OF IRON AND STEEL. 217 



they can do to keep up the regular teaching work of the 

 laboratories. 



The field covered by the mechanical testing of iron and 

 steel is too wide to be fully dealt with in this article ; it is 

 proposed therefore to confine attention to some of the more 

 important points, to briefly state the extent of our present 

 knowledge on them and to indicate the lines upon which it 

 is desirable that further advance should be made. 



Forms of Test Bars. It is certainly a matter of great 

 regret that some system of uniform proportions for speci- 

 mens for testing has not been adopted. As the matter now 

 stands it is often extremely difficult to compare the results 

 obtained by different experimenters, especially in the case 

 of tensile tests when dealing with what is usually termed 

 the ultimate extension. 



In hard materials this difficulty does not arise, because 

 after the limit of elasticity has been passed there does not 

 occur to any extent that remarkable change in the molecular 

 condition of the bar which is so pronounced in the ductile 

 irons and steels, the bar in fact does not pass into the 

 semi-plastic condition. 



With ductile materials, however, as soon as the Yield 

 Point has been reached the material passes into the semi- 

 plastic condition, and just before rupture occurs we are con- 

 fronted with a condition which most materially alters the 

 ultimate extension. Up to the point in the test at which 

 this occurs, it is when the maximum load which can be 

 carried is reached, the extension has been fairly uniform 

 all along the bar ; if the bar broke at this stage the per- 

 centage extensions of all bars would no doubt be very 

 strictly comparable, the extension being then proportional 

 only to the length of the original bar for material of the 

 same quality. But as soon as the maximum load is 

 reached the extension becomes local, and the specimen 

 draws down at the point at which fracture finally occurs ; 

 occasionally it draws down also at some other point to a 

 certain extent. This local extension being confined to a 

 very short length increases the ultimate extension of a short 

 test piece proportionately much more than a long one. 



