MECHANICAL TESTING OF IRON AND STEEL. 223 



more general law. This fact at once disposes of much of the 

 talk often indulged in during discussions on this subject 

 as to the value of high elastic limits when they are deter- 

 mined simply from a static tensile test ; and the importance 

 of the question is at once realised when we consider that 

 these so-called higher values are made the basis of a claim 

 for hio-her working stresses, even for cases where the loads 

 are alternating tension and compression. 



It is to be hoped, therefore, that this question of the 

 determination of the Natural Limits of Elasticity of different 

 qualities of iron and steel will be undertaken as a research 

 work by some competent observers having the requisite deli- 

 cate measuring apparatus at their disposal. Such a research if 

 carefully carried through would settle many important points 

 of the utmost practical value, and would at the same time 

 confirm experiments which, from their far-reaching con- 

 sequences, and from the fact that the deductions from them 

 depend on extremely small differences in measurements, 

 themselves very minute, much need confirmation. 



Carried out in a systematic manner, careful chemical 

 analysis being made of the material tested, we should obtain 

 data to make it possible to deduce the relations existing 

 between the values of the natural limits, and the propor- 

 tions of the various constituents of other elements present 

 in iron or steel, and probably be able to ascertain definitely 

 the best proportions for a material required for some given 

 purpose. It would also enable us to settle in a much more 

 rational manner the working stresses which could be allowed 

 in any given case ; this, a matter of the utmost consequence 

 in the case of parts of machines or structures subjected to 

 alternations of stresses, is equally of importance where there 

 are only variation of stresses or even only long-continued 

 statical loads, for we are still without information as to the 

 changes which may go on in the value of the Limit under 

 these conditions. 



That a steady change in the opinion of engineers in 

 regard to this question of working stress has been in 

 progress for the past few years is undoubted, but at present 

 we are still groping more or less in the dark, because though 



