282 



REPORTS ON THE STATE OF SCIENCE. — 1910. 



Table II. 

 Elastic and Plastic Limits. 



(!) Well defined limits. 



(?) In these cases there were too few observed stresses to make accurate estimates 

 of the ' drops.' 



(b) 111 defined limits. 



accuracy. The ultimate shearing strengths, S, were obtained from the 

 torsion tests as explained below. In these cases, too, the unsteady- 

 conditions at the ends of the tests interfered with the accuracy. Never- 

 theless we find that the ratio S/T,. is fairly constant but rather less than 

 5 as required by Guest's law. The ratios S/T vary from 0-756 to 946 

 and seem to be of no value. 



Table I. also contains the estimated ultimate strengths estimated 

 from three formulae there given. The agreement between the estimated 

 and actual stresses is remarkably good in the case of T. The two other 

 formulae are of interest because the constants for T,. are approximately 

 twice as great as those for S. 



Possibly with an increased number of experiments the agreement may 

 prove to bo a closer one. It should be noted that the influence of nitrogen 



