492 



RKPORTS ON THE STATE OF SCIENCE. — 1919. 



only slightly outside the ellips?s which they touch at H, it follows that 

 this and other evidence that supports de Saint Venant's hypothesis may 

 be regarded as giving almost equal support to the hypothesis of constant 

 strain-energy. When the two principal stresses are more nearly equal, 

 as in rotating discs, de Saint Venant's hypothesis overestimates the elastic- 

 limit and cannot be regarded as reliable. 



Tests by Batson,^ on two mild-steel tubes, are also represented in the 

 same diagram. In these tubes, which were tested in torsion and in tension, 

 the ratio between the elastic-limits is somewhat difierent from that between 

 the yield-points. This may have been due to the comparatively thick 

 walls of the tubes, the ratio between the internal and external diameters 

 being 0-72. These results were described as supporting Guest's hypothesis, 

 but are more closely in agreement with the hypothesis of constant limiting 

 strain-energy. 



Fig. 19. 



Figure 19 shows the portion of the diagram relating to combinations 

 of pull and push ; particularly with equal intensities, producing simple 

 shear. Three experiments, described in detail by Mason,'' are represented 

 by points close to the ellipses ; and a number of other experiments on steel 

 tubes, more briefly mentioned in the same article, give points between the 

 ellipses and the straight line HGr. A number of points represent other 

 experiments on steel tubes in torsion and pull, by Turner.^ The ratio 

 between the two elastic limits varies widely in the different specimens 

 tested ; but analysing the observations by the theory of probabilities — 

 as applied in problems of external ballistics — Sears ^ showed that the 



6 R. G. C. Batson, Proc. Inst.Mech.Eng., 1917. 

 ' Wm. Mason, Engineering^DecemheT 24, 1909. 

 8 Professor Turner, Engineering, February 1909. . 

 8 J. E. Sears, Engineering, February 19, 1909. 



