80 Mr. J. J. Guest on the Strength of 



under these conditions, experiments upon this part of the 

 range were deferred. 



21. Method of checking the Isotropism of the Material as 

 regards Yield-point Stresses. — This, however, is not the only 

 advantage of the internal pressure and tension tests, for by 

 their use we can check the isotropism of the material within 

 the range + ^p, p, by having the greater of the stresses first 

 parallel to the axis and then circumferential : to do this 

 within the range 0, +ip it would be necessary to use a com- 

 pressive load and internal pressure, as the longitudinal stress 

 due to the internal pressure must be added to the longitudinal 

 stress due to the externally applied load. 



In addition, tests under the combination of internal pressure 

 and torsion permit the principal axes of the stress to be 

 inclined at varying angles to the generating lines of the 

 specimen, and also, if a practical amount of isotropism has 

 already been demonstrated by the internal pressure and 

 tension tests, afford a check upon the general accuracy of the 

 work by comparing the results of these tests with the results 

 of experiments involving similar stresses differently applied. 



Torsion tests on thin tubes, it may be noticed, cause a 

 tendency in the wall to buckle, while the internal-pressure 

 and tension tests do not. No buckling occurred in the 

 experiments. 



Although combined stresses may be produced by simpler 

 methods, such as the loading of a cranked shaft or of a 

 helical spring, such methods, since a small portion only of 

 the material is carried to the yield-point condition, are hardly 

 likely to furnish decisive results. It is essential for definition, 

 both because of the elastic-limit effect and because of the small- 

 ness of the elastic strain, that the whole of the material 

 directly under observation should be similarly stressed. The 

 simplest methods of attaining this appeared to me to be those 

 indicated above, namely, the application of axial loads, 

 torques and internal pressures, singly or in pairs, to thin- 

 walled circular tubes. As the production of thin- walled 

 specimens otherwise than by drawing would be difficult, I 

 determined to use seamless drawn tubing as specimens, and 

 to remove the state of set left by the operation of drawing 

 by annealing. 



22. Objections to Tubes on account of want of Isotropy. — It has 

 frequently been urged that no such material as wires, tubing, 

 or sheet-metal can be regarded as isotropic ; but I think 

 such criticism may fairly be met by the consideration that in 

 the production of ductile materials, operations producing sets 

 and internal stresses, afterwards removed by annealing, are 



