Abstract
As to the yielding of materials the maximum shear stress theory is generally believed to be correct. But the author has a quite different idea. It is the same that the yielding of mild steel is due to shear, but it does not obey the maximum stress theory. It is a problem of stability. And so the maximum stress at yield point differs when the stress distribution is not similar. By the experiments of torsion of hollow bars this fact is ascertained. By the author's theory the moment at yield point by torsion must be proportional to the moment of area of cross section, and this fact agrees very well with experiments.
