In the 1st Report, the author presented a phenomenological fracture criterion of metals explaining the polyaxial fracture test results of various kinds of metals. It would result from the unknown stress and strain distributions together with the obscure fracture criterion that the problem or practical notch brittleness remains unexplained quantitatively. In this sense, tensile fracture tests of mild steel round bars having various grades of notch sharpnesses were urformed, and stress and strain distributions in the minimum section were analysed approximately with their average fracture values measured precisely by electric strain gauges. Since the plastic stress distribution has not such sharp stress gradient as the elastic one, it would be reasonable to apply the fracture criterion to each point on the fracture section. Consequently, basing on the idea that there exist the shear and tensile fracture modes, various fracture behaviors were able to be well explained and a view as to the notch brittleness was given.