Using the experimental results given in Part I as a guide, a linear model is developed in this Part which describes the interaction of a lattice defect and a dislocation. It is supposed in this model that the defect is free to diffuse between existing (or in‐grown) traps such as jogs or dislocation nodal points, and that if the defect is trapped, dislocation climb follows. From a comparison of the results given in Part I and the predictions of the model, values for the ``pipeline'' diffusivity and the strength of the binding of the defects to the in‐grown nodal points may be obtained. The numbers so obtained are consistent with values evolved from other techniques for studying atom motions. An estimate for diffusional jog‐pair energies is also given.