Abstract
The methods of analysing thermally activated dislocation glide using applied stress τ as an independent variable, (τ, T) analysis, or the effective local stress τ∗ as the variable, (τ∗, T) analysis, are compared. Each scheme is self-consistent and the two are mutually compatible. The (τ, T) analysis refers to a real thermo-dynamic system, the entire crystal. The (τ∗, T) analysis refers not to a real sub-system within this crystal but to a hypothetical model system, constructed on the basis of linear elasticity theory. It may be derived directly from a definition of the model system. A variety of averaging processes must be invoked in order to relate the equations of motion of an individual dislocation segment to the macroscopic deformation parameters of a crystal.