Abstract
The conventional approach to the interpretation of TSD data is shown to be inconsistent with experimental evidence in that it rigidly links the theory of TSD with the ideal isothermal Debye response. It is proposed that the correct analysis must introduce the concept of thermal relaxation under time-variable temperature, which is quite separate from the familiar isothermal relaxation and is, in most situations, very different in form. In the absence of positive experimental guidance, it is proposed that the thermal relaxation function should be assumed to be exponential in time and that it is a strong force dominating the weak isothermal relaxation.