Experimental data for the stepped current and open‐circuit transient methods were fitted to one form of the modeling equations given by Dignam. A good representation of the data was obtained except that extreme linearity was found for vs. data for the peak field in stepped current experiments where some degree of nonlinearity was expected. Another question involves the role of electrolyte incorporation in causing the observed curvature of steady‐state vs. plots. It is shown that if a model is preferred in which the history effects are due to structural changes in the glassy oxide rather than to the special current‐driven polarization model of Dignam, then one can obtain equations equivalent to Dignam's by writing the activation energy for ionic transport as a function of a parameter describing the structural condition of the oxide, except that some relations between parameters which follow naturally from Dignam's model do not then appear.