At present the relevance of percolative transport to traditional glasses is being established. The electronic glasses represent (relatively) well defined systems in which the effects of e.g. 'interactions' and 'disorder' may be isolated, and for which the percolative aspects of transport are relatively well understood. Thus these systems (spatially random, SR. and variablerange hopping, YRH) can be useful as models for glassy conduction processes. However, some inconsistencies in the published theoretical descriptions of such system exist, namely regarding the placement of lhe critical (scaling, of peak) frequency and the relationship of the frequency dependence of the AC conductivity to critical exponents from percolation theory. These questions are clarified here.'Furlher conclusions about the nature of the scaling relationships between the AC and DC conductivities are drawn. The role of stochasticity, or randomness in the definition of the scaling frequency, is emphasized.
Transport in glasses: theoretical basisThe questions of the appropriate form of glass transport equations [ I d ] , as well as of the best approximations to their solution 17-10] have generated enormous controversy. The term 'glass' is itself riddled with controversy [ll-151. Here the term is used in the sense of either disordered solids or liquids with continuous distributions of transition rates. The question of when a system is sufficiently disordered to be called a glass is left to other publications [16-181.Two fundamental perspectives exist regarding transport in disordered systems. One starts from the assumed validity of 'percolative' transport 11-33. In such a picture, space, while isotropic in the mean, is strongly inhomogeneous. Thus only certain portions of the system can be utilized for transport upon application of e.g. an electric field [Z] (the description of viscous flow is somewhat more complex [NI), but temporal constancy is,implied, i.e. the same transport path will be utilized in successive applications of an electric field. This is to be contrasted with diffusive transport, in which spatial homogeneity is guaranteed, but for which the exact path of transport is never repeated. Traditional glasses and/oq viscous liquids obviously cannot be rigidly classified into either 'of these two categories. Clearly, as the temperature is dropped, the transport becomes more nearly percolative in character.There is a growing suspicion [1,3,19] that the cross-over occurs above the traditional glass transition, in' particular at or near the mode-coupling temperature, Tc.While the question of the character of'transport'in glasses generally has not been re+olved, three facts about electronic glasses are clear., Electronic glasses exemplify
