Hopping transport in a magnetic field: Kadanoff-Baym-Keldysh approach and magnetoconductivity - Comment on

In a recent paper [Phys. Rev. B 51, 14 020 (1995)] Bleibaum, Bottger, Bryksin, and Schulz have reinvestigated the temperature dependence of the variable range hopping conductance. Their finding that, according to the effective-medium theory for hopping in d spatial dimensions, the de conductivity should follow sigma(T)=sigma(0)exp{-(T-0/T)(2/(d-2))} is shown to be invalid, caused by an improper averaging of the site energies. If properly averaged the conventional Mott law sigma(T)=sigma(0)exp{-(T-0/T)(1/(d+1))} is recovered. A scaling argument is given that supports the latter result.