Abstract
The zero-temperature conductivity sigma of a disordered three-dimensional metal is elaborated by means of a diffusion model discussed recently by Kaveh and Mott (1981) in the context of logarithmic corrections to sigma for two-dimensional systems. It is argued that the contribution from diffusion is of importance when analysing the conductivity as a function of electron density. The resulting expression for the conductivity is compared with recent measurements for Si:P.