The Townsend–Huxley transverse diffusion experiment has been modelled
using kinetic theory, with the aim of explaining the successes and
shortcomings of some current ratio formulae that are based on hydrodynamic
theory. These are shown to be essentially identical in form, apart from the
estimated magnitudes of the boundary layer corrections to the second and
fourth transverse position moments of the current density. The main reason for
the shortcomings is shown to be the inability of hydrodynamic theory to
describe the relaxation of the input electron energy distribution towards the
steady state free space form. As a byproduct of the studies, exact
eigensolutions have been found for the associated ‘infinite slab’
problem, using the constant collision frequency model with the two-term
approximation.