In a recent paper Hansen and Goertz (hereafter HG) [Phys. Fluids B 4,
2713 (1992)] considered the coupling between fast and Alfven modes in
a cold plasma containing a uniform magnetic field (B0x) extending betw
een two perfectly reflecting plane boundaries at constant x. The equil
ibrium medium is invariant in only one direction (y), and (importantly
) the density may vary along the equilibrium field lines, rho0(x,z). H
G sought solutions of the coupled governing PDEs (partial differential
equations) for linear perturbations of the form exp i(k(y)y-omegat).
The solution has been studied previously [Planet. Space Sci. 22, 483 (
1974); J. Geophys. Res. 79, 1024 (1974)] in the case when rho0 does no
t vary along the background field lines, when each Fourier mode in x d
ecouples from the others and may be considered separately-reducing the
problem to an ODE (ordinary differential equation). In this case a lo
garithmic singularity exists at the resonant field line where omega2 =
k(x)2V(A)2(z), V(A) being the Alfven speed (V(A)2 = B-0(2)/4pirho0).
HG claim the introduction of density variation along the equilibrium f
ield causes the modes in x to become coupled resulting in the singular
ODE solution becoming a nonsingular solution in the PDE case. If this
conclusion is true it is of great importance for researchers in many
areas such as solar corona and laboratory plasma heating, and magnetos
pheric pulsations. Indeed, it suggests that a large portion of the exi
sting literature in these fields is wrong. Clearly it is important to
decide whether the calculation of HG is correct or not. In this Commen
t the equations they set up are analyzed and are solved in a different
fashion to HG. The solution found is different from that of HG and in
agreement with the existing body of literature. Some sources of error
in HG's analysis are pointed out.