An error in the stability analysis for a nonstationary and inhomogeneous plasma in various previous papers is pointed out. The correct formula for the growth rate of the plasmon number is reobtained based on a geometric optics approximation.
