Jicamarca radar backscatter maps were made during four consecutive nights in March 1979. Two of these maps displayed single towering plumes extending to nearly 1000-km altitude. On a third night, discussed in detail here, six plumes were generated in clear association with a nearly sinusoidal oscillation of the height of the bottomside of the F layer. The vertical amplitude of the oscillation was several hundred kilometers, and the period about 100 minutes. The plumes were generated either when the bottomside of the F layer was at the highest altitude or in the descending phase of the motion. Families of curves are presented which correspond to the solution of the dispersion relation for gravity waves capable of initiating the observed bottomside oscillations via the spatial resonance mechanism. We conclude that the solutions thus derived are reasonable and present a criterion for how well matched the gravity wave phase velocity and plasma drift have to be to produce a given perturbation in the ionization density. This criterion indicates that although initiation by a gravity wave seems likely, the gravity wave interaction cannot yield the large displacements observed without further amplification by the Rayleigh-Taylor instability. Finally, we show that the preferential generation of plumes during the descending phase of the F layer height oscillation can be explained by a generalized Rayleigh-Taylor instability operating on the distorted ionosphere with the destabilizing effects of gravity, a zonal electric field, and a zonal neutral wind included.