Midlatitude F region winds velocities observed during main phases of magnetic storms have been shown by Hernandez and Roble to be correlated with the rate of energy input into the ring current as measured by d (Dst)/dt.

The attempt by Hernandez and Roble to relate the winds to the ring current decay mechanism of Cummings and Dessler fails because a correlation of wind velocity with positive values of d (Dst)/dt is necessary to support this mechanism rather than the observed correlation with negative values of d (Dst)/dt.

Thus an alternative mechanism for deposition of ring current associated energy into the midlatitude thermosphere is needed. The first event discussed by Hernandez and Roble shows a wind divergence at midlatitudes which is taken as evidence of direct energy input there, and there are several suitable mechanisms, some or all of which would act in addition to high latitude sources (clearly present in the absence of changes in Dst).

The mechanism of the precipitation of energetic neutral atoms arising from the charge exchange neutralization of ring current ions is suitable in all respects for a direct midlatitude energy input, except that the energy flux is several times too small, but there are reasons for the flux being greater than hitherto inferred.