Onsager et al. (1991) have put forward a model of the formation of the plasma sheet boundary which relies on a steady source of plasma from a spatially extended plasma sheet, together with steady equatorward and earthward E¿B convection of field lines due to reconnection at a downtail neutral line. This model is a synthesis of earlier proposals and it explains such features as an electron layer exterior to the ion boundary layer, ion velocity dispersion, counter streaming beams, low-speed cutoffs in the beams. It also explains the apparent evolution of the ion beams through ''kidney beam'' shaped velocity-space distributions toward quasi-isotropic shells without invoking pitch angle scattering or energy diffusion. In this paper we explore two ramifications of the mode: (1) low-speed cutoffs observed in the ion distribution functions contain formation about the distance to the neutral line downtail and (2) values of phase space density at the low-speed cutoffs are the same as those in the distribution just earthward of the reconnection site.

In principle we can map, as a function of time, the downtail neutral line distance and establish whether or not it is retreating during substorm recovery. We can also reconstruct the plasma distribution function near the neutral line to see if it is most consistent with mantle or plasma sheet plasma. We perform this analysis using ISEE Fast Plasma Experiment data for two plasma sheet recovery events, one on March 1, 1978, and the other on April 18, 1978. On March 1, 1978, we find evidence for an initial retreat from around 110 to 160 RE in the first 15 min; little further retreat occurs thereafter. On April 18, 1978, the neutral line location ranges from as little as 40 RE tailward of the satellite to as much as 200 RE, but there is no evidence for a systematic retreat. The reconstructed ion distributions for these events are most consistent with a plasma sheet origin for the March 1 case and possibly plasma mantle or low-latitude boundary layer for the April 18 case. ¿ American Geophysical Union 1995