The first Galileo flyby will be remembered for the insight it has provided into the remarkable complexity of the plasma distributions in the plasma sheet. Frank et al. [1993a] report that the plasma instrument provided the first fully three-dimensional determinations of the bulk motions of ion and electron plasmas in the plasma sheet near lunar distances. They find that the electron bulk speed may differ substantially from the ion bulk speed, with the differences principally in the field-aligned component of flow. The measurements are extremely challenging. The electron flow velocities, corrected for the spacecraft potential, are measured with errors of order 100 km/s, in plasmas with electron thermal speeds of order 20,000 km/s. Although it is evident that, where field-aligned currents are present, the positive and negative particles carriers must be moving relative to each other, it is very gratifying that the current-carrying population has actually been resolved. This technical achievement promises to change our understanding of the flows in magnetotail regions where primarily electron properties had been measured on earlier spacecraft.
The plasma distribution functions which appear in the mid-tail
region from about -90 R
to -35 R
also provide new
perspectives on the physics of the distant plasma sheet.
Frank et al. [1994a and b] report that the proton velocity
distributions are extremely non-Maxwellian. Characteristic
features that emerge in many of the observations are illustrated in
Figure 1, which presents samples of data from a current sheet
crossing near -50 R
. Using picturesque, self-defining
language, Frank et al. refer to lima bean distributions,
evident for example in the interval (e), and cold beam
distributions, present on the negative v
axis during interval
(i), for example. The lima bean distributions have typical spreads
corresponding to 
keV energy while the cold beams have much
lower thermal energies. The net negative z-velocity,
particularly evident for the cold beams, arises because the solar
wind flow was deflected southward during this interval [
Kivelson et al. 1993a] and this set the entire tail in motion.
Figure 1 shows that features of the velocity distribution may vary
on the time scale of a single 
19 second frame as in interval
(b) where the upper and lower half planes are sampled within 9.5 s
of one another. These times are of the order of the proton
gyroperiods (2
m/qB). More often variation is
significant from one frame to the next.
The different elements of the distribution function change intensity and flow direction in samples from the mantle, the plasma sheet boundary layer, and in samples from the current sheet near the neutral line, earthward of it, or tailward of it. In particular, in the immediate vicinity of the neutral line, the lima bean distributions are absent and complex patterns of cold beams are present. Frank et al. [1994 a and b] interpret the observations in ways that link to advances in the theory and computer simulation of particle trajectories in a steady magnetotail magnetic geometry [ Lyons and Speiser, 1982; Chen et al., 1990, Burkhart and Chen, 1991; Chen, 1992; Ashour-Abdalla et al., 1993]. Figure 2 shows schematically the difference between ion trajectories that emerge from the neutral sheet roughly within half a gyroperiod and those that move farther along the neutral sheet where they are substantially accelerated by the cross tail electric field before emerging. The former trajectories form the cold beam plasmas and the latter are thought to form the lima bean (heated) distributions. The observed distributions are remarkably similar to the forms described by Cowley and co-workers [ Cowley and Shull, 1983; Cowley, 1984] who used simple conservation laws to show the expected form of magnetotail distribution functions. When earthward (or tailward) convection is imposed by a uniform cross tail electric field, the curved field lines of the current sheet form moving magnetic mirrors that accelerate the ions in the inertial frame. Making allowance for pitch angle scattering which can arise either from non-adiabatic motion in the current sheet or the action of wave noise, Cowley and co-workers predict that lima beam distributions and cold beams may be simultaneously present. Figure 3, from their work, shows how these processes would affect tailward flowing plasmas and resembles greatly the lima bean plus cold beam distributions in tailward flowing plasma shown by Frank et al. [1994 a and b].
As Galileo moved earthward, it encountered different plasma regimes, some of them several times. For example, the spacecraft crossed the neutral sheet repeatedly, partly because the magnetotail moved northward and southward in response to changes in the solar wind flow direction [ Kivelson et al., 1993a]. Moments of the plasma distributions [ Frank et al., 1994a] show that the total particle pressure changed by more than an order of magnitude between the lobe or boundary layer and the central current sheet, but the total pressure (thermal plus magnetic) varied only by about 20% across the entire tail over a period of hours from 0400 to 1200 UT on December 8, 1990. This is illustrated in Figure 4. Although the distribution functions contained ever changing collections of cold beams and highly anisotropic energetic populations, their second moments continued to satisfy the requirements of force balance. One marvels at the creative ways in which nature arranges to satisfy conservation laws.