The Fokker-Planck equation is studied for an energetic ion beam injected into a magnetized plasma consisting of Maxwellian ions and electrons with υthi ≪υb≪ υthe. The time evolution of the fast ion distribution is given in terms of an infinite sum of Legendre polynomials for distributions that are axisymmetric about the magnetic field. The effect of charge exchange is included. The resulting ion distribution is somewhat isotropic for velocities much less than the injection velocity, however, the distribution is sharply peaked in both energy and pitch angle for velocities near the injection velocity. Approximate asymptotic expressions are given for the distribution in the vicinity of the injected beam and for velocities greater than the injection velocity. The effect of a weak parallel electric field is also given.

It is shown that in hydrogen dominated multi‐ion plasmas supporting coherent hydrogen cyclotron waves, the minority ion species with large M/Q are preferentially accelerated and the maximum energy achieved scales as (M/MH+)5/3. The importance of this scaling to O+ acceleration in the auroral zones and to other high energy heavy ion observations in the earth's and Jupiter's magnetospheres is discussed.

The pickup of newborn ions in the terest, the interstellar helium data obtained by solar wind by means of low-frequency, electromag-AMPTE and cometary ion pickup distributions obnetic waves, either due to the intrinsic turbu-rained by various spacecraft at comets Giacobinilence of the solar wind or induced turbulence due Zinner and Halley. to instabilities excited by the newborn ions, is discussed. The pickup process is envisioned to 1. 4 mp, •/• = 10 , Be = •p = 1, and •i = 0.01 where •. =• 8=n0kBT./B02 and VA2 = B02/4=n0mp) J J ß Near maximum growth the real frequency increases roughly linearly with k, • -kv•, with v•that the electrostatic and higher frequency elec-0.5 V 0 . The growth rate peaks at kc/• = 0.045

We use a fast, efficient method to trace charged particles through realistic magnetospheric electric and magnetic fields, greatly reducing computer simulation times. The method works for particles having arbitrary charge, energy, or pitch angle but which conserve the first two adiabatic invariants. We also apply an efficient method of classifying drift orbits, which greatly simplifies the task of identifying the last closed drift path or other drift boundaries. Finally, we calculate the time‐independent evolution of the bounce‐averaged phase space density along convective drift orbits. With these three tools, convective evolution of the particle distribution from the tail can now be described quantitatively, an essential step in understanding the production of unstable distributions in the magnetosphere. One can also categorize topologically different drift orbits, which is necessary to understand the unique particle signatures of the convecting plasma such as Alfvén layers and the plasmapause. These signatures can then be used to extract the electric and magnetic fields or to test the validity of the model fields. The method is particularly appropriate for particles in the energy range 0.01