If figure 5 we show the effect of increasing pressure on a stable negative polarity discharge. As is evident from the color photograph, the pink emission from ion impact photoemission (positive bias) is located at the equator, whereas the blue emission from electron impact photoemission (negative bias) dominates over the poles of the magnet. This two color scheme illustrates that electrons and ions have differing regions of importance, and greatly simplifies the interpretation of the floating potential.
Figure 6 plots the current-voltage profiles of a Langmuir probe which give us the floating potential at various locations in the dipole plasma. As can be seen in the data, the equatorial floating potential is more positive than the polar floating potential. Since all the field lines terminate on a nickel-plated magnet which is presumeably at constant potential, one would have to assume that there are field-aligned potential drops to account for this difference, even if one were unable to examine the same field-line at two different points.
Such field-aligned potentials are expected if the ions and the electrons have different pitchangle distributions. Since the ions and electrons have gyroradii differing in size by factors of several thousand, the large gradients in the fields are guaranteed to produce differing distribution functions. From the colors of figure 6 we infer that the electrons are produced on the nickel surface of the magnet and are therefore field-aligned, whereas the ions are produced by collisions in the trapping region near the equator and are therefore highly trapped with a pancake distribution. This combination of distributions sets up the field-aligned potential observed, and may account for space-charge and lightning-like discharges described earlier.