Small secondary peaks are frequently measured in energetic (25 eV<E<15 keV) electron distribution functions during rocket flights over bright auroras. One sample distribution function that was measured at 250 km is analyzed in this paper. The purpose of the analysis is to see if the observed secondary peaks are large enough to drive electrostatic instabilities. Two specific features of the distribution function are analyzed in detail: a secondary parallel peak, characterized by positive ∂f/∂&ngr;⊥. The parallel peak produces positive growth rates for electrostatic whistler and ion cyclotron waves, but these growth rates are too small to generate significant wave amplitudes within an arc. The perpendicular peak produces much larger growth rates for electrostatic electron cyclotron and upper hybrid waves when these two modes are interacting. We conclude that the perpendicular peak was the last feature to produce a strong instability as the electrons moved down to the rocket. The largest waves should be seen as rockets reach an altitude where &ohgr;UH=2¿&ohgr;ce‖. The plasma detected at 250 km appears to have evolved until the growth lengths of waves that are driven by the perpendicular peak are comparable to the dimension of an arc. Once generated, the upper hybrid and electron cyclotron waves will also resonate with electrons which have small pitch angles, and they can therefore spread the parallel peak well beyond the point at which it produces significant growth rates.