In the inner coma of a comet the total ion (or electron) density results from an equilibrium between ion production and ion destruction. The ions are mainly produced by photoionization of cometary gas, whereas the main destruction process is dissociative recombination of molecular ions with electrons. The reaction rate for the recombination depends on the temperature of the thermal electrons and decreases with increasing temperature. The electron temperature therefore determines (together with the photoionization rate and the gas density) the total ion density. During the flyby of the European spaceprobe Giotto at comet P/Halley an unexpected peak of the ion density at a cometocentric distance of 12,000 km was observed [Balsiger et al., 1986>. The ion density at the position of the maximum was even higher than at the nearest recorded spectrum at 1300 km from the nucleus. The observed peak of the ion density is explained by a steep increase of the electron temperature with distance [H¿berli et al., 1995; Eberhardt and Krankowsky, 1995>. We solved the coupled equations for the electron temperature and density in local energy balance. We further assumed that local photochemical equilibrium holds for the electron density, which is indeed justified in the region of interest. If one assumes the excess energy from photoionization as the only energy source for the electrons, the results are inconsistent with the measurements. The resulting increase of the electron temperature with distance is not sufficiently steep to produce a peak of the ion density around 12,000 km from the comet. The observed peak of the ion density can only be explained with this model if an additional energy source for the electrons is taken into account which locally even dominates over heating by photoionization. As an additional energy source, a population of suprathermal electrons with a typical energy of 10 to 15 eV is suggested. We present a mechanism which may lead to an enhanced flux of solar wind electrons in the vicinity of the nucleus due to the compression of the interplanetary magnetic field. The expected flux of suprathermal electrons at about 15,000 km from the nucleus is in good agreement with the electron measurements in the coma of P/Halley aboard the VEGA 2 spacecraft. This flux is also sufficient to heat the thermal electrons to the required temperature in order to explain the peak of the ion density. ¿ American Geophysical Union 1996