The Societas Meteorologica Palatina, or Meteorological Society of Mannheim, was set up in 1781 to coordinate observations of the weather on an international scale. In addition to temperature, pressure and humidity, observers connected to the network were instructed to record
various atmospheric phenomena, among these the aurora borealis. The 39 stations of the network reported about 1400 individual sightings of auroras during the Society's dozen years of existence. The reported sightings are subjected to a statistical analysis that brings out striking discrepancies
between the number of auroras that one would expect and the number that was reported. The statistical analysis is supplemented by an analysis of the theoretical and phenomenological comments in the Society's annual reports. The study suggests that observers on the Continent considered themselves
just as advantageously situated as observers further north when trying to solve the riddle of the northern lights. It also illustrates the variety of conflicting ideas about the aurora borealis that existed during the late Enlightenment, and how these might have influenced the number of reported
auroras. This lack of consensus contributed to many anomalies in the data presented in the Society's reports. By combining linguistic and scientific competence it is possible to shed light on these anomalies and on the historical context that shaped them.