Inconsistency in Classical Electrodynamics?*