The question as to whether the momentum of a transverse electromagnetic (TEM) wave packet in a simple, but selfconsistent, model of an electromagnetic material described by time- and field-independent scalars ε and μ, is given by ∫ D × B dV or ∫ (E × H)/c2 dV is re-examined. It is found that the momentum of the wave packet is given by neither of these two expressions. ∫ (E × H)/c2 dV has to be supplemented by a momentum attributable to the material. It is also found that the power flow density of the electromagnetic field E × H has to be supplemented by a mechanical contribution. The net energy momentum four-tensor is symmetric. These conclusions apply to a model of the electromagnetic medium forming a closed system, including Newton's law for the material. Finally, it is pointed out that a strict linearization of the equations describing the wave propagation, treating the system under consideration as an open system in that momentum interchange occurs with the higher order terms omitted in a linearization, leads to the assignment of ∫ D × B dV as the momentum of a wave packet. The dispute as to whether the Abraham or Minkowski energy stress tensor is the correct one is, hopefully, clarified by pointing out that self-consistency can be achieved only by a careful consideration of the assumptions contained in the model of the electromagnetic material. © 1969.