Gordon has not given a tensorial form of Dirac’s equation but his well-known decomposition of the Dirac current into a conduction and a polarisation part.