Abstract
Conventional non-relativistic quantum electrodynamics is used to calculate position-dependent corrections to the magnetic moment of an electron a distance z from a perfectly conducting surface. To leading order, delta mu /sub ///=-e3/32m2z and delta mu z=0. These results follow easily and without any special pleading from the usual Foldy-Wouthuysen Hamiltonian, and are validated by necessarily much more elaborate relativistic calculations using either the full quantum field theory, or Dirac single-particle theory. Together with a recent non-relativistic account of the free-electron anomalous moment, such agreement is interpreted as a plausible indication that non-relativistic quantum electrodynamics is at least qualitatively adequate to deal with magnetic radiative effects, on the same footing as it can deal with non-magnetic analogues like the Lamb shift.