THE approach of Hoyle and Narlikar1 to quantum electrodynamics provides a good explanation of spontaneous emission, and leads to an expression for the emission rate which is in agreement with the observed rate S, provided that the zero-point oscillations of the electrodynamic field are ignored. Previously it was considered2 that spontaneous emission was induced by these oscillations but, because there is only an energy of ½hν per state, the predicted rate in this case is only ½S.