This letter is a re-examination of the ethylene oxide (EO) temperature programmed desorption (tpd) peak shapes (reported in references 1 and 2) which were obtained by tpd after having adsorbed EO at 250 K on to Ag(110) and Ag(111) surfaces. In these papers, the peaks were deemed to originate from a unimolecular rearrangement of an adsorbed oxametallacycle, producing EO, which desorbed on formation. Consequently, on the basis of this thesis, the activation energy of the composite process of internal rearrangement of the oxametallacycle and desorption of EO was determined by solution of the first order Redhead equation [3] at the peak maximum temperature, using an assumed value of 10(13) s(-1) for the desorption pre-exponential term. The conclusions of this re-examination are that the m/z = 29 desorption peaks shown in Fig. 1 derive from a second order surface reaction of an adsorbed O atom and adsorbed ethylene molecule and that this is the rate determining step in the desorption of EO. An important corollary of these conclusions is that an oxametallacycle is not involved in the ethylene epoxidation reaction coordinate as the papers suggest.