According to the usual form of measured cross sections, the slope of an Arrhenius plot differs from the activation energy by an electron volt or so at high T. Thus, cross-section shape is needed before activation energy can properly be deduced from rate data.