Transient variations in responding to Pavlovian conditioned stimuli have implications for the mechanisms of “priming”