Fig. 6.  Memory performance in the first minute after encoding. In the Encoding Task, image pairs were presented 6, 27, or 63 s apart. Recognition of the second (old) presentation of the 27 and 63 s probes is shown as a ratio of recognition at 6 s. Random confounders such as fatigue or sequence effects are equally distributed, and thus differences would reflect a true memory effect. At 27 s, there was no significant effect of drug (P = 0.416), level (P = 0.828), or drug:level interaction (P = 0.537). Similarly, at 63 s, there was no effect of drug (P = 0.915), level (P = 0.914) or drug:level interaction (P = 0.704). Although our other results imply that the events leading to memory decay were established at the time of encoding, no effect on performance was detectible in the first minute. Error bars  represent SEM.