dure. These animals were subsequently tested according to a within-subject design in four different novel environments under the following conditions: (i) notreatment, (ii) saline vehicle injection, (iii) naloxone (2 mg/kg) administered at the beginning of the 5-hour delay immediately after the first four choices on the maze; and (iv) naloxone (2 mg/kg) delayed for 2 hours after the initial four choices. The assignment of animals to treatments in each novel environment (a room) was approximately counterbalanced. As there was no statistically significant difference between the no-treatment and saline-treatment conditions (Table 2), for each animal, mean trials to criterion and error scores were calculated. These mean control values were used in subsequent statistical analyses. As in the previous experiments, naloxone administered immediately after the first four choices significantly enhanced performance, as reflected in both trials and errors to criterion (14). The data from the naloxone-delay condition do not, however, differ significantly from either the control treatments or the naloxone treatment at no delay. These results indicate that naloxone does not produce an effect comparable to that of the no-delay condition when administration is delayed for 2 hours after training. As in the previous experiments, rotation of the maze on a test conducted for each animal after criterion performance indicated that performance was based on extra-maze cues provided in each environment 113). The results indicate that when performance on a spatial learning task is less than optimal, administering an opiate antagonist after training can enhance memory. Thus, the memory-enhancing effect of opiate antagonists seems not to be restricted to memories acquired through aversive training or associated with noxious events. This view of a more general role for opioid peptides in memory is congruent with prelminary clinical studies that have recently indicated that naloxone improves memory functions in patients with Alzheimer's -disease (15). Further research on the role of opioid peptides in memory may, therefore, have important implications for understanding the biological basis of both normal memory and its disorders. MICHELA GALLAGHER RICHARD A. KING NANCY B. YOUNG Department ofPsychology, University of North Carolina, Chapel Hill 27514