The purpose of the present study was to detect the phenomenal findings that had been detected in the partial reinforcement schedule in the case of the partial delay of reinforcement schedule.The findings were as follow:1. The superiority of the performance level of the partial reinforcement was detected in all parts of the runway. There was a tendency that the nearer to the goal box area the part in the runway was, the later the period was in which the superiority of the performance of the partial reinforcement group appeared.2. The effect of the detention time was also detected by the fact that Group P-10.5 (partial reinforcement group, detention time 10.5 sec) was equal to Group C-0 (continuous reinforcement group) and inferior to Group P-2.5 and P-30.According to the first purpose, the runway was divided into the four parts (i.e., the start box area, the first runway area, the second runway area, and the goal box area), the running time was measured in each part and the 120 training trials were used in the schedule of 10 trials per day for 12 training days.The four groups which consisted of five or six subjects (albino rats) were used. One group was no delay group and the other three groups were the partial delay (50%) groups. According to the second purpose, the one of the 3 lengths of the delay time in the goal box in delay trials (2.5 sec, 10.5 sec and 30 sec) was assigned to one group of the partial delay groups.The results were quite identical with those in the partial reinforcement except the result of Group D-2.5 (partial delay group, delay time 2.5 sec) in the first and second runway areas.The similarities between partial reinforcement schedule and partial delay of reinforcement schedule in the relationship to the continuous or no delay group in the performance levels suggested the possibility of same mechanism under which the effects of these schedules exerted on the performance level.The mechanism was assumed to be the frustration mechanism (Amsel 1952, 1958).