Sttmmary.-Rats ( N = 6 9 ) were trained in a runway under 0-, lo-, and 20-sec. delays and extinguished with 0-, lo-, 20-sec. delay-box confinements using one trial/day in each stage of the experiment. Duration of delay in training was inversely related to running speed in acquisition and in extinction. Duration of confinement in extinction was inversely related to running speed in extinction. Increases in delay from training to extinction led to fastec extinction; decreases in delay to slower extinction.

Snmmary.-72 rats were trained in a runway for 9 days. Group D l was delayed for 2 0 sec. on tke first 3 days, D 2 o n the second 3 days, and D 3 o n the last 3. All other trials were rewarded immediately. During 5 days of extinction half of Ss had 20-sec. delay-box confinement on all trials; the other Ss had no confinement. Delays in training produced sudden, pronounced changes in behavior. O n the average, Ss ran faster in both training and extinction, the earlier delay was introduced in training. Running speed in extinction varied inversely with the duration of corfinemenr in extinction. Shifts to increased delay from training to extinction produced faster extinction; shifts to decreased delay led to slower extinction in comparison with relevant unchanged groups.

Thirty-five rats were trained for 36 days (6 trials/day) in a runway with 20 sec delay (D) In two previous studies (Wike, Mellgren, & Wike, 1967) it was observed that rats with 20 sec delays on all trials initially ran very slowly in a runway but with continued training they tended to improve. The purposes of the present study were: (a) to determine whether or not the deleterious effects of delay can be overcome as a result of long term training; and (b) to investigate further the influence of delay-box confinement during extinction upon extinction performance. The work of Tombaugh (1966) and our previous studies suggest that increases in confinement duration from training to extinction hasten extinction, and decreases retard extinction. or no delay (ND). Each training group was divided in extinction.into 20 sec delay-box confinement (D) or no confinement (ND). Extinction lasted for 12 days (6 trials/day MethodThe Ss were 35 male Sprague-Dawley rats, 65 days old. One additional S died during the experiment. The Ss were randomly assigned to the four treatments described below.The L-shaped runway consisted of an 8 in. long start box, a 33.5 in. long alley, a 16.5 in. long delay box, and a 8 in. x 13.5 in. goal box. The unpainted 4.75 in. wide delay box had a Plexiglas cover; the rest of the 4 in. wide runway and the goal box were painted black and covered with hardware cloth. The response measure, running time, was the time between E's elevation of the start box door and S's breaking a light beam located 3 ft from the start door.During the first 14 days Ss were put on a 23 h hunger cycle with 1 h daily feedings, handled, adapted to the runway, and given two immediately rewarded (two 45 mg Noyes pellets) runs. Training extended over 36 days. Each S had six trials per day with the trials being rotated among six Ss. The 18 Ss in Group ND were rewarded immediately; the 17 Ss in Group D were delayed for 20 sec on all trials. In extinction the two training groups were divided Psychon. Sci., 1967, Vol. 9 (7A) EDWARD L. WIKE AND JUNKO McWILLIAMS THE UNIVERSITY OF KANSASrandomly with half of the Ss receiving no confinement in the delay box (ND) and half being confined for 20 sec (D). Thus, the experiment involved a 2 by 2 factorial design with 0 and 20 sec delay-box confinement in training and extinction. On all extinction trials Ss were kept in the goal box for 10 sec. Extinction extended over a 12 day period, with six rotated trials per day as in training. Results and DiscussionThe median running times from pretraining (P), training, and extinction are displayed in Fig. 1. The slight difference between Groups D and ND prior to training was in favor of D and was not significant (p> .05). It may be observed that after 216 training trials the two groups still remained separated. No statistical analysis was performed on the training data because the D group was obviously slower by inspection. For example, on the last five days of training the running times from the two groups formed two nonoverlapping distributions in ...