A standard procedure, originated by D. M. Page and used in the Columbia laboratory for establishing a discriminative response in the white rat, is one in which the positive stimulus (the SD) is presented after a period of no responding in the absence of this stimulus (S<1) . Thus, when the SD appears and the rat responds, as by pressing a bar, it is reinforced with a pellet of food, after which S<1 is presented. The SD is again presented following a period (say 30 sec.) in which no responses during S<1 have appeared. Any response during S<1 serves to postpone the presentation of SD for another 30 sec. The extinction of S<1 responding, essential to forming a discrimination, is accordingly compressed within the early minutes of training, and good discriminations (involving very few responses during S<1 and low latencies to the SD) are rapidly established.A similar procedure of reinforcement may be used when no specific discriminative stimuli are presented. In such a case, reinforcement is contingent upon a response that follows a period of nonresponse. For example, a bar press is reinforced only if it follows the preceding bar press by at least 30 sec.Skinner (3, p. 306) used such a technique in a successful attempt to break down high rates of responding that had previously been set up under periodic reinforcement. However, more information from animals with no previous conditioning history is desirable. The present study was undertaken to determine (a) whether the strength of a conditioned operant (bar pressing) may be maintained under such a reinforcing schedule, (b) what changes in rate, if any, may be expected when the delay intervals are progressively increased, and (c) the nature of extinction following such a procedure.
METHODThree male albino rats of Wistar strain, approximately five months old, were used. Prior to and during experimentation, they were housed in individual livingworking cages (2) and fed daily after 22~ hr. of food deprivation. Purina Laboratory Chow Checkers made up the basic diet, and the same food, in the form of I-gr. pellets, was used to reinforce the bar-pressing response.During experimental sessions, which were 1 hr. long, the living-working cages were placed inside separate lightproof, sound-resistant cubicles, which also contained the lever and the automatic food magazine. The animals were observed by means of a mirror placed over the cage and glass windows which replaced the lightproof doors. Illumination within the cubicle was provided by a 9-w. General Electric red lamp mounted 11 in. from the top of the cage. Previous work had shown this light to be nonaversive. The experimental room itself was dark.The recording apparatus, the lever, and the food magazine were modified forms of those described elsewhere (see 2). The bar portion of the lever was % in.in diameter. It required approximately 10 gm. of force for depression and moved through a vertical distance of 14 mm. Since there were slight differences in lever operation, care was taken to see that each animal used the same appa...