V A Hospi~aL, Coral Gables, FloridaS u m m a r y . 4 rats were reinforced with cream and sugar for traversing a 4-ft. runway which allowed right and left approaches to separate goal boxes. After each S had satisfied 3 criteria of stable responses to a goal, the path to that goal was blocked by a door, rhus requiring a discrimination reversal. All Ss met the same criteria of stabiliry through 2 to 4 reversals as a decreasing function of number of reinforced trials.Recent reviews of the literature on discrimination reversal in the rat (Paul, 1965;Sperling, 1965) show that most of the studies have used an arbitrary number of days or some criterion set for all Ss in a particular group to define original as well as post-reversal learning. Some of the discrepancies between studies designed to investigate the same or similar variables may be due to the fact that each S was not required to reach the same criterion prior co discrimination reversal. Requiring each S's behavior to meet such a baseline or criterion enables intrasubject comparisons of original and reversal learning in single organisms and, thus, reduces the total number of Ss needed in each study.This study was a follow-up of an earlier experiment (Davidson, 1966) which demonstrated baselines or criterion behavior of the kind proposed in single Ss (alligators). Since these earlier baselines were somewhat variable and the procedure used (heat-escape conditioning) was one rarely applied to rats, the method was replicated with the more common laboratory animal (rat) using more typical procedures (positive reinforcement) in a runway of the same design.
METHOD
SldbjectsFour male, hooded rats were maintained at 80 to 85% of their free-feeding weights beginning at about 120 days of age. Each S was weighed and fed Purina rat chow according to body weight. Food was delivered following daily trials. Water was provided ad lzbitz~m in the home cages.
ApparatusA 48-X 24-X 8-in. high, black, plywood alley, with a 16-in. partition dividing the goal end into right and left approaches, was the apparatus. These dimensions were proportional to the larger maze used with alligators (Davidson, 1966). Standard Hoeltge rat cages equipped with metal lids and cardboard guillotine doors were used as interchangeable start and goal boxes. The doors