IIai'vard i r H'ivci '.tily The effects of varying the interval between the conditioned and unconditioned stimuli in an avoidance learning situation have only recently been subjected to intensive experimental examination. Warner's (7) early experiment demonstrated that rats required more trials to achieve an avoidance criterion as CS-US interval was lengthened. However, there was no study of extinction, and many important indices of acquisition were not reported. Kamin's (3) recent experiment represents the first detailed analysis of the effects of the CS-US interval on the acquisition and extinction of an avoidance response. Using a trace-conditioning procedure, and CS-US intervals varying from 5 to 40 sec., Kamin found that the shorter the CS-US interval, the greater the case of acquisition and the greater the resistance to extinction. Using a delayedconditioning procedure, with a similar range of CS-US intervals, Brush, Brush, and Solomon (1) found no such marked effect from the same parameter. When the CS-US interval was held constant between the two experiments, somewhat faster acquisition and much greater resistance to extinction were found in the delayed-conditioning experiment.Although Brush, Brush, and Solomon attributed the differences between Kamin's results and their own to the difference between the trace and delay procedures, there were a number of other procedural differences between the two experiments. The CS in Kamin's experiment was the sounding of a buzzer for a maximum of 2 sec. He also used & jree-responding situation in which the dog was able to jump across the barrier at any lime. In the Brush, Brush, and Solomon experiment, the CS was the offset of the light in the dog's

In 6 ourarized dogs, the magnitudes of both the cardiac acceleration during shock stimulation and deceleration after stimulation were found to be (a) monotonic increasing functions of intensity (2, 4, 6, and 8 ma.) and (b) inverted U-shaped functions of duration (0.1, 0.5, 2.5, 5.0, and 10.0 sec.) of electric shock. Utilizing 32 curarized dogs with a discriminative classical conditioning procedure, (a) the presence of a warning signal did not affect the magnitude of the unconditioned cardiac response, and (b) previous experience with shocks of lower intensity did reduce the magnitude of the unconditioned cardiac response to subsequent high-intensity shock.A strong electric shock results in large changes in the dog's heart rate. During shock, heart rate increases abruptly; when shock is terminated, the heart .rate rapidly decreases to a rate below the preshock rate and then slowly returns to normal (Black, Carlson, & Solomon, 1962). The acceleratory phase has been thought to reflect subjective "shock severity," while the deceleratory phase has been thought to reflect subjective "relief." The purpose of our first experiment was to evaluate the degree to which these two cardiac phenomena could usefully serve as indices of severity of electric shock. The purpose of the second experiment was to evaluate the effects of prior experience with shocks of increasing intensity, and the presence of a warning signal on the cardiac indices of shock severity. Both experiments used dogs paralyzed by tubocurarine chloride in order to eliminate movements and to control respiration, two sources of variability which affect heart rate (Westcott & Huttenlocher, 1961).EXPERIMENT 1 Our initial assumption was that the severity of pain is positively related to both