The study involved three experiments. The first. a parametric investigation of nictitating membrane conditioning with eight constant intertrial intervals (lTls) between 5 and 120 sec, orthogonal to interstimulus intervals (ISis) of 250 and 750 msec plus three temporal conditioning control groups, revealed that performance improved rapidly with increasing ITI but stabilized at relatively low ITI values. At 75Q-msec lSI, a decrement in performance was found at 60•sec IT!. Experiment II, using constant ITIs of 45-75 sec in 5-sec steps, at 75Q-mseclSI confirmed the trend toward a performance decrement around 60 sec, although the trend was weak and highly variable. Experiment III evaluated the differences in performance between constant and variable ITI, using three ITI values and three conditions of variation at each value. Findings were discussed in terms of differences in conditioning resulting from both length and degree of variation of ITI and some subtle effects which may emerge only when constant ITIs are used.

A parametric investigation of nictitating membrane conditioning with interstimulus intervals (ISIs) of 250 or 1.000 msec. constant intertrial intervals (1TIs) of 30 or 90 sec. and 5. 15. 30. 60. or 120 trials per daily session revealed that acquisition rate decreased as a fUr;lction of both larger number of trials per session (NTS) and longer lSI. but with larger overall effects of NTS occurring at the longer lSI. The effects of lengthening the ITI were also greater at the higher NTS and varied as a function of lSI. Results were discussed in terms of several constructs used to explain time-related conditioning effects. namely reactive inhibition. stimulus fluctuation. consolidation. and rehearsal.The number of trials per session (NTS) has received considerable attention in the recent conditioning literature. In a study involving nictitating membrane response (NMR) conditioning. Hupka. varied interstimulus interval (lSI) and NTS and found that acquisition was directly related to NTS. with faster acquisition at smaller NTS. Moreover. there were only slight differences in acquisition rate as a function of lSI with larger NTS. a tinding which seems to be at variance with much of the conditioning literature. Levinthal and Papsdorf (1970). using only one trial per daily session. found a reversal of the usual lSI function. That is. at 1 TS. their 1.2SO-msec lSI group conditioned significantly more rapidly than their 2SO-msec group. In a later paper. Levinthal (1973) replicated the earlier finding. and in addition. using 1. 5. or 20 TS. found that the lSI reversal took place between 1 and 5 TS. In a more recent study. Kehoe and Gormezano (1974) employed 1. 5. 10. or SO TS, and confirmed the previous findings that acquisition rate varied systematically as a function of NTS.In addition to these studies. there is increasing evidence in the rabbit conditioning literature that conditioned responses (CRs) often emerge in the early trials-frequently the first trial--of a session and continue at a rate of nearly 100% even though no CRs had occurred in previous sessions. This phenomenon has been obse~ved frequently in our laboratory and has been noted in several recent studies (e.g., who referred to it as an "overnight gains effect" (1966, p, 211), offered no explanation. Hupka et at. suggested that the effect might result from over night consolidation of training, Regardless of the explanation. however, it seems likely that the effect would be related to such variables as NTS,The purpose of the present study was to investigate the relationship between NTS, lSI. and ITI over an extensive range of NTS values. METHOD SubjectsThe subjects were 145 naive male and female New Zealand rabbits, 70 to 100 days old at the start of the experiment. They were maintained on ad-lib food and water and housed in Iight-and climate-controlled quarters. ApparatusFour rabbits were run concurrently in a sound-attenuated room containing four separate cubicles. serviced independently by control and recording equipment located in an adjoining room....

The effects of different placements of electrodes used to present shock unconditioned stimuli on conditioning the rabbit nictitating membrane response (NMR) were examined. Fifty-one rabbits were conditioned with shock leads attached either ciroumorbitally, at the ear base, or at the ear tip. In one circumorbital shock group, nictitating membranes of both shocked and unshocked eyes were monitored. Circumorbital, ear base, and ear tip shock yielded best to poorest performances, respectively, and the shocked membrane yielded better performance than the unshocked one. Analysis of unconditioned response and conditioned response latencies and peak amplitude, as well as slow motion filma of responses, suggested that poorer performance resulted from the elicitation of other responses, of which the NMR may have been a secondary or indirect component.

No abstract