In each of three experiments, Ss received classical eyelid conditioning with either a corneal airpuff or a subjectively matched infraorbital shock as the ucs. It was expected that conditioning levels would be higher for the presumably more "modifiable" airpuff ucs. Contrary to expectation, infraorbital shock produced superior conditioning in all three experiments. However, Ss' ratings of airpuff and shock aversiveness in Experiments n and in were negatively correlated with anticipatory CHS, indicating that an instrumental reinforcement mechanism operates for both ucs modalities. The results were discussed in terms of contiguity and effect models of classical conditioning.

The importance of CS duration in concurrent trace conditioning of the GSR and eyelid response was investigated with 32 retardates. Four treatment groups completed a 2-day training procedure involving habituation, acquisition, and extinction. Irrespective of response class, a long CS (1000 msec.) resulted in better conditioning than did a CS with a shorter duration (100 msec.). This finding could not be accounted for by either interstimulus or trace interval differences.