EFFECTS OF PRENATAL RADIATION UPON LEARNING 653 



been deprived of food for 24 hours at the start of the preconditioning 

 period, for 48 hours at the start of the conditioning period, and again for 

 24 hours at the start of the extinction period. 



Results 



Each group was tested for a significant difference from every other group 

 by the Mann-Whitney U test on the following measures: number of re- 

 sponses during the 30 minute preconditioning; number of responses during 

 the 60 minute conditioning; number of responses to extinction; extinction 

 defined as a return to or below the preoperant level of responding over a 

 30 minute period; number of responses to extinction, extinction defined as 

 5 minutes without responding; time required for extinction, extinction de- 

 fined as a return to or below the preoperant level of responding over a 30 

 minute period; time required for extinction, extinction defined, as 5 minutes 

 without responding; number of responses during the first 5 minutes of 

 extinction. 



The entire group of 540-day-old animals was compared with the entire 

 group of 120-day-old animals for each of these measures; no significant 

 diflferences were found. Of 637 tests, 18 reflected significant differences 

 between the 200 r/1.5 day subjects and all but one of the other groups, 

 for numbers of preconditioned responses and numbers of responses during 

 the first 5 minutes of the extinction period. This consistent direction of 

 difference for the one group and the fact that only 23 other differences 

 were found between the various groups with no pattern of consistency, led 

 to an examination of the probability that all differences noted were due to 

 chance or artifact. Further tests provide evidence that 36 of the 41 diff"er- 

 ences noted could have been due to chance alone. 



Discussion 



The Learning T.asks 



Several inferences can be drawn from the data of the two experiments. 

 It is apparent that while the maze provided a learning task sufficiently 

 sensitive to detect behavioral aberrations, the bar pressing instrument did 

 not. This does not imply that the operant conditioning technique cannot 

 produce evidence of behavioral anomalies resulting from prenatal irradia- 

 tion, but it suggests that the schedule undertaken in establishing the condi- 

 tioned response and studying its extinction was inappropriate to discriminate 

 between prevailing deficiencies in the learning capabilities of the subjects. 

 Now the problem is to explore intensively sequences of schedules; after the 

 Skinner technique (Ferster and Skinner, 1957), to establish appropriate 



