AX ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 23 



compared with Table II it is seen that a sHght inhibition con- 

 tinued throughout this series of tests as indicated in the average 

 number of responses per 50 trials (Table III-B) and the per cent 

 of stimuli followed by response (Tables III and VI). The aver- 

 age number of reactions per response was also less than in the 

 normal food series. For the food-pressure response the average 

 number of reactions in the numerical order of the six snails was 

 3.96, 3.62, 1.96, 4.1, 1.58, 3.78, or a mean for all six snails of 

 3.18. The corresponding values for the normal food response 

 (Table II) are 3.48, 4.35, 6.5, 6.5, 3.4, 10.93, with a mean value 

 of 5.85 for the six snails. Only in the case of Snail No. 4 is there 

 a greater average number of reactions per response to the " con- 

 ditioned " stimulus than to the " unconditioned." Table III 

 shows further that the average duration of the reactions in the 

 food pressure response is less than that of the normal food re- 

 sponse. Expressed in the number of reactions per minute the 

 values are 51.84, 51.92, 65.45, 51.72, 56.7, 50.84, with a mean 

 of 54.74. The corresponding values for the normal food response 

 are 46.51, 45.9, 46.15, 46.51, 44.44, 48.00 with a mean of 46.25. 

 Under food-pressure stimulation the reactions thus occur at the 

 rate of 8.49 more per minute than under the normal food stimu- 

 lus. Along with the reduction in the number of reactions there 

 is in the food-pressure response an acceleration of the individual 

 reaction as compared with that of the normal food response. 

 The distribution of the reactions over the series of 250 trials 

 is shown for the six snails in Tables III-A and III-B and is 

 represented in graphs (figs. 2-13, Plates II-III). Table III-A 

 and figs. 2-7 (Plate II) show the great fluctuation in the number 

 of reactions per response. This fluctuation causes the reaction 

 curve to vary from the response curv^e although the two are in 

 general similar in direction. The response curve shows a some- 

 what more uniform rise than the reaction curve. A general fact 

 to be noted is the relative uniformity in the average duration of 

 the individual reactions throughout the series. The method of 

 timing the response duration by means of the stop watch is 

 necessarily inaccurate, especially with values which, as in this 

 case, are frequently in the neighborhood of a second. If a more 

 accurate method were available greater uniformity might be 

 expected. With the method employed it is noteworthy that 

 the agreement is so close. 



