PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 35 



Discussion of Table 8. — The figures given in this table were 

 obtained from an analysis of 4,795 trials that were given to 49 

 subjects, and since the relation of each trial to the trials that 

 preceded it was obtained by a necessarily somewhat compli- 

 cated method of calculation, a presentation of illustrative speci- 

 mens of this work will be reserved for the appendix. Table 8 

 presents in detail most of the data on which the following discus- 

 sion is based, but occasional reference to what will hereafter 

 be designated " calculations for table 8 " will be necessary. 



Alley position was evidently in itself a partial determinant of 

 some of the subject's first choices. Feeble-minded Girl 10, for 

 example, tried alley No. 1 first during each of 79 trials, and it 

 is probable that the mere position of this alley as the first one 

 to the left determined her preference for it as the first one to 

 try. Of course a marked tendency toward Type C reactions 

 (trying alleys in the order 1-2-J-4 or 4-J-2-1) would account for 

 a preference of alley No. 1 or alley No. 4 because of their ter- 

 minal positions with reference to the row of alleys. Neverthe- 

 less, table 8 shows that of the 4,795 first choices made by the 

 subjects, 1199 were for alley No. 1, 1,126 for alley No. 2, 1,093 

 for alley No. 3 and 1,377 for alley No. 4 — a finding w^hich suggests 

 that there was no marked disproportion between the average 

 reactive values of the terminal and the middle alleys. 



Twelve of the subjects have the maximum number of their 

 first choices in the alley No. 1 column, 7 in the alley No. 2 col- 

 umn, 11 in alley No. 3 column and 19 in alley No. 4 column. 

 Feeble-minded Girl 10 made no first choices whatsoever of either 

 alley No. 3 or alley No. 4. 



Recency as a factor in determining first choices is shown in 

 table 8 to have played a less important role than the factor of 

 frequency. It is to be remembered that my method is such that 

 it is to the subject's disadvantage to initiate a trial by choosing 

 first the alley that was last entered during the immediately pre- 

 ceding trial. The general tendency of mammalians to repeat 

 the most recently manifested activity on reencountering a given 

 situation might be expected to result in more than 25 of these 

 " recency first choices " as we may conveniently designate them, 

 per 100 trials in subjects who are incapable of learning to react 

 with reference to the disadvantage of such first choices. On the 

 other hand, a subject who has more or less capacity for learning 



