40 INDIVIDUAL DIFFERENCES AND FAMILY BESEMSLANCES 



this assumption. This case indicates that an experimenter can pre- 

 dict the actions of an animal to only a limited degree, for, after mak- 

 ing consistantly good records in any one task, some external or in- 

 ternal factors, unknown to the investigator, may become operative 

 and break up the expected sequence of learning. It also emphasizes 

 the importance of testing an animal in more than one task before we 

 finally grade its behavior. 



If space permitted it would be interesting to give in detail the 

 learning curves of mice that illustrate still other kinds of individual 

 differences. It may be well to call attention to the average records 

 of mouse No. 147 (see Table III.), that failed completely to make 

 a successful trip through the maze, although it was tried in every 

 one of the tests. It finally succeeded in making the poor average 

 record of 159 seconds in the multiple choice test, but when retested 

 at the time the retention test was given to the other mice, it again 

 failed completely in the maze test. Because of the great amount of 

 individual differences among the animals, it is difficult to find the 

 record of a single mouse that made a record that was similar to the 

 average for each task. Only No. 133, whose averages are given in 

 Table VI., and No. 189, in Table III., approximated an "average" 

 record. The record of No. 142, in Table III., is interesting in that 

 it shows consistently uniform performance throughout all the tasks. 

 This mouse made the following averages: 36 seconds in the initial 

 learning period, 24 seconds for the first two interference trials, 36 

 seconds for the last ten interference trials, 35 seconds for the mul- 

 tiple choice test, and 33 seconds for the retention test. Another 

 type of animal behavior is to be seen in the case of mice that made 

 very poor records in the initial maze tests, and good records in the 

 multiple choice and retention tests. 14 



The multiple choice test, which did not favor the production of a 

 stereotyped form of reaction, nevertheless showed some interesting 

 types of individual and group responses. It was found that the 

 animals exhibited three types of behavior, which were not definite, 

 but merged one into the other, so that an animal might use one of 

 them on a certain day and another type on the next. The types of 

 reactions were as follows : 



Type 1. Response by slowly and carefully "examining," visu- 

 ally or otherwise, each of the colored doors, very suddenly becoming 

 oriented and dashing through the successful red gate. 



Type 2. The animal would react by trying the gates in a definite 



14 See records of the following mice in Table III., Nos. 138, 153, 154, 165, 

 171, 178, 181, 187, 188, 206, 207, 209, 210, 211, 212, and in Table IV., No. 169. 



