V. EXPERIMENTAL RESULTS 



(a) TIME AND ERROR AVERAGES FOR ALL TESTS 



Tables I. to VII. give the time and the number of errors, i.e., the 

 number of cases in which the mouse tried to go through a locked 

 gate, which is a measure of the activity of the animal. In this paper 

 the average of the last fifteen trials is used as an index of perform- 

 ance for the first set of seventeen trials in the maze test, and the last 

 23 trials are used as an index for the 25 trials in the multiple choice 

 test, in each case the first two trials being eliminated for irregulari- 

 ties previously mentioned. The groups of two and ten trials each in 

 the interference test, and the group of ten in the retention test are 

 used as indexes for the respective cases. The above tables give the 

 complete records of the 183 mice tested, grouped in families as de- 

 scribed below. The average time is 54.12 P.E. 2.3 seconds per trial 

 for the last fifteen trials in the maze test; 60.26 P.E. 4.7 seconds 

 for the last ten trials of the interference test; 52.81 P.E. 4.7 sec- 

 onds for the retention test, and 39.47 P.E. .08 seconds for the last 

 23 trials of the multiple choice test. The distributions of the indi- 

 viduals in both experiments is shown in Fig. 3. 8 The distribution 

 for the animals in the maze test, based on the average speed attained 

 in the last fifteen trials is indicated by the solid line, and the distribu- 

 tion for those in the multiple choice test, based on the speed attained 

 in the last 23 trials, by the broken line. In the maze test 65 animals 

 took less than 20 seconds, in 47 cases the time was between 20 and 40 

 seconds, and there were 71 cases between 40 and 360 seconds. But 

 one mouse failed to learn the maze. The distribution in the multiple 

 choice test gave 30 cases in which the time was under 20 seconds, 20 

 cases between 20 and 40 seconds, and 26 cases between 40 and 280 

 seconds. None of the mice failed to learn the multiple choice test. 



(&) DIFFERENCE IN LEARNING BETWEEN FAMILIES OF WHITE AND 

 YELLOW MICE 



As reported in the preliminary account of this experiment, it was 

 found that certain strains of mice took considerably longer time to 

 learn the maze than others tested at the same time. Among the 



"Seventy-six of the 183 cases were tested in both the mazes given in these 

 distributions. 



