HEREDITY OF WILDNESS AND SAVAGENESS IN MICE 27 



3. Litters were equally divided and the individuals of 

 one group were tested every day or every two days until the 

 five tests were given, while the individuals of the other group 

 were tested at the usual times. 



The division of the mice in this experiment depended entirely 

 upon the order in which the mice were caught for their first 

 test. The first mouse was put in the cage of those to be tested 

 at the usual times and the second was put into the cage of those 

 to be tested daily. Hence, each litter was divided into two 

 practically equal groups which, having the same ancestry, 

 should form a basis for a fairly accurate estimate of the effect 

 of frequency of tests upon the lowering of the grades of wild- 

 ness and savageness. 



The first mice to be used in this experiment were the 12 wild 

 mice raised in the laboratory and later used as wild parental 

 stock (table 1). There were two litters of six mice each. One 

 litter had four males and two females and the other two males 

 and four females. They were therefore not divided according 

 to chance, as has just been mentioned but into two groups of 

 three males and three females each. The results of the group 

 with the usual time intervening between the tests (Group 1) 

 are given in table 16, and those receiving daily tests (Group 

 2) in table 17. In the case of wildness, when the results of the 

 males and females are considered together as are shown in the 

 first part of the tables, the averages are identical with the ex- 

 ception of that for the fifth test which is 0.34 grade greater 

 for the mice tested daily. The averages of the first test for 

 savageness are also the same for both groups, and the remain- 

 ing compare very favorably, the difference being in no case 

 greater than 0.25 grade. When the results for the males and 

 females are considered separately the females of Group 2 are 

 seen to grade, as a rule, slightly higher than those of Group 1. 



The sixth and seventh tests of the mice of Group 2 were given 

 at the same time the mice of Group 1 were receiving their fourth 

 and fifth tests. Assuming that the three factors, namely, 

 age, presence of the experimenter while feeding and testing, and 

 the handling during the tests, each have a certain effect on the 

 lowering of the grades in successive tests; a comparison of the 

 averages of the fourth and the fifth tests of the mice of Group 

 1 with the averages of the sixth and seventh tests of those of 



