14 CHARLES A. COBURN 



The first generation hybrids In each series were obtained 

 by crossing wild females with tame males and tame females 

 with wild males. The second generations of hybrids were 

 obtained by crossing the first generation hybrids among them- 

 selves without selection as to wlldness or savageness, except 

 in the cases where selective breeding was tried, the results of 

 which will be presented in separate tables. Likewise by cross- 

 ing the second generation hybrids among themselves the third 

 generation of hybrids were obtained. 



It was aimed to test each individual five times. However, 

 there was a number which because of death or other reason 

 did not receive that number. Those that received but one 

 test have not been considered in the results. On the other 

 hand, for reasons which will be discussed later, some mice 

 received six to eight tests. 



Deducting those Individuals used in experiments to deter- 

 mine the effect of age on the lowering of grades of wlldness 

 and savageness in the successive tests, the average age for the 

 first test was 38.8 days, and for the fifth test, 92.8 days. The 

 average time between the first and second tests was 17 days; 

 between the second and third, 15 days; between the third and 

 fourth and between the fourth and fifth, 11 days each. 



EXPERIMENTAL RESULTS 



A. General comparison of inheritance in the hybrid generations 



of hath series 



In table 3 Is presented a general summary of the results for 

 the three generations of both series. A comparison of the 

 averages in each of the generations shows a very great simi- 

 larity, the difference in any case being so small as to be practi- 

 cally negligible. However, the difference in the grades of 

 wlldness and savageness attained in the first and last test is 

 significant. This lowering of the grades is gradual in each 

 generation but Is greatest in both wlldness and savageness in 

 the F2's. The model grade, i.e., the grade which is the most 

 frequent, is either 3 or 4 for wlldness In each generation, and 

 for savageness ranges from 3 to 5. 



In tables 4 and 5 the results of the series are presented sepa- 

 rately. The similarities of the averages in the generations so 



