INHERITANCE OF WIIITE-SPOTTINC IX RABBITS. 23 



of the three categories used in testing heterozygous Enghsh males. It 

 will be noted that he produced only English young, however mated, 

 conclusive evidence of his homozygous dominant character. His hetero- 

 zygous English young were of mean grade 4.77 and 4.84 by homozygous 

 English and self does respectively. This male was a son of male 9532, 

 table 41, with whose genetic character his own was very similar, judg- 

 ing by the grade of their heterozygous English young. His line was 

 not continued. 



Table 46 enables one to survey at a glance the summarized results 

 of this entire selection experiment. The course of the experiment is 

 followed only in the male line, because only in the case of the males 

 is the number of young large enough to show beyond question the 

 genetic properties of the individual. From the beginning of the experi- 

 ment fluctuation was observed in the grade of the 3'oung produced, and 

 this fluctuation was in part at least genotypic, since the higher-grade 

 mothers have given higher-grade young in matin gs with the same 

 male. That the fluctuation was also in part lihenotypic is shown by 

 a comparison of the records made by different males of the same 

 grade when mated with the same group of females (tables 31-45). 



The entire selection race derives its English character from cf2711. 

 This animal was a heterozygote deriving the English character in a 

 single gamete from cf2545, who was also heterozygous. Hence the 

 English character had evidently changed in transmission from father 

 to son, a sufficient refutation of the idea of unit-character constancy. 

 AVhether the change resulted from a directly changed unit-factor (gene) 

 or from the introduction of one or more modifying factors is a matter 

 for further consideration. 



The advances made in the male line seem to occur as five successive 

 steps corresponding roughly with generations of selected ancestry 

 (table 46). The first advance comes with the selection of the (single 

 gamete) male 2711, founder of the race; the next in the selection of 

 his grandson, 5555 ; the third in two sons of 5555, viz, 6370 and 6072 ; 

 the fourth occurs in the selection of two sons of 6072, viz, 7699 and 

 9532; the fifth is seen in 1212 and 534, sons of 7699. The direct line 

 of advance is through 2711, 5555, 0072, and 7699. The amount of 

 advance at each step, as indicated by the average grade of the young 

 of these males, is shown in table 47. The rate of advance has evidently 

 decreased as the experiment progressed. 



That modification of the English pattern resulted imme<liately 

 from the cross with self individuals of an unrelated race is conclusi^•ely 

 shown in table 30. The original English male, 2545, i)roduced by Eng- 

 lish mates heterozygous English young of mean grade 2.32; by self 

 mates he produced heterozygous English young of mean gratle 2. SO, 

 practically half a grade higher. By most of such mates the young 

 were more than a grade in advance of those produced by English 



