INHERITANCE OF WHITE-SPOTTING IN RABHITS. 21 



does, but it must be borne in niind that the mothers were nf)t all 

 identical in the two cases. Black-and-tan does seemed in general to 

 give lower-grade oftsjiring than does of the self black and Himalayan 

 races, with which cf 2711 had been mated. 



We next used as sire in the selection experiment cfoSTo, generation 

 2^, grade 4|, a son of cf 5086 by his half sister. (8ee table 33.) He 

 was mated with 9 different heterozygous English does, all but one of 

 which had also been mated with his father. They produced English 

 young of somewhat lower mean grade than they had borne bv the 

 father, 6^5086. (See tables 32 and 33.) 17 homozygous English 

 young were of mean grade 1 .20; 29 heterozygous English young were of 

 mean grade 3.79; there were also 25 self young. Again the higher- 

 grade mothers produced the higher-grade young. Hence there wa.s 

 evidently material favorable for selection among the mothers, if not 

 among the fathers. This male was now discarded and replaced by 

 an own brother of slightly higher grade, viz, cf 5555, generation 2^, 

 grade 4f . (See table 34.) 



This male (cf5555) was bred more extensively than any of his 

 predecessors and produced higher-grade offspring. He was mated to 

 the same does as his father and grandfather and also to a number of 

 new ones which now^ became available. By all classes of does he pro- 

 duced higher-grade 3'oung than had an}- of his predecessors. He also 

 produced higher-grade young by his mates of higher grade than by 

 his lower-grade mates. (Table 34.) In his case, then, a second 

 advance had been made by selection in the male line and the necessary 

 variation was evidently' present to make possible similar advances by 

 selection in the female line. JNIale 5555 produced 41 homozygous 

 English young, 125 heterozj-gous English, 2 ungraded English, and 

 65 selfs, or all together 168 English and 65 selfs. 



The next sire tested was a son of cf5555; viz, 6^6370, grade 5, 

 generation 3, as regards selected ancestry. (See table 35.) He was 

 more advanced in grade and generations than any male thus far 

 tested and produced higher-grade young by females of the same 

 grade. Many of the older females had now been discarded, but 

 enough remained to form a standard of comi:)arison between the 

 genetic properties of this male and those of his predecessors. The 

 mean grade of the heterozygous English 3'oung of this male was 4.66; 

 the grade of his homozygous English young was 1.79. The corre- 

 sponding figures for his father were 4.40 and 1.36 respectively. 



Three other sons of cf 5555 were also tested by matings with sub- 

 stantially the same group of does, although tests in the case of the less 

 promising ones were terminated sooner. Male 6420 (table 36) was of 

 slightly lower grade than his father and was found to be probably 

 inferior to him and so was soon discarded. His heterozygous English 

 young were of mean grade 4.33. 



