CASTLE. — INBREEDING, CROSS-BREEDING, AND SELECTION. 751 



TABLE VI. 



Effect of a Cross between Two Different Lines (A and A') 

 of the Same Inbred Stock. 



Accordingly it appears that separation for eleven generations had not 

 made the two sub-families of the A race sufficiently unlike to profit by a 

 cross between them. The cross-breds were no more productive than the 

 more productive parental sub-family. 



But any two unrelated inbred families are clearly mutually benefited 

 by a cross between them. Their offspring are more productive than 

 either parental race (Tables III-V). 



Meetings of Cousins not more Fruitful than those of Brother and Sister. 



That a mating between cousins of an inbred race is not more fruitful 

 than a mating between brother and sister of the same race has been 

 shown by experiments among which only the following need be mentioned. 

 Twelve matings were made between cousins in generation 28 of the A 

 series. Three of these were sterile and two others produced incomplete 

 broods of young. The average brood of the remaining seven pairs was 

 105. The uncrossed A race, generation 29, reared under identical con- 

 ditions, consisted of four sterile pairs and nine fertile ones. The latter 

 produced broods averaging 153 in number. Further, the young of the 

 cousins bred inter se gave the following result. Four out of fourteen 

 pairs were sterile, the remaining ten produced an average of 112 young. 

 The uncrossed A race, generation 30, reared simultaneously and under 

 identical conditions, gave broods averaging 101, produced by five fertile 

 pairs, three pairs being sterile. 



Cyclical Variations in Fertility and their Relation to 

 External Conditions and to Inbreeding. 



One of the most striking features of the record of the A series is the 

 cyclical rise and fall of the productiveness. (See Figure 1.) This was 



