112 



Reseakch in Animal Breeding. 



[May, 



Fig. 4). The Fl generation is uniform, but when a further 

 generation is raised from these, fresh types appear. In addition 

 to black polled and red horned beasts there will be horned blacks 

 and polled reds, types distinct both from parents and grand- 

 parents, but evidently a recombination of characters found in the 

 grandparents. These four types appear in widely different pro- 

 portions, as indicated by the numbers in Fig. 4. It has been 

 pointed out already that polled and horned cattle form a pair of 

 alternative characters, of which polled is dominant ; and that 

 black and red form a similar pair, the black being dominant. 

 Knowing this, we should expect all the Fl beasts to be both black 

 and polled ; the F2 generation to consist of blacks and reds in the 

 ratio 8:1, and polled and horned in the same ratio. If we 

 suppose that the factors for the black-red and the polled-horned 

 pairs are transmitted in the same manner, but independently of 

 one another, we must obtain a F2 generation consisting of the 

 four classes black-polled, black-horned, red-polled and red-horned 

 in the ratio 9 : 3 : 3 : 1. This is the only ratio in which the 

 polled and horned appear in the proportion 3:1, and the blacks and 

 reds simultaneously in the same proportion, provided that each 

 pair is inherited independently. Though the ratio 9:3:3:1 

 has not been verified on a comprehensive scale for the cattle 

 cross, it has been worked out in all details in many cases for 

 smaller animals, where the expenses of breeding are far less. 

 There is reason for supposing that if a F2 generation of several 

 hundreds of cattle were bred from this cross, the four classes 

 mentioned would be obtained in the proportions given above. 



There has been a " break-up " of the parental types in that 

 the two new classes, horned-blacks and polled-reds appear in the 

 F2 generation ; and it is clear that these new classes arise, through 

 recombination of the two pairs of factors in which the original 

 parents differed. The " break-up," however, is not marked, 

 because the parents differed in two pairs of factors only. Had 

 they differed in ten pairs, the F2 generation would have been 

 very much more complex, and the feature of recombination, so 

 obvious in the simpler case, would have been obscured by the 

 great number of recombinations that would have appeared. 

 Nevertheless, on the evidence obtained from smaller animals, 

 there is good reason for supposing that the more complicated 

 case could be resolved on the same lines as the simpler one, 

 and that the same principle underlies both. 



A cross may be undertaken deliberately with the idea of com- 

 bining particular characters found in one breed with other cha- 



