MUTATION. 151 



there is always a very great difference between the size of the 

 horns in the two sexes. (Heather sheep). In those breeds 

 where the females have large horns those of the rams are 

 enormous — Dorset, Merino, Algerian. 



In all the breeds the males have a greater tendency to horn- 

 formation than the females, and horn-formation and shape of 

 the horn is much influenced by early castration. Animals of 

 one genotype differ in the amount of horn-formation according 

 to their sex. There is no evidence for the assumption, that all 

 those genes which must influence the growth of the horns, 

 must by their cooperation to the development tend to make 

 the horns grow longer. As we know that absence of horns 

 is more or less dominant in cattle and in goats, we are safe in 

 assuming that there can be genes whose cooperation to the de- 

 velopment tends to make the horns stay shorter, as well as 

 genes whose action results in a lengthening of the horns. 



Not to make the case too complicated, we may imagine a 

 wild breed of sheep, possessing two genes, influencing horn-for- 

 mation, one tending to reduce the size of the horns and one 

 tending to increase the size. We will call these genes A and 

 B, and we imagine the sheep to be all A ABB. The sheep are 

 in an intermediate position in regard to horns, and the males 

 are horned and the females horn-less. If now we imagine, that 

 two different breeds exist, differing from this wild species, 

 one by absence of A and one by absence of B, the result will be 

 that one of these breeds will be of the composition aaBB, and 

 the other will be A Abb. Therefore the horns of the first breed 

 come under the action of the factor B, promoting horn-growth, 

 and in this breed both sexes are horned, although the males 

 will of course have horns longer than the females. The second 

 breed will lack B, and at the same time it has A, which has a 

 reducing effect. This breed will be wholly hornless. If ever we 

 cross the two breeds, A Abb with aaBB, the hybrids will have 

 both A and B, and even if we assume that as heterozygotes 

 they show the effect of a presence of either factor as well as 

 homozygotes, we will see that they will in respect to horn-form- 



