1921.] Kesearch in Animal Breeding. 333 



that some of them may turn out to have an economic value. 

 For if a visible character, such as colour or pattern, were linked 

 with such a character as higher milk yield, or more succulent 

 mutton, breeding for these latter characters would be greatly 

 facilitated. 



Of our investigations into broodiness we can say little more 

 than that they have shown the character to be a complex one. 

 Hens vary greatly in this respect. Some go fully broody each 

 year; others go broody occasionally and for a few days only; 

 and all intermediate grades exist. Nevertheless our experience 

 affords grounds for supposing that the character can be analysed 

 and expressed in terms of definite factors, though it is clear 

 that the experimental work demanded would be both long and 

 tedious. For in broodiness, as in egg-colour, the case is compli- 

 cated by the circumstances that the factors are carried and 

 transmitted by the cock, though he neither goes broody nor lays 

 eggs. And the cock can only be analysed by mating him with 

 hens of known constitution, and testing the nature of his female 

 progeny — which takes time. 



Incidentally our experiments elicited a fact of some interest 

 to poultry keepers. It is well known that the typically non- 

 broody races lay white eggs, while the races that lay brown eggs 

 belong to the broody section. It has been held that broodiness 

 is necessarily correlated with the brown egg, and that it is not 

 possible to establish a non-broody brown-egg race. Our experi- 

 ments do not bear this out. It is true that the brown egg may 

 be correlated with broodiness ; nevertheless, the linkage between 

 the factors concerned, if it exists, is not complete, for we suc- 

 ceeded in combining the full brown egg with the non-broody 

 character. We do not doubt therefore that, by working on the 

 right lines, a non -broody race laying brown e^ztrs could be 

 established. 



As a by-product of the above investigation we obtained data 

 on the inheritance of leg-feathering. For the Langshan is a 

 breed with feathered legs, while the Leghorns and the 11am- 

 burglis are clean shanked. Our own data, taken in conjunction 

 witli tliosc collcctcMl hy other observers, have served to show that 

 at any rate two factors are concerned in connection with tliis 

 character. As with weight and egg-colour, the factors produce 

 a cumulative* efT(M't, and a continuous stories is to be found 

 ranging fi'om excessive* de\(*lo])uient in birds pure for both 

 factors, to absence of leg-feathering wIkm'c neither factor is 

 present. For a detailiMl discussion, howeni^i-. the rt^adtM* may be 



