116 



Research in Animal Breeding. 



[May,. 



for the animal breeder. It assures him that a character put into 

 a cross can be recovered from it by suitable procedure, even 

 though it may for a time appear to be submerged and lost. It 

 also offers the prospect of understanding, and so of controlling 

 fully the material with which he works ; but in order that he may 

 be in a position to do this he must first be provided with an 

 analysis of that material in terms of factors. The factor is for 

 the breeder much what the atom is for the chemist. Though 

 they may have a real existence, in practice both the atom and 

 the factor are used as symbols. The chemist analyses his 

 material in terms of atoms that he has never seen, but the con- 

 ception of their existence is justified by the control he obtains 

 when using the atomic theory as his guide. The theory guides 

 the analysis that enables him to build up a conception of the 

 chemical constitution of the substance he examines, and this 

 conception enables him to predict the behaviour of the substance 

 in its various reactions. Understanding the atomic nature of the 

 substance, he can thenceforth control it; so also, the biologist 

 is .attempting to analyse his material in terms of factors which 

 he has never seen. For if factors are something definite and 

 permanent, following a definite scheme of distribution in here- 

 dity, it is clear that the characters of living things can be brought 

 under accurate control by the breeder. They can be dissociated 

 and recombined, just as the chemist dissociates and recombines 

 atoms to make new substances. 



This work of analysing the living beast is only beginning. It is 

 only within recent years -that the factorial theory of heredity was 

 enunciated, and the scientific man is still busy, testing how far 

 it is sound. In simple cases such as those described above, it has 

 certainly borne the test. The skipping of characters for a genera- 

 tion — the persistence of the unwanted recessive even in most 

 highly pedigreed flocks and herds — the unfixable nature of 

 certain types — the explanation of the curious phenomenon of 

 reversion on crossing — the meaning of the break up of the type 

 in the second generation from a cross — the principles governing 

 the recombination of characters — all these things are now 

 straightforward, and will be found treated of in a text book 

 dealing with heredity. But can we interpret in terms of the fac- 

 torial theory those cases, where at first sight there appears no 

 suggestion of clear cut alternative pairs of characters — where a 

 cross seems to. result in a muddled blend — or must we confess 

 that no solution has been found? It is such problems as these 

 that have been engaging our attention at Cambridge for some 

 years past, and a brief account of them will be given in the two 

 following articles. 



