304 The Genetics of the Dutch Rabbit 



factorial hypothesis if we assume the existence of several factors each 

 behaving in a similar way. But in dealing with such characters as 

 those involving size and weight, which are so susceptible to alterations 

 in the environment, the difficulties in the way of complete analysis are 

 practically insuperable, and the interpretation in factorial terms must 

 rest largely upon analogy. 



The difficulties are not, however, so great in another group of cases 

 where we encounter these apparently continuous series, viz. the group of 

 pattern cases in animals, where the coat may range from almost white up 

 to self-colour: As compared with size and weight they offer the great 

 advantage that the pattern is practically constant throughout life, and 

 apparently independent of extraneous influences such as food and 

 warmth. Records can be obtained without the necessity of rearing all 

 animals to maturity. The best known and most fully investigated 

 example of a continuous series among pattern cases is that of Castle's 

 rats, where the evidence shews that animals exhibiting the recessive 

 hooded character may range from almost self-colour down to almost 

 white. Castle's experiments shewed that, by selecting the darker and 

 the lighter forms respectively, strains could be established in which the 

 mean grade of pigmentation was in the one case considerably darker, 

 and in the other considerably lighter than the extremes of pigmentation 

 exhibited by the original stocks From his experiments Castle con- 

 sidered that the factor itself underwent modifications as the result 

 of systematic selection, a view directly in opposition to that of most 

 genetic workers, who regard the conception of the stability of the 

 factor — relative, if not absolute — not only as an invaluable tool for 

 progressive construction, but as firmly based upon experimental fact. 

 The diflFerence between Castle and his critics, however, need no longer 

 be insisted upon, since, in his most recent contribution ^ he has definitely 

 retired from the position which he formerly took up. He admits that 

 the difference between a self-coloured rat and a hooded rat, whether of 

 the highly pigmented or of the nearly white form, is a difference of a 

 single factor: and further that this factor is not influenced by systematic 



1 The "original stock " consisted of relatively few individuals, and those probably with 

 markings near what is demanded by the fancy in a hooded rat, i.e. about midway between 

 the extremes of dark and light. Had it consisted of some thousands, allowed to breed 

 together promiscuously without interference on the part of the fancier, it would probably 

 have exhibited forms comparable with the extremes of dark and light eventually attained 

 by selection. 



2 Castle, W. E., Studies of Heredity in Rabbits, Rats, and Mice. Publ. No. 288, 

 Carnegie Inst., Wash., 1919. 



