R. C. PUNNETT 305 



selection as he had previously held. That the recessive hooded form 

 may shew such a wide range of variation, from an almost fully pig- 

 mented to a nearly white type, he attributes to " residual heredity." 

 This term, so far as I can understand it, I take to mean practically the 

 same thing as the " modifying factors " of Morgan and other critics 

 of Castle, When, by selection, a markedly darker strain of hooded 

 rats is formed from average individuals, we may suppose that a larger 

 proportion of factors whose modifying activity is towards fuller pig- 

 mentation has been accumulated in that strain, the members of the 

 strain remaining unchanged in respect of the factor that determines 

 the difference between self-colour and the hooded pattern. If this 

 is what is meant by saying that an alteration in " residual heredity " has 

 taken place in these selected rats, it is clear that Castle's position is 

 brought into line with a strict factorial interpretation of the continuous 

 series of grades of pigmentation exhibited by hooded rats. The next 

 step in the analysis of this celebrated case is obviously to attempt 

 to define the number and scope of these postulated modifying factors, 

 an experimental task which is likely to demand a very considerable 

 expenditure of time and labour. 



Meanwhile we may consider Castle's other series of experiments 

 with an animal shewing an apparently continuous series of graded 

 patterns, viz. that dealing with Dutch rabbits, of which an account 

 has just been published ^ I am led to discuss this series in some detail 

 because I have been investigating the case for the past twelve years^ 

 and while in general my results are in accordance with those obtained 

 by Castle I have been led to adopt an interpretation different from that 

 which he has just put forward. My experiments are not yet complete, 

 and probably several years must elapse before I am in a position to 

 present a full account. Nevertheless I have gained sufficient experience 

 to venture upon a tentative interpretation, and I feel that, in the interest 

 of the case itself, as well as of other workers on similar lines, it would 

 be of service if I put it forward in conjunction with the data presented 

 in Castle's valuable memoir. 



In order to facilitate discussion I have taken the liberty of re- 

 producing one of the plates accompanying his paper. On this (Fig. 1), 

 as will be seen, Castle arranges his animals in 18 grades with an almost 

 self-coloured animal at one extreme (grade 1), and at the other a white 



1 Op. cit. 



- My experiments were started in 1907. From 1911 till the outbreak of war I was 

 helped by the late Majof P. G. Bailey. Since then I have continued to carry them on 

 alone. 



20—2 



