MELANIN COLOR FORMATION. 335 



tissues and organs, together with environmental conditions ex- 

 ternal and internal, supply whatever else is concerned in color 

 production. 1 



In his work on mice Castle ('06) shows conclusively that in 

 these forms he is not dealing with the total absence of a character, 

 even in some cases where this seems apparent ; he there states 

 that the purity of gametes does not exist, and further, " no more 

 does the purity of factors exist." It would seem that this paper 

 by Castle is one of the best, if not the best document extant to 

 convince that no such things as factors exist. Yet, quite recently, 2 

 Castle has carried the factor hypothesis to its highest state of 

 complexity, to what is apparently its logical conclusion, if any 

 "particle" basis whatever could be granted for hereditary proc- 

 esses. Castle pictures his conception of the factors in melanin 

 color inheritance and their relations in the following diagram, 

 modelled after a chemical formula : 



S^U 3 



I C = Color, or chromogen factor. E = Extension factor. 



A C^ R F A = Agouti factor. I = = Intensity " 



B =. Black U = Uniformity " 



D ^p D= Dilution " S = Spotted 



This visualizes for us the body of factors which are to be 

 shuffled in the germs and " determine " the colors of the progeny. 

 From what has been said it is obvious how far this conception 

 leads us astray ; because this complex Medelian interpretation 

 and description of color inheritance leads us away from a simple 

 series of color developments due to the difference in degree to 

 which one substance may be oxidized. 



The placing of the " uniformity-spotted," "intensity-dilution," 

 and " extension " factors in these germs is a virtual surrender of 

 the whole theory of discontinuous variation ; and in reality puts 



'The many and accumulating additions, qualifications, " contaminations," " laten- 

 cies," etc., that have been attached to Mendel's ('65) original conception of domi- 

 nance and recessiveness, or to Cuenot's ('03) presence and absence of a factor 

 hypothesis, are but so many direct admissions that the "purity of gametes" concep- 

 tion is an error ; they are but so many secondary and tertiary hypotheses of completer 

 preformation made to bolster up a primal preformation hypothesis. Incidentally, they 

 give present-day students the opportunity to see the child of Weismannism recapitulate 

 the developmental history of the parent. 



2 Darwinian lecture, Baltimore, January I, 1909. 



3 Either one of the pair may be present (or active). 



