August 31, 1906.] 



SCIENCE. 



279 



black character, even when yellow is recessive 

 to black. 



The assumption which underlies the ex- 

 planation of color inheritance given by Cue- 

 not, and adopted by Bateson, is that recessives 

 lack altogether a certain factor necessary for 

 the production of the dominant pigment; that 

 albinos, for example, have one factor necessary 

 for the production of pigment, but lack a sec- 

 ond factor altogether. Now granting that two 

 such factors exist (they may or they may not), 

 it is perfectly certain that many albinos pos- 

 sess hath of them. For albino guinea-pigs 

 and Himalayan albino rabbits actually do 

 form hair pigments. There is nothing alto- 

 gether absent from them which is a necessary 

 factor in pigment production. In such cases, 

 what distinguishes an albino mammal from 

 any other sort, so far as our present knowl- 

 edge goes, is a peculiarity in the distribution 

 of the pigments over the body. Albino mam- 

 mals lack pigment in the eye; what pigment 

 they form is found at the extremities of the 

 body. 



Again, white-plumaged birds do not lach 

 altogether some factor necessary for pigment 

 formation. They invariably have pigmented 

 eyes, but commonly lack pigment in their 

 feathers. Nevertheless, the purest strains of 

 white fowls are proverbially prone to form a 

 ' black feather ' occasionally. Can we say 

 that the character, black plumage, is alto- 

 gether laching in white fowls ? It is not 

 present as a recessive character — I speak of 

 established races, not cross-breds. Shall we 

 say, with Cuenot, that an occasional black 

 feather is a variation not heritable? By no 

 mean^ In the white fowl with a black feather 

 the bi. ..c character is present, every factor of 

 it ! Those factors were likewise present in 

 the white-plumaged parents of the bird in 

 question, but they functioned less actively, so 

 that no feather may have shown the black, 

 though pigment was formed in the eye. The 

 hypothesis of absent factors is inadequate to 

 explain the observed facts, in at least a ma- 

 jority of known cases. By cross-breeding and 

 selection we can alter the proportions of the 

 different pigments in the coat without elim- 



inating any. By repeated crossing of black 

 with yellow, in guinea-pigs, we can weaken 

 the activity of the black, so that while black 

 pigment is still formed all over the body, it is 

 formed in less amount. The black pigment is 

 found in greatest amount at the extremities. 

 If the process of weakening the black pig- 

 mentation is capable of being carried to its 

 logical conclusion (a matter still under in- 

 vestigation), black pigment shoTild finally dis- 

 appear except at the extremities. Again, cross- 

 breeding albino guinea-pigs with blacks in- 

 creases the amount of black pigmentation 

 formed at the extremities by the albinos, and 

 induces a slight pigmentation of the coat gen- 

 erally, as I have elsewhere shown. How far 

 the contamination of the albinos can be car- 

 ried, I am unable as yet to say. 



Facts such as these are difficult to explain 

 on the hypothesis of two or more factors sepa- 

 rately heritable, imless we suppose further 

 that those factors are inherited in varying 

 degrees or amounts. Explanation is rendered 

 still more difficult when we come to consider 

 characters other than pigments, such, for ex- 

 ample, as polydactylism. For this reason I 

 have carefully avoided incorporating the ter- 

 minology of the ' factor ' hypothesis into my 

 published papers on heredity, and have re- 

 ferred to characters as more or less completely 

 active, or in some cases as latent, a usage in 

 harmony with that of Tschermak (1906). 

 Bateson, Saunders and Punnett (1906) have 

 recently criticized this usage, but I believe 

 time will justify present caution in the adop- 

 tion of the factor hypothesis. Morgan (1905) 

 has attempted to corabine the latency idea with 

 the factor idea, but with results not very satis- 

 factory from either point of view. 



It seems to me, rather, that we must recog- 

 nize, along with the fundamental principle of 

 unit characters in heredity, the fact that char- 

 acters may exist in varying states of activity. 



The presence of one character often inhibits 

 the activity of another. .When in guinea-pigs 

 the characters black and yellow are present 

 together, yellow is largely inhibited. If yel- 

 low is made more active by repeated crossing 

 with homozygous yellow, black is partially in- 



