THE PRODUCTION AND FIXATION OF NEW BliF.EDS. 



41 



(Fig. 3). A single cross of black with red (when not attended by 

 reversion, i. e., when the factor A is not present) produces commonly 

 black individuals of a reddish hue, as already stated. These individ- 

 uals form gametes of the two sorts, B and R. Repeated crossing, how- 

 ever, is apt to produce blacks which bear patches of red hair inter- 

 spersed with the black. Such animals we ma}* call brindles. Breed- 

 ing these together we may obtain (1) brindles which produce brindles 

 and blacks (B-BR, Fig. 3) ; (2) those which produce brindles and 

 reds (R-BR, Fig. 3) ; and (3) brindles which breed true, i. e., are 

 fixed (BR-BR, Fig. 3). The fixation is really due (1) to contamina- 

 tion of black with red, so that gametes containing black contain also 

 fragments of red along with the black; and (2) to the formation of a 

 zygote out of such contaminated gametes. 



To recapitulate, where Mendelian (or alternative) inheritance is 

 involved, new characters are produced by cross breeding in one of two 

 ways : First, a new character may be produced by the becoming visible 

 of an element previously invisible in one of the parents; or second, 

 the characters brought together by the cross may form a mosaic dif- 

 ferent from either parental condition by itself. Fixation of the new 

 character will consist in either case in securing gametes which sep- 

 arately contain all the factors necessary for the production of the 

 new character. 



The question now arises: How long will it take to secure fixation? 

 To this we may reply : All will depend upon how the new character 

 has been produced; if by bringing together factors previously sep- 

 arated (as in the agouti reversion), complete fixation may be secured 

 in two generations. If, however, the new character is a mosaic of 

 alternative conditions (as of self-red and self-black) fixation will 

 probably require much longer ; one cannot in advance say how long. 

 In this case, gametic contamination is involved, the production, out of 

 a B R zygote, of B gametes containing fragments of R, or of R 

 gametes containing fragments of B, instead of the theoretical pure B 

 and pure R gametes. We cannot say in advance how soon this con- 

 tamination will occur. All we can say is that the more sharply 

 alternative the inheritance and the more complete the dominance in 

 a first cross, the less likely is contamination to occur, for experience 

 shows that where dominance is most perfect, segregation is most com- 

 plete. Gametic contamination is non-Mendelian, it borders on blend- 

 ing inheritance, and will occur most speedily when the inheritance 

 deviates most from the Mendelian type. When the inheritance is com- 

 pletely blending no fixation at all of the hybrid character is required; 

 it is already fixed. 



