Wright: Color Inheritance in Mammals 



523 



three kinds of whites, the following per 

 cents of red, roan and white offspring are 

 expected. The percentages expected in 

 the simple one factor hypotheses that 

 roan is the heterozygote between red and 

 white are also given and finally, the per 

 cents actually found as given in Went- 

 worth's data quoted above. 



The striking fact that in a popula- 

 tion breeding nearly at random, roan by 

 roan produces little more than 50% roan, 

 the rest being divided between red and 

 white, while red by white produces 

 about 96% roan, can hardly be ac- 

 counted for on any theory of inheritance 

 other than a single main Mendelian 

 factor without dominance. The excep- 

 tions must be accounted for on hypo- 

 theses subsidiary to this, as is done by 

 Wilson and by Walther.^ The two 

 factor hypothesis gives expectations far 

 removed from the actual results and the 

 writer can devise no other hypothesis 

 involving multiple factors of anything 

 like equal influence which is appreciably 

 better. Finally, an additional test may 

 be noted which the one factor hypothesis 

 meets quite successfully. With 756 

 whites (WW) and 4,169 reds (ww) in a 

 population in which there is random mat- 

 ing, the number of roans (Ww) should 

 be twice the square root of the product, 

 or 3 , 5 5 1 . The actual number 3 , 7 80 is as 

 close as can be expected, considering the 

 departures from random mating in the 

 present data. 



VARIATIONS OF ROAN 



As for the discrepancies, part are 

 known to be due to errors in the herd 



books. A clue to the remainder is seen 

 in a consideration of other breeds which 

 show the nearly solid white condition. 



An excellent review of the subject has 

 recently been written by Lloyd-Jones 

 and Evvard.fi The famous Chilling- 

 ham cattle are white with dark points. 

 They occasionally throw black or dun 

 colored calves. Here white is clearly 

 dominant. A very similar white with 

 dark points crops out from time to time 

 in Pembroke and in Highland cattle, 

 although constantly selected against. 

 Here as Lloyd-Jones and Eward point 

 out white is clearly recessive. Thus we 

 have whites of very similar appearance 

 which are respectively recessive, domi- 

 nant or neither in Pembroke, Park cattle 

 and Shorthorns. Is it necessary to sup- 

 pose that three independent factors are 

 involved? The simplest explanation 

 seems to be that there is a continuous 

 series of physiological conditions be- 

 tween a very strong self tendency 

 through roan to a very strong white 

 tendency. The level in this series 

 possessed by a particular group of epi- 

 dermal cells of any animal is determined 

 perhaps by many factors, some merely 

 developmental (making it possible for 

 adjacent hairs to have different potencies 

 and so produce the roan effect) and some 

 hereditary. Among the hereditary fac- 

 tors, two hypotheses are at present 

 equally possible. There may be one 

 Mendelian pair (W,w of this paper) 

 which stands out in the importance of 

 the effect but which, nevertheless, co- 

 operates with lesser factors which deter- 

 mine the general level in the series; or 

 there may be a series of allelomorphs 

 such that a more potent white factor is 

 dominant over a self factor which is just 

 below the roan level and vice versa. The 

 two hypotheses would give very differ- 

 ent results but there seem to be no data 

 at present by which they may be distin- 

 guished. 



This roan factor quite certainly 

 belongs in class lai in spite of its variable 

 dominance. It is a factor which deter- 

 mines an inhibitor of color regardless of 

 the quality of the color. The truth of 



sWalther, Ad. 1913. Zeit. f. Ahst. u. Ver., \(i:\-A^. 



« Lloyd-Jones O. and Eward. 1916. Res. Bull. No. 30, Ag. Exp. Sta., Iowa State Coll. Agr. 



