No. 540] INHERITANCE OF COLOR IN CATTLE 731 



reconciled with gametic purity. The old hypothesis is, 

 therefore, abandoned. The facts demand the companion- 

 trait or unit-complex hypothesis modified to permit of 

 occasional intra-zygotic reactions in response to a definite 

 set of conditions, instead of the single-unit notion, and 

 the 45 case matings instead of the typical 6. 



The following table is compiled from data reported in 

 Biometrika, 1905, 1906, 3 by Amy Barrington and Karl 

 Pearson from Coates's Shorthorn Herd Book. 



TABLE VII 





Offspring 















1. Red by red 



156 

 >24?> 

 1 



104 



23 

 2 



1 







6 

 324 

 85 



11 



39 



69 



40 

 409 



398 

 49 

 







1 



196 

 656 



2. Red by roan 



3. Red by white 



4. Roan by roan 



5. Roan by white 



6. White by white 



In addition to this, special search 4 for white-by-white 

 matings yielded 91 cases, giving 1 red, 4 roan and 86 

 white offspring; in two cases, red-by-red matings were 

 reported to have given white offspring. 



The color distribution of this table practically parallels 

 that found in the Cruikshank herd as reported by Mr. 

 Bruce and that of the other compilations herein recorded. 



Barrington and Pearson then proceeded with the fol- 

 lowing criticism: 



... No simple Mendelian formula applies rigidly. We find our- 

 selves neglecting sensible percentages of occurrences incompatible with 

 the theory of the pure gamete. 5 



It is true that the first studies in color in Shorthorn 

 cattle suggested the single-unit color coat hypothesis and 

 as late as 1909 James Wilson, of the Royal College of 

 Dublin, in his interesting book on the ' ' Evolution of British 

 Cattle," suggested this hypothesis. It is not, however, 

 the sole possible Mendelian interpretation, but rather the 



'Ibid., pp. 427-464. 

 * Ibid., p. 441. 

 8 Ibid., p. 454. 



