42 THE INHERITANCE OF COLOR IN MICE. 



black pigmented animal to produce a blotched black-and-yellow appearance. 

 For the sake of convenience we may call animals possessing this "tortoise" 

 marking t. All self-black animals lacking this modification, and therefore 

 having no red blotches, are of the formula T. Experiments with guinea-pigs 

 (at present being carried on by Dr. Castle) shows that self-black is epistatic to 

 tortoise. This pair of characters is, however, independent from the factors, 

 E = extended black pigmentation, necessary for the production of any black hairs 

 on the coat, and its allelomorph e =the loss of black pigment from the coat, 

 producing 3 r ellow animals. 



Thus, if a race of tortoise animals existed, formula EEtt, we should have a 

 homozygous race of tortoise animals (which in fact exist). If, however, the 

 tortoise animals were of the formula Eett we should expect young of three 

 kinds: 



(a) EEtt tortoise. (b) Eett tortoise. (c) eett yellow.* 



Types (a) and (6) having E present would be able to form black pigment on the 

 coat; but type (c), lacking E, would be yellow in appearance, though it would 

 be potentially tortoise. 



If the factor for spotting with white (s) be added to the formula of a tor- 

 toise animal (Eett), we should have, if these animals were bred together, three 

 kinds of young. 



(a) EEttss tri-color (tortoise and white). 

 (6) Eettss tri-color (tortoise and white), 

 (c) eettss yellow and white.* 



Types (a) and (b) would be tri-color, that is, tortoise and white; type (c) 

 would be yellow and white. This might account for the appearance of yellow 

 and white individuals in a race of tri-colored animals. 



If, however, yellow were epistatic to black, as in mice, we should have to 

 explain things differently. Tortoise animals could then be of only one formula 

 and would breed true, being recessives, whereas self yellows might also produce 

 tortoise young. 



rrtt = tortoise. 



Rrtt = yellow (heterozygous in tortoise). 

 RRtt = yellow (homozygous). 



If, now, the modification s were added to such hypothetical tortoise 

 animals, we should have 



rrttss = tri-color (tortoise and white) homozygous. 



If such animals were crossed inter se, we should, on genetic grounds, have 

 to suppose a change of dominance to explain the occurrence of yellow-and- 

 white young. This does not seem likely. A simple mechanical fact is, how- 

 ever, able to account for the occurrence of yellow-and-white animals in an 

 homozygous or heterozygous race of tortoise-and-white (tri-color) animals. 

 Both forms of spotting are independent as to the regions of the coat which they 

 affect. If the white regions should, by chance, coincide with the black regions 



*Castle (1905), p. 34, has described such animals. 



