5<)4 



THE AMERICAN NATURALIST [Vol. XLIII 



of the grandparent carried the chocolate determiner since this 

 was known to be absent in the black grandparent. Hence the 

 yellow germ cells transmit the determiners for other colors. 



Cuenot has objected to this conclusion on the ground that the 

 chocolate grandchild was due to a diluting factor carried by the 

 yellow grandparent. This objection would be valid if choco- 

 late is the dilute form of black, but Miss Durham has shown that 

 the dilute form of black is not chocolate and that chocolate itself 

 has also a dilute form. This relation I have also seen in my 

 experiments. Furthermore had there been a diluting factor in 

 my original yellow, of which there is no evidence, I should have 

 obtained blues and silver fawns in some of the descendants that 

 were inbred for some time but this is not the case. 



It is probable therefore that the yellow color is not the allelo- 

 morph of the other colors but may be transmitted along with 

 them. Its allelomorph would be in Bateson's sense the absence 

 of yellow. Even this assumption fails however to show why pure 

 yellows do not appear, and we must look still further for an 

 explanation of the behavior of yellow in inheritance. 



Castle has made some important suggestions that bear on this 

 question. The gray coat of rabbits is due, according to his 

 analysis to at least five distinct unit characters represented in 

 the formula 



U 



A— C— B— E 

 I 



C is the color producer; A is the factor for ticking; B stands 

 for black; U for uniform (i. e., not spotted) distribution of 

 color; I is the intensifier or strengthener ; and E a factor that 

 governs the extension of black over the body. For a black 

 rabbit the same formula holds with A left out. For a" yellow 

 rabbit E is replaced by K, a factor that stands for the absence 

 of black. A sooty yellow rabbit is like the last with A absent. 



It will be noticed that there is no factor in these formulae for 

 yellow, because yellow is assumed to be present in all these 

 rabbits, but since it has never been lost its claim to be looked 

 upon as a unit character is not established. Castle believes that 

 yellow is always present if C is present. Yellow rabbits there- 

 fore differ from gray, as stated above in the absence of E (not 



