TWO CROSSES BETWEEN DIFFERENT RACES OF PIGEONS. 221 



series of progressive or orthogenetic variations and not with a 

 n-tro^n ive mutation. 



There is another result that seems to me to be especially 

 significant. The white on ihe top of the head of the starling is 

 dominant to the plain (colored) head of tin- turbit . Inn the white 

 win^ bars of the starling are recessive to the dark wing bars of 

 tin- turliit. I'nless these are two different kinds of white, one 

 dominant ' inhibitor of color) and the other recessive ( absence of 

 color: then-Milt is difficult to explain, unles> as I have suggested * 

 tin gation in mosaic types i- not germinal but ontogenetic. 



1 " 'I In- Intlm-: ;ity aiv! : :r"ninciu in I Vlri mining the- ('".it 



' ' . X X I.. Jul> . \<ii i . 



