BLASTOGENIC VARIATIONS. 145 



culiar to a rock or other pigeon, as such, is one unit. 

 Also let it be granted that the characters separating any 

 variety of the pigeon from the ancestral rock pigeon are 

 of the same value as those separating species of the same 

 genus, namely, one unit. Now in the gradual course of 

 evolution of a domestic variety of pigeon from a rock 

 pigeon, we may assume that the total amount of germ- 

 plasm bearing hereditary characters has remained prac- 

 tically constant, and hence, as one unit of determinants 

 has been added on to the rock pigeon germ-plasm, one 

 must have disappeared. Now did this unit of deter- 

 minants corresponding to the characters of the domes- 

 tic variety of pigeon replace that of the rock pigeon, or 

 was it superimposed on it ? Embryology seems to teach 

 us that once any character is, as it were, laid down in 

 the germ-plasm, it is fixed there, and as a rule only very 

 slowly dwindles away by a process of gradual dilution 

 by subsequent ontogenetic stages. It seems reasonable 

 to assume, therefore, that the determinants are re- 

 placed in proportion to the relative amounts of them 

 present, or that, on an average, -ffo of the replaced 

 unit concern the sum total of hereditary characters 

 which go to constitute the species pigeon, and y^ 

 those peculiar to the species blue rock pigeon. The 

 germ-plasm of a domestic pigeon will therefore be made 

 up of 1 unit of determinants corresponding to the char- 

 acters domestic pigeon, .99 of a unit corresponding to 

 the characters Hue rock pigeon, and 98.01 units cor- 

 responding to the characters species pigeon. It there- 

 fore follows that the hereditary characters of the an- 

 cestral rock pigeon are almost as strongly represented 

 in the germ-plasm of a domestic pigeon as they were 



