722 



THE AMERICAN NATURALIST [Vol. XLVII 



of a woody plant repeats little of the histological features 

 of its predecessors, but it does recapitulate the general 

 vascular topography of successive ancestral forms. The 

 developing organism has concentrated within it an essence, 

 so to speak, of the most conservative and therefore the 

 most salient characters which distinguished the ancient 

 members of its line. The fact that all plastic and highly 

 variable features have been swept away enables these his- 

 torical landmarks to stand out distinctly, and gives to the 

 structure of the animal embryo and of the young plant 

 a very important significance in the science of phylogeny. 



The principle that fixity of character increases with 

 differentiation, which we have regarded as of so much im- 

 portance in evolution, is easier to establish than to explain. 

 It is possible to regard the matter from the viewpoint of 

 genetics and to imagine that a "variable" species is a 

 "mixed population," the members of which are continu- 

 ally intercrossing, and that the appearance, in certain indi- 

 viduals, of definite discontinuous variations isolates such 

 individuals from the rest of the species and causes the 

 partial or complete establishment of each as a distinct 

 "pure line" with more closely defined characters. The 

 more numerous such discontinuous variations were, the 

 more complete the isolation of a given line would become 

 and the more purely, therefore, would it reproduce itself 

 until finally its characters became very sharply fixed. In 

 other words, fixity may be due to germinal segregation 

 and may depend directly on the proportion of factors 

 which are in a homozygous condition in the germ plasm of 

 the two parents. Complete homozygosity in both would 

 ensure complete fixity of parental characters in the 

 offspring. 



A comparison also suggests itself between the effects 

 of differentiation in ontogeny and in phylogeny. Experi- 

 mental work has shown that in the more primitive ani- 

 mals, where the power of regulation is best developed, any 

 part of a tissue or elementary organ, so long as it remains 



