EVOLUTION IN NATURE 167 



recapitulation theory, he cannot fail to admit that 

 a sequence which actually occurs in ontogeny must 

 at least be possible in phylogeny. 



One thing disturbs the perfection of this gradual 

 series. However clear the succession of major 

 structures may be, the microscopic cells producing 

 them present a discontinuity. Bone cells have 

 peculiar properties of their own, as also do cartilage 

 cells and those of the various other types of con- 

 nective tissues. A sharp division of the cells of 

 the neural tube into nerve cells and neuroglia oc- 

 curs during embryonic life. But these diverse ele- 

 ments may be traced to common predecessors also. 

 The apparent abruptness of their separation is 

 probably due to incomplete understanding of the 

 minute details of structure and functions, and on 

 the purely logical grounds of our interpretation of 

 the organism they cannot fail to express potenti- 

 alities of preexisting units. They do not present a 

 linear series such as may be traced in the phylo- 

 genetic history of a single structure, but rather 

 correspond with the divergent products of a single 

 ancestral form. They are not comparable with the 

 foot of the horse and its predecessors, but with all 

 vertebrate appendages, derived from the primitive 

 pentadactyl source. 



The development of discrete structures is some- 

 what closer to the appearance of new characters, 

 but even these are merely the product of preexisting 



