THE RECAPITULATION THEORY 357 



Metazoa ; that it occurs in all Metazoa, and that 

 when asexual reproduction, as by budding, &c., 

 occurs, this merely alternates with the sexual pro- 

 cess which sooner or later becomes essential. (!_ 

 the fundamental importance of sexual reproduction 

 to the welfare of the"speries be granted, and if it 

 be further admitted that Metazoa are descended 

 from Protozoa, then we see that there is really a 

 constraining force of a most powerful nature com- 

 pelling every animal to commence its life-history in 

 the unicellular condition, the only condition in 

 which the advantage of ^cross^fertilisation can be 

 obtained *.*., constraining every animal to begin 

 its development at its ^earlist_ancestral stage, at 

 the very bottom of its genealogical tree. 



On this view the actual development of any 

 animal is strictly limited at both ends ; it must 

 commence as an egg, and it must enoTirf the like- 

 ness of the parent. The problem ofrecapitulation 

 becomes thereby greatly narrowed ; all that remains 

 being to explain why the intermediate stages in the 

 actual development should repeat ^he^intermediate 

 stages of the ancestral history. Although narrowed 

 in this way, the problem still remains__one of ex- 

 treme difficulty.\ 



It is a consequence of the Theory of Natural 

 Selection that identity of structure involves com- 

 munity of descent : a given result_can only be 

 arrived at through a given sequence of events : 

 the same morphological goal cannot be reached 

 by two independent paths. A negro and a white 

 man have had common ancestors in the past ; 



