36 THE EVOLUTION OF THE METAZOA 



The situation is usually such that the grown-up animal 

 types— actually with widely different kinds of structure— live 

 on the floor of the sea, lead a benthonic life, while at the 

 same time their embryos have developed into planktonic 

 larvae. The way that this case is presented here shows that 

 we consider the life on the floor of the sea to be primary, 

 and that in the open water, i.e. in the plankton, to be a 

 secondary result of the development. In fact, the larval 

 stage represents a prolongation of the ontogeny. In no way can we 

 consider these larvae to be ancestors of any recent animals 

 that live on the floor of the sea. We have no recapitulation 

 of some stage reached by their ancestors, as is the case with 

 some other stages in ontogeny In spite of this the larval 

 stages can be of good use in our study of phylogeny. Larvae 

 of some animal groups, e.g. those of Crustacea, have very 

 characteristic forms and structures (in this instance the so- 

 called nauplius) . These forms are so firmly established in the 

 whole that they are present even when the grow^n-up forms had 

 been much changed so that they can hardly be identified — 

 whether this change be due to an adaptation of the benthonic 

 way of life or to a strong parasitism. A similar case can be 

 observed in Tunicata (Kowalevski). In such cases it is our 

 right to consider larvae as evidence that species do belong 

 together whose larvae have the same forms and structures. 

 It would, however, be completely wrong to conclude on 

 the basis of these facts that an ancestral form is being recapi- 

 tulated in a nauplius. The example of Crustacea itself shows 

 that it is not impossible that the larval stage, the nauplius, 

 disappears. In the transition from the sea to fresh water, the 

 free larval stage — the nauplius— becomes embryonaUzed. Fre- 

 quently, this change is hardly noticeable. The free living stage 

 of ontogeny had possibly originated in various levels which the 

 evolution of the animal had reached, and it can also depart 

 more or less from the "normal" type. Not infrequently (e.g. in 

 insects) we find the main part of the individual existence 

 transferred to the larval stage. In these instances the 



