CNIDARIA AS THE ONLY COELENTER.\TA 43 



are no recapitulation of some grown-up ancestors. In the 

 pilidium the first genuine metamorphosis can be observed: in 

 this "catastrophic" metamorphosis the worm develops out of 

 special and quite separate Anlagen and in an entirely different 

 orientation so that finally a considerable part of the body is 

 destroyed, (see Fig. 6). 



Fig. 6. Nemertinan planktonic larva, called the pilidium; (1), the gut, 



(2), primordium (A.nlage) of the proboscis; (3 — 5), primordia of 



the "worm-body". 



I repeat that comparative morphology which tries to find 

 homologies and evaluate relationship connections on the basis 

 of such homologies can with full right make use of onto- 

 genetic morphogeny and thus the larval stages. Yet, first, it 

 must not seek in these some recapitulation of the morphogeny 

 of their grown-up ancestors, and, secondly, it should not lay 

 too great an emphasis on the ontogenetic facts and on the 

 larval stages. It should not base the whole animal classi- 

 fication on the larval stages or even on the type of cleavage. 

 The main emphasis should be on comparisons of grown-up 

 forms both of the fossil as well as of recent animal species. 



