THE NEW GENEALOGICAL TREE 403 



into Other classes of Cnidaria and not into any other form 

 of the Eumetazoa. We think that the formation of the coelom 

 (we think here always of the perigastrocoele only) had original- 

 ly nothing in common with the intestine. The cases, however, 

 where the coelom sacks develop during ontogeny as a result 

 of an outgrowth of the intestinal wall (enterocoely) — a pheno- 

 menon which can be regularly observed among the higher 

 Eumetazoa and occasionally also among the lower Eume- 

 tazoa— must be understood as a type of development which 

 had evolved secondarily, as a rationalization, as a special 

 "mechanism" which had developed from the phylogenetically 

 older ecterocoelic type of development. Frequently, it is nei- 

 ther the ectoderm nor the endoderm which serves as a basis 

 for the development of the perigastrocoele, but rather the 

 transition area between the two. In the Polymeria the peri- 

 gastrocoele develops, with few exceptions that have already 

 been mentioned, without a participation of the Anlage of the 

 intestine, yet in widely different ways. Very frequently we can 

 find the Anlage of the segmented perigastrocoele developed 

 in the form of a pair of the so-called primitive mesodermal 

 cells which change into a pair of cell strings that finally develop 

 into pairs of sacks that grow frontwards. This can be observed 

 above all in the polychaetous Annelida, among the lowest 

 Polymeria. Thus we cannot find here an intermediate prelimin- 

 ary stage with a uniform perigastrocoele. On the other hand 

 we can frequently find in the derivative Annelida, Anlagen 

 of coelom sacks which later grow together forming a second- 

 ary uniform perigastrocoele. 



The ontogenetic type of development of coelom sacks had 

 been necessarily strongly modified because of the planktonic 

 larvae, the trochophores, which had been evolved by the 

 benthonic Annelida. This suggests that we cannot make any 

 conclusions regarding the original type of development by 

 means of a comparison between the development of the 

 mesohyl and of the coelom sacks in recent species; in this 

 point there can be no recapitulation in the narrow sense of 



