NEW INTERPRETATION OF CNIDARIA 295 



like layer. An intermediate space is developed in this way 

 which is surrounded by the terminal ends and which is filled 

 with a fluid. The androgametes (the migrating nuclei) of the 

 two individuals must thus leave their maternal bodies and 

 cross this intermediate space. Not only does a cellularization 

 of these karyons take place, during which the cytoplasm 

 which surrounds a nucleus becomes separated from the com- 

 mon cytoplasm of the conjugant, but also this cytoplasm 

 develops into a spermatozoa which is similar to a flagellate. 

 Thus equipped the androgamete which is now free (it deserves 

 fully this term), crosses by swimming the intermediate 

 space and penetrates actively into the cytoplasm of the partner 

 in order to reach in this way the gynogamete which has 

 kept the form of a karyon. 



This androgamete of Entodiniomorpha is not a typical 

 spermatozoan, its cytoplasmatic process which is similar to 

 a flagellum is supposed to be developed at the anterior end 

 of the cell and it does not correspond cytologically to the 

 flagellum of a spermatozoan; yet in spite of this— and contrary 

 to the interpretation proposed by Jagersten— it is an andro- 

 gamete and it is a homologue to a spermatozoan. Nor 

 can we suppose that it represents a line of evolution which 

 had prepared the way to the formation of a spermatozoan. It is 

 much more probable that we see in this special form of 

 androgametes a new formation which has never led— and 

 which will never lead— to the formation of a typical sperma- 

 tozoan. According to DoUo's rule of irreversibility, the andro- 

 gametes of Entodiniomorpha have not been able to recover 

 the characteristic properties of the typical spermatozoa which 

 they had previously lost; their androgametes have now been 

 developed in a new way. In this w^ay it becomes clear that I 

 do not derive the spermatozoa of Eumetazoa from such 

 androgametes that can be found secondarily developed in 

 Ophryoscohcidae: in fact I derive the Eumetazoa from such 

 infusorial ancestors that possessed free and typical androga- 

 metes. It is therefore unjust when Jagersten accuses me of 



20* 



