THE NEW GENEALOGICAL TREE 397 



the Priapuloidea, Nemertina, Endoprocta, and Mollusca (the 

 first and the last classes are placed among the Coelomata). 

 It should be mentioned in passing that in our study of 

 all these various animal systems we frequently get an impres- 

 sion that the pre-Darwinian custom of using a dichotomic 

 determinative key in the construction of an artificial system 

 still continues to work subconsciously even nowadays (cf. 

 the attempt made by Colosi which has already been discussed 

 above). The systematists must liberate themselves of this 

 influence if they wish to construct a truly natural animal 

 system. At present, it would be sufficient if these classes were 

 placed into such a linear order, or indicated in the genealogical 

 tree in such a way, that certain classes which are supposed 

 to be more closely related appear close to each other. It is 

 possible that after a certain time we will perhaps get a better 

 knowledge of their mutual relationships and it will be only 

 then that we will be able to begin with a more detailed work 

 on these subdivisions. Important progress has recently been 

 made just in this direction; as examples we can mention here 

 the work made by Steinbock (the relationship connections 

 between the Turbellaria and many other groups of the Ameria), 

 Golvan (the closer relationship connections between the Pria- 

 puloidea and the Acanthocephala), and my own studies Cni- 

 daria-Ctenophora, Turbellaria). 



The Transition from the Amerous to the Volymeroiis State 



We have seen that the phylum Ameria as the first phylum 

 of the Eumetazoa does not possess a perigastrocoele. When 

 we continue with our construction wdthin the framework of 

 one genealogical tree, we soon meet with the problem of 

 how to explain the emergence of a perigastrocoele (coelom 

 s.str.) and what its functional correlative could be. The 

 problem is whether this perigastrocoele had originally evolved 

 as a uniform cavity \^'hich had later become segmented, or 

 alternatively that there had been originally a segmented 



