THE POSITION OF SPONGES, ETC. 431 



independently from the class Choanoflagellata ". Sollas, 

 though formerly of a different opinion, came round to the 

 same view and proposed (1888, xcii.-xcv.) the name Parazoa 

 for sponges, as a name of the same genetic and systematic 

 value as Metazoa. He regards the possession of collar 

 cells, as well as certain developmental facts, as distinguish- 

 ing them from the Ccelenterata, the most primitive Metazoa, 

 which in their turn are distinguished by the possession of 

 nematocysts. 



Finally Delage (1892, p. 416) has come to the conclusion 

 on developmental grounds, that "the sponges show a pro- 

 gressive differentiation of their elements ; but this differentia- 

 tion does not take place from the beginning in the sense of 

 germ layers comparable to those of other animals ". He 

 thinks they may be descended from Protozoa, "but that 

 they have followed since the beginning an isolated develop- 

 ment at the side of the stem of the Cctlenterates and other 

 Metazoa ". 



III.— THE SPONGES AS METAZOA. 



The majority of authors, impressed by many facts of 

 sponge organisation, such as their being composed of 

 differentiated cells and tissues, their sexual reproduction, 

 with ova and spermatozoa, and their development through 

 a free swimming larva composed of distinct germinal layers, 

 have regarded sponges as Metazoa, that is to say as de- 

 scended from an ancestor which had passed the colonial 

 protozoan stage, and which was common to them and other 

 Metazoa. This does not mean however that the majority 

 of authors are united in their opinions ; far from it. We 

 have next to distinguish two different points of view, those 

 who regard sponges as coelenterates, and those who con- 

 sider them a phylum of Metazoa distinct from the coelen- 

 terates. It is more convenient to take the latter first. 



(ii) The Sponges Metazoa but not Ccelenterata. 



If sponges are not coelenterates, the next question which 

 arises is, are they composed of the same primary germ 

 layers, ectoderm and endoderm, as are coelenterates and 

 all other Metazoa ^ 



