ANATOMY OF THE MADREPORARIAN CORAL FUNGIA. 319 



gelatinous matter, poured in, as it were, between the ecto- 

 derm and endoderm to serve as a support for those tissues and 

 to give coherence and consistency to the body of the animal. 

 Where no cells are present (Hydromedusse, Charybdoea), a 

 third cell layer, a mesoblast, obviously cannot be spoken of. 

 In other forms we find cells derived from one or other of the 

 primary layers wandering into the gelatinous substance after 

 the formation of the latter, and retaining a constant position 

 there (Discomedusae, Actinaria). These cells ought to be con- 

 sidered epiblastic or hypoblastic according to their origin, just 

 as much as the central nervous system of the Vertebrata, 

 entirely surrounded by mesoblastic structures, is considered as 

 part of the epiblast. 



In the Alcyonaria the separation of the epiblast cells which 

 are destined to become the skeletogenous cells takes place 

 contemporaneously with the secretion of the gelatinous matrix 

 in which they are embedded. None the more is the layer thus 

 formed entitled to be called a germinal layer, or even a 

 separate cell layer, though a step has been made towards the 

 latter. The cells forming the skeleton are clearly epiblastic in 

 origin, are derived from the epiblast after its demarcation, and 

 are properly considered as its derivatives. The Alcyonarian 

 skeleton is really of epiblastic origin. 



In a further stage the cells which, in Ctenophora, are 

 destined to become stellate muscular cells embedded in the 

 gelatinous matrix of the supporting lamina, are, according to 

 Metschnikoff, marked out early in the embryo, at a period 

 when the endoderm is scarcely covered in by the ectoderm. 

 This is nearly the same thing as the formation of embryonic 

 mesoblast, and foreshadows it, but the ultimate history of the 

 cells ought to preclude our calling them mesoblastic. 



I do not wish to assert that the supporting lamina of Cce- 

 lenterata is not represented in the mesoblast of Coelomata; it 

 is highly probable that it is. The Alcyonaria and Ctenophora 

 are good examples of the tendency which muscular and con- 

 nective-tissue cells, primitively belonging to the external and 

 internal limiting layers, have to separate themselves from their 



