In the Actinozoa an epiblastic skeleton is exceptional, and according to most authorities absent. Quite recently however Koch ([167]) has found that the axial branched skeleton of most of the Gorgonidæ, viz. the Gorgoninæ and Isidinæ, is separated from the cœnosarc by an epithelium, which he believes to be epiblastic, and to which no doubt the axial skeleton owes its origin. A similar epithelium surrounds the axis of the Pennatulidæ.
In the Medusæ the epiblast also gives rise to a central nervous system, which however continues to form a constituent part of the layer, and to the organs of special sense[86].
A special differentiation of the hypoblast is found in the solid axis of the tentacles. This axis replaces the gastric prolongation found in many forms, and the cells composing it differentiate themselves into a chorda-like tissue, which has a skeletal function, and is no longer connected with nutrition. This axis is placed by many morphologists amongst the mesoblastic structures.
In all the higher Cœlenterata certain tissues become interposed between the epiblast and hypoblast, which may be classified together as the mesoblast.
The most important of these are:
(1) The various distinct muscular layers.
(2) The gelatinous tissue of the Medusæ and Ctenophora.
(3) The skeletogenous tissue of the Actinozoa.
In most cases the muscular fibres are connected with epithelial cells, but in certain forms amongst the Medusæ and in the majority if not all the Actinozoa they constitute a distinct layer, sometimes separated from the epiblast by a structureless membrane, Æquorea Mitrocoma. Such layers when on the outer side of the membrane separating epiblast and hypoblast are undoubtedly epiblastic in origin, but in some cases amongst the Actinozoa they adjoin the hypoblast, and are very probably derived from this layer.
The origin of the gelatinous tissue is still involved in much obscurity.
It originates as a homogeneous layer between epiblast and hypoblast, which in the Hydromedusæ never becomes cellular though traversed by elastic fibres.
In the Acraspeda it contains anastomosing cells in the main apparently (Claus) derived from the hypoblast, and in the Ctenophora it is richly supplied with muscular stellate cells for the most part of epiblastic origin, though some are stated by Chun to come from the hypoblast. On the whole it seems probable, that the gelatinous tissue may be regarded as a product of both layers; and there are some grounds for thinking that it is an immense development of the membrane always interposed between the two primary layers. It must however be borne in mind that a membrane, regarded by the Hertwigs as the equivalent of the ordinary membrane between the epiblast and hypoblast, can be usually demonstrated on both surfaces of the gelatinous tissues in Medusæ. The skeletogenous layer of the Actinozoa is probably the morphological homologue of the gelatinous tissue; but the evidence we have is on the whole in favour of the connective-tissue cells it contains being epiblastic in origin. It gives rise to the skeleton of the Hexacoralla, to the spicular skeleton of Alcyonium, the axial skeleton of Corallium, and the skeleton of the Helioporidæ and Tubiporidæ.