iv PHYLUM CCELENTERATA 227 



however, show evidence of degeneration, so that it is quite possible 

 to conceive them as having been derived from more highly 

 organised forms, instead of springing directly from simple polypes 

 of the scyphula type. The Semostomse, Cubomedusse, and 

 Rhizostomae clearly represent three grades of increasing com- 

 plexity along the same general line of descent, the Coronata 

 diverging somewhat. It is to be noted, however, that such a 

 supposed line does not lead towards the simpler Actinozoa, but 

 towards a type which diverges from the latter as well as from the 

 Lucernarida, Cubomedusae and Peromedusae in the absence of 

 septa or mesenteries in the adult condition. 



The close similarity of Edwardsia and the Alcyonaria in the 

 number and arrangement of the mesenteries seems to indicate the 

 derivation of both Zoantharia and Alcyonaria from a common 

 ancestor in the form of a simple actinozoan polype or actinula, 

 Edwardsia clearly leads us to the Hexactinise or typical Sea- 

 anemones, and the Madreporaria are undoubtedly to be looked 

 upon as skeleton-forming Hexactinia?. 



The relationships of the Ctenophora to the other Ccelenterata are 

 very doubtful. Ctenaria, one of the Anthomedusaj (Fig. 109, 1\ 

 presents some remarkable resemblances to a Cydippid, such as 

 Hormiphora. It has two tentacles, situated in opposite per-radii, 

 and each having at its base a deep pouch in the umbrella resem- 

 bling the sheath of Hormiphora. There are eight radial canals 

 formed by the bifurcation of four inter-radial offshoots of the 

 stomach, and corresponding with them are eight bands of nema- 

 tocysts diverging from the apex of the ex-umbrella. If these 

 striking resemblances indicate true homologies, we must compare 

 the whole sub-umbrellar cavity of Ctenaria with the stomodseum 

 of Hormiphora, the margin of the bell of Ctenaria with the mouth 

 of Hormiphora, and the mouth of Ctenaria with the aperture 

 between the stomodseum and the infundibulum of Hormiphora. 

 But, as we have seen, the gullet of Ctenophora is a true stomo- 

 doeum developed as an in-pushing of the oral ectoderm, and has 

 therefore a totally different origin from the sub-umbrella of a 

 medusa. Moreover, the tentacles of Ctenaria have no muscular 

 base contained in the sheath, but spring from the margin of 

 the umbrella as in other Hydrozoa : its gonads are developed in 

 the manubrium, not in the radial canals, and there is no trace of 

 an aboral sense-organ. 



Of Hydroctena, which has also been supposed to afford us a 

 connecting link between the Hydrozoa and the Ctenophora, almost 

 the same may be said. Hydroctena is bell-like, and provided with 

 a velum. At its apex is an ampulla bearing two lithites supported 

 on spring-like processes of the epithelium. From the apex of the 

 gastric cavity a canal is given off which extends to the sense- 

 organ, where it terminates blindly, and from the sides a pair of 



Q2 



