18 ON THE STRUCTURE OF CERATELLA FUSCA (GRAY). 



purpose which it can apparently serve is that of a covering layer v^^hich prevents 

 foreign objects from passing in between the meshes of the network and interfering 

 with the general welfare of the colony. It is strange however to note, if this be its 

 function, that most usually the internal coenosarc contain a far greater number of 

 thread cells than this external layer does in Ceratella. 



In Millepora Professor Moseley was miable to determine its exact relationship to 

 the zooids but in Ceratella by its means all the ectodermal structures Ij'ing on the 

 external surface are brought into direct continuation with one another. 



It may be noted in passing that though in the genus Clathrozoon* the branches 

 of the colony are formed of a somewhat similar network of soft parts there is nothing 

 present resembling this external layer the whole branch being in this instance covered 

 with a thin protective perisarc. 



Taking both the hard and soft parts we find the following points of agreement 

 to exist between the Hydractiniidae on the one hand and the Ceratelladae on the 

 other though it must be borne in mind that we only know as yet the structure of the 

 soft parts in one member of the latter family. 



(1.) The skeleton has the form of a branching chitinous network. 



(2.) The hydrophyton consists of a network of freely branching and 

 anastomosing co^nosarcal tubes. 



(3.) The zooids arise directly from this network and no true hydrothecae 

 or gonothecse are formed. 



(4.) A common external layer is present enclosing the whole colony. 

 The two differ from one another in the following points : — 



(1.) Hydractiniidae form encrusting masses with at most spinulose branches 

 arising from the surface which do not bear zooids. The Ceratelladae 

 always form freely branching masses either erect or procumbent : the 

 basal part which serves to attach the colony being alone of an 

 encrusting nature whilst even this has the form of intertwined 

 branches. 



(2.) The Ceratelladae always possess hydrophores or special developments of 

 the skeleton which serve as a support for the basis of the hydroid- 

 zooids and nothing similar to which is foimd in the Hydractiniidae. 



(3.) The hydroid zooids Ceratelladae possess scattered capitate tentacles 

 those of the Hydractiniid* being filiform and arranged in a single 

 circle beneath the mouth. 



* Tians. R.S., Victoria, ISilO, p. 121. PI. 18, Fig. 3; PI. 19, Fig. 12. 



