16 ON THE STRUCTUKE OF CERATELLA FUSCA (gRAY). 



ordinary ectoderm cell thongh stained very deeply and having a homogeneous 

 appearance ; in b. the cell has increased in size, the nucleus is much larger, and has 

 a clear space all around it between it and the cell wall ; this is clearly marked in c. 

 and d. where, in the former, a nucleolus is present and in the latter two darker thread- 

 like lines possibly indications of the commencing formation of the thread ; in c. what 

 is evidently a very young thread cell is seen— it is somewhat darkly stained without 

 a definite thick wall such as is seen clearly in later stages, and down the centre is a 

 lighter line corresponding to the thicker attached part of the thread. It has also the 

 shape of the thread cell but there is no trace of the clear space present in earlier 

 stages, a certain amount of stained protoplasm being attached to it. In/, and^. two 

 later stages are shown in which the thick cell wall is present and the coiled thread 

 can be clearly seen. These thread cells evidently resemble closely in structure the 

 three-barbed ones described by Professor Moseley in Millepora. 



The only other point to notice in regard to the ccenosarc is the structure of the finer 

 growing branches which are somewhat flattened out. A longitudinal section of one 

 of these is represented in figure 14. Up the centre runs a ceenosarc tube with a large 

 cavity and clearly-marked endoderm the ectoderm being as usual irregular. From 

 this central tube short branches are given off {D) which run outwards towards the 

 external layer with which, as at the point x, they may come into direct contact. At 

 this point the cells of the two layers are well marked, and in all probability this 

 shows us the earliest stage in the formation of a zooid. It has already been noted 

 that the ectoderm of the latter is in direct connection not with that of the ccenosarc 

 tubes but with tlie common external layer and this method of formation would 

 explain this otherwise somewhat curious fact. The endodermal process grows out 

 into a bud — the early stage of a zooid — carrying with it the external ectoderm layer 

 which thus, as further growth takes place, naturally gives rise to that of the zooid 

 itself. At the same time the coe.nosarc tube branches as the stem increases in size 

 and thus the zooid, if the branching be near the base of the latter, will become 

 connected with two or more tubes. 



Affinities of the CERATELLADiE. 



When Dr. Gray first described these forms there were only two specimens 

 available neither of which possessed the soft parts. The hard parts whilst agreeing 

 in important points differed from each other sufliciently to be regarded by him as 

 species of two distinct genera Ceratella and Dehitella. 



Mr. Carter, with more material at his disposal, recognised the fact that they 

 were hydroids and owing to similarities in their skeleton and that of Hydractinia 

 placed them in the family Hydractiniidae, thus abandoning Dr. Gray's family 

 CeratelladaB which had been created under the assumption that the two forms were 

 allied to the sponges. 



