36 EXPEDITION ANTARCTIQUE BELGE 



although there are, as I hâve shown avicularia, but even if \ve should not go so far as to place 

 it under Cheilostomata as now understood, it does not seem that it should remain with 

 Cvclostomata. The zoœcial chamber becomes much wider near the aperture, and has the 

 shape of that of several of the Cheilostomata ; the projections indicate a moveable operculum (*) 

 of the same character as those of Cc/laria, and further the shape of the zoœcia, and the 

 perforated front walls are similar in Cellaria. The central tubular structure is not known 

 in anj r living Cheilostoma and the ovicells as figured b} r Gregory do not resemble those of 

 Cheilostomata. The conclusion to which I am led is, that Cellaria and Onychocclla branched 

 from a common ancestor before the Cretaceous, and that this common ancestor and the 

 M elicerititidœ separated at a still earlier period. 1 mav add that if we had found Myriozoum 

 truncatum, in the chalk, and it allies were unknown, then its position would hâve been as 

 great a puzzle as is Meliceritites. 



Cellaria Dennanti MacG. 



(PI. II, figs. ga-f; PI. VIII. fig. A a 

 Cellaria Dennanti MacGillivkav, Tert. Polyzoa of Victoria, p. 3i. pi. XXII, fig. 14. 



In six of the bottles sent there were spécimens, though mostly in but small pièces ; 

 one spécimen, however, attains to about 40 mm in length, and in no case is there any sign of 

 articulation. In two colonies there are basai radicle attachments, from the proximal end of 

 the zoœcia. The zoœcia are rhomboidal, the oral aperture has two denticles on the straight, 

 proximal margin ; in some parts the distal edge of the zoœcium is rounded and in the oldest 

 parts is somewhat raised. 



The avicularium is rounded at the distal end, is about the same width as the zoœcia, 

 but longer, the opening is rounded above, and contracts to the two attachment ridges, which 

 are near the proximal end of the avicularium. The proximal edge of the avicularian opening 

 is either rounded or forms an obtuse angle ; the mandible, corresponding with the border 

 of the avicularian chamber, thus extends distally considerably over the avicularian opening. 

 The mandible, which is straight belovv, with straight sides, and rounded distal end, is among 

 the larger mandibles of the Cheilostomata, exceeded however by those of Flustra abvssicola 

 Sars, and M embranipora perfragilis MacG., ail of which are of the same type. 



The avicularian chamber as seen in transverse sections is enormous, reaching more 

 than half way across the zoarium (fig. gé). In a good spécimen (242) the border of the avicu- 

 larium is considerably raised, but this is not always the case. 



The opercula hâve a much raised ridge on each side for the muscular attachments, 

 and at each side a so-called foramina which fits over the opercular denticle, and in the covering 

 membrane there is a « trabeculum » on each side of the operculum. 



The ovarian opening does not occur in ail the spécimens, and is not very generally 

 found in any. It is quite at the distal end of the zoœcium and at first is semicircular then 

 becomes oval or nearly round. 



(1) The operculum of Euthyris clathrata Harmer works upon condyles very similar to the projections in 

 Meliceritites. 



