634 MOLLUSCOIDEA. 



appendages can hardly ever be preserved, though their former exist- 

 ence may be inferred from the presence of " special pores " which 

 are usually placed near the mouth, or the existence of thickened 

 tubular supports which penetrate the substance of the polyzoary and 

 are thus recognisable in thin sections (fig. 455, a). 



The " ovicells " or marsupial pouches of the Cheilostomata are 

 very characteristic structures, though they may be wanting, or may 

 be so deeply immersed in the skeleton as to be inconspicuous. 

 Most usually, the ovicells have the form of globular or helmet- 

 shaped sacs (fig. 480, b) appended to the anterior end of the cells, 

 and placed in communication with the cavity of the latter by special 

 openings. 



New cells are produced by budding from the anterior ends or 

 lateral margins of the pre-existing cells, and all the zocecia remain 

 to some extent directly connected with one another. In the Cheil- 

 ostomata generally contiguous cells are placed in communication 

 with one another by means of perforated portions of their cell-walls 

 (" rosette-plates "), the number and position of these structures 

 varying in different cases. These " rosette-plates " correspond with 

 the pores which are so commonly developed in the cell-wall of the 

 Cyclostomatous Polyzoa, but their presence can only be detected 

 in the fossil forms in cases where the condition of preservation is 

 exceptionally good. 



As regards the distribution of the Cheilostomata in time, it is 

 doubtful, as previously noted, if any representatives of this sub-order 

 have hitherto been detected in the Palaeozoic rocks, though there 

 are some Palaeozoic types (such as Paleschara) which may belong 

 here. On the other hand, in the Secondary rocks, from the Jurassic 

 onwards, we meet with an abundance of the remains of Cheilosto- 

 matous Polyzoa, while a vast number of Tertiary forms have been 

 described. As the characters which separate the different groups 

 of the Cheilostomata are for the most part difficult of recognition, 

 and as the classification of the group is still in a more or less 

 unsettled condition, it will be necessary here to deal with the 

 families of the sub-order very briefly. The arrangement here 

 followed is, in the main, that adopted by Hincks, but only those 

 families which from their size or pabeontological significance are of 

 special importance, are alluded to. 



In the family of the Cellular iidce, the polyzoary is erect and 

 dichotomously branched, with linear divisions, composed of cells 

 arranged in the same plane. The genus Cellularia has no fossil 

 representatives, but Scrupocellaria (fig. 457, e) is known both by 

 Recent and Tertiary species. 



In the family of the Cellariidce (Salicornariada of Busk), the poly- 

 zoary is erect and dichotomously divided, the branches being cylin- 



