ZOOLOGICAL AND BOTANICAL ASSOCIATION. 79 



Some of tho new forms throw some little additional light upon the 

 structure and development of the coencecium. In C. ITarveyi and in 

 C. a lata, the two membranes of which the cell wall is composed are re- 

 markably distinct. In the former species particularly the outer layer 

 seems scarcely to be in contact at any point with the inner, investing it 

 like a loose horny sac. The large M avicularian processes" are open and 

 cup-like, with ragged edges. In C. alata the two layers are also very 

 distinct ; but they are in contact over the greater part of the surface of 

 the cell. In both species the upper spine or cup and the lower division 

 of the lateral processes are formed of the outer membrane alone ; while 

 the true avicularian chambers, with the avicularia, are processes of the 

 inner layer, the true wall of the cell. Chambers formed of the outer film 

 are often entirely open or irregularly perforated with large apertures. 



The outer membrane seems to have something to do with the deve- 

 lopment of the cell. It is very possible that during the process of the 

 extension of the coencecium by gemmation, the outer layer may form a 

 dilatation expanding and filling with formative blastema, and that within 

 this sac the true cell wall and the organs of the polypide may be subse- 

 quently specialized. In C. ITarveyi, at a bifurcation, the young secondary 

 cell is entirely invested by this membrane during its early development, 

 and it remains permanently entire over the calyptriform ovicell, in both 

 these cases looking like the natural continuation and "finish" of the 

 abortive cup of the superior lateral process. 



In all cases where a secondary cell is the result of a further lateral 

 <!' \ i lopment of a primary cell, the former originates in the avicularian 

 chamber and process of the latter. In C. geminata one lateral process 

 of each axial cell is always developed into a secondary cell ; very pro- 

 bably the avicularian chamber, with its processes, whatever their direct 

 teleological object, may be the aborted indications of a constant tendency 

 towards development in this direction. 



Notwithstanding the numerous additions to the genus, Mr. Busk's 

 original subdivisions retain their natural integrity. C. alata fraternizes 

 with the typical Fenestratae. Busk's specimen of C. aurita must have 

 been poor. A good example differs so much from the Fenestrate group, 

 and so closely approaches C. geminata, which could not possibly be 

 associated with them, that it has been deemed advisable to put the two 

 species provisionally at the end of the list, thus indicating the tendency 

 of C. geminata towards the structure of the next genus. 



