150 



PEOr. ST. GEOHaE MIVART ON" 



drawn out at opposite poles and be irregularly constricted meri- 

 dionally and equatorially, and we get such a form as Botryocampe, 

 one of a series of Polycystinian forms. 



Let this be open at one pole, and we get such forms as Eucyr- 

 tidium, Gyrtocalpis ohliqica, and Cornutella (fig. 5). 



Let the margins of the open end be produced into spines, and 

 we get such forms as Lychnocanium and Dictyopodium (fig. 6) . 



Let one pole be greatly expanded and we have Eucecryphalus, 



Fig. 7. 



Eucccryphaliis Schultzei. (After Kolliker.) 



and finally we ha,\ e Litharachnium tentorium^ which latter exhibits, 

 in the delicacy of its skeletal bars, a return towards the circum- 

 ferential network of Aulacantlia. 



Fourthly, let the circumferential skeleton, as it is in Gyrtido- 

 sphcera^ be conceived as haying its apertures greatly reduced and 

 its solid parts augmented in massiveness while they are reduced 

 in number, and we get such forms k^ ZygostepJianus* , Dictyocha'\, 

 and Acanthodesmia. 



But the circumferential skeleton may not only form a single 

 layer investing the capsule, for there may be two, three, or even 

 six more or less completely developed skeletal spheres concen- 

 trically investing the capsule, as in Arachnosplicera. These com- 

 plex concentric structures, however, are intimately connected 

 with and more or less dependent upon radiating skeletal elements, 

 so that they will be best considered along with the radiating struc- 

 tures referred to. 



* L. c. plate xii. l\g. 2. t L. c. plate xii. figs. 3-6. 



