218 J- Playfair McMuerich, 



endocoels) ; tlie tliird set coiisists of one or more cycles of pairs with 

 their long-itiidinal muscles on their endocoelic faces, developed in all 

 the exocoels present at tlie time of their development. In conse- 

 quence of tliis mode of development of the mesenteries in cycles of 

 pairs the symmetry in usually hexamerous after the development of 

 the primary pairs, but occasionally one or more pairs of the second- 

 ary or later cycles may be suppressed, an octamerons, decamerous 

 or rarely an heptamerous symmetry being- tlius produced. The sym- 

 metry may also be disturbed by the formation of more than two 

 pairs of directive mesenteries or by one or both of those usually 

 present developing their longitudinal muscles on their endocoelic faces. 



Suborder Actiiiinae Andres 1883. 



Actiniaria with usually simple tentacles arranged in single or 

 altern ating- cycles corresponding to the mesenteries, any one endocoel 

 or exocoel liaving- but one tentacle communicating with it. 



The arrangement of the Actiniaria in suborders must be regarded 

 as provisional, especially so far as the suborder SticJiodactylinae is 

 concerned, since it seems not at all improbable that this is a com- 

 posite group, the forms usually assigned to it having possibly a poly- 

 phyletic origin. 



Family Edivardsüdae. 



Actininae with a rounded base (physa), and with eight, twelve 

 or fourteen perfect mesenteries; additional imperfect or rudimentary 

 mesenteries may be present in the upper part of the column. Sphincter 

 absent or weak, endodermal or mesogloeal, longitudinal muscles circum- 

 scribed; parieto-basilar and basilar muscles equally developed or 

 almost so to form a parietal muscle ; acontia wanting. 



The description by Faueot ^) of the existence of rudiment have been 

 regarded as typical of the Edwardsiae has necessitated the abolition 

 of that Order and has broken down the distinction between it and 

 the family Halcampidae. I have chosen the term Edwardsiidae to de- 

 note the family which must now include these two groups and the 

 genus Scijtophorus also, since that term seems to liave the priority as a 

 family name over Halcamindae and because the latter term has been 

 used with a much wider seuse than seems to me justiflable. I had 



1) Andres (1880) was really the first to observe the occurence of 

 rudimentary mesenteries in an Edivardsia [E. da2)aredvt), two secondary 

 couples being distinctly represented in his fig. 70. 



