2 McMURRICH 



HEXACTINI^ 



ACTININ.E 



Family SAGARTIID.E 



Actininae with adherent base ; with a mesogloeal (rarely a 

 weak endodermal ^sphincter) ; and with acontia which are 

 emitted either through the mouth alone or also through special 

 openings (cinclides) in the column wall. 



The family Sagartiida^, since its first establishment by Gosse 

 in 1858, has undergone certain changes which have for the 

 most part been fully discussed by various authors. Within 

 recent years there has been introduced a subdivision into sub- 

 families. The family Phellin^e of Verrill ('68) has been added 

 as a subfamily and the remaining forms assigned to the sub- 

 families Sagartiinse or Metridiinse according as they possessed 

 more than six pairs of perfect mesenteries or only that number 

 (Carlgren '93) ; a subfamily Chondractininae had previously 

 been proposed by Haddon ('89) ; Simon ('92), recognized two 

 subfamilies Aiptasiina^ and Sagartiinse, the former characterized 

 by possessing an endodermal sphincter or none at all ; and, 

 finally, Haddon accepts all the proposed subdivisions, admitting 

 the existence of no less than five subfamilies. 



It seems to me that the subdivision proposed by Carlgren 

 is that to be preferred. I do not think the recognition of a sub- 

 family Aiptasiinae is advisable since several undoubted Aiptasias 

 are known to possess a mesoglceal sphincter, as, for instance A. 

 pallida (Ag.), A. sp. (from the Bermudas, McM.)and A. litcida 

 (Duch. & Mich.) Duerden. In all these, it is true, the muscle 

 is exceeding weak, but it nevertheless is present, and its absence 

 in certain species is merely the fulfillment of the reduction of it 

 which is characteristic of the genus. A separation of the forms 

 with no mesogloeal sphincter would be an act of violence, and, 

 if this be avoided, the genus Aiptasia is properly referable to 

 the Metridiinae. 



In the second place, it does not seem to me that a recognition 

 of the Phelliinae and Chondractiniinae as distinct subfamilies is 

 necessary. Both lack cinclides, have a coriaceous column wall 



