560 P. CHALMERS MITCHELL. 
Thelaceros. Beyond this exact anatomical data fail. But 
Klinziger says of Discosoma, “ Mundscheibe nicht ein- 
stulpbar ;”’ of Cryptodendrum, “ scheibe nicht retractit.’’? For 
the most part observers entirely omit to make definite state- 
ments upon so striking a feature, and Andres sometimes omits 
even the scanty statements made by the authors he cites. 
But some of the forms under Stichodactyline and Thalas- 
sianthide have retractile tentacles, and so probably a sphincter. 
Such are the Aurelianide, which some state to be quite retrac 
tile; the Capreide, Heterocanthus (= Actinothrix of Andres), 
in which apud Klinziger the oral disc is ‘ fast ganz retractil.” 
It is therefore obvious that from want of accurate anatomical 
investigation so important a character as the presence or 
absence of asphincter is set at nought in the existing classifica- 
tions of this set of Hexactinians. It is equally obvious, con- 
sidering the almost undoubted presence of a sphincter in 
Corynactisand itsundoubted absence in Hertwig’s “Challenger” 
Supplement form, that this form must be definitely separated 
from the genus Corynactis. If this be done, from the mass of 
confusion in which the forms under consideration are involved, 
one group stands clearly defined ; 
Corallimorphide, Hertwig. 
Hexactinia without a sphincter, cinclides, or acontia, with 
accessory tentacles, so that more than one tentacle opens into 
a single radial chamber, with the normal and accessory tentacles 
knobbed. 
1. Corallimorphus. 
2. The form described by Hertwig in the “ Challenger” 
Supplement, and wrongly spoken of as a doubtful Corynactis. 
As yet all these forms are deep-water, while Corynactis is 
littoral. To this family I propose to add another equally de- 
finite family : 
Thelaceride. 
Hexactiniz without a sphincter, cinclides, or acontia, with 
numerous accessory rudimentary tentacles, so that more than 
