CLASSIFICATION OF ACTINIARIA 525 
Without going over the old arguments again, we take it 
that if the ideas advocated in this paper be accepted at all, it 
will have been made clear in Part I that forms exhibiting 
these grades of mesenterial development need separation. 
We have, therefore, at once three families, Condylanthidae, 
Myonanthidae, and Actiniidae sens. strict., and these will 
be defined in Part III. This gives a homogeneous and intel- 
ligible Actiniudae, and has the advantage of providing two 
TExtT-Fic. 10. 
ing numerous perfect mesenteries. 
families as links between the Actiniidae and Haleampidae, 
from near which they presumably arose. The three families 
might be compared with, for instance, the Diadumenidae, 
Metridiidae, and Sagartudae of Part I—in which we have 
the same three grades of mesentery development, but acontia 
and cinelides and mesogloeal sphincter in all. In our new trio 
there is a common absence of acontia and mesogloeal sphincters 
and also of vesicles—as to cinclides it is hardly safe to say 
anything. 
NO. 260 Nn 
