CLASSIFICATION OF ACTINIARIA 541 
senting four lines of evolution; and the ‘ Stichodactylines ’ 
form a compact group within one of these four sets. These 
four groups can be defined by the sums of their main characters, 
and clearly the Actiniine-Stichodactyline contrast must be 
used simply in connexion with a subordinate division of that 
one of the four groups in which it occurs—if it be used at all. 
This is only making it one degree more subordinate than 
Carlgren does in his scheme. It is evident that as primary 
subdivisions of Actiniaria the two groups are no longer adequate 
—they must be reduced in rank, at least, from tribes to less 
than sub-tribes. 
Carlgren’s scheme is : 
Tribe NYNANTHEAE. 
Sub-tribe 1. Actiniinea. 
a. Athenaria. 
b. Thenaria. 
Sub-tribe 2. Stichodactylinae. 
The grouping I wish to suggest, as expressive of the above- 
mentioned four main lines of Nynanthean evolution, is : 
Tribe NYNANTHEAE. 
Sub-tribe 1. Athenaria. 
Sub-tribe 2. Endocoelactaria. 
Sub-tribe 3. Mesomyaria. 
Sub-tribe 4. Endomyaria. 
a. Actiniinae. 
b. Stichodactylinae. 
I have put in the Actinunae and Stichodactylinae where they 
must come, if used, in this scheme—as subordinate to Endo- 
myaria. 
The Athenaria of this plan is Carlgren’s Athenaria without 
the Edwardsiuds. I fully agree that it is a good group—but 
it represents a line of evolution within Nynantheae, all of which 
are derivatives of a Halcampa-like stage, and needs no 
subordination to anything else. Nor is there any need for 
a contrasting group Thenaria; the other three tribes are 
mostly *Thenaria’, but they represent three evolutionary 
lines and are best kept independent (see p. 560 et seq.). 
NO. 260 00 
