554 T. A. STEPHENSON 
in terms of evolution. The view here taken of the evolution 
of the forms will now be further developed. 
In Part I reasons were given for thinking of a Halcampa- 
like form as more primitive than such a creature as Catadio- 
mene (though of course more advanced than Gonactinia), 
and it was concluded that whatever the detail, the main 
chrection of evolution would be in the direction Haleampa- 
form ——->Catadiomene and not the reverse, and that 
this would generally apply. Without going into it all again 
(see Part I, p. 487) it may be assumed that in dealing with 
such a group as the Endomyaria, some Halcampoides- 
like form is the end to start at, and Tealia or Phymactis, 
or some Stichodactyline the antithesis, for much the same sort 
of reason, with differences in detail. Before discussing the 
Endomyaria further, however, it will be well to try to get 
at the relationship of Endomyaria and Mesomyaria to other 
croups. 
If it is fairly clear that both these groups originated some- 
where near Halcampa, the same is still clearer of the 
Athenaria—i.e. the Haleampids themselves and their burrowing 
descendants. There is also a clear suggestion of origin from 
a Halcampa-like ancestor in the Endocoelactaria, and they 
must be thought of as Halcampa-stock diverging from the 
main lines. The Stichodactylina (excluding the Corallimor- 
phidae and Discosomidae) are to be thought of as specialized 
Endomyaria. ‘The first idea to establish then is that Endo- 
myaria, Mesomyaria, Endocoelactaria, and Athenaria are the 
outcome along different lines of a Halecampa-stage with 
strong retractors and with ciliated tracts on the filaments. 
That is, they are ‘ post-Halcampid’ and form a single class, 
Nynantheae s.s. as defined on pp. 540 and 552, and in Part III. 
Next, there are the Gonactinidae, Ptychodactidae, and 
Madreporaria to be considered. The idea I hope to work out 
in connexion with these is that they originated in an ancestor 
earlier and less advanced than Halcampa (it would of course 
also give rise to Halcampa itself), and in fact may he called 
‘ pre-Halcampid ’. 
