CLASSIFICATION OF AGTINIARIA 559 
Summarizing so far, we get the suggestion of an evolutionary 
course somewhat as follows : 
From a small, delicate, bilateral ancestor, with eight feebly 
muscular mesenteries, with some degree of differentiation of 
ciated tracts, and with generalized ectoderm, there arose 
(i) Edwardsiaria, the mesenteries of which never 
paired, but some of them attained muscularity (see 
Text-fig. 16, 8). 
(ii) Zoanthinaria, the mesenteries of which paired, but 
which went in for various curiosities (see ‘Text-fig. 16, r). 
(in) Dodecactiniaria, the mesenteries of which paired, 
and which developed along the familiar ‘ Hexactinian’ 
lines. 
There is just the possibility of an alternative view of the 
Kdwardsiaria to the one adopted in this paper—namely, that 
they might somehow be Nynantheae in which certain mesen- 
teries had been suppressed so that now there are only couples 
and not pairs. It is their histology which rather suggests 
Nynanthean affinities, but this idea is put forward very 
tentatively and further work would be required to ascertain 
how far it could be entertained as a possibility. 
The Dodecactiniaria split on the rock of sluggishness versus 
muscularity.t The Gonactinia-like ancestors experimented 
a little, and gave rise to the Gonactiniidae and Ptychodactidae, 
perhaps trial-lines, on the one hand, and to the corals on the 
other ; all these losing the ciliated tracts and never getting 
very muscular, the majority-forms going in for strict sedentari- 
ness and skeleton-building, often colonially. In a different 
direction there arose from one of the Gonactinia-like 
ancestors a muscular Halcampa-form; this, far from 
losing the ciliated tracts, developed them further, and gave 
rise to the individualized and typically muscular forms, which 
fell into four sets—Athenaria, Endocoelactaria, Mesomyaria, 
Endomyaria. 
' See in this connexion Chapter VIII in Thomson and Geddes, ‘ Evolu- 
tion’, 
Pp2 
