556 T. A. STEPHENSON 
Although these forms (Gonactiniids, Ptychodactids, Madre- 
pores) must be put down as pre-Halcampid, they have common 
features establishing them as distinct from Hdwardsiaria and 
Zoanthinaria, and they form one group, Dodecactiniaria— 
for instance, they have typically attained pairing of mesen- 
teries and equality of directives, and the pairs are not usually 
formed each of a macro- and a micromesenteric partner, nor 
do they usually develop in two lateral zones of increase only, 
after a certain point; there are no canals in the body-wall 
save in some of the skeleton-making Madreporaria. 
So that it may be said that the Dodecactiniaria present on 
the one hand descendants of a Gonactinia-like form, and 
these are poor in muscle and lack ciliated tracts ; and on the 
other hand descendants of a Halcampa-like form (itself, 
of course, the outcome of an earlier Gonactinia-like one), 
with the ciliated tracts stabilized and a tendency to muscularity. 
Is the ancestor of the Zoanthactiniaria, the group containing 
the Dodecactiniaria as well as the Edwardsiaria and Zoan- 
thinaria, simply the same sort of Gonactinia-like animal ? 
The whole situation suggests that it must have had a good deal 
in common with Gonactinia—it would surely be a small 
form with weak muscle and generalized ectoderm and only 
eight perfect mesenteries (see Text-fig. 16, a); the chief point 
of debate is, had its filaments ciliated tracts ? At first glance 
one would say No, the state without the tracts is more primi- 
tive: but there are other things which do not suggest that it 
was devoid of them. That the ancestor of all Anthozoa was 
without them seems certain, but that is even farther back 
than the one here visualized. Our Zoanthactiniarian ancestor 
gave rise to Edwardsians and Zoanthids as well as to Dode- 
cactiniaria, and both the former have ciliated tracts, even if 
they are not quite the same as those of the Nynantheae. This 
suggests that either (i) the Edwardsians and Zoanthids attained 
them independently, or else that (i) the Gonactinuds, Ptycho- 
dactids, and Madreporaria lost them, while Halecampa and 
its followers retained, stabilized, and developed them. (See 
Text-fig. 17 for the main types of filament here mentioned.) 
