CLASSIFICATION OF ACTINIARIA 519) 5) 
What forms are more primitive than Haleampa? It 
was suggested in Part [that Gonactinia and Protanthea 
are survivals of something very early (see pp. 493, 496-7, &c.). 
The grounds are these. The ‘Haleampa-stage’ in evolution 
may be defined as a stage with six pairs of perfect mesenteries 
(including two pairs of directives) bearing strong retractors, 
gonads, and filaments with ciliated tracts; any mesenteries 
beyond these six pairs would be rudimentary; there would 
probably be little or no base, a fairly narrow body, and little 
or no sphincter (cf. Text-figs. 8 and 7, c, p). This is not the 
Halcampa-stage sometimes used in an embryological sense, 
but is the way in which the term is usually taken for purposes 
of this paper. Now the Gonactiniids have paired mesenteries, 
but not six pairs perfect—only the eight protocnemial couples 
are fully developed. The filaments have no ciliated lobes, 
and the mesenteries have very weak musculature, not forming 
retractors as in the Haleampa-stage. Moreover, the body- 
wall, tentacles, disc, and actinopharynx approximate to each 
other in structure, at least as regards ectodermal muscle, and 
mostly spirocysts. This gives something much nearer a possible 
ancestor for the groups not specified as post-Haleampid than 
anything else. The consideration of Anthozoa generally, 
Suggests inevitably that mesenteries coupled before they 
paired, and the Gonactiniids still keep a vestige of the coupling 
which Halcampa has lost (see Text-fig. 16, By—and in 
a case like this the generalized musculature may be taken to 
indicate a stage before much differentiation of tentacles from 
body-wall, and of good retractors, had set in. 
There seems no reason to think that the Ptychodactidae or 
Madreporaria ever passed through a Haleampa-stage in 
the sense outlined above. They did not attain to much in the 
retractor line, and the Ptychodactids did not differentiate the 
parts of their ectoderm very markedly. They never have 
ciliated tracts on the filaments, and their whole organization 
and histology, especially of course in Madreporaria, suggests 
a difference of direction in evolution from that of the post- 
Halcampids. 
