THE ACTINIAE'S OF PORTO RICO. 
327 
The Protanihece will thus include Actinians which have retained their primi- 
tive histological and anatomical structure, wholly irrespective of modification along 
other directions, while the JSFynanthece will embrace forms more highly developed 
as to their histology, anatomy, and external characteristics. Included under the 
former are representatives from each of the two great divisions, Actinince and Sticho- 
dactylince . 1 
With the great progress which has been made in the morphological study of 
the Zoantharia (using the term in its usually accepted sense) within the last two 
decades, students of the group are in a better position than they formerly were for 
estimating which of the polypal characters are really of phylogenetic importance 
and should therefore be seized upon for the purpose of erecting a classification 
which will show their true relationships. Writers such as McMurrich (1898, page 
229), Haddon (1898, page Til), and van Beneden (1897, page 153) are disposed to 
regard the ectodermal columnar musculature as ancestral, and the forms in which 
it occurs as the lowest members of their own particular group, but they are not 
prepared to accord to it the importance of making it the one character upon which 
the grouping should be determined, to the exclusion of later divergences. It is 
rather regarded as a character which, sporadically, as it were, may appear in any 
group, the various species possessing it not necessarily representing a homogeneous 
or natural assemblage. 
No one doubts that the forms in which the columnar ectodermal muscle and 
nerve layers are associated with the absence of the basilar muscle, a weak parieto- 
basilar muscle, absence of the ciliated streak and gonidial grooves are more primi- 
tive than forms in which the muscle and nerve layers are absent, but it is doubtful 
as to how far their possessors represent a homogeneous group to be separated 
from others. To my mind they represent the most primitive members of various 
divergent groups, rather than a group to themselves. To separate them as a whole 
from others would be to neglect the facts of their subsequent development from the 
primitive type. 
At the outset we are confronted with the fact that Actinians have evolved in 
complexity of structure along man}’ different directions. In addition to that of the 
mesenterial plan there are divergences in the tentacular system, the musculature, the 
column-wall, as well as in many minor characters. Following Carlgren’s proposals, 
Actinians which retain certain primary characteristics will be grouped together, 
however divergent they may have otherwise become. Carlgren replies that structures 
1 In connection with the structural variation exhibited by the Actiniaria a comparison with the polyps of the closely 
allied Madreporaria is instructive. So far as I have observed in the course of an examination of the soft tissues of over 
twenty species of West Indian corals there are no certain indications of a columnar ectodermal musculature and ganglion 
layer, though such occur in the tentacles and disk. True gonidial grooves are always absent from the stomodreum. The 
mesenterial filaments are invariably formed of only a single median lobe, no lateral lobes with ciliated bands being 
developed. The internal musculature is everywhere very weak, and the sphincter remains endodermal; mesogloeal 
plaitings are rarely formed for the support of the parieto-basilar muscles, and basilar muscles arc absent. Marginal 
spherules and cinclides have not been met with; mesenterial filaments are freely extruded through any part of the column- 
wall and disk, but at no part of their course are they independent of mesenterial attachment, and hence are not true 
acontia. The Madreporaria are therefore at practically the same phylogenetic level as the Actiniaria which Carlgren 
includes within the Protanthcx , except as regards the apparent absence of the columnar ectodermal musculature and 
ganglion layer. . 
