326 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
is the presence or absence of a longitudinal ectodermal musculature on the column- 
wall. With this appears to be often associated the presence of a columnar ectodermal 
nerve layer, the absence of a basilar muscle, and a weak parieto-basilar; likewise the 
absence of the Flimmerstreifen or ciliated streaks from the mesenterial filament, and 
the gonidial grooves from the s to modicum. Taxonomic studies of the Actiniae may 
thus be said to have passed through three' phases in the endeavor to secure a mor- 
phological and phylogenetic system. 
Undoubtedly the best arrangement would be one in which the results from al 
three sources were combined, but it is doubtful if more than an approximation to 
this is possible, divergences having taken place along many different lines, often in 
the same species. Hence the great difference of results according as one or another 
feature is accorded special prominence in systematic studies. 
Carlgren commences with the supposition that the earliest Actiniae were free and 
possessed of a well -developed ectodermal musculature and ganglion layer 1 throughout 
the body-wall and stomodseum, and the column-wall, tentacles, and disk were much 
alike in structure. The internal musculature was very weak or absent, and the mes- 
enterial filaments possessed only a single median lobe. The Flimmerstreifen or 
ciliated streaks and gonidial grooves were absent. With the assumption of a sed- 
entary habit the external musculature diminished in importance, the internal basilar, 
sphincter, and mesenterial muscles increasing, while the ciliated bands and gonidial 
grooves appeared to strengthen the internal circulation. In some polyps the original 
ectodermal columnar musculature persists, while in others it has become altogether 
lost, though it remains in the tentacles and disk. Wherever the ectodermal mus- 
culature is persistent in the column-wall Carlgren would consider its possessor as 
a primitive type and, however divergent the species may have become in other 
directions, would classify it along with polyps in which the same structure occurs. 
The classification followed by Carlgren in his “Ostafrikanische Actinien,” with 
the principal characters in each subdivision of the Actiniaria , is as follows: 
I. Ceriantharia. 
II. Actiniaria. 
A. Tribe Protanthcx. Actiniaria with a longitudinal musculature and ganglion layer in the column-wall and usually in 
the stomodeeum; basilar muscle absent; filaments usually without ciliated streak; without acontia, marginal 
spherules, and cinclides; sphincter muscle either absent or very weak, and always endodermal. 
1. Subtribe Protactininie. Protanthe* with only one tentacle in each radial chamber. 
2. Subtribe Protostichodactylinx. Protanthe® with more than one tentacle in some of the radial chambers. 
B. Tribe Nynanthcx. Actiniaria in which the column-wall and usually the stomodreum arc devoid of a longitudinal 
musculature and ganglion layer. Basilar muscle and ciliated streaks usually present. 
1. Subtribe Actimnx. Nynanthe® with the tentacles arranged in alternating cycles, not in radial series. Each radial 
chamber with only one tentacle. 
2. Sub tribe Stichodcictylinx. Nynanthese in which all or part of the tentacles are arranged in radial rows or groups; 
all or part of the radial chambers bear more than one tentacle. 
III. ZoA NTII ARIA. 
1 Attention may be drawn to the strong development of the nervous layer which occurs at the aboral region in the 
larval, free-swimming stage of many anemones and corals. McMurrich (1891, p. 317) has described such in the larva of 
Rhodactis sancti-thomx, and 1 have found a somewhat similar condition in the larva of Lcbrunia curalligcns (1899). Also 
in the larvse of the Madreporarians Fcivia fragum, Isnphyllia dipsacca, and Agaricia agaricites I have discovered a well- 
developed aboral nervous layer. Probably the formation represents part of an aboral sense-organ and disappears 
on fixation, the larvte settling by that extremity. I am not prepared to say if it has any phylogenetic significance. 
No ectodermal muscle fibers have been distinguished in the larva;, though on Carlgren’s assumption they may be expected 
to occur. 
