354 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
characteristic of the buds of the coral Cladocora a.rbuscula , while in the development of larval polyps 
of Sideraslrxa radians the succession is from the dorsal to the ventral side. 
The brothers G. F. and A. F. Dixon (1889, p. 322) first obtained evidence of the successive 
development followed by the members of the second cycle of mesenteries from freshly extruded larvae 
of Bunodes verrucosa. Their series of sections from above downward first showed the dorso-lateral 
pairs on each side, then the middle pairs, and lastly the ventro-lateral pairs. The succession in 
growth was therefore from the dorsal to the ventral side. 
At one region or another mesenterial filaments are present on all the mesenteries. On the mem- 
bers of the third cycle, which throughout their course are free from the stomodseum, the middle lobe 
first appears as a simple, more deeply-staining enlargement of the epithelium at the free end of the 
mesentery. But the lateral lobes make their appearance a little below, and the typical trilobed 
Actinian filament is then presented, the three lobes being very distinctly separated from one another 
(fig. 38). The four areas, which elsewhere (1900, p. 145) I have spoken of as the glandular, ciliated, 
intermediate, and reticular streaks, are well defined by their differences in histological detail, and do not 
require full description. 
As the mesenteries of the second and first cycles cease their connection with the stomodseum 
they are tipped with filaments which, both in longitudinal and transverse sections, are shown to be 
continuous with the stomodaeal ectoderm, the middle or glandular streak most nearly corresponding 
in histological detail. The strongly ciliated lateral lobes are limited to the region at which the mesen- 
teries become free; below the stomodseum only the simple median lobe is retained (fig. 39), and this 
is the case also with the filaments of the third-cycle mesenteries. Where the filament is simple the 
mesenterial epithelium immediately behind becomes much swollen, as shown in fig. 39. The species 
is somewhat exceptional in the short course through which the lateral or ciliated streaks persist. In 
the" lower region of the polyp the mesenteries are greatly contorted at their free edge, which is provided 
with the simple filament all the way. 
No gonads were present in either of the specimens. 
Family SAGARTID/E Gosse. 
Actiniaria with a contractile pedal disk; body-wall smooth, or provided with verrucse or tuber- 
cles, and usually perforated by cinclides, with or without a cuticle. Tentacles usually numerous and 
retractile, usually not very long, simple, and generally entacmseous. Sphincter muscle usually well 
developed and mesogloeal, occasionally diffuse endodermal, or even absent ; at least six pairs of perfect 
mesenteries; the first cycle of six pairs of mesenteries may be fertile or sterile. Acontia present. 
The above definition is mostly taken from Haddon (1898, p. 448), who in 1889 made one of the 
first anatomical studies of the group. In his latest paper he recognizes five subfamilies, as also does 
Carlgren: Aiptasinse, Sagcirtinse, Phellime, Metridinse, Chondraclininie. The first and fourth have each 
a representative in the Porto Rican collection. 
Subfamily AIPTASINSE Simon. 
Sagartidx in which the sphincter muscle is either absent, or very feebly developed and mesogloeal, 
or diffuse and endodermal. 
Genus AIPTASIA Gosse. 
Aiptasici, Gosse, 1860, p. 151: Andres, 18.83, p. 373; McMurricli, 1889, p. 6; Haddon, 1898, p. 447. 
Bartholomea, Ducha'ssaing et Miehelotti, 1866, p. 133. 
Aiptasinse with an extensile soft body; column smooth, with cinclides arranged in from one to 
several horizontal rows around the middle of the column; no verrucse or tubercles; tentacles numerous, 
long, smooth, or with thickened bands or tubercles, strongly entacmseous; sphincter muscle either 
absent or very feebly developed and mesogloeal, or diffuse and endodermal. 
McMurricli (1889, p. 6) gives a full history of the genus, which is exceptional in the family 
Sagartidx in that some of its members are entirely wanting a mesogloeal sphincter muscle, in conse- 
quence of which the disk and tentacles are never infolded. I have added the character of the ten- 
tacular wall to Haddon’s definition of the genus. 
