80 ARVID R. MOLANDER, ALCYONACEA. 
in the anthocodiae within the latter species, besides the small anthocodiae and the 
slenderer spicules in anthocodia. 
The trunk of the colony is well developed, somewhat membranous at the base. The 
branches proceed from the whole length of trunk, but do not cover it. This becomes 
still more noticeable from the fact that large and small branches first begin to ramify 
in the upper part of the branches (terminally), and usually dicotomically, too. The an- 
thocodiae, therefore, collect on the points of the branches and the lower parts are almost 
devoid of anthocodiae. In consequence of this arrangement, the anthocodiae form close 
communities, which appear as semispherical, cushion-like parts. 
The anthocodiae are 1—1.s mm. in length, none club-shaped, and have, in general, 
weak armature. In the anthocodiae, however, this may occasionally be fairly strong, but 
transitions from these species to more weakly armed ones are numerous. In the tentacles, 
as well in the anthocodia body, this armature mostly consists of rods, 0.1—0.14 mm., and 
of spindles, a few clubs and crosses. A great number of spicules are situated at the base 
O0O00- TBGOOON = 
EN 
388 Zz 
Fig. 14. a) Diagrammatic figure of the ramification in the case of Eunephthya rosea (KOR. and DAN.).: I, II, III sections 
through different parts of branches. b) Diagrammatic figure of the reduction of the gastral tubes in the case of 
Eunephthya rosea. Section I is drawn at the base of the poiyps, II and III through branches of different sizes. 
Sections I, II, and III 'also correspond in the two figures. 
of the tentacles. As a rule, spicules are wanting in the anthocodia shaft, and if they are 
found there, it is only on the dorsal side. "The bark of the branches and the upper trunk 
are almost entirely free from spicules, and it is only in the basal parts of the trunk that 
we find 0.07—0.1 mm. rollers with girdles and crosses with large thorns. In the walls 
between the coelenterons we find similar spicules, although smaller ones. This is accord- 
ing to KUKENTHAL's observations; the specimens I have investigated have been without 
spicules in the inner coenenchym. 
The indirect canal system is the only one that occurs; it has the shape of rather 
broad canals, which often produce large spaces between the coelenterons. The umbellate 
ramification has produced an extraordinary development of the canal system. On every 
small branch there are placed on an average 4—6 anthocodiae. In the shaft of the small 
branches on which the anthocodiae are placed the coelenterons of the anthocodiae run 
side by side. The small branches again are packed two and two on larger branches, the 
coelenterons of which, consequently, are a direct continuation of those in the small 
branches. On the coalescence of two of these small branches to a larger one, there should 
in this case, if no reduction in the number of canals took place, be found from 8—14 
coelenterons in the larger branches. This does not occur, however, because, where two 
