62 ARVID R. MOLANDER, ALCYONACEA. 
may be repeated many times on the same branch. In the lower parts of the branches 
and in the trunk, where the number of the coelenterons is more constant, this change 
from direct to indirect canal system disappears, and the indirect system predomi- 
nates. (PI. III, figs. 30, 32.) 
In those forms where the branches are reduced and a number of polyps rest on 
Small, short branches, the connection always becomes direct. This occurs, too, in the 
trunks of specimens belonging to forma arctica (DAN.) and forma loricata (MAR.), where 
the coenenchym between the coelenterons, at least with forma loricata, have become 
almost reduced. KÖKENTHAL (1906) mentions this in connection with certain species, 
which he classifies with Gersemia, but which are homologous with Gersemia fruticosa 
forma arctica and loricata. I have noticed the same characteristic in numerous speci- 
mens. 
The canal system is, consequently, not a constant characteristic of the species, 
but is found with all the varieties in the construction of the colony that distinguish the 
species. 
Forma typica. 
The colony is provided with numerous branches; the latter never cover the obvious 
trunk. The branches are often divided, and the anthocodiae are generally situated on the 
branches of the second or third order. The anthocodiae, which vary in size, are somewhat 
scattered and are found not only on the points of the branches, but also along their sides, 
although not very close. The calyx is rather common, hut the anthocodiae are com- 
pletely retractile. In young and small specimens, where the branches are small and short, 
and the trunk low, the existence of a calyx is general. These small colonies, which have 
small, short branches, sometimes nearly covering the trunk, ought, however, to be dis- 
tinguished from those of forma arctica, where branches are wanting, and the anthocodiae 
placed directly on the trunk. 
The armature is powerful. In the upper part of the anthocodia exist long spindles 
rather scantily provided with thorns. The lower part of the anthocodia (the polyp shaft) 
is provided with eight rows of rods, lying more or less closely together, and with uniformly 
arranged and rather powerful thorns. Often every row is double, with the spicules half 
impressed into each other, and some of the spicules are considerably narrower and 
more delicate than the others. In the branch bark there are rollers without distinet 
girdles. Double stars (dumb-bells) are wanting as a rule. In the trunk bark the spicules 
resemble those in the branch bark, but are somewhat smaller and have more delicate 
thorns. (For measurements see the table.) Within these general types there exists 
ample variety among the spicules, both of the anthocodia and from other parts of the 
colony, and transitions are found, towards both Gersemia clavata and G. mirabilis. 
The base of the trunk is generally membranously developed. In the specimens, 
which are found on a loose, clayish bottom, the basal part of the trunk is often formed 
like a bag, which contains a larger or smaller amount of mud or clay. In this way the 
colonies obtain a means of securing themselves on the loose bottom. In general the 
specimens of G. fruticosa are attached to some hard and fast substratum. 
I 
| 
| 
4 
| 
. 
d 
