18 ARVID R. MOLANDER, ALCYONACEA. 
thropodium, KOR. and DAN.), Alcyonium antarcticum (WR. and StuD.) and, in all proba- 
bility, Alcyonmium (= Erythropodium) contortum, Alcyonmium membranaceum, Alcyonium 
reptans show the same peculiarity. AÅlcyonium coralloides is referred by KÖKENTHAL 
to Erythropodium, but according to my contention this cannot be done, as Alcyonmium 
coralloides, on another substratum, can assume a typical Alcyonium form. The same is 
the case with Alcyonium digitatum var. norvegicum, too. This variety, acknowledged by 
KUKENTHAL and BROCH as a species — Hrythropodium norvegicum — shows itself, on closer 
examination, to have, in normal cases, a fully typical Alceyomium colony. An antarctic 
specimen, the above-mentioned Alcyonium antarcticum (WRIGHT and STUDER), often 
shews a membranous extension of parts of the colony, where the substratum renders it 
necessary, but otherwise it is typical. TIn all probability a great proportion of the Ery- 
thropodium-species, Alcyonium membranacewm, Alcyonium reptans, and others too, could 
be classed as abnormal forms of what are, on another substratum, fully normal forms. 
For this reason, the sub-genus Hrythropodium should only contain membranous transfor- 
mations of otherwise typical species of Alcyonium, and, consequently, it possesses no claim 
Fig. 7. Gersemia fruticosa (SARS) forma membranea (GRIEG) 
= Sympodium hyalinum GRIEG 1887. 
to be considered as a separate sub-genus. The change of form of colonies is, consequently, 
an adaptation to the substratum, and has probably been produced from statical causes, 
where the insignificant substratum is not a sufficient attachment for larger colonies. 
If we compare the membranous forms belonging to the genera Gersemia and Alcyo- 
mum, we find certain differences. Generally, the colonies of Alcyonium have the polyps 
more closely packed; often, as in the case of A. digitatum var. norvegicum, they form a 
continuous covering over the whole colony, except in the extreme peripheral parts, or 
else they occur in close-lying groups. This distribution of the polyps on the membranes 
is a consequence of the distribution of the polyps on the normal, well developed colonies. 
A Gersemia has the polyps more widely scattered, and if the polyps are forced together 
into closer groups, these latter are often small and irregular. This, in turn, stands in 
definite connection with the formation of branches within the Gersemia-colony and with 
the distribution of the polyps. 
With regard to the canal system, the types of Alcyonium and Gersemia present a 
great similarity, as I have previously shown. When the polyps are scattered, the mem- 
branous spaces between them are filled with horizontal, shorter or longer indirect canals. 
In the investigated cases, at least, these indirect canals seems to be more numerous, and 
nearly fill the membrane in the species of Gersemia, while they are much fewer in number 
in the Alcyonium-type, and the indirect canal system mostly consist of endodermal cell- 
