50 ARVID R. MOLANDER, ALCYONACEA. 
at the two terminations of the spicule and leave a middle part of the spicule free from 
thorns. According as the girdles are more or less provided with thorns, we call these 
spicules simple or double stars. In the inner coenenchym nearly the same types appear 
as in the branches and trunk, but they are of a more simple construction and are gener- 
ally somewhat irregular. 
With reference to the variation in the shape of the spicules and the distribution of 
the thorns, it may be pointed out that the more closely the spicules are placed together, 
the more richly are they provided with large thorns or proce ses. As an instance of this 
may be pointed out the d fference between the spicules in the upper parc of the antho- 
 codia of the poorly armed Gersemia rubiformis, and those from the same part of the 
anthocodia of the considerably more strongly armed Gersemia uvaeformis, and the 
difference between the spicules in the upper and lower parts of the anthocodia of G. fru- 
ticosa. "The distinct difference in shape between the spicules in the upper and lower 
parts of the anthocodia, in the branches, and in the trunk,is due to their different positions 
and their different functions. The spicules situated in the bark of branches, together 
with those in the lower part of the anthocodia, — these spicules are also to be found in 
the parts which form the calyx — are short, with strongly developed thorns and processes. 
They are a protection, first of all, against pressure from without. These spicules lie very 
close together, the thorns are entwined closely together, and the whole results in an almost 
homogenous layer of spicules. The spicules in the upper part of the anthocodia are long 
and have few and feeble thorns. They are situated in the direction of the length of the 
anthocodia. They support the anthocodiae and they lack strong thorns, as such would 
prevent the contraction and retraction of the anthocodia. In the inner coenenchym the 
spicules are irregular, and they often have no definite longitudinal direction. The thorns, 
which are less frequently arranged in girdles, are long and slender. In the coenenchym 
the spicules lie in all directions. 
WOODLAND (1905), who has investigated the development of the spicules in Alcyo- 
nacea, states that spicules which are developed in proximity to ecto- or endodermal 
surfaces, have a more monaxonic shape, while the spicules are irregular and have long 
thorns, which are situated in the inner parts. He demonstrates further that the irre- 
gularity in the shape of the spicules is a consequence of their being situated in meso- 
dermal substance, far away from all surfaces, and of their vicinity to other spicules, ca- 
nals, etc., in the surrounding mesoderm; they thus become subject to influences which 
tend to produce irregularities; or these are produced since, for a growing body in a sur- 
rounding resistant medium, it is most effective to develop more or less sharp processes. 
In this case, it may be remarked that the spicules situated in the surface of the branches, 
in the bark, in the lower parts of the anthocodiae, and in the walls of the stomodaeum, 
are by no means regular and have at least numerous, strong thorns. That more or less 
crowding among the spicules produces a greater or slighter development of thorns is, 
however, correct, but such crowding also occurs in the proximity of ecto- and endodermal 
surfaces. As concerns the alternative possibility proposed, this is based on the above- 
mentioned crowding between the growing spicules. This case is, according to my opinion, 
by no means different from the one first demonstrated. In brief, we may say that the 
