19 ARVID R. MOLANDER, ALCYONACEA. 
KÖUKEKTHAL has not made clear here, any more than he has done in the case of the 
preceding family, what is meant by indirect and direct connection. Apart from this, 
however, and also because the canal system in the case of Alcyoniidae is far from being 
so homogenous as KUKENTHAL considers it is, a closer investigation of Hunephthya  (sensu 
KÖUKENTHAL) and Gersemia (sensu KÖKENTHAL) gives an altogether different conclusion. 
The genus Hunephthya is divided, if we follow KÖKENTHAL, into two groups: the 
alcyoniformes and the nephthyrformes. The former is distinguished by a richer coe- 
nenchym in the points of the branches, from which follows a complete retraction of the 
anthocodia, the latter being characterised by a reduction in the coenenchym of the 
branch points. A retraction of the anthocodia is, therefore, not possible. To the alcyo- 
niformes belong the greater number of the species of the genus Gersemia (sensu KUKEN- 
THAL), as Gersemia arctica (DAN.), G. loricata MARENZ., G. Damelsem (STUDER), G. 
crassa (DAN.). 
On the above-mentioned reduction in the coenenchym of the branch-points in the 
case of the nephthyiformes there does not follow a corresponding reduction in the coenen- 
chym between the coelenterons. "The walls of the coelenterons are well-developed, and 
sometimes very strong, and their reduction is far less than in the case of species belonging 
to the alcyoniformes, as for instance, Hunephthya rosea (KOR. and DAN.), p. 14 and 84. 
Even this shows that a reduction of the outer coenenchym is not necessarily followed by 
a reduction in the coenenchym of the walls of coelenterons, so that direct connection 
results. 
Even within Kunephthya VERRILL (sensu KUKENTHAL) — divided by me into Ger- 
semia MARENZ. and Hunephthya VERRILL — we meet with a well-developed canalicular 
communicating with the inner coenenchym — an indirect system of canals. It consists, 
in this case too, of wide canals and cell-strings. These go in different directions, but run 
for the most part in a longitudinal direction — parallel with the coelenterons. Broad, en- 
dodermal direct canals, however, also exist between the coelenterons, just as in the case 
of the families immediately preceding. These have arisen from various causes and are 
secondary. 
The coelenterons, which run through 'branches and trunk, terminate at different 
heights. According to BourRNE (1900), there are found, even among the upper polyps, 
some whose coelenterons reach down to the base of the colony, but, by examining a series 
of sections from different species of the genus Gersemia, I have found that, in most cases 
at least, the older polyps do not accompany the growth of the colony so far, but that it 
is mostly the younger polyps that occupy the upper parts of the trunk and branches. 
This is pointed out as it is of importance for the arrangement of space between the po- 
lyps in the various parts of the colony. 
The endodermal canals which run in the coenenchym between the coelenterons 
are often very wide, and have considerable extension, parallel with the coelenterons, 
and forming larger or smaller cavities. How the direct canal system has developed here 
is shown by an account of some different types. 
In the case of Gersemia rubiformis, Gersemia uvaeformis, and Gersemia clavata, 
the points of the branches are strongly swelled, in consequence of a copious development 
