312 G. ©. J. VosMARR, 
the other hand they form at the periphery very distinet brushes of 
diverging spicules (Taf. 15 Fig. 5). 
III. Acanthella. 
This genus is again established by Oscar Schaumipr (1862, p.64—65), 
who gave the following diagnosis: “Halichondriae ramosae et fruti- 
cosae, tanquam spinis obsitae. Cutis laevis, porosissima, quae in 
ramis crassioribus sola pigmento infecta est et verae pellis instar a 
parenchymate distinguitur. Parenchyma spisse impletum spieulis 
simplicibus longioribus, substantia firmiori non inclusis.” ScHMIDT 
described two species, A. acuta and A. obtusa; by later authors the 
number increased to 19. As the type of the genus I will take the 
species, which ScHmipr described first, viz. A. acuta. Both species, 
acuta and obtusa are very well described and illustrated by SCHMIDT; 
they are easily recognised. The difference between them is, however, 
greater than might be supposed by the original desceriptions. Not 
only is the canalsystem different, but also the skeleton, although the 
elements of the latter, viz. the spieules, resemble each other very 
much. The difference in their arrangement is best seen in prepara- 
tions the soft parts of which are removed. Fig. 3 and 4 (Taf. 16) 
represent a part of the skeleton of A. acuta, Fig. 1 and 2 that of 
A. obtusa. I arrived at the conclusion that the differences are large 
enoush to justify even a generic distinction. Consequently we have, 
with regard to Acanthella only to do with specimens corresponding 
to SCHMIDT’s A. acuta. 
The skeleton of Acanthella shows no distinetion between an axial 
and an extra-axial part. Herewith the genus is at once distinguishable 
from Azxinella and Phakellia. We find that the spicules are united 
by spongin in order to form funiculi, which, by their ramifications 
and anastomoses form a network, composing in this way a funis 
(Taf. 15 Fig. 6 and 7). The meshes between the funiculi are rather 
small; hence the funis is compact, and as the composing spieules 
hardly project beyond it, the funis is on the whole smooth (Taf. 16 
Fig. 3). At the base of the sponge the funis is thick; it soon divides 
itself into two or three — at any rate a few — branches, slightly 
less in diameter than the basal stem. Each branch likewise divides 
itself and so on till the periphery is reached. Here, by gradually 
diminishing the diameter the branches are composed of only a few 
spicules. As the branches in ramifying always diverge and never 
