LUCERNARIANS OF PORT ERIN. 257 
ment of tentacles having been disposed of as far as the 
groups of secondary tentacles were concerned (as detailed 
above), there only remained the octagonal disc and absence 
of primary tentacles in Depastrwm to separate it from 
Carduella. 
With regard to the first point, as Clark points out, 
there is in Sars’ figures (and in Allman’s also) an octa- 
gonal area, the corners of which correspond to the inter- 
vals between the tentacular groups and whose outline is 
marked out by a distinct line of brown pigment; the ten- 
tacles arise from immediately without this line, and there 
can be little doubt that this 1s the octagonal margin of 
the disc spoken of by Gosse, but the real margin of the 
disc is formed by the circular muscle situated outside the 
origin of the tentacles, some of which are usually curved 
over it when the animal is fully expanded, the primary 
tentacles being invariably so apparently. ‘This circular 
muscle is in life a delicate translucent structure and may 
have escaped Gosse’s notice. Then as to the absence of 
primary tentacles in Depastrum. These in Carduella are 
similar in form and structure to the secondary tentacles of 
the groups, unlike those of Haliclystus awricula which 
have been modified into the so-called marginal anchors 
and function as adhesive organs. They are in Carduella 
not very conspicuous, being usually somewhat smaller 
than the secondary tentacles and constantly (as I have 
remarked above) reverted over the circular marginal mus- 
cle and closely applied to its surface. It is accordingly 
not unlikely these too escaped Gosse’s observation, espe- 
cially if his specimens had been kept long, as the tentacles 
seem apt to slough away when the animal is removed 
from its natural conditions. In addition I may mention 
that Depastruwm, and apparently other members of this 
eroup also, 1s subject to much variation, especially in 
