138 Allman, on Carduella cyathiformis> 
canal at the base of the tentacles towards the central 
stomach. 
In the ova (fig. 5), the germinal vesicle and germinal spot 
are distinct ; and the spermatozoa (fig. 6) present a charac- 
teristic hydrozoal form, consisting of conical corpuscles^ with 
the caudal filament attached to the broad end of the cone. 
Homologies. — As to the true import of the structure now 
described^ it will be easily seen that we have in it a genuine 
hydrozoal type, notwithstanding a certain superficial resem- 
blance to the structure of the Actinozoa. The point which 
at first sight seems to remove it most widely from the 
Hydrozoa, and approximate it to the Actinozoa^ will be found 
in the presence of the vertical lamellse which connect the 
stomach with the outer wall of the animal. A little atten- 
tion, however, will show that these must on no account be 
confounded with the radiating lamellae of an Actinia, from 
which they differ entirely in their arrangement and relations. 
The axile stomach of Carduella is exactly the manubrium 
of a Medusa, while the external body-walls correspond to 
the umbrella; and if I am coiTCct in my interpretation of 
the appearances which lead me to believe in the existence of 
longitudinal and circular canals, we have in them the exact 
representatives of the radiating and circular canals of the 
gastro-vascular system of a Medusa. 
As to the homology of the oral disc, I cannot help seeing 
in this muscular membrane the representative of the muscular 
velum of a gymnophthalmic Medusa, which, instead of being, 
as in the Medusa, free towards the axis of the animal, is here 
united to the stomach, while it is at the same time extended 
and so folded into plaits, as to form by the union of these 
plaits alternately to the stomach within, and to the umbrella 
without, the four pairs of vertical lamellae ; and although 
what we know of the development of the velum in the Medusa 
can scarcely be said to give any direct support to this view, 
it certainly is not inconsistent with it. 
It will now be seen that it is with the gymnophthalmic , 
rather than with the steganophthalmic Medusa, that the 
affinities of Carduella are to be sought. We have_, indeed, 
only to conceive of a gymnophthalmic Medusa, with its 
stomach (manubrium) united to the umbrella along four 
equidistant longitudinal lines through the medium of a pecu- 
liarly plaited velum, in order to convert it, so far as regards 
the most important points of its structure, into a Carduella, 
