10 
might it find there? Surely only microscopical algæ or bacteries, 
and hydroids do not, as a rule, take vegetable food ; moreover the 
pteropod shells are usually fairly clean, so that a hydroid colony 
consisting of several persons could hardly find sufficient food there. 
Finally there is the possibility that the hydroid eats the mucus 
or the epithelium of the pteropod. In such case the only possible 
place for attack is the upper surface of the foot, including the 
wings. The polyps placed on the front margin of the shell cannot 
dip into the ventral mantie cavity of the pteropod. Indeed, their 
obvious outward direction indicates that they really attack the 
wings. This must be done in that way, that the muscular mouth 
of the polyps adheres to the wings, the whole body of the polyp 
following the constantly flapping movement of the wings. The pol¬ 
yps placed behind the shell opening should then attack the upper 
surface of the hind part of the foot. And, indeed, in one of the 
colonies from the old material, one of the polyps on the ventral 
side of the shell, behind the shell mouth, is bent forwards, the 
mouth opening tightly adhering to the surface of the partly re- 
tracted foot of the pteropod. I have cut sections of this foot in 
order to see, whether the tissues might be in any way destroyed 
by the parasite. The tissues are remarkably well preserved, con- 
sidering the age of the material. On the dorsal surface of the hind 
part of the foot (which in the present case has been particularly 
exposed to the attack of the parasite) the epithelial cells are en- 
tirely uninjured. Towards the hind and lateral edges the cells are 
still covered with a gelatinous cuticula, but on the greater central 
part of the foot the cuticula has been rubbed o ff. This may be 
due to preservation, though it is worth noticing that, on account 
of the retraction of the animal, one part of the foot was concealed 
and protected inside the shell, and that the cuticula is in the same 
condition inside as well as outside the shell. Thus it seems quite 
likely that the cuticula has been eaten off by the hydroid, where- 
as the latter cannot afford to penetrate the very cell-layers of the 
host. 
This may hold good or not. In any case there can be no doubt, 
but that Kinetocodium danæ really feeds on the surface of the foot 
of the pteropod. From their position around the opening of the 
shell the nutritive polyps attack the dorsal surface of the expanded 
