472 
SEITARO GOTO 
As to the relation between the gastropod shell and the chitin- 
ous skeleton of the hydroid, Stechow thinks that the former is 
dissolved by the latter. In his full paper he says/ '^Bei unserer 
Form breitet sich die druchweg chitinose, nicht kalkige, hell — bis 
dunkelbraune Skelettschicht krustenformig liber die Schnecken- 
schale aus, die sie offenbar sehr schnell auflost. Ich fand nirgends, 
nicht einmal bei jungen Kolonien, auch nur Spuren von Kalk, 
weder in den Stacheln noch in der eigentUchen Schale." This is, 
as already suggested, a mistake. It is true that in most cases 
the larger part of the ^'sheir' is composed entirely of the chitinous 
skeleton of the hydroid, but if one breaks open its apex a small 
gastropod shell, Columbella^ or other genera is invariably found 
retaining its natural surface intact. It may therefore be inferred 
that the hydroid colony starts on the surface of a gastropod shell 
probably already inhabited by Eupagurus constans, and that 
after covering up the whole surface of the shell it grows out beyond 
its mouth. The hermit crab also growing hand in hand with the 
hydroid colony, and having no need of migrating into a new and 
larger shell to accomodate its growing body, there is thus estab- 
lished a permanent symbiosis between the two animals. That 
this is the case is proved by specimens occasionally met with, in 
which the hydroid apparently started on a comparatively large 
shell and has not succeeded in covering up its whole surface but 
has left a part still more or less exposed. Such a one is shown in 
fig. 2, in which the shell is a species of Buccinidse, and in which 
the exposed part is seen to have many small circular holes. That 
these are, however, not due to the action of the hydroids is very 
clear from the fact that they are most numerous where the hydroid 
colony is still very thin. In view of the destructive action of 
Hydractinia echinata (in this case in symbiosis with a Pagurus) 
on the shell which it covers, as was observed by Carter,^ it is never- 
theless not impossible that our hydroid may have a certain action 
on the substance of the shell, although such is not apparent in 
nbid: '09, p. 22. 
* For the determination of the shells mentioned in this paper I am indebted to 
my friend, Prof. T. Iwakawa of the Imperial Household Museum. 
•Carter: 73, p. 2. 
