INHABITING MULTILOCULAR SHELLS. 39 
be imagined that, being connected with the animal, it must necessa- 
rily live or die with its owner; but this is not always the case. 
Muscular fibres extend from parts of the Pinna, and other molluscs, 
which, when beyond the shell, lose all their vitality, and become 
mere strings of attachment, without feeling or sensation of any kind. 
Of course, before any very long time has elapsed, the little inhabitant 
will have grown too large for his cell, and be ready for other and 
wider accommodation. In order to obtain this, there seems to be a 
process analogous to that of casting the shell in crustacea ; but, in- 
stead of a complete separation, the siphuncle remains fixed, and being, 
perhaps, itself grown, and also a little extensible, it allows the body 
to be removed to the wide-open end of the shell. Immediately upon 
this there is a fresh accumulation and exudation of shelly matter, in- 
creasing the shell in magnitude, and depositing, besides, a wall sepa- 
rating the original habitation from the new dwelling of the animal. 
It is clear, however, that since the mantle is pierced where the si- 
phuncle emerges, there will be in that place a corresponding aperture 
in the wall; thus forming an air-tight chamber, with a tube passing 
through it. This process of enlargement being repeated an indefinite 
number of times, we obtain the complete shell, consisting of a num- 
ber of empty air-chambers, with the tube passing through all of them. 
The cup-like form of the animal accounts for the shape of the septa, 
and the part of the septum where it has the aperture left for the si- 
phuncle is one very useful character for distinguishing genera. 
Now if the siphuncle were merely an organ of attachment by which 
the animal was kept in its place during the formation of new shell, 
although it might still be an important point to know its situation, 
there would be little interest attaching to it. But it is not so— 
There is every reason to believe it to be a contrivance adapted for a 
much higher object, and one helping, perhaps, to connect the animal 
still more nearly with the vertebrated classes. In fishes, and all the 
swimming reptiles and mammalia, there is always more or less a 
power of rising and sinking in the water. It is true that this power 
is much more limited in fishes than people commonly imagine ; but in 
the molluscs generally, and in the less highly organized classes, it is 
rarely found to exist at all. There can, however, be little doubt that 
the siphuncle is a means by which the Cephalopods provided with 
multilocular shells are enabled to alter their specific gravity, and so 
alter their relative depth. 
The whole nature of the contrivance may be thus explained :— 
The deserted chambers in which the animal once dwelt are filled with 
