MOLLUSCA DISSOLVE THEIR SHELLS. 483 



turreted shells, and the septa of the regularly chambered 

 shells, are all deposited in the same manner, the body repre- 

 senting the model on which they are formed. But it is not 

 so easy to understand why such cavities should be left in 

 tliese shells, especially in the upper valves, as it is to account 

 for the existence of the analogous structure in turreted and 

 chambered shells, the flat form of their valves enabling the 

 animals of the former, as we might suppose, to extend the 

 diameter of the existing cavity, when larger space was re- 

 quired for their accommodation without constructing one 

 altogether new. In the Etheriac, cavities in the form of 

 small vesicles, or very thin bladders, are also left between 

 the plates. The cavities in the Ostreae, Spondyli and Ethe- 

 riae, are, I have reason to believe, filled with water when 

 the animals are alive ; and this also appears to be the case, 

 from Mr. George Bennett's account, with the chambers of 

 the Nautilus ; but the water soon evaporates through the 

 pores of the shell, if kept in a dry place. I have never 

 observed this peculiarity except in those bivalve shells 

 which are immediately attached by their outer surface to 



other bodies. 



Many shells are composed entirely either of the rhombic 

 crystalline or of the concretionary structure ; but I know only 

 of a single instance (and that occurs in the tube of a shell) 

 in which the whole mass affects the prismatic crystalline 

 structure. In all other shells of this latter texture, the 

 inner and front part, which is occupied by the body of the 

 animal, is always covered with a coat of the laminar con- 

 cretionary texture. 



4. On the Power possessed hy MoUusca of dissolving Shells, 



Rocks, ^'c. 



It has been generally believed, and indeed sometimes most 

 positively asserted, that Molluscous animals do not possess 

 the power of re-absorbing the matter of their shells when 

 once deposited. The following observations, I think, will 

 distinctly prove that this assertion is quite unsupported by 

 fact. 



If a cone, an olive, or any shell whose last whorl nearly 

 envelopes and protects the rest, and whose cavity is much 

 compressed, allowing only a small space for the convolutions 

 of the body of the animal, be slit down, either from the 

 apex to the front of the axis, or across the body volution, 

 at a little distance before the suture, it will be observed that 

 all the septa between the different whorls are extremely 



F F 



