426 STRUCTURE OF SHELLS. 



they are formed are crystallised, in the other they are 

 granular. These differences in structure coi'respond with 

 differences in their chemical character; the shells in which 

 the lime is crystallised appearing to contain less animal 

 matter than the others. The primary division of shells 

 according to their structure consequently agrees with the 

 division proposed by Mr. Hatchett, in his paper on their 

 chemical composition ; the porcellaneous shells of that 

 distinguished chemist being crystalline, and the nacreous 

 granular. 



Shells of the crystalline structure are themselves of two 

 very distinct kinds ; the crystals being rhombic in some, and 

 prismatic in others.* 



Those of the rhombic crystalline structure exhibit, when 

 broken, three distinct layers of calcareous matter. On 

 examining fragments of most of the spiral univalves, there 

 will be observed on two of the fractured sides of the cubic 

 pieces into which they generally break, flat surfaces on the 

 inner and outer edges, separated from each other by a shelv- 

 ing portion in the centre ; and on the two intermediate 

 broken sides shelving external and internal edges, connected 

 by a flat central portion ; these differences of surface being 

 produced by the different position of the crystals of the 

 different layers. 



Each of the three layers, thus rendered obvious, is com- 

 posed of very thin laminae, placed side by side, as high as 

 the thickness of the plate and perpendicular to its surface. 

 When these laminae are minutely examined, they will be 

 found marked with obscure oblique lines ; in the direction 

 of which they separate, when broken, into long narrow 

 rhombic crystals. The lines of cleavage in the succeeding 

 laminae are placed in contrary directions, so that when two 

 of these plates united are examined under the microscope, the 

 lines of cleavage apj)ear to cross each other at right angles, 

 whilst those of the alternate laminae follow the same direction. 



The laininjfi of the outer and inner plates are always di- 

 rected from the apex of the cone of which the shell is formed 

 towards its movith ; in the spiral shells they consequently 



* Woodward and Poll described the prismatic structure ; and the 

 Compte de Bournon, in his " Traite de Mineralogie," described the rhombic 

 structure in the Strombus and Cyprsea, but liis work has been overlooked by 

 all conchologjsts, and was not known to me until long after this paper was 

 published. It is curious enough that it has not been referred to by any of 

 the numerous persons who have written to me about these observations. 

 The Compte de Bournon only considers the shell mineralogically, while I 

 have referred to it physiologically, and in its living relation to the animal. 



