666 IMOLLUSCA. 



structure may be detected on fhe surface; and this is troquently so thick as to makeup a great part of" 

 the shell, as in Pinna, Avicula, and a large proportion of Cuvier's family Ostracece. The internal 

 layer is usually more compact, and presents less distinct traces of cellular structure. In the shells of 

 the Cuvierian families, Camacece^ Cardiacece, and Inclusa, the greater part of the thickness is formed 

 by the internal layer. In the Mutilacecc^ both layers are usually well seen, the inner layer, howevtr, 

 being usually the tliiekor. The successive additions to the shell are not made on the same plan in the 

 two layers. The L^uter layer is merely extended by the junction of an additional portion to its margin ; 

 but the inner htyer receives an addition to its whole internal surface, so that its thickness is increased, 

 fis well as its extent. 'J'his is well seen in the common Oi/.s/er, in wliicli the succe?.si\'e layers of the 

 shell remain unusually distinct. The white inner portion, o^ wliich the greater part of each valve is 

 composed, is made up of a number of laniinge, each of v\'hich extends beyond the one external to it; 

 and thus the outer lamiuie are at tlie same time the oldest and smallest. Each of the layers of this 

 substance is covered at its edge by an elastic yellowish-brown margin, which is so arranged that if 

 the successive layers were closely adherent to each other, this substanre would form a complete ex- 

 ternal covering to the shell. A section of the shell of an Unio^ in wiiieh the layers do thus adhere, 

 shows thac the mode of growth of the two layers of a compact shell is essentially the same. 



In many Bivalve shells of the class BRACHroPODA, especially belonging to the genus Terehratida 

 and its allies, a very curious arrangement exists; the shell being perforated by a number of minute 

 apertures, extending from the internal to the external surface; and these canals being filled with 

 prolongations of the soft tissue of the animal itself. 



A number of curious varieties of shell-structure havti been described by Dr. Carpenter (Inc. cit.) ; 

 who has also shown that, in many families and genera of bivalves, the structure of the shell affords 

 characters of great importance in classihcation ; and that it is possible in several instances to recog- 

 nize the family, or even the genus, to which a specimen belonged, by the microscopic examination of 

 but a very minute fragment of it. 



In the Univalve shells formed by MoUusca of the class Gasteropoda, the arrangement is generally 

 different, especially in the porcellanous shells of many of the PectinihrancJiiata. These are of great 

 density, and contain very little animal matter. They have three layers instead of two, and these three 

 layers are similar to each other in structure, being composed of a series of rhomboidal plates, disposed 

 vertkallij to the surface of the shtll, and giving an appearance of crystaillne structure when the shell is 

 broken across. The ilirtction of these plates is the same in the inner and outer layers ; but it is 

 transverse in the middle ln;yer, by which arrangement a greatly increased strength is gained. Each 

 plate is made up of a set of long narrow cells filled with carbonate of lime, and adherent to each other 

 at their edges. The shells of Gasteropoda do nut, for the most part, undergo any thickening by suc- 

 cessive additions; tlic enlargement required by the growth of the animal being effected by the exten- 

 sion of the margin, with no further addition to the previous internal layer than is sufficient to give it 

 the requisite smoothness along the line of junction of tlie new and old portions. 



In the General Classification of the Mnllusca, given by Cuvier, the chief alteration required by the 

 progress of knowledge is the entire removal of the Cirrhopoda to the sub-kingdom Articulata ; their 

 affinity to which, perceived by Cuvier, has been since placed beyond doubt by the discovery, that they 

 issue from the egg in a condition resembling that of certain low forms of Crustacea, having eyes and 

 active powers of locomotion, and only acquiring the form and condition of Cirrhopods after a series of 

 metanvirphoses, in whirh the eyes are lost, Ahe locomotive organs altered in character, and the shell 

 formed. This shell, in the Dalaiins and other sessile Cirrhopods, is composed of an outer and an 

 inner plate, separated by a diploe or cancellated texture, which is sometimes of considerable thickness. 



Perliaps the best primary division of the true MoUusca is into the encephalous, or those provided 

 with a head, and the acephalous^ or headless ; in which last the mouth is not placed on a prominent 

 pui'tion of the body, but is concealed by the projection of the mantle. 



The first division includes the three classes of Cephalopoda, Pteropoda, and Gasteropoda, 

 from which last the Heteropoda are detached by some Zoologists, to be ranked as a distinct class. 



The second division may also be arranged into three classes, namely — the Conciiifera, furming 

 the testaceous division of the Acephala of Cuvier; the Brachiopoda ; and the TuMC ata, fornung 

 tlic naked er shell-less ilivision of the Acephala of Cuvier. The hr::^t two of these classes both possess 

 bivalve shells; but the structure of the anima! is very diQ'ereut, the respiration being carried on in the 



