258 SHELLS AND SHELL-FISH. PART I. 



barely be sufficient for a few general remarks on the 

 other families ; and these will be confined to the forms 

 most characteristic of the primary types of bivalves. 



(243.) The Tiibulihranchia, or tubular tribe, may 

 be traced in almost every one of the above families : 

 it is shown by an unusual elongation of the shell, which 

 is at the same time very narrow, aiid inclined to be 

 cylindrical. The whole of the Solens have this 

 analogy, as weU as the genus Anatina, in the Myadts. 

 Psammohia and Donax are modifications of the same ; 

 and so also is Cardissa: most of these also have the 

 anterior side remarkably short, precisely as in all the 

 PholadcE. Coralophaga is another example ; and it is 

 even apparent in one of the sub-genera of Cardium. 

 This type of form is equally, and even more strongly, 

 developed in the order Atrachia : thus, we detect it in 

 Iridina, Byssoarca, Lithophaga, and Avicula; aU of 

 which are the most elongated in general shape, and the 

 most abbreviated at their anterior extremity. In an- 

 other type, which evidently runs through the whole, it 

 is characterised either by the absence of all teeth to the 

 valves, or by there being but two, placed in the centre, 

 and resembling the letter V reversed. This is almost 

 always found in that division of a group which is most 

 aberrant in its own circle, as the SaxicavidcB and the 

 Etheridcd; the Anodontincp among the river bivalves ; 

 as well as the genera Pholadomya, Thracia, Acardo, 

 Loripes, Pinna, Placuna, Plicatula, Trigonia, Pan- 

 dora, &c. A third type of form is evidently intended 

 to represent the Branchiopoda ; its chief distinction 

 being that of having an internal cartilage — not, indeed, 

 protruding through an opening in the valves, but 

 situated between the cardinal teeth. This is well 

 known to be the predominant characteristic of all the 

 MactridcB, and of their prototypes scattered in other 

 groups, as Gnafhodon, Erycina, Nucula, UnguUna, 

 Pecten, Pedum, &c. Now, it is a remarkable fact, that, 

 on comparing our analysis of the whole of these circular 

 groups, we almost invariably found that these three 



