CHAP. II. TESTACEOUS MOLLUSKS GENERALLY. 29 



monary circulation describes a separate and distinct 

 circle. As for their manner of reproduction, it is even 

 more variable and diversified than their mode of re- 

 spiration. In some, as with Gasteropoda, there is a 

 head with tentacula and eyes, either slightly or very 

 fully developed ; while in the Dithyra, or bivalves, all 

 these parts are wanting. Nearly all, however, have a 

 developement of the skin which covers their body, which 

 bears more or less resemblance to a mantle : but even 

 this assumes innumerable shapes; sometimes it is di- 

 lated in the form of wing-like fins ; in others, as the car- 

 nivorous Gasteropoda, it becomes a syphon by which the 

 animal breathes. In the naked Testacea, the mantle is 

 simply membranaceous, coriaceous, or fleshy. It is, in 

 fact, utterly impossible to give such a general detail of 

 this diversified class, either in respect to external or 

 internal anatomy, as will not be subject to innumerable 

 exceptions at every step. All the modes of mastication 

 and deglutition, as Cuvier says, are found among these 

 animals ; their stomachs are sometimes simple, some- 

 times multiple. They also present examples of all the 

 modes of generation ; and the same learned anatomist 

 confesses that " these varieties of the digestive and ge- 

 nerative processes are found in the same order, and 

 sometimes in the same family." * Hence it inevi- 

 tably follows, that any system founded solely on any 

 one or more of these anatomical considerations, is sure 

 to be not only unnatural, but perfectly unintelligible 

 to the great bulk of naturalists, who are referred to the 

 soft parts of an animal which they never saw, and 

 cannot procure. In regard to . the nervous system, 

 Mr. MacLeay has well observed, that " the most ge^ 

 neral notion we can form, at present, of the nervous 

 system of the Mollusca, is, that the medullary collar, in 

 the more typical groups, must always, in its circum- 

 ference, contain four ganglions, which may either be 

 united two and two, as it is probable they are in the 



* Griff. Cuv. p. 4. 



