INSTITUTE OF SCIENCE, PHILADELPHIA. 5I3 



specialized fo nns inhabit permanent burrows of their own excavation. They 

 are sometimes commensal in the burrows of other animals. Similarly, though 

 often byssiferous, especially when young, few of them fix themselves perma- 

 nently by a byssus. With the exception of a few specialized forms they pos- 

 sess a pair of direct and reflected branchial laminae on each side of the body, 

 frequently united behind the foot, forming an anal chamber ; the two sets on 

 one side usually of unequal size, and of the reticulate type. None are known 

 with typically foliobranch or filibranch gills, though some abyssal forms have 

 archaic subfoliobranchiate ctenidia. 



The Paleoconcha. 



Prototypic Pelecypods, with thin shells, a simple or obscure pallial line, 

 subequal adductor scars placed high in the valves; dorsal area absent or am- 

 phidetic ; ligament external, variable ; hinge margin edentulous or with poly- 

 morphous teeth formed by modifications of the margin and not set upon a 

 hinge-plate (Neumayr). 



While the forms included here are not always actually the most ancient, 

 yet in their modifications they indicate clearly the origin of many subsequently 

 developed structures found in Pelecypods of a more modern type ; and by 

 their undifferentiated polymorphic character are difficult to assign a place in 

 any classification based on more fully-developed forms. Neumayr included in 

 the group the following families : 



Vlastidae. Protomyidas (including 

 Cardiolidae. • possibly the recent 6b/i??«j/a). 



Antipleuridae. Solenopsidse. 



Lunulicardiidae. Grammysiidje. 



Praecardiidae. Posidonomyidee. 



Silurinidse. Daonellidoe. 



Of some of these there is little doubt that they show taxodont affinities, 

 and others recall Plioladoinya, but the final discussion of these puzzling forms 

 awaits greater knowledge of them and other early bivalves. 



The subdivisions of the orders will now be considered. No linear ar- 

 rangement is practicable to show all the relationships. The earlier members 

 of the older groups converge, as a matter of course, and are assigned places 

 by an estimation of the preponderance of affinities, which cannot be infallible, 

 and upon which opinions will and must justly differ among students. It is 

 quite possible that, especially ainong the older forms, the number of families 

 will have to be increased with a future greater knowledge of the fossils. In a 

 few cases, where the author has not felt able to form an opinion, some groups 

 proposed by others have been omitted. In giving a list of genera included in 

 the families adopted, it must be borne in mind that these lists are not intended 

 to be exhaustive, and are given merely to exemplify the types associated under 

 the family name, without pretending to enumerate them all. The larger sub- 



