874 PROCEEDINGS OF THE NATIONAL MUSEUM. vouxxi. 



siinilar in other points of structure, except where we know to the con- 

 trary. We find, moreover, that the dentition is freciuently indistinctly 

 developed, or somewhat amorphous, rendering it difficult to make out 

 the homologies of the diiferent parts of the hinge. It is certainly unsafe 

 to assume, as Bernard has sometimes done, that the position of a dental 

 lamina is sufficient to settle its homology. The dynamic reactions of 

 the teeth upon each other are, 1 am confident, of the utmost importance 

 in the development of the hinge. As in the vertebrate skeleton, pres- 

 sure and friction in localized areas will produce directly a response in 

 facets and buttresses. In fact, to the eye trained to take such matters 

 into account, every hinge shows more or less evidenceoftbe mutability 

 of hinge structure and its responses to stress, as well as to inherited 

 tendencies of form. In no group are these more obvious than in the 

 Leptonacea. 



The prototypic hinge of the grou]), or that which witli slight modifi- 

 cations will exhibit any of the various types of hinge-structure found 

 in the grouj), is very simple and has been figured by Bernard in his 

 illustrations of a minute form which he has named Pachyldlya. Ilis 

 invaluable researches upon the early features of the hinge have shown 

 that among the Teleodesmacea the so-called laterals and cardinals are 

 dissevered parts of originally single lamina? sharj^y bent at the proxi- 

 mal, or umbonal, end and having somewhat the form of a figure seven 

 (7). In Pachyl-ellya the hinge is composed of an internal resilium not 

 obviously separated from the ligament and inclined obliquely backward 

 as in many nepionic Teleodonts. On each side of this in each valve is 

 a pair of the N.-shaped lamella?, of which most have developed more or 

 less distinctly the proximal or cardinal "hook." The lower ones are 

 less engaged in the various stresses to which the laminte are subjected 

 in use, and hence, as might be expected, the hook is less evident or 

 even undeveloped. 



From this type of hinge all the others can be developed by trifling 

 modifications. The lamin.ne may be long or short; when the outer limb 

 is short we have a A-shaped tooth; if the angle proceeds to that stage 

 of development when its continuity is lost we may have a hinge like 

 that of Cyaraiomactra ; the severed hook may be modified by pressure 

 to a petaloid shape, which again by degeneration may be reduced to 

 two obscure minute conical projections, as in some species of Galeomma. 

 Any part or the whole of the hinge may become obsolete; the resilium 

 and ligament may separate or continue in connection; the latter fre 

 quently becomes external and often obsolete, though traces of it almost 

 always exist. 



The arrangement of the groups must, in our ijresent state of knowl- 

 edge, be provisional. 



No linear arrangement will show the exact interrelations of the difl'er- 

 ent genera, -and yet we are confined to a linear arrangement. 



The present tentative scheme is based on our present insufficient 



