INSTITUTE OF SCIENCE, PHILADELPHIA. 5OI 



modern bivalves. Its gradual disappearance may be traced in various Priono- 

 doiit genera. In Uiiio, for instance, the ligament is parivincular and wholly 

 opisthodetic, but in front of the beaks in most species is a depressed space, 

 homologous with the lunula of Venus, over which the epidermis is raised in 

 dense elevated lamellae which, to a careless inspection, seem like a continua- 

 tion forward of the ligament. It was probably this which led Neumayr to regard 

 the Naiades as having an amphidetic ligament. A careful inspection, especially 

 of fresh specimens, will convince any one that the functional ligament is wholly 

 posterior. Trigonia is in the same condition as far as the ligament is con- 

 cerned. 



The insertion of the ligament on the cardinal margin and the action of 

 natural selection lead to the formation of nymphs or ligamental fulcra in thin 

 shells, and the development and prominence of these often increases as the 

 shell grows thicker. There seems to be a point, however, beyond which in- 

 creased nymphal surface adds nothing to the security of the hinge, and conse- 

 quently is not promoted. As forms with thicker shells develop, the solidity 

 of the valves is sufficient to resist all strains, and the nymphse are not corre- 

 spondingly increased. But in many such cases the margins of the groove in 

 which the insertion of the ligament lies are elevated, thus protecting the joint 

 externally as well as internally. Dosinia is a good instance of external pro- 

 tection, and many heavy Uniones of the internal. As a large part of the growth 

 of the ligament is at the ends, there is generally a thin, more or less flexible 

 free edge or point medially behind; and, especially in Unio,B. thin edge of liga- 

 mentary substance at the hinder end of each nymph. In Unio this may usually 

 be noted, in the heavier and particularly in the alate forms, as a sinus in the 

 pearly surface of the posterior dorsal margin, covered by a brownish veneer of 

 ligamentary substance. 



It is hardly necessary to repeat the description of the mechanical process 

 by which the resilium is separated from the ligament when the latter is thick 

 and short.* The gradual submergence of the resilium in many forms is read- 

 ily accounted for by the agency of natural selection, when we remember that 

 the nearer the center of the animal the axis of motion is placed, the more 

 effective and rigid its operation with relation to the apposition of the valvular 

 margins. With a well-developed hinge, and deep-set resilium, the cardinal 

 functions are often so well performed that the ligament, if parivincular, may 

 degenerate and become no longer functional, or, if narrow, may (as in Rangia 

 or Midinid) follow the cartilage below the outer surface of the hinge-plate. In 

 this case it may descend into the same pit (as in Midinid), or into a separate 

 cavity (as in Cyclomactrd). These peculiarities will be more fully discussed in 

 treating of the Mactracea. Crassatellites offers a parallel example. 



In connection with the submergence of the resilium we may consider the 

 reinforcement of the pits in which the ends of the organ are attached to the 



* See Am. Journ. Sci. xixviii, 1889, pp. 448-451. 



