Hinge-teeth are probably, in origin, derived from the crenulations or ribbings of the surface of the shell, the upper ends of which impinge upon the dorsal margin and mark it in a way which is quite recognisable when the shell is thin. Similar crenulations, resulting in interlocking of the valves, are not uncommon upon the ventral margin in certain genera (Fig. 190). The mechanical effect of these continued riblets, when fitted together on the opposing valves, would be to prevent the valves sliding upon one another while closing, or after being closed. Thus there would be a probability of their surviving, even after the ribbing had disappeared from the surface of the shell, the increased strength given by the hinge compensating for, and making it possible to do without, the extra strength supplied by the ribs. It is therefore possible that the teeth of the Nuculidae and Arcadae, which have no distinction between cardinals and laterals, represent a very ancient type, from which have been evolved the various forms of hinge in which cardinals and laterals are distinguished. Even in some forms of Arcadae (comp. _Pectunculus_) we get a hint how the transverse teeth of the typical _Arca_ may have become transformed into the longitudinal tooth of the normal lateral.[350] The developed hinge-teeth, then, ensure the opening of the valves in one direction; they also secure their accurate closure upon one another in exactly the same plane. Exposed shells and gaping siphons matter little to animals which are protected by their burrowing propensities, but to those which live in material which can be easily penetrated by their foes, it must be of advantage to be able to buckle their armour absolutely tight. The edentulous hinge of _Anodonta_ is a degeneration from a dentate type, which retains its teeth (in _Unio_, etc.) when subject to the jar of rapid streams, but tends to lose them in the stiller waters of canals, lakes, and ponds. _Other processes in the bivalve shell._--In _Anatina_ each umbo is fissured and strengthened on the inside by a kind of umbonal plate which carries the ligament. Some forms of _Liligna_ develop a strong internal umbonal rib, which serves as a buttress to strengthen the shell. In _Pholas_, the so-called falciform process serves as a point of attachment for the muscles of the foot and viscera. There is no ligament or hinge-teeth, the place of the latter being taken by the anterior adductor muscle, which is attached to the hinge-plate, the latter being reflected back into the shell. In _Septifer_ the anterior adductor muscle is carried on a sort of shelf or _myophore_, and in _Cucullaea_ the posterior adductor is partly raised on a similar and very prominent formation.