CONCHIFERA. 251 



ridge of Dlceras and CardUia resembles a lateral tooth, and in the extinct 

 genus Radiolites both adductors were attached to large tooth -like processes 

 of the opercular valve ; but, as a rule, the muscles deposit less shell than the 

 mantle, and their impressions deepen with age. 



The paHial line (fig. 177, m) is produced by the muscular fibres of the 

 mantle-margin; it is broken up into irregular spots in the monomyary 

 bivalves, and in Saxicava and Panopeea Norvegica. 



The sifhonal impression, or pallial sinus (fig. 14, p. 26,) only exists in 

 those shells which have retractile siphons ; its depth is an index to their 

 length. The large combined siphons of Mya (fig. 170) are much longer than 

 the shell ; and those of some TelHnidce three or four times its length, yet they 

 are completely retractile. The small siphons of Cyclus aad Dreissena cause 

 no inflection of the pallial line. The form of the sinus is characteristic of 

 genera and species. 



In the umbonal area (within the pallial line) there are sometimes fur- 

 rows produced by the viscera, which may be distinguished from the mus- 

 cular markings by absence of polish and outline. (See Lucina, PI. XIX. 

 fig. 6.) 



Fossil bivalves are of constant occurrence in all sedimentary rocks ; they 

 are somewhat rare in the oldest formations, but increase steadily in number 

 and variety through the secondary and tertiary strata, and attain a maximum 

 of development in existing seas. 



Some families, like the Cyprinidee and Lucinidce are more abundant 

 fossil than recent; whilst many genera, and one whole family (the Hippuri- 

 tidce), have become extinct. The determination of the affinities of fossil 

 bivalves is often exceedingly difficult, owing to the conditions under which 

 they occur. Sometimes they are found in pairs, filled up with hard stone ; 

 and frequently as casts, or moulds of the interior, giving no trace of the 

 hinge, and very obscure indications of the muscular markings. Casts of 

 single valves are more instructive, as they aff'ord impressions of the hinge.* 



Another difficulty arises from the frequent destruction of the nacreous 

 or lamellar portion of the fossil bivalves, whilst the cellular layers remain. 

 The Aviculidce of the chalk have entirely lost their pearly interiors ; the 

 Spondijli, Chamas, and Radiolites are in the same condition, their inner 

 layers are gone and no vacancy left, the whole interior being filled with chalk. 

 As it is the inner layer alone which forms the hinge, and alone receives the 

 impressions of the soft parts, the true characters of the shells could not be de- 

 j termined from such specimens. Our knowledge of the extinct RadioUte is 

 derived from natural moulds of the interior, formed before the dissolution of 



j * These impressions may be conveniently moulded with gutta-percha. M. Agassiz 



I published a set of plaster-casts of the interiors of the genera of recent shells, which 

 ( may be seen in the Brit. Museum. [Memoire sur les moules des Mollusques, vivans 

 etfossiles, par L. Agassiz, Mem. Soc. Sc. Nat. Netichafel, t. 2.]. 



