74 PALEONTOLOGY 



at first unconsciously, and later consciously, a tendency 

 towards phyletic classification that is to say, a classifica- 

 tion expressing the phytogeny or evolutional genealogical 

 tree of the group to be classified. All morphological 

 classifications must, in fact, be more or less artificial, 

 since they proceed by definitions of structure, not of 

 derivation : only by becoming phyletic do they become 

 completely " natural." Conversely, the more natural 

 a classification, the more difficult is it to express it by 

 morphological definitions, since the most important dis- 

 tinctions between two diverging lineages are not fully 

 developed in the earlier members of the lineages, and 

 also they may be lost in aberrant members whose con- 

 nexion with a lineage is clearly shown by secondary 

 features not characteristic of the whole of the lineage. 

 Hence modern taxonomic definitions are full of such 

 qualifications as "usually," " in typical genera," " in all 

 but the earliest members," etc. 



An early classification of Lamellibranchia was into 

 Monomyaria and Dimyaria, the latter being divided into 

 Heteromyaria and Isomyaria, and these last into Inte- 

 gripalliata and Sinupalliata. If the definitions are not 

 applied too rigidly, this forms a fairly natural classifica- 

 tion, and has the advantage to palaeontologists of being 

 based on shell-characters. But Heteromyaria are more 

 nearly allied to Monomyaria than to Isomyaria, and the 

 sinupalliate condition has been arrived at independently 

 along so many different (at least five) lines of descent, that 

 it cannot be used to define a single group. Another classi- 

 fication by shell-characters was that of Neumayr, based 



