THE LAMELLIBRANCIIIA 75 



upon the hinge-teeth : the class was divided into orders 

 Taxodonta, Heterodonta, etc. This is in large measure 

 a natural scheme, but not entirely, as it brings together 

 Nucula and Pectunculus as taxodonts, although they differ 

 very greatly in other respects, and associates Trigonia and 

 Unto as schizodonts more closely than seems really 

 justified. 



Zoologists, however, dealing with recent lamelli- 

 branchs, and not having only the shell to examine, have 

 proposed as the soundest basis of classification the 

 structure of the gills. Thus a few genera such as Nnwla 

 have plume-like gills like those of the other classes of 

 Mollusca : these evidently form a primitive order (Proto- 

 branchia). Others, as Pectunculus, Trigonia, Pteria (Fili- 

 branchia), have gills transitional in character between the 

 last and the typical sack-like gills which give the name 

 to the whole class. Many others, including the re- 

 mainder of those so far described, have the typical 

 lamellate gill (Eulamellibranchia). Lastly, a few genera 

 have the gills modified into a simple septum across the 

 mantle chamber (Septibranchia). 



This plan is unsatisfactory to palaeontologists. Not 

 only is it based on morphological characters that cannot 

 be discovered in an extinct genus, but it separates genera 

 (like Pecten and Lima] which shell-characters show to be 

 nearly related, and unites them with others with which 

 they have no near relation (Pecten with Trigonia, Lima 

 with the Heterodonts and Desmodonts). The develop- 

 ment of the typical lamellibranch gill is a process which 

 must have gone on independently in different lineages, 



