90 THE NAUTILUS. 



cies represent the rule, or typical condition. While we thus do not 

 know the type, we know, on the other hand, which species should not 

 be the type. 



Now if any subsequent author is to select a type species, this latter 

 surely should correspond to the original diagnosis, and should repre- 

 sent the rule but not the exception. Haas cites a rule of the inter- 

 national code of nomenclature (section 35), which says that no species 

 should be selected as type which has only " doubtfully " been as- 

 signed to the genus by the original describer. Of course, taken 

 verbally, this rule does not entirely fit the present case, but without 

 much difficulty it might be stretched so as to cover it. If U. mar- 

 garitiferus is selected as type, as Oken does, a species is taken which 

 is abnormal and does not fully correspond to the original diagnosis, 

 while a number of species which do fit the original diagnosis are 

 thrown out. While Bruguiere's change in the diagnosis consists 

 only of the dropping of the word " most" (plurimis], thus throwing 

 out the exceptional case only, Oken's definition of the genera involves 

 a complete change, for instead of having lateral teeth " in most 

 cases," as formerly, Unio now has " never any " lateral teeth, prac- 

 tically the opposite. 



These two considerations are, as far as I can see, consistent not 

 only with common sense, but also with the rules of nomenclature. 

 There is no rule which says that an author has no right to change 

 the concept of a genus by modifying the diagnosis, as long as one or 

 some of the original species remain included, and this is what Bru- 

 guiere has actually done, and we see it black or white before us. 

 And further, in doing this, Bruguiere simply carried out an idea 

 already suggested by Retzius, namely, that the genus Unio consists 

 of a number of species representing fully the normal condition of the 

 genus, and of an additional one which forms an exception. 



Consequently Bruguiere has the priority, and Lymnium of Oken 

 becomes simply a synonym of Unio, as restricted by Bruguidre. For 

 the remaining species ( U. margariferus) Schumacher's name Mar- 

 garitana is to be used. 



I may mention here incidentally that a number of North American 

 species are retained under the genus Unio by Simpson. I do not 

 think that they should remain congeneric with the European forms, 

 for reasons which will be set forth in another paper. For most of 

 the American forms the generic name Elliptio Rafinesque, 1819, 



