THE NAUTILUS. 23 



writer's monograph are slight, they have thus far proven reason- 

 ably stable. In any classification it must be the sum-total of 

 characters that provide diagnostic features of value, and even if 

 these be small they are sufficient if they include certain groups 

 to the exclusion of others. No single character, as the radula 

 or the genitalia, will prove satisfactory. A case in point is 

 Radix, in which the lateral teeth of the radula are now known 

 to be either bi- or tricuspid. As the writer stated in his mono- 

 graph diagnostic features must be based on the sum of the char- 

 acters presented by the shell, radula, genitalia, or other organs. 

 Colton says " our present knowledge will not allow us to make 

 a comprehensive classification of the Lymnaeids based on the 

 anatomy of the snail." This result cannot be attained by re- 

 casting the data now available, but rather by the addition of new 

 data on old species or on species now unknown anatomically. 



The writer can by no means agree with the statement made 

 twice in this paper that generic names should not be added un- 

 less based on undebatable grounds because of the inconvenience 

 to the cataloger. If this criticism should be recognized we 

 would revert to the use of many of the older names in the Pul- 

 monata as well as in the Naiades. It is recognized, of course, 

 that generic subdivision can be overdone, but in the advance- 

 ment of science the convenience of the cataloger or teacher is 

 not considered. Generic or other divisional names are simply 

 for the purpose of bringing together groups of similar organisms 

 which we designate as genera, subgenera or sections. In some 

 cases the criteria for the separation of these genera will be of a 

 distinct and decided character; while in others, where there are 

 many species of similar characteristics, these distinctions will 

 necessarily be founded on data of a less decided character. 

 Such a condition obtains in the family under discussion and 

 whatever the criteria used for the separation of genera or other 

 groups, they can apparently be of only quantitative character. 

 The relative value of these criteria will vary with the import- 

 ance ascribed to them by different authors. 



We welcome all additions to knowledge and we know full 

 well that the work of yesterday is rendered obsolete by the work 

 of to-morrow, but the writer cannot see how the reduction to 



