XIV 



the dentition is the most important per se, but because, as a 

 matter of fact and experience, it has been determined that the 

 modifications thereof are the co-ordinates of corresponding, though 

 perhaps not as readily recognizable, modifications of other parts, 

 and being so, they are taken advantage of for diagnostic purposes. 



In like manner, as a matter of experience, the groups of the 

 Pectinibranchiate mollusks agreeing in the dentition of the 

 radula appear to agree in other important respects, and there- 

 fore the modifications of the teeth of the radula have been made 

 use of as the prime characters, because they appear to be the 

 exponents of the sum total of structure, and until it is shown, 

 by a study and co-ordination of the modifications of the entire 

 structure, that there are other characteristics that are of more 

 importance and better indexes of affinity, and the application has 

 been actually made, it is not evident what other better combina- 

 tions capable of demonstration and diagnosis the true criteria 

 can be made. Undoubtedly we have much yet to learn concern- 

 ing the affinities of all the mollusks, and undoubtedly very con- 

 siderable, and perhaps fundamental, modifications of classification 

 will be required ; but, in addition to objections against a given 

 system, suggestions for reform are at the same time desirable, 

 and then a comparison of the respective merits of the competing 

 systems can be instituted. 



As it is evident that the differences of dentition in the Placen- 

 tal and Implacental mammals is of very unequal value, it is no 

 more than might be expected that the dentition in the class of 

 Gasteropods should also vary in value, and it is actually found 

 that while in the Pectinibranchiates the dentition is an excellent 

 index of affinities, it is not so in the Tectibranchiates or Nudi- 

 branchiates. In this admitted fact, however, there is no more 

 valid argument against its value in the Pectinibranchiates than 

 in the corresponding case in mammals. 



EXTINCT FORMS. 



With respect to the extinct forms, the compiler has deemed it 

 advisable to accept the views of the most approved students of 

 the groups as to their relations, but has felt obliged to apply to 

 them that indefinite but generally appreciated standard of value 

 which has been used for the living forms, and consequently the 



