of distinguishing; Species and Higher Groups. 201 



Although, in the class Conchifera, the shells do not furnish 

 good ordinal characters for the five orders of Mr. Gray's list, 

 yet all of the forty-two families which are distributed through 

 these orders may be easily distinguished by their shells. Also 

 the six families of the Brachiopoda, and those four of the 

 Pteropoda which are testaceous, are easily distinguished by 

 their shells. 



Thus of the 95 testaceous families of Mollusca, 88, or more 

 than 92^ per cent., are distinguishable by their shells. 



For the generic value of the shells, we may take the re- 

 sults in Mr. Gray's list, which is remarkable for its multipli- 

 cation of genera; 810 recent genera are recognized. Of these 

 152 are naked, and 648 are testaceous. Of the latter, the 

 shells of two genera are said by Mr. Gray to be indistinguish- 

 able from certain annelides ; and of seven others it is Isaid to 

 be impossible to characterize them by the shells only. This 

 list, although small, is susceptible of reduction ; for M. Des- 

 hayes has pointed out the difference between the shells of Ver- 

 metus and of Serpula, and Dr. Gould has discovered generic 

 characters in the shells of Lottia, distinguishing them from 

 Patella. Of 648 testaceous genera, 641, therefore, may be 

 distinguished by their shells. This amounts nearly to 99 per 

 cent. 



Although the soft parts of 5,000 species of Mollusca, includ- 

 ing both the naked and the testaceous species, are known, we 

 are not aware of the existence of a single testaceous species^ 

 which is not distinguishable by the shell with as much facility 

 as by the soft parts. The examples are very few, indeed, if 

 any exist, in which the species are not recognized by their 

 shells with much greater ease than by the soft parts. 



Of the 20,000 species of shells which are known, the spe- 

 cific value of many is more or less doubtful. It is barely pos- 

 sible that a knowledge of the soft parts will aid materially in 

 determining such species. But this is not probable, since the 



