70 INTRODUCTION. 



necessarilj produced by a corresponding difference in 

 the structure of the animal. This principle, which 

 is exceedingly plausible at first sight, ought, if true, to 

 be followed by another axiom which depends upon it, 

 namely, that great diversities of the animal structure 

 are attended by great differences in the shell ; and both 

 ought to be confirmed by well-known facts. But, in 

 truth, neither of these principles is practically of 

 much value in the diagnosis of species, and every one 

 famihar with shells, knows, that very considerable dif- 

 ferences occur in all the particulars we have named, 

 while the animal remains unchanged ; and on the other 

 hand, it is equally well understood that considerable 

 modifications of the animal structure may exist, without 

 anything hke a corresponding amount of variation in 

 the shell. A remarkable instance of this occurs in the 

 marine genera Lottia and Patella, in which the shells 

 resemble each other very nearly, but in which a separa- 

 tion of the two genera has been justified by important 

 differences in the animal structure alone. In the branch 

 which we are considering, though not particularly among 

 the species of this country, the naturahst can find no 

 differences of animal structure to sustain the generic 

 distinctions which, if he relied upon variations of the 

 shell, he often would found upon these external char- 

 acters. 



The second remark which we make, as resulting 

 from the preceding observations, is, that no one is in 

 a position to estabUsh species with confidence, even 



