CHAP. IV. VOLUTA TYPES OF FORM. 113 



an instance of that inequality in analogical relations, 

 which, in our former volumes, we have so frequently ad- 

 verted to. This inequality, strangely enough, has been 

 somewhere urged as an objection to the theory of repre- 

 sentation. Such reasoners seem to suppose that, unless 

 all groups possess the same degree of resemblance to 

 each other, the evidence is inconclusive. On this plan, 

 an eagle and a lion may perhaps be admitted as analo- 

 gous ; but a carnivorous insect and a carnivorous beast 

 cannot be so, because the analogy is remote or obscure. 

 But if there are any analogical resemblances in nature, 

 it follows, as a necessary consequence, that such resem- 

 blances are strong or faint, near or remote, according 

 to the proximity or distance, the similarity or the dis- 

 similarity, of the objects compared. Thus it is in the 

 present instance : the groups being remote, their analo- 

 gies are not strong ; nevertheless they are substantially 

 true, because they are perfectly verified through the 

 medium of other or intervening groups, which, from being 

 more alike, render the analogies, in the same proportion, 

 more obvious or direct. 



(104,) We may now proceed a step further, and 

 apply the same description of proof to the sub-genera, 

 or types of form,, of the typical genus Valuta, upon the 

 affinities of which we have already said so much. We 

 leave it to be determined by others, whether these types 

 of form should be designated as simple divisions, or as 

 sub-genera, and therefore to be distinguished by a pa- 

 tronymic name. That this will ultimately be done, we 

 have no doubt ; because they are of the same rank as 

 the sub-genera of the other families. The shells which 

 we view as types of form in the restricted genus Valuta, 

 are as follows : Valuta Neptuni, V. imperialis, V. 

 Scapha, V. angulata, and V. magniftca. We shall now 

 place these in one column, the two extremes of which, 

 as we have already shown, meet, and form a circle : 

 the next column is composed of the genera of the La- 

 marckian volutes, which also form another, circle. Now, 

 if the contents of one represent the contents of the other, 



