CHAP. V. ANALOGIES OF THE COLUMBELLIN^. 153 



explaining the singular mixture of characters concen- 

 trated in this sub-family j for the only one by which 

 all its genera are bound together,, consists in the outer 

 lip being invariably inflexed, and either striated or 

 toothed. Some of the analogies are weak, others strong; 

 but as they follow each other in the same series as that 

 of the primary divisions of the family, this regularity, 

 of itself, is quite sufficient for our purpose. The sin- 

 gular tuberculated tooth at the bottom of the aperture, 

 in several of the Nitidellce, their flattened pillar ter- 

 minated by an internal fold, the absence of a definite 

 inner lip, and the effuseness of the aperture, are all 

 explained, by the following table. 



Analogies of the Genera of the COLUMBELLIN^U 



Genera Sub-families Genera Sub-families 



of the Analogies. of the of the of the 



COLUMBELLIN.E. VOLUTIN^E. VOLUTIN.E. MURICID.E. 



Colurnbella. Sub-typical. MITRA. CYMBIOLA. MDUICINJE. 



Pustostoma, Typical. VOLUTA. VOLUTA. CASSINJB. 



C Outer lip much } 

 Crassispira. < thickened, and > M A RGINELLA. SCAPHELLA. BUCCININ^. 



C notched above, j 



f Obsolete plaits af| 



| the base of the | 



Nitidella. 4 PJJJJJ.J ap 2$JJ" > ANCILLARIA. VOLCTILITHES. PURPURINJE. 



I times with an in- 1 

 I ternal tooth. J 



r Spire short; inner} 

 Corudea^ 4 lip much deve- > OLIYA. HARPULA. NASSINJE. 



1 loped. 3 



(142.) But the Columbettince possess other resem- 

 blances, among which that to the mitres has been thought 

 so strong, as to induce the early reformers of conchology 

 to place them both in the same genus ; hence it was 

 taken as a resemblance of affinity. We have seen, how- 

 ever, by the analysis of the Volutidfe, that this idea is 

 erroneous. The impression undoubtedly originated in 

 the strong resemblance between the strombiform division 

 of the mitres, forming our genus Harpella, and the 

 typical ColumbellfB. Yet even this, although a striking 

 instance of relationship, is only analogical, as will be 

 seen more clearly by the following table : 



