CHAP x. UNIONIN.E. THELIDERMA. 273 



where the form of the shell is nearly semicircular, so that 

 we again return to irrorata, from whence we first set 

 out. Four types of the group are now disposed of, but 

 another is obviously wanting ; because, although there 

 are representations of four genera of the Unionida, we 

 have not yet noticed the prototype of Anodon. Now, 

 this type is in the Manchester Museum, and was sent to 

 us, many years ago, by our friend professor Rafinesque, 

 without a name. By the annexed figure (d), the con- 

 chologist will perceive that, in its outward appearance, 

 no one would suspect it was really different from the 

 Alasmodon rugosa of Barnes ; yet, on opening the valves, 

 we find the perfect cardinal and lateral teeth of Theli- 

 derma. The anodons are remarkable for the smallness 

 and the depression of their summits j and this is the 

 only shell of its own group which possesses this cha- 

 racter, while its form is exactly intermediate between 

 cylindrica and plicata. Condensing the results of these 

 remarks, we find the sectional types of this sub-genus 

 will represent alT those of the entire family, and in the 

 following manner : 



Analogies of the Sectional Types of THELIDERMA. 



^rtitmnJ Genera Genera 



T, Analogical Characters. of of 



Types - Unionidte. Unio. 



Irrorata. Semicircular, but not winged. UNIO. UNIO. 



Metanevra. Semicircular, and winged. HYRIA. 



Pustulata. pval^winged; summit, very j ANODM , LYMNAMA . 



PUcata. * "<> iMl A-HO . PoTo, D , 



Thus we find that every thing apparently anomalous 

 in T. cylindrica is at once explained by its being the 

 link between Ligumia and Theliderma; for if a shell so 

 formed had not been discovered, this link would have 

 been wanting. It has been stated that the crassidens, 

 Peruviana, and rariplicata of Lamarck, the undulatus 



