CHAP. X. 



UNIONIN^. 



LYMNADIA. 



281 



Anodon : it is also so perfectly connate^ that Mr. Lea, look- 

 ing only to that circumstance, actually places it close to 

 L. alata. It is nevertheless a distinct type, and seems to 

 represent Anodon in this sub-genus : the cardinal teeth, 

 in short, strengthen us in the belief that Pofomida suc- 

 ceeds to Lyynnadia. We shall therefore introduce it here, 

 leaving its natural affinities still open to future inquiry. 

 (263.) In regard to the shells provisionally placed in 

 the genus Potomida, we have some doubts. We feel al- 

 most convinced that the types do not belong to the genus 

 Myfica ; and yet some of the aberrant species approach 

 so closely to such shells as M. pictorum and ovata, that 

 the gradation, to ordinary eyes, appears perfect. We have 

 one or two species from Brazil, and others from North 

 America, but most of the types seem to be European 

 These present us with the following distinguishing charac- 

 ters : the form, of course, is variable in the species ; but 

 it is never so long and attenuated a.s 3fysca; the breadth, 

 also, is greater, and the extremities more obtuse. The ty- 

 pical species, P. corrugata 

 ( ^^.57.):'isnearly asround 

 as the generality of the 

 Venus -shells : the cardinal 

 teeth are always two, — not 

 thin, lengthened, and com- 

 pressed as in Mysca, but 

 thick, short, and strong, so 

 as to resemble the true 

 American unios ; like these latter, also, the inner cardinal 

 tooth is decidedly the largest — but then they are divided 

 like those of Mysca and ^glia. The size of the teeth, 

 with the general shape, and robustness of the shell, in- 

 duce us to think that Potomida is not naturally associated 

 with either Mysca or Unio. The littoralis of Lamarck, 

 common in the rivers of France, w411 show all these cha- 

 racters ; but in another Continental species, the Unio 

 sinuata Lam,, the form is more lengthened ; and by the 

 direction of the cardinal cleft, this shell seems to approxi- 

 mate still more to the typical Unios. Next we have the 



