12 Lnizersity of Michigan 



ognizable species (see below) and is congeneric with U. pli- 

 catiis Say, the type of Crenodonta Schlueter '36, Amblema 

 Raf. '19, thus supersedes Crenodonta. 



Amblema costata Rafinesque, 1820 

 Type locality: Ohio River and small rivers of Kentucky. 



Amblema costata Rafinesque, '20, p. 315, pi. 82, figs. 13-14. 



Unio undulatus Barnes, '2^, p. 120, pi. 2, f. 2. 



U. costatus Raf. = f/. undulatus Bar., Conrad, '34, p. 68; Ferussac, 

 '35, P- 27; Conrad, '2,6, p. 17. 



Quadrula undulata (Bar.), Simpson, '14, p. 819. 



Qiiadrula costata (Raf.) = Q. undulata (Bar.), Vanatta, '15, p. 556 

 ("type" examined). 



Amblema costata Raf. =1 U. undulatus Bar., Frierson, 'i(fi, p. 62. 



Amblema plicata costata (Raf.) = f/. undulatus Bar., Utterback, '16, 

 p. 39; Ortmann, '18, p. 538. 



Quadrula undulata (Bar.) =: A. costata Raf. or U. rariplicata Lam., 

 '19, Walker, 'i8c, p. 170. 



There is unanimity among all authors that A. costata of 

 Rafinesque is one of the "plicate" forms of mussels. This is 

 absolutely clear from the original description and figure, and 

 most authors (all except Walker) have unhesitatingly identified 

 it with U. undulatus Bar. of authors. 



Rafinesque calls his species in the description "applati" 

 (flattened or compressed), and thus it is evident that we 

 should refer costata to that form of the plicata group (genus 

 Amblema) which dififers from the others in the first place by 

 its compressed shell, and this is the shell commonly called 

 undulatus Bar. 



Thus, A. costata is identifiable from the original description, 

 which in addition has been verified by Vanatta by the exam- 

 ination of the so-called Rafinesque-Poulson type. The name 

 is valid. 



As to undulatus Bar., see under Megalonaias gigantea Bar. 



