CATFISH GENUS NOTURUS RAFINESQUE 53 



he left no figure or specimen of Silurus mollis in the Strasbourg collec- 

 tion, and may never have handled the specimen. 



4. The "Musei Humphrediani" was "A [sale] Catalogue of the Large 

 and Valuable Museum of Mr. George Humphrey. . . ." It is dated 

 1779, and item 33 on page 131 reads "a globe fish, another American 

 fish, and a snake, in three bottles." [Copy seen in the British Museum 

 of Natural History.] Lists of European natural history collections do 

 not indicate the disposition of this collection, nor its present location, 

 and I have failed to locate Hermann's specimen. Humphrey apparently 

 was not the collector of the fish; he was primarily a dealer in shells. 



5. The meaningful characters in the description are: the name Silurus, 

 a posterior dorsal fin that is adipose, twelve anal rays, and a very soft 

 and flabby belly. In themselves, these are not diagnostic. In addition, 

 it is stated that there are four cirri of the upper jaw and that the 

 specimen came from America. America could mean either North or 

 South America. The meaning of the four cirri of the upper jaw is un- 

 certain, as no lower barbels are indicated for the specimen. The body of 

 the description "radio primo . . . mollibus candidis," is confused and 

 apparently cannot be meaningfully translated. It indicates among 

 other things something that is soft and (?) white and that a spine is 

 associated with the head, neither character applying to Noturus 

 gyrinus. 



To summarize, the description of Silurus mollis is indefinite; it may 

 not apply to an ictalurid; and the specimen from which the description 

 was made may be lost forever. Thus, until shown to be otherwise, I 

 agree with Fowler (1945, p. 123) that the nanrfe is unidentifiable; it 

 should be removed from association with Noturus gyrinus. 



The original description of Silurus gyrinus as published by Mitchill 

 (1817, p. 289) is perhaps inadequate, and led to some confusion among 

 early workers. The later and more complete description given by 

 Mitchill (1818, pp. 241, 322-323) unquestionably identifies his ma- 

 terial with this long known form. Characteristic are the locality, the 

 pectoral formula, the pectoral spine, the tail, and the easily overlooked 

 adipose fin. 



The type of Noturus sialis Jordan has not been located and probably 

 was destroyed by fire at Indiana University. The name, however, has 

 long been correctly placed in the synonymy of Noturus gyrinus as 

 confirmed by the descriptions and by Jordan's action. 



According to Mitchell (1904), the name SchUbeodes punctatus is a 

 lapsus for SchUbeodes gyrinus. 



Etymology. — The name gyrinus (Greek) means tadpole, in allusion 

 to its resemblance to a larval salientian. 



Relationship. — Among the species of Noturus, N. gyrinus appears 

 to be more closely related to N. lachneri than to any other. Both 



