REPORT ON ICHTHYOLOGY. 413 



lurus given by Rafinesqne is, perhaps, the best description of a genus given in his 

 work, and is thought worthy of being copied: 



"Head depressed, with eight barbs, one at each corner of the mouth, longer than 

 the others, four under the chin, and two on the snout behind the nostrils. Teeth in 

 two patches, acute and file-shaped. Pectoral fins and first dorsal fin tinned with an 

 anterior spine. First dorsal trapezoidal and before the abdominals: second opposite 

 the anal. Body compressed behind, vent posterior and sub-medial. Operculum simple.'' 



By the above limitation, the subgenus Ictalurus is seen to partly correspond with 

 that of Pimelodus of Cuvier, the teeth being said to be in two patches or only on the 

 jaws. By the description of the condition and position of the fins and the number of 

 barbels, it includes only a small section of the Cuvierian subgenus. 



The name Ictalurus must be then reserved for some of our Siluroids — for all, if 

 they should be found to be congeneric — for a section, if it is ascertained that several 

 genera are embraced under the subgenus. 



Our studies of the Siluroids have convinced us that there are four natural genera 

 found in the United States, three of which were included by Rafinesque in his sub- 

 genus Ictalurus, but placed at the same time in sections, which received from him 

 various scientific names. 



The sections established by Rafinesque were chiefly characterized by the form of 

 the "tail" or caudal fin, and of the eyes, and the number of rays in the abdominal or 

 ventral fins. 



The first section was named Elliops, and included fishes with the "tail forked. 

 Eyes elliptical. Abdominal fins with less than nine rays." 



This group exists in nature, and is of generic value, but the characters given by 

 Rafinesque are not those which essentially characterize it, nor can the name Elliops be 

 retained for it. 



The name given to a group as a whole must be preserved, and if that group is 

 divided into sections, one of those sections must retain the name of the greater group 

 In Rafinesque's system, Ictalurus is the greater group, and in it are included all the North 

 American Pimelodi, with the exception of Noturus. When Rafinesque divided that 

 group into sections, he should, therefore, have still retained that name for one of them. 

 Such has not been done, but upon each of his sections was conferred another name. 

 As this is in opposition to the rules of nomenclature, Ictalurus must be restored to one 

 of liis sections, and it is advisable to retain it for his first, and reject the name of 

 Elliops. The section with this name is now accepted as a genus; its diagnosis will be 

 hereafter given. 



The name Pimelodus, it is true, was applied to all the Ictaluri, and by that name 

 only are they called. If Pimelodus had been of Rafinesque's creation, that name 

 should, therefore, have been adopted; but as Rafinesque has only taken it from La- 

 cepede, with the characters given to it by its founder, it is to be supposed that he 

 intended it to be otherwise restricted. It appears to us that it is no valid argument 

 against the acceptation of Rafinesque's names for genera, if his sections should prove 

 to be such, that he did not apply them specifically. 



The section called Elliops, on comparison with its type Pimelodus cmrulescens of 



