PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 93 



to fiud specific differences. They are probably sijecimens which have de- 

 veloped iiuder more favorable circumstances than are usually accorded 

 to the species. 



33. Enneacanthus obesus (Grd.) Gill. 



34. Enneacanthus gloriosus* (Holbr.) Jor. 



35. Enneacanthus eriarchus (Jor. ) Mc'lvay, 



10. Genus Micropterus Lacepede. 



36. Micropterus sahnoides (Lac.) Hensliall iu "Book of tlio Black Bass" (advance 



sheats) =^ 2Iicro2)teri(s paUidus Gill & Jor. 



37. Micropterus dolomieu Lacdpede = lficrq/>/er«s salmoidcs Gill. 



Indiana University, 



Bloomington, Inch, March 10, 1881. 



A REVIEW OF THE GENUS CENTIIBUS, SWAINSON. 



By ROBERT RIOGWAY. 



INTRODUCTION. 



The collection of the United States National Museum contains ex- 

 amples of all the known species of the genus Centums, excepting C. hypo- 

 polius (Wagl.), C. ruhriventris Swains., and C. terricolor Berlepsch, the 

 two latter being of some what doubtful status. More or less confusion 

 has hitherto existed regarding the nomenclature of several of the Mid- 

 dle American forms, and it was the desire to clear away as much of this 

 confusion as possible that prompted the investigations upon which this 

 review is based, and which have led to the discovery of relationships 

 which were before quite unsuspected, at least by the writer. 



Of the fourteen forms treated of in this paper as sufficiently distinct 

 for definition, not more than six, or less than one-half, can be said to 

 be perfectly isolated, or to possess the requirements of perfectly distinct 

 species ; at least the abundant material which has been examined in this 

 connection i)roves beyond question the intergradation of four so-called 

 species, while it suggests more or less strongly the probability or possi- 

 bility of such relationship with regard to five of the remaining ten. 

 Those which appear to be unquestionably distinct are the three West 

 Indian species, C. radiolatus (Wagl.), C. svperciliaris (Temm.), and C. 

 siriatus (MiilL), and three continental species, G. uropygialis, Baird, C. 

 hypopolius (Wagl.), and G. elegans (Swains.). Those which certainly in- 

 tergrade, and are therefore to be united under one specific designation, 

 are G. aurifrons (Wagl.), G. santacruzi Bp., G. dubius (Cabot), and G. lioff- 

 manni, Caban., all of which are, however, strongly characterized geo- 

 graphical races or sub species. The five forms of doubtful relationship 

 are (1) G. carolinus (Linn.), which may possibly grade into G. ruhriventris, 

 but which is probably distinct; (2) G. ruhriventris (Sw.), with which I 



