II. DESCRIPTION OF A COLLECTION OF FOSSIL FISHES 



FROM THE BITUMINOUS SHALES AT RIACHO DOCE, 



STATE OF ALAGOAS, BRAZIL. 



By David Starr Jordan. 



Tlie collection of fossil fishes described below was made under the 

 direction of Dr. John Casper Branner in 1907 for the Carnegie Museum 

 of Pittsburgh. The types described in this paper belong to that museum. 

 Dui)licates of the species are in the paleontological collections of the 

 Department of Geology at Stanford University, California. The ac- 

 compan>ing drawings arc by Mr. Sckko Shimada. 



Family CLUPEID^. 



Genus Diplomystus Cope. 



Diplomyslus Cope, Bull. U. S. Geol. Surv. Terr., Ill, 1877, 808 {denlatus) (not 



Diplomystus Bleeker, a genus of catfishes) . 

 Copeichthys Dollo, Results Voyage Belgica. 1904, 159 {denlatus) (substitute for 



Diplomystus, considered as preoccupied). 



A large, deep-bodied compressed herring from the Green River Eocene 

 shales in \V>oming has been made the type of the genus Diplomystus. 

 The characteristic features of the genus are the very strong ventral plates, 

 and the presence of similar smaller plates on the dorsal line before the 

 dorsal fin. At the same time another herring, obviously related, but 

 ditTering in the slender form, short anal, and fewer vertebrae, besides 

 other characters, was associated with this species. This species, Cliipea 

 humilis Leidy = Clupea pusilln Cope, both names preoccupied, became 

 later, under the name of Kvightia eoccena, the type of the genus Knightia. 

 More or less intermediate between Diplomystus and Knightia are several 

 species from the fossil beds of Europe and Asia. 



Among Dr. Branner's Brazilian collections from Riacho Doce, are two 

 new species allied to Diplomystus. Another, related to these, called 

 Diplomystus longicostatus, has been already known from the Cretaceous 

 of Brazil. These Brazilian species have the general traits of Diplomystus 

 dentatus, with the short anal and fewer vertebrae of Kvightia, while at 

 the same time their squamation seems to be different from both. For this 

 group I suggest the name of Ellipes. We may thus recognize among 



23 



