138 TRANSACTIONS OF THE [MAY 9^ 



tremity, and widening to four inches behind. It is marked on 

 the upper margin by a series of broad shallow pits for the inser- 

 tion of comparatively few but large teeth. The teeth are coni- 

 cal, acute, and striated at the base, where they have a convoluted 

 structure like those of Dendrodus. A large bone, probably a 

 coracoid, is two inches by one and a quarter in diameter in the 

 centre, expanded at either extremity and must have been orig- 

 inally about one foot in length. Like the jaws, this consists of 

 a thin shell of bone surrounding a large area which was un- 

 doubtedly occupied by cartilage. This structure is peculiar 

 and suggested the name Ccelosteus (hollow bone) given to the 

 genus. Dr. Newberry said this fish was apparently allied to 

 Dendrodus and Rhizodus, which had mandibles somewhat similar 

 in form, set with few large teeth implanted in broad, shallow 

 sockets. In Rhizodus, however, the dentary bone was segmented, 

 while in Cmlosteus it was entire. The specific name given to 

 the new fish was feroz, with reference to its formidable dentition. 



Br. Henry H. Rusby read a paper 



FURTHER NOTES OF A SCIENTIFIC JOURNEY ACROSS SOUTH 



AMERICA, THROUGH THE VALLEYS OF THE BENI 



AND MADEIRA RIVERS. 



May 9, 1887. 



Stated Meeting. 



The President, Dr. J. S. Newberry, in the Chair. 



Twenty-three persons present. 



A list of accessions to the Library was read by the President. 



In reply to questions by Mr. Chittenden, the President 

 discussed the ''Taconic System" of Emmons, and affirmed the 

 truthfulness, in general, of that author's generalization. 



The following paper was read by title : 



