244 C 



portcil \)\ all of the basilar Ijoiics. Six of" the latter articulate in the fossae of 

 (he groove of the scapula already described. They are flat, contracted at tlic 

 middle, and expanded at the extremities. In front of these are two others, 

 of a short, thick, cylindric form, one applied to the superior, the other to the 

 inferior fiicets of the scapula above mentioned, while tlie tuberosity rises 

 pedestal-like between them. This structure gives a slight hinge-movement, 

 like the opening of the blades of a knife, and entirely unlike the rotary hinge- 

 movement characteristic of the Siluridfe. 



The spine is composed of parallel rods in close apposition. The ante- 

 rior edge being oblique, the extremities of the rods terminate successively al 

 tlie border, which is trenchant, constituting the ofTensive part of the spine. 

 The edge is hardened and the adjacent parts of the spine thickened, and in 

 some cases roughened by a deposit of a hard substance resembling enamel. 

 It is either straight, or regularly undulate or serrate, with recurved, acute, 

 tooth-like processes. The smaller species exhibit the serrate character ; the 

 larger, the regular border. In either case, a most formidable weapon is indi- 

 cated, not less admirable than those already described from Paleozoic rocks. 

 There is a considerable resemblance between the serrate type and the spines 

 of the Carboniferous genus Edestus, where the teeth are more developed and 

 denticulate. 



Spines of this type were referred by Professor Agassiz to tiie cestra- 

 ciont sharks of the genus Ftychodus ; but the structure of the scapular arcii, 

 now first described, demonstrates the incorrectness of this association. The 

 increased number of the basilar bones shows clearly that they belong to a 

 peculiar family, which I call the Pdecopteridce. The principal genus receives 

 the name Pclccoiiterus, leaving the question of its relationship to ErisiclUlie 

 an open one. The supposition that these spines belong to Portheus and its 

 allies is, therefore, no longer entertained. 



PELECOPTERUS, Cope. 



The characters of this genus are only known from the bones of the 

 scapular arch and the spine of tlie pectoral fin. As these have been already 

 described under the head of the liunily, it remains to discriminate the species. 

 Three of these have come under my observation, which are fishes of large 

 size. They may be readily distinguished by the character of the cutting- 

 edge and the relative size and direction oi'tlic coniponeni rods. 



