PISCES. 239 



palaeontologist. The family of Cephalaspides exhibited features so bizarre, 

 as to cause them to be placed anywhere else than among fishes : it was 

 Agassiz who first recognised their true character, and placed them in the 

 position to which they properl}' belong. The essential character is to be 

 found in the broad, tesselated, bony plates, which encircle the head and a 

 part of the trunk. The form, number, and arrangement of these, vary 

 greatly in the different genera, although they possess one general character 

 in the enamel coating, the smooth inner face, and the variously marked or 

 granulated exterior. The head is covered by a simple or compound, 

 always liat, buckler, of various characters. The body, like the head, is 

 flat, and variously covered with plates. The fins exhibit a peculiar 

 development, the ventrals are entirely absent, the pectorals of a narrow 

 rayless plate, situated behind the head, produced more or less into a wing. 

 With a single exception, the caudal fin is entirely wanting, and the dorsal 

 and anal, when developed, never attain to any considerable size. In the 

 structure of the skeleton they exhibit some close affinities to the sturgeons. 

 No species have as yet been found in North America. 



The concluding family of the Chondrostei, Holoptychii, is composed of 

 fishes with slender and powerful bodies, thick heads, wide jaws, and well 

 developed fins. The jaws are furnished with small, sharp teeth, at the 

 edge, and with a few others that are very large, strongly conical, at 

 considerable distances apart ; these, with the fins, indicate a highly 

 predacious character. All the teeth are covered with vertical folds, which 

 become lost towards the apex. The scales, in form and arrangement, 

 resemble those of the true cycloids, and overlap each other in oblique 

 series. Their coating of enamel, however, indicates clearly their ganoid 

 structure. Even the bones of the head are covered with enamel, and 

 variously sculptured on the surface. The fin rays and bones, as far as 

 these exist, possess internal cavities, an unique character peculiar to these 

 fish. The genera belong to the old red sandstone and the Devonian, but it 

 is doubtful whether any species, either of this family or of the preceding, 

 occur in North America. The reported occurrence of Holoptychius 

 7iohilissimus in Pennsylvania and New York wants confirmation. 



Before commencing the consideration of the trul}^ cartilaginous fishes 

 forming the division Selachii, as distinguished from the Teleostei and 

 Ganoidei, it will be necessary to dwell for a moment upon the ninth order 

 of the tabular classification placed at the head of our article, and 

 constituted by a single family, the Sirenoidei. This family includes two 

 species, of, perhaps, two different genera, the Lspidosiren paradoxa from 

 Brazil, and Lepidosiren, perhaps Protopterus annectens, from the Gambia 

 River, Africa. By most Continental naturalists the Lepidosiren is 

 considered to be a reptile, while Professor Owen is confident as to its 

 icbthyal character. It in fact combines the characters of both reptile and 

 fish, to a most remarkable degree, the African species inclining more to the 

 latter, the South American to the former. Deferring further consideration 

 of the subject until we come to the class of Reptiles, we proceed to the 

 subject of the Selachii, above referred to. 



443 



