FISHES OF THE OLD RED SANDSTONE. 



595 



vated edges of the successive plates of growth. A clear proof of this is, that on making a trans- 

 versal cut through the shell, nothing is perceptible on one side but small perpendicular clefts which 

 terminate abruptly, and on the other, notwithstanding the regularity of these striae, they are some- 

 times distinctly bifurcate. Finally, their similarity on both sides of the disk, without the least trace 

 of a longitudinal division on the inferior layers of the substance, is a character which does not in 

 any way accord with what is known to us in the enlargement of mollusks, and in the scales of the 

 tail of crustaceans ; whilst the analogy of these disks, with the heads of the Cephalaspis which 

 have been just described, is more striking the more they are examined in detail. 



" The head of this species is a more obtuse oval than the preceding j its anterior edge is completely 

 circular : we can, however, evidently see the impression of an ethmoid bone similar to that of 

 C. Lewisii, but narrower and longer ; it was even decidedly longer than it appears to be in fig. 9 ; 

 the original of which is broken on its anterior edges. The posterior border of the disk of the head 

 is cut off obliquely as in C. Lewisii, but it is not raised up. The lateral edges are inclined uni- 

 formly to one side, and follow the general curvature of the arch of the head, which is nowhere de- 

 pressed on its anterior part, as in the preceding species. Besides the coat of enamel which forms 

 the exterior surface of the disk, and which is particularly well seen on the right and interior side 

 of fig. 9, as well as on a fragment preserved towards the posterior edge of this same specimen, 

 there may also be distinguished two other layers of very different structure : one, which is the 

 centre of the shell, has a granulous structure similar to that of the bones of the chondropterygians, 

 and perfectly identical with that of the plates in C. rostratus ; the other, which is the inferior 

 layer, decomposes in laminae superposed one to another, like the plates of growth in molluscous 

 shells. This last layer is thicker than the other three. In the specimen (fig. that in which 

 the form is best preserved, no plates are seen except those of this third coat, which in part cover 

 the raised impression of the inferior surface of the head. In fig. 9, on the contrary, at the edge of 

 the disk, we recognize first the exterior coat, under which the middle coat is concealed, and a little 

 further in the interior of the cast, the plates of the inferior coat. Finally, in fig. 8, is seen a frag- 

 ment so broken as to present the three coats in their natural order of superposition. 



"There is no trace visible on the sides of the head of any horn-shaped lateral prolongation any- 

 more than in the C. Lewisii. These two spiecies ought perhaps , therefore, to he generically separated 

 from C. Lyellii and C. rostratus. 



" From what has already been said, the structure of the head of the Cephalaspis resembles singu- 

 larly that of the shell of crustaceans, which possess also an exterior coloured layer, under which is 

 found a layer of a granulous structure, and then a layer of lamellar structure ; and it was not 

 without long deliberation that I decided on considering the fossils represented in figs. 6, 7 and 9., 

 as the heads of Cephalaspis, rather than the terminal scales of some unknown crustacean. It is very 

 extraordinary that shields of which the ichthyological characters cannot be mistaken, in the species re- 

 presented in figs. 1, 3 and 5, one of which at least has been found with its body and fins, should have 

 exactly the same structure with other disks that might be supposed to be the tails of trilohites. The 

 difficulty of deciding on these fossils is greater, because the buckler and tail of many species of trilo- 

 hites have also the outer layer of their shell ornamented with furrows similar to those on the head of 

 the Cephalaspis, and their edges sometimes raised up as a border, like the posterior edge of C. Leiv- 

 isii. The constant presence, however, of the piece which seems to me to be the ethmoid bone, and 

 the longitudinal ridge along the middle of the disk, seems to decide the question, and compel us to 

 place all these specimens in the class of fishes, and to look upon them as the heads of Cephalaspis, 

 or of some nearly allied genera. Their particular structure requires the greatest circumspection in 



