ON CEPHALASPIS AND PTERASPIS 515 



it could be for a moment confounded. Its relations with Cephalaspis, 

 on the contrary, are very close. In each the shield is excessively 

 thin, and composed of three or four layers : — ist, an "internal," com- 

 posed of lamellae parallel with the surface, and traversed more or less 

 obliquely by vascular canals ; 2nd, next to this is a " middle layer," 

 containing the network of wide canals or areolje ; 3rd, the " reticular 

 layer," described in CepJialaspis as part of No. 2, from which it is not 

 distinctly marked in that genus ; 4th, the " external layer," consisting 

 of a cosmine-like substance raised into ridges or tubercles. 



The " bony granules," or " prisms," supposed to be characteristic 

 of Pteraspis, the " polygonal ossicles " and the " fibrous bony layer," 

 supposed to be peculiar features of CepJialaspis, have, as I have shown, 

 no existence. Supposing that the shield of Pteraspis, like that of 

 Cephalaspis, covered the animal's head (though there may be some 

 ground for entertaining a doubt on this point), then it may be said 

 that the presence of orbits in one, and their absence in the other, 

 indicates a wide difference between the two genera. It must be 

 remembered, however, that there is precisely the same difference 

 between Pterichthys and Coccosteus, which are admitted by all to be 

 closely allied. 



Though I have had no opportunity of examining the Russian 

 species, I believe I do not err in regarding what Pander describes as 

 the teeth of CepJialaspis as merely an excessive development of the 

 marginal tubercles of the outer layer. It does not appear to me that 

 there is any evidence that the mouth was situated at the margin of 

 the shield ; on the contrary, the inward prolongation of the reflected 

 ventral layer leads me to suspect that the under surface of the head 

 ■of CepJialaspis resembled that of Loricaria or of Acipenser. 



Zoological position of CEPHALASPIS and PTERASPIS. 



Leaving for the present Professor Pander's " Conodonts " out of 

 view, CepJialaspis and Pteraspis are among the oldest, if they are not 

 the very oldest, of known fishes ; and it is therefore highly interesting 

 to inquire into their position in the scale of ichthyic nature. 



Palaeontologists in general, following Agassiz, classify them as 

 " Ganoids ; " but it is to be feared that few persons who have not paid 

 special attention to recent Ichthyology and to Comparative Anatomy 

 have a clear conception of what is meant by the term " Ganoid." 



The founder of the Order, allowing himself to attach an undue 

 weight to mere secondary characters, included under the head of 



L L 2 



