FOSSIL FISHES. H 



and somewhat more rounded in outline in Diniohthys that in Ooccosteus, 

 but less so than might be inferred from the figures; as in that of the plas- 



VKNTEiL Shield of Coooosteus decifieks, Agass. 



ITatnial size ; after Pander. 



tron of Ooccosteus — which is taken from Pander's restoration — the pos- 

 terior lateral plates are represented as overlapped by the anterior laterals, 

 and thus their anterior margins are concealed. One of these plates is also 

 made by Pander in his restoration to overlap the other, and the posterior 

 end of the plastron is thereby much narrowed. What evidence of such 

 overlapping he had does not appear, as no indications of it are seen in 

 his figures of the fossils themselves, either the individual plates, or the 

 groups in which they are shown. The posterior lateral angles of the 

 hinder pair of ventral plates are produced in both Dinichthys and Coccos- 

 teus, but much more so in the former than in the latter. In Dinichthys 

 this portion of the plate is greatly thickened, and its interior face is exca- 

 vated in a deep longitudinal furrow, which must have served to hold and 

 protect the nerves and blood-vessels that passed out to the ventral fins. 

 This furrow is shown in the figures of the post-ventro-lateral plates, which 

 are given, of the natural size, on Chart VI.* 



* I have elsowhere called attention to the fact that in position and form these bones 

 correspond so closely with the pubic bones of PledoBwwrus as to suggest homology with 

 them. It will also be noticed that the anterior pair of ventral plates offer us striking 



