44 H. B. POLLARD. 



and concluded that the structural coincidencies in the two forms 

 " must lead us to assign a place near, if not among the Siluroidei 

 to Coccosteus." 



This view has not met with general acceptance and Traquair writes 

 " Undoubtedly, the weakest point in Professor Huxley's ' Essay ' is the 

 attempt which he made to show by comparison of the exoskeletal 

 plates of Coccosteus with the bonej visible on the exterior of the 

 skeleton of many recent Siluroids, that there was a possibility at least 

 of the enigmatical group of the Placodermata turning out to belong to 

 the great order of Teleostei, or ordinary bony fishes, ' hitherto supposed 

 to be entirely absent from formations of palaeozoic age.' Recent dis- 

 coveries in the palaeozoic rocks of America point, as we shall presently 

 see, to another, and, perhaps more probable solution of the question." 



The " perhaps more probable solution " is given by the discovery of 

 Dinichthys. Newberry discovered that Dinichthys has a dentition like 

 that of Protopterus, and therefore concludes that it is allied to the 

 Dipnoi, Dinichthys being also allied to Coccosteus, it follows that 

 Coccosteus is allied to the Dipnoi. 



Nevertheless, too much stress must not be laid on a single feature. 

 Fusion of teeth to form great dental plates has occurred over and 

 over again in the Vertebrates as for example in Plertognathi and in 

 Hatteria. The jaws of Dinichthys have far less resemblance to those 

 of the more archaic Ceratodus, where the teeth lie on the inside on 

 the lower jaw, than to the more modified Protopterus. Therefore, the 

 resemblance may be due to convergence. Other features forbid entirely 

 the close alliance of Dinichthys and Protopterus. The latter has no 

 structures corresponding to the spines of the former. The whole 

 dermal armature is entirely different, and finally the distribution of 

 the lateral line system, as figured by Cliypole, is in no respect like 

 that of Ceratodus or Protopterus (I have examined both of the latter 

 animals on this point), while it bears a remarkable similarity to that of 

 Clarias, as figured by me, 



Returning to Coccosteus, I may state that I have examined a number 

 of specimens and the dermal armaUire certainly shows no affinity with 

 Dipnoi. Elsewhere I have endeavoured to prove that the lateral line 

 of Clarias closely resembles that of Coccosteus, thus offering confirmation 

 of Huxley's view. 



The Siluroids are therefore not classed with the Cephalaspidae and 



