472 On an undescribed Fish. 



Its anterior border has an articulation witli tlie rostrum 

 ventrally, and dorsally is cut out into two shallow notches 

 which must have formed the posterior margin of a^ incom- 

 pletely divided nostril. 



Parts of four circumorbitals are shown, three of which 

 are unfortunately not in their natural positions. The most 

 dorsal and largest of these bones projects outward from the 

 frontal, so as to carry tlie orbit away from the middle line. 

 It is quite certain that the orbit was small. 



The maxilla and the preoperculum are as Mr. Gill has 

 described them, and there is in addition a small triangular 

 bone lying between the latter element and the supratemporal, 

 which appears to be homologous with the accessory pre- 

 operculum described by Traquair in Cheirolepis. 



The bones of the operculum are well exposed on the right 

 side ; tliey are a large opercular ornamented with sparse 

 longitudinal ridges, a small subopercular, and a long series 

 of branchiostegals, the uppermost not much smaller than the 

 subopercidar. It is impossible to see if large lateral gulars 

 and a median gular are present or not. 



That Mr. Gill is correct in referring this remarkable fish 

 to the Palseoniscidse is obvious, but, as he has brought out, 

 it represents a very aberrant type. The whole build of the 

 fish, its squamation, the character of the tail-shaft, and 

 especially the short-based pelvic fin and the small number 

 and large size of the unjointed dermal rays in the fins are 

 totally unlike those of any other Palaeoniscid, and suggest 

 most strongly a higher fish and especially a Semionotid. 



The skull and shoulder-girdle, however, are conclusive 

 evidence of Palseoniscid affinities, and I believe that w^e must 

 see in Phanerorhynchus another case of the precocious 

 acquirement of characters in a short-lived side-line, which, 

 more slowly reached by a conservative stock, lead on to the 

 establishment of a new order. In other words, this fish, 

 although itself not ancestral to any later form, shows that 

 the Palseoniscids had latent in them the potentiality of 

 giving rise to the Lepidosteoid Ganoids, and hence to all 

 higher bony fish. 



It is obvious that Phanerorhyyichus differs from all other 

 fish so greatly as to deserve at least family distinction ; but 

 as to remove it from the family Palj^eoniscidse would obscure 

 its close relations to those fish, and as that family as 

 at present constituted includes forms presenting a very 

 great range in structure, I do not propose to establish a new 

 group for its reception. 



