Remains of the Pycnodont Fish, Mesturus. 13 



The bony shield of the cranial roof is remarkable for its 

 close resemblance to the corresponding shield in Acipenser 

 and certain Siluroids. It is, in fact, the first instance of this 

 arrangement of plates recorded among fishes which exhibit a 

 trunk and fin-skeleton on the same biological level as the 

 Pycnodont type. Moreover, the disposition of the sutures is 

 evidently as capricious as that in Acipenser ; for not only do 

 the parietal plates of no. 1 lack bilateral symmetry, but two 

 specimens so closely similar in other respects as nos. 1 and 24 

 differ completely in the arrangement of the median roof- 

 elements. It is also to be noted that in the allied genera 

 Mesodon, Microdon, and Gyrodus there is only one parietal 

 plate on each side, while from this in the two first-mentioned 

 genera there arises a large posteriorly directed process with 

 digitate extremity, which extends immediately beneath the 

 dorsal scales of the trunk. In the Cretaceous Coccodus all 

 the elements of the cranial roof are fused together. The ex- 

 ternal bones of the Pycnodont skull are thus of an anomalous 

 and variable character. 



The opercular apparatus is equally remarkable, and only 

 parallelled among deep-bodied Teleostean fishes in which the 

 jaws are displaced far forwards (e. g. Lampris and the Gym- 

 nodonts). The small operculum has hitherto been described 

 as " supraclavicle ; " but if any evidence of its identity is 

 required beyond that described above, it is only necessary to 

 examine a specimen of Gyrodus in the British Museum 

 (no. 37108), where the element may be seen suspended by 

 the opercular process of the hyomandibular *. The identifi- 

 cation of the preoperculum (" operculum " of previous authors) 

 is confirmed by its demonstrated relationships in Anomoeodus 

 Willetti. The two branchiostegal rays and the tesserae of the 

 gular region have previously been noticed in Gyrodus. 



In estimating the taxonomic value of these and the gene- 

 rally-known characters of Pycnodonts it is necessary to 

 remember two facts : — Firstly, that Acipenser and its living 

 allies are proved by palaeontology to be the degenerate descen- 

 dants of ganoid fishes which had the cranial roof-bones 

 arranged in a normal manner (cf. Chondrosteus) ; secondly, 

 that the disposition of the opercular apparatus, as also the 

 firm fixation of the ptery go- palatine arcade, observed in the 

 Pycnodonts, is likewise met with in more than one group of 

 the bony fishes — is, in fact, merely a physiological correlation 



(1) with the reduction in size of the outlet of the gill-cavity, 



(2) with the forward displacement of the mouth, and (3) with 



* This fossil is described and figured in the recently published vol, iii. 

 of the Catal. Foss. Fishes B. M. p. 235, pi. xvi. fig. 3. 



