504 MR. A. SMITH WOODWARD OS THE [Juiie 20, 



tion of Dapedtiis already published by Dr. Ti-caquair ' suffices for 

 the purpose ; but the cranium itself is showu only iu one specimen 

 from Lyine Regis (Brit. Mus. no. P. 3541), which seems worthy 

 of description as an appendix to the foregoing account of Lqjidotus. 

 As seen in side view (Plate L. fig. 3), the basicranial axis of 

 Dapedius is sharply bent upwards iu front of the otic region — an 

 arrangement perhaps correlated with the deepened form of the 

 fish. As in Lepidotus, the cranial cartilage is well ossified, and 

 there seems to have been a complete, or nearly complete, interorbital 

 septum. The hasloccipltal element (h.occ.) is deep, much exca- 

 vated behind for the notochord («./.), and longitudinally grooved 

 below for the basicranial canal (hex.). The divisions between the 

 elements of the cranium and the situation of the foramina for the 

 nerves are unfortunately not distinguishable ; but it is clear that 

 there was an ossified supraoccipital (s.occ), with a vertical median 

 ridge behind, and there are robust ossifications in the lyrefroaUd and 

 postfrontal regions. The ethmoidal region (e.) terminates in front 

 in a small blunt process, pierced transversely by a large foramen ; 

 and it expands on each side, in advance of and below the pre- 

 frontal, into a great mass that would be sheathed by the vomer. 

 The olfactory nerve evidently passed through a foramen (/.) 

 between the prefrontal and ethmoid, there beiiig no elongation of 

 this foramen into a canal. Of the membrane-bones of the cranial 

 roof, thejMrietals, frontcds, and squamosals are fused into a continuous 

 plate (jrfs.); while, as in Lepidotus, a narrow rim of the cranium 

 projects behind the covering thus formed. Viewed from behind 

 (Plate L. fig. 3 a, m.), there is seen to be a small cavity on each 

 side between the cranium and the squamosal portion of the roof, 

 this being evidently the reduced temporal fossa. 



III. Conclusion. 



From the observations now recorded it would be premature to 

 make any very general deductions, the characters of the skull 

 having yet to be discovered in the majority of the Mesozoic fishes. 

 The new facts, however, are interesting as tending to confirm a 

 conclusion that must have impressed everyone who has deeply 

 studied these extinct fishes, namely, that it is impossible in Jurassic 

 and early Cretaceous formations to recognize any absolute sub- 

 division of the so-called Ganoids into " Lepidosteoidei " and 

 " Amioidei." The skulls of Lepidosteiis and Dapedius differ from 

 those of existing " ganoids " in exhibiting the backward extension 

 of the basicranial canal ; and the cartilaginous cranium of Dapedius 

 is remarkably similar in every respect to that of the modern salmon 

 (Sahno), except somewhat more ossified. Both Lepidotus and 

 Dapedius agree with Lepidosteus and Amia in the fact that the 

 membrane-bones of the roof do not extend quite to the occipital 



^ R. H. Traquair, " On the Structure aud Affinities of the Platysomida;," 

 Trans. Eoy. Soc. Edinb. vol. x\ix. (1879) pi vi. fig. 13. 



