i8 9 6. SOME NEW BOOKS. 4 11 



their geological distribution. They arise in the Upper Permian and 

 attain their maximum development in the Trias, Lias, Jurassic, and 

 Lower Cretaceous. The osteology of most of the genera is treated 

 with great completeness, and the skull and skeleton of Lepidotus are 

 specially described in a remarkable way. A restoration of the 

 complete fish and several drawings of the head, leave very little more 

 to be learned concerning this grand and widely-distributed fish. The 

 genus Dapedius is treated in a similar manner, new and important 

 facts being added concerning the skeleton of the trunk. The com- 

 paratively small family of Macrosemiidae includes only seven genera 

 with slender trunk, enamelled scales, and pointed teeth, especially 

 distributed through the Jurassic rocks (Ophiopsis,Histionotus, Legnonotus, 

 Macvosemius, Petalopteryx, Propterus, and Notagogus). The systematic 

 position of the interesting Pycnodontidae, ranging from the Jurassic 

 to the Eocene, has hitherto been misunderstood by almost all authors. 

 From their singularly deep and laterally compressed trunk, their 

 peculiar scales, and their dentition, these notochordal fishes have 

 mostly been regarded as a separate order equivalent to that of the 

 Lepidostei. On account of a superficial resemblance they have often 

 been considered as descendants of the Platysomidae. The careful re- 

 searches of Woodward leave no longer any doubt that the Pycnodon- 

 tidae belong to the typical Protospondyli, and that they bear almost 

 the same relation to the Semionotidae and Macrosemiidae as the Platy- 

 somidae to the Palaeoniscidae. The two families form analogously 

 differentiated branches of the Heterocerci and Protospondyli respec- 

 tively, which were apparently not capable of further development and 

 left no descendants. Traquair had already emphatically denied all 

 genetic connection between the Platysomidae and the Pycnodontidae. 

 A. S. Woodward determines the systematic position of the latter with 

 absolute certainty, and the view of Egerton, which I had likewise 

 adopted, is finally abandoned. The twelve genera of Pycnodontidae, 

 like those of the other families, are very carefully revised, a number 

 of superfluous genera and species are relegated to the synonymy, and 

 the whole of the widely scattered material is arranged in a masterly 

 manner. Several genera are illustrated by finely executed figures. 



Under the denomination Eugnathidae, Woodward includes a 

 number of Mesozoic fishes with imperfectly ossified vertebral column, 

 a deeply cleft mouth, pointed teeth, and slender trunk, which begin 

 in the Trias and become extinct in the Chalk. The ganoid scales are 

 rhombic, sometimes rounded at the hinder angles, and of very variable 

 thickness. The genera belonging here (EugnatJius, Hetevolepidotus, 

 Allolepidotus, PtycJwlepis, Caturus, Calloptcrus, Eurycormus, NeorJiombolepis, 

 LopJiiostomus) have hitherto been separated in different families, and 

 their osteology has been insufficiently known. Some of the species 

 until now referred to Pholidophorns, are transferred to EugnatJius or 

 Hetevolepidotus, and the great similarity between the skull of Caturus 

 and that of EugnatJius is pointed out in a convincing manner. 

 Woodward assigns small importance to the characters of the scales ; 

 among the Eugnathidae we therefore find genera with thick polished 

 rhombic ganoid scales and others with thin, cycloidal, deeply 

 imbricating, enamelled scales. As among the Semionotidae and 

 Macrosemiidae, the latest genera are characterised by the most 

 advanced ossification of the vertebral column, which, as in Amia, 

 results in the formation of complete centra. 



Between the Eugnathidae and Amiidae the distinction is slight ; 

 the vertebral column in the latter is generally better ossified, and the 

 dermal skeleton always consists of thin cycloidal scales. The two 



2 G 2 



