UNE. 1896. SOME NEW BOOKS. 409 



closely related that hitherto no agreement has been arrived at 

 concerning the definition of the larger sub-divisions. In the first, and 

 in part of the second, volume of his Catalogue, A. S. Woodward has 

 treated the fossil Elasmobranchii ; in half of the second volume he has 

 dealt with the completely extinct Ostracodermi, the Dipnoi, the 

 Crossopterygii, and part of the Chondrostei. The classes, sub-classes, 

 orders, families, genera, and species, are well defined ; all the fossil 

 forms, including those not represented in the British Museum, are 

 enumerated ; the synonymy is critically determined and the literature 

 completely summarised. The third volume contains the remainder of 

 the Chondrostei, all the other ganoid fishes, and part of the Mesozoic 

 bony fishes. In a further volume the later Teleostei are to follow. 

 The most interesting groups from the phylogenetic and morphological 

 points of view are included in the three volumes now published, so 

 that the significance of A. S. Woodward's work can already be 

 completely appreciated. Since the author follows no theoretical 

 course, and contemplates no systematic reform according to his own 

 ideas, but attempts more especially to present the facts in well- 

 arranged and general order according to the latest phase of our 

 knowledge, his Catalogue will be of permanent value in the literature 

 of palasichthyology, and will serve as an inexhaustible source of 

 information, not only for palaeontologists and geologists, but also for 

 zoologists and embryologists. 



A space of four years intervenes between the appearance of the 

 second and third volumes — a short period of time for the arrangement 

 of the varied and difficult materials which are treated in Part III. 

 For the Chondrostei, it is true there are available the important 

 earlier works of R. H. Traquair ; among the Protospondyli, Aethfeo- 

 spondyli, and Isospondyli, Woodward in many cases has had to 

 determine the fundamental method of classification, and to submit to 

 careful critical revision the families, genera, and species represented. 

 Traquair's remarkable researches among the Palaeozoic heterocercal 

 scaly ganoids have led to the unexpected result that there exists so 

 great an anatomical resemblance between these fishes covered with 

 typical ganoid scales and the Acipenseroids, that no doubt can exist as 

 to the genetic connection between the two. A. S. Woodward adopts 

 Traquair's views completely, and unites under the heading Chondrostei 

 both the heterocercal-tailed Palaeoniscidae, Platysomidae, and Catop- 

 teridae, and also the recent sturgeons and paddle-fishes (Polyodontidae), 

 besides an extinct Mesozoic family (Belonorhynchidae) with a diphy- 

 cercal tail. As common characters are enumerated the incomplete 

 ossification of the vertebral column, the smaller number of supports 

 than rays in the dorsal and anal fins, and the more distinct segmenta- 

 tion of the pectoral girdle by the formation of an infra-clavicle and 

 clavicle ; or, as Gegenbaur has lately correctly named these bones, a 

 clavicula and a cleithrum. In addition to these principal characters 

 there are still others of less importance, namely, the presence of 

 baseosts in the pelvic fins, the small extension of the preoperculum on 

 the cheek, and the sub-division of the hinder expansion of the 

 maxilla. 



With the exception of the Belonorhynchidae all the Chondrostei 

 have a beautifully heterocercal tail. In the dermal skeleton there is 

 the greatest possible variation. In the Palaeoniscidae, Platysomidae, 

 and Catopteridae, the body and tail are completely covered with 

 rhombic, very rarely cycloidal scales ; only the aberrant Permian 

 Dovyptenis appears to have been scaleless. In the Belonorhynchidae 

 true scales are wanting, but four rows of small bony scutes extend 



2 G 



