﻿XX11 INTRODUCTION. 



paired fins are truly archipterygial, whether elongate or abbreviate ; 

 while in the existing Polypteridae the pectoral fins have lost all 

 trace of the original branched arrangement of the cartilages (pre- 

 cisely like the Sharks), and in Polypterus itself the pelvic fins are 

 approximately in the same condition as those of one of the Actino- 

 pterygian Chondrostei. Among the early families, the characters 

 of the median fins lead to the recognition of two or three divisions. 

 It is probable that one type in which the median fin remains 

 undivided and more or less in its primitive condition will eventually 

 be met with, even if it be not already known. This group has 

 received the name of Haplistia, and we provisionally assign to it 

 the problematical Tarrasiidae. The second and third types, though 

 now clearly definable, are not satisfactorily formulated in the some- 

 what fluctuating classifications of Cope ; and the terms Rhipidistia 

 and Actinistia are selected on the present occasion from those 

 already proposed by that author, as being most expressive and 

 accurate. For their diagnosis and description, reference may be 

 made to the Catalogue itself; and it only seems necessary here to 

 remark upon the extraordinary manner in which the specialized 

 dorsal fins of the Bhipidistia resemble paired limbs (see especially 

 fig. 50, p. 335). When subdivided, the dorsal fin invariably 

 degenerates to two portions, and these are supported on a plan that 

 does not differ much from that of an abbreviate archipterygium. 



The great group of Actinopterygian Teleostomi is that con- 

 cerning which palaeontology affords most extensive information ; 

 but as only the typically Palaeozoic families of Palaeoniscidae and 

 Platysomatidae are comprised in the present volume, it will be con- 

 venient to defer general observations on their relationships until the 

 completion of Part III. 



In conclusion, there is little to add concerning the details of the 

 plan of the Catalogue to the statement already made in the Intro- 

 duction to Part I. Family names derived from generic names 

 terminating in -aspis and -lepis occur now for the first time ; and, 

 from the point of view of euphony, it has been deemed advisable 

 to omit the reduplication of " id" which would be necessitated by a 

 strict adherence to classical rule. There is already justification for 

 this procedure in the universal adoption of the term Crossopterygii 

 instead of the strictly accurate Crossotopterygii. More descriptive 

 sections have been included than in the former volume, on account 

 of the importance of the Palaeozoic types to the modern Biologist, 



