8 



SIRENOID AND CROSSOPTERYGIAN GANOIDS. 



bulbus arteriosus. Ceratodus now comes in to bridge the interval. It has the functional 

 lung of Lepidosiren , the transverse arterial valves of Polypterus. 



The relations of Dipnoi and Ganoidei consequently require fresh adjustment. It 

 seems at present most satisfactory to treat the Dipnoi as a part of the Ganoid order, 

 whose definition may be extended to include fishes having longitudinal valves in the 

 bulbus arteriosus. 



Dr. Giinther has proposed 1 a further step, viz. to unite the reconstituted order Ganoidei 

 with the Elasmobranchii, so as to form a new subclass — Palceichthyes. Solid reasons 

 are adduced for this step, but its discussion is not needful in this place. The Ganoidei 

 and Dipnoi may, for reasons subsequently given, be united, and their place is plainly 

 next the Elasmobranchii. That the relations of the three groups are close will not be 

 disputed by any zoologist. 



The following list of recent Ganoids is taken, with trifling alterations, from Dr. Gunther's- 

 'Catalogue of Eishes,' vol. viii. 



Protopterus, Ow. 



. 1 



species 



Lepidosiren, Fitz. 



. 1 



j> 



Ceratodus, Ag. 



. 2 



5) 



Polypterus, Geof. St. Hil. . 



. 1 



;> 



CalamoicJdJiys, J. A. Smith 



. 1 



5> 



Lepidosteus, Lacep. 



. 3 



J) 



Acipenser, Art. 



. 19 



> J 



Scaphirhytichus, Heck. 



. 1 



J) 



Polyodon, Lacep. . 



. 2 



?> 



Ami a, L. 



. 1 



) ) 



We rank, then, among the Ganoidei the recent genera here enumerated, and such 

 fossil forms as are more nearly connected with them than with other recent fishes. What 

 these forms are it will be our task in the following pages to ascertain. 



It is plainly necessary to follow some uniform nomenclature in describing the bones 

 of the skull, which are usually the most instructive fragments of fossil Ganoids. It will 

 also be conceded that the best available principle of classification of the elements, 

 whether of the skull or of any other part of the animal framework, is that which rests 

 upon embryology. The method of identification by relative position is too uncertain, 

 and wholly fails us in very aberrant skulls, and sometimes even in the comparison 

 of typical skulls. A nomenclature based upon the vertebrate theory of the skull is 

 elegant and philosophical in form, but the serial homologies which it presumes have 

 no demonstrated basis of fact. To found upon the development of the skull means, in 

 the present state of our knowledge, to found upon Prof. Parker's memoirs. These 

 invaluable writings, in which the views suggested by Prof. Huxley's ' Croonian Lecture' 



1 ' Phil. Trans.,' vol. clxi, pt. ii, p. 554 (1872). 



