1876.] AND ON THE CLASSIFICATION OF FISHES. 57 



In accordance with the view thus suggested, the humerus ia the 

 chiropterygium is the homologue of the proximal mesomere or joint 

 of the axis of the archipterygiam, while the radius and the ulna are 

 the homologues of the proximal ends of prseaxial and postaxial para- 

 meres of the archipterygium. 



The confirmation or refutation of this hypothesis is to be sought 

 in development, and in the condition of the limbs in those Palae- 

 ozoic Amphibia which may have more nearly approximated to Dipnoi 

 than any existing or extinct forms at present known. I suggest it 

 mainly in the hope of stimulating investigation in both these direc- 

 tions. 



IV. Taxonomy q/" Ceratodus, and Remarks on the Classification of 



Fishes. 



The indications afforded by the brain, the skull, and the limbs of 

 Ceratodus are sufficient to show that it occupies a curiously central 

 position among the Ichthyopsida, being allied on one side to the 

 Amphibia, on another to the Chimseroidei and Plagiostomi, and on 

 yet another to the Ganoidei — especially to that group of the 

 Ganoids which I have termed Crossopterygidce, and to the affinities 

 of which with Lepidosiren I called attention in 1861. 



But even Dipterus, which approaches Ceratodus and Lepidosiren 

 so closely in its dentition and in the form of its fins, is far more similar 

 to Polypterus and Amia in other respects ; and there is, at present, no 

 reason to believe that any of the Crossopterygian Ganoids possessed 

 other than a hyostylic skull, or differed from Polypterus in those 

 respects in which Polypterus differs from the existing Dipnoi. All 

 known Crossopterygians have jugular plates, of which there is no trace 

 in the Dipnoi. And as to the position of the anterior nares, which 

 appear to have been situated on the under face of the broad snout, 

 not only in Dipterus, but in Osteolepis and Diplopterus, I have 

 shown above that, so far from being; a diagnostic character of the 

 Dipnoi, it is simply an embryonic feature retained in them, the Se- 

 lachians, and very probably in m.iny of the early Ganoidei. On the 

 other hand, in Amia, there is an even closer approximation between 

 the Ganoids and the Teleosteans than can at present be shown to 

 exist between any Ganoids and the Dipnoi ; while the differences 

 between the Dipnoi and the Cliimseroidei, and between the Chimse- 

 roidei and the Plagiostomi respectively, are not less than those between 

 the Ganoids and the Dipnoi. 



It seems to me, therefore, that by forming the Dipnoi, Ganoidei, 

 Chimseroidei, and Plagiostomi into a group of" Palseichthyes," from 

 which the Teleostei are excluded, as Dr. Giinther proposes to do, the 

 differences between the Teleostei and the other hyostylic fishes are 

 brought into undue prominence, and that it is better to retain the 

 Miillerian groups of Dipnoi (Sirenoidei, Miiller), Ganoidei, Teleostei, 

 Plagiostomi, and Chimaeroidei {Holocephali, Miiller) as equivalent 

 and distinct natural assemblages. 



In discussing any system of classification, however, it must be 



