XXIX 



of the radii direct with the scapular arch) the expressions of matters of 

 fact, and partly the interpretation of homologies. 



It is assumed (1) that the external basal element of the limb in Chon- 

 drosteans is equivalent to the median element (when differentiated) of the 

 plagiostome Elasmobranchiates, and is, therefore, the mesopterygium, and 

 (2) that the propterygium is not developed. 



It is not evident, however, why the external element should not be 

 homologous, in part at least;, with the propterygium of the Elasmobran 

 chiates. The latter affords a better basis for identification, and it would 

 seem more justifiable if it must needs be identified with a single element 

 to refer it to the propterygium rather than to the mesopterygium. The 

 mesopterygium may (1) either be represented, in the Chondrosteans, by an 

 independent element (&quot;r&quot; in Gegenbaur s Untersuchungen), crowded out 

 of place by the intervention of the rays (as in certain Raise), or (2) it may 

 be entirely suppressed through atrophy, or (3) it may be fused with the 

 propterygium (as in the Heterodontidae and Scymnidse). 1 In the first case, 

 the expressed differences of the Crossopterygians would be confined to the 

 exclusion of the actinosteal element from direct articulation with the scapu 

 lar arch. But in the most teleosteoid of the Ganoids (Amia), we find even 

 that condition approximated, only one (of the seven actinosts) being artic 

 ulated directly with the arch, the rest being connected with the metaptery- 

 gium. 



But even supposing that the mesopterygium is an element entirely 

 wanting in the Hyoganoids and Chondrosteans, two elements (metaptery- 

 gium and propterygium) are developed in those forms in common with the 

 Crossopterygians, and which are wanting in the Teleosts. It is not evi 

 dent why the development of a mesopterygium should be of importance 

 so much superior to that of the other two elements, or why the mere fact 

 that the articulation of the actinosts with the scapular arch should be of 

 such paramount significance as to justify the combination of all forms 

 agreeing therein (including the Chondrosteans and all Teleost fishes), 

 and the separation therefrom, as co-ordinate terms, of forms not agreeing 

 therein. 



But it is true that the evidence appears to be somewhat contradictory 

 as to the relations of the forms distinguished by the structure of the pec 

 toral limbs as well as the scapular arch. On the one hand, the Chondros- 

 teaus (rather than Crossopterygians) agree with the Hyoganoids in the con 

 struction of the paraglenal element as well as the pectoral member ; on the 

 other hand, the Crossopterygians appear to agree more with the Elasmo- 



1 Most naturalists would probably prefer either of these interpretations to the ho- 

 mological representation in Choudrostean, by a mesopterygium disintegrated and 

 represented by apparent rays. 



