XXX 



branchiates, and less with the Hyoganoids, in these 'respects. But the 

 Crossopterygians agree with the latter much better in the composition of 

 the skull and squamation, and the question, therefore arises whether it is 

 more probable that the Crossopterygians should have attained that special- 

 ized similarity to the Hyoganoids from an independent origin, or whether 

 they should have departed (after having received such characters from a 

 common progenitor) in the structure of the scapular arch and the pectoral 

 member, and whether the apparent greater similarity in those respects to 

 the Elasmobranchiates is not rather adaptive, or the result of simplicity of 

 structure of the paraglenal. Possibly, the following hypothesis may ap- 

 proximate the truth, and account for the divergencies of the several types. 

 The Acanthodeans of the devonian and following epochs may be the 

 nearest of kin known to the representatives of the direct line of descent 

 from the typical Elasmobranchiates ; the development of two marginal 

 (external and internal) spines in the pectoral limb may lend significance 

 to the specialized condition of the metapterygium and propterygium in 

 the pectoral limbs of the succeeding forms, as may also the character of the 

 scales for those of the typical " Ganoid" type. 



CHONDROGANOIDS. 



The Chondrosteans furnish the most satisfactory evidence of closest re- 

 lationship with the ancestral stock iu the histological condition of the ske- 

 leton, the generalized and little concentrated brachial and hyoid apparatus, 

 and the structure of the fins. At the same time they are considerably 

 removed from the direct line of descent. 1 



BBACHIOGAXOID AND DIPNOAN OFFSHOOTS. 



From the ancestral stock, somewhat more specialized than that from 

 which the Chondrosteans originated, but with approximately the same 

 pattern of pectoral limb, forms may have been developed in which the me- 

 tapterygium and propterygium (converging towards the base in the Chon- 

 drosteans — and Acanthodeans ?) finally approximated and grew together ; 

 the intervening cartilage (mesopterygium) became ejected and projected 

 backwards, bearing the specialized actinosts on a convex periphery. 



(1) From such an ancestor a long line may have descended which finally 

 culminated in the specialized Crossopterygians now known. 



(2) From an equally ancient stock, and deviating less in histological 

 characters, the Dipuoans may have descended : in such forms, the metaptery- 

 gium and propterygium, instead of diverging backwards, may have con- 



1 There are lome reasons for thinking that the Selachostomi are the most general 

 ixed group of G*aoids. 



