J. K. Tliacher — Ventral Fins of Ganoids. 239 



Before reviewing the results obtained, I wish to call attention 

 again to the shark ventrals, and to insist on the comparison between 

 them and the Chondrostean ventrals. If the reader will place the 

 figures 1, 2, 3 and 7, side by side with figui'es 67 and 69 of my paper 

 of last February, or with any of the figures of shark ventrals given 

 by Gegenbaur, there can be no question as to the way in which they 

 are related to one another, and the way in Avhich the latter were 

 formed. If we compare fig. 67 for example, with fig. 3, PI. I, there 

 can be no hesitation in recognizing the corresponding parts of each ; 

 in seeing that the solid "basale metapterygii" of 67 answers to the 

 hinder jjart of the row of proximals, for the most part separate, of 

 fig. 3, and that the part hd in the latter figure represents, more or 

 less exactly, one half the pelvic girdle in the former. The iliac part 

 or parts are not developed in the shark. Thus the most essential 

 difference between the Chondrostean and Elasmobi'anch, as regards 

 their ventral fins, would be obliterated by that process of concrescence, 

 whose absolute insignificance is shown by the varying degrees in 

 which it shows itself, here in the sturgeon ventrals and elsewhere. 

 I think I may regard the formation of the shark ventrals from a series 

 of three jointed rays in the way exhibited diagrammatically in my 

 former 2:)aper, as established fully and finally. It will moreover be 

 observed, that this method of forming the sharks' ventrals, resulting 

 as it did in a " uniserial archipterygium," is incompetent to form a 

 " biserial archipterygium," and that consequently we have in the 

 shark ventral no approach to the Ceratodus fin. 



The Relations of the Groups of Gnathostomes and the Modification 

 of the Fin-skeleton in those Groups. 



Views regarding the history of any part or organ determine to a 

 certain extent the views which are held regarding the genealogical 

 connection of tlie groups in which that part or organ is exhibited ; 

 and conversely genealogic connections already established may throw, 

 and usually do throw, a certain weight of evidence for or against a 

 view of the development of an organ. And inasmuch as from the 

 history of one part light is thrown u])on the genealogic connection of 

 the groups, and as from this genealogic connection inferences can be 

 drawn respecting the history of other parts, it is evident that in this 

 way the grounds for accepting or rejecting a view of the history of 

 one organ may become the grounds for accepting or rejecting a view 

 proposed for the history of another organ ; and be the weight of this 

 evidence great or small, there is no way to bring our belief in one 



