76 FlN-RAYS IN CLASSIFICATION OF FISHES. 



ences aroso. na])[)ily such taxoiiomic inetliods are becomiug a matter 

 of the past, and it will not be long before it will not bo the misfortune 

 of tlie concholoiiist to name the "si)at'' stages of the oyster as distinct 

 si)ecies,or ibr the iciithyologist to erect ''families" upon the characters 

 presented by larval fishes. 



The truth is, that if embi-yological, internal or external, anatomical 

 charai'ters are each taken separately they will lead to diverse results, 

 and just in proportion to the superliciality of the characters upon which 

 names are based in just that proportion will there be uncertainty as to 

 the relations of the discriminated forms in the minds of subsequent in- 

 vestigators who may have other forms to compare or better s[)ecimeiis 

 to study. So it will not be by means of embryology through a study 

 of a single character or group of characters or with the help of external 

 or ot internal anatomy alone that we can be guided, but by all three 

 combined, with such help from distribution in space as may be accessi- 

 ble in the case of living organisms. The method of the palteontologisr 

 is necessarily ditierent, but even that does not aflbrd an apology for the 

 treatment of now-living forms according to a widely prevalent but fun- 

 damentally wrong method. 



To cite an instance in illustration, most persons would suppose that 

 there was no reasi)u to sus[tect that the ventral tins of a Sea robin or a 

 Toadtish had not always been inserted in advauceof theijectoral. There 

 have been theorists who have thought diiierently, Owen amongst the 

 nund)er, but it was not until the embryological observations of A. 

 Agassiz i)ut us in possession of the data from which to formulate it as 

 a fact that in some instances at least the ventral fins of the embryos of 

 PhysQclist fishes originate behind the pectorals.and are suddenly shoved 

 forward, below and in advance of the pectoral. This information gained, 

 we are in a position to state with i)ositive certainty that the relative 

 position of the paired fins of Physoclist fishes was preceded in time by 

 one in which they were urnre nearly in the same relative position as in 

 the existing physostonu)ns forms. Xo i)Ossible construction of the facts 

 of the anatomy of the adults could have given such conclusive evidence 

 in favor of what becomes an obvious truth in the light of ontogenetic 

 investigation. Moreover, the facts of the anatomy of Physoclists be- 

 come at once of greater interest, for the crossing of the nerves which 

 sui)ply the paired fins is satisfactorily exi)lained. 



To return, however, to the discussion of the fact that parts of struct- 

 ures vanish in an order exactly the reverse of that in which they ap- 

 peared, we may recur to the Salmon, in which the preanal fin-fold atro- 

 phies or disappears in a manner exactly the reverse of that presented by 

 its appearance. During its outgrowth it slowly widens or becomes 

 higher, while during its atrophy it becomes gradually lower and nar- 

 rower, until all outward evidence of its existence vanishes. 



Where new comi)lications of develoi)ment occur, when new structural 

 details are added to pre-existing ones, as in the case of the fins, it is very 

 dear that this process is often actually accompanied by one of atrophy, 



