188 M. Agassiz on the Classification of Fishes. 



against tlie objections which some are desirous to draw from 

 the imperfect state of our knowledge respecting the entire se- 

 ries of fossils embedded in the strata of the whole globe. But 

 it is evident, that the knowledge we have already acquired in 

 this respect, ought to have an influence on our classifications ; 

 and that authors will thus come always to take more into ac- 

 count the order of the succession of types in their systematic 

 arrangements. 



I have already had more than one occasion to draw atten- 

 tion to the striking analogy which exists between certain em- 

 bryonic forms, which are transitory in the development of in- 

 dividuals, and the constant characters of numerous genera be- 

 longing to different families, which have but few representa- 

 tives in the existing creation, or have become wholly extinct. 

 It cannot be doubted, therefore, that these considerations ought 

 to exercise, in their turn, an influence on the position to be 

 assigned, in the system, to these same genera. When resum- 

 ing my researches on the conformation of the skeleton of fishes, 

 I have shewn, at different times, how far the results of embry- 

 ology agree with those of paleontology. I have thence become 

 convinced that embryological researches, prosecuted with the 

 view of appreciating the value of organic forms, as zoological 

 characters, ought, in like manner, one day to exercise a gi'eat 

 influence on our methods. It will, no doubt, be the same with 

 microscopic investigations, which are now pursued with so 

 much ardour in every branch of natural science. 



Do the relations between organized beings, thus varied as 

 they are, admit of being expressed by linear series I I think 

 not. I am more inclined to believe that naturalists will revert 

 to the idea of well-defined divisions, placed after each other, to 

 admit, as the expression of the varied relations of organized 

 beings, of graphic pictures, in the centre of which the best 

 known types will be placed, and around which will be ranged, 

 according to their greater or less affinity, other types, which 

 may become, in their turn, the centre around which other se- 

 condary types will gravitate. And the better we become ac- 

 quainted with the entire details of one great division, the better 

 will we group all its members, according to their diverse affi- 

 nities. If we are considering the Echinodermata, for example, 



