114 THE TRANSFORMATIONS OF THE ANIMAL WORLD 



certainties. The genealogical trees we are able 

 to draw up by relying upon morphology and on 

 chronological series are subjective to the feeling of 

 each observer. 



Another group of facts seems to allow us to es- 

 tablish by a different road the genealogical relations 

 of fossil animals. I refer to the ontogenic method, 

 that is, the study of embryonic characteristics, or, 

 more exactly, the characteristics of youth, which, 

 transitory among recent types, persist in the 

 adult state among the earliest forms of the same 

 natural groups. This method, first observed by 

 Agassiz, one of the most fervent adversaries of the 

 descent theory, has throughout been in high favour 

 amongst the philosophers of transformism, and 

 has received from Haeckel a precise formula under 

 the name of fundamental biogenetic law. " The 

 history of individual development, or ontogeny," 

 he says, " is but a short recapitulation of the long 

 palaeontological history or phytogeny." Our existing 

 embryology should then be, if this law be exact, 

 equal to the reconstruction, in at least an approxi- 

 mate manner, of the fossil precursors of each group, 

 and if these precursors were susceptible of preserva- 

 tion, they should be discovered in the terrestrial 

 strata. 



If we appeal to palaeontology, it must be recog- 

 nized that this hypothesis is by no means verified. 

 There do exist here and there a few fossil genera, 

 which have retained all their lives certain youthful 

 characteristics apparent in their living descendants ; 

 but when it comes to reconstructing whole series 

 chronologically continuous, grave contradictions are 



