ORGANIC EVOLUTION 257 



The development of the salamander is but one example 

 of that correspondence between embryonic forms and 

 general plans of adult structure that is exhibited in all the 

 groups. The study of this correspondence by embryologists 

 gave rise to the "biogenetic law" (which is, rather, a rule 

 with many exceptions), that "Every animal in its develop- 

 ment tends to repeat in embryonic stages the successive 

 types of structure of animals lower in the series to which it 

 belongs." But this is only a tendency, due to common 

 origin, and a common mode of development. The corres- 

 pondence is always remote, as we have seen; not to 

 details of adult structure, but to the simplest expression of 

 the structural type, and it may be perverted by any cause, 

 internal or external, that can modify developmental stages 

 independently. It is commonly obscured: 



1) By abbreviation of the ontogenetic record. Develop- 

 mental stages, normal to the higher members of a phylum, 

 may be dropped out of ontogeny. Thus the free-swimming 

 nauplius stage, common to most of the Crustacea, does not 

 appear (as a free-swimming stage) in the development of the 

 crawfish ; it is passed over in the egg before hatching. 



2) By the coming into predominance in growth of some 

 part of late acquisition in phylogenetic history. The 

 development of the brain in the higher vertebrates is cer- 

 tainly of this precocious sort. The huge brain of the 

 embryo of a bird or a mammal can by no means be regarded 

 as primitive, although it early develops to great size in the 

 embryo. 



3) By independent specialization of some of the develop- 

 mental stages. This is of the commonest occurrence with 

 free living larvae, which may be specialized in relative 

 independence of adults. The balancers of salamander 

 larvae, present in one species of Ambystoma (A . punctatum) 

 and absent in another closely allied species ( .4 . tigrinum) 



