486 SCIENCE PROGRESS. 



life. He instances as the best examples of this the aquatic 

 larvae of May-flies and dragon-flies, which have gills con- 

 sisting of thin laminae abundantly supplied with tracheae. 

 He then proceeds to consider in some detail the numerous 

 larval forms of Crustacea and Echinodermata, and to draw 

 conclusions concerning their ancestral sicrnificance. 



We have then to rely for our conclusions concerning the 

 evolutionary history of a form, or the course of modification 

 it has passed through in descent, on the evidence of com- 

 parative zoology and palaeontology, and in tracing lines of 

 evolution and comparing them with the changes exhibited 

 in the development of the individual, we are necessarily led 

 to inferences concerninp^ the modes in which modifications 

 are produced. Conditions and structures exhibited in the 

 embryo which are absent in the fully developed organism 

 are in some cases, as we have seen, repetitions of a larval or 

 temporary condition of the immediate ancestor. Further 

 we know that in numerous cases such a larval stage is 

 secondary and not itself an ancestral condition. Therefore 

 neither the embryonic transient stage nor the larval stage 

 by any means necessarily exhibits a condition of any adult 

 ancestor. The general truth which we do find to hold good 

 is that well-marked larval stages are associated with condi- 

 tions of life differinof from those to which the adult is sub- 

 jected. Metamorphosis in the structure of the organism is 

 observed to correspond to a more or less sudden and 

 considerable change in the conditions of life. An organism 

 may be adapted to one set of conditions throughout its 

 independent life, or successively to different sets. 



The fact that all the terrestrial Vertebrata, the Saurop- 

 sida and Mammalia, exhibit gill slits in the embryo at a 

 certain stage is explained by the fact that they are all 

 descended from ancestors which passed through an aquatic 

 larval stage. The Amphibia at the present day pass 

 throucrh such a staee. It does not follow that the Amniota 

 are descended from ancestors which were aquatic and 

 branchiate in their adult condition. The actual history of 

 the Alpine Salamander {Salainandra atrd), shows us how, as 

 an adaptation to terrestrial conditions, the offspring may be 



