Chap. XT. EVOLUTION. 115 



period will generally be advantageous, and their re- 

 tarded appearance disadvantageous ; the former, when 

 it appears accidentally, will be preserved by natural 

 selection. It is the same with every change which 

 gives to the larval stages, rendered multifarious by 

 crossed and oblique characters, a more straightforward 

 direction, simplifies and abridges the process of deve- 

 lopment, and forces it back to an earlier period of life, 

 and finally into the life of the egg.^ 



As this conversion of a development passing through 

 different young states into a more direct one, is not the 

 consequence of a mysterious inherent impulse, but de- 

 pendent upon advances accidentally presenting them- 

 selves, it may take place in the most nearly allied 

 animals in the most various ways, and require very 

 different periods of time for its completion. There is 

 one thing, however, that must not be overlooked here. 

 The historical development of a species can hardly 

 ever have taken place in a continuously uniform flow ; 

 periods of rest will have alternated with periods of 

 rapid progress. But forms, which in periods of rapid 

 progress were severed from others after a short dura- 

 tion, must have impressed themselves less deeply upon 

 the developmental history of their descendants, than 

 those which repeated themselves unchanged, through a 

 long series of successive generations -in periods of rest. 

 These more fixed forms, less inclined to variation, will 

 present a more tenacious resistance in the transition 

 to direct development, and will maintain themselves 

 in a more uniform manner and to the last, however 



I 2 



