ii LIMITS ON VARIATION 109 



new forms are always as sharply defined 

 as the old, differing from them by char- 

 acters which are as well marked and as 

 constant as all their predecessors in the 

 wonderful processions of organic life. It 

 helps us very little to remember that in 

 the existing world some varieties do 

 occur in certain species varieties which 

 are sometimes sufficiently well marked 

 to raise the question among classifiers 

 whether they are, or are not, sufficiently 

 constant to deserve the name of separate 

 species. But this does not help us 

 much, because such varieties are very 

 limited in extent, and are almost always 

 confined to such superficial features as 

 the colour of hair or of feathers. They 

 never, so far as I know, affect organic 

 structure, and no accumulation of them 

 would account for the very different 

 kinds of variation which are conspicuous 

 in the successions of organic life. 



