THE BIOLOGY OF THE MUD-DAUBING WASPS 63 



purely, we would have been sidetracked on some feature of the 

 insect's anatomy which is entirely irrelevant to its ability to 

 work out its own salvation. But if now the variation studies 

 could be taken up in the light of the preceding paragraphs, 

 they might lead us to profitable problems of the existence or 

 nonexistence of correlation between variation and elimination, 

 or in other words, to add one more word to the hypothesis of 

 selective elimination. 



