138 NATURAL SELECTION AND 



general term, but because we are looking for 

 concrete particular explanations, that we insist 

 on the reality of " survivals " in institutions. 

 The fact that a custom occasionally outlives the 

 conditions which originally favoured its growth, 

 needs no explanation. The tendency of human 

 beings is to go on doing what they have been 

 accustomed to do, unless there is a very strong 

 reason for giving it up ; and frequently even 

 then. Natural selection does not eliminate 

 disadvantageous customs in coherent human 

 societies as rapidly as it eliminates disadvanta- 

 geous characteristics among the lower animals. 

 The disappearance of the circumstances, which 

 produced any particular custom originally, make 

 it easy, of course, fc r the custom to die out ; 

 but, as a rule, some positive and considerable 

 inconvenience is necessary to rouse people 

 sufficiently to make them shake off any old 

 habit. Occasionally something purely "acci- 

 dental " (" accidental," of course, only in the 

 same sense in which we speak of " spon- 

 taneous " variations) is sufficient to put an end 

 to an old custom : thus the death of the holder 

 of some antiquated office may give the occasion 

 for discontinuing it. If an old custom dies out 



