Permanence and Evolution. 91 



to me, and which seems to me very probable, 

 though not yet proved, namely, that except in 

 points in a large sense hygienic there is no 

 variation whatever ; that is to say, the children 

 of A and B are the joint result of the characters 

 first, of their parents in varying degrees (which 

 alone would give rise to infinite shades of 

 difference between human beings) ; secondly, 

 often of some one or more of their four grand- 

 parents, C, D, E, F, also in infinitely varying 

 degrees and proportions ; thirdly, sometimes of 

 remoter reversions, the same rule applying to 

 physical and mental characters, the origin of 

 primary types being unknown. 



But however this may be and I do not 

 consider this as anything but a conjecture 

 I feel sure that it is important to begin at 

 this end ; instead of starting from variability 

 and asking how far does inheritance extend, 

 we" ought to start from permanence and 

 ask how far does variation extend. Most, by 



