22 Permanence and Evolution. 



ass. The facts of inheritance are so well known, 

 and those of reversion so well established, that 

 we are entitled, or rather compelled, to treat 

 any case of apparent non-inheritance as due to 

 reversion till the contrary be clearly shown. 



. To obtain absolute demonstration on this point 

 is probably impossible. It is only of a few 

 privileged animals that pedigrees are kept ; and 

 there is no certainty exactly how far reversion 

 may operate, but if we find that most characters 

 of animals are accounted for by direct inherit- 

 ance, that of the remainder, a large proportion, 

 can be immediately shown to be the result of 

 reversion, and that in proportion as the circum- 

 stances are more distinctly known, this more 

 and more appears to be the case, we are justified 

 in presuming that inheritance is at all events 

 at present within the sphere of our knowledge 

 absolute, and that the small residuum of in- 

 stances which we cannot resolve is due to our 

 ignorance or mistake. For it can never be 



