92 Permanence and Evolution. 



far, of the characters of any living being are 

 directly inherited from one or other parent* 

 When this is not so, they can usually be traced 

 to a grandparent or some near ancestor ; when 

 this cannot be done, the matter can often be 

 explained by more distant reversion ; but if 

 experiments were tried with the object of ascer- 

 taining whether there was any residuum, and if 

 it was ascertained that there are sometimes new 

 characters which arise without inheritance and 

 without any reasonable chance of reversion, then 

 we should define what these characters are and 

 under what conditions they arise ; and when we 

 had done this, we should know what we are 

 talking about when we speak of variation and 

 evolution. 



I recommend such experiments to all na- 

 turalists, specially evolutionists, if they are 

 anxious to found their hypothesis on facts, not 



* So even Darwin, "Animals and Plants," vol. ii. p. 57. 

 " Hence we are led to look at inheritance as the rule, and non- 

 inheritance as the anomaly." 



