204 ETHICAL ASPECTS OF EVOLUTION 



give them. Human actions would resemble those of moths, 

 if men required experience many times repeated before 

 they arrived at the general belief that fire burns. For 

 tunately for them, they can both recollect a single experience, 

 and associate with its idea the ideas of the concomitant 

 sensations and feelings. This primitive belief, or feeling, 

 when it becomes fully conscious, is found on analysis to 

 be resolvable into two essentially distinct propositions 

 one, that when on a future occasion fire is touched, a burn 

 will ensue ; the other, that it is the fire that causes the burn. 

 The first is in the nature of a prediction ; the second, of 

 an explanation. The first is, as we shall see, the law for 

 all objective ; the second for all subjective series. 



No doubt, the frequent repetition of the same sequence 

 tends to strengthen our belief that it is invariable ; that 

 is the result of summation of stimuli. But the more common 

 and by far the more important function of experience is 

 not to confirm, but to contradict. It is experience that 

 shows us that our crude belief that like follows like is as 

 often wrong as right, and compels us to examine the ground 

 on which we affirm its applicability to any single instance. 

 A man opens an egg, and finds it pleasant to scent and taste ; 

 he then opens another, and expects it to have the same 

 properties, but is disappointed. His belief receives a shock ; 

 but chance shows him that some eggs float, whereas others 

 sink ; and this discovery, if used as the clue to experiment, 

 leads to the further discovery that all eggs that float, if not 

 equally well-flavoured, are at any rate eatable. He thus 

 finds that his mistake lay, not in his general belief that like 

 follows like, but in the supposition that all eggs are alike. 

 Further experience, supplemented by the artificial experi 

 ence which we call experiment, enables him at last to frame 

 a general definition of likeness. In order that the result 

 he expects may follow from a given fact, all the conditions 

 of that fact, except its position in time and space, must be 

 the same as those of the instance on which his expectation 

 is based. Moreover, in order to state his law, he must 

 define the result with similar exactness. It is not until he 



