40 THE METHOD OF SCIENCE 



we have a choice between two hypotheses, two inferences, two 

 statements of what may be the relations between the phenomena. 

 First, we may infer that parental ill-health tends to injure the 

 germ-plasm, and, therefore, that the children are innately puny, 

 innately incapable of becoming robust. If this hypothesis is true, 

 then it follows as a necessary consequence that, if the children con- 

 tinue to dwell in the unhealthy environment, the weakened germ- 

 plasm will be still more injured and the grandchildren will be still 

 more puny, and therefore that the race will degenerate step by 

 step. At any rate, if it can be shown that races become innately 

 degenerate under such conditions, the hypothesis may be regarded 

 as a theory the truth of which has been established. Here, 



(a) The facts are that the children of unhealthy parents are very 

 frequently puny. 



(fr) The hypothesis, the induction, is that parental ill-health tends 

 to injure the germ-plasm, and so render the children and their 

 descendants innately puny. 



(c) The deduced consequences are that families, classes, and races, 

 exposed to unhealthy conditions, tend to become degenerate. 



(d) The appeal to reality is the ascertaining whether populations 

 exposed to unhealthy conditions actually do become degenerate. 



By this means we test our thinking, our hypothesis. Second, we 

 may infer that the puniness of the children is due, not to the 

 deterioration of the germ-plasm, but merely to inferior development 

 resulting from the unhealthy conditions to which the children 

 themselves are exposed. That is, we suppose that the puniness is 

 not innate. If this second hypothesis is true, the race will not 

 grow degenerate through being exposed to unhealthy conditions. 

 On the contrary it will become more resistant to those conditions 

 through the weeding out of the naturally less resistant individuals. 

 Having drawn these two contradictory inferences, we have, as I 

 say, to ascertain which of them is true by appealing to reality in 

 the world around us. 1 



general condition, then, may be considered as involving three subordinate 

 conditions : 



(1) That the hypothesis be self -consistent, and in harmony with all the other 

 laws included in the conceived system of reality. 



(2) That it furnish a basis for rigorous deductive inference of consequences. 



(3) That these inferred consequences be in agreement with reality. 



Of these conditions the first two are applicable to the formation of every 

 hypothesis, no matter how provisional a character it may have ; the third is a 

 condition of the acceptance of a hypothesis as true." (Welton, Manual of 'Logic , 

 voi. ii. pp. 95-6.) 



1 See 733. 



