NATURAL SELECTION 61 



be considered merely as a generalisation of the effects 

 that are perceived to flow, uniformly and invariably, 

 from the action of a causal principle. It may be 

 said that apart from the mere collection of facts and 

 phenomena, the object of science is to generalise, or, 

 in other words, to establish laws or general principles 

 that may be applied to theoretical or practical uses. 

 But often the amount of knowledge attained in some 

 field of inquiry does not enable the scientist to feel 

 himself justified in designating as a law of Nature a 

 generalisation that is suggested to his mind by an 

 apparent trend of direction in the facts with which 

 he is dealing. Yet he considers himself justified in 

 making it a provisional generalisation or hypothesis, 

 with a view to its future verification or rejection. 

 But, as a scientific principle, every legitimate 

 hypothesis must be a generalisation of known facts 

 or phenomena, that is to say, it must have its foot- 

 hold in experience and observation, and, as a justifica- 

 tion for its existence, it must have a body of facts 

 apparently trending in its direction. 



A hypothesis, according to its etymology, ought to 

 be a principle of arrangement which brings a body of 

 facts under its domination. Whereas that which 

 is termed a hypothesis, but which merits no better 

 designation than a mere guess, or speculative incursion 

 into the region of the theoretical, evolved from a 

 man's inner consciousness, and launched into the 

 world unattached to any body of known facts, is 

 a concept abhorrent from the first principles of 



