6o THE GRAMMAR OF SCIENCE 



2. We may infer what we cannot verify by direct 

 sense-impression only when the inference is from known 

 things to unknown things of the Hke nature in similar 

 surroundings. 



Thus we may not infer an " infinite " consciousness 

 outside the physical surroundings of finite consciousness ; 

 we may not infer a man in the moon, however like in 

 nature to ourselves, because the physical surroundings in 

 the moon are not such as we find man in here, etc., etc. 



3. We may infer the truth of tradition when its con- 

 tents are of like character and continuous with men's 

 present experience, and when there is reasonable ground 

 for supposing its source to He in persons knowing the 

 facts and reporting what they knew. 



The tradition that Wellington and Blucher won the 

 battle of Waterloo fulfils the necessary conditions, while 

 the miracle of Karl the Great and the adder fulfils 

 neither condition. 



4. While it is reasonable in the minor actions of life, 

 where rapidity of decision is important, to infer on slight 

 evidence and believe on small balances of probability, 

 it is opposed to the true interests of society to take as 

 a permanent standard of conduct a belief based on in- 

 adequate testimony. 



This canon suggests that the acceptance, as habitual 

 guides to conduct, of beliefs based on insufficient evidence, 

 must lead to the want of a proper sense of the individual's 

 responsibility for the important decisions of life. I have 

 no right to believe at seven o'clock that a cab will be on 

 the stand at eight o'clock, if my catching the train at 

 half-past is of vital importance to others. 



^ 1 1 . — TJie External Universe 



Before we draw from our present discussion any con- 

 clusions as to the facts of science we must return once 

 more to the immediate sense-impression and examine its 

 nature a little more closely. We are accustomed to talk 

 of the " external world," of the " reality" outside us. We 



