298 THE GEEAT UNKNOWN. 



tion of impossibility on the other. If eye-witnesses (or 

 those who present themselves as such) could decide the 

 points, they would have been decided long ago ; but those 

 who are beKeved to be best acquainted with natural laws 

 claim that theoretic impossibilities should overpower even 

 ocular demonstration. There is far more justice in this 

 claim than appears at first sight. The power of drawing 

 correct inferences from what we see, and even of knowing 

 what we do really see, and what we only imagine, is vastly 

 augmented by the rigorous training of the faculties which 

 long habits of observing certain classes of phenomena 

 induce ; and every man of science must have met with 

 numberless cases in which statements egregiously false 

 have been made to him in the most perfect good faith ; 

 his informant implicitly believing that he was simply 

 telling what he had seen with his own eyes. A person 

 the other day assured me, that he had frequently seen 

 humming-birds sucking fl^owers in England : I did not set 

 him down as a liar, because he was a person of indubitable 

 honour ; his acquaintance with natural history, however, 

 was small, and he had fallen into the very natural error 

 of mistaking a moth for a bird. 



It is quite proper that, when evidence is presented of 

 certain occurrences, the admission of which would over- 

 turn what we have come to consider as fixed laws, or 

 against which there exists a high degree of antecedent 

 improbability, — that evidence should be received with 

 great suspicion. It should be carefully sifted ; possible 

 causes of error should be suggested ; the powers of the 



