RARITY OF PERFECT FREEDOM FROM ERROR. 141 



For additional illustration of the errors of scientific 

 research and experiment, I must refer the reader to Jevons's 

 6 Principles of Science,' Vol. I. Chap XV., and to the 

 various works on logic. I do not profess, in this limited 

 treatise, to speak of the special errors peculiar to par- 

 ticular sciences. 



CHAPTER X. 



ON THE CERTAINTY OF SCIENTIFIC KNOWLEDGE. 



ALL truth must be equally certain in itself, because 

 certainty is one of its characteristics ; that which is not 

 intrinsically certain is nob true. But all truths do not 

 appear equally certain to us, because we cannot perceive 

 them with equal clearness and force. To an infinite mind, 

 all things are equally simple, but the human mind is 

 extremely finite, and to us the system of nature is one of 

 infinite complexity. The certainty of truth is a great 

 moral quality ; and however true, great, beautiful, or ad- 

 vantageous to us, a sta+ement may be, if we cannot prove 

 its truth with certainty, our confidence in it is blind and 

 very defective. 



Our certainty of the truth of a scientific statement 

 depends largely upon the nature of the case ; negative 

 and universal conclusions are often exceedingly uncertain, 

 and it is frequently impossible to prove them, because no 

 man can exhaust the universe of truth. They are only to 

 be trusted when we know that if the fact existed it would 

 have been certain to be noticed. According to Jevons, 1 

 ' the results of geometrical reasoning are absolutely cer- 



1 Principles of Science, i. 268. 



