106 HOViTM OLGAffUM. [BOOK L 



and enslave the world to human thoughts, and thoughts to 

 words. But demonstrations are in some manner themselves 

 systems of philosophy and science j for such as they are, and 

 accordingly as they are regularly or improperly established, such 

 will be the resulting systems of philosophy and contemplation. 

 But those which we employ in the whole process leading from 

 the senses and things to axioms and conclusions, are fallacious 

 and incompetent. This process is fourfold, and the errors are in 

 equal number. In the first place the impressions of the senses 

 are erroneous, for they fail and deceive us. We must supply 

 defects by substitutions, and fallacies by their correction. Se 

 condly, notions are improperly abstracted from the senses, and 

 indeterminate and confused when they ought to be the reverse. 

 Thirdly, the induction that is employed is improper, for it deter 

 mines the principles of sciences by simple enumeration,* without 

 adopting exclusions and resolutions, or just separations of nature. 

 Lastly, The usual method of discovery and proof, by first esta 

 blishing the most general propositions, then applying and proving 

 the intermediate axioms according to them, is the parent of error 

 and the calamity of every science. But we will treat more fully 

 of that which we now slightly touch upon, when we come to lay- 

 down the true way of interpreting nature, after having gone 

 through the above expiatory process and purification of the 

 mind. 



LXX. But experience is by far the best demonstration, pro 

 vided it atffiere to the experiment actually made, for if that 

 experiment be transferred to other subjects apparently similar, 

 unless with proper and methodical caution it becomes fallacious. 

 The present method of experiment is blind and stupid ; hence 



1 Bacon is hardly correct in implying that the enumerationem per 

 simplicem was the only light in which the ancients looked upon induc 

 tion, as they appear to have regarded it as only one, and that the least 

 important, of its species. Aristotle expressly considers induction in a 

 perfect or dialectic sense, and in an imperfect or rhetorical sense. Thus 

 if a genus (G), contains four species (A, B, c, D), the syllogism would lead 

 us to infer, that what is true of G, is true of any of the one four. But perfect 

 induction would reason, that what we can prove of A, B, C, D, separately, 

 we may properly state as true of G, the whole genus. This is evidently a 

 formal argument as demonstrative as the syllogism. In necessary matters, 

 however, legitimate induction may claim a wider province, and infer of 

 the whole genus what is only apparent in a part of the species. Such 

 are those inductive inferences which concern the laws of nature, the 

 immutability of forms, by which Bacon strove to erect his new system 

 of philosophy. The Stagyrite, however, looked upon enumerationem per 

 simplicem, without any regard to the nature of the matter, or to the com 

 pleteness oi the species, with as much reprehensive caution as Bacon, 

 and guarded his readers against it as the source of innumerable errors, 

 Ed. 



