The Inductive Theories. 19 



should be controlled and fortified against error. According as 

 the problem of induction was looked at from its theoretical or 

 its practical side, it was stated by him on the one hand to be the 

 discovery of a Form, or necessary accompanying characteristic 

 of a class of phenomena; or, on the other, to be the " freeing of a 

 direction " (Val. Ter. C. 11), that is, the formation of a rule of 

 action (N. O., II, IV); but as he remarks, the two phrases des- 

 cribe identically the same process viewed in different aspects. 

 The rules for the interpretation of Nature which were given by 

 Bacon (N. O. II, X), are in two divisions; first, those concern- 

 ing the manner of forming general propositions from experi- 

 ence; second, those regulating the derivation of new experiments 

 from these generalizations. The rules for the formation of gen- 

 eral propositions are divided into three parts: (1) RuIqs for 

 observing. (2) For arranging our observations in reference to 

 the particular inquiry we are upon; and (3) Rules for true and 

 legitimate Induction. The method of such true Induction is 

 sketched in the XVItli axiom and consists of the exclusion from 

 the list of possible Forms of the phenomenon to be investiga- 

 ted all qualities which are not found where the given phenome- 

 non is, or are found when it is not, or are found to increase 

 when it diminishes or vice versa. "Then indeed," he contin- 

 ues, "after the rejection and exclusion has been duly made, 

 there will remain at the bottom, all light opinions vanishing 

 into smoke, a Form affirmative, solid and true and well 

 defined." 



In the light of further experience in Inductive investigation, 

 the theory of Induction thus laid down may be criticized as fol- 

 lows: First: Bacon assumes that. every nature or phenomenon 

 of a given kind has a Form or invariably accompanying quality; 

 and that the multifarious phenomena of the universe are the 

 result of the combination of such Forms, which are compara- 

 tively few in number. This doctrine of Forms except, in so far 

 as it is allied to the idea of causation has not formed a part of 

 later theories of induction. Experience has not brought to 

 view any such invariability in the co-existences of phenomena; 

 and even as an approximation to a doctrine of causation, the 



