Ixii Appendix, 



is annexed the distinction into the two classes of sciences, as follows : — " That 

 science which deals with the succession of reason and consequent, is called an 

 abstract science ; wliile that which deals with cause and effect is called, 

 for the most part, a natural or physical science." To the foregoing definition 

 there appear to me to be several weighty objections, some of them insuperable. 

 In the first place, the wording of the definition is vague and loose : phenomena 

 are said to be influenced by " reasons," these being distinguished from " causes "; 

 and, again, it is not easy to understand what is meant by the " legitimate con- 

 nection and interdependence " of phenomena. The definition, moreover, does 

 not adequately recognize the sciences called Classificatory, such as Botany. 

 These sciences do not deal exclusively with the ascertainment of causes and 

 effects J they are conversant mainly with the determination of concomitants. 

 Furthermore, there is not afforded, in the explanation referred to, 'any defini- 

 tion of Science, taken collectively, but rather two definitions, incongruously 

 combined, pertaining severally to the two classes of sciences, the demonstrative 

 and the inductive. 



It has been said to be the characteristic of Science, that it predicts ; but 

 this statement requires limitation. Not all science predicts ; not all that 

 predicts is science. Demonstrative or abstract science does not predict ; and 

 the same may be said as to much of mixed science ; albeit the application of 

 the sciences enables us to make predictions. But so also does the application 

 of unscientific knowledge. An untutored and unreflective savage can foretell 

 many of the phenomena of nature. The predictive character of science is a 

 fact to be noted, and is exceedingly interesting ] but still the power of predic- 

 tion does not so characterize Science as to furnish its definition. 



All science is knowledge, although not all knowledge is science. Of the 

 things that we know, some are individual events, and others are general truths, 

 or, as I prefer to express it, for the purpose of avoiding ambiguity, generic 

 truths. Of the generic truths that we know, a portion is self-evident ; but these 

 do not of themselves constitute any science ; and all the others are matters of 

 inference. Science, therefore, in the abstract, is the valid inference of generic 

 truth. By an inference being valid I mean, that the data or premises from 

 which a conclusion is derived are known to be true and that the sequence of 

 the conclusion from those data is perceived to be cogent. Wherever these 

 conditions are realized, whatever the subject-matter treated of, there we have 

 scientific knowledge j and if any of these conditions be wanting, there is an 

 absence of science. All valid inference of generic truth is essentially scientific; 

 but, in order to constitute what is termed a science, there must be a series of 

 generic truths connected by some special relations. A science^ therefore, in 

 the concrete, we may define a formulated deparimeiit of generic truths legiti- 

 mately assumed and inferred. 



