12 SCIENTIFIC METHOD 



But like mere mathematical deduction, mere physical 

 induction has its limits. The former breaks down on 

 the subtlety of Nature, the latter on its imperceptibility. 

 As we pass from the simple to the complex there are 

 too many alternatives for us to deduce one ; and the 

 further we proceed from experience beyond it the more 

 difficult is it to get instances to induce universals. 

 Curiously enough the nature of heat, the very thing 

 which Bacon took to illustrate his inductive method, is 

 precisely a case in point : it is beyond induction. Induction 

 will reveal external causes of heat, such as conduction, 

 convection, friction, percussion. It will reveal the ex- 

 ternal effects of heat : Bacon himself made experiments 

 with the air and water thermometer, and showed how 

 air when heated expanded and depressed the water. But 

 he did not in any of his instances show what heat is, 

 because none of them contained that imperceptible 

 oscillating motion of particles which, as Bacon knew 

 but not by induction, is essentially the nature or essence of 

 heat. Something more then is wanted. By the empirical 

 method we know what are called the ' phenomena' oT 

 heat. Now these phenomena are similar to those which 

 we otherwise know to be consequences of motion. By 

 combining these two kinds of knowledge we conclude 

 that, similarly, the phenomena of heat are consequences 

 of heat being a mode of motion. In other words, we 



\ infer the nature of heat, not by induction, but by that 

 kind of deduction which combines ' phenomena ' with 



V laws '. 



Deduction, when based on scientific induction, has 

 several superiorities over its basis. ^ We have already 

 shown that it is more definite, discovering as it does 

 particulars beyond experience which are inaccessible 

 to the generalities of induction. ^Secondly, we have 



