Mt Glover on Forms of Induction. 181 



same thicknesses, exhibit the same colours ; lastly, it was dis- 

 covered that the transparent body viewed obliquely, did not 

 exhibit the same colour at the same place, as when viewed di- 

 rectly, and that bodies undergo changes in colour by alterations 

 in their mechanical condition (as well as can be observed) ac- 

 cording to the law, which at this part of the inquiry might be 

 supposed to exist. This famous experiment enabled Newton to 

 frame a law, which states that the causes of the different colours 

 of bodies exist in the sizes of their component particles. 



When one instance does not afford a sufficient display of 

 properties to allow of an extensive inference being drawn, other 

 instances must be got together, which in a mass have such varied 

 characters as to make up a strong case. Thus, in Dr Wells's 

 theory of dew the great doctrine of which is, that in all bodies 

 on the earth's surface, the dew-attracting power bears a uniform 

 ratio to the power of radiating heat ; the author proves his main 

 fact, by taking platinum, gold, silver, lead, charcoal, grass, and 

 gravel, with such like instances ; cases which, put together, may 

 be supposed to afford a fair specimen of the relations preserved 

 by the properties in question throughout all bodies whatever, 

 there being among these instances every conceivable variety in 

 radiating power from the zero of burnished silver to the maxi- 

 mum of porous charcoal ; and he finds, that, in the cases expe- 

 rimented on, the relation between the properties is regular. He 

 therefore draws an inference for all bodies in the circumstances 

 of those which he has tried. 



When induction, therefore, takes cognizance of the more es- 

 sential properties of bodies, and investigates those by means of 

 instruments, an inference including all facts in certain contin- 

 gent circumstances can be drawn from a few analogous facts 

 in such a condition ; and the mind, in framing this it may be 

 universal generalization, does not necessarily make use of any 

 fiction, logical or physical, but proceeds upon what may often 

 with reason be believed to be a sure fact. 



But where the properties are not of this essential and inti- 

 mate kind, and do not admit of being investigated experiment- 

 ally, so as to allow of their relations being determined, a uni- 

 versal conclusion cannot be drawn without the use, by the mind, 



i2 



