12S Mr Glover on Forms of Induction. 



Avarrant the conclusion. Dr Whately, in terms synonymous with 

 those of Gassencli, asserts that an inductive inference, drawn 

 from a part of a class with regard to the whole, can only be sup- 

 posed legitimate through a species of logical fiction, in assum- 

 ing that one or two of a small number of instances do ade- 

 quately represent the class to which they belong. Now, we 

 maintain, that, in the greater number of inductions performed 

 in physical inquiry, such a supposition involves no fiction, but 

 rather a positive fact ; a class of facts, the individuals of which 

 are in external aspect somewhat dissimilar, being often suffi- 

 ciently represented for all the purposes required in the induc- 

 tion by a very few instances. 



The more essential properties of bodies are the objects of 

 scientific investigation ; and it is probably only where the in- 

 duction has regard to such properties, that one fact can be ta- 

 ken as a specimen of others analogous with it. Now, there are 

 instances created in nature with the properties common to their 

 class, so highly developed in them, as that the relations of those 

 can be more readily discerned. And often by the aid of expe- 

 riment, those properties can be so tested as to enable it to be 

 known, that, in the instance experimented on, where one pro- 

 perty is placed in a certain situation, another will attend it in 

 a certain order. And the mind having such a knowledge of 

 a class as to be able to divest its individuals of their accidental 

 properties, and to discover in them one essential arrangement, 

 defines the whole class to possess that arrangement, — which be- 

 ing found, in a certain number of well-marked cases, dependent 

 upon the relationship of some properties, is supposed, upon the 

 oTound of an intuitive conviction of like effects being owing to 

 like causes, to be dependent upon the same properties in all 

 the other cases. If the primary definition did not include the 

 whole class, neither should the last inference affect the whole 

 class ; in that case, the extension of the conclusions beyond the 

 individuals known to possess a defined arrangement, would be a 

 mere presumption. But let it be clearly understood that there 

 is a power in induction to determine the nature of individual 



* See Whately '3 Logic, Art. Induction throughout ; also some strictures 

 of Iklr Stewait in the 2d vol. cf his Lectures, p. 345, upon a passage from Dr 

 Wallis of Oxford. 



