202 THE COLOUR-SENSE IN FISHES 



equal, other colours and mixtures of colours 

 would not answer as well. The delusion results 

 from pursuing what, in inductive logic, is called 

 the method of agreement, and not checking its 

 results by the method of difference.' 



This passage emphasises admirably the vague 

 attitude of thought of even educated men on 

 such questions, and the fallacy of drawing 

 general or universal conclusions from isolated 

 and exceptional instances. If we are to arrive 

 at a more exact apprehension of the powers and 

 limits of the vision offish, and of the impressions 

 produced in them by colours, we must first have 

 a series of long and carefully conducted experi- 

 ments. The results, duly noted and tabulated, 

 would form a ground-work for investigations 

 that should lead to conclusions of greatest 

 interest to anglers and naturalists. 



