'200 GENERAL CONDITIONS OF SCIENTIFIC RESEARCH. 



results of an unusual combination of known conditions; 

 or they are merely opposite instances. Some apparent 

 exceptions, also, are the results of an additional law or 

 condition which interferes ; others seem to limit or even 

 contradict a known law when they really confirm it. 1 



In order to determine whether an apparently excep- 

 tional case is a real one, and our statement of the law 

 which governs it is wrong, we usually require to make an 

 original research. If something remains unexplained 

 which we cannot possibly explain by the aid of our present 

 knowledge, then new experiments and observations of quite 

 a novel kind must be made. Many really exceptional facts 

 always exist for us to investigate, and such facts always 

 show their so-called governing law not to be a law by 

 contradicting it ; they also always indicate the existence 

 of a new law or of a wider one. 



It is usually more intrinsically important to discover 

 an exceptional phenomenon than a conspicuous one ; but 

 having once found a really exceptional one, the next im- 

 portant step is to discover a conspicuous one of the same 

 kind. An extreme or conspicuous instance is of great 

 value in a new research ; because by yielding a powerful 

 effect, it enables us to investigate more clearly and com- 

 pletely the particular phenomenon in all its detail ; it is 

 also often of considerable value in technical applications 

 of the discovery. 



Extreme instances, like really exceptional ones, are 

 usually also discovered only by means of extensive or 

 exhaustive research ; i.e., by taking such a large number 

 of instances as to be certain of including some of the most 

 conspicuous ones. The employment of rare substances is 

 often a likely means of meeting with conspicuous degrees 



1 Jevons, Principles of Science, vol. ii. ch. xxix. 



