354 PERSONAL PREFAB ATION FOR RESEARCH. 



we can always prove a thing by disproving its negative or 

 contradictory, so in a research we often prove a thing by 

 disproving all other things (i.e. causes and explanations) 

 that lie within the sphere of possibility. Every existence 

 has its proof ; and indirect inference often indicates to us 

 what a term is by showing us what it is not. 



Indirect inference assists us to perceive all the logical 

 conclusions or alternatives possible in the most complex 

 case, and helps us to render explicit the whole of the 

 available truth in any series of statements, and exhibit it 

 in other forms of conclusion ; and as the phenomena of 

 nature are usually complex, this process is very extensively 

 used in scientific investigation. We employ it in most 

 scientific researches also, because in the majority of in- 

 vestigations the phenomenon we are examining is attended, 

 not only by its necessary causes and conditions, but also 

 by unessential and unnecessary circumstances, which can- 

 not be separated without simultaneously excluding essen- 

 tial or necessary ones. Most experiments, even in their 

 simplest form, especially those involving molecular action, 

 include a number of inseparable circumstances which do 

 not materially affect the particular result. 



The indirect method is a troublesome one, because in 

 most of the cases in which we employ it the conditions 

 are several or numerous, and the number of possible alter- 

 native hypotheses increases at a very rapid rate with each 

 additional condition ; and we only employ it because it is 

 the method suitable for unravelling complex phenomena. 



In an investigation we always reduce the particular 

 experiment to its simplest state as soon as we can, by 

 separating as completely as possible all unessential cir- 

 cumstances ; and having done this, a limited number only 

 of observed conditions remain. We now imagine as com- 

 pletely as we can, by means of mental analysis, or by com- 



