FALLACIES OF TESTIMONY. 243 



which, within my own experience, what seemed the most trust 

 worthy testimony has been completely discredited by the appli 

 cation of such methods, I shall give some examples hereafter. 



I would by no means claim for myself or any other scientific 

 man an immunity from idolatrous prepossessions ; for we must 

 all be guided in our researches by some notion of what we expect 

 to find ; and this notion may be very misleading. Thus, when 

 no metal was known that is not several times heavier than water, 

 it was not surprising that Dr. Pearson, as he poised upon his 

 finder the first globule of potassium produced by the battery of 

 Davy, should have exclaimed, &quot; Bless me, how heavy it is ! &quot; 

 though, when thrown into water, the metal floated upon it. 

 But while the true disciple of Bacon is on his guard against 

 &quot; idolatry,&quot; and is constantly finding himself rudely handled (as 

 Dr. Pearson was) by &quot; the irresistible logic of facts &quot; if he falls 

 into it, the pledged upholder of any religious system can be 

 scarcely other than, in some degree, an &quot; idolater.&quot; The real 

 philosopher, says Schiller, is distinguished from the &quot; trader in 

 knowledge&quot; by his -always loving truth better than his system.&quot; 



Bacon s classification of &quot;idols&quot; is based on the sources of 

 our prepossessions ; and although his four types graduate in 

 sensibly into each other, yet the study of them is very profitable. 

 Sir John Herschel is, I think, less successful when he classifies 

 them as (i) prejudices of opinion and (2) prejudices of sense ; 

 because an analysis of any of his &quot; prejudices of sense &quot; shows 

 that it is really a &quot;prejudice of opinion.&quot; My first object is to 

 show that we are liable to be affected by our prepossessions at 

 every stage of our mental activity, from our primary reception 

 of impressions from without, to the highest exercise of our reason 

 ing powers ; and that the value of the testimony of any individu.il, 

 therefore, as to any fact whatever, essentially depends upon his 

 freedom from any prepossessions that can affect it. 



That our own states of consciousness constitute what are, to 

 each individual, the most certain of all truths in a philosophical 

 sense (as J. S. Mill says) the only certain truths will, I suppose, 

 be generally admitted ; but there is a wide hiatus between this, 

 and the position that every state of consciousness which repre- 



