106 GENERAL VIEW AND BASIS OF SCIENTIFIC RESEARCH. 



or error ; but fallacies are frequently beacons to discovery, 

 and the recognition of them is often a great step towards 

 the attainment of truth, because it enables us to diminish 

 the number of uncertain points to be settled. 



Next to the discovery of truth, the most important 

 circumstance to attend to in original scientific research 

 is the avoidance of fallacy, to 'unmask falsehood, and 

 bring truth to light.' l Fallacy is the worst kind of error. 

 If also we hold an erroneous belief we must abandon it' 

 before we can receive the truth ; and if that belief is a 

 firmly fixed one, the removal of it is a very difficult 

 matter. As also in original research there are always very 

 many ways of going wrong, and usually only one (or at 

 most a few) of going right, we need to be continually on 

 our guard lest we may make a mistake. 



In such research we are liable to fall into error at 

 every step. At the very outset, the statement we assume 

 to be a fact, and which we wish to investigate, may not be 

 a fact at all; or the hypothesis we have imagined, and 

 which we intend to examine, may have no possible 

 counterpart in nature. In testing a statement or hypo- 

 thesis also, we may arrange our tests in an improper way, 

 make the experiments carelessly or imperfectly, mistake 

 our sensations or impressions for the true effects of the 

 experiments ; observe the results with the aid of inaccu- 

 rate instruments, or whilst influenced by a biassed or 

 prejudiced state of mind, or an ill state of bodily health, 

 and thus receive false perceptions of them. We may also 

 make false comparisons, and too extensive generalisations 

 of our observations, and draw illogical or imperfect in- 

 ferences from them. 



To an inexperienced investigator, in particular, a study 



1 Shakespeare. 



