196 GENERAL CONDITIONS OF SCIENTIFIC RESEARCH. 



There always exist many known phenomena which we 

 cannot explain; and man^ of these must remain unex- 

 plained until science is farther advanced in other depart- 

 ments. Unexplained phenomena may be classed into 

 anomalous, exceptional, contradictory, extreme, con- 

 spicuous, residual, &c., and all these kinds are fre- 

 quently the source of new and important discoveries. 

 Crookes's discovery of the thermic repulsion of bodies 

 in highly rarefied gases arose from an observation of 

 some irregular phenomena whilst weighing bodies in 

 vacua. 



Anomalous, exceptional, and contradictory phenomena 

 are not really so, but only appear so in consequence of 

 our ignorance ; they are exceptions, not to the laws of 

 nature, but only to our imperfect or wrong statements of 

 those laws. Contradictions do not exist in nature, but 

 only in our conceptions of it. Truth cannot conflict with 

 truth. A law of nature cannot fail ; it always fulfils itself, 

 and has no real exceptions. A law or principle is not real 

 unless it is true in all cases ; if it includes exceptions, it is 

 either wrongly given or overstated. A single real excep- 

 tion will overturn the strongest theory ; for instance, the 

 phenomena of optical interference overthrew Newton's 

 corpuscular theory of light. The fact, however, must be 

 proved to be irreconeileable, because it may be only 

 an opposite result of the very law itself, as the rising 

 of a cork and the sinking of a stone in water, are 

 each results of the attraction of the earth ; or it may be 

 merely a case of some interfering circumstance, &c. 



The greatest truths often require a comparison of the 

 greatest number of instances, because they only appear, or 

 are forced upon our attention, by exceptional cases, and 

 such cases are usually met with only during the exami- 

 nation of a large number of instances. In nearly every 



