IN THE LAST HALF-CENTURY. 33 



state at any preceding moment. Another 

 is that any of the rules, or so-called ' laws 

 of nature,' by which the relation of phe- 

 nomena is truly defined, is true for all 

 time. The validity of these postulates 

 is a problem of metaphysics ; they are 

 neither self-evident nor are they, strict- 

 ly speaking, demonstrable. The justifica- 

 tion of their employment, as axioms of 

 physical philosophy, lies in the circum- 

 stance that expectations logically based 

 upon them are verified, or, at any rate, 

 not contradicted, whenever they can be 

 tested by experience. 



Physical science therefore rests on veri- and uses 



J hypoth- 



fied or uncontradicted hypotheses ; and, escs. 

 such being the case, it is not surprising 

 that a great condition of its progress has 

 been the invention of verifiable hypothe- 

 ses. It is a favorite popular delusion 

 that the scientific inquirer is under a sort 

 of moral obligation to abstain from going 

 beyond that generalisation of observed 



