ch. in.] THE TEST OF TRUTH. 61 



from its attribute without destroying the conception alto- 

 gether. So we cannot conceive that a lump of iron will 

 float in water. Why ? Because our conception of iron, 

 formed solely by experience, is that of a substance which 

 sinks in water ; and to imagine it otherwise is to suppress the 

 conception, either of iron or of water, and to substitute some 

 other conception in its place. We may try the experiment 

 for ourselves. Try to imagine a lump of iron floating in 

 water, and you will find that you cannot do it, without 

 mentally endowing either the iron or the water with other 

 attributes than those by virtue of which these substances are 

 what they are, and thus your attempt destroys itself. Yet no 

 Kantian would deny that your conception of iron or of water 

 is wholly formed by experience. Your conception is just what 

 experience has made it, and you cannot alter it without de- 

 stroying it, simply because you cannot transcend experience. 



Here then we come to a conclusion quite the reverse of 

 that maintained by the Kantians. " The irresistible tendency 

 we have to anticipate that the future course of events will 

 resemble the past, is simply that we have experience only of 

 the past, and as we cannot transcend our experience, we 

 cannot conceive things really existing otherwise than as we 

 have Kiwwn them. The very fact of our being compelled 

 to judge of the unknown by the known — of our irresistibly 

 anticipating that the future course of events will resemble 

 the past — of our incapacity to believe that the same effects 

 should not follow from the same causes — this very fact is a 

 triumphant proof of our having no ideas not acquired through 

 experience. If we had a priori ideas, these, as independent 

 of, and superior to, all experience, would enable us to judge 

 the unknown according to some other standard than that of 

 the known. But no other standard is possible for us." * 



The same general considerations will apply to the truths of 

 mathematics, which some Kantians regard as the necessary 



1 Lewes, History of Philosophy, 2nd edition, p. 668. 



