83 



nection with future cases. Simple enumeration of cases can 

 have reference only to the past; they never necessarily refer 

 to the future. As far as ' objective impossibility ' is concerned, 

 on the 'experience hypothesis', the expression is valueless, 

 for it is just as possible that a variation may occur the ten 

 millionth time as the second time, though, to be sure, not more 

 probable. In fact, on the basis of mere enumerationem 

 simplicem, there is no warrant whatever for positing 'neces- 

 sities of thought', but only expectations of thought. Past 

 experience, no matter how long, does not imply any necessity 

 whatever. In short, experiences which do not all possess 

 necessary relations cannot produce such relations. 



Let it be supposed, for example, that in one thousand 

 experiences an intuition is not "fixed", but that in the thou- 

 sand and first it is. On the 'experience hypothesis', there is 

 no way to account for this, for that which has made a necessary 

 relation out of the thousand and first case must be some 

 peculiar characteristic of that particular case which is absent 

 from all the rest. It is not the result of the previous one 

 thousand cases, for in the thousandth case, and in all previous 

 cases, there was no necessity. 



How many times, one might ask, is it necessary to ex- 

 perience a certain relation before it becomes a " fixed intuition"? 

 Suppose it were possible, from Spencer's standpoint, to name 

 a certain number of experiences which would be required to 

 produce a 'necessity of thought'. If this number were one 

 thousand and one repetitions, then on the thousand and 

 first repetition we should have a necessary relation. But 

 that only means, on Spencer's basis, that the gray matter of 

 the brain has been impressed in a certain way for one thousand 

 and one times. Past experience may be said to be a good 

 guarantee that what has happened has happened, but, in 

 the matter of necessity, the past has no word for the future. 

 Now, it may be that on the thousand and second occurrence 

 of this relation, the whole thing may be changed, and the order 

 of occurrence of the relation in question be different. Ne- 

 cessity will then have disappeared until one thousand and one 

 cases of the new relation have been experienced. And so 

 on. In other words, necessity itself on the 'experience hy- 

 pothesis' becomes contingent. 



Not only is Spencer's hypothesis untenable on the above 

 basis, but such a position is directly contrary to numerous 

 facts of every-day experience. For example, through countless 

 ages, night and day have followed one another unceasingly; 

 the sun, moon, and stars have shone; the seasons have 



