102 : H LIMIT-ATIO-NS OF SCIENCE 



tween the spheres. It really amounts to endowing 

 these cosmic fragments with the intelligence of Leib- 

 nitzian monads; if impact and grinding produced them, 

 why are they also not true spheres? The sand on the 

 seashore is as spherical as the pebbles. Then too, if 

 the original parts of space were reduced in size by at- 

 trition, why did they attain a certain size and then 

 cease to be worn away? Or are we to suppose that 

 the whole universe earth, stars, and interstellar spaces 

 are still grinding themselves away until some day all 

 its parts will be reduced to the size of the cosmic dust 

 which he calls fire ? We might go still further and let 

 the universe actually grind itself into nothing, and 

 simply vanish. 



But even if we showed, step by step, that the scheme 

 not only was not true, but even not capable of resisting 

 the most cursory criticism, we should be met by the 

 answer; that as knowledge increases, details which are 

 erroneous will be abandoned and new ones substituted 

 which better approximate to the truth. This counter- 

 criticism seems aside from the question; it would be 

 strange if the efforts made to discover new phenomena 

 and laws, and to correct false ones, did not increase our 

 knowledge. But is this aim furthered by such hypo- 

 thetical systems, which attempt to describe the mechan- 

 ism of these phenomena and laws, and which, at bot- 

 tom, assume that their authors are the creators of the 

 universe and not observers of one whose laws and 



