EM A NCI PA TION FROM SCIENTIFIC MA TERIA LISM. 427 



chanical conception of the universe, namely, the assumption 

 that the resolving of phenomena into a system of moving 

 particles is the goal to be aimed at in our explanation of 

 nature. Should, however, this foundation fall — and we 

 have seen that it must — then with it goes the ignorabimus, 

 and science is once more free to move onward. 



I do not believe that this result will be received with 

 astonishment, for judging by my own experience, no in- 

 vestigator of nature has seriously believed in the ignora- 

 bimus, although it was not clear where the weak point 

 of the argument lay. Hence the result gained by the 

 rejection of the mechanical conception of the Universe, 

 namely, the banishment of that menacing spectre, may 

 well be of some value to many a thinker unable to find 

 a flaw in the resistless logic of du Bois-Reymond's 

 argument. 



What has here been set forth for the sake of clearness 

 with respect to special discussions such as the foregoing, 

 has, however, a far wider significance. The doing away 

 with the mechanical construction of the universe goes 

 down to the very foundations of. the whole materialistic 

 conception of things, taking the word materialistic in its 

 scientific sense. If it appears a vain undertaking", ending 

 with every serious attempt in final failure, to give a me- 

 chanical representation of the known phenomena of physics, 

 we are driven to the conclusion that similar attempts in the 

 incomparably more complicated phenomena of organic lite 

 will be still less likely to succeed. The same fundamental 

 contradictions occur here also, and the assertion that all the 

 phenomena of nature can be primarily referred to me- 

 chanical ones cannot even be designated here as a practical 

 working-hypothesis ; it is simply incorrect. 



This error appears more clearly when viewed in the 

 light of the following fact. The equations of mechanics 

 all possess the property that they still hold good when the 

 sign of the quantity denoting time is changed. That is to 

 say, theoretically perfect mechanical processes can take 

 place just as well backwards as forwards. In a purely 

 mechanical world there would be, therefore, no Before and 



