SCIENCE 367 



since the beginning of this century. Many things are 

 now known with absolute certainty which before it were 

 unknown. No one will probably deny that science is 

 still advancing,, Nevertheless such advance would be 

 impossible if we could not, by observations and reason- 

 ings, become so certain with respect to some facts, as to 

 be able to make them starting-points for fresh observa- 

 tions and reasonings as to other facts. It is impossible 

 to deny that we may repose with absolute confidence and 

 entire certainty upon a variety of such newly ascertained 

 facts ; e.g., that as to the mode of reproduction in 

 ferns.* 



The laws of the reasoning process and the conditions 

 of its validity, have been explained in an elementary 

 manner in the chapter on logic. But however excellent 

 and admirable our process of reasoning may be, the 

 process must stop somewhere. For in order to prove 

 anything by reasoning, we must show that it necessarily 

 follows, as a consequence from antecedent truths, which 

 therefore must be deemed yet more indisputable. But 

 this process cannot go on for ever. We cannot prove 

 everything. However long our arguments may be, we 

 must at last come to statements which have to be taken 

 for granted, otherwise we should have to reason for ever, 

 which is, in effect, to affirm that nothing at all can ever 

 be proved. If we could not know some things without 

 their being proved, we could never reason validly at all. 

 So again with respect to the validity of the reasoning 

 process itself. When the laws of logic have been duly 

 complied with, the validity of that process is a truth 

 which we can see to be true by studying it, and which 

 is also implied in the fact that science progresses. If 



* See ante, p. 205. 



