102 'HUXLEY MEMORIAL LECTURES 



mathematical certainty (which mathematical cer- 

 tainty, in such a case, must always be deceptive), 

 but will be content (like a good number of 

 sciences at the present time) with a sufficiently 

 high degree of probability, with a probability 

 capable of being pushed farther and farther till 

 it becomes so great that it may end by becoming 

 practically equivalent to certainty. In short, I 

 am of opinion that there is no absolutely certain 

 principle from which the answer to these ques- 

 tions can be deduced in a mathematical way. 

 Nor does there exist a privileged fact, or a col- 

 lection of privileged facts, from which the answer 

 can be inferred, as, for example, occurs in a 

 problem in physics or chemistry. But it seems 

 to me that in a great number of different fields 

 there is a great number of collections of facts, 

 each of which, considered apart, gives us a direc- 

 tion in which the answer to the problem may be 

 sought a direction only. But it is a great thing 

 to have even a direction, and still more to have 

 several directions, for at the precise point where 

 these directions converge might be found the 

 solution we are seeking. What we possess mean- 

 while are lines of facts, none of which goes far 

 enough, none of which goes right up to the point 

 which interests us and at which we want to place 

 ourselves ; but these lines may be more and more 

 prolonged, and they, moreover, already suffi- 

 ciently indicate to us, by the ideal prolongation 



