62 SCIENCE AND METHOD. 



desired solution. The liberated atoms will then 

 experience collisions, either with each other, or with 

 the atoms that have remained stationary, which 

 they will run against in their course. I apologize 

 once more. My comparison is very crude, but I 

 cannot well see how I could explain my thought 

 in any other way. 



However it be, the only combinations that have 

 any chance of being formed are those in which one 

 at least of the elements is one of the atoms deliber- 

 ately selected by our will. Now it is evidently 

 among these that what I called just now the riglit 

 combination is to be found. Perhaps there is here 

 a means of modifying what was paradoxical in the 

 original hypothesis. 



Yet another observation. It never happens that 

 unconscious work supplies ready-made the result of 

 a lengthy calculation in which we have only to apply 

 fixed rules. It might be supposed that the sub- 

 liminal ego, purely automatic as it is, was peculiarly 

 fitted for this kind of work, which is, in a sense, ex- 

 clusively mechanical. It would seem that, by think- 

 ing overnight of the factors of a multiplication sum, 

 we might hope to find the product ready-made for 

 us on waking ; or, again, that an algebraical calcula- 

 tion, for instance, or a verification could be made 

 unconsciously. Observation proves that such is by no 

 means the case. All that we can hope from these 

 inspirations, which are the fruits of unconscious 

 work, is to obtain points of departure for such 

 calculations. As for the calculations themselves, 

 they must be made in the second period of conscious 

 work which follows the inspiration, and in which 



