COINCIDENCES 73 



In one sense the chance of the molecules returning 

 to one half of the vessel is too absurdly small to think 

 about. Yet in science we think about it a great deal, 

 because it gives a measure of the irrevocable mischief 

 we did when we casually removed the partition. Even 

 if we had good reasons for wanting the gas to fill the 

 vessel there was no need to waste the organisation; as 

 we have mentioned, it is negotiable and might have been 

 passed on somewhere where it was useful.* When the 

 gas was released and began to spread across the vessel, 

 say from left to right, there was no immediate increase 

 of the random element. In order to spread from left to 

 right, left-to-right velocities of the molecules must have 

 preponderated, that is to say the motion was partly 

 organised. Organisation of position was replaced by 

 organisation of motion. A moment later the molecules 

 struck the farther wall of the vessel and the random 

 element began to increase. But, before it was destroyed, 

 the left-to-right organisation of molecular velocities was 

 the exact numerical equivalent of the lost organisation 

 in space. By that we mean that the chance against the 

 left-to-right preponderance of velocity occurring by 

 accident is the same as the chance against segregation 

 in one half of the vessel occurring by accident. 



The adverse chance here mentioned is a preposterous 

 number which (written in the usual decimal notation) 

 would fill all the books in the world many times over. 

 We are not interested in it as a practical contingency; 

 but we are interested in the fact that it is definite. It 

 raises "organisation" from a vague descriptive epithet 

 to one of the measurable quantities of exact science. 

 We are confronted with many kinds of organisation. 



* If the gas in expanding had been made to move a piston, the 

 organisation would have passed into the motion of the piston. 



