How Science Works 41 



applicable in practice, because the balls begin to run into 

 one another too frequently. Methods have, therefore, been 

 devised for giving the probable results of movement within 

 the system in terms of the number of balls which will hit 

 a given length of cushion in a given length of time, the 

 total number of impacts of all the balls involved, and so 

 forth. By sacrificing one's interest in the behavior of an 

 individual ball one gains a grasp of the behavior of the 

 total system which will be accurate enough for many pur- 

 poses. 



A practical example in everyday life is the prediction 

 of the number of automobile accidents over a given week- 

 end. It is, of course, impossibly difficult to predict when 

 or where a given driver in a given automobile is going to 

 run into another. Nevertheless, the National Safety Council 

 can tell with grisly accuracy how many people are going 

 to be killed next Labor Day. Similarly, the physicist can- 

 not tell which individual radium atom is going to break 

 down into lead and radiant energy, but he knows very 

 accurately what proportion will do so in any selected length 

 of time. Indeed, much of our modern knowledge of the 

 nature of matter has been made possible by the develop- 

 ment of statistical methods for analyzing this type of situ- 

 ation. 



It is much more difl5cult to analyze what Weaver calls 

 "systems of organized complexity." In such systems the 

 parts are so closely interrelated that a modification in any 

 one of them is likely to modify the behavior of all the 

 rest. Living organisms provide the best examples of such 

 systems and it is precisely their organized complexity that 

 contributes most to making the life sciences different from 

 the physical sciences. The parts of the human body are 

 much more closely interrelated than the atoms in a gram 

 of radium or the automobiles on the highways of the United 



