44 WHAT IS SCIENCE? 



raised by finding that there is really no such thing as 

 " steel." It has been implied that there are certain 

 properties common to all steels. If we make the word 

 steel mean anything which has these common properties, 

 whatever other properties it may have, then, since these 

 properties are invariably associated, the proposition that 

 there is steel (in this sense) will be a true law. But if 

 we examine the matter closely enough, we find that there 

 are not really any properties common to all steels ; we 

 can find common properties only if we overlook distinc- 

 tions which are among the most important in science. 

 All steels, we may say, contain iron and carbon and all 

 are capable of being tempered. But they do not all con- 

 tain the same amount of iron and carbon, nor are their 

 tempering properties all the same ; and the variation 

 in the amount of carbon they contain is associated with 

 important variations of their tempering properties. As 

 we shall see if, indeed, it is not obvious one of the 

 most important distinctions which science makes is 

 between objects or substances which have all the same 

 property, but have it in different degrees ; the study of 

 such distinctions is measurement, and measurement is 

 essential to science. The deeper we inquire, the less we 

 shall feel inclined to regard the statement, there is 

 steel, as a law, asserting invariable associations We 

 shall want to break this law up into many laws, one 

 corresponding to each of the different kinds of steel that 

 we can recognize by the most delicate investigation ; 

 when we have pushed these distinctions to the utmost 

 limit, then, and not till then, we shall have arrived at 

 laws stating truly invariable associations between the 

 various properties of these different kinds of steel. 



Here we meet with a process in the development of 

 science precisely contrary to that which we considered 

 before. We were then considering the process by which 

 science, starting from a relatively small number of laws, 



