THE APPLICATIONS OF SCIENCE 175 



we must first differentiate the various kinds of steel 

 and seek laws which involve each kind separately. 



Neglect of this precaution is one of the most frequent 

 causes of a failure to detect and to cure troubles encoun- 

 tered in industrial processes. The unscientific manager 

 regards as identical everything that is sold to him as steel ; 

 he regards as " water " everything that comes out of 

 the water main ; and as " gas " everything that comes 

 out of the gas main. He does not realize that these 

 substances, ' though called by the same name, may have 

 very different properties ; and when his customary 

 process does not lead to the usual result, he will probably 

 waste a great deal of time and money on far-fetched 

 ideas before he realizes that he did not get the same 

 result because he did not start with the same materials. 

 He can expect his processes to be governed by laws and 

 to lead invariably to the same result only if all the 

 materials and operations involved in those laws and 

 employed in those processes are themselves invariable ; 

 that is to say, if their constituent properties and events 

 are themselves associated by invariable laws. This is 

 a very obvious conclusion when it is pointed out, but 

 there is no conclusion more difficult to impress finally 

 on the practical man ignorant of science. He is misled 

 by words. Words are very useful when they really 

 represent ideas, but are a most terrible danger when they 

 ;iot. A v presents ideas, in the sense important 



1 applications of science, only when the 

 things which it is used to denote are truly collections of 

 properties or events associated by laws ; for it is only 

 tin. -n that the word can properly occur in one of the ! 

 <>n U those applications depend. Perhaps tin- 



most i! ' service which science can render to 



prac insist that laws can only be expected 



to hold between things which an- themselves the expres- 

 sion of laws. 



