62 



every day increases the degree of refinement which characterizes the work. The 

 research laboratory stands as a response to these conditions. It is its function to 

 investigate, in a scientific manner, problems which arise in practice or which may 

 be suggested by practical experience. The fields of science and the field of 

 engineering combined make up its proper domain. Its equipment, therefore, em- 

 braces the delicate apparatus of the scientist and the ponderous machinery of 

 the engineer, and its lines of investigation may be chemical, metallurgical, 

 structural, pneumatic, hydraulic, or thermodynamic. Its methods eliminate the 

 complicating conditions of service and allow efTects to be traced singly to their 

 causes. For example, efforts to determine the power and efficiency of locomotives 

 while in service upon the road extend back through more than three decades, 

 with no general result that is satisfactory. But the difliculties and inaccuracies 

 which appear in the process of road testing entirely disappear when tests are 

 made in the laboratory, for here it is possible to maintain for an indefinite period 

 an unvarying condition of speed and load, and to employ sensitive apparatus in 

 observing the performance of the machine. 



There have been many instances where locomotives on the road have left bent 

 rails in the track behind them, but it required the laboratory to demonstrate that 

 under conditions not uncommon in practice, the drive-wheels of a locomotive 

 leave the track at every revolution. This being proved, the matter of the bent 

 rails was easily explained. 



Again, it has been assumed for years that the draft produced by the exhaust 

 steam in a locomotive was the result of an action similar to that of a pump ; that 

 each puff from the cylinders supplied a ball of steam which filled the stack as a 

 pump piston fills its barrel, and pushed before it a certain volume of the smoke- 

 box gases until it passed out at the top of the stack. Believing this view to be 

 the true one, designers have shaped the details of locomotive draft appliances 

 accordingly, and the value of proposed improvements has been measured by the 

 completeness with which they have satisfied the conditions of the accepted theory. 

 But the processes of the laboratory have disproved this whole assumption. They 

 have shown that the steam does not fill the stack except at its very top, and that 

 the action of the jet is clearly one of induction. In accordance with these 

 results a new theory has been formulated, and although it is but a few months 

 old, the laboratory facts which sustain it are so conclusive that it has already 

 been generally accepted. These illustrations, drawn from a single field of inves- 

 tigation, will serve to show something of the character of the work done by the 

 research laboratory. They might, with equal justice, have been selected from 

 any one of the many different departments into which engineering research may 



