850 Traxsacttoxs of the Amerwax Institute. 



gives little idea of the actual power exerted, for in every mannfaetory 

 the load is constantly changing. It is more than probable that the 

 excellent results claimed in many cases are obtained by calculating 

 the power from a diagram taken with the full load on, and the cost 

 of the power from the average coal, or worse yet, from the coal which 

 is thrown in the furnace in any particular hour, without noting 

 whether the lire is heavier at the beginning of the hour than at the 

 end. A manufacturers coal bills always tell him what his steam- 

 power has cost for a given time, but his one hundred horse-power 

 engine might have been exerting, on the. average, only fifty horse- 

 power ; so, without actual and careful observation, no results can be 

 obtained of any value to the engineering profession. The only true 

 way is to make thorough trials, and repeat them until the results 

 practically coincide. 



"When the power of the engine is measured by a dynamometer, the 

 s^me care should be taken to frequently record the revolutions of the 

 engine and the indications of the instrument, so as to be able to cal- 

 culate the true average power. Fuller reasons for such pi'ecautions 

 have already been given in the preceding discussion. 



We are now prepared to select the methods and means necessary 

 for a scientific trial of the economy of steam machinery, which shall 

 be complete and above criticism. We must first be^^r in mind that it 

 is the economy that we wish to test, and not the excellent manner in 

 which some device controls the speed of the engine under varying 

 loads. Special trials may be made of each detail, if desired, but only 

 one thing can be tested at a time. To get accurate results, great uni- 

 formity is necessary. The closer the resemblance between the records, 

 at different times, the more correct will be the averages. It is essen- 

 ■ tial, then, to carry a luiiform pressure of steam, and to have a uniform 

 load and speed to the engine. In regular practice the load is neces- 

 sarily varied somewhat, which can only be provided against by fre- 

 quent observations ; but our remarks are more particularly applicable 

 to an establishment fitted up especially to test steam machinery, and 

 in other trials details must be varied according to circumstances. In 

 such case the boiler should be of ample size to do the work, and the 

 pressure should be regulated by a steam damper. The resistance 

 should consist of wind or water M-heels, or pumps. We prefer high 

 speed fans or blowers, as the resistance can then be easily regulated by 

 varying the size of the discharge openings. Tanks should be provided 

 for measuring the feed water of the l)oiler, and it would be well. 



