34 ANNUAL OF SCIENTIFIC DISCOVERY. 



absence of frictional resistance in some engines contrasting 

 strangely with the expenditure of power absolutely wasted in 

 others. It is not the mere loss of fuel alone although that is 

 bad enough that has to be considered in dealing with this sub- 

 ject. We find engines unable to do their work overloaded and 

 worn out ; boilers burned and overtaxed ; grease and oil wasted ; 

 indeed, we go so far as to hold that every horse-power unneces- 

 sarily spent in overcoming the frictional resistance of a steam 

 engine costs three times as much as if it were spent in doing 

 useful work, and this without taking at all into account the fact 

 that useful work returns money, while what we may term the in- 

 ternal work of the steam engine returns none. 



The difficulties which lie in the way of ascertaining by actual 

 experiment what the frictional resistance of an engine is are 

 very great, and to this cause no doubt is to be attributed the 

 greater portion of the existing ignorance of the subject. The 

 obstacles in the way are of two kinds. In the first place, it is 

 very difficult to put a dynamometer or brake on large engines, 

 whereby to ascertain their dut} r ; and, in the s&eond place, the 

 amount of friction varies not only in different engines, but in the 

 same engines, in a very extraordinary way. As regards the first 

 difficulty, we can, in the case of pumping engines, ascertain pre- 

 cisely how many foot-pounds of work an engine actually gives 

 out in the shape of useful effect, while the indicator shows the 

 work done on the piston ; but from these data it is impossible to 

 calculate engine friction exactly, because our calculations are 

 complicated by the greater or less efficiency of the pumps. It is 

 possible that nothing can be more deceptive than the results 

 obtained from pumping engines, and therefore we have no 

 hesitation in rejecting their aid in dealing with questions of 

 engine friction. Practically speaking the only generally available 

 test is the indicator, used with the engine light and the engine 

 loaded ; but diagrams taken thus do not account for the extra 

 friction due to the performance of work, though useful to some 

 extent in their way ; but no investigation of the qualities of an 

 engine can be regarded as complete unless the dynamometer is 

 used as well as the indicator. 



As regards the variation in the loss by friction in the steam 

 engine, a very great deal might be said which we shall not 

 attempt to say now. It may induce others to experiment for 

 themselves, however, if we place a few facts curiously illustrative 

 of the peculiar phenomena of engine friction before our readers. 

 In one case we conducted the experiment personally ; for the 

 results of the other we are indebted to a gentleman who, in 

 superintending the replacement of ordinary boilers by the now 

 well-known Howard boiler, has occasion to indicate a very large 

 number of engines, and on whose accuracy we can rely with crr- 

 tainty. In the first experiment which w r e shall cite we found the 

 full power exerted by a rolling-null engine in the north of Eng- 

 land, -- where, it is unnecessary to specify, to be 291.5 horse. 

 This included the resistance due to fly weighing 30 tons, a bar- 

 mill with 2 pairs of rolls working on heavy orders, and the 



