18 ENGINE TESTS. 



taking steam. If it leaks, allowance should be made for this 

 leakage. If there is considerable piping, and it pitches toward 

 the throttle valve, it is also necessary that allowance be nxule 

 for the steam which condenses in the pipes and collects at the 

 throttle valve. In some cases it will be seen that the condi- 

 tions may be such that the determination of the correction for 

 leakage may be a difficult matter; but it is a subject which 

 ought always to receive attention when the object of the test, 

 as in the present instances, is to determine the quantity of 

 steam used by the engine alone. 



Whatever method of feed-water measurement is employed, it 

 is necessary that the height of water in the gauge glass should 

 be the same at the end of the allotted time of the test as at the 

 beginning. It is important also that the condition of the fire 

 should be the same at one time as at the other, because, as 

 elsewhere noted, the height of the water may be more or less 

 affected. For example, if the test begins just before firing and 

 with the damper closed, or nearly closed, it should also end 

 just before firing and with the damper likewise closed. It is 

 better to overrun the allotted time or even to cut it short, and 

 have these conditions right, than to overlook them in the desire 

 to make the duration of the trial a predetermined immlier of 

 hours to the exact minute. If the height of the water at the 

 end of the test is different from what it was at the beginning, 

 the necessary correction estimated from the corresponding vol- 

 ume of water is applied to the quantity weighed. This correc- 

 tion is determined with sufficient accuracy, in most cases, by 

 calculation from the known exterior measurements of the boiler. 



INDICATING. 



It is unnecessary for the purposes of this volume to go into 

 a description of steam-engine indicators, for the books on the 

 subject of the indicator furnish an ample amount of informa- 

 tion of this character. It will suffice to say that for most of 

 the tests here reported the instruments used were either of the 

 Tabor or the Crosby pattern, or both. The methods of apply- 

 ing the instruments, however, the means of driving them and 



