HOW THE TESTS WERE CONDUCTED. 33 



be learned from the examination of the steam accounted for at 

 cut-off than that accounted for at release. The difference in 

 the work done during expansion is not proportional to the dif- 

 ference in the steam accounted for; and, consequently, the 

 actual loss of economy due to cylinder condensation and leak- 

 age is more closely measured by the percentage which is 

 accounted for at cut-off than by the percentage accounted for 

 at release. Between the two, if only one computation is to be 

 made, it is better to use the cut-off point than the release point. 



Fig. 1. 



The proportions "c" and "h" are found by measuring the en- 

 tire length of the diagram, first erecting perpendiculars at the 

 extreme points, and then measuring the length up to the point 

 marked, dividing one by the other, and ascertaining the result- 

 ing proportion expressed in a decimal. The proportion "e" for 

 the clearance may be found either by measurement of the 

 clearance spaces from drawings of the cylinder and valves or 

 from actual test. The latter is to be preferred ; for drawings, 

 however correct in themselves, do not show the exact measure- 

 ments of the material, especially -of the ports and passages 

 which are in the state of rough casting. 



To measure the clearance by actual test, the engine is carefully 

 set on the centre, with the piston at the end where the measure- 

 ment is to be taken. Assuming, for example, a Corliss engine, 

 the best method to pursue is to remove the steam-valve so as 

 to have access to the whole steam-port, and then fill up the 



