HOW THE TESTS WE HE CONDUCTED. 25 



piston at different points. If the end of the exhaust pipe is 

 open to view, as would be the case with a non-condensing 

 engine, the steam which leaks through can be observed at the 

 open outlet. This can also be done in the case of a condensing 

 engine where there is a branch exhaust pipe leading to the at- 

 mosphere. Where the engine is condensing, and no such branch 

 is provided, and there is no other opening in the exhaust pipe 

 in front of the condenser, a pretty good idea can be obtained of 

 the general facts by observing the amount which the condenser 

 is heated by the steam which leaks. 



With the piston in any given position in a Corliss engine, 

 the leakage on such tests embraces the leakage of one exhaust 

 valve, one steam valve, and the piston. To investigate the 

 leakage of the other steam valve and the other exhaust valve, 

 the test must be made with the piston taking steam on the 

 opposite stroke. In either case, if the previous inspection of 

 the two steam valves shows them to be leaking, this fact must 

 be considered in drawing conclusions as to the leakage of the 

 piston and exhaust valves. 



There is another method of testing the leakage of piston and 

 exhaust valves, namely, the " time method." The fly-wheel is 

 blocked, as before, with the piston at some distance from the 

 beginning of the stroke, the throttle valve is opened, and steam 

 is admitted at full pressure until the cylinder is thoroughly 

 warmed. Then the throttle valve is shut, and the length of 

 time is observed which is required for the steam to escape 

 through the leaking openings. To conduct the test properly, 

 an indicator is attached to the cylinder at the end containing 

 the steam, and a mark is made on a blank card at intervals of, 

 say, one-quarter of a minute from the time the throttle valve is 

 closed ; and by this means the rate of fall of pressure and escape 

 of steam is recorded. This test, like the others, is qualitative, 

 and not quantitative. The relative condition of the engine 

 determined from results of the time tests must be judged by 

 comparing with other cases where known conditions of excel- 

 lence prevailed. In a leaking engine the fall of pressure on a 

 test of this kind is very rapid. If the leakage is serious, the 



