22 ENGINE TESTS. 



sumption is made that the loss of steam from condensation and 

 drips during the time the engine is first starting and attaining 

 its working speed counterbalances the deficiency of load be- 

 tween the time when the speed is attained and the working- 

 load is actually applied. In factory work, the interval of time 

 between the attainment of the working speed and the applica- 

 tion of the full load is usually less than three minutes. 



In taking diagrams from an engine with the object of deter- 

 mining its power, it is not desirable to limit the diagram to a 

 single revolution. The marking-point of the indicator should 

 be applied long enough to obtain four or five diagrams, cor- 

 responding to that number of successive revolutions, in order 

 that the effect which the fluctuations in the governing mechan- 

 ism has upon the diagrams may be provided for. In working 

 up the diagrams, then, the mean pressure is obtained for the 

 average diagram, and not for any single one. By pursuing 

 this method, the average power which is determined relates to 

 several times as many diagrams as it would if it were confined 

 to a single revolution in each case. Instances are frequently 

 met where the fluctuations in the cut-off for half a dozen suc- 

 cessive diagrams varies from 2 to 5 per cent of the length of 

 the stroke, and in such cases this matter is of considerable im- 

 portance. As a convenience in working up the diagrams, a 

 good plan to follow is to go over each one with a pencil, and 

 trace with dotted lines the diagram which represents an aver- 

 age of those made by the indicator, and in the subsequent 

 calculations to use this dotted diagram. When a load is ex- 

 tremely fluctuating, this system should be carried further. The 

 period of taking the diagram should extend over at least a full 

 minute, though it is unnecessary to make it a continuous dia- 

 gram for this length of time. The marking-point can be pre- 

 ferably applied for three or four revolutions at the beginning 

 of, say every ten seconds of a minute, and in that way the 

 record applies to some twenty revolutions spread over the full 

 period. Having these diagrams now on the same card, an 

 average line can be dotted in by hand, using the best judgment 

 after examining the appearance of the various diagrams and 

 their location. 



