Proceedings of the Polytechnic Association. 1088 



divisions of the diagram, we take the terminal pressure in the dividers, 

 or on a slip of paper, and find with which of the vertical lines it cor- 

 responds, at the secant corresponding with the number of parts into 

 which the diagram is divided ; and each of the other pressures will 

 be found at the intersection of its corresponding secant with the same 

 vertical line, and may be transferred directly to the divisions of the 

 diagram. Should the number of divisions exceed ten, the dotted 

 secants are used. They also serve for semi-divisions of the diagram, 

 where accuracy requires them ; as where the point of cut-off is quite 

 early in the stroke. It is not essential to determine the terminal 

 pressure in all cases, but frequently it will be sufficient to take the 

 pressure at the ninth, or any convenient division, and from that point 

 lay off the curve each way. 



Fig. 5 is a device for laying off t'he divisions of the diagram, inclu- 

 ding clearance, the length, of diagram and percentage of clearance to 

 stroke being known. It is a diagonal scale with one of the divisions 

 subdivided in a progressive ratio. In using it, the length of the 

 indicator-diagram is taken on a slip of paper, and so placed upon 

 the scale, that while horizontal, one end rests upon the subdivision 

 which represents the percentage of clearance, while the other end ia 

 at the extreme line of the scale ; the several principal divisions of the 

 scales are then marked on the slip of paper, the distance of the exte- 

 rior line from the percentage subdivision mark being the clear- 

 ance in the same scale as the stroke. When the length including 

 clearance is given, this scale is used simply as a proportional scale 

 of parts. "With this is given a scale 14.7 pounds in various scales of 

 pressure, a very convenient adjunct for laying down the line of no 

 pressure or perfect vacuum. 



Let us now consider briefl}'', the effect of condensation and leakage 

 of the piston or exhaust valves ; and whether these so modify the 

 results so obtained by this mode of comparison as to render it of little 

 value. 



Of course, any cause which lowers the terminal pressure, will 

 decrease the size of the theoretical diagram in any given case, and 

 thus give the engine a higher than its proper rate. Leakage from the 

 cylinder produces this effect, and when this exists to any consider- 

 able extent, the diagrams are worthless for any other purpose than 

 a measure of power. As the amount of such leakage is not measur- 

 able upon the diagram, it is important, when diagrams are taken for 

 the above investigation, or in fact for any purpose, that this point 



