Gas-Engine Indicator- Diagram. 75 



C G H I F represents the total loss of heat to the cold cylinder 

 during the stroke. We are in a position to speak of the loss 

 of heat from X=0 to \=1'2 (see § 9). 



This loss must be represented on the diagram to the same 

 scale as the indicator-diagram represents work done, and it is 

 expended in the part of the working stroke from \=0 to 

 X=l*2. Hence if k is the mean ordinate of such a diagram, 



£=2-31 x 61-52 x l-330-i-l-2=158. 



Now the mean ordinate of the curve G H I (fig. 6) from 

 \ = to \=1*2 being taken and found different from 158, all 

 the ordinates of G H I (fig. 6) have been diminished in the 

 proportion of 158 to the mean ordinate of G H I (fig. 6) to 

 get the diagram X L Y of fig. 4. This diagram X L Y repre- 

 sents the rate of loss of heat by conduction and convection 

 from the fluid to the cylinder during the working stroke until 

 the exhaust- valve opens. 



13. Rate at which Combustion goes on during the Stroke. — 

 The curve EFGrH represents the rate at which heat is actu- 

 ally gained by the fluid, and X L Y shows the rate at which 

 heat is wasted to the cylinder ; so that the curve I J K Y 

 shows at every point the rate at which heat is being generated 

 in the fluid by combustion. It is obvious, then, that combus- 

 tion is not complete at the end of the explosion part of the 

 curve, although, as Mr. Clerk's experiments prove, the mix- 

 ture of air and gas is in the proper proportions for explosion 

 immediately behind the piston at all periods of the compres- 

 sion-stroke. The diagram IJKY is specially valuable, as 

 showing the effect of dissociation of the products of combus- 

 tion at such temperatures as obtain in the gas-engine, and are 

 shown in Table VIII. 



Postsckipt, added June lQth. — We have assumed in the 

 paper that the rate of loss of heat by radiation and convection 

 is proportional to the difference between the mean temperature 

 of the fluid and the temperature of the cylinder. When we 

 have more information concerning the distribution of heat in 

 the fluid, and the way in which a heated fluid loses its heat to 

 a cold enclosing vessel, a more accurate assumption may be 

 made; and it is easy to see what alteration this will introduce 

 in our method of obtaining the curve X L Y (fig. 4). 



It is known from the experiments of MM. Dulong and Petit 

 that the rate of cooling by radiation and convection of a solid 

 body increases more rapidly than the difference of tempera- 

 ture, and that it is greater at greater pressures of the gaseous 

 medium between the hot body and the surrounding cold vessel. 



