638 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



The second part of the problem proposed for this discussion is equivalent 

 to the question, which point in a V T area is best for the purposes in 

 hand. The answer is given by the other two sets of curves. 



The £ curves are shown in Figure 6, and are straight lines, the 

 efficiency of a cycle of this type being independent of its breadth. 

 Figure 7 shows the TF curves. The line P = 1 is a part of the curve 

 W=^ because it is also the curve ^=0. The P axis is also a part 



400 460 500 



QinBTU 



Figure 5. 

 T>/pe A. VT area. This area (of which only the lower part is included within 

 the limits of the figure) contains all the adiabatic cycles which satisfy the conditions 

 Te ^ T, and F< V. The curves within give the efficiency and work of these 

 available cycles and enable one to choose between them. 



of the curve W=0, because a type A cycle whose point is on that axis, 

 has zero area. Each of the other curves has these lines as asymptotes. 



In Figure 5 the full lines are efficiency curves, and the dotted lines 

 are work curves. If efficiency is the important thing, the best point 

 is evidently as far up in the area as possible, that is at the intersection 

 of the curve V= Fwith the P axis. The corresponding cycle is a null 

 cycle {Q = W = 0) and must be regarded as a limiting case in the 

 direction of increasing efficiency. If, on the other hand, the important 



