the vacuum, or separate the gases preparatory to diffusion, requires an expenditure of energy at least equal to the mechanical effect to he derived.
Since the reversible engine is as efficient as any heat-engine, and since all reversible engines of whatever construction and whatever the working substance have the same efficiency, it is allowable, in discussing the question as to the amount of mechanical effect derivable under given conditions from a given amount of heat, to assume any form of reversible engine, using any working substance which may be most convenient. And it makes no difference whether the engine assumed be practically possible, so long as we know the properties of the working substance well enough to determine its action under the assumed conditions. Sir W. Thomson, before 1851, assuming Carnot's engine with air as the working substance, furnished us with a very complete discussion of this question. The properties of air in relation to heat are very simple. Heat expands and cold contracts it with great uniformity. Compression heats and expansion cools it according to a well-known law. The effects of any change of volume or of temperature in the cylinder of an engine can, therefore, be exactly predicted.
Suppose a given mass of air to be compressed and the heat developed by compression removed, so that its temperature remains constant. The pressure exerted by it will increase, as shown graphically in the annexed diagram, where O a, measured on the horizontal axis