Prof. Barnard on a modification of the Ericsson Engine. 245 
pressed air has been inferred as if it depended upon nothing but 
altered density, according to the law of Mariotte. But this is to 
compute the power of a caloric engine, by disregarding a material 
part of the caloric in the case. 
It is evident that, if ¢ is the tension of the air in the reservoir, 
then the air in the supply cylinder will open the valves and begin 
to enter the reservoir (in consequence of the elasticity due to heat 
of compression), before the density reaches is’ In Poisson’s 
Traité de Mecanique, we find the following formula adapted to 
case. ‘ 
o\7 
p'=p (£) ’ (I) 
In which p and p’, 9 and ¢&, denote the pressures and densities 
of the same air before and after compression, respectively, and y 
expresses the ratio between the specific heats of air at constant 
pressure and at constant volume. The mean of the valves given 
for y by Poisson, is 1-36.* 
following, which denotes the sensible temperature of the air 
after the change ; 
a=(040)(£)"" 0, (Il) 
In which 6 and 6 express the temperatures before and after the 
change of density, © represents the number of degrees of increase 
of temperarature required to double the bulk of air taken origi- 
nally at the temperature of 32° Fah., (=491° F., according to 
Regnault, ) and ¢ and & are used as before. 
But, assuming these several elements to be variable, becomes - 
a fraction of all of them. We may find a general expression for 
its value, and thus, if we please, eliminate it, as follows. 
In the London and Edinburgh Phil. Mag., for June, Mr. Rankine employs 1-41 
; : : rege ir 
- Scie 
Rot fallen under the notice of the writer, at the time of its preparation. com- 
putations which follow, stand as originally made, with the value of 7 =1°36. 
