238 Prof. Barnard ona modification of the Ericsson Engine. 
ing the capacity, predetermined, to bring it to the density of the 
reservoir. Finally, by the descent of the piston P”, it is to be 
discharged through V into the reservoir, R. 
oattempt will be made here to discuss difficulties of a merely 
mechanical nature. Some such are foreseen, and also the means 
of obviating them; but if the proposed modification offers a real 
improvement, it will be time to consider them seriously when the 
subject acquires a practical importance. 
The disadvantages of the suggested plan are obvious at first 
sight; and therefore they may as well be enumerated at once, 
in advance. They are, 
Ist. A multiplication of parts, always to be avoided, if possible. 
n increase, to some extent, of weight. 
3d. The friction of an additional piston. 
The advantages are recognized only upon a more careful con- 
sideration. It will be seen, first, that the task of driving the aur, 
after condensation, into the reservoir, is, by this arrangement, dis- 
tributed equally over the entire stroke, instead of being confined, 
as before, to a little more than one-half of it. The total expendi- 
ture of force necessary in effecting this object, is, of course, neither 
increased nor diminished by the change. In the second. place, 
the air of the condensing cylinder acquires its maximum of den- 
sity only at the very termination of the stroke; and hence the 
maximum opposition which it offers to the motion of the piston 
is deferred to the latest possible moment, instead of growing Up 
rapidly before the completion of the half stroke, as in the present 
engines. In all this, there is-no gain of power, certainly, but 
there is a great gain in equality of action. It might seem, in- 
deed, at first thought, that there is a loss of power, since the elas- 
tic force of the air undergoing condensation constitutes a resistance, 
increased at mid-stroke, and will be more uniformly diffused overt 
the entire stroke, not, indeed, doing away with, but deferring, the 
great resistance at the close. ? 
All this can be made to appear only by the aid of mathematical 
formule. The expression for mean pressure, though constructed 
upon somewhat different data from the former, assames identically 
the Same shape in the end, and is 
P=15a (mn —1)+ 15am ( (1—n) hl Z—hl mn) 
