290.  The circumstance of so large a portion of the air(3*) driven into furnaces being not merely useless, but acting really as a cooling, instead of a heating, cause, added to so great a waste of mechanical power in condensing it, amounting, in fact, to four-fifths of the whole, clearly shews the defects of the present method, and the want of some better mode of exciting combustion on a large scale.  The following suggestions are thrown out as likely to lead to valuable results, even though they should prove ineffectual for their professed object.

291.  The great difficulty appears to be to separate the oxygen, which aids combustion, from the azote which impedes it.  If either of those gases becomes liquid at a lower pressure than the other, and if those pressures are within the limits of our present powers of compression, the object might be accomplished.

Let us assume, for example, that oxygen becomes liquid under a pressure of 200 atmospheres, whilst azote requires a pressure of 250.  Then if atmospheric air be condensed to the two hundredth part of its bulk, the oxygen will be found in a liquid state at the bottom of the vessel in which the condensation is effected, and the upper part of the vessel will contain only azote in the state of gas.  The oxygen, now liquefied, may be drawn off for the supply of the furnace; but as it ought when used, to have a very moderate degree of condensation, its expansive force may be previously employed in working a small engine.  The compressed azote also in the upper part of the vessel, though useless for combustion, may be employed as a source of power, and, by its expansion, work another engine.  By these means the mechanical force exerted in the original compression would all be restored, except that small part retained for forcing the pure oxygen into the furnace, and the much larger part lost in the friction of the apparatus.

292.  The principal difficulty to be apprehended in these operations is that of packing a working piston so as to bear the pressure of 200 or 300 atmospheres:  but this does not seem insurmountable.  It is possible also that the chemical combination of the two gases which constitute common air may be effected by such pressures:  if this should be the case, it might offer a new mode of manufacturing nitrous or nitric acids.  The result of such experiments might take another direction:  if the condensation were performed over liquids, it is possible that they might enter into new chemical combinations.  Thus, if air were highly condensed in a vessel containing water, the latter might unite with an additional dose of oxygen, (4*) which might afterwards be easily disengaged for the use of the furnace.

293.  A further cause of the uncertainty of the results of such an experiment arises from the possibility that azote may really contribute to the fusion of the mixed mass in the furnace, though its mode of operating is at present unknown.  An examination of the nature of the gases issuing from the chimneys of iron-foundries, might perhaps assist in clearing up this point; and, in fact, if such enquiries were also instituted upon the various products of all furnaces, we might expect the elucidation of many points in the economy of the metallurgic art.