Numerous attempts have been made to supersede steam as a motive power, with the view to avoid the loss of heat by its absorption in the steam in a latent state. Mercury vapour and spirit vapour have been tried, in the expectation that as they possess much less capacity for heat, an equal pressure might be obtained, with a diminished loss of heating power. Several gaseous agents have been applied to the same purpose, of which carbonic acid gas seemed to present the best prospect of success, because it becomes expanded with a comparatively small increase of temperature. None of these attempts to produce a motive power superior to steam have yet proved successful. They have all, after a short season of promise, dropped out of notice; and the only one that is still in the field, struggling for superiority, is the air engine.

The first known air engine was invented by Sir George Cayley, in 1803. In his engine the air was heated by passing directly through the hot coals of the furnace, which some engineers yet consider to be the best mode of expansion; but its operation did not answer expectations. Mr. D. Stirling, of Dundee, afterwards improved on Sir George Cayley's plan, and introduced a method of regaining the heat from the expanded air, after it had done its work in the cylinder, and of applying it to expand the air again. Engines on this construction have been for some years working in Scotland, and in 1850 Mr. Stirling took out a patent for an improvement in the arrangement, which is stated to have been very successful.

Though Sir George Cayley and Mr. Stirling were the first in the field as inventors of air engines, the name of Mr. Ericsson, an American, is more closely associated with the invention, as he has for many years been conducting experiments on a large scale, and has tried his "caloric engine" on land, and on a ship of large burthen, built for the purpose.

The principle and the working of Mr. Ericsson's caloric engine is nearly the same as Mr. Stirling's; but as it has been brought most prominently into notice, we shall direct attention more particularly to its construction and performances. Mr. Ericsson obtained a patent for his caloric engine in this country in 1833, and a subsequent patent for improvements on it was taken out in 1851. During those years, and to a late period, he was indefatigably working out the principle, and numerous highly favourable reports have from time to time been made of the results of the experiments; but the advantages to be derived from the air engine remain nevertheless very questionable.

The object attempted to be gained is to make the same heating power do its work again and again. Atmospheric air, after being expanded by passing over an extensive hot surface, exerts the force thus acquired to raise the piston of a large cylinder, and it is then attempted to abstract the heat as the air issues out, and to apply it to the expansion of a further quantity.

The practicability of this plan has undergone much discussion; its friends and foes being equally confident in their opinions. The former pronounce it to be one of the most valuable inventions of the age, being calculated to economize heat, and to give greatly additional impulse to navigation; whilst its opponents declare that the calculations are erroneous, the experiments fallacious, and that the expanded air consumes more heating power than steam.

In one of the favourable notices of Mr. Ericsson's engine in an American publication, it is thus described:—"Two caloric engines have been constructed in New York, one of 5-horse power, the other of 60. The latter has four cylinders; two of 6 feet diameter, placed side by side, surmounted by two of much smaller size. Within are pistons, so connected that those in the lower and upper cylinders move together. A fire is placed under the bottom of the large cylinders, called the working cylinders; those above are called the supply cylinders. As the piston in the supply cylinder moves down, valves at the top admit the air. As they rise, those valves close, and the air passes into a receiver and regenerator, where it is heated to about 450°, and entering the next working cylinder, it is further heated by a fire underneath to 485°. The air is thus expanded to double its volume; and supposing the supply cylinder to be half the size of the other, the air, when expanded, will completely fill the larger cylinder. As the area of the piston of the smaller cylinder will be only half that of the larger, and as the air will be of the same pressure in both, the total pressure on the piston of the large cylinder will be double that on the small one. This surplus furnishes the working power of the engine. After the air in the working cylinder has forced up the piston within it, a valve opens; and as the air passes out, the piston descends by gravity, and cold air rushes in, and fills the supply cylinder.

"The most striking feature is the regenerator. It is composed of wire net, placed together to a thickness of about 12 inches. The side of the regenerator, near the working cylinder, is heated to a high temperature. The air passes through it before entering the working cylinder, and becomes heated to 450°. The additional heat of 30° is communicated by the fire underneath to the large cylinder. The expanded air forces the cylinder upwards, valves open, and it passes from the cylinder, and again enters the regenerator. One side of the regenerator is kept cool by the air on its entering in the opposite direction at each stroke of the piston; consequently, as the air of the working cylinder passes out, the wires abstract its heat so effectually, that when it leaves the regenerator, it has been robbed of all except about 30°. In other words, as the air passes into the working cylinder, it gradually receives from the regenerator about 450° of heat; and as it passes out, this is returned to the wires, and it is thus used over and over again; the only purpose of the fires beneath the cylinders being to supply the 30° of heat which are lost by radiation and expansion.

"The regenerator in the 60-horse engine measures 26 inches in height and width. Each disc of wire composing it contains 676 superficial square inches, and the net has 10 meshes to the inch. Each superficial inch, therefore, contains 100 meshes, and there are 67,600 in each disc; and as 200 discs are employed, the regenerator contains 13,520,000 meshes, with an equal number of small spaces between the discs as there are meshes; therefore, the air is distributed into 27,000,000 of minute cells. The wire in each disc is 1,140 feet long; and the total length of wire in the regenerator is 41½ miles, or equal to the surface of four steam boilers, each 40 feet long and 4 feet diameter."

The accounts received from America of the great success that had attended the working of Mr. Ericsson's air engine, on the ship "Ericsson," attracted much attention in this country, and formed the subject of two evenings' discussion in the Institution of Civil Engineers. The most prevalent opinion was, that it is impossible to regain the heating power without corresponding loss of mechanical force or the addition of heat, and that there must have been some fallacy in the reports of the work done and of the quantity of fuel consumed.

It is, indeed, evident that nothing approaching the amount of heat said to have been recovered could be regained by passing through the regenerator; for as the apparatus becomes heated by the first portions of air passing through it, the temperature of the quantity that afterwards passed must at least be equal to that of the heated wires, and the last portions of air would consequently scarcely part with any caloric to the regenerator, previously heated to nearly its own temperature. Experience has since proved that the notion of regaining the heat by the regenerator was fallacious, for in the last improvements in Mr. Ericsson's engine, it is stated that the regenerator has been abandoned, and the plan has been adopted of cooling the air as it issues from the large cylinder, by passing it through tubes surrounded by cold water, and then using the same air over again.

One great practical inconvenience in the use of the air engine was the necessity of having enormously large cylinders to attain the required power, with the low amount of pressure that can be procured by the expansion of the air. The consequent friction increased the loss of power, and the difficulty of lubricating the pistons added to the practical objections to the air engine. To overcome these objections, the air in Mr. Stirling's engine is compressed before it is heated, by which means an equal amount of pressure is obtained on a smaller piston.

The air engine would in many respects possess advantages over the steam engine, if it could be worked economically. The space occupied by the boilers would be saved, and the danger of explosions would be avoided; for hot air does not scald, and the quantity at any time expanded would be too small to do much injury. A patent has since been obtained by Messrs. Napier and Rankine, for improvements in the air engine, which they anticipated would remove the objections that have been raised to the engines of Stirling and Ericsson. The heating surface has been greatly increased by employing tubes; and other defects in the former engines, to which their want of complete success is attributed, have been remedied, so that Mr. Rankine, in his description of the improvements at the meeting of the British Association at Liverpool, confidently anticipated to effect a great saving of heating power, combined with the other advantages of the air engine. He estimated the consumption of fuel by a theoretically perfect air engine on Mr. Stirling's principle at 0·37 lbs. per horse power per hour; whilst a theoretically perfect steam engine would consume 1·86 lbs. The actual average consumption of a steam engine is, however, 4 lbs. of fuel per horse power per hour, and the actual consumption of Stirling's engine is stated by Mr. Rankine to have been 2·20 lbs, and that of Ericsson's 2·80 lbs. It appears from this statement, therefore, that the air engines of Messrs. Stirling and Ericsson are superior in point of economy of fuel to steam engines; and if Mr. Rankine's anticipations of the superiority of his air engine be realized, it will effect still greater economy. In Messrs. Napier and Rankine's engine, the air is compressed before expansion, so that the size of the cylinders may be reduced to even smaller dimensions than the cylinders of steam engines of equal power.