Means of cooling the air in factories, public buildings, and private dwellings have long been a subject of study and experiment. The methods proposed may be classed under two heads: (1) fresh air is introduced into a given space, or (2) the conduits by which the air is to pass are kept at a low temperature. Under the first head come the 4 systems proposed by Peclet: (a) mechanical action is employed, and the new air is compressed, and expands at the moment of escape; (6) evaporation is used by passing the air over moist surfaces; (c) it is refrigerated by circulation through pipes artificially cooled by ice, etc.; (d) it is passed through subterranean channels whose temperature, nearly always invariable, is actually equal to the mean temperature at the surface. The 4 systems may be described in detail.

(A) By Mechanical Action

Much was anticipated from the ingenious principle illustrated in lectures of natural science by the ordinary air match (briquet). It is known that when a gas is condensed, it becomes heated; on the other hand, when it is expanded, it cools. Suppose a certain amount of air has been condensed in a receiver, and that after having left the apparatus alone for some time, the temperature of its capacity, and the gas contained within it, should be equalised with that of the surrounding air; if the air be liberated from the receiver, it will be placed in the same condition as a gas expanding to resume its normal state of atmospheric pressure, and it would follow that it would become cooler in proportion to its degree of condensation. This experiment has led several inventors to applications for the production of cool air.

In order to render this means efficacious, it is necessary that the sensible heat developed in the compression should be destroyed by the refrigeration of the receiver containing the compressed air, otherwise the new air would be introduced at an equal temperature to that of the exterior atmosphere. Suppose this condition to be fulfilled. Let t' be the temperature of the compressed air, which will thus be that of the external temperature; P', its pressure; t and P, the temperature and pressure of the expanded air when it reaches the interior; 7, the relation of the capacity of the air at pressure and mean volume; we have, according to a formula of Poisson:

1+at' = (p) r-1'

1+at p r the quantity P is evidently equal to the atmospheric pressure; taking it as unity, and observing that t and t' never have anything except feeble values, this equation may be put into the following form: p' = {1+a (t-t)} r-1 , r and replacing a and r by their values: -

P = { 1 + 0.00867 (t' - t) } 3.375, for a lowering of temperature by 5°, t' - t = 5, we deduct P'= 1.06.

This compression of atmosphere 6/100, or, in mercury, of 46 mm., at first sight appears small. At the hospital Lari-boisiere, in a space containing 306 beds, the new air immediately after the action of the ventilator was compressed to 28 mm. of water, or about 2 mm. of mercury. Now this compression, although it is excessively slight, requires, from the great quantity of air to be compressed, a 12-horse steam-engine. If the same quantity of air had to be compressed by 46 mm. (mercury), it would require for an apparatus more useful than that of the Lariboisiere Hospital a steam-engine 15 to 18 times as powerful, or of 180 to 240 horse-power. A power of 3/5 to 3/5 of a horse-power would thus be required for each invalid. Without occupying oneself with other difficulties presented by this method, it may be seen, from this consideration alone, that it is impracticable.

In a mine in South Wales, an apparatus suggested by Piazzi Smith, despite the imperfection of the system, has met with some success. This apparatus is thus composed: - (1) of a pump for the compression of the air; (2) of a refrigerator, composed of a long tube or series of tubes, within which circulates a current of water, while the compressed air passes round, the disengaged heat of which is thus absorbed; (3) of a detention cylinder, where the air expands and cools to a temperature nearly proportionate to its primitive temperature: this cylinder was utilised to work the pump. The process is not only impracticable in consequence of the considerable mechanical labour required, the refrigeration produced by expansion not being nearly as high as that indicated by the calculation, in consequence of the heat developed in the air by the jets of compressed gas and the heat furnished by the casing.

The principle of this mode of refrigeration has been elegantly carried out by a beautiful experiment of Thilorier. It has been not less happily applied to the fabrication of ice on a large scale, which proves, as a general result, that the process is capable of inducing intense cold. Gome's machine has received much notice in America; Windhausen's machine in Germany; in England, the machines of Harrison and of Siebe; in France, the processes of Carre' and of Tellier are still more remarkable, and their reputation has long since been established.

(6) By Evaporation

By this means fresh air is easily cooled, and almost without expense; but its composition is altered by the introduction of a further degree of humidity, which is a fatal inconvenience in many cases.

One gramme of water in evaporation absorbs or renders latent a quantity of heat enough to cool 1 cub. metre of air by 2° C. Let p be the weight in grammes of the vapour of water contained in 1 cub. metre of external air, let π represent the weight of the water introduced by evaporation; by pv the weight of 1 cub. metre of internal air, and by y the resultant lower temperature, we obtain p +π = Pv y = 2π, and p + y = pv.

The evaporation and lowering of the temperature at an end, the air is saturated. The value of p and the exterior temperature t being known, this limit may be determined by tables giving the weight of the saturated vapours at different temperatures.