Duvoir's apparatus thus acted well, but did not fulfil all the required conditions, owing to the causes already mentioned. Peclet subsequently improved upon this system by causing the air to be cooled to pass into an apparatus formed of a number of little moistened tubes not in the interior, as with Duvoir's plan, but at the exterior. Each tube was surrounded by a sheet kept constantly wet, and the evaporation was stimulated by the action of a strong ventilator. Still the necessary amount of moisture was not introduced.

Among apparatus of this kind, Baum-bauer's plan is particularly to be noticed. Its simplicity attracted much attention.

It was composed of a cylinder traversed by a minutely divided current of water, a current of air proceeding in the same direction, and then flowing into a chimney draught in which a gas-jet was burning-; this cylinder was enclosed by another open at both ends, containing transverse metallic gauze, by which the cold was to pass to the air descending between the cylinders by the effect of the lower temperature.

(C) By Ice

Ice as a refrigerant may be placed either within or without the shafts bringing in fresh air. In the first case, generally preferred by the inventors, it melts, and afterwards evaporates in the fresh air. The cold resulting from the fusion and warming of the water produced not being more than 1/6of that due to evaporation, it follows that the amount of moisture introduced into the air is about 1/7, or nearly as much as that of evaporation alone. The refrigeration obtained has the same inconveniences as before, and is equally ineffectual. When the ice surrounds the pipe without being in contact with the fresh air, it gives, by fusion and the heating of the water thus produced, about 100 calories of cold per kiln. (2 .2 lb.). In a hospital, for instance, a specific ventilation of 100 cub. metres with a temperature of 5° C, would require per bed and hour 100 X 0.24 X 5 = 120 calories of cold, to be raised by reason of loss to 200 calories, or 2 Mo. (4.4 lb.) of ice. In warm countries, this refrigeration would be wanted, generally speaking, at least fur 5 hours a day during 100 days; therefore it would need, per bed per year, the employment of about 1 ton of ice.

In temperate climates it would always be possible to obtain that quantity of ice, despite its importance; it would always cause a high amount of expense in consequence of the ice-houses to be constructed, and even the harvesting of the ice during winter. But in hot climates, precisely when the consumption would be the greatest, it would be almost impossible to obtain it without enormous expense. In point of fact, this method of refrigeration must be considered impracticable when ice is not very cheap, and cold cannot be produced as inexpensively as heat.

The apparatus shown in Fig. 12 and based upon this principle, may be used in some cases; its simplicity has gained it a certain degree of support. The air conduit c passes through a casing a 6, formed of a double lining. The interior space d surrounding the air conduit contains ice. The next space e is filled with tan, or, still better, down or wool, which acts as a non-conductor of cold. A tap / lets off the water formed by the melting of the ice into a receiver g. The air conduit c is fitted with mechanical fly-wings A, which increase the contact of the air with the sides cooled by the ice. These metallic fly-wings are fixed to a vertical axis, and in successive rows, but in different planes, which multiplies the surface over which the air has to pass. A mixture of ice and salt is better than pure ice, should ice be dear, or the casing a b be small in comparison with the degree of refrigeration required. This contrivance, which manifests ingenious details of construction, may have been applied with success, but it is far from being sufficiently cheap.

It has been calculated that the cost of cooling the wards of the Lariboi-siere Hospital in summer by this apparatus would equal the cost of warming them in winter, taking the minimum price of 1/4d. per lb. of ice.

Refrigerating mixtures in point of economy, so far as they have been realised, have turned out altogether illusory. True a considerable degree of cold has been attained, but the expense is too high. An apparatus of this kind would only be accessible to amateurs who could afford to pay a fancy price, but may be adopted in some particular instances. If it were possible to obtain inexpensive chemical mixtures for producing great degrees of cold, this model might be advantageously employed.

(D) By Underground Channels

The temperature of the soil to a depth of 6, 10, or 20 yd., according to its nature and the local position, is on an annual average nearly the same as the temperature of the air at the surface, but the variations to which it is subject are much less, and decrease rapidly with the depth; at 3 or 4 yd, these variations are not more than 2° or 3° C, and the temperature may practically be regarded as constant. By causing currents of air to pass through vaults built at that depth, they will be perceptibly cooled in summer if they are of any length. As such conduits are almost necessarily a part of any ventilation by injection, the cooling of air is effected by this system without any additional expense - that is to say, naturally. In two experiments mode at the Necker Hospital the following results were obtained: -


Temperature of the Hall.

Temperature of the Fresh Air at Its Introduction.


Temperature in the shade

Difference, or

Hefrigeration obtained.





22.3 0


20. 6°












The subterraneous channel is an aqueduct of masonry, built under the cellars at a depth of about 4 yd.; the distances traversed in this conduit are on the average 15 yd. for the first inlet orifice, 35 yd. for the second, and 55 yd. for the third. On the first 15 yd. of the conduit the cooling is 2-9° C in the first experiment, 3.4° C. in the second experiment: that is to say, 3.l0

Fig. 13.

D By Underground Channels 40014