The cream, being suitably prepared, is placed in the can, and the tub is filled with ice and salt. The scrapers are inserted in place, and the lid is attached. In the side of the tub is cut a recess, through which a pinion on the vertical shaft c enters, and engages a circular rack on the can. When these parts are brought into gear, the tub is held in place by the pin d. The vertical shaft c is now rotated by bevel gear connected with the main horizontal shaft, which last is turned by the crank shown. The can is thus revolved until the cream becomes quite thick. The paddle, which is secured to the disc on the left, is now thrown into operation by the lever e, on moving which, gearing connected with said disc is engaged with gearing on the main shaft. The oscillations of the paddle are continued until the cream becomes stiff and hard. The can is open during the entire operation, and hence its contents are always under the eye of the operator. The inventor (C. L. Dexter, Philadelphia) states that a boy of 14 years can easily make 30 qt. of ice cream at a time without assistance.

The cans may hold 12 to 40 qt., and there is no churning of the cream into butter by this apparatus, which may be operated by steam, if desired.

Fig 18.

Cooling Air Part 11 40023

2. By Evaporation of Liquids. - The evaporation and recondensation of a liquid may be utilised in 2 ways for the production of cold. Typical of the first method is the well-known laboratory ammonia apparatus of Carre. This consists of two vessels, which may be called a and 6, capable of resisting considerable pressure, and joined by a pipe. The first step in the process is the preparation of liquid ammonia. Vessel a contains a solution of ammonia in water, and is artificially heated; b is kept cool, and the air is effectually excluded, if there be any leakage, by a water jacket, and in it the ammonia condenses under the pressure caused by the heat of a. When sufficient ammonia is condensed a is transferred to a vessel of cold water; the ammonia vapour is rapidly absorbed by the cold water within the vessel a. Under the reduced pressure in 6, the ammonia boils, absorbing much heat and producing considerable cold. The second method is the exact converse of the condensing steam-engine. In the steam-engine, water is converted into steam in a vessel at a high temperature; it expands in the cylinder of the engine, losing a portion of its heat, which becomes useful mechanical work; it then passes to a second vessel maintained at a lower temperature and pressure, in which it is condensed, giving up the balance of the heat it absorbed in the boiler.

Imagine the water replaced by ether, the temperature of boiler and condenser appropriately lower, and the direction of rotation of the engine reversed by the application of external power, so that, in fact, it becomes a pump. The ether evaporates in the condenser, absorbing heat and causing cold; the ether steam passes to the pump, where it is compressed, converting mechanical power into heat, and, under the pressure exerted, it is condensed and forced at a higher temperature and pressure into the vessel corresponding with the steam boiler, where it gives up its heat as may be arranged. Upon the choice of the liquid used will depend the pressure and temperatures in the two vessels or chambers. At 32° F. (0° C), the tension of water vapour is 4.6 mm. mercury; of ether, 183.3; of sulphurous anhydride, 1165•1; of ammonia, 3162•9. To produce 1 litre (1 3/4 pint) of water vapour at 32° F. (0 C), requires 0.0029 units of heat, the unit being the heat required to raise 1 kilo. (2.2 lb.) of water 1° C.; to produce 1 litre of ether vapour at the same temperature requires 0.073 units. At 32° F. (0° C), each stroke of a pump will abstract by ether vapour nearly 30 times as much heat as by water vapour.

A glance at these figures shows an obvious advantage in using liquids having low boiling-points; a pump of small capacity will remove a large quantity of heat, but all such substances are too costly to be wasted, and are offensive if any of the vapour escapes. Water presents obvious advantages, in the fact that we need not care what becomes of the vapour when condensed. But the use of water demands the power to produce and maintain a near approach to a perfect vacuum; the barometric pressure must be reduced from the normal of about 760 mm. of mercury, to less than 4, and for every unit of heat removed, at least 350 litres of vapour must be withdrawn and condensed. Water may be used in either of the methods already mentioned. It may be used in a manner exactly corresponding with Carre's ammonia apparatus, the water taking the place of the ammonia, and some hygroscopic substance, such as sulphuric acid, taking the place of the water, the pressures of course being always very much lower. Or if we can find a sufficiently perfect pump to produce and maintain a vacuum of less than 4 mm. of mercury, we may realise the precise reversal of the condensing steam-engine; but to produce any quantity of ice, the pump must not only be very perfect, but have a great capacity.

A combination of the two methods answers best. (Dr. Hopkinson.)

In selecting bodies for abstracting and absorbing heat with the object of producing refrigeration on an extensive scale, several points require to be taken into consideration. (1) The first is the amount of latent heat absorbed by 1 lb. of the body in changing its state, being 966*1 heat units for watery vapours, 900 for gaseous ammonia, 364.3 for alcohol vapour, 162*8 for ether vapour.

The amount of artificial cold produced will be in inverse ratio: thus the formation of 1 ton of ice will necessitate the evaporisation of about 395$ lb. of water, 424f lb. of liquid ammonia, 1049 1/4 lb. of alcohol, or 2348 lb. of ether. (2) The next important consideration is the degree of facility with which the bodies are vaporised, and the range of temperature within which the evaporisation can be readily accomplished, or, in other words, the boiling-point of the body and the tension of its vapour. It is sought to obtain a body having the former as low as is convenient, combined with the latter also moderately low. Many practical difficulties have been encountered through selecting bodies possessing the former quality, without much regard to the latter. Thus, at a temperature of 75° F. (24° C), which is often exceeded in town waters in warm countries, the tension of liquid ammonia will be 150-160 lb. a sq. in.; methyl chloride, about 80 lb.; methylic ether, 78 lb.; sulphur dioxide (sulphurous anhydride or oxide), 60 lb.