ICE is the solid crystalline form which water assumes when exposed to a sufficiently low temperature. It is frequently precipitated from the air as hoarfrost, snow, or hail; and in the glaciers and snows of lofty mountain sys-tems or of regions of high latitude it exists on a gigantic scale, being especially characteristic of the seas and lands around the poles, which consequently have hitherto been practically inaccessible to man. Also in various parts of the world, especially in France and Italy, great quantities of ice form in caves, which, in virtue of their depth below the earth's surface, their height above the sea-level, or their exposure to suitable winds, or to two or more of these conditions in combination, are unaffected by ordinary climatic changes, so that the mean annual temperature is sufficiently low to ensure the permanency of the ice. The great ice supply for the island of Teneriffe is obtained from such a cave, which is 100 feet long, 30 feet broad, and from 10 to 15 feet high, and which is situated on the Peak some 10,000 feet above the sea-level. Accord-ing to the Rev. S. Browne (Brit. Ass. Report, 1864), such cave-ice is generally peculiar in its columnar appear-ance, and apparently less easy to melt than ordinary surface ice.

In the mutual transformations of water and ice, many remarkable physical phenomena occur. Thus, during the process of melting a block of ice or of freezing a quantity of water, no change of temperature can take place so long as there is a thorough mixture of water and ice. Consequently, the " freezing-point" or temperature at which water freezes is a temperature so readily determined that it is conveniently employed as one of the standard temperatures in the gradua-tion of ordinary thermometer scales, such as the centigrade, the Fahrenheit, and the Reaumur. The centigrade scale, whose zero corresponds to this freezing-point of water, is the temperature scale that is employed throughout this article. In the act of freezing, water, though its tempera-ture remains unchanged, undergoes a remarkable expansion or increase of bulk, so that ice at 0° C. is less dense than water—a fact demonstrated at once by its power of floating. "Ground-ice" or "anchor-ice," which forms in certain cir-cumstances at the bottom of streams, is only an apparent exception to this relation between the densities of water in its solid and liquid states, being retained there by the cohesion between it and the stones or rocks which compose the river's bed. When forcibly released from this contact with the bottom, the ice at once ascends to the surface. Ground-ice may thus be the lowest stratum of the once completely frozen mass of water, adhering to the bottom during the thawing and melting of the ice at the surface ; or it may even be formed under favourable conditions below briskly flowing water, probably by the action of eddies, which draw the surface water down through the warmer but denser liquid, and thus cool the stones and rocks at the bottom. As water then expands on freezing, so con-versely ice contracts on melting; and the ice-cold water thus formed continues to contract when heated until it has reached its point of maximum density. Joule, from a series of careful experiments, determined the temperature at which water attains its maximum density to be 39°T Fahr., or very nearly 4° C. Hence water contracts as its temperature rises from 0° C. to 4° C; but at higher temperatures it behaves like the great majority of other substances, expand-ing with rise of temperature. At no temperature, however, does water in the liquid state become less dense than ice: as the following table of relative densities shows:—
Density of ice at 0° C. - "9175
water at 0° C. = "99988
„ 4°C. = 1-00000
„ 10° 0. - -99976
„ 100° C. • -95866
Under the influence of heat, ice itself behaves as most solids do, contracting when cooled, expanding when heated.

According to Plucker, the coefficient of cubical dilatation at moderately low temperatures is -0001585. From a series of elaborate experiments, Person deduced '505 as the specific heat of ice, or about half that of water; in other words, the heat required to raise 1 lb of water 1° C. will raise 2 lb of ice through the same range of temperature or 1 ft) of ice through 2° C.
Though no rise of temperature accompanies the melting of ice, there is yet a definite quantity of heat absorbed, and a corresponding amount of work done—mainly in altering the physical condition of the substance. The heat which disappears is transformed into other and less evident forms of energy,—as, for example, the energy of translatory motion, which is the chief characteristic, according to the recognized molecular theory of matter, of the molecule in the liquid as compared with the molecule in the solid. The heat which is thus absorbed during the melting of unit mass of ice is called the latent heat of water, and its value in ordinary heat-units is 79'25, according to the determina-tion of Person. Hence as much heat is required to trans-form 1 So of ice at 0° C. into water at the same tempera-ture as would raise in temperature 1 tt> of water through a range of 79°'25 C, or 79'25 D> of water through a range of 1° C. The same amount of heat which is absorbed when ice becomes water is evolved when water becomes ice, so that the melting of ice is accompanied by the abstraction of heat from surrounding objects, that is, by a cooling effect; and the freezing of water by a heating effect. These thermal effects are generally masked by the pro-cesses whereby the change of state is effected; but the cooling which accompanies the melting of ice may be observed when pressure is used as the agent for accom-plishing the change. That ice can be so melted by increase of pressure was first pointed out by Professor James Thomson (now of Glasgow) in a paper published in the Transactions of the Royal Society of Edinburgh for 1849 ; previous to that time the temperature of melting ice was believed to be absolutely constant under all conditions. Thomson showed that, since water expands on freezing, the laws of thermodynamics require that its freezing-point must be lowered by increase of pressure; and, by an ap-plication of Carnot's principle, he calculated that for every additional atmosphere of pressure the freezing-point of water was lowered by -0075 of a degree centigrade. This remarkable result was soon after verified, even to its numerical details, by his brother, Sir William Thomson (Proceedings of the Royal Society of Edinburgh, 1850). The Thomsons and Helmholtz have since then successfully ap-plied this behaviour of ice under pressure to the explanation of many curious properties of the substance. When two blocks of ice at 0° C. are pressed together or even simply laid in contact, they gradually unite along their touching surfaces till they form one block. This regelation, as it is called, is due to the increased pressure at the various points of contact causing the ice there to melt and cool. The water so formed tends to escape, thus relieving the pres-sure for an instant, refreezing, and returning to the original temperature. This succession of melting and freezing, with their accompanying thermal effects, goes on until the two blocks are cemented into one. Thus it is that a snowball is formed ; and in virtue of the same succession of pheno-mena does the glacier mould itself to its rocky bed and flow down the valley, behaving in many respects like a viscous fluid.

Ice forms over fresh water if the temperature of the air has been for a sufficient time at or below the freezing-point; but not until the whole mass of water has been cooled down to its point of maximum density, so that the subsequent cooling of the surface can give rise to no convection currents, is the freezing possible. Sea-water, in the most favourable circumstances, does not freeze till its temperature is reduced to about - 2° C.; and the ice, when formed, is found to have rejected four-fifths of the salt which was Originally present. In the upper provinces of India, water is made to freeze during cold clear nights by leaving it overnight in porous vessels, or in bottles which are enwrapped in moistened cloth. The water then freezes in virtue of the cold produced by its own evaporation or by the drying of the moistened wrapper. In Bengal the natives resort to a still more elaborate forcing of the conditions. Shallow pits are dug about 2 feet deep and filled three-quarters full with dry straw, on which are set flat porous pans containing the water to be frozen. Exposed overnight to a cool dry gentle wind from the north-west, the water evaporates at the expense of its own heat, and the consequent cooling takes place with sufficient rapidity to overbalance the slow influx of heat from above through the cooled dense air or from below through the badly conducting straw.

The growing demand for ice for domestic, medicinal, and ioe. other purposes has led, not only to the development of a machines, regularly organized ice-trade, but also to the invention of machines for the manufacture of ice in countries which do not possess a sufficient home supply. The various types of machines which have been or are in use call for a brief description. Freezing-mixtures, such as the familiar snow and salt or the mixture of sulphate or phosphate of sodium and dilute nitric acid, may be dismissed with a word, since they are restricted in use to the production of intense cold for a brief period of time, and are incapable of economic application to the formation of large quantities of ice.

All ice-machines which have proved of practical utility may be grouped under two great classes:—those which utilize the lowering of temperature that accompanies the rapid expansion of a compressed gas, and those which make use of the like thermal effect that results from the vola-tilization of some liquid. In machines of the first type, the gas usually employed is atmospheric air, which is first compressed to three or four atmospheres, and kept cool by circulating water or by other suitable means. It is then allowed to expand, and the heat necessarily absorbed during the expansion is drawn either from the water to be frozen or from a solution of brine which does not freeze at the ordinary freezing temperature, and thus becomes, so to speak, a vehicle for the cold. In 1849 Gorrie constructed such a machine, which, however, was unsatisfactory in its action, probably because the compressed air was not sufficiently cooled and dried. More efficient in their action were Kirk's machine (patented in 1863), and Windhausen's (1870), one of which at the Vienna exhibition produced 30 cwts. of ice per hour, at the cost of Is. per cwt. The mode of action of Windhausen's is as follows. A piston works to and fro in a cylinder, compressing the air in the one end and allowing it to expand in the other. The compressed and therefore heated air forces its way through a valve to the cooling chambers, from which it is led towards the other end of the cylinder. Here the inlet valve is so arranged that it closes at a certain position of the receding piston, thus permitting what air has entered to expand and cool. At the return stroke this cooled air is forced out through easily opening valves,-—part going to cool the chambers into which the heated compressed air enters from the cylinder, and part passing to the refrigerator, from which after serving its purpose it is pushed on by the fresh supply of cooled air to the compressing end of the piston chamber. Such machines, to work economically, require large cylinders, tight-fitting pistons working with little friction, and perfect regulation in the motions of the various parts—conditions so difficult to fulfil that refrigeration by means of compressed air may be regarded as a practical failure. The machines constructed by the Bell-Coleman Mechanical Refrigeration

Company (Glasgow) utilize as the cooling agent a mixture of certain hydrocarbon gases which are obtained from the distillation of carbonaceous shale. The gas is compressed to a pressure of about 8 atmospheres, and, after being cooled by expansion, is carried off and consumed as fuel. These machines are not specially intended for the production of ice; but, as refrigerators, they are successfully employed for preserving meat on board ship.

Among machines of the second group there is a great variety of construction, because of the great differences which exist in the properties of the liquids used. Thus water, sulphuric ether, bisulphide of carbon, ammonia, methylic ether, sulphurous acid, and other substances have been employed as refrigerating agents. In all cases, it is the so-called latent heat of vaporization that is utilized; and did the efficiency of the method depend only on this, water would undoubtedly be the best material on account of the great latent heat of its vapour. But as important from a practical point of view are the vapour pressures that come into play throughout the range of temperature employed. Thus at 10° C. the pressure of water vapour is so small, only -012 of an atmosphere (and at lower tem-peratures of course it is still smaller), that, to make the evaporation of water an efficient means of refrigeration, the process must be conducted under a very much diminished pressure. As early as 1755, Dr Cullen managed to freeze water by its own evaporation in a vacuum; but this method, though greatly developed by Nairne, Leslie, and Vallance, can be applied to the production of ice in small quantities only.





The same objection applies, of course, to sulphuric ether, bisulphide of carbon, or any substance which boils under ordinary atmospheric pressure at a temperature above that of the air. Ether boils at 340,8 C, and bisulphide of carbon at 46°-2 C.; and their vapour pressures at 10° C. are respectively -377 and '267 of an atmosphere. They thus volatilize much more readily than water, and require a comparatively slight vacuum to render their evaporation sufficiently rapid for refrigerating purposes. In the ether machine, which may be taken as a type, the ether, on being vaporized in the refrigerator under a partial vacuum, is drawn over and compressed to the liquid state in the con-denser, which is kept cool by circulating water. From the condenser it is then led back to the refrigerator, to be re-evaporated. Perkins's machine (1834), Twining's patent of 1850, Harrison's machine (1857), Siebe's machine (1862), and Siddeley and Mackay's apparatus are ether-machines ; and all except the first, which is hardly adapted for exten-sive freezing, surround the refrigerator with brine, which when cooled flows easily around and between the cases containing the water to be frozen. Van derWeyde (1869) substituted naphtha, gasolin, or chimogene for the ether; and in Johnston and Whitelaw's machine bisulphide of carbon is used somewhat similarly. The great difficulty in machines of the ether type is to prevent leakage, so as to keep the partial vacuum really efficient; and moreover ether, which is in most respects superior to all the other substances employed, has an awkward tendency, under the influence of frequent condensations and rarefactions, to transform itself into less volatile isomers.

The great characteristic of ice-machines which employ ammonia, methylic ether, or sulphurous acid, as compared with those of the ether type, is that they work at increased instead of diminished pressures, since these substances are gaseous at ordinary temperatures and pressures, and require for their liquefaction either the production of a low tem-perature or the application of a high pressure. For facility of reference the boiling points and vapour pressures at three different temperatures for these substances are given in the following table.

== TABLE ==

The best known of the ammonia machines is Carre's (1859), the principle and construction of which are remark-ably simple. Two strong metal vessels, the boiler and refrigerator, are connected above by a tube. In the boiler a saturated solution of ammonia is raised to 130°-150° C. The ammonia is driven over under high pressure into the refrigerator, round which cold water circulates, and in which the ammonia is condensed to a liquid. The boiler is then placed in cold water, and as its temperature falls the pressure in the apparatus is relieved and the liquid ammonia in the refrigerator vaporizes rapidly, thereby producing intense cold, and redissolves in the boiler. The temperature to which the boiler must be raised at first is determined by the condition that the pressure in the boiler must correspond to the pressure of the ammonia vapour at the temperature of the condenser. Now the pressure of ammonia vapour increases from 8J atmospheres at 20° C. to 11£ at 30° C. ; and this higher pressure is extremely difficult to keep up in such an apparatus as Carre's, because of inevitable leakage. In warm countries, accordingly, the ammonia-machine is practically useless because of the high pressures required; andin temperate climates, where natural ice can be stored throughout summer, an ice-machine is not in such great demand. One great drawback to the efficient working of Carre's machine is the difficulty of keeping the refrigerating liquid free of water—only 75 per cent, of it being ammonia. To remedy this defect Reece invented his machine (1869). The essential part of this ingenious apparatus is an upright cylinder in which a descending current of strong ammonia solution, drawn originally from the boiler, is met by an ascending current of steam. The ammonia is thus separated from the water, and is driven off into a rectifier, from which, after being freed from any small quantity of water it may have carried along with it, it passes into a condenser where it is kept liquid by its own pressure. It is then allowed to collect in the refrigerator, where at the required moment the pres-sure is relieved, permitting the ammonia to vaporize and escape to a separate chamber to be redissolved. Brine flowing through a coiled tube within the refrigerator is used as the vehicle for the cold produced; or even mere water may suffice if the object is simply to get a diminished temperature without freezing. Linde's ice-making machine, some twenty-two of which were in operation at the Diissel-dorf Exhibition of 1880, is the latest form of ammonia machine ; and its inventor claims for it superiority over all others as an economical refrigerator. The danger of explosion, one of the great disadvantages of ammonia, is obviated by carrying the liquefied gas through narrow iron tubes and by employing only a small quantity of the substance at one time. Blocks of ice are formed between the spokes of a revolving drum, which, cooled internally by the evaporating liquid, dips into a tank of water. Methylic ether is in some respects better than ammonia, having a higher boiling point, and requiring smaller pres-sures, without the necessity of heating. In Tellier's machine (described in the Annates de Chimie et cle Physique for 1874), which is specially suitable for use on board ship, the methylic ether evaporates in a closed metallic vessel, the sides of which are in immediate contact with the water to be frozen or chilled.

Sulphurous acid, first successfully employed as a refrigerating agent by Pictet of Geneva (1876), and thereafter applied by Gamgee to the formation of his glaciarium or artificial skating rink, is in many respects far superior to any other known refrigerator. Thus it is more easily liquefied than ammonia and methylic ether, exerting a vapour pressure of only 4|- atmospheres at 30° C. : it has no chemical action upon metals or fats; it is incombustible; it is obtainable at small expense; and it has, besides, good lubricating properties:—in short, it seems to possess all the essentials of an efficient and economical refrigerator. In Pictet's machine, the liquid sulphurous acid passes under pressure from the condenser to the refrigerator, where on the pressure being relieved it vaporizes, cooling to -7° C. a current of brine which then flows round the tanks containing the water to be frozen. The sulphurous acid gas in the refrigerator is drawn over by an aspirating force-pump and recondensed in the condenser, which is kept cool by an ample supply of cold water. By a special modification of the sulphurous acid machine, Pictet obtained as low a temperature as -73° C.; under this low tempera-ture he then compressed carbonic acid gas to a liquid, by the evaporation of which he produced such intense cold as to enable him to liquefy the so-called permanent gases under a pressure of several hundreds of atmospheres (BibliotJieque Universelle, 1878). Gamgee uses as his congealing liquid a solution of 4 parts of glycerin in 6 parts of water, which is conveyed in pipes beneath the water-surface to be frozen.

Machines which are capable of freezing water may in certain circumstances be much more efficiently employed to produce cooling without freezing. For instance, in curing-houses, breweries, sugar refineries, provision stores in hot climates, and in ships engaged in the transport of meat, • where it is of importance to have the temperature moderately cool, it is usually by no means necessary to obtain ice. In many such cases, indeed, the production of ice would be a mere waste of labour. In tropical and subtropical climates refrigeration is of high importance from a sanitary point of view ; and there seems little doubt that if a simple, economical, and thoroughly efficient means of cooling were discovered, houses would be cooled in warm weather with the same care and regularity with which they are when necessary heated. At present, however, the manufacture of ice and the artificial production of cold are arts still in their infancy, which have a powerful rival in the extensive and increasing ice-trade that has sprung up within the last half century. Ice trade. The idea of trading in ice first occurred to a Boston merchant, named Tudor, who in 1805 shipped ice to Martinique. In 1833 American ice began to be imported into Calcutta, where it was sold for 3d. per pound—exactly half the price of the Bengal manufactured ice. In America, which was for long the great ice-exporting country of the world, supplying especially the West Indies, India, and China, the cutting and storing of ice form an important industry during the winter months. When the ice is suffi-ciently thick, 9 to 12 inches for home consumption, 20 inches for exportation, the surface is scraped free of all porous ice, and is marked out into squares of 5 feet each way. Along these lines the ice is grooved to a depth of 3 inches by means of a plough. An instrument like a harrow is drawn over the grooves so as to deepen them; and, after the surface has been divided into smaller squares, the ice is cut up into blocks by means of handsaws. The blocks are then removed to large double-walled storehouses, many of which are capable of containing thousands of tons of ice. It is estimated that, in America, 2,000,000 tons of ice are cut and stored annually by companies supplying New York and the middle States. New York city alone consumes as much as 500,000 tons per annum. A considerable quantity of ice from Wenham Lake near Boston was at one time imported into Britain, but now the whole supply comes from Drobak near Christiania in Norway. The N orwegian ice is remarkably solid and pure, and is superior in its staying power to English ice or to manufactured ice. The total quantity
imported into the United Kingdom may be estimated
roughly at 150,000 tons per annum, of which the greater
part is consumed in London, where it is retailed at from
2s. 6d. to 3s. 6d. per cwt. At present Norway is undoubtedly
the great ice-store for the Old World ; and quite recently
(1880) Norwegian ice has been sold in the United States
more cheaply than native ice. The transport on board ship
offers practically no difficulty, since, as long as the hold is
kept dry and cool, there is very little loss, and in the lading
no special care need be taken. For the storing in houses,
see ICE-HOUSE. (c. G. K.)