PRESSING OUT THE OIL.
As soon as a "heater" is ready to be emptied, the meal is taken out and put into six hair sacks, corresponding to the six pans in the press. There are six hair mats about one foot wide and six long, one side of each being coated with leather. The hair mat is about an inch thick. Now the hair sack, containing ten and a half to eleven pounds of heated steaming meal, is placed on one end of the mat, and the meal distributed so as to make a pad or cushion of uniform thickness. The pad of meal is not quite three feet long, a foot wide, and three inches thick, and the hair mat is folded over, sandwiching the pad and leaving the leather coating of the pad outside. In this form the six loads are put into the six pans, and by means of a powerful hydraulic press the pans are slowly pressed together. The oil begins trickling out at the side, slowly at first, and then suddenly it begins running freely. The pressure on the "loads" is 350 tons. After being pressed about five minutes, the pressure is eased off and the "loads" taken out. What had been a mushy pad three inches thick is a hard, compact cake about three-quarters of an inch thick, and the sack is literally glued to the cake. The crude oil has a reddish muddy color as it runs into the tanks.
To one side were lying great heaps of sacks of yellowish meal—the cakes which have been broken and ground up into meal. That, as explained above, forms the body of all fertilizers. The following is a summary of the work for the eight months' season at the Atlanta mills:
Fifteen thousand tons of seed used give:
Fifteen million pounds of hull.
Ten million three hundred and thirty-one thousand two hundred and fifty pounds of meal.
Four million six hundred and sixty-eight thousand seven hundred and fifty pounds of oil.
Three hundred thousand pounds of lint cotton.
The meal is worth at the rate of $6 for 700 pounds, or $88,603.58.
The oil is worth thirty cents a gallon, or seven and a half pounds, or $186,750.
The lint is worth $18,000, making a total of $293,353, and that doesn't include the 15,000,000 pounds of hull.—Atlanta Constitution.
MANUFACTURE OF PHOTOGRAPHIC SENSITIVE PLATES.
Quite recently Messrs. Marion & Company, London, began on their own account to manufacture sensitive photographic plates by machinery, and the operations are exceedingly delicate, for a single minute air bubble or speck of dust on a plate may mar the perfection of a picture. Their works for the purpose at Southgate were erected in the summer of 1886, and were designed throughout by Mr. Alexander Cowan.
Buildings of this kind have to be specially constructed, because some of the operations have to be carried on in the absence of daylight, and in that kind of non-actinic illumination which does not act upon the particular description of sensitive photographic compound manipulated. Glass and other materials have therefore to pass from light to dark rooms through double doors or double sliding cupboards made for the purpose, and the workshops have to be so placed in relation to each other that the amount of lifting and the distance of carriage of material shall be reduced to a minimum. Moreover, the final drying of sensitive photographic plates takes place in absolute darkness. Fig. 1 is a ground plan of the chief portion of the works. In this cut, A is the manager's private office, B the counting house, C the manager's laboratory, and D his dark room for private experiment, which can thus be conducted without interfering with the regular work of the establishment. E is the carpenter's shop and packing room, F the albumen preparation room, G the engine room, with its two doors; the position of the engine is marked at H. The main building is entered through the door, K; the passage, L, is used for the storage of glass, and has openings in the wall on one side to permit the passage of glass into the cleaning room, M; this room is illuminated by daylight. The plates, after being cleaned, pass into the coating rooms, N and O, into which daylight is never admitted; the coating machine is in the room, N, and three hand coating tables in the room, O; both these rooms are illuminated by non-actinic light.
The walls of N and O are of brick, to keep these interior rooms as cool as possible in hot weather, for the making of photographic plates is more difficult in summer time, because the high temperature tends to prevent the rapid setting of the gelatine emulsion upon them. At the end of these rooms and communicating with both is the lift, P, by which the coated plates are carried to the drying rooms above, which there cover the entire area of the main building; they consist of two rooms measuring 60 ft. by 30 ft., and are each 30 ft. high at the highest part in the center of the building; these rooms are necessarily kept in absolute darkness, except while the plates are being stored therein or removed therefrom, and on such occasions non-actinic light is used. After the plates are dry, they come down the lift, Q, into the cutting and packing room, R, which is illuminated by non-actinic light. In the drying rooms the batches of plates are placed one after the other on tram lines at one end of the room, and are gradually pushed to the other end of the building, so that the first batches coated are the first to be ready to be taken off when dry, and to be sent down the lift, Q. The plates in R, when sufficiently packed to be safe from the action of daylight, are passed through specially constructed openings into the outside packing room, S, where they are labeled. The chemicals are kept in the room, T, where they are weighed and measured ready for the making of the photographic emulsion in the room, U. The next room, V, is for washing small experimental batches of emulsion, and W is the large washing room. The emulsion is then taken into the passage, X, communicating with the two coating rooms. A centrifugal machine in the room, Y, is used for extracting silver residues from waste materials, also for freeing the emulsion from all soluble salts. Washing and cleaning in general go on in the room, Z.
The glass for machine coating is cut to standard sizes at the starting, instead of being coated in large sheets and cut afterward—a practice somewhat common in this industry. The disadvantage of the ordinary plan is that minute fragments of glass are liable to settle upon the sensitive film and to cause spots and scratches during the packing operations; any defect of this kind renders a plate worthless to the photographer. When any breakages take place in the cutting, it is best that they should occur at the outset, and not after the plate has been coated with emulsion. The cutting when necessary is effected by the aid of a "cutting board," Fig. 2, invented by Mr. Cowan, and now largely in use in the photographic world. This appliance is used to divide into two equal parts, with absolute exactness, any plate within its capacity, and it is especially useful in dimly lighted rooms. It consists of four rods pivoted together at the corners and swinging on two centers, so that in the first position it is truly square, and in other positions of rhomboid form, the two outer bars approaching each other like those of a parallel ruler. The hinge flap comes down on the exact center of the plate, minus the thickness of the block holding the diamond. By this appliance plates can be cut in either direction. Fig. 3 represents a similar arrangement for cutting a number of very small plates out of one large one; in this the hinge flap is made in the form of a gridiron, and the bars are spaced at accurate distances, according to the size of the plate to be cut, so that a plate 10 in. square, receiving four cuts in each direction, will be divided into twenty-five small plates.
Before being cleaned all sharp edges are roughly taken off those plates intended for machine coating by girls, who rub the edges and corners of the plates upon a stone; the plates are then cleaned by any suitable method in use among photographers. The plates, now ready for the coating room, have to be warmed to the temperature of the emulsion, say from 80 deg. F. to 100 deg. F., before they pass to the coating machine, the inventor of which, Mr. Cadett, having come to the conclusion that, if the plates are not of the proper temperature, the coating given will be uneven over various parts of the surface. The plate-warming machine is represented in Fig. 4; it was designed by Mr. A. Cowan, and made by his son, Mr. A. R. Cowan. It consists of a trough 7 ft. long by 3 in. deep, forming a flat tank, through which hot water passes by means of the circulating system shown in the engraving. To facilitate the traveling of the glass plates without friction the top of the tank is a sheet of plate glass bedded on a sand bath. An assistant at one end places the glasses one after the other on the warm glass slab, and by means of a movable slide pushes them one at a time under the cover, which cover is represented raised in the engraving to show the interior of the machine. After having put one glass plate on the slide, another cannot be added until the man in the dark room at the other end of the slide has taken off the farthest warmed plate, because the slide has a reciprocating movement. This heating apparatus is built at right angles to the coating machine in the next room, in order to be conveniently placed in the present building; but it is intended in future to use it as a part of the coating machine itself, and to drive it at the same speed and with the same gearing, so that the cold plates will be put on by hand at one end, get warmed as they pass into the dark room, at the other end of which they will be delivered by the machine in coated condition. Underneath the heating table is a copper boiler, with its Bunsen's burner of three concentric rings to get up the temperature quickly and to give the power of keeping the water under the heating slab at a definite temperature, as indicated by a thermometer. The cold water tank of the system is represented against the wall in the cut.
Fig. 5 represents the hot water circulating system outside the coating rooms for keeping the gelatine emulsions in these dimly lighted regions at a given temperature, without liberating the products of combustion where the emulsion is manipulated. The temperature is regulated automatically. It will be noticed where the pipes enter the two coating rooms, and Fig. 6 shows the copper inside one of them heated by the apparatus just described. The emulsion vessel in the copper is surrounded by warm water, and the copper itself is jacketed and connected with the hot water pipes, so forming part of the circulating system.
Fig. 7 is a general view of the coating machine recently invented by Mr. Cadett, of the Greville Works, Ashtead, Surrey. The plates warmed in the light room, as already described, are delivered near the end of the coating table, where they are picked off a gridiron-like platform, represented on the right hand side of the cut, and are placed by an assistant one by one upon the parallel gauges shown at the beginning of the machine proper; they are then carried on endless cords under the coating trough described farther on. After they have been coated they are carried onward upon a series of four broad endless bands of absorbent cotton—Turkish toweling answers well—and this cotton is kept constantly soaked with cold water, which flows over sheets of accurately leveled plate glass below and in contact with the toweling; the backs of the plates being thus kept in contact with fresh cold water, the emulsion upon them is soon cooled down and is firmly set by the time the plates have reached the end of the series of four wet tables. They are then received upon one over which dry toweling travels, which absorbs most of the moisture which may be clinging to the backs of the plates; very little wet comes off the backs, so that during a day's work it is not necessary to adopt special means to redry this last endless band. What are technically known as "whole plates," which are 8½ in. by 6½ in., are placed touching each other end to end as they enter the machine, and they travel through it at the rate of 720 per hour; smaller sizes are coated in proportion, the smaller the plates the larger is the number coated in a given time. The smaller plates pass through the machine in two parallel rows, instead of in a single row, so that quarter plates, 4¼ in. by 3¼ in., are delivered at the end of the machine at the rate of 2,800 per hour, keeping two attendants well employed in picking them up and placing them in racks as quickly as they can do the work. The double row of cords for carrying two lines of small plates through the machine is represented in the engraving. Although the plates touch each other at their edges on entering the machine, they are separated from each other by short intervals after being coated; this is effected by differential gearing. The water flowing over the tables for cooling the plates is caught in receptacles below and carried away by pipes. Between each of the tables is a little roller to enable small plates to travel without tilting over the necessary gap between each pair of bands.
The feeding trough of Cadett's machine is represented in Fig. 8. The plates, cleaned as already described, are carried upon the cords under a brass roller, the weight of which causes sufficient friction to keep the plates from tilting; they next pass under a soft camel's hair brush to remove anything in the shape of dust or grit, and are then coated. They afterward pass over a series of accurately leveled wheels running in a tank of water kept exact by an automatic regulator at a temperature of from 80 deg. Fah. to 100 deg. Fah., by means of a small hot water circulating system. The emulsion trough is jacketed with hot water at a constant temperature. This trough is silver plated inside, because most metals in common use would spoil the emulsion by chemical action. The trough is 16 in. long; it somewhat tapers toward the bottom, and contains a series of silver pumps shown in the cut; the whole of this series of pumps is connected with one long adjustable crank when plates of the largest size have to be coated; when coating plates of smaller sizes some of the pumps are detached. A chief object of the machine is to deliver a carefully measured quantity of emulsion upon each plate, and this is done by means of pumps, in order that the quantity of emulsion delivered shall not be affected by changes in the level of the emulsion in the trough; the quantity delivered is thus independent of variations due to gravity or to the speed of the machine. These pumps draw the emulsion from a sufficient depth in the trough to avoid danger from the presence of air bubbles, and the bottom of the trough is so shaped that should by chance any sedimentary matter be present, it has a tendency to travel downward, away from the bottoms of the pumps. There is a steady flow of emulsion from the pumps to the delivery pipes, then it passes down a guide plate of the exact width of the plate to be coated. Immediately in front of the guide plate is a fixed silver cylinder, kept out of contact with the plate by the thickness of a piece of fine and very hard hempen cord, which can be renewed from time to time. These cords keep the cylinder from scraping the emulsion off the plate, and they help to distribute it in an even layer. There would be two lines upon each plate where it is touched by the cords, were not the emulsion so fluid as to flow over the cut-like lines made and close them up.
The silver cylinder to a certain extent overcomes the effects of irregularities in the glass plates, for the cylinder is jointed somewhat in the cup and ball fashion, and is made in two or more parts, which parts are held together by lengths of India rubber.
The arrangement is shown in section in Fig. 9, in which A is the hot water jacket of the emulsion vessel; B, the crank driving the pumps; C, a pump with piston in position; D, delivery tube of the pump; E, the silver guide plate to conduct the emulsion down to the glass; F, the spreading cylinder; G, the cords regulating the distance of the cylinder from the glass plates; H, soft camel's hair brush; K, friction roller; L L L, three plates passing under the emulsion tank; M, knife edged wheels in the hot water tank, N; the "plucking roller," P, has a hot water tank of its own, and travels at slightly greater speed than the other rollers; R is the beginning of the cooling bands; T, the driving cords; and W, a level of the emulsion in the trough. Y represents one of the bucket pistons of the pumps, detached. The construction of the crank itself is such that, by adjustment of the connecting rods, more or less emulsion may be put upon the plates. Mr. Cowan, however, intends to adjust the pumps once for all, and to regulate the amount of emulsion delivered upon the plates by means of driving wheels of different diameters upon the cranks.
Fig. 10 is a section of the hollow spreading cylinder, made of sheet silver as thin as paper, so that its weight is light. For coating large plates it is divided in the center, so as to adapt itself somewhat to irregularities in the surface of each plate. In this case it is supported by a third and central thread, as represented in the cut. Otherwise the cylinder would touch the center of the plate. Its two halves are held together by a slip of India rubber.—The Engineer.
THE USE OF AMMONIA AS A REFRIGERATING AGENT.[1]
By Mr. T.B. Lightfoot, M.I.C.E.
Within the last few years considerable progress has been made in the application of refrigerating processes to industrial purposes, and the demand for refrigerating apparatus thus created has led to the production of machines employing various substances as the refrigerating agent. In a paper read by the author before the Institution of Mechanical Engineers, in May, 1886, these systems were shortly described, and general comparisons given as to their respective merits, scope of application, and cost of working. In the present paper it is proposed to deal entirely with the use of ammonia as a refrigerating agent, and to deal with it in a more full and comprehensive manner than was possible in a paper devoted to the consideration of a number of different systems and apparatus. In the United States and in Germany, as well as to some extent elsewhere, ammonia has been very generally employed for refrigerating purposes during the last ten years or so. In this country, however, its application has been extremely limited; and even at the present time there are but few ammonia machines successfully at work in Great Britain. No doubt this is, to a large extent, due to the fact that in the United States and in Germany there existed certain stimulating causes, both as regards climate and manufactures, while in this country, on the other hand, these causes were present only in a modified degree, or were absent altogether. The consequence was that up to a comparatively recent date the only machine manufactured on anything like a commercial scale was the original Harrison's ether machine, first produced by Siebe, about the year 1857—a machine which, though answering its purpose as a refrigerator, was both costly to make and costly to work. In 1878 the desirability of supplementing our then existing meat supply by means of the large stocks in our colonies and abroad led to the rapid development of the special class of refrigerating apparatus commonly known as the dry air refrigerator, which, in the first instance, was specially designed for use on board ship, where it was considered undesirable to employ chemical refrigerants. Owing to their simplicity, and perhaps also to their novelty, these cold air machines have very frequently been applied on land, under circumstances in which the same result could have been obtained with much greater economy by the use of ammonia or some other chemical agent. Recently, however, more attention has been directed to the question of economy, and consideration is now being given to the applicability of certain machines to certain special purposes, with the result that ammonia—which is the agent that, in our present state of knowledge, gives as a rule the best results for large installations, while on land at any rate its application for all refrigerating purposes presents no unusual difficulties—promises to become largely adopted. It is hoped, therefore, that the following paper respecting its use will be of interest.
In all cases where a liquid is employed, the refrigerating action is produced by the change in physical state from the liquid to the vaporous form. It is, of course, well known that such a change can only be brought about by the acquirement of heat; and for the purpose of refrigeration (by which must be understood the abstraction of heat at temperatures below the normal) it is obvious that, other things being equal, that liquid is the best which has the highest heat of vaporization, because with it the least quantity has to be dealt with in order to produce a given result. In fact, however, liquids vary, not only in the amount of heat required to vaporize them (this amount also varying according to the temperature or pressure at which vaporization occurs), but also in the conditions under which such change can be effected. For instance, water has an extremely high latent heat, but as its boiling point at atmospheric pressure is also high, evaporation at such temperatures as would enable it to be used for refrigerating purposes can only be effected under an almost perfect vacuum. The boiling point of anhydrous ammonia, on the other hand, is 37½° below zero F. at atmospheric pressure, and therefore for all ordinary cooling purposes its evaporation can take place at pressures considerably above that of our atmosphere. Some other agents used for refrigerating purposes are methylic ether, Pictet's liquid, sulphur dioxide, and ether. In this connection it should be stated that Pictet's liquid is a compound of carbon dioxide and sulphur dioxide, and is said to possess the property of having vapor tensions not only much below those of pure carbon dioxide at equal temperatures, but even below those of pure sulphur dioxide at temperatures above 78° F. The considerations, therefore, which chiefly influence the selection of a liquid refrigerating agent are:
1. The amount of heat required to effect the change from the liquid to the vaporous state, commonly called the latent heat of vaporization.
2. The temperatures and pressures at which such change can be effected.
This latter attribute is of twofold importance; for, in order to avoid the renewal of the agent, it is necessary to deprive it of the heat acquired during vaporization, under such conditions as will cause it to assume the liquid form, and thus become again available for refrigeration. As this rejection of heat can only take place if the temperature of the vapor is somewhat above that of the cooling body which receives the heat, and which, for obvious reasons, is in all cases water, the liquefying pressure at the temperature of the cooling water, and the facility with which this pressure can be reached and maintained, is of great importance in the practical working of any refrigerating apparatus. Ammonia in its anhydrous form, the use of which is specially dealt with in this paper, is a liquid having at atmospheric pressure a latent heat of vaporization of 900, and a boiling point at the same pressure of 37½° below zero F. Water being unity, the specific gravity of the liquid at a temperature of 40° F. is 0.76, and the specific gravity of its vapor is 0.59, air being unity. In the use of ammonia, two distinct systems are employed. So far, however, as the mere evaporating or refrigerating part of the process is concerned, it is the same in both. The object is to evaporate the liquid anhydrous ammonia at such tension and in such quantity as will produce the required cooling effect. The actual tension under which this evaporation should be effected in any particular case depends entirely upon the temperature at which the acquirement of heat is to take place, or, in other words, on the temperature of the material to be cooled. The higher the temperature, the higher may be the evaporating pressure, and therefore the higher is the density of the vapor, the greater the weight of liquid evaporated in a given time, and the greater the amount of heat abstracted. On the other hand, it must be remembered that, as in the case of water, the lower the temperature of the evaporating liquid, the higher is the heat of vaporization. It is in the method of securing the rejection of heat during condensation of the vapor that the two systems diverge, and it will be convenient to consider each of these separately.
The Absorption Process.—The principle employed in this process is physical rather than mechanical. Ordinary ammonia liquor of commerce of 0.880 specific gravity, which contains about 38 per cent. by weight of pure ammonia and 62 per cent. of water, is introduced into a vessel named the generator. This vessel is heated by means of steam circulating through coils of iron piping, and a mixed vapor of ammonia and water is driven off. This mixed vapor is then passed into a second vessel, in order to be subjected to the cooling action of water. And here, owing to the difference between the boiling points of water and ammonia, fractional condensation takes place, the bulk of the water, which condenses first, being caught and run back to the generator, while the ammonia in a nearly anhydrous state is condensed and collected in the lower part of the vessel.
This process of fractional condensation is due to Rees Reece, and forms an important feature in the modern absorption machine. Prior to the introduction of this invention, the water evaporated in the generator was condensed with the ammonia, and interfered very seriously with the efficiency of the process by reducing the power of the refrigerating agent by raising its boiling point. In the improved form of apparatus, ammonia is obtained in a nearly anhydrous condition, and in this state passes on to the refrigerator. In this vessel, which is in communication with another vessel called the absorber, containing cold water or very weak ammonia liquor, evaporation takes place, owing to the readiness with which cold water or weak liquor absorbs the ammonia, water at 59° Fahr. absorbing 727 times its volume of ammonia vapor. The heat necessary to effect this vaporization is abstracted from brine or other liquid, which is circulated through the refrigerator by means of a pump. Owing to the absorption of ammonia, the weak liquor in the absorber becomes strengthened, and it is then pumped back into the generating vessel to be again dealt with as above described.
The absorption apparatus, as applied for cooling purposes, consists of a generator, which is a vessel of cast iron containing coils of iron piping to which steam at any convenient pressure is supplied; an analyzer, in which a portion of the water vapor is condensed, and from which it flows back immediately into the generator; a rectifier and condenser, in the upper portion of which a further condensation of water vapor and a little ammonia takes place, the liquid thus formed passing back by a pipe to the analyzer and thence to the generator, while in the lower portion the ammonia vapor is condensed and collected; and a refrigerator or cooler, into which the nearly anhydrous liquid obtained in the condenser is admitted by a pipe and regulating valve, and allowed to evaporate, the upper portion being in communication with the absorber.
Through this vessel weak liquor, which has been deprived of its ammonia in the generator, is continually circulated, after being first cooled in an economizer by an opposite current of strong cold liquor passing from the absorber to the generator, while, in addition, the liquor in the absorber, which would become heated by the liberation of heat due to the absorption and consequent liquefaction of the ammonia vapor, is still further cooled by the circulation of cold water. As the pressure in the absorber is much lower than that in the generator, the strong liquor has to be pumped into the latter vessel, and for this purpose pumps are provided. Though of necessity the various operations have been described separately, the process is a continuous one, strong liquor from the absorber being constantly pumped into the generator through the heater or economizer, while nearly anhydrous liquid ammonia is being continually formed in the condenser, then evaporated in the refrigerator and absorbed by the cool weak liquor passing through the absorber.
Putting aside the effect of losses from radiation, etc., all the heat expended in the generator will be taken up by the water passing through the condenser, less that portion due to the condensation of the water vapor in the analyzer, and plus the amount due to the difference between the temperature of the liquid as it enters the generator and the temperature at which it leaves the condenser. In the refrigerator the liquid ammonia, in becoming vaporized, will take up the precise quantity of heat that was given off during its cooling and liquefaction in the condenser, plus the amount due to the difference in heat of vaporization, owing to the lower pressure at which the change of state takes place in the refrigerator, and less the small amount due to the difference in temperature between the vapor entering the condenser and that leaving the refrigerator, less also the amount necessary to cool the liquid ammonia to the refrigerator temperature. When the vapor enters into solution with the weak liquor in the absorber, the heat taken up in the refrigerator is imparted to the cooling water, subject also to corrections for differences of pressure and temperature. The sources of loss in such an apparatus are:
a. Radiation and conduction of heat from all vessels and pipes above normal temperature, which can, to a large extent, be prevented by lagging.
b. Conduction of heat from without into all vessels and pipes that are below normal temperature, which can also to a large extent be prevented by lagging.
c. Inefficiency of economizer, by reason of which heat obtained by the expenditure of steam in the generator is passed on to the absorber and there uselessly imparted to the cooling water.
d. The entrance of water into the refrigerator, due to the liquid being not perfectly anhydrous.
e. The useless evaporation of water in the generator. With regard to the amount of heat used, it will have been seen that the whole of that required to vaporize the ammonia, and whatever water vapor passes off from the generator, has to be supplied from without. Owing to the fact that the heating takes place by means of coils, the steam passed through may be condensed, and thus each pound can be made to give up some 950 units of heat. With the absorption process worked by an efficient boiler, it may be taken that 200,000 thermal units per hour may be eliminated by the consumption of about 100 lb. of coal per hour, with a brine temperature in the refrigerator of about 20° Fahr.
Compression Process.—In this process ammonia is used in its anhydrous form. So far as the action of the refrigerator is concerned, it is precisely the same as it is in the case of the absorption apparatus, but instead of the vapor being liquefied by absorption by water, it is drawn from the refrigerator by a pump, by means of which it is compressed and delivered into the condenser at such pressure as to cause its liquefaction at the temperature of the cooling water. It must be borne in mind, however, that allowance must be made for the rise of temperature of the water passing through the condenser, and also for the difference in temperature necessary in order to permit the transfer of heat from one side of the cooling surface to the other. In a compression machine the work applied to the pump may be accounted for as follows:
a. Friction.
b. Heat rejected during compression and discharge.
c. Heat acquired by the ammonia in passing through the pump.
d. Work expended in discharging the compressed vapor from the pump.
But against this must be set the useful mechanical work performed by the vapor entering the pump. The heat rejected in the condenser is the heat of vaporization taken up in the refrigerator, less the amount due to the higher pressure at which the change in physical state occurs, plus the heat acquired in the pump, and less the amount due to the difference between the temperature at which the vapor is liquefied in the condenser and that at which it entered the pump. An ammonia compression machine, as applied to ice making, contains ice-making tanks, in which is circulated a brine mixture, uncongealable at any temperature likely to be reached during the process. This brine also circulates around coils of wrought iron pipes, in which the liquid ammonia passing from the condenser is vaporized, the heat required for this vaporization being obtained from the brine. A pump draws off the ammonia vapor from the refrigerator coils, and compresses it into the condenser, where, by means of the combined action of pressure and cooling by water, it assumes a liquid form, and is ready to be again passed on to the refrigerator for evaporation. The ammonia compression process is more economical than the absorption process, and with a good boiler and engine about 240,000 thermal units per hour can be eliminated by the expenditure of 100 lb. of coal per hour, with a brine temperature in the refrigerator of about 20° Fahr.