Precautions.—In order not to strain the rubber of the bags or of the anti-pulsators, it is advisable to place a stop-cock in advance of these devices so that they can not be filled while the motor is at rest.

The capacity of the rubber bags that can be bought in the market being limited, it is necessary to place one, two, or three extra bags in series (Figs. 48 and 49), for large pipes; but it should be borne in mind that the total section of the branch pipes should be at least equal to that of the main pipe. It is also advisable to extend the tube completely through the bag as shown in Figs. 48 and 49.

If there be two branch pipes the minimum diameter which meets this requirement is ascertained as follows: Draw to any scale a semicircle having a diameter equal or proportional to that of the main pipe (Fig. 50). The sides of the isosceles triangle inscribed within this semicircle give the minimum diameter of each of the branch pipes.

Sometimes engines are provided with a cock having an arrangement by means of which the gas feed is permanently regulated, according to the quality and pressure of the gas and according to the load at which the engine is to run. This renders it possible to open the cock always to the same point (Fig. 51).

Air Suction.—In a special chapter the precautions to be taken to counteract the influence of the suction of the engine in causing vibration will be treated. The manner in which the suction of air is effected necessarily has as marked an influence on the operation of the engine as the supply of gas, since air and gas constitute the explosive mixture.

Resistance to the suction of air should be carefully avoided, for which reason the length of the pipe should be reduced to a minimum, and its cross-section kept at least equal to that of the air inlet of the engine. Since the quality of street-gas varies with each city, the proper proportions of gas and air are not constant. In order that these proportions may be regulated, it is a matter of some importance to fit some suitable device on the pipe. Good engines are provided with a plug or flap valve. Generally the air-pipe terminates either in the hollowed portion of the frame, or in an independent pot, or air chest. The first arrangement is not to be recommended for engines over 20 to 25 horse-power. Accidents may result, such as the breaking of the frame by reason of back firing, of which more will be said later. If an independent chest be employed, its closeness to the ground renders it possible for dust easily to pass through the air-holes in the walls at the moment of suction, and even to enter the cylinder, where its presence is particularly harmful, leading, as it does, to the rapid wear of the rubbing surfaces. This evil can be largely remedied by filling the air-chest with cocoa fiber or even wood fiber, provided the latter does not become packed down so as to prevent the air from passing freely. Such fibers act as air-filters. Regular cleaning or renewal of the fiber protects the cylinder from wear. In a general way, care should be taken, before fitting both the gas and air pipes, to tap the pipes, elbows, and joints lightly with a hammer on the outside in order to loosen whatever rust or sand may cling to the interior; otherwise this foreign matter may enter the cylinder and cause perturbations in the operation of the engine. Under all circumstances, care should be taken not to place the end of the air-pipe under the floor or in an enclosed space, because leakage may occur, due to the bad seating of the air-valve, thereby producing a mixture which may explode if the flame leaps back, as we shall see in the discussion of suction by pipes terminating in the hollow of the frame. On the other hand, sand or sawdust should not be sprinkled on the floor.

Exhaust.—For the exhaust, cast-iron or drawn pipes as short as possible should be used. Not only the power of the engine, but also its economic consumption, can be markedly affected by the employment of long and bent pipes. Resistance to the exhaust of the products of combustion not only causes an injurious counter-pressure, but also prevents the clearing of the cylinder of burnt gases, which contaminate the aspired mixture and rob it of much of its explosiveness. The necessity of evacuating the cylinder as completely as possible is, nevertheless, not always reconcilable with local surroundings. To a certain extent, the objections to long exhaust-pipes are overcome by rigorously avoiding the use of elbows. Gradual curves are preferable. In the case of very long pipes it is advisable to increase their diameter every 16 feet from the exhaust. The exhaust-chest should be placed as near as possible to the engine; it should never be buried; for the joints of the inlet and outlet pipes of the exhaust-chest should be easily accessible, so that they may be renewed when necessary. The author recommends the placing of the exhaust-chest in a masonry pit, which can be closed with a sheet-metal cover. For engines of 20 horse-power and upward, these joints should be entirely of asbestos. Pipes screwed directly into the casting are liable to rust. Exposed as they are to the steam or water of the exhaust, they cannot be detached.

The water, which results from the combination of the hydrogen of the gas with the oxygen of the air, is deposited in most cases at the bottom of the exhaust-chest. It is advisable to fit a plug or iron cock in the base of the chest. Alkaline or acid water will always corrode a bronze cock. In order that the pipes may not also be attacked, they are not disposed horizontally, but are given a slight incline toward the point where the water is drained off. If pipes of some length be employed, they should be able to expand freely without straining the joints, as shown in the accompanying diagram (Fig. 52), in which the exhaust-chest rests on iron rollers which permit a slight displacement.

For the sake of safety, at least that portion of the piping which is near the engine should be located at a proper distance from woodwork and other combustible material. By no means should the exhaust discharge into a sewer or chimney, even though the sewer or chimney be not in use; for the unburnt gases may be trapped, and dangerous explosions may ensue at the moment of discharge.

The joints or threaded sleeves employed in assembling the exhaust-pipe should be tested for tightness. The combined action of the moisture and heat causes the metal to rust and to deteriorate very rapidly at leaky spots.

When several engines are installed near one another, each should be provided with a special exhaust-pipe; otherwise it may happen, when the engines are all running at once, that the products of combustion discharged by the one may cause a back pressure detrimental to the exhaust of the next.

It is possible to employ a pipe common to all the exhausts if the pipe starts from a point beyond the exhaust-chests, in which case Y-joints and not T-joints are to be used.

The manner of securing the pipes to walls by means of detachable hangers, lined with asbestos, is shown in a general way in the accompanying Fig. 53. The object of this arrangement is to render detachment easy and to prevent the transmission of shocks to the masonry.

The precautions to be taken for muffling the noise of the exhaust will be discussed later.

The end of the exhaust-pipe should be slightly curved down in order to prevent the entrance of rain. Exhaust-pipes are subjected to considerable vibration, due to the sudden discharge of the gases. To protect the joints, the pipes should be rigidly fastened in place.

Legal Authorization.—In most countries gas-engines may be installed only in accordance with the provision of general or local laws, which impose certain conditions. These laws vary with different localities, for which reason they are not discussed here.

CHAPTER IV

As in the construction of all foundations, care should be taken to excavate down to good soil and to line the bottom with concrete, in order to form a single mass of artificial stone. A day or two should be allowed for the masonry to dry out, before filling in around it.

When the engine is installed on the ground floor above a vaulted cellar, the foundation should not rest directly on the vault below or on the joists, but should be built upon the very floor of the cellar, so that it passes through the planking of the ground floor without contact.

When the engine is to be installed on a staging, the method of securing it in place illustrated in Fig. 55 should be adopted.

Although a foundation, built in the manner described, will fulfill the usual conditions of an industrial installation, it will be inadequate for special cases in which trepidation is to be expected. Such is the case when engines are to be installed in places where, owing to the absence of factories, it is necessary to avoid all nuisance, such as noise, trepidations, odors, and the like.

Vibration.—In order to prevent the transmission of vibration, the foundation should be carefully insulated from all neighboring walls. For this purpose various insulating substances called "anti-vibratory" are to be recommended. Among these may be mentioned horsehair, felt packing, cork, and the like. The efficacy of these substances depends much on the manner in which they are applied. It is always advisable to interpose a layer of one of these substances, from one to four inches thick, between the foundation and the surrounding soil, the thickness varying with the nature of the material used and the effect to be obtained. Between the bed of concrete, mentioned previously, and the foundation-masonry and between the foundation and the engine-frame, a layer of insulating material may well be placed. Preference is to be given to substances not likely to rot or at least not likely to lose their insulating property, when acted upon by heat, moisture or pressure.

Here it may not be amiss to warn against the utilization of cork for the bottom of the foundation; for water may cause the cork to swell and to dislocate the foundation or destroy its level.

The employment of the various substances mentioned does not entail any great expense when the foundations are not large and the engines are light. But the cost becomes considerable when insulating material is to be employed for the foundation of a 30 to 50 horse-power engine and upwards. For an engine of such size the author recommends an arrangement as simple as it is efficient, which consists in placing the foundation of the engine in a veritable masonry basin, the bottom of which is a bed of concrete of suitable thickness. The foundation is so placed that the lateral surfaces are absolutely independent of the supporting-walls of the basin thus formed. Care should be taken to cover the bottom with a layer of dry sand, rammed down well, varying in thickness with each case. This layer of sand constitutes the anti-vibratory material and confines the trepidations of the engine to the foundation.

As a result of this arrangement, it should be observed that, being unsupported laterally, the foundation should be all the more resistant, for which reason the base-area and weight should be increased by 30 to 40 per cent. The expense entailed will be largely offset by saving the cost of special anti-vibratory substances. In places liable to be flooded by water, the basin should be cemented or asphalted.

When the engine is of some size and is intended for the driving of one or more dynamos which may themselves give rise to vibrations, the dynamos are secured directly to the foundation of the engine, which is extended for that purpose, so that both machines are carried solidly on a single base.

The foregoing outline should not lead the proprietor of a plant to dispense with the services of experts, whose long experience has brought home to them the difficulties to be overcome in special cases.

It should here be stated, as a general rule, that the bricks should be thoroughly moistened before they are laid in order that they may grip the mortar.

After having been placed on the foundation and roughly trimmed with respect to the transmission devices, the engine is carefully leveled by means of hardwood wedges driven under the base. This done, the bolts are sealed by very gradually pouring a cement wash into the holes, and allowing it to set. When the holes are completely filled and the bolts securely fastened in place, a shallow rim, or edge of clay, or sand is run around the cast base, so as to form a small box or trough, in which cement is also poured for the purpose of firmly binding the engine frame and foundation together. When, as in the case of electric-light engines, single extra-heavy fly-wheels are employed, provided with bearings held in independent cast supports, the following rule should be observed to prevent the overheating due to unlevelness, which usually occurs at the bushings of these bearings: That part of the foundation which is to receive such a support should rest directly on the concrete bed and should be rigidly connected at the bottom with the main foundation. When the foundation is completely blocked up, the fly-wheel bearing with its support is hung to the crank-shaft; and not until this is effected is the masonry at the base of the support completed and rigidly fixed in its proper position.

For very large engines, the foundation-bolts should be particularly well sealed into the foundation. In order to attain this end the bricks are laid around the bolt-holes, alternately projected and retracted as shown in Fig. 54. Broken stone is then rammed down around the fixed bolt; in the interstices cement wash is poured.

Air Vibration, etc.—Vibration due chiefly to the transmission of noises and the displacement of air by the piston should not be confused with the trepidations previously mentioned.

The noise of an engine is caused by two distinct phenomena. The one is due to the transmitting properties of the entire solid mass constituting the frame, the foundation, and the soil. The other is due to vibrations transmitted to the air. In both cases, in order to reduce the noise to a minimum, the moving parts should be kept nicely adjusted, and above all, shocks avoided, the more harmful of which are caused by the play between the joint at the foot of the connecting-rod and the piston-pin, and between the head of the connecting-rod and the crank-shaft.

Although smooth running of the engine may be assured, there is always an inherent drawback in the rapid reciprocating movement of the piston. In large, single-acting gas-engines, a considerable displacement of air is thus produced. In the case of a forty horse-power engine having a cylinder diameter and piston-stroke respectively of 13³⁄₄ inches and 21³⁄₅ inches, it is evident that at each stroke the piston will displace about 2 cubic feet of air, the effect of which will be doubled when it is considered that on the forward stroke back pressure is created and on the return stroke suction is produced.

The air motion caused by the engine is the more readily felt as the engine-room is smaller. If the room, for example, be 9 feet by 15 feet by 8 feet, the volume will be 1,080 cubic feet. From this it follows that the 2 cubic feet of air in the case supposed will be alternately displaced six times each second, which means the displacement of 12 cubic feet at short intervals with an average speed of 550 feet per minute. Such vibrations transmitted to halls or neighboring rooms are due entirely to the displacement of the air.

In installations where the air-intake of the engine is located in the engine-room, a certain compensation is secured, at the period of suction, between the quantity of air expelled on the forward stroke of the piston and the quantity of air drawn into the cylinder. From this it follows that the vibration caused by the movement of the air is felt less and occurs but once for two revolutions of the engine.

This phenomenon is very manifest in narrow rooms in which the engine happens to be installed near glass windows. By reason of the elasticity of the glass, the windows acquire a vibratory movement corresponding in period with half the number of revolutions of the engine. It follows from the preceding that, in order to do away with the air vibration occasioned by the piston in drawing in and forcing out air in an enclosed space, openings should be provided for the entrance of large quantities of air, or a sufficient supply of air should be forced in by means of a fan.

The author ends this section with the advice that all pipes in general and the exhaust-pipe in particular be insulated from the foundation and from the walls through which they pass as well as from the ground, as metal pipes are good conductors of sound and liable to carry to some distance from the engine the sounds of the moving parts.

Exhaust Noises.—Among the most difficult noises to muffle is that of the exhaust. Indeed, it is the exhaust above all that betrays the gas-engine by its discharge to the exterior through the exhaust-pipe. The most commonly employed means for rendering the exhaust less perceptible consists in extending the pipe upward as far as possible, even to the height of the roof. This is an easy way out of the difficulty; but it has a bad effect on the operation of the engine. It reduces the power generated and increases the consumption, as will be explained in a special paragraph.

Expansion-boxes, more commonly called exhaust-mufflers, considerably deaden the noise of explosion by the use of two or three successive receptacles. But this remedy is attended with the same faults that mark the use of extremely long pipes. The best plan is to mount a single exhaust-muffler near the discharge of the engine in the engine-room itself, where it will serve at least the purpose of localizing the sound.

The employment of pipes of sufficiently large cross-section to constitute expansion-boxes in themselves will also muffle the exhaust. A more complete solution of the problem is obtained by causing the exhaust-pipe, after leaving the muffler, to discharge into a masonry trough having a volume equal to twelve times that of the engine-cylinder (Fig. 56). This trough should be divided into two parts, separated by a horizontal iron grating. Into the lower part, which is empty, the exhaust-pipe discharges; in the upper part, paving-blocks or hard stones not likely to crumble with the heat, are placed. Between this layer of stones and the cover it is advisable to leave a space equal to the first. Here the gases may expand after having been divided into many parts in passing through the spaces left between adjacent stones. The trough should not be closed by a rigid cover; for, although efficient muffling may be attained, certain disadvantages are nevertheless encountered. It may happen that in a badly regulated engine, unburnt gases may be discharged into this trough, forming an explosive mixture which will be ignited by the next explosion, causing considerable damage. Still, the explosion will be less dangerous than noisy. It may be mentioned in passing that this disadvantage occurs rarely.

A second arrangement consists in superposing the end of the exhaust-pipe upon a casing of suitable size, which casing is partitioned off by several perforated baffle-plates. This casing is preferably made of wood, lined with metal, so that it will not be resonant. The size of the casing, the number of partitions and their perforations, and the manner of disposing the partitions have much to do with the result to be obtained. Here again the experience of the expert is of use.

Various other systems are employed, depending upon the particular circumstances of each case. Among these systems may be mentioned those in which the pipe is forked at its end to form either a yoke (Fig. 57) or a double curve, each branch of which terminates in a muffler (Fig. 58).

It should be observed that, under ordinary conditions, noises heard as hissing sounds are often due to the presence of projections, or to distortion of the pipes near the discharge opening. Consequently, in connecting the pipes, care should be taken that the joints or seams have no interior projections. Occasionally, water may be injected into the exhaust-muffler in order to condense the vapors of the exhaust, the result being a deadening of the noises; but in order to be truly efficient this method should be employed with discretion, for which reason the advice of an expert is of value.

CHAPTER V

Running Water.—A water-jacket fed from a constant source of running water, such as the water mains of a town, is certainly productive of the best results, the supply, moreover, being easily regulated; but the system is not widely used because the water runs away and is entirely lost. If running water be employed, the outlet of the jacket is so disposed that the water gushes out immediately on leaving the cylinder, and that the flow is visible and accessible, in order that the temperature may be tested by the hand. Apart from the relatively great cost of water in towns, the use of running water is objectionable on account of its chemical composition. Though it may be clear and limpid, it frequently contains lime salts, carbonates, sulphates, and silicates which are precipitated by reason of the sudden change of temperature to which the water is subjected as it comes into contact with the walls of the cylinder. That part of the water-jacket surrounding the head or explosion-chamber, where the temperature is necessarily the highest, becomes literally covered with calcareous incrustations, which are the more harmful because they are bad conductors of heat and because they reduce and even obstruct the passage exactly at the point where the water must circulate most freely to do any good. If the circulating water be pumped into the jacket, it is preferable, wherever possible, to use cistern water, which is not likely to contain lime salts in suspension. If river water be used, it should be free from the objections already mentioned, which are all the more grave if the water be muddy, as sometimes happens. The water-jacket can be easily freed from all non-adhering deposits by flushing it periodically through the medium of a conveniently placed cock. It is always preferable to pass the water through a reservoir where its impurities can settle, before it flows to the cylinder. In the case considered, the water usually has an average temperature of 54 to 60 degrees F., under which condition the hourly flow should be at least 5¹⁄₂ gallons per horse-power per hour, the temperature rising at the outlet-pipe of the cylinder to 140 and 158 degrees F., which should not be surpassed. However, in engines working with high compression, 104 to 122 degrees F. should not be exceeded.

If the water-jacket be fed by a reservoir, it is essential that the reservoir comply with the following conditions:

In horizontal engines the water-inlet is always located in the base of the cylinder, while the outlet is located at the top. By providing the inlet-pipe extending to the cylinder with a cock, the circulation of water can be regulated to correspond with the work performed by the engine. Another cock at the end of the outlet-pipe near the reservoir serves, in conjunction with the first, to arrest the circulating water. When the weather is very cold or when the cylinder must be repaired, these two cocks may be closed, and the pipe and water-jacket of the cylinder drained by means of the drain-cock V (Fig. 59), mounted at the inlet of the engine's water-jacket. In order that the pressure of the atmosphere may not prevent the flowing of the water, the highest part of the pipe is provided with a small tube, T, communicating with the atmosphere.

On account of the importance of preventing losses of the charge in the pipes the author recommends the utilization of sluice-valves of the type shown in Fig. 60, instead of the usual cone or plug type.

Water-Tanks.—The reservoir is mounted in such a way that its base is flush with the top of the cylinder; it should be as near as possible to the cylinder in order to obviate the use of long inlet and return pipes. This fact, however, does not necessarily render it advisable to place the reservoir in the engine-room; for such a disposition is doubly disadvantageous in so far as it does not permit a sufficiently rapid cooling of the circulating water by reason of the high temperature of the surrounding air, and in so far as it is liable to cause the formation of vapors which injuriously affect the engine. Consequently, the reservoir should be placed in as cool a place as possible, preferably even in the open air; for the water is not likely to freeze, except when it has been allowed to stand for a considerable time. The reservoir should be left uncovered so as to facilitate cooling by the liberation of the vapors formed on the surface of the water.

Circulation being effected solely by the difference in specific gravity or density between the warmer water emerging from the cylinder and the cooler water which flows in from the reservoir, the slightest obstruction will impede the flow. Hence, the cross-section of the pipes should not be less than that of the inlet and outlet openings of the cylinder of the engine. Good circulation cannot be attained if the water must overcome inclines or obstacles in the pipes themselves. Instead of elbows, long curves of great radius, limited to the smallest possible number, should be employed. This is particularly true of the return-pipe extending from the cylinder back to the reservoir. For this pipe a minimum incline of 10 to 15 per cent. should be allowed, in order that the water may run up into the reservoir. The height of the water in the reservoir should be from 2 to 4 inches above the discharge of the return-pipe. In order to maintain this level it is advisable to use some automatic device such as a float-valve, in which case the reservoir should not be allowed to become too full.

The size of a reservoir is determined by the engine; it should be large enough to enable the engine to run smoothly at its maximum load for several hours consecutively. Under these conditions, the reservoir should have a capacity of 45 to 55 gallons per horse-power for engines with "hit-and-miss" admission, and 55 to 65 gallons for engines controlled by variable admission. It is not advisable to employ reservoirs having a capacity of more than 330 to 440 gallons, the usual diameter being about 3 feet.

If the power of the engine be such that several reservoirs are necessary, then the reservoirs should be connected in such a manner that the top of the first communicates with the bottom of the next and so on, the first reservoir receiving the water as it comes from the cylinder (Fig. 61).

Intercommunication of the reservoirs by means of a common top tube (a) is objectionable; and simultaneous intercommunication at top and bottom (a and b) is ineffective, so far as one of the reservoirs is concerned (Fig. 62).

The reservoirs are true thermo-siphons. Consequently the water should be methodically circulated; in other words, the hottest water, flowing from the engine into the top of the first reservoir and having, for example, a temperature of 104 degrees F., is cooled off to 86 degrees F. and drops to the bottom of the reservoir, thence to be driven, at a temperature sensibly equal to 86 degrees F., to the second reservoir, where a further cooling of 18 degrees F. takes place. In passing on to the following reservoirs the temperature is still further lowered, until the water finally reaches its minimum temperature, after which it flows back to the engine-cylinder.

In order to effect this cooling, the reservoirs can be connected in several ways. The most common method, as shown in Fig. 63, consists in connecting the reservoirs by oblique pipes. This is open to criticism, however, since leakage occurs, caused by the employment of elbows which retard the circulation. A less cumbrous and more efficient method of connection consists in joining the reservoirs by a single pipe at the top, as shown in Fig. 61; but care must be taken to extend this pipe at the point of its entrance into the adjoining reservoir by means of a downwardly projecting extension, or to fit its discharge-end with a box, closed by a single partition, open at the bottom.

In order to prevent incrustation of the water-jacket surrounding the cylinder, a pound of soda per 17 cubic feet of the reservoir capacity is monthly introduced, and the jacket flushed weekly by a cock conveniently mounted near the cylinder (Fig. 59). The jacket is thus purged of calcareous sediments, which are prevented by the soda from adhering to the metal. The flushing-cock mentioned also serves to drain the water-jacket of the cylinder in case of intense or persistent cold, which would certainly freeze the water in the jacket, thereby cracking the cylinder or the exposed pipes.

In order to regulate the circulation of the water in accordance with the work performed by the engine, a cock should be fitted to the water supply pipe at a convenient place.

In engines of large size, driven at full load for long periods, cooling by natural circulation is often inadequate. In such cases, circulation is quickened by a small rotary or reciprocating pump, driven from the engine itself and fitted with a by-pass provided with a cock. This arrangement permits the renewal of the natural thermo-siphon circulation in case of accident to the pump (Fig. 64).

Coolers.—The arrangement which is illustrated in Fig. 65, and which has the merit of simplicity, will be found of service in cooling the water. It comprises a tank B surmounted by a set of trays E, formed of frames to which iron rods are secured, spaced 1 to 2 feet apart, so as to form superimposed series separated by 1¹⁄₂ to 2¹⁄₃ feet. On these trays bundles of tree branches are placed. The cold water at the bottom of the tank is forced by the pump Pi into the water-jacket, from which it emerges hot, and flows through the pipe T, which ends in a sprinkler G, formed of communicating tubes and perforated with a sufficient number of holes to enable the water to fall upon the trays in many drops. Thus finely divided, the water falls from one tray to another, retarded as it descends by the bundles of tree branches. It finally reaches the tank in a very cold condition and is then ready to be pumped to the engine. Birch branches are to be preferred on account of their tenuity.

Great care should be taken to cover the tank with a sheet-metal closure in order to prevent twigs and foreign bodies from entering and from being drawn into the pump.

In the following table the dimensions of an operative apparatus of this kind are given,—an apparatus, moreover, that may be constructed of wood or of iron:—

In order that the water may not drop to one side, the base of the apparatus should be made 10 to 12 inches less in width than the tank.

The size of these apparatus may be considerably reduced by constructing them in the form of closed chests, into the bottom of which air maybe injected by means of fans in order to accelerate cooling (Fig. 66).

CHAPTER VI

The piston is the essential, vital element of an engine. Upon its freedom from leakage depends the maintenance of a proper compression, and, consequently, the production of power and economical consumption. As it travels forward and as it recedes from the explosion-chamber, it uncovers more and more of the frictional surface constituting the interior wall of the cylinder. This surface, as a result, is regularly brought into contact with the ignited, expanding gases after each explosion. For this reason the oil which covers the wall is constantly subjected to high temperatures, by which it is likely to be volatilized and burned. Therefore, the first condition to be fulfilled in properly lubricating the piston is a constant and regular supply of oil.

Quality of Oils.—For cylinder lubrication only the very best oils should be used; perfect lubrication is of such importance that cost should not be considered. Besides, the surplus oil which is usually caught in the drip-pan is by no means lost. After having been filtered it can be used for lubricating the bearings of the crank, the cam-shaft, and like parts.

Cylinder-oil should be exceedingly pure, free from acids, and composed of hydrocarbons that leave no residue after combustion. Only mineral oils, therefore, are suitable for the purpose. Those oils should be selected which, with a maximum of viscosity, are capable of withstanding great heat without volatilizing or burning. The point at which a good cylinder-oil ignites should not be lower than 535 degrees F.

Whether an oil possesses this essential quality is easily enough ascertained in practice without resorting to laboratory tests. All that is necessary is to heat the oil in a metal vessel or a porcelain dish. In order that the temperature may be uniform the vessel is shielded from the direct flame by interposing a piece of sheet metal or a layer of dry sand. As soon as gases begin to arise a lighted match is held over the oil. When the gases are ignited the thermometer reading is taken, the instrument being immersed in the oil. The temperature recorded is that corresponding with the point of ignition.

For cylinder lubrication American mineral oil is preferable to Russian oil. The specific gravity should lie somewhere between .886 and .889 at 70 degrees F. Oil of this quality begins to evaporate at about 365 degrees F. Ignition occurs at 535 degrees F. The point of complete combustibility lies between 625 and 645 degrees F. Oil of this quality solidifies at 39 or 41 degrees F. Its color is a reddish yellow with a greenish fluorescence. Compared with water its degree of viscosity lies between 11.5 and 12.5 at a temperature of 140 degrees F.

Before lubricating other parts of the engine with oil that has been used for the piston, heavy particles and foreign matter, such as dust, bearing incrustations, and the like, should be filtered out. The piston-pivot and the connecting-rod head are preferably lubricated with fresh oil, because their constant movement renders inspection difficult and the control of lubrication irksome. A good, industrial mineral oil of usual market quality will be found satisfactory. In order to bring home the importance of employing good cylinder-oil and of proper lubrication the author can only state that in his personal experience he has frequently detected losses varying from 10 to 15 per cent. in the power developed by engines poorly lubricated.

Types of Lubricators.—Among the more common apparatus employed for automatically lubricating the cylinder, the author mentions an English oiler of the type pictured in Fig. 67 which is driven simply by a belt from the intermediary shaft, and which rotates the pulley P secured on the shaft a of the apparatus, at a very slow speed. The shaft a is provided at its end with a small crank, from which a small iron arm f is suspended, which arm dips in the oil contained in the cup G of the oiler. When the shaft a is turned this arm, as it sweeps through the oil-bath, collects a certain quantity of oil which it deposits on the collector b. From this spindle the oil passes through an outlet-pipe opening into the bottom of the oiler, and thence to the cylinder. The entire apparatus is closed by a cover D which can be easily removed in order to ascertain the quantity of oil still remaining in the apparatus. Many other systems are utilized which, like the one that has been described, enable the feed to be controlled. Often small force-pumps are employed as cylinder-lubricators. Whatever may be the type selected, preference should be given to that in which the feed is visible (Fig. 68).