Figs. 88 and 9 show two views of a single motor-car engine cylinder, the water-jacket forming part of the cylinder casting. In the figures C is the cylinder barrel or bore; J the water-jacket; I the inlet for the jacket water; O the outlet for the jacket water; D is for the compression tap; S for the sparking plug; V₁, V₂ are the valve seats; G₁, G₂ are the valve stem guides; H₁, H₂ are caps which may be removed when the valves are being put in or taken out; f₁, f₂, f₃, f₄, f₅ are called flanges. The flange f₁ is used for attaching the cylinder to the crankchamber; while it is quite true that the force of the explosion within the cylinder drives the piston downwards, it is equally true that it also tends to force the cylinder head off or to blow the cylinder casting upwards off the crankchamber, unless it is securely fastened to it by means of screws or bolts passing through the flange f₁. The flanges, f₂, f₃ are for the inlet and outlet water pipe attachments, and f₄, f₅ are for the induction pipe and exhaust pipe connexions. Generally the pipes will have flanges and be held tight against the flanges on the cylinder casting by means of screws or studs. Figs. 10, 11, and 12 show how two metal flanges are held in contact by means of screws or studs or bolts, and they also show the packing materials between the metal surfaces which keep the joint tight and prevent water or gas leaking across the flanges and escaping to the outside air, or air leaking in if the internal pressure is below that of the atmosphere.

In Figs. 8 and 9 the valves are placed one on each side of the cylinder, this form of cylinder being known as a T-headed cylinder, but it is rather more usual here in England to place both valves on the same side of the cylinder and next to each other as indicated in Fig. 13, this form of cylinder being known as an L-headed cylinder. The chief object is of course to avoid the use of two valve shafts and also to produce a neater looking engine, but the T-headed design is better cleaned or scavenged by the passage of the inlet and exhaust gases. When a motor-car engine has two cylinders we frequently find them both in a single casting, having a common water-jacket, and then we say they are cast in pairs. A four-cylinder engine may thus have: (1) Cylinders cast separately; (2) Cylinders cast in pairs; (3) Cylinders cast en bloc; or all four in a single large casting. The third method is cheapest in first cost, but in the event of breakage will become the most expensive. The second method is a sound compromise.

An example of a built-up cylinder and water-jacket is shown in Fig. 14, the cylinder barrel being of steel tube with steel flanges, and the water-jacket being formed by copper tube slipped over the outside of the steel cylinder. Its great advantage lies in the reduction of weight, and it is thus largely used for aeroplane work. The valves would then be fitted in the top cover of the cylinder and driven by overhead gearing.

CHAPTER III
ENGINE DETAILS

The Piston is perhaps the most important detail to consider, for it is on the piston that the force of the explosion acts when the heat energy is converted into mechanical energy. It must be made sufficiently strong to withstand the bursting effect of successive explosions, and yet if we make the metal too thick it will retain too much of the waste heat and the piston may seize in the cylinder due to expansion. To understand why the piston is likely to seize in the cylinder we have only to remember that when a metal body is heated it gets larger in every direction, but if cooled it returns to its original size. Now if we make the metal of the piston too thick so that the waste heat cannot pass quickly through it and dissipate itself at cooler parts of the engine, then the bulk of this heat will be concentrated in the piston head, causing it to expand and become a tight fit in the cylinder, as the cylinder walls are fairly thin and in contact with the jacket water which keeps them fairly cool and prevents them expanding much above their normal size. The actual amount of expansion is very small of course, but there is very little clearance between the piston and the cylinder walls, even when the engine is all cold—perhaps five-thousandths of an inch. The piston therefore must be strong, yet as light as we can make it, having regard to the necessity for its being amply stiff and rigid and able to stand up to its work.

Generally it will be an iron casting in the form of a small cylinder (see Fig. 15), having provision in it for the packing rings P, and the gudgeon pin G, with its fastening screws S₁, S₂. The piston itself, as we have observed, must be a nice sliding fit in the bore of the cylinder without any shake or side play when there are no packing rings in the grooves. The packing rings are turned to size so as to fit the cylinder exactly and prevent any gas leaking past the piston into the crankchamber. These rings are very light, are made from cast iron, and arranged to break joint, as indicated, by cutting the middle ring in the opposite direction to the two outer ones. Bosses are cast on the inside of the piston and afterwards bored out to receive the steel gudgeon pin or wrist pin G. This pin is best made of plain parallel cylindrical form ground true, and the bosses in the piston should be reamered out to the exact size of the pin. When the pin has been inserted the tapered screws are screwed hard up by means of a special spanner and bear against the pin, preventing it from coming loose or from shaking or knocking. There are many other methods of fixing the gudgeon pin which are not shown here; each has some special point in its favour, but the one illustrated is undoubtedly the best and affords a positive adjustment for wear.

An enlarged view of one of the bosses, showing the taper pin in detail and how the split pin Q prevents it from slacking back by contact with the wall of the piston, is shown in Fig. 16. Sometimes the lower part of the piston is made lighter by drilling holes through the walls. It is very important to reduce the weight of the piston as much as possible, otherwise the engine cannot attain a high speed, so that it becomes essential to bear this in mind when constructing engines for racing purposes. Frequently we find steel pistons used, as they may be made lighter for the same strength, and then steel piston rings may be used; they are not much in favour for ordinary motor-car engines because the steel pistons expand at a greater rate than the cast iron of the cylinder, so that there is more liability to seizure. The crown of the piston is sometimes curved upwards and at other times curved downwards, but more often it is flat as shown in Fig. 17. The gudgeon pin is sometimes made of mild steel, and the surface is then case-hardened in the centre where the connecting rod end bears. At the present time it is quite as common to find gudgeon pins made of special nickel steel or other steel alloys that do not require case-hardening. On the whole these special steels make the best gudgeon pins and stand the hardest wear.

One end of the connecting rod moves up and down with the piston and oscillates (or swings to and fro) on the gudgeon pin, while the other end of the connecting rod travels in a circle, being pivoted at the crankpin and rotating in a circle which has for its centre the centre line of the engine crankshaft. This is clearly indicated in Fig. 18. On the suction stroke of the engine the piston has to be pulled down, as we have already seen; on the explosion stroke the greatest pressure acts on the piston and pushes the connecting rod down. Thus sometimes the connecting rod is being pulled and at other times it is being pushed; in each case it has to overcome the resistance of the engine and drive the car. It is evident, therefore, that the character of the load carried by a connecting rod is just about as complex and dangerous as it is possible for a system of loading to be, and great care has to be taken in the design of such rods to ensure adequate strength without undue weight, as this would tend to keep down the maximum speed of the engine. Another important consideration is the cost of production, and for this reason one often finds it in the form of a phosphor bronze stamping of I section, although the ideal form is a round section of steel with a straight taper from gudgeon pin to crankpin end, and having a hole bored right up the centre to reduce the weight without sacrificing much strength. When the rod is made in the form of a stamping between dies it is possible to turn out great quantities at very low cost and at a very rapid rate, whereas the round steel rods would require to be machined from the solid bar to compete in price with the others. When phosphor bronze is used it is only necessary to bore out carefully and face the bearings at the two ends for the gudgeon pin and crankpin; the bearing at the crankpin end is always formed with a removable cap to facilitate fitting it nicely to the crankpin, journal and also to allow for adjustment as the bearing wears. With steel rods it is necessary to cast a white-metal lining in the crankpin end and then bore it out to form the bearing, but the crosshead bearing is usually formed by a phosphor bronze bush. It is evident, therefore, that the steel rods are more expensive, but they make a splendid mechanical job. A steel connecting rod is shown complete in Fig. 19. Stamped steel rods of I section are also commonly used and are much better and stronger than those made entirely of phosphor bronze.

The Crankshaft, as its name implies, is a shaft with one or more cranks or right-angled bends in it. Its function is to convert the sliding motion of the piston into the rotary motion of the flywheel and revolving shaft. A crankshaft with a single throw (or single crank) is shown in Fig. 20; a four-throw crankshaft is shown in Fig. 21; and Fig. 22 shows how an equivalent motion can be obtained by a single pin fixed into the face of a flywheel. This device (Fig. 22) is frequently used for motor-cycle engines. Crankshafts are always made of steel; sometimes mild steel is used, but more usually special alloys of steel containing chrome, nickel, vanadium, etc., are used. The general practice at the present time is to machine the crankshaft direct out of a solid bar of steel; this requires special jigs for holding the work and special tools for cutting the metal, but is the quickest, cheapest, and most satisfactory method to adopt. A few firms specialize in this class of work with high-grade steel and can supply crankshafts from stock.

It is easily seen by examining Fig. 18 that the crankshaft is being twisted in overcoming the engine resistance, while Fig. 20 shows that the crankshaft is being bent under the push from the connecting rod, so that we say the material of a crankshaft is subjected to combined bending and twisting, and as such a combination is not easy to resist we see now why special steel alloys are required for safety, combined with economy in material and reduction of weight. In Fig. 20 the crankpin is shown at A, the crank cheeks or webs at B₁, B₂ and the crankshaft proper at C. The portions of the crankshaft C which work in the bearings D₁, D₂ are termed journals. A crankshaft must be very stiff and not bend or twist sensibly, otherwise the shaft will vibrate when the engine runs up to speed—which would be very undesirable. It must be perfectly true with all the bearings absolutely in line and the journals well bedded down in their respective brasses (or bearings), otherwise mechanical troubles will arise. Each crank with its crankpin and webs forms a lop-sided or unbalanced mass, so that either (1) each crank must have its own balance weight as in Fig. 23, or (2) special balancing masses must be fitted at each end of the crankshaft. A convenient method of balancing the crankshaft is to have a fan pulley at one end and the flywheel at the opposite end, so that by drilling holes in the faces of these discs an amount of metal may be removed from them sufficient to balance the excess weight of the respective crankpins and webs. In Fig. 24 the shaded area indicates that portion of the crank which constitutes an unbalanced mass. Crankshafts for high-speed engines have always to be very carefully balanced, otherwise the engines will never run satisfactorily, the want of balance being greatly aggravated as the speed of rotation increases. Fig. 25 shows how the crankshaft of a two-cylinder engine may be balanced by drilling holes in the flywheel and fan pulley respectively, but the same effect may be produced by attaching balancing masses—this latter method would, however, be more inconvenient and expensive. The crankpins and journals are ground truly circular after being turned in the lathe as true as possible.

The Flywheel.—We have already described how the force driving the piston of a motor-car engine varies during the four strokes of the cycle, but we must note that it also varies considerably during each individual stroke. Thus, on what is known as the explosion stroke (or power stroke) of the cycle, the pressure at the commencement of this stroke may be exceedingly great and yet towards the end of the stroke the gases have expanded and the exhaust valve has been opened, so that the pressure acting on the piston is then very small. Again, on the compression, suction, and exhaust strokes, the piston has to be pushed or pulled by some means, as no power is being generated. Therefore, if the engine is to be self-acting and run continuously, some means must be provided for storing up the great force of the explosions and giving it out again on the idle strokes. The function of the flywheel is to store any energy given to it over and above that required to drive the car and to give it out again when required for performing the functions of compressing, exhausting, and sucking in gas, as well as to keep the car running steadily. It is simply a heavy wheel mounted on the end of the crankshaft which, when once started revolving at a high speed, is not easily stopped, and which will give up part of its energy each time its speed is reduced owing to the demands of the engine; but when the engine is generating power the wheel will speed up and store the excess—the mere fact that the wheel is heavy causes these changes in speed to occur slowly, and therefore on the whole the fluctuation of speed is not great when a suitable flywheel is fitted. The flywheel does not limit the maximum speed of the engine, as it could go on slowly increasing in speed if no resistance was encountered until the wheel finally burst or flew to pieces. Thus the flywheel does not regulate the speed of the engine; it merely smooths out the inequalities in the several strokes of the “cycle.” Flywheels of motor-car engines are now always made of steel, so that they can be safely run at speeds up to 3,000 revolutions per minute without fear of the rim bursting. All parts of the rim tend to fly off radially in the direction of the arrows as shown in Fig. 26 under the action of centrifugal force. A built-up flywheel is shown in Fig. 27, and one made from a single stamping of steel is shown in Fig. 28. Generally speaking, when a coned clutch is fitted one portion of it is formed on the inside of the flywheel rim as indicated in these two figures. When the construction shown in Fig. 28 is adopted the lining would be inserted after the clutch cone had been put into place; very often the lining is made up of sections which can be readily inserted or withdrawn after the cone is in position.

CHAPTER IV
THE VALVES

Poppet Valves.—Valves are provided for the purpose of controlling the admission of the mixture to the cylinder and also for controlling the exhaust or ejection of the burnt gases at the end of the firing stroke. The most common form of valve is the mushroom or poppet type of valve shown in Fig. 29, in which A is the valve head, B is the valve stem, C is the valve seating, and D is the cotter hole for the cotter E. It will be seen that the general appearance of the valve is a disc of steel with a fine stem to it similar to a mushroom in general outline—hence its name. The valve has a coned face which is kept pressed down on a coned seating by means of the pressure of a powerful spring F acting on the washer G, which bears against the cotter E and thus presses down the valve stem. To ensure that the valve shall always come down correctly on its seating and make a gas-tight joint, the valve stem guide M is provided.

The cam H raises the valve off its seat at the required instant when the motion of the camshaft brings the cam under the roller R. The cam lifts the roller vertically and with it the tappet or push rod K, which slides vertically upwards in the guide P and lifts the valve. The tappet is provided with an adjustable head S kept in position by the locknut T. To adjust the clearance between the head of the tappet and the underside of the valve stem the locknut T must first be slackened back and then the head S can be screwed up or down as desired, the best clearance being about 1/64 of an inch; the locknut is then tightened down again. During this operation the valve must be down on its seat. Sometimes to reduce the noise arising from the tappet striking the valve stem, the head of the tappet is padded with some material such as hard vulcanite fibre, but this wears down more quickly than steel and requires frequent adjustment. The latest device for reducing the noise arising from the valve mechanism consists in totally enclosing the valve gear and springs either by metal plates bolted to the cylinder casting or by extending the crankchamber to cover it all in, and then it is certain to be well lubricated. The exhaust valve is always liable to give trouble either from leakage or seizure or other causes due to the great heat of the exhaust gases, so that the valves are often made now of tungsten steel alloy which is not much affected by heat. If a mushroom type valve leaks it can be ground in and made a tight fit on its seating, provision usually being made for this in the form of a slot cut in the valve head, as shown in Fig. 32, for the insertion of a screwdriver or special tool. To grind in a valve, remove the cap Q by unscrewing it, raise the spring F by pushing up the washer G and then withdraw the cotter E. Lift out the valve and smear the coned face with fine emery powder and oil (or water). Put the valve back and turn it to and fro on its seating by means of the screwdriver, keeping a firm pressure down on it; continue the operation until by examining the valve you ascertain that it touches on the seating all the way round, then couple up the spring again, after carefully removing all traces of the emery powder.

Sleeve Valves.—Another form of valve which has come very much into favour is the sleeve valve, two views of which are shown in Figs. 30 and 31. In this case the gases enter the cylinder through ports or slots P cut in the cylindrical cast iron sleeves S₁, S₂, which are placed between the piston K and the walls of the water-jacketed cylinder C. These sleeves are moved up and down inside the cylinder, while the piston travels up and down inside the inner sleeve S₂ just as though it constituted the cylinder C. Some engines have two sleeves, as shown in the figure, but others have only one sleeve, and there is very little to choose between the two types on the score of efficiency. The great claim made for the sleeve valve is that it is almost noiseless in action and gives very much fuller openings for inlet and outlet of the gases. The piston has the usual number of packing rings to keep it gas-tight, and there is also a deep packing ring provided in the head of the cylinder H to keep the sleeve S₂ gas-tight and prevent loss of compression pressure. The head of the cylinder is usually detachable, and has often separate water connexions in the form of pipes leading from the cylinder jackets. The sleeves receive their reciprocating motion from eccentrics and rods attached to pins shown at the bottom right-hand corner of each sleeve. It might be expected that the sleeves would get very hot or very dry and seize up, or the piston might seize, but in actual practice this has not occurred to any great extent, and on the whole they have been very successful. It is, however, necessary to keep the engine well lubricated, especially when the sleeves get worn, as the oil prevents loss of gas by leakage past the sleeves and piston. In Fig. 31 the two sleeves have come together in such a position that the ports coincide with the exhaust ports cut in the cylinder walls and therefore the exhaust is full open, and as the sleeves travel at times in opposite directions quick opening and closing of the ports is secured. The cylinder head is held down to the cylinder casting by screws or bolts and can be readily detached for cleaning or inspecting the interior of the cylinder. The great objection raised against the sleeve valves is their excessive weight and the unmechanical manner in which they are operated.

The Camshafts and Eccentric Shafts.—These are usually made from the same material as the crankshaft and machined from the solid bar, the projecting cams or eccentrics being afterwards cut to the correct shape. In the case of a camshaft it is very important that the shape of the cams should be such that they lift the valves quickly off their seats to the full extent of their opening (or lift), keep them open for as long a period as desirable, and then allow them to close quickly but without shock. Cams which have straight sides are more in favour than those with curved sides, but if the action of the cams is to be theoretically correct the side of the cam should be curved in such a manner that the valve is lifted at first with a uniformly increasing speed and afterwards with a uniformly decreasing speed, so that it will be at rest in its top position. If this is not done the valve tappet may jump a little above the cam each time the valve is lifted. In Fig. 33 the cam A is intended for the inlet valve and the cam B for the exhaust valve, the essential difference being that the exhaust valve must be kept open longer than the inlet valve, and therefore the exhaust valve cam is the wider of the two. The timing of the inlet and exhaust valves of an up-to-date engine may be explained by considering the crankpin circle as divided into 360 parts or degrees. If there were no lag or lead in the opening of the valves, then they would open when the crank was on its dead-centre and close when the crank was on its dead-centre. The inlet valve would open when the crank was on its top dead-centre and close when it had reached its bottom dead-centre, this representing the suction stroke of the engine. Then would follow compression and explosion, giving two strokes or one revolution before the exhaust valve commenced to open. The exhaust valve would then open when the crank was on its bottom dead-centre and close when the crank reached its top dead-centre corresponding to the completion of the exhaust stroke. It is very important that the pressure of the gases above the piston when it commences to move upwards on the exhaust stroke should be as low as possible, and this can only be secured by opening the exhaust valve towards the end of the explosion or power stroke, thus allowing the bulk of the gases to escape and leaving the piston with little resistance to encounter on its upward exhaust stroke. Therefore we give the exhaust valve a lead of about 30 degrees, which means that it begins to open when the engine crank is 30 degrees from the bottom dead-centre on the downward explosion stroke, and we give it a lag of about 5 degrees in closing. This means that we keep the exhaust valve open until the crank has moved 5 degrees over the top centre, so that we may fully utilize the momentum of the gases to clear out the cylinder or scavenge it. As the piston moves rapidly up the cylinder on the exhaust stroke it pushes the gases in front of it out through the exhaust opening, but when it gets to the top of its stroke the piston stops and then comes down again for the suction stroke, whereas the gases will tend to keep on moving if they are not unduly restricted in their passage through the exhaust system, so that we can generally obtain some slight advantage by giving the exhaust valve a small amount of lag in closing.

The pressure of the gases in the cylinder after the exhaust valve closes will nearly always be a little above atmospheric pressure, and therefore nothing is gained by opening the inlet valve immediately the exhaust closes—we generally allow an interval of 5 degrees, which means that the total lag of the inlet valve is 10 degrees in opening, or the inlet valve does not begin to open until the crank has moved 10 degrees off its top dead-centre on the downward suction stroke. At the end of the suction stroke the piston will again come to rest before moving up on the compression stroke, but the gases will continue to rush into the cylinder from the carburettor owing to their momentum if we leave the inlet valve open a little longer, hence we generally give it a lag of 20 degrees in closing, which means that the inlet valve does not close until the crank has moved 20 degrees up from the bottom dead-centre on the compression stroke.

The camshaft requires to be well supported in bearings to prevent it from sagging or bending under its load. If the shaft and the cams are not made from nickel steel or high-grade steel alloy, they require to be case-hardened (hardened on the surface) to prevent wear on the surfaces due to the pressure of the valve springs, which is considerable and may reach 100 lb. per valve easily; the same applies to the rollers of the tappets. When sleeve valves are fitted to the engine, eccentric sheaves must be used instead of cams, as no springs are employed. An eccentric sheave with its strap and rod are shown in Fig. 34. The valve shaft or lay shaft is shown at C, and the sheave with the hole bored eccentrically is shown at A, and B is the combined eccentric strap and rod. The pin D operates the sleeve valve, giving it a reciprocating motion in a vertical direction, the angular movement being taken up by the oscillation of the rod about the pin D, which would be fixed into the sleeve. Sometimes a groove is formed round the periphery of the eccentric disc or sheave to keep the strap in position and prevent end movement. As the weight of the sleeves is very considerable, the pin D and the eccentric rod must be well proportioned to prevent breakage or undue wear.

The Timing Wheels.—As there is only one suction stroke and one exhaust stroke in every two revolutions of the engine crankshaft, it will be clear that the camshaft or eccentric shaft must be driven at half the speed of the engine crankshaft. This may be done by the use of two gear wheels or wheels having teeth cut on their periphery, such wheels when used for this purpose being called timing wheels, because the positions of the cams on the camshaft (or the eccentrics on the eccentric shaft) relative to the engine crankshaft when the teeth of the timing wheels are put into mesh determines the timing of the inlet and exhaust valves, i.e., the instant at which they will open or close. A pair of timing wheels is shown in Fig. 35. The pinion A has twelve teeth and is keyed to the engine crankshaft, but the wheel B, which is keyed to the valve shaft, has twenty-four teeth, and hence the valve shaft runs at half the speed of the crankshaft. The wheels shown are spur gears, and the teeth run straight across the rim of the wheel; it is, however, quite common to find wheels with curved or helical teeth, as these run quieter. Sometimes when spur gearing is used, one of the wheels is made of fibre and the other of steel, but when helical gears are used the wheels are generally made from nickel steel of high tensile strength. The finer the pitch of the teeth (i.e. the distance between the centres of consecutive teeth) the quieter the gears will run, but the question of strength and the cost of production must also be considered. The latest practice is to use a silent chain drive; this originated with the introduction of the sleeve valve and eccentric shaft. When chains are used for the timing wheels provision must be made for taking up slack in the chain owing to stretching of the links, and as this cannot be done in the usual manner (by sliding the sprocket wheels further apart) owing to the centres of the crankshaft and the valve shaft being rigidly fixed by the bearings, a small jockey pulley (with teeth on it similar to those on the chain sprocket wheels) is provided attached to a short shaft or spindle, which can be raised or lowered at will, and thus keep the correct tension on the chain. The chain drive must be more expensive and require more attention; moreover, it cannot be so very much quieter in action than good well-cut helical gearing.

Two views of a crankchamber of modern design are shown in Figs. 36 and 37. In these figures A is the top half of the crankchamber which rests upon the chassis or framework of the car, being bolted to an underframe at B and C. The cylinders are attached to the chamber at the flange H by means of studs and nuts. This portion, the top half of the crankchamber, requires to be very strong and stiff, because the upward pressure of the explosions acts on the crown of the cylinder and tends to tear the cylinder off the flange H, while at the same time it exerts a great force on the piston, pushing it downwards and tending to force the crankshaft down out of its bearings. In the best practice the whole weight of the crankshaft is supported from the top half of the crankchamber and is carried on the bearing bolts as shown at S, so that they also receive the downward thrust of the piston and in their turn transmit it to the main casting.

The bottom half of the crankchamber then becomes merely an oil container, or reservoir, and dust cover; it should be so arranged and situated that it may be readily removed for inspection of shaft and bearings from underneath. Sometimes the crankchamber has long arms, which can be attached directly to the side members of the chassis, or it may be supported in the chassis by a tubular cross member.

In Fig. 37 the camshaft is shown at T; the magneto would be carried on the bracket E and driven by gearing from the crankshaft. The facing at G is for the water pump, which, in this case, is intended to be mounted on an extension of the camshaft T. The oil pump would be fixed at F, preferably towards the rear of the engine, so as to secure an adequate supply of oil for the pump when the car is climbing a steep hill. The oil could be drawn off and the reservoir emptied by unscrewing the large plug shown in the centre of D in Fig. 37. The timing wheel housing or casing is shown at Q; the oil ducts and connexions for supplying the main bearings with oil are not shown in these drawings, nor are the inspection openings and covers. The upper half of the crankchamber frequently becomes very hot, due to conduction of heat from the metal of the cylinders, and for this reason it has from time to time been proposed to draw the air supply of the carburettor through the crankchamber to serve the dual purpose of cooling the bearings and heating the air supply to the carburettor; but the idea has not found favour, as there is considerable risk of dust and grit finding its way into the bearings and causing trouble due to abrasion.

CHAPTER V
THE CARBURETTOR AND CARBURATION

A carburettor is a contrivance for supplying an explosive mixture of air and petrol vapour to a petrol engine. Petrol, although a liquid fuel, is a combination of carbon and hydrogen which, when supplied with the necessary air, can be burnt and thus evolve heat, which heat is turned into work inside the engine cylinder. What we have to supply to the engine is really a mixture of air and petrol vapour in certain proportions, such a mixture being often spoken of as carburetted air on account of the carbon contained in it. About two parts of petrol vapour (by volume) are required to every one hundred parts of mixture, or fifteen pounds of air to every pound of petrol vapour (by weight). This carburetted air must be of the required strength and form a homogeneous mixture in the form of a vapour. The problem of carburation consists in forming a mixture of the correct strength and character. Air may be carburetted by passing it over the surface of liquid petrol in a surface carburettor, or by drawing it over or among wicks saturated with liquid petrol as in the wick type of carburettor, but both these methods have been largely superseded by the use of what is now known as a jet or spray type of carburettor, in which the petrol is sprayed from a fine jet and mixes with air which is passing up rapidly round the outside of the jet. In all cases, however, the liquid petrol must be vaporized before entering the engine, and to do this heat must be supplied to the mixture, just as water has to be heated before it can be vaporized and turned into steam. Under ordinary circumstances sufficient heat can be obtained from the incoming air to effect vaporization of the liquid petrol if it issues in the form of a very finely divided spray, but when the demand for mixture, from the engine, is great the air cannot supply the requisite heat without its temperature falling below the vaporization point; hence most carburettors of up-to-date pattern are fitted with a mixing chamber surrounded by a hot-water jacket. The essential features of the carburetting plant are shown diagrammatically in Fig. 38, in which A is the petrol tank fitted with the petrol tap G, to which is coupled the petrol pipe F. Some form of petrol filter as indicated at B should be placed between the tank and the carburettor C. The throttle valve of the carburettor is shown at H, the extra-air valve at E, and the engine induction pipe at D.