How does it work? asks the reader. Centrifugal force[37] is the governing principle. To explain its application we append a sectional illustration (Fig. 191) of Messrs. Burmeister and Wain's hand-power separator, which may be taken as generally representative of this class of machines. Inside a circular casing is a cylindrical bowl, D, mounted on a shaft which can be revolved 5,000 times a minute by means of the cog-wheels and the screw thread chased on it near the bottom extremity. Milk flows from the reservoir R (supported on a stout arm) through tap A into a little distributer on the top of the separator, and from it drops into the central tube C of the bowl. Falling to the bottom, it is flung outwards by centrifugal force, finds an escape upwards through the holes a a, and climbs up the perforated grid e, the surface of which is a series of pyramidical excrescences, and finally reaches the inner surface of the drum proper. The velocity of rotation is so tremendous that the heavier portions of the milk—that is, the watery—crowd towards the point furthest from the centre, and keep the lighter fatty elements away from contact with the sides of the drum. In the diagram the water is represented by small circles, the cream by small crosses.

As more milk enters the drum it forces upwards what is already there. The cap of the drum has an inner jacket, F, which at the bottom all but touches the side of the drum. The distance between them is the merest slit; but the cream is deflected up outside F into space E, and escapes through a hole one-sixteenth of an inch in diameter perforating the plate G. The cream is flung into space K and trickles out of spout B, while the water flies into space H and trickles away through spout A.

THE "HYDRO.,"

used in laundries for wringing clothes by centrifugal force, has a solid outer casing and an inner perforated cylindrical cage, revolved at high speed by a vertical shaft. The wet clothes are placed in the cage, and the machine is started. The water escapes through the perforations and runs down the side of the casing to a drain. After a few minutes the clothes are dry enough for ironing. So great is the centrifugal force that they are consolidated against the sides of the cage, and care is needed in their removal.

Chapter XIX.

THE TANK SYSTEM

is shown diagrammatically in Fig. 192. The boiler is situated at the back of the range, and when a "damper" is drawn the fire and hot gases pass under it to a flue leading to the chimney. The almost boiling water rises to the top of the boiler and thence finds its way up the flow pipe into the hot-water tank A, displacing the somewhat colder water there, which descends through the return pipe to the bottom of the boiler.

Water is drawn off from the flow pipe. This pipe projects some distance through the bottom of A, so that the hottest portion of the contents may be drawn off first. A tank situated in the roof, and fed from the main by a ball-cock valve, communicates with A through the siphon pipe S. The bend in this pipe prevents the ascent of hot water, which cannot sink through water colder than itself. From the top of A an expansion pipe is led up and turned over the cold-water tank to discharge any steam which may be generated in the boiler.

A hot-water radiator for warming the house may be connected to the flow and return pipes as shown. Since it opens a "short circuit" for the circulation, the water in the tank above will not be so well heated while it is in action. If cocks are fitted to the radiator pipes, the amount of heat thus deflected can be governed.

A disadvantage of the tank system is that the tank, if placed high enough to supply all flows, is sometimes so far from the boiler that the water loses much of its heat in the course of circulation. Also, if for any reason the cold water fails, tank A may be entirely emptied, circulation cease, and the water in the boiler and pipes boil away rapidly.

THE CYLINDER SYSTEM

(Fig. 193) is open to neither of these objections. Instead of a rectangular tank up aloft, we now have a large copper cylinder situated in the kitchen near the range. The flow and return pipes are continuous, and the cold supply enters the bottom of the cylinder through a pipe with a siphon bend in it. As before, water is drawn off from the flow pipe, and a radiator may be put in the circuit. Since there is no draw-off point below the top of the cylinder, even if the cold supply fails the cylinder will remain full, and the failure will be discovered long before there is any danger of the water in it boiling away.

Boiler explosions are due to obstructions in the pipes. If the expansion pipe and the cold-water supply pipe freeze, there is danger of a slight accumulation of steam; and if one of the circulation pipes is also blocked, steam must generate until "something has to go,"[38] which is naturally the boiler. Assuming that the pipes are quite full to the points of obstruction, the fracture would result from the expansion of the water. Steam cannot generate unless there be a space above the water. But the expanding water has stored up the heat which would have raised steam, and the moment expansion begins after fracture this energy is suddenly let loose. Steam forms instantaneously, augmenting the effects of the explosion. From this it will be gathered that all pipes should be properly protected against frost; especially near the roof.

Another cause of disaster is the furring up of the pipes with the lime deposited by hard water when heated. When hard water is used, the pipes will sooner or later be blocked near the boiler; and as the deposit is too hard to be scraped away, periodical renewals are unavoidable.

HOW A LAMP WORKS.

From heating we turn to lighting, and first to the ordinary paraffin lamp. The two chief things to notice about this are the wick and the chimney. The wick, being made of closely-woven cotton, draws up the oil by what is known as capillary attraction. If you dip the ends of two glass tubes, one half an inch, the other one-eighth of an inch in diameter, into a vessel of water, you will notice that the water rises higher in the smaller tube. Or get two clean glass plates and lay them face to face, touching at one end, but kept slightly apart at the other by some small object. If they are partly submerged perpendicularly, the water will rise between the plates—furthest on the side at which the two plates touch, and less and less as the other edge is approached. The tendency of liquids to rise through porous bodies is a phenomenon for which we cannot account.

Mineral oil contains a large proportion of carbon and hydrogen; it is therefore termed hydro-carbon. When oil reaches the top of a lighted wick, the liquid is heated until it turns into gas. The carbon and hydrogen unite with the oxygen of the air. Some particles of the carbon apparently do not combine at once, and as they pass through the fiery zone of the flame are heated to such a temperature as to become highly luminous. It is to produce these light-rays that we use a lamp, and to burn our oil efficiently we must supply the flame with plenty of oxygen, with more than it could naturally obtain. So we surround it with a transparent chimney of special glass. The air inside the chimney is heated, and rises; fresh air rushes in at the bottom, and is also heated and replaced. As the air passes through, the flame seizes on the oxygen. If the wick is turned up until the flame becomes smoky and flares, the point has been passed at which the induced chimney draught can supply sufficient oxygen to combine with the carbon of the vapour, and the "free" carbon escapes as smoke.

The blower-plate used to draw up a fire (Fig. 194) performs exactly the same function as the lamp chimney, but on a larger scale. The plate prevents air passing straight up the chimney over the coals, and compels it to find a way through the fire itself to replace the heated air rising up the chimney.

GAS AND GASWORKS.

A lamp is an apparatus for converting hydro-carbon mineral oil into gas and burning it efficiently. The gas-jet burns gases produced by driving off hydro-carbon vapours from coal in apparatus specially designed for the purpose. Gas-making is now, in spite of the competition of electric lighting, so important an industry that we shall do well to glance at the processes which it includes. Coal gas may be produced on a very small scale as follows:—Fill a tin canister (the joints of which have been made by folding the metal, not by soldering) with coal, clap on the lid, and place it, lid downwards, in a bright fire, after punching a hole in the bottom. Vapour soon begins to issue from the hole. This is probably at first only steam, due to the coal being more or less damp. But if a lighted match be presently applied the vapour takes fire, showing that coal gas proper is coming off. The flame lasts for a long time. When it dies the canister may be removed and the contents examined. Most of the carbon remains in the form of coke. It is bulk for bulk much lighter than coal, for the hydrogen, oxygen, and other gases, and some of the carbon have been driven off by the heat. The coke itself burns if placed in a fire, but without any smoke, such as issues from coal.

Our home-made gas yields a smoky and unsatisfactory flame, owing to the presence of certain impurities—ammonia, tar, sulphuretted hydrogen, and carbon bisulphide. A gas factory must be equipped with means of getting rid of these objectionable constituents. Turning to Fig. 195, which displays very diagrammatically the main features of a gas plant, we observe at the extreme right the retorts, which correspond to our canister. These are usually long fire-brick tubes of D-section, the flat side at the bottom. Under each is a furnace, the flames of which play on the bottom, sides, and inner end of the retort. The outer end projecting beyond the brickwork seating has an iron air-tight door for filling the retort through, immediately behind which rises an iron exit pipe, A, for the gases. Tar, which vaporizes at high temperatures, but liquefies at ordinary atmospheric heat, must first be got rid of. This is effected by passing the gas through the hydraulic main, a tubular vessel half full of water running the whole length of the retorts. The end of pipe A dips below the surface of the water, which condenses most of the tar and steam. The partly-purified gas now passes through pipe B to the condensers, a series of inverted U-pipes standing on an iron chest with vertical cross divisions between the mouths of each U. These divisions dip into water, so that the gas has to pass up one leg of a U, down the other, up the first leg of the second pipe, and so on, till all traces of the tar and other liquid constituents have condensed on the inside of the pipe, from which they drop into the tank below.

The next stage is the passage of the scrubber, filled with coke over which water perpetually flows. The ammonia gas is here absorbed. There still remain the sulphuretted hydrogen and the carbon bisulphide, both of which are extremely offensive to the nostrils. Slaked lime, laid on trays in an air-tight compartment called the lime purifier, absorbs most of the sulphurous elements of these; and the coal gas is then fit for use. On leaving the purifiers it flows into the gasometer, or gasholder, the huge cake-like form of which is a very familiar object in the environs of towns. The gasometer is a cylindrical box with a domed top, but no bottom, built of riveted steel plates. It stands in a circular tank of water, so that it may rise and fall without any escape of gas. The levity of the gas, in conjunction with weights attached to the ends of chains working over pulleys on the framework surrounding the holder, suffices to raise the holder.

Some gasometers have an enormous capacity. The record is at present held by that built for the South Metropolitan Gas Co., London, by Messrs. Clayton & Son of Leeds. This monster (of which we append an illustration, Fig. 196) is 300 feet in diameter and 180 feet high. When fully extended it holds 12,158,600 cubic feet of gas. Owing to its immense size, it is built on the telescopic principle in six "lifts," of 30 feet deep each. The sides of each lift, or ring, except the topmost, have a section shaped somewhat like the letter N. Two of the members form a deep, narrow cup to hold water, in which the "dip" member of the ring above it rises and falls.

AUTOMATIC STOKING.

The labour of feeding the retorts with coal and removing the coke is exceedingly severe. In the illustration on p. 400 (made from a very fine photograph taken by Mr. F. Marsh of Clifton) we see a man engaged in "drawing" the retorts through the iron doors at their outer ends. Automatic machinery is now used in large gasworks for both operations. One of the most ingenious stokers is the De Brouwer, shown at work in Fig. 198. The machine is suspended from an overhead trolley running on rails along the face of the retorts. Coal falls into a funnel at the top of the telescopic pipe P from hoppers in the story above, which have openings, H H, controlled by shutters. The coal as it falls is caught by a rubber belt working round part of the circumference of the large wheel W and a number of pulleys, and is shot into the mouth of the retort. The operator is seen pulling the handle which opens the shutter of the hopper above the feed-tube, and switching on the 4 h.p. electric motor which drives the belt and moves the machine about. One of these feeders will charge a retort 20 feet long in twenty-two seconds.

A GAS GOVERNOR.

Some readers may have noticed that late at night a gas-jet, which a few hours before burned with a somewhat feeble flame when the tap was turned fully on, now becomes more and more vigorous, and finally may flare up with a hissing sound. This is because many of the burners fed by the main supplying the house have been turned off, and consequently there is a greater amount of gas available for the jets still burning, which therefore feel an increased pressure. As a matter of fact, the pressure of gas in the main is constantly varying, owing partly to the irregularity of the delivery from the gasometer, and partly to the fact that the number of burners in action is not the same for many minutes together. It must also be remembered that houses near the gasometer end of the main will receive their gas at a higher pressure than those at the other end. The gas stored in the holders may be wanted for use in the street lamps a few yards away, or for other lamps several miles distant. It is therefore evident that if there be just enough pressure to give a good supply to the nearest lamp, there will be too little a short distance beyond it, and none at all at the extreme point; so that it is necessary to put on enough pressure to overcome the friction on all these miles of pipe, and give just enough gas at the extreme end. It follows that at all intermediate points the pressure is excessive. Gas of the average quality is burned to the greatest advantage, as regards its light-giving properties, when its pressure is equal to that of a column of water half an inch high, or about 1⁄50 lb. to the square inch. With less it gives a smoky, flickering light, and with more the combustion is also imperfect.

Every house supply should therefore be fitted with a gas governor, to keep the pressure constant. A governor frequently used, the Stott, is shown in section in Fig. 199. Gas enters from the main on the right, and passes into a circular elbow, D, which has top and bottom apertures closed by the valves V V. Attached to the valve shaft is a large inverted cup of metal, the tip of which is immersed in mercury. The pressure at which the governor is to act is determined by the weights W, with which the valve spindle is loaded at the top. As soon as this pressure is exceeded, the gas in C C lifts the metal cup, and V V are pressed against their seats, so cutting off the supply. Gas cannot escape from C C, as it has not sufficient pressure to force its way through the mercury under the lip of the cup. Immediately the pressure in C C falls, owing to some of the gas being used up, the valves open and admit more gas. When the fluctuations of pressure are slight, the valves never close completely, but merely throttle the supply until the pressure beyond them falls to its proper level—that is, they pass just as much gas as the burners in use can consume at the pressure arranged for.

Governors of much larger size, but working on much the same principle, are fitted to the mains at the point where they leave the gasometers. They are not, however, sensitive to local fluctuations in the pipes, hence the necessity for separate governors in the house between the meter and the burners.

THE GAS-METER

commonly used in houses acts on the principle shown in Fig. 200. The air-tight casing is divided by horizontal and vertical divisions into three gas-chambers, B, C, and D. Gas enters at A, and passes to the valve chamber B. The slide-valves of this allow it to pass into C and D, and also into the two circular leather bellows E, F, which are attached to the central division G, but are quite independent of one another.

We will suppose that in the illustration the valves are admitting gas to chamber C and bellows F. The pressure in C presses the circular head of E towards the division G, expelling the contents of the bellows through an outlet pipe (not shown) to the burners in operation within the house. Simultaneously the inflation of F forces the gas in chamber D also through the outlet. The head-plates of the bellows are attached to rods and levers (not shown) working the slide-valves in B. As soon as E is fully in, and F fully expanded, the valves begin to open and put the inlet pipe in communication with D and E, and allow the contents of F and C to escape to the outlet. The movements of the valve mechanism operate a train of counting wheels, visible through a glass window in the side of the case. As the bellows have a definite capacity, every stroke that they give means that a certain volume of gas has been ejected either from them or from the chambers in which they move: this is registered by the counter. The apparatus practically has two double-action cylinders (of which the bellows ends are the pistons) working on the same principle as the steam-cylinder (Fig. 21). The valves have three ports—the central, or exhaust, leading to the outlet, the outer ones from the inlet. The bellows are fed through channels in the division G.

INCANDESCENT GAS LIGHTING.

The introduction of the electric arc lamp and the incandescent glow-lamp seemed at one time to spell the doom of gas as an illuminating agent. But the appearance in 1886 of the Welsbach incandescent mantle for gas-burners opened a prosperous era in the history of gas lighting.

The luminosity of a gas flame depends on the number of carbon particles liberated within it, and the temperature to which these particles can be heated as they pass through the intensely hot outside zone of the flame. By enriching the gas in carbon more light is yielded, up to a certain point, with a flame of a given temperature. To increase the heat of the flame various devices were tried before the introduction of the incandescent mantle, but they were found to be too short-lived to have any commercial value. Inventors therefore sought for methods by which the emission of light could be obtained from coal gas independently of the incandescence of the carbon particles in the flame itself; and step by step it was discovered that gas could be better employed merely as a heating agent, to raise to incandescence substances having a higher emissivity of light than carbon.

Dr. Auer von Welsbach found that the substances most suitable for incandescent mantles were the oxides of certain rare metals, thorium, and cerium. The mantle is made by dipping a cylinder of cotton net into a solution of nitrate of thorium and cerium, containing 99 per cent. of the former and 1 per cent. of the latter metal. When the fibres are sufficiently soaked, the mantle is withdrawn, squeezed, and placed on a mould to dry. It is next held over a Bunsen gas flame and the cotton is burned away, while the nitrates are converted into oxides. The mantle is now ready for use, but very brittle. So it has to undergo a further dipping, in a solution of gun-cotton and alcohol, to render it tough enough for packing. When it is required for use, it is suspended over the burner by an asbestos thread woven across the top, a light is applied to the bottom, and the collodion burned off, leaving nothing but the heat-resisting oxides.

The burner used with a mantle is constructed on the Bunsen principle. The gas is mixed, as it emerges from the jet, with sufficient air to render its combustion perfect. All the carbon is burned, and the flame, though almost invisible, is intensely hot. The mantle oxides convert the heat energy of the flame into light energy. This is proved not only by the intense whiteness of the mantle, but by the fact that the heat issuing from the chimney of the burner is not nearly so great when the mantle is in position as when it is absent.

The incandescent mantle is more extensively used every year. In Germany 90 per cent. of gas lighting is on the incandescent system, and in England about 40 per cent. We may notice, as an interesting example of the fluctuating fortunes of invention, that the once doomed gas-burner has, thanks to Welsbach's mantle, in many instances replaced the incandescent electric lamps that were to doom it.

Chapter XX.

The clepsydra, or water-clock, also of great antiquity, was the first contrivance for gauging the passage of the hours independently of the motion of the earth. In its simplest form it was a measure into which water fell drop by drop, hour levels being marked on the inside. Subsequently a very simple mechanism was added to drive a pointer—a float carrying a vertical rack, engaging with a cog on the pointer spindle; or a string from the float passed over a pulley attached to the pointer and rotated it as the float rose, after the manner of the wheel barometer (Fig. 153). In 807 A.D. Charlemagne received from the King of Persia a water-clock which struck the hours. It is thus described in Gifford's "History of France":—"The dial was composed of twelve small doors, which represented the division of the hours. Each door opened at the hour it was intended to represent, and out of it came a small number of little balls, which fell one by one, at equal distances of time, on a brass drum. It might be told by the eye what hour it was by the number of doors that were open, and by the ear by the number of balls that fell. When it was twelve o'clock twelve horsemen in miniature issued forth at the same time and shut all the doors."

Sand-glasses were introduced about 330 A.D. Except for special purposes, such as timing sermons and boiling eggs, they have not been of any practical value.

The clepsydra naturally suggested to the mechanical mind the idea of driving a mechanism for registering time by the force of gravity acting on some body other than water. The invention of the weight-driven clock is attributed, like a good many other things, to Archimedes, the famous Sicilian mathematician of the third century B.C.; but no record exists of any actual clock composed of wheels operated by a weight prior to 1120 A.D. So we may take that year as opening the era of the clock as we know it.

About 1500 Peter Hele of Nuremberg invented the mainspring as a substitute for the weight, and the watch appeared soon afterwards (1525 A.D.). The pendulum was first adopted for controlling the motion of the wheels by Christian Huygens, a distinguished Dutch mechanician, in 1659.

To Thomas Tompion, "the father of English watchmaking," is ascribed the honour of first fitting a hairspring to the escapement of a watch, in or about the year 1660. He also introduced the cylinder escapement now so commonly used in cheap watches. Though many improvements have been made since his time, Tompion manufactured clocks and watches which were excellent timekeepers, and as a reward for the benefits conferred on his fellows during his lifetime, he was, after death, granted the exceptional honour of a resting-place in Westminster Abbey.

THE CONSTRUCTION OF TIMEPIECES.

A clock or watch contains three main elements:—(1) The source of power, which may be a weight or a spring; (2) the train of wheels operated by the driving force; (3) the agent for controlling the movements of the train—this in large clocks is usually a pendulum, in small clocks and watches a hairspring balance. To these may be added, in the case of clocks, the apparatus for striking the hour.

THE DRIVING POWER.

Weights are used only in large clocks, such as one finds in halls, towers, and observatories. The great advantage of employing weights is that a constant driving power is exerted. Springs occupy much less room than weights, and are indispensable for portable timepieces. The employment of them caused trouble to early experimenters on account of the decrease in power which necessarily accompanies the uncoiling of a wound-up spring. Jacob Zech of Prague overcame the difficulty in 1525 by the invention of the fusee, a kind of conical pulley interposed between the barrel, or circular drum containing the mainspring, and the train of wheels which the spring has to drive. The principle of the "drum and fusee" action will be understood from Fig. 201. The mainspring is a long steel ribbon fixed at one end to an arbor (the watchmaker's name for a spindle or axle), round which it is tightly wound. The arbor and spring are inserted in the barrel. The arbor is prevented from turning by a ratchet, B, and click, and therefore the spring in its effort to uncoil causes the barrel to rotate.

A string of catgut (or a very fine chain) is connected at one end to the circumference of the drum, and wound round it, the other end being fixed to the larger end of the fusee, which is attached to the driving-wheel of the watch or clock by the intervention of a ratchet and click (not shown). To wind the spring the fusee is turned backward by means of a key applied to the square end A of the fusee arbor, and this draws the string from off the drum on to the fusee. The force of the spring causes the fusee to rotate by pulling the string off it, coil by coil, and so drives the train of wheels. But while the mainspring, when fully wound, turns the fusee by uncoiling the string from the smallest part of the fusee, it gets the advantage of the larger radius as its energy becomes lessened.

The fusee is still used for marine chronometers, for some clocks that have a mainspring and pendulum, and occasionally for watches. In the latter it has been rendered unnecessary by the introduction of the going-barrel by Swiss watchmakers, who formed teeth on the edge of the mainspring barrel to drive the train of wheels. This kind of drum is called "going" because it drives the watch during the operation of winding, which is performed by rotating the drum arbor to which the inner end of the spring is attached. A ratchet prevents the arbor from being turned backwards by the spring. The adoption of the going-barrel has been made satisfactory by the improvements in the various escapement actions.

THE ESCAPEMENT.

The spring or weight transmits its power through a train of cogs to the escapement, or device for regulating the rate at which the wheels are to revolve. In clocks a pendulum is generally used as the controlling agent. Galileo, when a student at Pisa, noticed that certain hanging lamps in the cathedral there swung on their cords at an equal rate; and on investigation he discovered the principle that the shorter a pendulum is the more quickly will it swing to and fro. As has already been observed, Huygens first applied the principle to the governing of clocks. In Fig. 202 we have a simple representation of the "dead-beat" escapement commonly used in clocks. The escape-wheel is mounted on the shaft of the last cog of the driving train, the pallet on a spindle from which depends a split arm embracing the rod and the pendulum. We must be careful to note that the pendulum controls motion only; it does not cause movement.

The escape-wheel revolves in a clockwise direction. The two pallets a and b are so designed that only one can rest on the teeth at one time. In the sketch the sloping end of b has just been forced upwards by the pressure of a tooth. This swings the pallet and the pendulum. The momentum of the latter causes a to descend, and at the instant when b clears its tooth a catches and holds another. The left-hand side of a, called the locking-face, is part of a circle, so that the escape-wheel is held motionless as long as it touches a: hence the term, "dead beat"—that is, brought to a dead stop. As the pendulum swings back, to the left, under the influence of gravity, a is raised and frees the tooth. The wheel jerks round, and another tooth is caught by the locking-face of b. Again the pendulum swings to the right, and the sloping end of b is pushed up once more, giving the pendulum fresh impetus. This process repeats itself as long as the driving power lasts—for weeks, months, or years, as the case may be, and the mechanism continues to be in good working order.

COMPENSATING PENDULUMS.

Metal expands when heated; therefore a steel pendulum which is of the exact length to govern a clock correctly at a temperature of 60° would become too long at 80°, and slow the clock, and too short at 40°, and cause it to gain. In common clocks the pendulum rod is often made of wood, which maintains an almost constant length at all ordinary temperatures. But for very accurate clocks something more efficient is required. Graham, the partner of Thomas Tompion, took advantage of the fact that different kinds of metal have different ratios of expansion to produce a self-compensating pendulum on the principle illustrated by Fig. 203. He used steel for the rod, and formed the bob, or weighted end, of a glass jar containing mercury held in a stirrup; the mercury being of such a height that, as the pendulum rod lengthened with a rise of temperature, the mercury expanded upwards sufficiently to keep the distance between the point of suspension and the centre of gravity of the bob always the same. With a fall of temperature the rod shortened, while the mercury sank in the jar. This device has not been improved upon, and is still used in observatories and other places where timekeepers of extreme precision are required. The milled nut S in Fig. 203 is fitted at the end of the pendulum rod to permit the exact adjustment of the pendulum's length.

For watches, chronometers, and small clocks

THE SPRING BALANCE

takes the place of the pendulum. We still have an escape-wheel with teeth of a suitable shape to give impulses to the controlling agent. There are two forms of spring escapement, but as both employ a hairspring and balance-wheel we will glance at these before going further.

The hairspring is made of very fine steel ribbon, tempered to extreme elasticity, and shaped to a spiral. The inner end is attached to the arbor of the balance-wheel, the outer end to a stud projecting from the plate of the watch. When the balance-wheel, impelled by the escapement, rotates, it winds up the spring. The energy thus stored helps the wheel to revolve the other way during the locking of a tooth of the escape-wheel. The time occupied by the winding and the unwinding depends upon the length of the spring. The strength of the impulse makes no difference. A strong impulse causes the spring to coil itself up more than a weak impulse would; but inasmuch as more energy is stored the process of unwinding is hastened. To put the matter very simply—a strong impulse moves the balance-wheel further, but rotates it quickly; a weak impulse moves it a shorter distance, but rotates it slowly. In fact, the principle of the pendulum is also that of the hairspring; and the duration of a vibration depends on the length of the rod in the one case, and of the spring in the other.

Motion is transmitted to the balance by one of two methods. Either (1) directly, by a cylinder escapement; or (2) indirectly, through a lever.

THE CYLINDER ESCAPEMENT

is seen in Fig. 204. The escape-wheel has sharp teeth set on stalks. (One tooth is removed to show the stalk.) The balance-wheel is mounted on a small steel cylinder, with part of the circumference cut away at the level of the teeth, so that if seen from above it would appear like a in our illustration. A tooth is just beginning to shove its point under the nearer edge of the opening. As it is forced forwards, b is revolved in a clockwise direction, winding up the hairspring. When the tooth has passed the nearer edge it flies forward, striking the inside of the further wall of the cylinder, which holds it while the spring uncoils. The tooth now pushes its way past the other edge, accelerating the unwinding, and, as it escapes, the next tooth jumps forward and is arrested by the outside of the cylinder. The balance now reverses its motion, is helped by the tooth, is wound up, locks the tooth, and so on.

THE LEVER ESCAPEMENT

is somewhat more complicated. The escape-wheel teeth are locked and unlocked by the pallets P P¹ projecting from a lever which moves on a pivot (Fig. 205). The end of the lever is forked, and has a square notch in it. On the arbor of the balance-wheel is a roller, or plate, R, which carries a small pin, I. Two pins, B B, projecting from the plate of the watch prevent the lever moving too far. We must further notice the little pin C on the lever, and a notch in the edge of the roller.