SAFETY IN RAILROAD TRAVEL

By H. G. PROUT.

In 1829, when Ericsson's little locomotive "Novelty," weighing two and a half tons, ran a short distance at the rate of thirty miles an hour, a writer of the time said that "it was the most wonderful exhibition of human daring and human skill that the world had ever seen." To-day trains weighing four hundred tons thunder by at seventy-five miles an hour, and we hardly note their passage. We take their safety as a matter of course, and seldom think of the tremendous possibilities of destruction stored up in them. But seventy-five miles an hour is one hundred and ten feet a second, and the energy of four hundred tons moving at that rate is nearly twice as great as that of a 2,000-pound shot fired from a 100-ton Armstrong gun. This is the extreme of weight and speed now reached in passenger service, and, indeed, is very rarely attained, and then but for short distances; but sixty miles is a common speed, and a rate of forty or fifty miles is attained daily on almost every railroad in the country. We cannot tell from the time-tables how fast we travel. The schedule times do not indicate the delays that must be made up by spurts between stations. The traveller who is curious to know just how fast he is going, and likes the stimulus of thinking that he is in a little danger, may find amusement in taking the time between mile-posts; and when these are not to be seen, he can often get the speed very accurately by counting the rails passed in a given time. This may be done by listening attentively at an open window or door. The regular clicks of the wheels over the rail-joints can usually soon be singled out from the other noises, and counted. The number of rail-lengths passed in twenty seconds is almost exactly the number of miles run in an hour.

But if one wants to get a lively sense of what it means to rush through space at fifty or sixty miles an hour, he must get on a locomotive. Then only does he begin to realize what trifles stand between him and destruction. A few months ago a lady sat an hour in the cab of a locomotive hauling a fast express train over a mountain road. She saw the narrow bright line of the rails and the slender points of the switches. She heard the thunder of the bridges, and saw the track shut in by rocky bluffs, and new perils suddenly revealed as the engine swept around sharp curves. The experience was to her magnificent, but the sense of danger was almost appalling. To have made her experience complete, she should have taken one engine ride in a dark and rainy night. In a daylight ride on a locomotive, we come to realize how slender is the rail and how fragile its fastenings, compared with the ponderous machine which they carry. We see what a trifling movement of a switch makes the difference between life and death. We learn how short the look ahead must often be, and how close danger sits on either hand. But it is only in a night ride that we learn how dependent the engineer must be, after all, upon the faithful vigilance of others. We lean out of the cab and strain our eyes in vain to see ahead. The head-light reveals a few yards of glistening rail, and the ghostly telegraph poles and switch targets. Were a switch open, a rail taken up, or a pile of ties on the track, we could not possibly see the danger in time to stop. The friendly twinkle of a signal lamp, shining faintly, red or white, tells the engineer that the way is blocked or is clear, and he can only rush along trusting that no one of a dozen men on whom his life depends has made a mistake.

When one reflects upon the destructive energy which is contained in a swiftly moving train, and sees its effects in a wreck; when he understands how many minute mechanical details, and how many minds and hands must work together in harmony to insure its safe arrival at its destination, he must marvel at the safety of railroad travel. In the year 1887, the passengers killed in train accidents in the United States were 207; those injured were 916. The employees killed were 406, and injured 890.[20] These were in train accidents only, it must be remembered, and do not include persons killed at crossings, or while trespassing on the track, or employees killed and injured making up trains. As will be seen later, the casualties in these two classes are much greater than those from train accidents. The total passenger movement in 1887 was equal to one passenger travelling 10,570,306,710 miles. That is to say, a passenger might have travelled 51,000,000 miles before being killed, or 12,000,000 miles before being injured. Or he might travel day and night steadily at the rate of 30 miles an hour for 194 years before being killed. Mark Twain would doubtless conclude from this that travelling by rail is much the safest profession that a man could adopt. It is unquestionably true that it is safer than travelling by coach or on horseback, and probably it is safer than any other method of getting over the earth's surface that man has yet contrived, unless it may be by ocean steamer. If one wants anything safer he must walk.

In considering the means that have been adopted to make railroad travel safe, it must be remembered that there are very few devices in use that are purely safety appliances. Nearly everything used on a railroad has an economic or mechanical value, and if it promotes safety that is but part of its duty. The great source of safety in railroad working is good discipline. Of all the train accidents which have happened in the United States in the last sixteen years, nearly ten per cent. were due to negligence in operation, and seventeen per cent. were unexplained. Of these no doubt many were due to negligence, and many that were attributed to defects of track and equipment would have been prevented, had men done their duty. The value of mechanical appliances for safety is perhaps as often overrated as underrated. Undoubtedly the best, and in the long run the cheapest, practice will be that which combines in the highest degree both elements—disciplined intelligence and perfection of mechanical details.

First in importance among the mechanisms which demand attention here is the brake. From the beginning of railroads the necessity for brakes was apparent, and in 1833 Robert Stephenson patented a steam driver-brake (the brake on the driving-wheels). This was but four years after the Rainhill trials, which settled the question of the use of locomotives on the Liverpool & Manchester Railroad. This early brake contained the principle of the driver-brake, operated by steam or air, which has in late years come into wide use. The apparatus is so simple that the cut representing it hardly needs explanation. Admission of steam into the cylinder raised the piston, which through a lever and rod raised the toggle-joint between the brake-blocks and forced them against the treads of the wheels. Essentially the same method of applying the retarding force can now be seen on most passenger engines, and often, but not so commonly, on engines for freight service. For various reasons Stephenson's driver-brake did not come into use.

Innumerable devices for car-brakes have been invented, but they divide themselves into two groups: those in which the retarding force is applied to the circumference of the wheel, and those in which it is applied to the rail. The class of brakes in which the retarding force is applied to the rail has been little used, although various contrivances have been devised to transfer a portion of the weight of the car from the wheels to runners sliding on the rails. There are many objections to the principle, and it will probably never again be seriously considered by railroad men. The apparatus is necessarily heavy, the power required to apply it is great, and its action is slow. When brought into action it is not as efficient as the brake applied to the tread of the wheels, and the transfer of the load increases the chance of derailment.

Many different devices have been used to apply the brake-shoes to the wheels, and various sources of power. Hand-power brakes have been used, worked by levers, or by screws, or by winding a chain on a staff; or, in still other forms, springs wound up by hand are released and apply the brakes by their pressure. The momentum of the train has been employed to wind up chains by the rotation of the axles. This is the principle of the chain-brake, very much used in England. This same source of power has been utilized by causing the drawheads, when thrust in as the cars run together, to wind up the brake-chains. Hydraulic pressure has been used in cylinders under the cars; and finally air, either under pressure or acting against a vacuum, has been found to be the most useful of all means of operating train-brakes. Early forms of hand-brakes are seen in the illustrations of some old English cars. The coach shows a hand-brake operated by a screw and system of levers. By turning a crank the guard puts in operation the system of levers which apply the brake with great force; but the operation is slow. The common hand-brake of the United States is too well known to need illustration. With this brake a chain is wound around the foot of a staff, and the pull of this chain is transmitted by a rod to the brake-levers. This apparatus is simple, and when a train is manned by a sufficient number of smart brakemen it is capable of doing good service. This simple form of hand-brake will probably be used in freight-car service until it is replaced by air-brakes, and the various forms of chain and momentum brakes do not appear likely to be much more used in the future than they have been in the past. Therefore, no further space will be given to them.

The expression, electric brake, is now often heard, and requires a word of explanation. There are various forms of so-called electric brakes which are practicable, and even efficient, working devices. In none of them, however, does electricity furnish the power by which the brakes are applied; it merely puts in operation some other power. In one type of electric brake the active braking force is taken from an axle of each car. A small friction-drum is made fast to the axle. Another friction-drum hung from the body of the car swings near the axle. If, when the car is in motion, these drums are brought in contact, that one which hangs from the car takes motion from the other, and may be made to wind a chain on its shaft. Winding in this chain pulls on the brake-levers precisely as if it had been wound on the shaft of the hand-brake. The sole function of electricity in this form of brake is to bring the friction-drums together. In a French brake which has been used experimentally for some years with much success, an electric current, controlled by the engine-driver, energizes an electro-magnet which forms part of the swinging-frame in which the loose friction-pulley is carried. This electro-magnet being vitalized, is attracted toward the axle, thus bringing the friction-drums in contact. In an American brake lately exhibited on a long freight train, a smaller electro-magnet is used, but the same end is accomplished by multiplying the power by the intervention of a lever and wheel. The other type of so-called electric brake is that in which the motive power is compressed air, and the function of the electric device is simply to manipulate the valves under each car, by which the air is let into the brake-cylinder or allowed to escape, thus putting on or releasing the brakes. All of these devices have this advantage, that, whatever the length of the train, the application of the brakes is simultaneous on all the wheels, and stops can be made from high speed with little shock. Up to two years ago it seemed as if this advantage might be a controlling one, and compel the introduction of electric brakes for freight service. Since then the new "quick-acting" form of the air-brake has been developed, by which the brakes are applied on the rear of a fifty-car train in two seconds, and there is no longer any necessity to turn to other devices. It is doubtful, therefore, if the additional complication of electricity is widely introduced into brake mechanism for many years, if ever.

It is now universally held that the brake, both for freight and for passenger service, must be continuous; that is, it must be applied to every wheel of every car of the train from some one point, and ordinarily that point must be the engineer's cab. With the valve of an efficient continuous brake constantly under his left hand, the engine-driver can play with the heaviest and fastest train. Without that instrument his work is far more anxious, and much less certain.

The continuous brake which to-day prevails all over the world, is the automatic air-brake. In the United States much the largest part of the rolling stock used in passenger service is equipped with the Westinghouse automatic brake. A few roads peculiarly situated use the Eames vacuum-brake. That brake is used on the elevated roads of New York, and on the Brooklyn bridge roads. The Westinghouse brake is also largely used in England, on the Continent of Europe, in India, Australia, and South America. In the United States it is being rapidly applied to freight cars also. This brake, therefore, being the highest development of the automatic air-brake, and the one most widely used, will be briefly described, as best representing the most approved type of the most important of all safety appliances.

The general diagram which is given on pages 196–97 shows all of the principal parts as applied to a locomotive, a tender, and a passenger car. The diagram is reduced from one prepared by Mr. M. N. Forney for a new edition of his "Catechism of the Locomotive." In the plan view are shown very clearly the hand-wheels, the chains, the rods, and the levers by which the brake is applied by hand. In passenger service the hand-wheels are rarely used, but they are retained for convenience in switching cars in the yard, and for those rare emergencies in which the air-brakes fail. Under the middle of the car the ordinary pull-rod of the old hand-brake is cut and two levers are inserted. One lever is connected with the brake-cylinder, and the other with the piston which slides in that cylinder. When air is admitted to the cylinder the piston is driven out, and the brakes are applied exactly as they would be were the chains wound up by turning the hand-wheels. Compressed air is supplied to the cylinder from the reservoir near it, in which pressure is maintained at from 70 to 80 pounds per square inch by a pump placed on one side of the locomotive. The pump fills the main reservoir on the engine, and also the car-reservoirs, by means of the train-pipe which extends under all the cars. When the brakes are off there is a full pressure of air in all of the car-reservoirs and train-pipes. It is a reduction of the pressure in the train-pipes which causes the brakes to be applied.

This fact must be borne in mind, for it is on this principle that the automatic action of the brakes depends. If a train parts, or if the air leaks out of the train-pipe, the brakes go on. This automatic principle is a vital one in most safety appliances, and it is secured in the case of the air-brake by one of the most ingenious little devices that man ever contrived, that is, the triple valve, which is placed in the piping system between the brake-cylinder and the car-reservoir. This triple valve has passages to the brake-cylinder, to the car-reservoir, to the train-pipe, and to the atmosphere. Which of these passages are open and which are closed depends upon the position of a piston inside of the triple valve, and the position of that piston is determined by the difference in air-pressure on either side of it. Thus, when the pressure in the train-pipe is greater than that in the car-reservoir, the triple valve piston is forced over, say to the left, a communication is opened from the train-pipe to the car-reservoir, and the air pressure in the latter is restored from the main reservoir on the locomotive. At the same time a passage is opened from the brake-cylinder to the atmosphere, the compressed air escapes, the brake-piston is driven back by a spring, and the brakes are released. If the pressure in the train-pipe is reduced, the triple-valve piston is driven to the right (we will assume) by the pressure from the car-reservoir, the passage to the atmosphere is closed, air flows freely from the car-reservoir to the brake-cylinder, and the brakes are applied.

The function of the engineer's valve is to control these operations. Naturally the runner's left hand rests on this instrument, which is fixed to the back head of the boiler. To apply the brakes he turns the handle to such a position as to allow air to escape from the train-pipe; to release, he turns it to allow air to pass from the main or locomotive reservoir into the train-pipe, and thence into the car-reservoir. It is hardly necessary to say that the operation of the brake, which has been described for one car, is practically simultaneous throughout the train. The brakes on the driving-wheels of the engine are also automatically applied at the same time as those of the cars and the tender.

In the plan on page 197 the several different positions of the handle of the engineer's valve are indicated, and among them the service-stop and the emergency-stop positions. The quickness of the stop can be to some degree controlled by the rapidity with which the air-pressure in the train-pipe is reduced. To make a stop in the shortest possible time, the runner moves the throttle lever with his right hand and shuts off steam, and with his left hand moves the handle of the engineer's valve to the emergency position, then pulls the sand-rod handle to let sand down to the rails, and finally, if the engine is not fitted with driver-brakes, he must reverse the engine and again open the throttle. These movements must be made in order and with precision; and to make them instantly and without mistake in the face of sudden danger requires coolness and presence of mind. It sometimes happens that an engine-runner reverses his engine before shutting off steam, in which case the cylinder-heads will very likely be blown out and the engine be instantly disabled. Then, if there are no driver-brakes, the locomotive is worse than useless, for instead of aiding in making the stop, its momentum adds to the work to be done by the train-brakes. Again, if the air-pressure in the brake-cylinders is so high, and the adjustment of the levers such that an instant application of the full pressure will stop the rotation of the wheels, and cause them to slide on the rails, the stop will take longer than if the wheels continued to revolve. The maximum braking effect is obtained when the pressure on the wheels is as great as it can be without causing them to slide, and it may happen that a quicker stop can be made by putting the engineer's valve to the service-stop position than by trying to make an emergency-stop. The runner must, therefore, be familiar with the special conditions of his brakes, and must have that kind of mind which can be depended upon to work clearly and quickly in a moment of tremendous responsibility. Fortunately, such minds are not very rare. The world is full of heroes who want only discipline, habit, and opportunity.

The pressure of air in the main reservoir and the train-pipe is maintained by the air-pump on the locomotive, the speed of which is automatically regulated by an ingenious governor. It is the throbbing of this vigilant machine which one hears during short stops at stations. The air-pressure has been reduced in applying the brakes, and the governor has set the pump at work.

All of those parts of the air-brake apparatus which are shown in the diagram (pp. 196–97) can be easily seen on a train standing at a station; but the curious traveller must be careful not to mistake the gas-tank carried under some cars for the car-reservoir. The gas-tank is about eight feet long; the car-reservoir is about thirty-three inches.

Although the air-brake can almost talk, it is still not perfect. There are several fortunes to be made yet in improving it. For instance, it is desirable, in descending long and steep grades, that the brake-pressure should be just sufficient to control the speed of the train, and should be steadily applied; otherwise the descent will be by a succession of jerks which may become dangerous. With the automatic the brakes must be occasionally released to recharge the reservoirs, or when the speed of the train is too much reduced; and it is difficult to keep a uniform speed. So far, the means devised to overcome this difficulty and keep a constant and light pressure on the wheels have been thought too costly or complicated for general use. With hand-brakes long trains are controlled by the brakes of but a few of the cars in any one train. It follows that in the descent of grades the braked wheels must often run for miles with the pressure as great as it can be without sliding the wheels. The rim of the wheel is rapidly heated by the friction of the brake-shoe, and the unequal expansion of the heated and the unheated parts of the wheel causes a fracture. This is why so many broken car-wheels are found at the foot of grades—of all places the worst for such an accident to happen. With "straight air," that is, with the pressure from the main reservoir, or the air-pump, going directly to the brake-cylinder, the engineer can apply the brakes to all the wheels of his train simultaneously, and with great delicacy of graduation; and by turning a three-way cock which is placed in the piping of each car, the air can be used "straight." This is regularly done on some mountain-roads. At summits the trains are stopped and the brakes are changed from "automatic" to "straight." This practice is dangerous, however, and is not approved by the best brake-experts, for if a hose bursts, or through some other accident the air in the train-pipe escapes, the brakes are useless. The automatic arrangement by which a reduction of air-pressure in the train-pipe applies the brakes, as previously explained, is much preferred, although no entirely satisfactory means has yet been devised for automatically regulating the air-pressure in the brake-cylinder.

There is not space here to enter into the history of the air-brake. It was first practically applied to passenger trains in 1868. The first great epoch in its subsequent development was the invention, by Mr. George Westinghouse, Jr., of the triple valve. The introduction of the triple valve at once reduced the time of full application of the brake for a ten-car train from twenty-five seconds to about eight seconds. This means, at forty miles an hour, a reduction by more than one thousand feet in the distance in which a train can be stopped. The next great epoch in the history of the air-brake was made by the celebrated Burlington brake-trials of 1886 and 1887. These trials were undertaken by a committee of the Master Car-builders' Association, to determine whether or not there was any power-brake fit for freight service. For general freight service the brake must be capable of arresting a very long train, with cars loosely coupled, running at a fair average passenger speed, without producing objectionable shocks in the rear of the train. The two series of trials were carried out in July, 1886, and May, 1887. The competing brake-companies brought to the trials trains of fifty cars each, equipped with their devices. Skilled mechanical engineers from various railroad and private companies assisted both years. These trials were most exhaustive, and have contributed more to the art of braking than any that preceded or have followed them. The first year's trials developed the fact that the air-brakes could not be applied on the rear of a fifty-car train in less than eighteen seconds, whereas the head of a train moving twenty miles an hour could be completely stopped in fifteen seconds. The result was that disastrous collisions between the cars of any one train were produced in the act of stopping. Men in the rear cars were thrown down and injured, and much damage was done to the cars. At the end of nineteen days the brake-companies went home to work another year over the new problem. In 1887 they reappeared on the same ground, and in eighteen days proved that no simple air-brakes, as then operated, could prevent disastrous shocks in a long train; but it was shown that by bringing in electricity to actuate the air-valves, the application of the brakes could be made practically simultaneous throughout the train. Mr. Westinghouse, however, during the summer following, made such modifications in the triple valve and in the train-pipe that he succeeded in applying the brakes throughout a fifty-car train in two seconds. That settled the matter. He at once equipped a train of fifty cars, and in October and November, 1887, that train made a journey of about three thousand miles, making exhibition stops at various cities. The journey was a splendid and conclusive demonstration that the air-brake is now a thoroughly efficient and reliable contrivance for freight as well as for passenger service. The result has been a very rapid application of the new quick-acting brake to freight cars. The performance of this train was to railroad men most impressive. A freight train of fifty cars is about one-third of a mile long. To see such a train, running forty miles an hour, smoothly stopped in one-third of its own length, without shock or fuss, was an object-lesson that no one could fail to understand or to remember. Some of the stops made by this train will give a fair notion of the relative power of hand- and air-brakes for quick stops. The following figures are averages of stops made in six different cities. They give the distances run in feet from the instant of applying the brakes till the train was brought to a stand-still:

With twenty cars at twenty miles an hour even shorter stops were made than those recorded above. In the Burlington trials the hand-brake stops, with fifty-car trains at forty miles an hour, were made in from two thousand five hundred to three thousand feet.

The air-brake is somewhat complicated, but the complicated mechanism is strong, has little movement, and is securely protected from dirt and the elements. It is therefore little liable to derangement. It is, however, becoming better understood that brake-gear must be good, and employees carefully instructed in the care and use of the air-brake to get its best results; and in recent years two or three elaborate instruction-cars have been fitted up for the education of the enginemen and trainmen.

Space does not permit more than an allusion to driver-brakes, which are operated by steam and by air. The forms in constant use are made by the Eames, the American, the Westinghouse, and the Beals companies. Nor can much be said here of the water-brake, used to some extent on locomotives working heavy grades. It consists of a simple arrangement of admitting a little hot water, instead of steam, to the cylinders. The engine is reversed and the cylinder-cocks are opened to the air. The cylinders then act as air-pumps, and the retarding effect is due to the back pressure. The use of the water is to prevent overheating of the parts.

Semaphore Signal with Indicators.
(One arm governs several tracks. The number of the track
which is clear is shown on the indicator disk.)

If it is important to have efficient means of stopping trains, it is scarcely less important to have timely information of the need of stopping them. To give such information is the function of signals, which, among safety appliances, must stand next after brakes. Signals fall naturally into two great classes: Those which protect points of danger and govern the movements of engines in yards, and those which keep an interval of space between two trains running on one track. For the protection of switches, crossings, junctions, and the like, signals in immense variety have been used, and, unfortunately, are still used; but in the last ten or fifteen years the semaphore signal has become the general standard in the United States, as it long has been in England. This consists of a board, called the blade or arm, pivoted on the post, and back of the pivot is a heavy casting which carries a colored glass lens, either green or red. On the post is hung a lantern. The danger position is with the blade horizontal. In this position the lens is in front of the lamp, and the light shows red or green, as the case may be. The safety position is with the blade hanging about sixty degrees from the horizontal. In this position the light of the lantern shows white. Red is the universal danger color, and green the color of caution. Therefore, a semaphore signal at a point of danger shows by day a blade painted red, with the end of the blade cut square. At night it shows a red light. At a position some distance from the point of actual danger, but where it is desirable to warn an engine-runner that he is likely to find the danger signal against him, a caution signal is placed. This is a semaphore blade painted green, with the end notched in a V-shape, or, as it is called, a fish-tail. At night this signal shows a green light. There is nothing very remarkable about a piece of board arranged to wag up and down on a pin stuck through a post, but it is wonderful how much of good brains and good breath have been expended in getting these boards to wag harmoniously, and in getting railroad officers to understand that a plain board, having two possible positions, is a better signal than any more complicated form.

The arrangement of a group of signals and switches in such a way that their movements are made mutually dependent one upon the other, and so that it is impossible to make these movements in any but prearranged sequences, is called, in railroad vernacular, "interlocking," and in this sense the word will be used here. Interlocking has become a special art. The objects which it is sought to accomplish by interlocking, and the admirable way in which those objects are attained, may best be understood from an actual example. For that purpose we shall take a double-track junction completely equipped with signals, facing-point locks, and derailing switches (p. 205).

A general view of an interlocking frame was given on page 171 of this volume. Two levers from such a frame are here shown. The normal position of the levers is forward, as lever A. When pulled back, as lever B, the lever is said to be reversed.

Let it be supposed that a main-line train is to be passed eastward in the direction of the arrow B. The first movement of the signalman in the signal-tower would naturally be to lower signals 1 and 2. He attempts to pull over lever 1, but cannot move it, and, in spite of any effort or ingenuity on his part, that signal remains at danger. The reason is that lever 2 when normal locks lever 1 normal. The logic of this will be at once apparent. Clearing signal 1 is an indication to the engineer that the way is clear, and that he may pass the junction at speed. So long as this signal (which, it must be remembered, is a caution signal) stands at danger he knows that he may pass it, but must be ready to stop before he reaches No. 2, the home-signal. Therefore No. 1 must never be lowered till all is arranged for passing the junction at speed. As the signalman cannot lower signal 1, he attempts to lower signal 2. Again he finds that he cannot budge the lever. It is locked by lever No. 3. This lever works a facing-point lock, which must be described just at this point (p. 206).

The front rod of the switch, that is, the rod which connects the points of the two moving rails of the switch, is pierced with two holes placed a distance apart just equal to the throw of the switch. In front of these holes is a bolt which is worked by a lever in the signal-tower. After the switch is set the lock-lever is reversed and the bolt enters one of the holes, thus securely locking the switch in position. There is one other interesting feature of this facing-point lock. It has happened very often that a switch has been thrown under a moving train, splitting the train and derailing more or less of it. This class of accidents is especially likely to happen when train movements are very frequent, and may be prevented by the use of the "detector-bar." This is a bar about forty feet long, placed alongside the rail, and carried on swinging links, like those of a parallel ruler, in such a way that any effort to move the bar lengthwise of the rail must raise it above the top of the rail. This bar is moved by the same lever which moves the locking-bolt. So long as there is a wheel on the rail above the detector-bar it cannot be moved, therefore the locking-bolt cannot be withdrawn, and the switch cannot be moved until the train has passed completely off it.

We left the signalman trying to lower signal No. 2; vainly, because No. 3 lever was still normal and the switch unlocked (Diagram, p. 205). Probably he would not have begun his operations in the bungling way that has been supposed, but would have first reversed lever 3. That locks the switch by the facing-point lock, and locks also switch-lever 4 in the frame in the signal-tower and releases lever 2. Then he reverses lever 2. That locks lever 3 and releases lever 1. Then he reverses lever 1, which locks lever 2. Now the way is made for a train to pass east on the main line, and the signals are clear. The last signal could not have been lowered until the chain of operations was complete; none of the levers can now be moved until lever 1 is again put normal and signal 1 made to show danger. There is one point of great danger in this particular train-movement which has not been mentioned; that is, the crossing of main-line east-bound track B by the branch-line west-bound track C. It will be noticed that with the levers normal, derailing switch 5 is open, and it is impossible for a locomotive to pass beyond it. Lever 5 is interlocked in the tower with lever 4 in such a way that, before 5 can be reversed to let a train pass west from C, lever 4 must be reversed to trap any train on B and turn it down the branch D. It must not be understood that the use of "derailers" is universal. In fact, they are not recommended by the best signal engineers, except in special conditions. In the absence of derailer No. 5, signals 11 and 12 would be interlocked with switch 4, so that, so long as that switch stands open for the main line a clear signal cannot be given to a train coming west on C. It will be noticed that signal 2 carries two semaphores on one post. The upper one is for the main line and the lower one for the branch. Both are operated by one lever, 2, and whether reversing lever 2 lowers the main-line signal or the branch signal depends on the position of the switch. The switch is made to pick out its signal by an ingenious but very simple little arrangement, called a selector, which is placed somewhere in the line of ground connections.

It would be an interesting study, were there space, to follow the possible and proper combinations of movements to pass trains over the various tracks. It will be seen that, by concentrating the levers which move switches and signals in one place and interlocking them, it is made mechanically impossible for a signalman to give a signal which would lead to a collision or a derailment within the region under his control. The only danger at such points is that an engineer may overrun the signals. This description of the objects and the capacity of the system of interlocking is no fancy sketch. The system has been in use for many years, doing just what has been here described, and more. A recent close estimate gave the number of interlocked levers now in use in the United States as about eight thousand, and the number is rapidly increasing. Recent official reports showed that in Great Britain and Ireland there were thirty-eight thousand cases in which a passenger line was connected with or crossed by another line, siding, or cross-over. In eighty-nine per cent. of these cases the levers operating the switches and protecting signals were interlocked.

The example of interlocking which has been given is one of the simplest; the principle is capable of almost indefinite expansion, and any one lever may be made to lock any one or more levers among hundreds in the same frame. The greatest number of levers assembled in any one signal-tower in this country is one hundred and sixteen, at the Grand Central Station in New York. In the London Bridge tower there are two hundred and eighty levers. This is probably the greatest number in any one tower in the world. All of these levers may be more or less interlocked. The same principle is applied to the locking of two levers at a single switch, and to the protection of drawbridges and highway crossings.

The mechanism by which the interlocking is done is strong and comparatively simple, but a detailed description of it seems out of place here. Two levers from a Saxby & Farmer machine are shown on page 204, with lever A normal and B reversed. The locking mechanism is in front of the levers, and is actuated not by the levers themselves, but by their catch-rods. It follows that it is not the actual movement of a signal which prevents the movement of other signals, or of switches, but it is the intention to move that signal. This principle of "preliminary locking" is one of great importance.

Switches and signals are often worked at such distances from the tower that it is impossible for the operator to know whether or not the movement contemplated has taken place. The British Board of Trade does not permit switches to be worked more than 750 feet away. In this country there is no limit, but probably 800 feet is very rarely exceeded. Signals are worked in England up to 3,000 or 3,500 feet very commonly, and they are even worked a mile away, but not satisfactorily. This is with direct mechanical connection, by rod or wire, from the levers. It is obvious that a break in the connections between the lever and the switch or signal might take place, and the lever be pulled over, without having produced the corresponding movement at the far end. The locking mechanism in the tower would not be affected by such an accident, and consequently conflicting signals might be given. Even this contingency is provided against with almost perfect safety. If a signal connection breaks, the signal is counter-weighted to go to danger. The worst that can happen is to delay traffic. If a switch connection breaks, the locking-bolt, in the latest form of facing-point lock, will not enter the hole in the switch-rod, and consequently warning is given in the tower that the switch has not moved. Electric annunciators are often placed in the signal-tower, to show on a board before the operator whether or not the movements of switches and signals have taken place.

Considerable work must be done in the movement of each lever. The ground connections must be put down with great care, as nearly straight and level as may be, well drained, and protected from ice and snow. All of these difficulties have been overcome in a beautiful pneumatic interlocking apparatus which has been introduced within the last two or three years. In this system the motive power is compressed air. Near each switch is a small cylinder, containing a piston which is attached directly to the switch movement. Compressed air admitted to one side or the other of this piston moves the switch one way or the other. But, as it would take some time for the necessary quantity of air to flow from the signal-tower to a distant switch, a small reservoir is placed near the switch, and the air from this reservoir is admitted to one end or the other of the switch cylinder according to the position of a valve. For transmitting the motion from the tower to the valve compressed air might be used, but, as air is elastic, a quicker movement is got by using in the pipes some liquid which does not readily freeze, and which, being practically non-compressible, transmits an impulse given at one end almost instantly to the other. The signals are worked in essentially the same manner as the switches, except that the pneumatic valves are moved by electricity. The tower apparatus of a pneumatic system in the yard of the Pennsylvania Railroad at Pittsburg is shown in the engraving opposite. In the front of the apparatus is seen a rank of small handles, which can be turned from side to side with as much ease as the keys of a piano can be depressed. Turning one of these handles admits compressed air to the end of a pipe containing liquid. Instantly the pressure is transmitted 500 or 1,000 feet to the valve at the switch to be moved. The small levers are interlocked perfectly, and in that particular perform the duties of the ordinary machine. A model of the tracks controlled is placed before the operator, showing the switches and signals, and when a movement is made on the ground it is at once repeated back by electricity and duplicated on the model. This beautiful system is due to the same genius that gave us the perfected air-brake and the triple valve, and is the greatest improvement that has been made in interlocking in the last dozen years.