The American Time Table and Train Order System

The telegraph played a key role in this uniquely North American system of railway operation

Rules of the Road

Road vehicles can steer to the right or left at the command of the driver, and can stop relatively quickly if needed. When vehicles meet on the road, they pass on one side or the other according to local custom. If the road is narrow, they squeeze by, or use turnouts that are provided at frequent intervals. If one driver sees the other in time, one can take the turnout and wait for the other. If not, the vehicle closest to a turnout, or the one with the driver of smaller stature, backs to the nearest passing place. In Scotland, there are still paved roads of this kind, with turnouts marked by white diamonds. When one vehicle comes up behind another in the same direction, it stops or slows until it can overtake at a turnout. When roads are wide enough to take two vehicles side by side, there is no problem with meeting vehicles, and little problem in overtaking. Rules like this have been observed since Roman times, and are still the basis for operation on today's motorways, suitably elaborated, of course, to agree with the urge to boss people around.

The classic rules of the road also applied on horse-powered tramways. Here, the vehicle was guided by the tram plates, so to enter or leave a turnout the driver or his helper had to move the plates to guide the wheels properly. Aside from this, everything was exactly the same as it was on the common road. Before the late 18th century, single wagons were the rule, especially where they had to be controlled on gradients. Trains came later, on coal tramways where brakesmen controlled the wagons as they descended by gravity, or on level tramways in agricultural districts. It was rather difficult to back a train, except by running the line team around the train and pulling from the other end. When steam engines pulled trains of coal, horse-drawn traffic took refuge in the turnouts, warned by the smoke of the slowly approaching locomotive. The Stockton and Darlington of 1825 was this kind of railway. Tramway locomotives ran at 3-4 mph, their drivers often walking beside them. Canals, incidentally, were even slower.

All this changed with the Liverpool and Manchester Railway in 1830, and with the similar railways that rapidly came into existence both in Britain and the United States. The trains were heavy, fast and with poor stopping power. They included passenger, freight and work trains, of different characteristics and speed. Passenger trains ran at 30-40 mph, ten times the familiar speeds. The old rules of the road, if used, would limit trains to low speeds and erratic movement at all times, which was unsatisfactory. To avoid the problem of opposing trains on the same track, two tracks were provided, one for each direction. To avoid the problem of a fast train running into the back of a slower, time tables were issued so that slow trains would not be sent off ahead of faster ones. At stations, where trains stopped, passengers and freight were handled, and trains were made up and broken apart, all trains had to move at low speed and enter the station only when signalled to do so. This system worked very well with the speeds and traffic of the time. Early accidents were almost always due to failure of equipment or the road, not to collisions of trains.

American Railroads

In the United States, this plan was followed as far as possible. A very good example is the Philadelphia and Columbia Rail Road of the State of Pennsylvania. This was a double-tracked line, on which private individuals could move their own wagons. Scheduled passenger service was provided by the company, drawn by steam locomotives. The company would also haul private wagons by locomotive, for an extra fee, on certain schedules. Private wagons had to keep out of the way of the scheduled locomotive services. Other early lines operated in much the same way, and most of the freight service was horse-drawn, as on the Baltimore and Ohio. Private wagons and drivers were troublesome when steam locomotives took over most of the traffic, and were abolished by the end of the 1830's. A relic of this was that most coal traffic was handled in 4-wheeled, private-owner wagons until the 1850's and later, though drawn by the railroad company's locomotives.

It was not possible in the early United States to raise enough money by the sale of stock to build a railway. The company had to issue bonded indebtedness (usually limited to the value of the capital stock), and then service this debt out of the proceeds of operation. Many companies failed to do this at the first attempt, and were sold at foreclosure or reorganized in a few years, after which traffic had picked up enough to support vital functions. American railroads were built into unsettled lands, where traffic prospects were only wild guesses. It is not surprising that American railroads were light single-track lines on wooden bridges, not heavy double-track lines on stone viaducts as in Britain, where traffic was even heavier than anticipated.

Single-track lines presented a very real problem. Although speeds were lower than in England, perhaps 10 mph for freight and 20 mph for passenger, it is only folklore that opposing trains could, as a regular practice, feel their way along single track, and use the nearest siding when they encountered one another. Traffic was often light enough that a single engine could start from a terminus at a scheduled time, take all the traffic that was offered, and work along the line to the other terminus, when it would turn around and do the same thing in the opposite direction. "One engine in steam" is a satisfactory solution to the problem under these conditions. American traffic was light, but soon was not that light, and something had to be done.

The Time Table

What was done was that the time table was created. On it was shown the times of each of the regular trains, the days they ran, and the stations at which opposing trains would meet, or pass by each other, one train taking the siding, which was the designation of the track used for this purpose. Most trains usually stopped at all stations. From the first, it was recognized that some trains were more important than others, meaning that delays to these preferred trains should be minimized. Passenger trains should not be delayed by freight trains, and neither should be delayed by company service trains, such as those carrying wood (fuel) or ballast. Therefore, trains were classified, establishing a pecking order. Trains of one class moved according to their time table schedule disregarding trains of inferior classes. Inferior trains had to be clear of the track when a superior train was due, and stay out of the way until it had passed. This applied to trains in the same direction as well as to opposing trains. Classification of trains greatly reduced delays to passenger trains, and ensured that slow trains would not get in the way of fast trains.

Regular trains were numbered arbitrarily. Later, even numbers for north and east bound trains, odd numbers for west and south bound, were assigned. First class trains were usually passenger trains, second class through freight, third and fourth class local freight trains. A schedule became "due" at the time given in the time table at any station, and was in effect until the train arrived and left, or for 12 hours, whichever came first. The 12-hour rule didn't actually mean one often waited until a train became this late, but merely set a limit for determining whether a certain train had arrived and left. You didn't have to check that a day-old train had ever come. The schedule of a train gave its number, class, and times at stations, usually leaving times. Any train not assigned to a schedule was an extra or wild train. The time table included practically all the passenger trains that were run. Extra freight trains were not rare, though the scheduled freight trains were the backbone of the service.

We'll see that American train operation is a series of superimposed layers, and the time table is one of those layers. It is a fundamental one, providing for the movement of regular trains, and, at least in principle, requires no further interference to carry out that function. The lowest, layer zero, however, on which even the time table rests, is the principle that any train in a position to be run into by another train, in either direction, must be protected by a flagman far enough away to stop any other train before collision occurs. A train can always proceed if a man walks ahead of it at a safe distance with a flag. By flag, we mean an actual red flag, a red lamp by night, a man waving his arms, a red fusee, or a torpedo (detonator) on the track: anything that will be noticed. In American operation, flagging was always used to protect a train from being run into by a following train, and was a common, everyday affair. Other steps were taken to reduce the danger of rear-end collision, but this basic one was never neglected.

Flagging and the time table are a practical and safe way to run a railroad, but not a satisfactory one. There are two things it does not handle well: delays to trains, and extra trains. Delays, in the United States, could be very annoying when trains covered great distances, hundreds of miles, exposed to all the annoyances of traffic and weather. One delayed important train held up everything else on the line. Extra trains were needed when the traffic exceeded what one train could handle, or when unexpected traffic appeared. Company trains, which were almost always unscheduled, had merely stayed out of the way of scheduled trains, and knew vaguely of each other, moving slowly and with flag protection. This meant that they were quite in the way of important extra trains.

Delays were handled by taking away the rights of the delayed train when it became a certain number of minutes late. Then, it became inferior to opposing trains of the same class, and had to wait for them. This got things moving, but what happened when an opposing train was also delayed? When it lost class, things started to freeze up, as stationary trains waited for each other. More elaborate rules were thought up to solve this problem, and it became quite a mental puzzle for a crew to decide whether it had the right to the track or not. The final resort was to proceed behind a flag, at walking speed. Only a few serious errors were made with this system, but people (notably passengers) were generally unhappy about the whole problem.

Extra trains were moved by having a preceding train carry signals on its locomotive to show that an extra train was following. A red flag meant that a train with precisely the same rights was following. The red flag was used for important extras, and ones immediately following the train carrying the signals. If the extra train was less important, or was not immediately following, a white flag was carried instead of a red one. This meant the following train would keep out of the way of opposing regular trains. The white flag warned opposing extras or company trains to keep clear or else to protect themselves with a flag. If two white flags met, someone--a station employee or a flagman specially detailed--would have to make sure that neither of the extras went beyond that point, but waited for each other there. If the necessity for an extra was foreseen, notice of its schedule and precedence could be bulletined at all points, and then it was treated as a regular train. A train could always venture out behind a flag in an emergency.

Time tables depended on time, and time was a definite problem in those days, when reliable watches were expensive. Often only the conductor carried a watch, usually furnished by the company. He compared it first with the Standard Clock at the terminus before departing, and then compared it with the watches of each stationmaster and other conductors as his train proceeded.

We see that the time table could handle delays and extra trains to some degree, but real flexibility was not there. Officials at terminals saw that trains were started on time and handled emergencies as well as possible, but officials at different points could not coordinate their activities, and there were often surprises. Train crews were responsible for determining if they had "right of track" or not, and for acting accordingly. This established a tradition for American train operation. However, the time table system worked, and did not depend on rapid communications, which, indeed, were not available.

The Telegraph Arrives

No telegraph was available anywhere in the United States until 1846. Some of first commercial telegraph lines generally followed railways, some common roads. Mostly, telegraph lines struck out where there were no railways to follow. These lines were often out of service, operated only during the day, the offices were not on railway property, and the operators were not railway employees. Where it was convenient, railway messages were sent as normal commercial traffic. Such messages might be very useful, but could not be relied upon for train movements. However, the value of the telegraph was realized at once.

The New York and Erie constructed a telegraph line beside its tracks as early as 1851, while it was still under construction, encouraged by its General Superintendent, David Mc Callum. The completed line had a telegraph over its entire length, New York to Dunkirk, with offices in its stations and operators that were railway employees. The Pennsylvania Railroad named a Superintendent of Telegraphs, David Brooks, in 1852, and before the line was declared "complete" in 1855, also had a telegraph over its entire length, used for train movements and traffic messages, as was the Erie line. Both companies had to arrange for licensing the Morse patents in order to build their lines. These lines were of better quality than the usual commercial line at the time, and were reliable, operating day and night. It is quite erroneous to think that railway companies refused to use the telegraph after it was present beside their tracks because of some innate conservatism. Nevertheless, in some situations the old methods remained in use where they gave satisfaction.

After 1855, Judge Caton of the Illinois and Mississipi Telegraph Company, and Anson Stager of Western Union, worked out forms of contractual agreement between telegraph and railway companies where the Morse patent rights were made available in exchange for placing telegraph wires beside the tracks. The railway was a much safer place for a telegraph line than the public road, and maintenance was much easier. As railways were built, telegraph lines were moved from their original locations to trackside. By 1860, most railway lines were accompanied by telegraph lines, and the telegraph was used for railway business, including train movements.

Use of the Telegraph

The officials that looked after train movements (Superintendents, Stationmasters and Trainmasters) could now make their wishes felt instantaneously at a distance. They could receive information of delays and make suitable arrangements. If a station wired that a certain train was delayed, the official in charge might ask the stationmaster to hold it until a certain opposing train arrived. Then the official might tell the conductor of a train that the superior train would wait at the station in question, and send the train on its way. Suppose this train was then disabled at a certain point, and the conductor wired the official the news. The official then could wire the station where the superior train would wait, and tell it to proceed, and instruct the disabled train to wait and wire when it was ready to go. The schedule of an extra train could be sent to all stations concerned in a few minutes, and certain trains could be instructed to wait for it at certain points. All of the annoying delays and uncertainty of pure time table operation were avoided. If, for any reason, communication failed, then things could go on under the time table, perhaps not as conveniently, but they could go on.

Certain evils crept insidiously into this picture. First, different officials might issue conflicting orders if they were not in the closest correspondence. No one had an overall view, and everyone responded to local conditions and emergencies. Instructions were given to stationmasters or telegraph operators, who were supposed to pass them on verbally to others. Here was a fruitful source of mistakes and confusion. Employees engaged in the same maneuvers could easily be working under instructions differently worded, and subject to different interpretations, perhaps even issued by different officials. A train might proceed on being told that an opposing train was waiting for it, but the message to the opposing train might never have been received. Things like this happen in ordinary life, but they must not happen when trains are involved.

The Standard Code

Most railway companies united in the General Time Convention to introduce Standard Time for railway operation, which was effective in November 1883. This seemed a good opportunity for concerted effort to establish a recommended code of operating rules. Most companies had had rule books since the 1850's, but practices differed greatly, and the handling of telegraphic orders was not standardized. In many cases, the practices were recognized as hazardous or clumsy, and a need for reform was evident. A Committee on Train Rules was formed with K. H. Wade as chairman, including E. B. Thomas, C. D. Gorham, J. T. Harahan, H. Walters, D. J. Chase, M. S. Belknap, H. B. Stone and R. Pitcairn.

Train rules were adopted 14 April 1887. The basic principles recognized by these rules were not changed in later revisions; only the wording and details of procedure were altered, and rules that proved inessential were dropped. The Standard Code, comprising train and other rules, was adopted in July 1889. The General Time Convention became the American Railway Association, and later the Association of American Railroads. The idea of cooperation had proved valuable, so the organization continued. The Standard Code was only recommended practice. Individual roads could and did differ from it, but generally accepted the Code. Other codes, such as the Uniform Code of Operating Rules, were derived from it to suit the prejudices of groups of companies. The Standard Code was amended at irregular intervals, but usually in April every two or three years until the early 20th century.

The rules were numbered, and arranged in the order of general rules, time and time tables, signals, train movement, and movement by train orders. They were relatively few, and dealt only with the matters at hand, based on precise definitions. The fundamental rule for the protection of trains by flagmen was originally Rule 399. It specified the conditions under which protection was required, and set out the detailed actions of the flagman, specifying what signals he should use, and at what distances they should be placed. The flagman put two torpedoes a rail length apart a half-mile from his train as a warning, and one torpedo at a quarter-mile, as an audible stop signal, where he waited with a red flag, or red and white lamps by night. This is proper protection if trains move at reasonable speeds, the track is level, and the weather is good. Many officials exercised their ingenuity in devising detailed instructions for flagmen that would work under all conditions, and would not require judgment or experience. These "long rules" never proved satisfactory.

Certain deprecated procedures familiar from the early days still survived on some roads in the 1880's, but were rejected by the Standard Code. For example, some roads used green and white flags on an engine to show that an irregular train followed. Green flags by then showed that a section followed, and white flags that the train carrying them was an extra. Combining the two seemed like a good idea to some. The irregular train would keep out of the way of opposing regular trains. Another practice was to run regular trains ahead of their time by train order. This was done, for example, to get all the sections of a train into the terminus by the scheduled time. Yard crews might have to be notified, and all meets arranged by train orders.

Other bad practices were too convenient to deny. The running of sections behind a green flag was one, and the designation of certain stretches of the main track by "yard limits" where operation would be by sight only, clearing passenger trains, which could proceed normally. Allowing yardmasters to create sections of freight trains on their own was another. After 1899, all extras had to be created by the dispatcher. Sections and yard limits persisted, however, and are largely responsible for any defects the Code may have.

In 1899, the rules were renumbered and modernized. Rule 399 became Rule 99, under which number it has become famous. The new Rule 99 was a "short rule" that said only that "when a train stops or is delayed under circumstances in which it may be overtaken by another train, the flagman must go back with stop signals a sufficient distance to ensure full protection." This rule depends on the judgment and experience of the conductor, who instructs his flagman what to do in questionable cases. Torpedoes are now used only in twos as a warning, and the single torpedo stop signal is abolished. At this time, red was defined as the "stop" signal rather than the "danger" signal, while the colors for caution and proceed (previously "safety") were not specified. Some important roads continued to use white for proceed and green for caution. Most, however, had adopted green for proceed, while the color for caution was debatable. Some roads used red and green side by side for caution. Later, of course, yellow was adopted. The change of green to yellow for caution was not complete. Many green flags and green signs that advised caution remained this color. The Pennsylvania completely replaced green by yellow, but retained white in switch targets to avoid confusion with green for proceed.

1915 saw many changes in the Code. Daytime markers, which had traditionally been green flags (for caution!) became yellow. The red fusee, an excellent signal, was recognized as a help to flagging, though it had been around for a while. The green fusee for caution, however, disappeared. Procedures for transmitting train orders by telephone were specified. The use of even hours (such as 4 00 pm) was abolished. The zeroes were often so badly written they could be mistaken for other digits. However, it's never been really clear to me why even hours were prohibited. Bad handwriting should have been prohibited, but never was. The rule that engineman and fireman had to call signals to each other appeared in this year.

The long Rule 99 was restored in 1924, only to be retired again in 1928. It remained a continuing problem, since there is no way that such a rule can be constructed. Protection of the rear by flagging is simply inexpedient for fast trains following at close intervals (or intervals that can become close in certain circumstances). Where it fails, some sort of block system is essential.

Red flags as markers, and the familiar grade crossing signal long-long-short-long are late additions to the Code. Most references to train parted were finally removed, long after they had any significance. When cars had no brakes, a parted train meant the danger of cars rolling freely, so the first order of business was to avoid having it run into you. After air brakes, the parted brake line ensured a quick application on the breakaway part, and a heavy drag on the part with the locomotive, as its feed valve attempted to keep the train line pressure up.

Since the Rules differ in details with time and with different companies, I will assume a kind of general rules environment in what follows, so that the general idea will be communicated. This will save a lot of qualification, that would be tedious and uninformative. In any actual case, one follows the rules in effect, of course.

The Dispatcher

A division was a length of 100 to 150 miles of railway, approximately a day's work for a freight train. Crews and locomotives were usually changed at the division points on through trains, and other trains were assembled and broken down there as well. It was a natural unit of train operation, managed by a Superintendent who was responsible for everything. When the necessity for unity of command was realized, the Superintendent became responsible for issuing train orders. Sometimes the Superintendent even worked at the Morse key, but at least had his own operator available. Since this turned out to be a full-time job, an official, always a qualified operator, was named Dispatcher to assume the responsibilities. A Night Dispatcher handled the night shift, so that there was someone on duty for 24 hours. If several divisions were controlled from the same point, there might be a Chief Dispatcher and a staff of Dispatchers to handle the different divisions. In railroad organization, a division came to be a geographical area, not a single operating division, so operating divisions were called Subdivisions or Districts. Here, I'll continue to use the word division with its earlier meaning.

Every staffed station along the line had an operator, sometimes the only employee, who handled other duties in addition to train orders. Many of these stations were open only days. At certain stations, night operators were also employed, to ensure that there were sufficient open offices through 24 hours. About 20 miles between open offices can be considered an average, but they were more frequent in busy areas, less frequent at night and where traffic was sparse. There were always stations at critical points such as the ends of double track, or junctions. Most stations had a siding, a track used for meeting and passing trains, where a train could stand clear of the main track. There was a train order signal. When this signal was at stop, it meant that a train could not leave the station without the operator's permission, and usually that there were orders for some train. When it was at proceed, it meant that there were no further orders to be delivered, and trains could proceed.

The train order signal was held at stop when there were any orders for any train at the office. It had to be assured that no train was allowed to pass for which there were orders, so each train had to be positively identified. Instead of changing the signal to proceed, a Clearance Card was issued, on which the train was identified. This meant that every train had to stop when the train order signal displayed stop, and report to the office for orders or a clearance card. If their orders were the only ones there, the train order signal could be changed to proceed to allow them to go. Otherwise, a clearance card was necessary.

Each office was cut into a wire that ran the length of the division and included the dispatcher's instrument. Every office could hear all the traffic on the wire, and thereby keep current on developments. Originally, there may have been only one wire, but later there were alternates, and wires devoted to through company messages, as well as a Western Union wire for public messages, for which the operator earned a commission. Each office had a call, usually two letters, which the operator easily recognized. Whenever anyone wanted to send, he opened his key switch and all the sounders clicked, alerting everyone. If the dispatcher had to interrupt some business, he would open his key for a while, called breaking, which would stop everything. Then, he would identify himself and send his message.

The dispatcher kept two essential daily records, the Train Sheet and the Order Book. When a train left at an open office, the operator would report to the dispatcher the train identification and the time of departure (and arrival, if so instructed). The dispatcher then wrote this OS ("on sheet") report in the row or column corresponding to the train in the Train Sheet, which gave him a synoptic view of movements on the division. A more graphic representation of the Train Sheet that was very convenient represented trains by lines on a graph where time ran horizontally and distance (stations) vertically. The Train Sheet was the indispensable tool of the dispatcher. Trains in one direction were shown on one page, trains in the opposite direction on the facing page.

Train orders were composed in the Order Book, required by rule, which could be an ordinary blank book. Each order was dated, and numbered serially for each day. The addresses were shown, first the superior train, whose rights were being restricted, then below the inferior trains, who took advantage of the restrictions, with the offices at which the order was to be delivered. The responses of each operator, and their times, were noted, as well as when the dispatcher transmitted other procedural signs to them. As each operator repeated the train order, the dispatcher underlined the words in the order. When an order was fulfilled, superseded or annulled and thereby became void, this was also recorded. Each order was issued over the intials of the Superintendent, representing authority, or later over the dispatcher's initials.

Train Order Principles

Gradually, certain principles to guarantee safety in the use of telegraphic messages gained general recognition. It would be very interesting to know the details of this development, but the details have been lost, and can only be presumed. It is easiest to state the final principles, and to realize that they were the result of a long process of trial and error. There are six fundamental principles.

The first principle is the unity of command, embodied in the dispatcher. Now that instructions passed directly from the dispatcher to the train crews through the operator, the position of Stationmaster disappeared, or else lost all direct connection with train movements.

The second principle is that all instructions are issued to the men who are to execute them, and in writing. The operator is not told to tell so-and-so this or that, but to prepare an accurate written message so that the person who must act knows exactly what to act upon. Previously, it was the practice (as it is, for example, in the Army) to give the instructions to the senior man present, who then verbally instructed his subordinates. Written messages also produce a "paper trail" that helps the investigation when things go wrong.

The third principle is that the instructions are issued in the same words to everyone involved. This ensures that everyone knows exactly what the others know. A message instructing a train to proceed is also given in exactly the same words to the trains opposing it, who know that they are restricted to exactly the same extent that the other train is helped. Too many times a train was told to wait at C, while another was told to proceed to D, and the trains collided between C and D. This principle is also usually violated in everyday life, where instructions are modified for each participant telling only the minimum.

The fourth principle is that every message is written as it is transmitted, and repeated to ensure accuracy. This means that every copy that will be needed must be made at one time and repeated. If it was necessary to make additional copies, they were traced from the original, and even repeated again, just like the original copies.

The fifth principle is that no message restricting the rights of a train can be acted upon until it is known that the message will be delivered to the train restricted. This originally required that the signature of the conductor of the restricted train be taken on the order and transmitted back to the dispatcher. Only then could the dispatcher send messages taking advantage of the restriction. This extremely onerous requirement was later relaxed when experience showed that delivery of train orders was reliable, and it was only necessary to ensure that the train restricted would be stopped if anything went wrong.

The sixth principle is that messages concerning the movement and safety of trains, and subject to the other principles, should be clearly distinguished from ordinary messages by their appearance. Such messages are then called train orders, while everything else is just a message. Train orders were written on yellow or green tissue paper (so that manifold copies could be made) forms, and all other messages on some other medium, usually common white paper. The carbon paper used to make the copies had carbon on both sides, so that it transferred to the back of the order as well as to the front. Such an order could be clearly read when held up to the light of the firebox door or the caboose stove, since the paper was translucent.

Train Order Delivery

Train orders are addressed to Conductor and Engineman (C&E). Originally, the operator copied the order and repeated it to the dispatcher, the time of repetition being written on the order. Earlier, this was called the OK, given by the dispatcher if the repetition was correct, and acknowledged with "i i OK" by the operator. The conductor called at the office for orders when the train order signal was seen at stop. The operator had the conductor sign the order on the line provided, and the signature was sent to the dispatcher. The dispatcher would record the signature in the Order Book, then send "complete" to the operator, who would write this on the order with the time, then hand copies of the order to the conductor, one for himself, and one for each engineman. The operator would file the copy on which the signature was taken. The conductor would then deliver the orders to his enginemen. An order had to be "complete" before it could be acted upon.

If the dispatcher had to wait until he gave "complete" to an order to a superior train whose rights were restricted thereby, before he could complete orders helping inferior trains, the delay would be very burdensome. This was the original idea, but was dropped even in the first version of the Code. It was found that once an order was repeated, there was very little chance that it would not be delivered before the superior train left the office, being held by the train order signal. The necessity of taking signatures guaranteed this. Rules were relaxed so that the dispatcher could complete the order for the inferior trains, and they could use it, before complete was given for the superior train. The only case where signatures were invariably required was when an order was sent to a train in care of another. Then the signatures were necessary to prove delivery. Originally, orders delivered "in care of" were intended only to help trains from a siding where there was no operator, not to restrict them.

The procedure for delivering an order was time-consuming, especially for heavy freight trains. It required at least a quarter of an hour, and with the time spent stopping and starting, pumping up brakes, walking along the train and so forth, half an hour could be wasted receiving an order. Some orders were such that if they were not delivered, only delay would result, not danger. If these orders could be handed to the conductor and engineman while the train was in motion, a great deal of time could be saved. A new type of train order was introduced to facilitate this. The old form was called a "31" order, and the new form a "19" order. 31 orders were written on yellow tissue, 19 orders on green tissue, so they could easily be distinguished from each other.

The operator would now stand beside the track as the train approached preparing to stop because of the train order signal at stop, perhaps with a green or yellow lantern hanging from the signal post to make it clear that 19 orders were to be delivered, and hand up the orders to the fireman and conductor as the train passed. Better ways than hand-to-hand transfers were soon developed. The orders could be fastened to a hoop made from a flexible withe that was held up by the free end, and the fireman or conductor could put his arm through the hoop. The orders would be removed, and the hoop thrown down so the operator could retrieve it. A much later method used a Y-shaped fork with the orders tied to a string looped around the Y and held by spring catches. The man on the train pulled off the string loop with the orders, while the operator continued to hold the fork. Somtimes the fork could be held in the proper position by a post beside the track.

It was gradually noticed that delivery of 19 orders was no less reliable than the delivery of 31 orders. There was, however, one other matter that had to be considered. When a conductor came to the office, it was easy to make sure what his train was, so that the orders were delivered to the correct train. This was not possible with 19 orders. Also, the train had to pass a train order signal at stop when receiving 19 orders. The clearance card was already in use to allow a train to pass the signal at stop. It showed the train identification, and now a list of the order numbers to be delivered to the train. The engineman and conductor could check the clearance card to ensure that they had the correct orders for their train.

The setting up of meets between opposing trains had required 31 orders. At least one 31 order was always required, and between extras, two. To overcome this inconvenience, the dispatcher was permitted to use 19 orders to set up the meet, provided that he first sent a 19 order to the operator at the station where the trains were to meet. This meant that the order signal at the meeting station would prevent either train from proceeding if either of the 19 orders failed to be delivered. This proved an excellent idea, since it protected against train crews' forgetting orders, and other contingencies, as well as nondelivery of orders. The order sent to the operator at the meeting point was called a "middle order."

As evidence for the reliability of order delivery built up, it became evident that there was really no need for the 31 order at all. Those few cases where signatures were required could be handled simply by taking the signatures on the 19 form. Some companies abolished the 31 order completely, going back to the generic train order of the distant past (but with different rules). Others kept the 31 form, but allowed the free use of the 19 order in most cases.

The original train order signals were rotating targets or simply a lamp, and applied to both directions indiscriminately. Semaphore train order signals, first appearing in the 1880's, usually had two arms, one for each direction, sharing a single lamp and a single post. Each arm displayed "stop" when horizontal, "proceed" when inclined. By night, a red light or a white light was displayed. Now, trains in a direction for which there were no orders could pass the signal without stopping or receiving a clearance card. A green or yellow flag or lamp at a regular location was used to mean that 19 orders were to be delivered without stopping the train. Later, three-position semaphore arms displayed proceed (green), 19 order (yellow) or 31 order (red). The Pennsylvania Railroad always displayed a yellow (earlier green) flag or lamp for train orders, using the block signals to stop or slow trains for delivery. Trains did not usually actually stop at a stop train order signal, as they would at an interlocking or block signal, but according to their convenience.

Using Train Orders

Let's see how a dispatcher would handle running an extra train. Under the Standard Code, all the red and white flags are dispensed with, except that the extra train now carries white flags on its own locomotive, simply to show that it is not a regular train. Suppose the yardmaster calls up and says that engine 475 is ready to go for Z. The dispatcher enters the extra on his Train Sheet, and checks if there are any opposing extras on the division. If not, he writes the order "Engine 475 run extra A to Z" and addresses it to the Conductor and Engineman (C&E) Engine 475 at some office, say A. Before 1906, the wording was "will meet" instead of simply "meet," and this applies to all the orders we will quote below. Orders are numbered to distinguish them, so let's call this Order No. 1. The dispatcher sends the order to the operator at A, or may deliver it himself. When the conductor calls at the office A for orders, he also receives a clearance card initialed by the dispatcher, as well as his "running order." The purpose of the clearance card is to ensure that all orders for the train are delivered. Originally, a train had to have either an order or a clearance card to start from its initial station, but later the clearance was required. If A to Z is westward, the train is now officially Extra 475 West, and can proceed, while keeping clear of all regular trains in the time table. The conductor of Extra 475 West must also know before starting that the regular trains that should have arrived and left by this time have actually done so. He cannot just assume that they have. Usually, at a station where trains originate, there is a book where each train records its passage, called the train register, and a check of it will give the information he needs. If there is no register, the dispatcher will provide the necessary information in a train order or message. The conductor informs his engineman of the result of the register check, usually on a standard form. Now, Extra 475 West can set out for Z, knowing that it has "right of track."

Suppose now that the dispatcher receives word that an extra is ready to come east at Z, with engine 1120. Again, he enters it on the Train Sheet, but now notes that Extra 475 is on the road. He now guesses where the trains will meet if they proceed as expected, say at station K. Now he writes the order [Order No. 2] "Engine 1120 run extra Z to A and meet Extra 475 West at K" and sends it on form 31 to the operator at E for delivery to Extra 475 West, which he notes has just left D. At the same time, he sends it as a form 19 to Z. E repeats the order, then Z, to whom he gives "complete." The operator at Z delivers the order to Engine 1120, creating it Extra 1120 East. After checking the register, the extra can leave. Extra 475 West finds the train order signal at E at stop. The conductor goes to the office and, after reading the order, signs it, and his signature is sent to the dispatcher. The dispatcher then tells the operator at E that the order is "complete" and the operator furnishes a copy for both the conductor and the engineman. Order No. 1 is not delivered to Extra 1120; it is enough that the opposing extra is mentioned in Order No. 2.

A better procedure would be for the dispatcher to send the order first as a form 19 to the operator at K, and give him complete. Now he sends the order as form 19 to E and Z, and gives complete in that order. Extra 475 would not have to be stopped at E for the delivery of the order. Whether a 19 or a 31, the operator at E would not accept the order if Extra 475 had passed his station, or if it was at all doubtful if he could stop it. If Extra 475 was already at E when the dispatcher sent the order, the operator would get the conductor in his office before he repeated it. Extras 475 and 1120 are now approaching each other on the main track, and know they have a meet at K.

The "meet" order is the fundamental way to run trains in both directions on single track. It can be issued for any trains, whether regular or extra, and has a single, unequivocal meaning. Neither of the trains involved (there are usually only two, but more can be included) can move beyond the meeting point named, and must approach that point under full control. The form of the meet order is specified in the Standard Code, as are the forms of other orders. These forms must be used when applicable, but the dispatcher is not restricted as to what he can order, except that certain types of orders may not be allowed.

The time table directs that of trains of the same class, trains in one direction are superior to those in the opposite direction. Let's suppose eastward trains are superior to westward trains of the same class on this division. Extra 1120 East then proceeds to K, holding the main track because it is in the superior direction, and stops short of the switch at the eastern end of the siding there, called the initial switch. His head brakeman usually opens the switch for the westbound train to enter the siding. This is not just courtesy; it protects his train in case Extra 475 approaches too enthusiastically. Extra 475 West approaches the east switch at K with caution. If Extra 1120 East were not already there, he would open the switch, and proceed to the west end of the siding and wait clear of the main track, closing the switch at the east end when he was through with it. Since Extra 1120 is already there, he moves onto the siding and proceeds to the switch at the west end. When he clears the east end, Extra 1120 closes the switch and proceeds. Extra 475 opens the switch at the west end, returns to the main track, closes the switch, and proceeds. There must never be any uncertainty about which train takes the siding at a meet. Superiority by direction only affects extras with regard to which should take siding, and in no other way.

The order given to Extra 1120 is actually two orders combined. One is the running order (key words, "run extra") and the other is a meet order (key word, "meet"). Here, the combination is a good idea, since it reduces the chance of one train or the other not receiving the meet order. Now suppose Extra 1120 is delayed, and Extra 475 would have to wait an hour or so at K for it. The dispatcher wants to move the meet farther west to R. He writes the order [Order No. 3] "Extra 1120 East meet Extra 475 West at R instead of K." Who should he send the order to first? Extra 1120 is being restricted, since it can proceed only as far as R, not K, so it must receive the order first, say at T. Extra 475 can receive a 19 order at, say K. Although he must approach with care, there is no reason to stop. This is a pure meet order, but contains the key words "instead of" that are used when a previous order is superseded. The instructions pertaining to the meet in the previous order are now void.

I have set things up to show a strange, but necessary and logical, conclusion. Suppose that word comes to the dispatcher that Extra 1120 will not be delayed after all, so the meet at R is not appropriate, and will delay the train unnecessarily. It is possible to annul an order, making it invalid. If the dispatcher writes [Order No. 4] "Order No 3 is annulled," and sends it to the extras before they reach K or R, then what is the outcome? Well, it's not good! Order No. 3 superseded the part of Order No. 2 that referred to the meet at K, so this part of the order became void. Order No. 4 then wiped out Order No. 3 that put the meet at R, so now there is no meet at all! An order is in effect until fulfilled, superseded or annulled. Life cannot be breathed back into an order that has experienced one of these shocks. There is an excellent reason for this; if it were not so, whether an order was revived or not would be subject to analysis and discussion, which is dangerous. Once dead, always dead is the rule. The dispatcher must write "Extra 475 West meet Extra 1120 East at K instead of R" if he wants to restore the previous meeting point.

What do you think would happen if he wrote "Extra 475 West meet Extra 1120 East at K" and stopped there? Now, Extra 475 cannot pass K, since it is ordered to meet there. Extra 1120 cannot pass R, since it was ordered to meet there, and it reaches R before K. An order is not superseded unless the words "instead of" appear explicitly, so Order No. 3 is still valid. Otherwise, one would not know what is being superseded.

Suppose the dispatcher knew beforehand when both extras would be running, and would meet at K around 2 30 pm if everything went well, but that Extra 1120 East had an unknown amount of work to do between Z and K, and could well be late. In that case, he did not want Extra 475 to be unnecessarily delayed, since it was a rather important train. He might arrange things as follows: [Order No. 1] "Engine 475 run extra A to Z with right over Extra 1120 East" and [Order No. 2] "Extra 475 West wait at K until 2 30 pm for Extra 1120 East." These orders would be given to both extras, and Engine 1120 would in addition receive its order to run extra Z to A. The first new form of order here is the "right" order making one extra superior to the other. The same form of order can be used to change time table superiority, or any other. Extra 1120 must keep out of the way of Extra 475. The other new form of order is the "wait" order. If Extra 1120 East could reach K before 3 20 pm (actually 5 minutes earlier to allow for variation in watches and other contingencies) it would enter the siding at the west switch at K. It could proceed no farther, since the right order makes Extra 475 superior to it. Extra 475 West would approach K any time before 2 30 pm prepared to stop and wait at the west switch until that time, or until Extra 1120 arrived. After 2 30 pm, Extra 475 would have the right to proceed without worrying about Extra 1120, which would have to take siding and be in the clear wherever it happened to be between K and Z. Note that Extra 1120 has no right to the track after 2 30 pm, no matter how far it is from K at that time. If Extra 475 arrived at 2 00 pm and found Extra 1120 already in the siding, it could depart at once, and not wait for 2 30 pm. No other train but Extra 1120, mentioned in the order, can take advantage of the waiting time. The wait order does not affect the superiority of the trains. If it were issued alone, Extra 1120 could proceed against Extra 475 at any time. As it is, it cannot proceed against Extra 475 except as helped by the wait order.

We know that superiority means that the inferior train must keep out of the way of the superior train, clear of the main track, as if the superior train could travel at the speed of light. One train can be superior to another by class or direction, specified in the time table, or by train-order. Train order superiority is called right. At one time, "right" was used more generally, but it came to refer only to superiority granted by train order. Should an order state "Extra 475 West has right over No 2 Engine 1005 A to D," then Extra 475 West can proceed to D disregarding No 2's time, while regular train No 2 has to wait at D until Extra 475 arrives. (No 2 is an eastbound regular train moving from Z to A.) Without this order, Extra 475 would have to be in the clear five minutes before No 2 is due at any station. Under the order, if Extra 475 does not find No. 2 at D, it must clear its time at any station beyond. When Extra 475 arrives at D, it must expect to take the siding if No 2 is due, since it becomes the inferior train at that point.

The five minutes we have kept mentioning where time is involved is called "clearance time" intended to cover variations in watches, and to allow a flagman to get out if something goes wrong. An inferior train must clear a superior train by five minutes at meeting and passing points, where the time is given by time table or specified in a train order. The superior train should not have to stop, or even slow down at these points. If the trains are of the same class, both must approach scheduled meeting points prepared to stop, and no clearance time is necessary. Later, five minutes was regarded as too little, considering the increased speed of trains that meant a flagman had to go out farther, and ten minutes became the usual clearance time. One train had to keep at least five minutes, later ten minutes, behind a preceding train. The preceding train was normally either a superior train, or a preceding section of the same schedule.

Time orders have many useful variations. Delayed regular trains can be handled with "run late" orders. For example, "No 2 Eng 1005 run 30 mins late A to D 20 mins late D to M and 10 mins late M to Q." Such an order can be given to opposing trains to help them proceed against No. 2's schedule. An extra train may be given a "schedule" in its running orders, by being directed to wait at a series of stations until specified times. Originally, this was only "with right over all trains" but was so useful that it was allowed in general, without specifying the trains to be waited for.

All trains, extra trains included, are expected to move in a single direction, from initial to terminal station, and must never move in the opposite direction (except behind a flagman in an emergency). The rules are predicated upon this fact, and safety depends on it. However, there is one important exception, and special rules apply. This is the work extra that is allowed to move in any way, in either direction, within certain specified limits. An order creating a work extra would read something like" "Eng 805 works extra 9 01 am until 3 30 pm between F and H." This means that engine 805 and its train may do whatever it thinks fit on the main track between F and H, so long as it clears all regular trains, and puts out flagmen in both directions to protect against extra trains. The time the track is occupied, and the length of track occupied, are to be as small as possible. The order must be given to all extra trains that will use the main track between F and H, so that they will know to look out for the work extra and not be surprised by it. Regular trains do not need the information, since the track must be clear for them.

Superior trains can pass and run ahead of inferior trains, and trains of the same class can pass and run ahead of each other, since they usually run at least as fast. Since extra trains run at the same speed as scheduled freight trains, extras are usually allowed to pass and run ahead of other freight trains, usually second and third class trains. The dispatcher can order trains to pass when necessary by an order like: "No 17 Eng 1006 pass No 1 Eng 1010 at M." We presume both trains are first class, and after passing the following train will not move more quickly than the preceding. Indeed, there may be some reason why No 1 cannot proceed at its normal speed. The problem arises when a slower train passes a faster for some unusual reason. The dispatcher must then make sure things are safe, with some order like" "Extra 475 run ahead of No. 77 A until overtaken." No 77 will look out for Extra 475 ahead of it, not running much faster than Extra 475 can, until it is seen ahead. Then Extra 475 will allow No 77 to pass at the next siding. The flagman of Extra 475 will keep a close watch on things. This is not as hazardous as it might seem to the uninitiated.

We have now given examples of nearly all the standard train order forms, except for those by which the dispatcher creates and manages sections. A section is one of two or more trains running on the same schedule. The dispatcher creates a section by an orders like: "Eng 850 display signals A to Z for Eng 1020" or "Eng 850 and Eng 1020 run as 1st and 2nd No 24 Z to A." A section displays green flags for a following section, and these flags are the only protection for the following section unless extra steps are taken. The superior train displaying signals must now become interested in inferior trains, since it must be sure they are warned that a section follows. This is generally done by whistle signals, the engine carrying signals giving a long and two short, answered by two short from the inferior train. If the signals are carried only to an intermediate point and there taken down, there is the great danger that the following section is in peril if it has not already arrived, so great efforts were exerted that opposing trains were properly notified. The only really safe way to handle this is to carry the flags through to the terminus, and annul the section beyond the station where it drops out. Handling sections is very intricate. Sections were intended to be used for similar trains running close together, where it is as safe as it can be. Some roads, however, insisted on running freight trains as sections of passenger trains, and running sections hours apart, and other doubtful procedures. Sections are a legacy from time table days, but most managements could not give them up.

One reason for running an extra passenger train as a section of a regular, first-class passenger train is to permit it to run at normal speed within yard limits. It was originally thought that the dispatcher could give orders to yard engines with respect to the main track, something like "Extra 475 West has right over yard movements at K" so that Extra 475 could run through at normal speed. This, however, proved impracticable. Yardmasters did not like receiving such orders from dispatchers, it seemed. Therefore, yard limits continued to represent a gap in the Standard Code. Many companies misused yard limits by establishing them at many stations without yard engines and such, merely to permit trains to stand without protection at stations, and for similar purposes. The Pennsylvania was finally able to abolish sections, and effectively, yard limits, when it made the block system universal on main tracks.

You may have noticed that in orders, No 2 is used, not No. 2, and 3 20 pm, not 3:20 pm. Train orders do not usually contain punctuation. Also, note that when a regular train is mentioned, its engine number is also given for positive identification. Sometimes, time is given as "3 20 three twenty pm," and even "No 2 two." The rules originally specified that time was to be expressed by figures only in train orders, even if the numbers were spelled out when transmitting an order.

Observations on the Train Order System

It is obvious that the use of train orders demanded literacy. Not just the dispatcher and operators had to use language effectively, but also the train crews, especially enginemen and conductors. These people not only had to read written instructions, but had to understand them perfectly and carry them out without fail. Very few jobs in similar industries have ever demanded so much literacy, and it is astonishing that men and women were not only available, but were expected to perform adequately. Train crews were surrounded by written instructions--the time table, bulletins, train orders and the rule book, for example. Dispatchers had to create instructions and make them not only correct, but comprehensible. The population as a whole was not characterized by great literacy, but railway employment attracted people of intelligence and understanding.

The men who created the Standard Code, and developed it in the succeeding years, were indeed men of intelligence and understanding, who strove for logic and correctness. In later years, after perhaps 1930, one notes less intelligence and less understanding in the development of the Code, and a failure to develop it effectively in the face of changing conditions. Perhaps the decline in ability to use written language, especially by train service employees, made the train order system difficult to maintain, and encouraged a change to simpler methods that involved less writing. Modern communication methods make this possible, so that now the dispatcher can talk directly with train crews without the intermediaries of operators and written instructions. However effective this is, it is much less interesting than train orders.

Psychological factors have a great deal to do with safe operation. The Standard Code has unconsciously encountered certain of these psychological phenomena. A fascinating one is the tendency of a man at a switch to turn the switch incorrectly in the face of a speeding train. I have discussed the reason for this apparently irrational action elsewhere, and here I can only mention that it cannot be cured by extra attention, which may even cause it. The only cure is to station the man far enough from the switch that he cannot reach it in his panic. This was never countered by a Standard Code rule, but many companies enforced a rule 104 (A) that required a man at a switch to stand on the opposite side of the track at a certain distance while the approaching train passed.

Another curious phenomenon occurred when train orders were transmitted by telephone. For accuracy, numbers and station names were both spelled and pronounced. The rule was that the letters were sent first, then the word was pronounced. If the word was pronounced first, and misunderstood, the spelling would do no good, the recipient hearing what he expected to hear, not what he actually heard. Remarkably, in later versions of the Code, and perhaps to this day, the instructions are to do it in the reverse, incorrect way. In fact, in the Army and other similar places, it is always done the wrong way, and this careless practice has now replaced the fruits of experience and insight.

Finally, we observe that the fundamental nature of the train order system is a single individual aware of everything that is taking place, and issuing orders to those who execute them in the field. It is exactly the same in principle as a dispatcher sitting before a CTC board on which trains are represented by lights, and issuing instructions electronically to command trains to stop or proceed by signals, and turning the switches as required. The difference, of course, is in the means of communication. There is more romance in the key and sounder, and the demands of written messages, however.


  1. American Railway Association, The Standard Code, January 1928.
  2. H. W. Forman, Rights of Trains (New York: The Railway Gazette, 1904).
  3. P. Josserand, Rights of Trains, 5th ed. (New York: Simmons-Boardman, 1957).

Return to Telegraph Index

Composed by J. B. Calvert
Created 28 April 2001
Last revised