I'll try to make a very much simplified but expanded explanation of train orders and timetable operation typical of US practice. The early US railroads were heavily influenced by British practice, but I'm not aware how fast these practices diverged or if the British had a comparable train order system.
On single track, there must be some form of system that prevents trains from colliding. In this era, main track authority on single track was granted by train register (dead end branches), staff system or a manual block system with train order and timetable being the most popular, being common until largely replaced by DTC and TWC by 1990.
Modern dispatching methods in dark territory are dependent upon reliable radio communications, soon to have constant electronic oversight by big brother PTC. Since radios did not exist in 1916, there had to be a way that trains could run without having constant communications to obtain additional authority from the dispatcher. As it was, many of SVRy's trains were long enough that a hand or lantern signal would not be visible the length of the train, leaving the engine whistle and conductor's valve as the means of communication between the head and rear end.
To prevent opposing trains from colliding head-on, each train is given a rank making it either superior or inferior to every opposing train using the same single main track. A train's superiority is determined by right, class and direction. Designated by the timetable, a First Class train (passenger) is superior to any opposing Second or Third Class train (freights); Second Class trains are superior to Third Class trains and so forth. If opposing trains are of the same class, the timetable will indicate which of them is in the superior direction.
If a train is superior to opposing trains, it may proceed observing its schedule (can't leave early) expecting to find any inferior train clear of the main track. An inferior train can proceed against a superior train's time, but must be in the clear of the main track for the superior train by at least 5 or 10 minutes. (Now you know why the railroads made such a big deal about accurate watches.)
For superiority of trains in the same direction, the rules are a little different. A second class train running ahead of a first class train must be in the clear of the main track before the leaving time of the First Class train at the station/siding to the rear. If the Second Class train can make the next station/siding ahead of the First Class train's time to the rear, then it can keep going. Head on collisions weren't that common, but rear end collisions were. Anytime a train was outside of yard limits running at slow speed or stopped on the main, there was a possibility that a following train could rear end it. The rear brakeman was actually the flagman, and his job was to protect his train by giving sufficient warning to a following train to be able to stop short. This could simply mean dropping fusees that would keep a following train 10 minutes behind, or going back placing torpedoes, lighting fusees and a flag or light to stop a flowing train. The distance varied depending on maximum speed and grades, but could be over 2 miles in high speed (60+) territories.
Have you ever heard the ballad and legend of Casey Jones? He was the engineer of a delayed passenger train trying to make up time when he rear ended another train standing on the main. It turns out several trains tried to meet at the same station and there wasn't room for them all to clear the main. This is one of the drawbacks of the train order system, and usually the conductors are able to figure out how to eventually saw everyone by. It was never clearly determined if the flagman was "short flagging", not being far enough back to stop a following train, or if Casey missed the stop signal or was simply running excessively fast and couldn't stop short.
If all the scheduled trains stay on time, there isn't a problem. If a superior train falls behind schedule, the inferior trains can't move until they meet the superior train, which may or may not even show up. A schedule is in effect until the train really late, typically 12 hours. The dispatcher can alter the right of trains, which is conferred by train order. For example the dispatcher can add time to a train's schedule or have a train wait until a given time at certain points, so inferior trains receiving the order can use the extra time to proceed against the superior train. The dispatcher can issue a meet to two or more trains which implies right for an inferior train to proceed to the designated meeting point. If desired the dispatcher can annul all or part of a train's schedule.
Obviously the timetable schedule can't accommodate all the possibilities, so the dispatcher can authorize (again by train order) extra trains. Unless provided by train order, an extra train is inferior to all regular trains shown in the timetable. Before authorizing an extra train, the dispatcher must provide protection against any opposing trains within the same territory, since the opposing trains may or may not be aware of the other and there is no superior direction between extras. The dispatcher make the extra order not in effect until after the arrival of opposing trains at the origin station, set a fixed meet, or give an extra right over any one or more trains. If the dispatcher gave an order stating ENG 50 RUN EXTRA MASON TO WHITNEY WITH RIGHT OVER EASTWARD EXTRA TRAINS then no eastward extra train could move against the superior Extra 50 West they knew it had been met. If the dispatcher added AND WAIT AT MASON UNTIL 415P S WYE UNTIL 445P LARCH UNTIL 535P then an opposing inferior train could potentially proceed against Extra 50 West, clearing its time as required by the rules.
There are several "forms" of train orders that can vary by railroad. These are basically standardized formats with their own specific rules the dispatcher is supposed to use when issuing train orders so that they are clear and concise. Without a SVRy rulebook, I can only use commonly accepted forms of train orders for example.
To differentiate regular trains from extra trains, the leading engine of extra trains is required to display a pair of white flags (by day) or lights. Also common on larger railroads is the ability to run multiple sections of a regular train. For example, an inferior train is waiting in a siding for a superior train that arrives displaying green flags or lights. This indicates that there is a following section of the train which has the same superiority and rights as the first section. If the following train is also displaying green signals, then you're stuck for another following section. When the last section shows up it will not be displaying green signals. So if you are given an order EXTRA 720 EAST MEET NO 1 AT S WYE, then Extra 720 East must meet train No 1 and all, if any, of its sections. To add to the confusion, the dispatcher could have issued a train order stating EXTRA 720 EAST MEET SECOND NO 1 AT S WYE which changes everything. Extra 720 West can proceed against No 1's time to meet it, but not beyond S Wye. After it meets No 1, if it wasn't at S-Wye Extra 720 East would have to proceed to S-Wye to meet Second No 1 to fulfill the train order, plus any additional following sections, if any.
And don't forget that a train may be juggling several orders covering several different extras in addition to the regular trains that are superior, not to mention slow orders. If for some reason you break down, stall or get delayed – you have to provide flag protection to the front and rear depending on what superior train's time you are on.
Another type of extra was a work extra. A work extra had no direction, and could move either direction within its limits as often as it wanted. In modern practice, a work extra owns the track, but under train orders a work extra is the lowest of the low in superiority. It not only has to clear regular trains, but it can only move on the main track under flag protection in case it was to encounter an extra train. Typically the dispatcher would add to the order giving it right over various trains or modifying the requirement to provide flag protection. For example, the dispatcher needs to allow the Whitney Day Helper to make numerous turns between Whitney and Larch: ENG 16 WORKS EXTRA UNTIL 230P BETWEEN WHITNEY AND LARCH NOT PROTECTING AGAINST EXTRA TRAINS. This order would allow engine 16 to make as many trips as it wanted, clearing regular trains as usual, but relieved from providing flag protection against extra trains. It is the dispatcher's responsibility not to authorize or to hold any other extras from going through Extra 16's working limits before 2:30pm. Since a work extra can never fulfill its train order by arriving at a destination, there is almost always an expiration time otherwise the dispatcher has to annul the work extra. If Extra 16 found itself at Alder Springs at 2:20, it would have to clear the main track immediately before it loses all right to occupy the main track and then contact the dispatcher for new authorization as an extra.
The actual rules are far more extensive and complex than here. If you consider the modern use of Track Warrant Control (TWC), it is similar to train orders except that there is no schedule or superiority of trains, and the dispatcher is just giving you authorization in small bites and only for the next section of track that is clear. (Mother may I …) An option used to be to issue a new warrant to a train stating "NOT IN EFFECT UNTIL AFTER THE ARRIVAL OF (train)" which allowed the dispatcher to issue a warrant contingent upon meeting another train. Too many crews either misidentified the train they were supposed to meet and forgot it all together and ended up going head on. I'm not sure whether it is due to lack of experience or intelligence, but most railroads have been forced to quit doing this or require that the other train must actually confirm they are already clear before issuing the next warrant.
