The Working and Management of an English Railway/Chapter 5

From Wikisource
Jump to navigation Jump to search

CHAPTER V.

Signals and Interlocking.

An attempt will now be made to give some idea of the elaborate system of signalling the trains, which, in conjunction with the use of the electric telegraph, alone enables the traffic of a great railway to be carried on with safety and despatch, and which has grown up from the rudest beginnings side by side with the growth of the railway system itself.

There is, of course, a very obvious and primary necessity, on any railway, for some visible indication by means of which the drivers of the trains may be warned when they may proceed, and when they must come to a stand, and shortly after the opening of the Stockton and Darlington, which was the earliest railway constructed, one of the station masters is said to have adopted the simple expedient of placing a lighted candle in the window of the station-house when it was necessary for a train to stop. From this rude beginning to the complicated system of signals and interlocking which may be seen, for example, at Clapham Junction or Waterloo is a very long step in advance, and it has, of course, only been achieved by a gradual process of evolution. Thus when the Liverpool and Manchester Railway was first opened in 1830, the only arrangement made for signalling the trains was a flag by day or a lamp by night, held in the hand of the pointsman, as he was then called, and an old print from the "Illustrated London News" shows us the pointsman, or policeman, in the long-tailed coat and tall hat of the period, standing outside his hut, and making the prescribed; motions with his flag; but this simple device was soon found to be insufficient, and, about four years after the line had been opened, stout posts were provided, upon which lamps were placed by the pointsman. This really marks the first step in the creation of the present system of signalling; but about the same time the Grand Junction Railway was approaching completion, and the necessity for some improved form of fixed signal seems to have been realised, since we find that the line, at its opening in 1837, was provided with station signals in the form of discs, carried on poles about 12 feet in height, with a lamp at the top, the disc and lamp being arranged to turn through a quarter circle by the pointsman actuating a lever at the base. If the disc was turned so as to face an approaching train, or the lamp showed a red light, it implied a signal to stop; while, if the disc was turned edgeways to the driver, or the lamp showed a white light, it meant "go on." These signals, which may still be seen in use in some remote country districts, were fixed on the platforms at stations, or outside the pointsmen's huts at junctions, and corresponded in position with the home signals of to-day, the "distant signal" being, at that time, a thing unknown.

Some five years later, about 1842, a semaphore signal, somewhat similar to those in use at the present time, was introduced, and it appears probable that this now universally-adopted type of signal very rapidly superseded the disc signals of the earlier period. The: old semaphores showed three positions, "all right," "slacken speed," and "danger," the arms being actuated by the pointsman by means of a lever on the post. Up to this time, although the necessity must have existed, as it does to-day, for a driver to be warned as to the state of the line some time before actually reaching the point of obstruction, where he was required to stop, no attempt seems to have been made to meet the want; but at length accident proved once more the parent of design, for in the year 1846 a pointsman, who had to attend to two station signals, placed some little distance apart, in order to save himself the trouble of walking to and fro between them, procured some wire, which he attached to the levers of the signals, using a broken iron chair as a counter-weight, and by this simple expedient found himself able to work both signals without leaving his hut. Thus was demonstrated the possibility of working a signal at a distance, and this man's primitive contrivance doubtless hastened the introduction of distant signals, the use of which rapidly became universal. "Starting signals" were the development of a later period, the necessity for them having only been brought about by the application of the Block-Telegraph system some years after.

The next step in advance was the adoption, in a very elementary form, of the principle of interlocking, and this seems to have been forced upon the attention of railway engineers at a very early period, since, as early as 1843, we find that the levers for working the signals at a junction were provided with a simple mechanical device to prevent the main-line signal being lowered at the same time as that for the branch line; but, at that time, no attempt was made to extend the locking to the levers which actuated the points. In the following year Mr. C. F. Whitworth patented a scheme for locking points and signals on the ground by means of locks worked by wire, but there was no suggestion for concentrating the levers. The apparatus was complicated and difficult to work, and apparently it was never brought to bear, but there is no doubt that the proposal contained the germ of the interlocking system as we have it to-day.

In 1856 a successful attempt was made by Mr. John Saxby, at the Bricklayers' Arms Junction (London), to concentrate and interlock the levers working both points and signals, and although the apparatus employed was crude as compared with the perfect mechanism now in use, it represented the earliest practical application of the principle of interlocking. In 1859 the first interlocking frame was fixed on the London and North Western Railway, at Willesden Junction, by Mr. Austin Chambers, who patented his arrangement in 1860, and from this point the interlocking of the London and North Western system proceeded rapidly; for thirteen years later, in 1873, it is recorded that 13,000 interlocked levers were in use on that railway.

At this point it may be worth while briefly to describe the meaning and application of the different kinds of signals and locking appliances in use upon the principal railways at the present time.

The form of signal most generally adopted is the "Semaphore," which is no doubt familiar to most of my readers, and consists of a timber or iron pole varying in dimensions according to circumstances, but sometimes as much as 70 feet high, with an arm about 5 feet long, capable of assuming two positions when actuated by mechanical force. When this arm is horizontal and at right angles to the post it signifies "stop"; when it is nearly vertical it indicates "go on." As the arm cannot be seen at night, a lamp is provided on the post, and working with the arm is a frame containing coloured glasses, and termed "spectacles," which cause the lamp, to show a red light to indicate "stop," and a white or green light to indicate "safety," or "caution."

The signals are distinguished as "home signals," "distant signals," "starting signals," "advanced starting signals," and "disc signals." The home signal, as its name implies, is placed close to the point at which it is desired a train shall stop. The distant signal is placed at varying distances behind the home signal according to the gradient of the line and other circumstances, but on a straight and level railway it would be about 1,000 yards back, and its function is to repeat the action of the home signal, so that if the latter is at danger the driver has timely intimation of the fact, and can at once reduce the speed of his train so as to stop at the home signal. Distant signals are distinguished by a notch cut in the end of the arm. In working a Military railway it would be quite practicable to dispense with visible signals altogether and rely entirely upon the use of the telegraph; or, at all events, the two simple signals, the home and distant, as described, would probably suffice; but for the purposes of ordinary railways the introduction of the block telegraph system has necessitated the provision of starting and advanced starting signals. Starting signals are usually placed at the end of the platform at a station, and they indicate to the driver, by the lowering of the arm, when he may start his train and enter the section in advance. The advanced starting signal is usually placed about 300 yards from the cabin, in front of the starting signal, and enables a train which has passed the platform starting signal for the purpose of picking up waggons from sidings, or to clear the section in the rear, to be brought to a stand without entering the section in advance. It may be explained that all semaphore signals are fixed in such a way that on approaching them the arm appears on the left-hand side of the post, and this arrangement enables the same post sometimes to be used for carrying both "up" and "down" signals.

The disc signal is used to indicate to a driver whose train is in a goods siding which joins the main line when he may leave the siding and pass out on to the main line. It is really a disc of metal, with a lamp, carried on a spindle fixed near the ground, and which turns on its axis. The position of the disc by day and the colour exhibited by the lamp at night furnish the required indications as to the state of the line. Fig. 10 shows a simple arrangement of signals at a station, and Fig. 11 shows the ordinary signals required at a simple junction where there is no complication of lines.

At a small roadside station, where there are no sidings connecting with the main line, the signals are easily worked by detached levers, placed in any convenient position; but at larger stations, where there are many signals and points to be actuated, a means of concentrating the working becomes necessary, and this is done by means of the "signal cabin," which brings us to the question of interlocking. The signal cabin contains a most complicated piece of mechanism, called the "locking
Fig. 10 Signals required at a station; Fig. 11 Signals required at a simple junction

frame," consisting of a row of levers by means of which the signalman actuates every pair of points and every signal under his control. By a mechanical contrivance, the points and signals are interlocked—that is to say, supposing, for example, that the signalman has moved a lever that opens a pair of points, to enable a train to come out of a siding on to the main line, the home and distant signals must be at "danger" to stop any train from approaching on the main line, and it is physically impossible for the signalman to lower them. The manner in which this is accomplished was described some years ago in the Engineer, and that description is so complete and lucid that probably it would be difficult to improve upon it. The writer says:—

"Mounting by an iron ladder to the signal platform, we enter the glass-house. One side of this building is occupied by a row of strong iron levers standing nearly upright from the floor, and placed at equal distances along one side of the apartment; on the opposite side are many electric telegraph instruments,and the rest of the width forms a gangway or passage from end to end for two stalwart and serious-looking men, whose time is entirely occupied in looking through the glass sides of their cell, and pulling this way or pushing that way some of the levers which are arranged before them. These levers work all the points and signals, and their number is found to correspond exactly with the number of points and signal levers outside. Every lever is numbered, and on the floor beside it there is fixed a brass plate, engraved with its name and use. Sets of them are also distinguished in a way that readily catches the eye, by being painted in strong colours. Thus, for example, all the point levers may be black, the up signals red, the down signals blue and the distant signals green. The row of levers thus presents a diversified pattern to the eye, which is readily caught by the parti-coloured groups, and, having once got the key, distinguishes quickly and correctly between their different classes.

"On examining the levers somewhat more closely, we perceive that many of them have numbers painted on their sides, not one number only, but in some cases half a dozen or more; and one naturally asks what can be their meaning. These numbers involve the whole secret of the safety which is secured by the mechanism, as will be readily understood on examining the principles on which it has been devised.

"The keys and pedals of an organ, as every one knows, command numerous valves admitting air from a wind-chest to the pipes which it is desired to sound. The key-boards are sometimes double or triple, and are occasionally arranged so that the performer sits with his back to the instrument. The pipes are generally spread over a large space, and sets of them are sometimes enclosed in separate chambers. There thus arises considerable complexity in the mechanism by which the several keys are made to operate on their respective air-valves. Nevertheless, by means of rods, cranks, and levers, such a connection is effected that, on depressing a C key, not one C pipe only, but it may be twenty C pipes are made to sound, in whatever part of the instrument those pipes may be situated. And so it is with the point and signal levers. The whole row may be considered to form a key-board, every key of which is connected by suitable cranks and rods to some one of the points and semaphores which have to be played upon. In the organ, a touch of the finger serves to depress a key, for the movement has only to admit a puff of air to certain pipes; but here the keys require a strong and steady pull, for they have to move ponderous point bars, or broad semaphore arms, and their movements have to be conveyed round many corners and over considerable distances. In both cases the mode of communicating motion is the same, the two mechanisms differing only in size and strength; and thus far the organ and the signal instrument exactly correspond. Now, however, we come to a point in which they differ toto cœlo. A performer on the organ can touch any keys he pleases in any order or in any number; he can 'discourse most eloquent music,' or he can rend the ears of his audience by abominable discord. Not so the signalman. Concord he can produce at will, but discord is utterly beyond his powers. He cannot open the points to one line and at the same time give a safety signal to a line which crosses it; and the points must be properly set, close home to the stock, or fixed rail, or the signal for a train to pass cannot possibly be given, and the least obstruction occurring to prevent the full and true opening or closing of the points is at once discovered, even with connecting rods of the greatest length practicable. Moreover, while a train is actually travelling through the points, it is itself master of the situation; not even the signalman can, either intentionally or inadvertently, change their position or disturb them until the whole train is safely passed. When he gives a clear signal for a main line, he cannot open a point crossing to it; when he gives a clear signal for a crossing he must show danger for all the lines which it crosses. And this is the meaning of the numbers marked on the different levers. No. 10, let us suppose, has 5, 7, and 23 marked on Its side. He may pull at No. 10 as long as he pleases, but he cannot move it till Nos. 5, 7, and 23 have first been moved; and so throughout the whole system. No signal lever can be moved to safety unless the point levers corresponding with it have first been moved, and no point levers can be moved while there stands at safety any signal lever that ought to stand at danger. Every lever is under lock and key, each being a part of the key which unlocks some of the others, and each forming a part of the lock which secures some of the others against possible movement, while each is at the same time subject to the control of all those which are related to it.

"This result, complex and difficult as it seems, is achieved by mechanism of great simplicity and beauty. Immediately under the floor of the platform, and just in front of the levers, are arranged several series of vibrating and sliding bars, somewhat like the tumblers of a lock, placed horizontally. These bars have projections here which stand in front of certain levers as obstacles to their motion, and notches there which permit certain levers to travel. Some of these have sloping faces, so that when a lever moves along them it edges them to one side, and this transverse motion being communicated to others of the series, brings the proper projections or notches in front of those other levers to which the moving lever is related. Thus by the movement of one lever some others are stopped, and some are left free, and this simple principle, carefully applied to all, combines them in a system incapable of discord."

There are several supplementary interlocking appliances now generally in use which may be briefly described. For instance, there is the "Facing-Point Lock," which is a bar of iron working in connection with facing points—that is, points by which one line diverges from another in the same direction. This bar, which is actuated by a separate lever in the cabin, works with, and securely locks, the points, and not until this has been done can the signal be lowered. The bar holds the points firmly in position, so that, even if the signal has been reversed, they cannot be moved while the train is passing over them. Then we have the "Locking Bar," somewhat similar to the facing-point lock, but actuated by the same lever that works the points. This is chiefly applied to siding points to prevent their being moved while a train is passing over them. "Detector Bars" are employed on parts of the line which cannot be seen by the signalman, to prevent the signals being lowered when the line is occupied by a train. "Detector Locks" are applied to facing points, and are worked by the wire that works the signals, and if the points should be injured or out of position the fault is at once detected, and the signal is locked at "danger."

Again, there is the "Point-rod Compensator," which automatically compensates for the expansion or contraction from heat or cold of the rods which actuate the points, and an "adjusting apparatus," which enables the signalman to adjust his signal wires without leaving his cabin, together with numerous other ingenious contrivances for working and interlocking with the signals the gates of level crossings, turn-tables, and canal drawbridges, and for many other purposes. There is, moreover, in use on some of the branch railways, a system of what is known as "Key Interlocking" for siding points, which merits a particular, description, as it is both simple and effectual, and might prove valuable in the working of temporary or Military railways. By means of this system, while the large outlay required for laying down rods to work the points is avoided, all the security of interlocking is retained. A simple frame of levers is provided for working the signals, with a number of keys. If a certain siding is required to be used, the key which applies to that siding is withdrawn, and this has the effect of mechanically locking all signals for the road with which the siding connects. The key is then carried to the siding, and with it the points are unlocked, but when they are once opened the key cannot be withdrawn until they are closed and locked again, and the signals, of course, cannot be lowered until the key has been carried back and inserted in the lever frame, so that the security is complete.

During the interval between 1859, when the first interlocking apparatus was fixed on the London and North-Western Railway, and 1873, when, as before stated, considerable progress had been made towards interlocking the whole of the system, many alterations and improvements in the method, both of signalling and interlocking, had from time to time been introduced with the inevitable result that the signal plant and apparatus in use had come to be of very mixed types, causing great difficulty and unnecessary expense in maintenance and repairs.

This state of things led to the introduction of a new system altogether, with regard to signalling the railway, the organisation of which was entrusted to Mr. F. W. Webb, Mem. Inst. C.E., the Company's chief mechanical engineer. Up to this time, all signal work upon the railway had been provided and kept in repair under contract by the different firms of railway signal manufacturers, but chiefly by Messrs. Saxby & Farmer, of which firm Mr. John Saxby, before referred to as having been the inventor of the first effective system of interlocking, was an active partner. The Company now determined in future to make and keep in repair their own signals and interlocking apparatus, so as to secure, not only efficiency and economy, but uniformity of pattern throughout the line. The scheme was brought to bear, and proved to be a success. The Company had already ample facilities in the locomotive works at Crewe for turning out all kinds of iron work; carpenters' and fitters' shops for making the cabins, signals, and other woodwork were added, a superintendent and a staff of artisans appointed, and the thing was done, the payment of royalties to inventors being soon dispensed with by the Company's engineers designing and patenting new systems of their own.

The signal-fitting shop at Crewe is 280 feet long, and 85 feet wide, the machine tools being driven by a fine Otto gas engine of 48 indicated horse-power. There is, of course, plenty of steam power available in the works, but the signal shop is often busy at a time when the other machinery is not working, and the gas engine is, for this reason, found to be a convenient and economical arrangement. One side of the shop is devoted to the preparation of the signal-posts, which are received from the saw-mill cut to the proper dimensions, and ready for fitting. The posts are of Baltic timber, varying in size from 9 inches to 15 inches square at the bottom end, and tapering to 6 inches square at the top, the length varying according to circumstances. About seven feet of each post, at the bottom, is left the full size of the timber from which the post is cut, and this portion is first charred in a furnace and afterwards well coated with gas tar, while upon the top of the post is placed a galvanised cast-iron cap. The arm of the signal is formed of a thin steel plate, stiffened by two corrugations running lengthways, and is riveted to a galvanised cast-iron centre, upon which is fixed an adjustable frame for the spectacle-glasses, made sufficiently heavy to counterbalance any accumulation of snow on the arm, and keep it in its "danger" position should any of the fittings fail. The arm works on a spindle squared to receive it at one end, and on the back light lever at the other end, except when the signal is electrically repeated, in which case a racked segment is substituted for the back light lever, and it is geared with a similar segment, in connection with the electrical contacts of the repeater apparatus. The spindle turns in a casting fastened to the post by four ⅝-inch bolts, which also serve to secure the lamp bracket fixed at the back of the arm, this bracket being of sufficient width to form a "stop" for the arm when in its "danger position."

There are several other distinctive patterns of signals employed for special purposes, amongst them being the "disc" signals, previously mentioned, for controlling the exits from sidings, and the miniature semaphore signals placed low down on the post, which at certain busy parts of the line are used instead of a hand signal when it is necessary to call an engine cautiously past a signal at "danger," and which are termed "calling on arms."

On an average the fitting shop at Crewe turns out about nine signals per week, including composite or bracket and gantry posts, some of which have been constructed to carry as many as twenty-four signal arms.

On the other side of the shop are arranged a number of erecting frames, on which the interlocking apparatus is built up. The frame used by the Company is known as lever actuation, with horizontal and vertical bar locking. The horizontal bars are of Bessemer steel, rolled to a shallow channel section, the locking studs being squared to fit the channel and secured by riveting. The vertical bars are also of steel, notched, the notches being punched out cold. Each vertical bar is guided by a channel the length of itself, one wall of which is made deeper than the other; notches are cast in this to act as guides for the horizontal bars, these notches being faced by a special drifting machine at the rate of thirty-five notches per hour. The apparatus is characterised by great strength and simplicity, and, an unusual amount of special machine work being employed in its construction, a perfect interchangeability of parts is secured which is highly essential to economical maintenance. This is sufficiently demonstrated by the fact that, since the introduction of the present form of apparatus and methods of construction, the annual cost of maintenance has been largely reduced.

Signals are worked from the signal cabins by wires, but the points are actuated by rods. Trailing points are worked at a distance of 220 yards from the cabin, but in the case of facing points the distance is restricted by the Board of Trade to 150 yards, increased in some exceptional cases to 180 yards. At such distances as these, it is obviously of great importance to have a strong and thoroughly trustworthy means of connection between the points and the cabin, and for this purpose ordinary iron gas-piping was formerly used, but latterly it has been found that a steel channel section of rodding is far superior for the purpose, and the success attending this innovation has led to its adoption by several other railway companies. The chief advantage over gas-piping is that rods weakened by corrosion are readily detected and can be removed. Moreover, being manufactured to a standard in the works at Crewe, they are of uniform excellence, both in material and workmanship. About 6,000 yards of this rodding is turned out on an average every month, is sawn off to 18-feet lengths, and, when laid down, is jointed by means of fish-plates, secured by six-inch bolts, the fish-plates being rolled to a section to fit the channel of the rod, and the bolt holes drilled by a multiple drill. After being laid down, the rodding receives a coating of red lead and a second coat of red oxide paint, which gives it a smart appearance, and is found to be a good preservative.

For lighting the signal lamps at night, gas is used in the neighbourhood of large towns, where it is readily obtainable; but by far the greater number are lighted with petroleum oil, a large number of spare lamps being kept always on hand to provide for cases of failure. The signalling of the London and North-Western Railway as it exists to-day has involved the erection of 1,400 signal cabins, containing some 30,000 levers, and of 16,000 signals. The steel rodding used to actuate the points would form a continuous rod from the Land's End to John 0'Groat's, while the wires which work the signals would stretch from Liverpool to New York. The largest interlocking apparatus ever constructed in the Crewe works is that at the South Junction cabin at Rugby, which contains no less than 180 levers.

The maintenance of all this complicated array of signals and interlocking apparatus, over upwards of 1,800 miles of railway, is in itself a work of great magnitude, and one requiring the closest and most anxious attention, having regard to the consequences that might result from a single point rod breaking or getting out of gear; or from a single signal arm failing to respond to the action of the lever. For the purposes of maintenance, the entire system is divided into eleven districts, each of which is in charge of an inspector, these eleven inspectors being assisted by twenty-one sub-inspectors and foremen, and having under their orders sixty-eight chargemen and 430 workmen. The districts are sub-divided into lengths, and each length is placed in charge of a "chargeman" and an assistant, who visit every signal cabin on their length once a fortnight, clean and oil the fittings of each signal and point, execute any small repairs or renewals that may be required, and which it is possible to carry out during their visit, and report to the inspector of the district any repairs or renewals which they may find to be necessary, but which they are unable to deal with upon the spot. In every signal cabin the name and private address of the chargeman who is responsible for it is posted, and it is the duty of the signalman to report to him every failure or defect that may become apparent during the intervals of his visits, sending at the same time a duplicate of the report to the inspector of the district.

In addition to this, every cabin is visited once a month by a fitter and his assistant, whose duty it is to carefully examine, clean, oil and, if necessary, repair the locking apparatus, particulars of the work done being entered in a monthly report, which is sent through the district inspector to the head office at Crewe.

Heavy repairs and renewals are executed by an extra gang, attached to each district, in charge of a responsible foreman, the most stringent rules being laid down as to the manner in which the work is to be carried out, so as to provide for the safe and uninterrupted conduct of the traffic during the time it is going on.

In order to secure uniformity in the sighting of the signals, the following rules are, as far as possible, adhered to, but these rules are always subject to modification so as to meet special circumstances such as, for instance, where bridges, trees, or other natural objects, or sharp curves, would otherwise obscure the sight of the signals.

Home Signals.—To be fixed within sight of the distant signal, and to be erected as near the signal cabin as possible, so that a driver, whose engine is standing at the signal, may be verbally communicated with, if necessary, by the signalman.

Distant Signals.—To be fixed at a uniform distance of i,ooo yards from the home signal, unless the gradient is a rising one, in which case a distance of 8oo yards is considered sufficient. In special cases, a distant signal may be fixed a maximum distance of 1,200 yards from the home signal, where the circumstances admit of a sufficiently direct route for the wire connections to enable the signal to work freely.

Junction Signals.—The same rules apply to these as to home signals. They are to be fixed near the facing points to which they apply, and in no case at a greater distance from them than 200 yards, unless repeater signals or duplicate locking bars are provided.

Junction signals for protecting a junction in the trailing direction (i.e., a junction formed by the connection of one line with another, not by meeting or facing points, but by trailing points) to be fixed a sufficient distance from the fouling point to afford some margin for a driver who may accidentally draw past the signal.

Junction signals are not in any case to be placed on the same post one above another, but a separate post is to be provided for each signal.

Platform Starting Signals.—These are sighted at a height of about 15 feet above rail-level, and so as to be seen by a driver the whole length of the platform. This rule, however, is not strictly applicable if the same post carries the distant or home signal for a cabin in advance, in which case the signals are sighted as home or distant signals, as the case may be.

Advanced Starting Signals.—These are to be sighted of a uniform height with platform starting signals, and are to be seen by a driver from the platform starting signal. They must not in any case be at a greater distance in advance of the cabin than 350 yards, and must be clearly in view of the signalman.

General Rules.—Bay starting and other subordinate signals are to be of a uniform height of 15 feet from rail level, and are to be provided with short arms and purple lights instead of red.

Lamps are never to be placed nearer together than 7 feet, except in the case of subordinate signals.

Signals elevated more than 45 feet above rail-level are to have a lower arm on the same post, 15 feet above rail-level, so as to be visible in fogs or thick weather.

Signal cabins are to be so placed that the signalman may have a good sight of all points and signals actuated from his cabin, more especially facing points.

All signals are to be fixed on the left-hand side of, and as near as possible to, the road to which they apply, no signal or cabin, however, being erected at a less distance than 4 feet 6 inches from a passenger line of rails.

A distant signal for a cabin in advance must never overlap a home or starting signal for a cabin in the rear, but must be fixed on the same post, 7 feet below the home or starting signal, and be controlled by the home or starting signal, to prevent the distant arm being "off" when the home or starting arm is at "danger." When a distant signal is placed below a home or starting signal, no other signal must be placed on the same post, excepting "fog" or "calling-on" arms.

When the distant signals for a junction have to be combined with the home or starting signals for a cabin in the rear, then the starting signal post must be nearest the line with the distant arm for the right-hand junction placed under it, and the distant signal for the left-hand junction must be carried by a separate post to the left of the home or starting post. Both distant arms must be controlled by the home or starting signal. This rule will not apply if the junction to the right is a goods loop or other subordinate line; in this case, the distant signal must be carried on a separate post to the right of the home or starting post.