William Henry Preece, MICE, presented a paper 'On Railway Telegraphs' before the Institution of Civil Engineers on 13 January 1863 that is very interesting for the view that it gives of railway signalling at that time of rapid change. The 'time interval' method of operation is reported to be 'practically universal' at that time, and regarded by railway managers as the only way the traffic can be handled, while Board of Trade inspectors are recommending the 'space interval.' The telegraph and fixed signals are used as an additional safeguard to warn trains of trains close ahead, and guards are expected to protect trains stopped out of course. It has been little more than twenty years since the electromagnetic telegraph first appeared, and the progress has been remarkable. The information in this paper that is of greatest interest to the historian of signalling is presented here, with comments and elaborations furnished from our present viewpoint.
The then-recent Clayton Tunnel accident on the Brighton on 25 August 1861 is brought up, since it was adduced as an example that the telegraph could cause accidents as well as prevent them. Preece points out quite correctly the errors in the method of using the telegraph that were responsible for the disaster. The telegraph had been installed here about 1851, a typical date for the first use of the telegraph through long tunnels, on steep inclines (such as the Lickey), and at junctions. The method of use is clearly explained. A needle telegraph instrument connects signalmen at the portals of the tunnel, and is used for both lines. A deflection of the needle to the left signifies 'train in,' and a deflection to the right, 'train out.' When a train enters the tunnel, the signalman displays Danger and sends 'train in' to the man at the other portal. When this train leaves, 'train out' is returned, and the signal turned to Safety. The signals are momentary, and may be interspersed with conversation or signals relating to a train in the other direction. Preece enunciates the important principles that the block wire should be used for nothing but signalling, and the currents should be continuous, not momentary. Preece estimates that 90% of the early isolated blocks were at tunnels.
The Eastern Counties single line from Norwich to Yarmouth was operated by the telegraph from 1844, initially by the Cooke and Wheatstone 5-needle instrument, which was well proportioned to the five stations on the line. However, maintenance of all the wires was a nuisance, so a more normal needle telegraph replaced the 5-needle instrument. It was still eleven years before the terrible accident at Norwich, caused by misuse of the telegraph, not by the telegraph, like Clayton Tunnel. In 1863, the Board of Trade required the telegraph for working single lines, except for one engine in steam. It seems that single lines were worked by a variety of methods, from staff and ticket to timetable with telegraphic orders to alter crossing places. There were more single lines that is usually appreciated these days.
The Board of Trade had just required the LSWR to double the line between Southampton and Dorchester, but the rest of the line, soon to reach Exeter, was single. It was operated in 50-mile divisions, with a single dispatcher for each, issuing crossing orders that were repeated and given to the drivers in written form. This was, as Preece points out, identical to the method then being developed in the United States for single lines, and there was apparently close contact between the two shores at that time. The LSWR is notably absent from L. T. C. Rolt's list of accidents in Red For Danger, so the system was well constructed and managed. The only thing that seems to be absent from what was in the mature American system was the requirement that orders be given in the same words to all concerned. The most important requirements were that orders be given by a single authority, and that they be repeated for accuracy before delivery. The muddle of the Radstock accident on the S&D shows how the system can be abused through incompetent management.
Edwin Clark wrote a 'Letter to Captain Huish, as to proposed improvements in the Electric Telegraph system, for the service of the London and North Western Railway Company' in March 1854 with his proposals for a block system. Clark believed in fail-safe operation, and in continuous currents to maintain a permanent reminder of the state of the block in front of the signalman. He used the available 2-needle speaking telegraph, one for each line, devoted entirely to signalling. A deflection to the left indicated train on line, a deflection to the right line clear, and no current meant line blocked, which was then to be considered a danger signal. The block instrument would normally show line clear, and would be changed to train on line when a train passed. When the train left the block, the instrument would be pegged back at line clear. A separate alarm bell was used to call attention of the signalmen. The telegraph wires were looped down certain poles, where they could be cut, sending the instruments to line blocked, and raising the alarm. the system was installed in 1855 between Euston and Rugby, 83 miles, with blocks 2-1/2 miles in length. Originally, it was operated as an absolute block, but this seemed to interfere with the numerous goods trains that ran at night, and it was altered to a permissive block. Passenger trains may still have been handled by an absolute block, or something close to it, but Preece does not say. There was no electrical signalling between Liverpool and Manchester at all, at this time.
Clark's system is obviously the ancestor of that which became classic, but initially had some differences. There were no bell codes, and the block instruments were controlled by the signalmen at both ends of the block, recalling the early train in - train out system. The drop wires were especially deprecated by Preece, and indeed they were a thoroughly bad idea, totally unnecessary with absolute block and little help with permissive block. However, the term 'Line Blocked' was retained for the neutral position in the later block system, although the meaning was absent from the term. Because Clark used continuous currents, he required a suitable closed-circuit battery, and so used the Daniell copper sulphate and zinc cell.
Preece then tries to describe the system used by the Great Northern, which also used 2-needle telegraphs, but in a way 'only understood by themselves.' There is one instrument for both lines, and the two needles refer to passenger and goods trains, respectively. Signals are momentary, and the working is permissive. He remarks that every line from London used a different system, which was quite true.
The New Street Tunnel at Birmingham required 8 wires for its working, which Preece says could be reduced. In the discussion, Mr Culley says it's only five wires, and gives the reason for so many wires. Also, electrical signal repeaters were definitely in use by this time, as was the Red Cap Pilot means of working a temporary single line..
Preece has an invention to advertise, and that is his block instrument that looks like a miniature semaphore signal, operated by an electromagnet, and controlled by the signalman in advance. He considers the section signal to be essentially a 'distant' signal for the signalman in advance, and his block instrument shows what the state of that signal should be to the man who must operate it mechanically. This, of course, is precisely the later view. The miniature signal shows Clear when current flows. Interruption of the current sends it back to Stop. This was the prototype of the 2-position block instruments in the form of miniature semaphores that became so popular in the South, while the Clark telegraphs evolved into the 3-position block instruments seen elsewhere. Preece's instruments were used on the LSWR.
Edward Tyer and C. V. Walker, of the South Eastern Railway, thought that Preece had not been completely fair to their systems. Walker invented a system using single-stroke bells only that was first installed in January 1852. These bell signals signified train in/train out as well as the character of the train, and he said they were quite satisfactory on lines of light or moderate traffic. Tyer had a block telegraph that appeared strikingly like the later three-position instruments, but was actually quite different. It was operated by momentary currents, which then displayed permanent reminders of the last signals sent and received. These were used on lines of heavy traffic, such as the North Kent and the Mid Kent lines. Tyer used a bell code, which may be of interest: 1-acknowledge, 2-passenger train, 3-goods train, 4-express or light engine, 5-obstruction, 6-testing bells. Walker said that as of 1863, 275 miles were worked by bells, 11 miles by Tyer's instruments. These were, of course, not the same as the later Tyer's instruments. Walker said his bells were suitable for heavy traffic as well, since they were used on the line from London Bridge, which had very heavy traffic.
Preece reports that German railways use bells, but very differently. Bells (apparently vibrating bells) are placed at frequent intervals along the line for the information of gatekeepers, platelayers, and others. When a train is ready to leave, the bells are rung all along the line to the next station. This obviously serves as a 'train in' signal, but how the 'train out' is given is not told. A French railway uses a two-needle instrument, but deflecting to one side only, and only on momentary currents. He assesses its operation as defective.
The disc block instruments of C. E. Spagnoletti, of the Great Western, are then briefly mentioned. There is a cut of one in the discussion on the paper. These instruments were installed on the Metropolitan Railway, which had opened that very month in 1863. The young Spagnoletti had been appointed Telegraph Superintendent only in 1855, but was already making a name for himself. Spagnoletti made the important invention of the induced needle, as well as a block instrument that replaced the hard-to-see needle by a rectangular opening showing sectors of a disc showing Train On Line by a red panel, Line Clear by a white one, and Line Blocked by half red, half white. The importance of the induced needle needs some further explanation.
A needle telegraph was a pivoted permanent magnet inside a coil of wire. When current passed through the coil in one direction, the magnet, and therefore the needle, were rotated to one side. If the current were reversed, the needle would deflect to the other side. With no current in the coil, the needle hung vertically. This was very convenient for the 3-position block instrument, which could show line clear or train on line by simply reversing the polarity of the current, and line blocked by cutting it off altogether. Now suppose the polarity of the magnet is reversed. This would reverse the direction of deflection, turning train on line into line clear.
It was not long after the first telegraph lines were strung along poles that telluric currents were discovered. These are currents generated by small fluctuations in the earth's magnetic field, especially strong when the Aurora Borealis is observed, or during 'magnetic storms' produced by solar flares. These currents can be strong enough to operate the telegraph without a battery, and occasionally destructively strong. The movement of thunder clouds, and more obviously, a lightning strike anywhere in the vicinity of a telegraph wire, causes a strong current surge. A direct strike, which is fortunately rare, may completely destroy the apparatus. All these effects are intensified on circuits using a ground return ('one-wire' circuits). Lightning does not easily travel down a telegraph wire, but takes the first reasonable path to earth. Wherever a telegraph wire enters a building, a 'protector' is installed to bypass dangerously high voltages to earth. The protector has no effect on lesser disturbances, which may still involve currents much greater than normal. All these disturbances were attributed to 'lightning,' regardless of the actual cause. Preece was well aware of those associated with thunder clouds, and recognises telluric currents, if not their cause.
When such an unusually large current passes through the coil in the telegraph, it not only causes a vigorous deflection, but also may magnetize the magnet in the direction of the magnetic field it excites. The most common effect was, however, to weaken the magnet so it would not deflect strongly enough. Telegraphers were provided with a strong permanent magnet to remagnetize the needle magnet if this happened. In some cases, the magnetization was reversed, which would make the block instrument read incorrectly. The error would probably soon be detected, but the situation was a dangerous one, frequently quoted by the anti-telegraph crowd.
Spagnoletti eliminated the problem by making the magnet connected with the needle out of soft iron, so it would not magnetize permanently either way. A permanent magnet, well away from the coil, was arranged to induce magnetic poles in the soft iron, and these were always correct to give the required deflections. An excessive current through the coil would cause a strong deflection, but would not change the magnetic polarity.
The paper occasioned a great deal of discussion, which occupied the next two weekly meetings of the Institution, on 20 and 27 January. Captain Huish said the telegraph was fine as an auxiliary, but should not be depended upon for safety. He thought train staff was the only way to operate a single line, and felt uncomfortable riding on the LSWR with its telegraphic orders. Mr Hemans, from Ireland, said the train staff was a very inconvenient system, even with light traffic, and he preferred telegraph. Captain Tyler, RE, said only the train staff could be trusted. He also said speaking telegraphs should be provided to reduce the temptation to use the block instruments for that purpose. Mr Adley said single lines were worked by telegraph in India with safety, and that all Indian railways had the telegraph. The Board of Trade had prepared rules for the train staff system, but had not done the same for the telegraph system, a strange circumstance. Mr E. A. Cowper reminded everyone that he had invented the fog signal. George Hawkins of the Brighton expressed the opinion that safety measures resulted in a relaxation of the attention of drivers, who might come to rely on them.
President Hawkshaw, in his summing-up, was of the opinion that the train staff was 'an approach to doing away with railway travelling,' and that the telegraph had not produced a relaxation in the vigilance of drivers, and was, in fact, a great time-saver, unlike the train staff, which was a time-waster.
Composed by J. B. Calvert
Last revised 19 September 1999