William F. Cooke, co-inventor of the electromagnetic telegraph, published a pamphlet in 1842 in which he showed how the telegraph could be used to prevent collisions on railways. The line was to be divided into sections called blocks. At each point of division signalmen, in communication by telegraph, were to be stationed at block posts, with fixed signals to control the trains. When a train wished to enter a block, the signalman at its entrance communicated with the signalman at the exit of the block. If they agreed that the block was clear of trains, the signalman at the exit prevented any train from entering at that end, while the signalman at the entrance admitted the train, protecting it with a signal behind it. When the train reached the other end of the block, the signalman there advised the signalman at the entrance that the train had left the block, which was now clear.
A written record was kept of the signals exchanged between the signalmen and the times they were sent or received in the block record. This helped the signalmen to know the state of the block at all times. This system works as well on single track as on double. Since the messages exchanged are few in number, they can be sent by bells as well as by the speaking telegraph. The first block systems, appearing about 1851 in England, used single-stroke bells. Isolated blocks for the protection of tunnels were introduced about the same time, using the needle telegraph. The desirability of a visual indicator of block status was recognized, and about 1855 the block instrument, based on the needle telegraph, made its appearance. On a double-track line, it was controlled by the signalman at the exit of a block.
In the beginning, the normal position of the block instrument on a double-track line where trains ran in a single direction on each track was Line Clear. The signalman could admit a train to the block, then notify the block post in advance that a train had entered the block. That signalman would change the block instrument to Train on Line. When the train arrived at his post, he would send Train Out to the block post in the rear, and change the block instrument to Line Clear in readiness for the next train.
This, obviously, was inappropriate for single track, and it was thought better to require the cooperation of two signalmen to admit a train to the block. Therefore, the normal position was called Line Blocked. The signalman in the rear would offer at train to the signalman in advance. If he agreed that the block was clear, he would accept the train and change the block instrument to Line Clear. The signalman in the rear could then clear his signals. When the train entered the block, he protected it with his signals, and advised the signalman in advance with Train Entering. This signalman changed the block instrument to Train On Line. When the train arrived, he sent Train Out to the post in the rear and changed the block instrument to Line Blocked. This became the standard method of operating the manual block in Britain, where it was universal by 1889.
The first continuous block system was installed on the United Railroads of New Jersey between Philadelphia (Kensington) and New Brunswick in 1865. The company resolution authorizing the installation is dated 27 September 1865. The system was devised by Ashbel Welch, a civil engineer with the United Railroads of New Jersey. In 1867 the system was extended from New Brunswick to Jersey City, so that the complete line from Philadelphia to New York was covered. There were possibly experiments as early as 1863, but that cannot be considered the date of installation of the block system. Isolated blocks for protecting tunnels and other dangerous places were already in use, on the Erie and the Philadelphia and Reading, for example. Welch visited England in 1854, and again in 1869. The block system had just come into use on his first vist, and he probably did not observe it. The later visit prompted the importation of interlocking frames. When the URRNJ was taken over by the Pennsylvania in 1871, Welch's block system and the plans for interlockings were retained with some enthusiasm. The Pennsylvania's own block system was devised and appeared in the Rule Book of 1874. The banner signal used for this block was different from Welch's, and it was operated differently, much like the British block system. By 1876 it had been extended from Philadelphia to Pittsburgh on the double-track main line.
Welch's system was an improvement on the earlier practice of telegraphing notice of a delayed train ahead to hold opposing trains. Welch probably did not know the principles of the British block system, and even later displayed an almost total ignorance of it. A box signal was used with glass-covered openings for each direction, with a black baffle between them. A white reflector 2 ft in diameter, illuminated by a lamp in front of it, and seen through the opening, was the Safety signal, while the absence of the disc was the Danger signal. This disc fell out of sight by gravity when it was not positively raised by means of a vertical rod operated by the signalman. When a train passed the station, a peg was inserted in the vertical rod preventing its operation to clear the signal. Each successive train caused an additional peg to be inserted. As trains were reported out of the block at the other end, pegs were removed until, when all were gone, the signal could then be cleared. The system was thoroughly permissive, the absence of a Safety signal advising trains to move at restricted speed into an occupied block. It included an equivalent to the block instrument, giving a visual indication of the state of the block by means of the pegs.
When a train approached within a half mile, it would give a long, loud whistle. The signalman would then pull on the rod and exhibit the Safety signal, if he could. The train would then give a single whistle sound (at that time, two was the signal for brakes) to acknowledge the signal. Otherwise, the engineman would call for brakes and stop. As the train passed, the signalman would let go of the rod and insert a peg in it. When the train was more than 500 yd in advance of the signal, he would telegraph Train Out Of Block to the post in the rear. A wire was dedicated to the block system.
The signal was later modified, at first to hold the lamp fixed, and then to display a red Danger signal. The stages of the evolution into the Pennsylvania banner signal are not known, but this signal retained the active clear aspect; that is, the signal had to be held at Safety. These systems were used on double track to prevent rear-end collisions. Much of the popular information on Welch's block system came from the recollections of his assistant, Robert Stewart, in 1893, but these recollections were very faulty. Welch himself told the true story much earlier in the Journal of the Franklin Institution.
The Pennsylvania's lead was admired, but not imitated. Only the Erie made a similar commitment to manual block later in the 1880's. While the Pennsylvania did not used the bells and block instruments typical of the British manual block, the Erie at least used the bells. Some changes were made to reflect the different conditions in America. The most important was that freight trains were admitted to blocks already occupied by a preceding freight train, after a warning of the fact. This is permissive blocking, and indeed was also used in Britain at first. Another was the use of only a single block signal instead of separate home and starting signals. The Pennsylvania did use advanced signals later, to allow a train to clear a block so that a following train could be admitted while the train did station work. The block instrument was considered a superfluous expense in the United States, and was not used. Reliance was placed on the block record.
The automatic block arrived with the formation of the Union Switch and Signal Company by George Westinghouse in 1880. The Union Signal Company owned the patents on the track circuit, which were acquired with this purchase. US&S offered the electropneumatic semaphore controlled by track circuits, as well as the clockwork or banner signal inherited from Union. The clockwork signal was much cheaper than the electropneumatic, and though not promoted vigorously, eventually sold 1055 examples, mainly used in the northeast. The Hall Signal Company greatly improved their enclosed disc signal, which joined the competition in 1890. All this interest in automatic block signals, and the threats of the Interstate Commerce Commission to become interested in railway safety, set many railway managements to thinking. While automatic block signals cost good money, the manual block could be implemented with essentially no additional investment, except the printing up of some new forms.
The manual block then burgeoned, with train order signals assuming the role of block signals as well, with telegraph operators for signalmen. This generally meant rather long blocks, 15 or 20 miles, instead of the 1-2 mile blocks normal in Britain. These long blocks almost demanded permissive operation, or line capacity would be seriously reduced. Many companies seriously blocked only passenger trains, which could be a selling point, leaving freight trains to their own devices. In every case, manual block was only supplemental to operation by time table and train order. This fact should be taken into account when comparing American and British manual block operation. Most manual block was later replaced by automatic block in the 1930's.
It was not lost on railway managers that operation by signal indication could offer some distinct advantages, such as the elimination of the tedious delays often associated with train orders. Rules 251-254 provided for the movement of trains by signal indication with the current of traffic on double track, and Rules 261-264 for movement of trains in both directions by signal indication. This is exactly what is done in Britain, where signal indications are authority to proceed. The normal manual block system was seldom trusted to support these rules on single track, but the controlled manual block was. In controlled manual block, a signalman cannot clear a signal unless it is unlocked by the signalman in advance, and in some cases the progress of trains is monitored by track circuits. The final development along this line was Centralized Traffic Control, where the progress of trains was electrically reported, and the dispatcher himself controlled switches and signals.
The Southern Railway operated manual block on the single-track St. Louis and Louisville Divisions between Mt. Carmel, Illinois and Duncan, Indiana, 111.1 miles. From Duncan to New Albany, 5.7 miles, the train staff was used, to protect a tunnel. In 1950, the train staff was abolished and automatic block signals came into use between New Albany and Milltown, 27.7 miles, with manual block on the remaining 89.3 miles. Manual block rules were eliminated in the new Rule Book of 1 August 1956.
Day and night offices were maintained at Mt. Carmel, (11.9) Princeton, (12.3) Oakland City, (24.0) Huntingburg, (15.0) Birdseye, (8.2) Taswell and (17.9) Milltown. The numbers in parentheses are the lengths of the blocks in miles. By day, additional offices were open at Francisco between Princeton and Oakland City, Winslow between Oakland City and Huntingburg, St. Anthony between Huntingburg and Birdseye, English and Marengo between Taswell and Milltown.
The block record form is shown below. On the Southern, this was Form 648A, printed on 6" x 22-3/4" paper, punched with two holes at the right margin. In addition to the words shown, there was a place to the left of the title for seven Employees on Duty, with Name, Occupation, Went on Duty, Went off Duty entries. To the right of this was a list of Undelivered Train Orders, with spaces for Order No. and For. On a line below were four spaces for Position of Train Order Signal. To the right of the title were the Instructions, and a box for the weather at 6 am, 12 noon, 6 pm and 12 midnight. On each side there were 20 or 21 columns, each for one train. The left-hand side was for south or west bound trains, the right-hand side for north or east bound.
I have the block record for the day office at Marengo, Indiana for 21 November to 28 November 1950. There are no entries for Thanksgiving, 23 November, or for the weekend, 25 and 26 November. The entries for different days are separated by vertical ruled lines. There were only from 2 to 4 trains during the day in each direction. The list of employees on duty was not complete, though the initials were all there. There was no train order information. The signals for closing the block station were not recorded.
SOUTHERN RAILWAY SYSTEM
Station Record of Train Movements
At ________________________Block Office
for 24 hours, commencing 12:01 A.M. __________________19_____
South or West Bound ___________________Division North or East Bound
|938||Block Entered||Block Cleared|
|948||Departed||Clear Block Received|
|930||Clear Block Received||Departed|
Let's see how the block record was filled out. As an example, the entries for Train No. 24, a westbound passenger train, are shown for 27 November 1950. The first thing that would happen was Milltown would ask Marengo for the block, with the message "1 for 24". Marengo would reply "13, 2 for 24" and enter the time, 9.29 am, which was 10 minutes before No. 24 was due to leave Milltown. The signal "13" meant "I understand." Marengo's eastbound train order signal was already at Stop, since this was its normal position when an operator was on duty. Marengo almost immediately sent "1 for 24" to English, who replied "13, 2 for 24" at 9.30 am. No. 24 entered the block at Milltown at its time table departure time there, 9.38 am. Milltown sent "46" to Marengo at 9.38 am. No. 24 arrived Marengo at 9.46 am and departed at 9.48 am. Marengo sent "46" to English at this time, and "2" to Milltown. The operator recorded these times, and sent the leaving time to the dispatcher as the OS report. The letters CTH are the initials of the Marengo operator on duty for this train. Marengo received "2" from English at 10.02 am, which was the last entry for No. 24. The final box is for AM or PM. If the telephone was used, words replaced the numbers in the messages, as "Block for No. 24," "Right; Block clear for No. 24."
Freight trains were allowed to follow freight trains into an occupied block by permissive signal or card. To offer such a train, "17 for Extra 4143" could be used. The reply "5 of 56, 13 for Extra 4143" means the block is not clear of No. 56, block asked for Extra 4143. The block signal could then display the Permissive aspect for Extra 4143. If the reply were "56 of 24" the extra could not be admitted. In general, it was necessary to distinguish between passenger trains and other-than-passenger trains. Instead of "1 for 24", the block would be requested with "36 for 24". "36" is block wanted for passenger train, while "3" is block wanted for train other than passenger. In order to accept the passenger train, the response would have to be "2 for 24". If there were a train in the block, "5 of ..." would mean block not clear of identified train other than passenger, while "56 of ..." would mean block not clear of passenger train. In either case, a passenger train could not be admitted to the block. However, if the block was requested by "3", a freight train could be admitted to the block if the response was "5 of ...".
Centralized Traffic Control and radio communication made operators superfluous, while the disappearance of passenger trains eliminated the necessity for stations and agents. Stations, operators, train order signals and companies disappeared after about 1970, and with them the manual block.
For more information on block systems, see the article on that subject on this website.
W. F. Cooke, Telegraphic Railways (London: 1842). A copy is in the Science Museum Library, London.
B. B. Adams, The Block System of Signaling on American Railroads (New York: Railroad Gazette, 1901).
J. A. Anderson, The First Block Signal System in America, Jour. Rly. Signal Assoc. 6, 13-19 (1909).
Composed by J. B. Calvert
Created 12 August 2004