The American Railway: Its Construction, Development, Management, and Appliances

by Thomas Curtis Clarke

Derailing Switch.

We left the signalman trying to lower signal No. 2; vainly, because No. 3 lever was still normal and the switch unlocked (Diagram, p. 205). Probably he would not have begun his operations in the bungling way that has been supposed, but would have first reversed lever 3. That locks the switch by the facing-point lock, and locks also switch-lever 4 in the frame in the signal-tower and releases lever 2. Then he reverses lever 2. That locks lever 3 and releases lever 1. Then he reverses lever 1, which locks lever 2. Now the way is made for a train to pass east on the main line, and the signals are clear. The last signal could not have been lowered until the chain of operations was complete; none of the levers can now be moved until lever 1 is again put normal and signal 1 made to show danger. There is one point of great danger in this particular train-movement which has not been mentioned; that is, the crossing of main-line east-bound track B by the branch-line west-bound track C. It will be noticed that with the levers normal, derailing switch 5 is open, and it is impossible for a locomotive to pass beyond it. Lever 5 is interlocked in the tower with lever 4 in such a way that, before 5 can be reversed to let a train pass west from C, lever 4 must be reversed to trap any train on B and turn it down the branch D. It must not be understood that the use of "derailers" is universal. In fact, they are not recommended by the best signal engineers, except in special conditions. In the absence of derailer No. 5, signals 11 and 12 would be interlocked with switch 4, so that, so long as that switch stands open for the main line a clear signal cannot be given to a train coming west on C. It will be noticed that signal 2 carries two semaphores on one post. The upper one is for the main line and the lower one for the branch. Both are operated by one lever, 2, and whether reversing lever 2 lowers the main-line signal or the branch signal depends on the position of the switch. The switch is made to pick out its signal by an ingenious but very simple little arrangement, called a selector, which is placed somewhere in the line of ground connections.

It would be an interesting study, were there space, to follow the possible and proper combinations of movements to pass trains over the various tracks. It will be seen that, by concentrating the levers which move switches and signals in one place and interlocking them, it is made mechanically impossible for a signalman to give a signal which would lead to a collision or a derailment within the region under his control. The only danger at such points is that an engineer may overrun the signals. This description of the objects and the capacity of the system of interlocking is no fancy sketch. The system has been in use for many years, doing just what has been here described, and more. A recent close estimate gave the number of interlocked levers now in use in the United States as about eight thousand, and the number is rapidly increasing. Recent official reports showed that in Great Britain and Ireland there were thirty-eight thousand cases in which a passenger line was connected with or crossed by another line, siding, or cross-over. In eighty-nine per cent. of these cases the levers operating the switches and protecting signals were interlocked.

The example of interlocking which has been given is one of the simplest; the principle is capable of almost indefinite expansion, and any one lever may be made to lock any one or more levers among hundreds in the same frame. The greatest number of levers assembled in any one signal-tower in this country is one hundred and sixteen, at the Grand Central Station in New York. In the London Bridge tower there are two hundred and eighty levers. This is probably the greatest number in any one tower in the world. All of these levers may be more or less interlocked. The same principle is applied to the locking of two levers at a single switch, and to the protection of drawbridges and highway crossings.

The mechanism by which the interlocking is done is strong and comparatively simple, but a detailed description of it seems out of place here. Two levers from a Saxby & Farmer machine are shown on [page 204], with lever A normal and B reversed. The locking mechanism is in front of the levers, and is actuated not by the levers themselves, but by their catch-rods. It follows that it is not the actual movement of a signal which prevents the movement of other signals, or of switches, but it is the intention to move that signal. This principle of "preliminary locking" is one of great importance.

Switches and signals are often worked at such distances from the tower that it is impossible for the operator to know whether or not the movement contemplated has taken place. The British Board of Trade does not permit switches to be worked more than 750 feet away. In this country there is no limit, but probably 800 feet is very rarely exceeded. Signals are worked in England up to 3,000 or 3,500 feet very commonly, and they are even worked a mile away, but not satisfactorily. This is with direct mechanical connection, by rod or wire, from the levers. It is obvious that a break in the connections between the lever and the switch or signal might take place, and the lever be pulled over, without having produced the corresponding movement at the far end. The locking mechanism in the tower would not be affected by such an accident, and consequently conflicting signals might be given. Even this contingency is provided against with almost perfect safety. If a signal connection breaks, the signal is counter-weighted to go to danger. The worst that can happen is to delay traffic. If a switch connection breaks, the locking-bolt, in the latest form of facing-point lock, will not enter the hole in the switch-rod, and consequently warning is given in the tower that the switch has not moved. Electric annunciators are often placed in the signal-tower, to show on a board before the operator whether or not the movements of switches and signals have taken place.

Considerable work must be done in the movement of each lever. The ground connections must be put down with great care, as nearly straight and level as may be, well drained, and protected from ice and snow. All of these difficulties have been overcome in a beautiful pneumatic interlocking apparatus which has been introduced within the last two or three years. In this system the motive power is compressed air. Near each switch is a small cylinder, containing a piston which is attached directly to the switch movement. Compressed air admitted to one side or the other of this piston moves the switch one way or the other. But, as it would take some time for the necessary quantity of air to flow from the signal-tower to a distant switch, a small reservoir is placed near the switch, and the air from this reservoir is admitted to one end or the other of the switch cylinder according to the position of a valve. For transmitting the motion from the tower to the valve compressed air might be used, but, as air is elastic, a quicker movement is got by using in the pipes some liquid which does not readily freeze, and which, being practically non-compressible, transmits an impulse given at one end almost instantly to the other. The signals are worked in essentially the same manner as the switches, except that the pneumatic valves are moved by electricity. The tower apparatus of a pneumatic system in the yard of the Pennsylvania Railroad at Pittsburg is shown in the engraving opposite. In the front of the apparatus is seen a rank of small handles, which can be turned from side to side with as much ease as the keys of a piano can be depressed. Turning one of these handles admits compressed air to the end of a pipe containing liquid. Instantly the pressure is transmitted 500 or 1,000 feet to the valve at the switch to be moved. The small levers are interlocked perfectly, and in that particular perform the duties of the ordinary machine. A model of the tracks controlled is placed before the operator, showing the switches and signals, and when a movement is made on the ground it is at once repeated back by electricity and duplicated on the model. This beautiful system is due to the same genius that gave us the perfected air-brake and the triple valve, and is the greatest improvement that has been made in interlocking in the last dozen years.