THE TRAIN-ORDER SIGNAL.
A method much used for signaling a train to stop for orders is to display a flag or light of suitable color, after receiving the direction to "hold the train." This is often done by holding the signal in the hand or placing it on the platform or ground or in some fixed place. If placed on the platform, without attendance, it is liable to be obscured or removed by persons about the place. If held, in the hand of the agent or operator it is a poor arrangement for performing so important an office. The operator is usually required to report that the signal is displayed. He evidently cannot do this without leaving the signal unattended, and in fact when he is alone he must so leave it, as, after it is displayed, he must return to the office to receive the order, and he must also often be engaged in his office while expecting a train. It will frequently occur that trains will pass his station after he has received an order for some subsequent train; in which case he must temporarily remove the signal, or stop a train which might otherwise not be required to stop. When this plan is used all trains that arrive before that for which the order is held are actually stopped. A serious accident occurred some years since from the hand-lamp going out as it was swung as a signal to stop a train for which orders had been received. The signal failed, and the train went on and collided with the opposing train. Lanterns and flags are the only available movable signals to be put in the hands of train and track men, but they should not be relied upon where anything better can be used. The evils attending this use of hand signals are so manifest that the practice is fast disappearing, and the reference to it here may before long be only a reminder of what has been done.
A signal for this purpose should be distinctive and of the most substantial character. A fixed signal manipulated from within the office is greatly to be preferred. Several such have been devised. The signal should be such as to be distinctly seen at proper distances; it should be as little as possible liable to confusion with other objects, and it should be an adornment rather than a disfigurement to the landscape in which it forms a prominent feature. The most satisfactory signal within the author's knowledge is the simple semaphore arm, extending horizontally from a post and showing a red light to signify "stop," and inclined and showing a white light to signify the opposite, and operated by a handle within the telegraph office.
Much discussion has been had in the past as to whether a danger signal, which this preeminently is, should stand normally at safety or danger. The earlier practice favored the former, as indicated, above, the absence of a signal, in the plan described, being the rule. In more recent years the weight of opinion has been that in all systems of danger signals the normal position, and that to which such signals should automatically move, is that indicating danger. So arranged, the indicator will always be in a position to stop trains unless it is moved to another position to show that there are no orders for them. It becomes a standing order to "hold," and, when an order is forwarded for a train, the fact of its receipt requires that the signal be simply left in its normal position and the train thus stopped. It will be then the rule and the habit of trainmen to observe all these signals and to stop when they are not placed, on their approach, in the position permitting them to proceed.
The rules of many railroads still indicate a usage contrary to this. The lamp, flag, or other stop signal is displayed only when a train is to be stopped for orders. It appears that under some circumstances, especially where the duties of the agent and operator are performed by the same person, the telegraphic duties being comparatively small, it is thought better to retain this method, and the rules of the Time Convention were so framed as to provide for either, leaving the choice to those concerned. Under the "normal at danger" plan, when an order is received in advance of the arrival of the train for which it is designed, and has been properly verified and prepared for delivery, it remains in the hands of the operator until the train arrives, the signal showing "stop." If, in the mean time, other trains pass for which there are no orders, the signal must be placed, as they approach, so as to indicate that they may pass. But there is then the danger that the operator may inadvertently allow the train to pass for which he has an order. This has actually occurred, and should be provided against. This should be done by requiring that, as soon as an order for a train not arrived is ready for the signatures of the trainmen, or for delivery when signatures are not taken, the copies designed for them shall be removed from the book, folded, and marked with the train number, and put in a designated place and in such position that the signal handle cannot be moved without the eye and hand being directed to the orders. This is readily effected by a rack to hold the orders placed on a small door closing by a spring and catch over the handle by which the operator moves the signal. The handle cannot be moved without unfastening the door and so opening it as to bring the orders, which are on it, under the eye and hand of the operator. This precaution may appear trivial, but while it is of great importance to adopt such routine that its mechanical performance will lead to a correct result, it is equally important to interpose such obstacles as are necessary to prevent a mechanical inadvertence that may lead to disaster. The same kind of risk exists in the use of block signals, and several plans have been used to obviate it by suitable mechanical means. In the other use of the train-order signal there is, to a certain extent, the same liability to this unconscious movement when it has been placed at danger, and a like precaution is needed to guard against it. It often happens that there are orders on hand for several trains. A definite place for them prevents their getting mixed with each other or with other papers; and removing them from the book avoids the necessity of leafing them over to find the particular order which men may be waiting to sign, and possible mistake in getting the wrong order.
The only reason of apparent moment that could be assigned for leaving the orders in the book is that the trainmen may sign all the copies. There does not appear to be any good reason for requiring their signatures on their own copies, and the manifold writing by them would be unsatisfactory. Again, it will often happen that more than one train is to receive a copy of the order, in which case the same signatures are not wanted on all the copies. The point here urged as of paramount importance is that the order itself shall be interposed between the operator and the instrument by which he might give a signal permitting a train to pass improperly. In this view the discussion of the point is pertinent to the subject of "Signal." It may be added that the final indorsement of "complete" after signature on each copy takes but a moment, and perhaps no longer than a careful writing of it over several copies in the manifold-book; and as the men should read and compare their copies before the final steps, it is difficult to see how they could do this properly if the orders remain in the book.
The train-order signal should be used for no other than its legitimate purpose. It will not be inconsistent with this to use it for holding a train the required time after the passage of another train in the same direction.
Upon some roads, trains passing while the stop signal is shown receive a "clearance" card stating that orders in hand are not for them. This is included as a part of the plan presented in the Time Convention rules for the use of the signal with its normal position at safety. It would seem to be necessary with this method; and in any case where it can be used it is a valuable precaution, the only objection being that it requires the stopping of fast or heavy trains which it might be quite objectionable to stop. This would seem, however, to be proper for any train stopped by the signal for time.
Where the plan is adopted of keeping the train-order signal normally at safety it should still, as in the other system, be so arranged that it will move automatically to danger if any of the mechanical parts fail. If this is not done and dependence is placed on fastening it at danger, the fastenings or some of the connections may fail and the signal move to safety without the fact being observed. One important advantage of the other plan of using the signal is that it is never at safety excepting when held in that position by the operator. Where the usual position is safety it cannot be arranged for the operator to actually hold the signal while it occupies the danger position.