UC-NRLF Eb 315 Block and Interlocking Signals. BY W. H. ELLIOTT, SIG. ENG., C., M. & ST. P. R.R. WHAT THEY ARE FOR. WHAT THEY DO. UNIVEBSITY ] H W THEY DO IT. LOCOMOTIVE ENGINEERING, NEW YORK. 1896. Copyrighted, 1890, by SINCLAIR & HILL, New York. CONTENTS. PAGE CHAPTER I. Block Signaling: What it is for, What it does, How it does it - - i CHAPTER II. Methods of Operation and Rules 17 CHAPTER III. Construction The Telegraphic Systems 33 CHAPTER IV. Construction The Controlled Manual Systems . ... 51 CHAPTER V. Construction The Automatic Electric Systems 65 CHAPTER VI. Construction The Automatic Electric Systems, con- tinued 83 CHAPTER VII. Construction The Automatic Mechanical and the Staff Systems 101 CHAPTER VIII. Installation and Care of Automatic Electric Signals, with a Comparison of the Cost of the Different Block Signal Systems 123 CHAPTER IX. What They are For and How They are Operated .... 143 CHAPTER X. For Junction Points and Drawbridges Construction of the Improved Saxby & Farmer Interlocking Machine 165 CHAPTER XI. The Stevens Machine, and How a Switch is Moved and Locked 183 CHAPTER XII. Details of Construction 199 CHAPTER XIII. The Westinghouse Electro-Pneumatic and the Gibbs Electric Street Railway Systems 220 CHAPTER XIV. Agreements, Contracts, Specifications, Installation and Repairs 241 CHAPTER XV. Switch Signals 257 OF THK UNIVERSITY BLOCK SIGNALING. WHAT IT IS FOR. WHAT IT DOES. HOW IT DOES IT. By W. H. ELLIOTT, SIGNAL ENGINEER, C., M. & ST. P. R.R. CHAPTER L "What are we stopping for, conductor, out here in the woods? This is a limited train. What! stopped by a signal, a block signal, you say? Why, what is that? Oh, I see! You have a red blade projecting from the top of a pole to indicate to the engineer when the blade is moved up or down whether he may enter the block or not, the block being the piece of track extend- ing to the next signal. So, then, when we are stopped by such a signal it means that another train is in the block, and we will have to wait until it has passed out." And thus it is that to-day trains are being run through towns and cities, over mountains and prairie, through bridges and tun- nels, in cuts and around curves with absolute safety, a fact not fully appreciated by the traveling public, but which becomes to the engineer, whose responsibility is lightened and from whom anxiety is removed, a guiding star, telling him that the track is his and that there will be no one to dispute it with him, for such little arguments, you know, are sometimes disastrous. Block signaling, though limited in extent in this country, in proportion to the miles of track operated, is so rapidly being ex- tended, not only from the natural increase of business and conse- quent demands for a safe method of operation, but from the general knowl- edge being acquired of the advantages to be gained from such a system, that I believe an article on the subject would be both inter- esting and instructive. To the man well posted on sig- nal matters, little that is new will be found, as this article is written more for those who are constantly guided by a signal, but have little idea of its construction. The commencement of signaling may be said to begin with the use of the locomotive, for it soon be- came manifest that some- thing would have to be de- vised, not only to prevent collisions between trains, but to give information to engineers regarding the position of switches and the right to go ahead. Many forms and devices were used in these early days, few of them being seen to-day, but which, as in the development of the locomotive, became step- ping stones to things much better. As each engineer pre- 1. Home Block Signal -"All Clear." 2. Home Block Signal " Danger, Stop." 3. Distant Signal " All Clear. 4 Home Block Signal-' 1 Danger. ferred his own devices to those of others, it followed, as a matter of course, that the practice was very varied, so much so in some cases that the safety signal on one road became the danger signal of another. Naturally enough, this state of things brought about many serious accidents, and finally resulted in a meeting being held by those interested, for the adoption of a standard form of fixed signal to be used by all the roads. The choice fell upon the "semaphore," a signal designed by Mr. Gregory in 1841, which indicates by position and not by its form whether the track is clear and the train has a right to proceed. It was decided that a horizontal position of the blade should indicate "danger" or "stop;" a vertical position, "all clear" or "go ahead," and a position midway between these two, making an angle of forty-five degrees with the horizontal, "caution" or "pro- ceed carefully." Its construction was very much the same as that used to-day, consisting of a blade pivoted at the top of a pole and capable of being turned through about a quarter of a circle. The colored glasses for giving the night indications were carried in a separate frame pivoted lower down on the pole, instead of being held, as in modern practice, by the casting to which the blade is fastened. The blades for governing train movements in one direction were always put on the same side of the pole. In this country, the blade projecting on the right-hand side of the signal pole, as looked at from an approaching train, is the one that governs. In England, where all trains run on the left-hand track, signal blades projecting to the left side govern. The signals were operated under what is called the -"time interval system;" that is, not allowing one train to follow another into the block until the lapse of a certain period of time. When a train entered the block the signal was put at danger and kept there for five minutes, when it was pulled to a cautionary position, and after the lapse of five minutes more the signal was "cleared," giving the right to the next train to proceed. Experience with this method of operation soon demonstrated that the principle was not correct. For while a train may have passed, a certain length of time, the signal gave no indication of how far it had gone. The many accidents occurring under this system stimulated the invention and adoption of electric indi- cators and telegraph instruments as a means of communication between signalmen, making it possible to keep a space interval between trains, and not allowing a second train to enter a block while it was occupied by the first. This method of block- ing, keeping a space interval between trains, is the end to which all the modern systems of signaling are designed, although the methods by which this result is obtained vary considerably. The next important principle to be developed was that the normal position of a signal should be at danger, not at safety ; or, in other words, to assume that danger existed unless known to the contrary. For with a signal constructed as were those first used, on an accident happening to the apparatus, or in case of any of its parts becoming disconnected, it would at once fall to the "all clear" position, and no protection would be afforded a train in the block. An engineer, of course, not knowing that the signal was out of order, would take the indication as one intended for him and proceed accordingly, a result very likely to cause trouble, but for which he could not be blamed. With the signal always remaining in the danger position^ and being so constructed that any accident or breakage of the appa- ratus would cause it to assume the danger position, no accidents from this cause can happen, as trains would be stopped instead of being allowed to proceed. In the most approved systems, the signal automatically re- turns to the danger position immediately upon a train entering the block, thus making it impossible for a second train to enter, as might easily occur under the old system should the signal- man fail to return the signal to the danger position. As showing the commencement of block signaling in this country, the exhibit of the Pennsylvania Railroad at the World's Fair, of a pole and ball signal used on the Newcastle & French- town Railroad in 1832, is very interesting, the following infor- mation being obtained from a letter written by Mr. A. Feld- pauche, principal assistant engineer of the P., W. & B. Ry., in regard to its use. The road was about 20 miles long, and the signals seem rather to have been used for conveying information from one end of the line to the other than for that of a block. "When the train was just starting from Newcastle, the man in charge of the signal at that point raised the ball to the top of the pole. The man at the next station, seeing the white ball raised by the first man, raised his ball to half the height of his pole. The men at the other stations, each on the lookout with his telescope, which, you will see in the cut, were placed in the guides provided for the purpose on the side of the pole, also raised their balls to half mast, thus conveying the information throughout the line that the train had started. OF THF. UNIVERSITY "When the train reached the first station, the man would immediately raise his ball to the top of the pole, as a signal both ways that the train had reached him, lowering his ball when the the train reached the next man ahead, this being repeated suc- cessively at each of the four stations. "When a train, having passed one station, did not arrive at the next, or was seen to be in trouble in any way, the man at the station next nearer Newcastle would lower his white ball and substitute therefor a black ball, kept at hand for the purpose, and would raise it to the top of his pole as a signal to be successively transmitted to Newcastle, whence a relief train would be dis- patched to the assistance of the regular train." Block signaling may be said to be practiced in this country in two ways that of "absolute" blocking, in which one train only is allowed to occupy a given block, and "permissive" block- ing, where, under certain regulations, more than one train is allowed to enter. From the definition of a block "a section of a track between two signal stations, the use of which is controlled by fixed sig- nals" it is seen that where absolute blocking is maintained, both head and rear-end collisions are impossible, and were it not for the expense and occasional delay to traffic, such would be more generally practiced. It is a fact, which experience is demon- strating every day, that more trains can be run over a given piece of track and with greater safety by a properly arranged block sys- tem than by any code of rules that can be devised, and although one cannot show in figures how much can be saved to a road by immunity from accidents, it will certainly repay any investments made in apparatus and operators' wages. More particularly is the problem of block signaling increasing in importance as the traffic on many roads is becoming too dense to be handled without such a system. Managers who thought that the results obtained by the English with absolute blocking could be ignored on account of the different conditions of opera- tion in this country, are gradually finding it to be the only safe way to operate their roads. In point of fact the means are already at hand, and any road having a telegraph wire and operators can, at a moment's notice, put the absolute blocking of all trains into effect, should condi- tions arise under which it would be desirable to do so. That many roads do not take advantage of this and train their operators is much to be regretted, and it is to be hoped that with the spread of information regarding the working of block systems and the safety to be gained by their use, managers will come to a full appreciation of their merits. With roads using a block system, where, from considerations of expense, the blocks are of a greater length than is advisable, "permissive" blocking has, with certain restrictions, come very much into use. Although it is an abandonment of the "space" for that of a time interval, the results obtained are such as to make its use in many cases a matter of good business judgment. Before describing the different methods of operating block FIG. 5. signals, it will be well to describe the construction of a sema- phore signal and to discuss the interesting questions connected therewith. As will be seen in Fig. 5, a casting is pivoted at the top of a pole which holds the blade and colored glass; this casting is called the arm plate. The blade is a thin board 5 feet long, taper- ing from 7 inches in width where bolted to the casting to 10 inches at its outer end. The end of the blade is often pointed, to more easily distinguish the block signal from other semaphore signals. The height of the pole, ordinarily, is about 25 feet above the ground, and a ladder bolted to its side allows of easy access to all the parts. About midway of the pole is an iron lever called the balance lever, to which are attached the wires for operating the signal. An up-and-down rod connects this lever with the arm plate cast- ing, being attached to the casting on the opposite end from the blade. By putting a weight on the balance lever, it is seen that the blade will be greatly overbalanced, and will, of course, be held in the horizontal position until a force is exerted on the lever sufficient to lift this weight and lower the blade. This construc- ton fulfills the requirement that the "normal position of the signal should be at danger," as it calls for a direct effort on the part of the signalman to change the signal to the "all clear" position. To give the different indications at night, a lamp is so placed that the light, when the signal is at danger, will show through a colored glass held in the arm plate casting, and will show an un- obscured white light when at "all clear." While the signals for use in the day-time were going through the various changes from one of "form" to that of "position," those for use at night simply became a question of color, as nothing has as yet been found which compares with it in distinctness and simplicity. This ques- tion of the proper color to be used for the different night signals is one of great interest to all railroad men, and one which is being widely discussed. What engineer has not had to pass an examination for color blindness, or has not felt somewhat "wrathy" when a sleepy op- erator has allowed his signal lamp to get low or go out? Who has not felt his pulse quicken when, on some dark night, a red light has suddenly appeared on the track ahead, even though it should prove to be only a "wide-awake" drummer anxious to get out of town, and who has stopped the train by putting a match inside of a red bottle? 9 Red is used for the danger or stop signal everywhere, as it makes the greatest impression -on the sense of sight. White is used for the "all clear" and green for the "caution" signal on most roads in this country, although a few follow the English practice of using green for the "all clear," white not being used for any signal. The use of white to indicate "all clear" and green for "caution," at first sight appears to answer all the requirements; but so strong are the arguments against this arrangement, that if a more satisfactory color could be found for the cautionary signal, green would be universally used for the "all clear" signal and the use of white abandoned. These arguments are, first, that the glass fastened in the arm plate casting may break and show a white light when the signal stands at danger. Second, engineers may mistake a light in some street or dwelling for the signal light and run by it. Fe\v accidents have happened from the glasses breaking, but the possibilities of a serious collision are always present. For this reason stationmen should always "keep an eye" on the signal to see that the glass is in its place, and engineers should, when- ever possible, see that the position of the blade corresponds with the indication given by the lamp. The use of green for the all clear signal overcomes both ob- jections against white, but leaves no available color for a caution- ary signal. The practice on one road using green for the "all clear" is to show both a red and a green light for the caution signal, blotting out the red for "all clear." On another road three lights are used, two showing in a horizontal line to indicate "caution," and two in a vertical line for "all clear." The color the blade is painted has nothing to do with the indications given, for while the blade may change its position the color does not, and consequently only one indication could be made. The blade is painted red for distinctness, that being the most easily discernible color; but on this point opinions differ, as one very prominent road paints them yellow. It may be well to note here that there is one system of blocking, "an automatic electric," in which the different indications are made by a change of color, or rather the appearance of a red disc for "danger" and its absence, thereby showing a white background, safety. 10 The next to be considered (having seen how a signal is con- structed) are the different signals to be found at a block station, with the meaning each is intended to convey; not that all are to be found at every station, but that it may be better understood how a good block signal system is operated. At first the only signals used were those at the entrance of a block, these, for economical reasons, being put at stations or at points where switch tenders were already stationed. As traffic increased and the speed of trains became greater, it often hap- pened when the signal was at danger that trains would run by, owing to the location of signal, or to the conditon of the weather being such as to prevent engineers from seeing the signal soon enough to stop. The result of this was that engineers were forced to slacken speed and approach the signal very carefully, so that if found at danger they could stop before passing it. To make this unnecessary, a second signal was erected which would give the same indication as the first signal gave, being placed some distance down the track from which the train was approaching; the engineer, by this arrangement, being informed of the position the controlling signal would be found in, some time before reaching it. For the sake of distinction, the more important signal the one controlling the block is called the "home" signal, and the other, or caution signal, the "distant" signal, these being the names by which they are known to-day. The possibility of operating the distant signal from the same place as the home signal, strange as it may seem to us now, was not thought of for some time ; the distant signal being placed only so far down the track as it was possible for a man to run, after first putting the home signal in the clear position, before the arrival of the train. A bright switchman, anxious to save himself the trouble of so much running, was the first to think of connecting this distant signal by means of a wire to a lever in the tower in which he was stationed. That this should have escaped the engineers and have been thought of by a switchman, recalls the invention of the valve motion by the lad who, getting tired of working by hand the steam engine valve, attached it to the end of the engine shaft by a stick. II Both signals were originally of the same form, but owing to the necessity of making a distinction of some kind between them, a notch was cut in the end of the distant signal. To make a still further difference, the distant signal blade was painted green a color which expresses the character of the indication given by the signal. A very amusing anecdote is told by Mr. W. J. Williams, traffic superintendent of the Brighton Railway, England, of the origin of the notch. Thinking that a distant or caution signal should be different from a home or stop signal, he sent a workman to a station on the main line between London and Brighton to cut a notch out of the end of the distant signal blade. The Brighton tracks are at this point used by the Southeastern Railway Co., and two or three days after the notch had been cut he received an in- dignant letter from that company, asking why he allowed his signals to get into such a state of disrepair that large pieces were actually chipped out of the end of them. Soon after the distant signal came into use, the cautionary indication, as given by the home signal blade, became, to a great extent, discontinued, owing to engineers not properly observing it. For a "cautionary indication" is really permissive blocking, a time interval between trains, and, unless great care is used by the engineer, accidents are very likely to happen. Another fact that made its use objectionable was the difficulty of keeping the signal properly adjusted, for, unless it was always lowered to the same position, engineers would be in doubt as to the exact meaning intended. And if the safe side were not taken, serious consequences were likely to follow. Where permissive blocking is used the best systems do away with the inclined position, and either stop the train to give a "per- missive card" or else use two signals on the same pole, the upper one being used for the "danger" or stop signal and the lower for the permissive or cautionary indication. This arrangement con- sists in placing a permissive arm, painted green and notched on the end, or a green light at night on the pole below the block arm and to work in connection with it, as shown in Fig. 6. The indications as given by these two blades on one pole are plain and unmistakable, and are as follows : 12 Block and permissive arms horizontal, or upper light red and lower light green, signifies "Danger, stop!" Block arm vertical and permissive arm horizontal, or upper light white and lower light green, signifies "Caution, proceed slowly!" Block and permissive arms vertical, or upper and lower light white, signifies "All clear, go ahead!" Danger. It is here that attention should be called to the two ways in which caution signals may be read, for unless clearly under- stood, the indications as given by each signal will not be correctly interpreted. Not that the indication for "caution" does not mean to exercise due vigilance and care in either case, but that the extent to which caution is to be observed varies greatly. One indication for caution is given by the signal blade in the inclined position, or when the lower arm of a two-blade signal is in the horizontal position, the upper one being vertical; the other as that given by a distant signal when at danger, in- dicating the position in which the home signal will be found. The 13 one is permission blocking, pure and simple; the other a warning to the engineer to use caution in approaching the home signal, expecting to find it at danger. The two signals at a block station, the home and the distant, were for a long time all that were necessary to properly handle trains without causing serious delays. But with the congestion of traffic as found on many roads, these have proved deficient, and a third signal has been added in many instances, thereby in- creasing the number of trains it is possible to run over a division in a given time. The third signal is aptly named the "advance" signal, from the position in which it is placed, being put far enough in advance of the home signal to allow a train to clear the latter at least 300 feet before being brought to a stop. The indications of the advance signal are positive, the same as those of the home signal the horizontal position of the blade mean- ing "Danger, stop!" a vertical or inclined position, "All clear, go ahead!" Without this signal it was often found that trains working at stations delayed following trains, from the fact that not having passed the home signal the block was not clear, and until it was the other train had to wait. With the use of the advance signal it was possible to so locate the home signal as to make a short block of the track between these two signals. By making this short block include the station and side tracks where switching was done, a train standing at the station would have cleared the block just behind it and at the same time remain under the control of the signalman. The location of the home and advance signals at any station is pretty well defined by the character and amount of the business transacted. That of the distant signal, however, is one that will vary with each locality, and calls for the exercise of care and good judgment, for on the position in which it is placed, more than with any other signal, will its usefulness depend. As the signal is intended to repeat the indications of the home signal, it is necessary that the signal be placed at such a distance as to enable a train after passing it to stop before reaching the home signal, no matter what the conditions are. Common practice in this respect is to put the distant signal 1,200 feet from the home signal, unless the conditions are such as, from the speed of the trains, or on account of grades and curves, it cannot be seen. There is a limitation to the distance it is possible to operate such a signal mechanically, owing to the difficulty of properly caring for the expansion of the wire and also the power required to "clear" the signal. The lever for this signal must also be interlocked with those of the home and advance signals, so that the signal cannot be pulled to "all clear" until they have both been cleared, thereby making it impossible for the signalman to make a mistake. Being placed at a distance from the home signal and the first seen by the engineer, it has become, in practice, the governing signal, allowing trains to keep a uniform speed under all conditions of operation. The number of signal poles in use at any station or tower varies, of course, with the system used, from that where the two signals for trains running in opposite directions are carried on one pole, to where six poles are used, each signal being placed in the best position with reference to the track which it governs. Before taking up the different methods of operating block signals, it is perhaps advisable to call attention to the principal points which have just been considered, and which it would be well to bear in mind. "Absolute blocking," or the maintenance of a "space" interval between all trains, is the only sure method of preventing collisions. The indications of a semaphore signal are made by the posi- tion of the arm, and not by its form or color. The normal position of all signals must be at danger. A semaphore arm displayed to the right of the signal pole, as seen from an approaching train, is the one that governs. A horizontal position of the semaphore arm, or a red light at night, means "Danger, stop!" A horizontal position of a semaphore arm that is notched in the end, or a green light at night, means "Caution, go slow!" A vertical position, or one nearly so, of a semaphore arm, or a white light at night, means "All clear, go ahead!" A block is a section of track between two signal stations, the use of which is controlled by fixed signals. A home block signal is a fixed signal at the entrance of a block to control trains entering said block. A distant block signal is a fixed signal of special form used in connection with the home block signal, and placed at such a dis- tance as will enable all trains to stop between the distant signal and the home signal. An advance block signal is an auxiliary fixed signal, placed in advance of a home block signal to control trains that have entered the block. ^^^c^TToT?^^ ^ese J B^ V* OF THK ^r UNIVERSITY & CHAPTER II. METHODS OF OPERATION AND RULES. There are three general methods of operating block signals, under which all the different systems may be classed. These are respectively called the "Telegraphic," the "Controlled Manual" and the "Automatic," the last-named including the "Automatic Mechanical," as well as the "Automatic Electric." Of these let us first consider the telegraphic method, as, from its simplicity and cheapness, it is in use on more miles of road than any other; its name being derived from the means by which communication is had between the different block stations for the purpose of ascertaining whether or not the block is "clear." The equipment of a station consists, primarily, of a signal for controlling trains, which, although preferably of the semaphore type, very often is not; of a lever for working the signal, placed in a position most convenient to the operator; of a wire used in connection with the ordinary telegraph instruments, or an electric bell for conveying the information necessary to properly work the block. The telegraph instruments of a division may all be put on the same wire, in which case it can only be used by one operator at a time, and every other operator can hear what is being said; or else the wire may run from one station only to the next, and thus be a local wire and ready for use at all times. That there is an essential difference in the manner in which these two arrangements are operated can be seen at once, although the result desired is the same in both. With the first arrangement the train dispatcher is expected to keep track of the operators and see that they properly report to the stations on either side of them the arrival and departure of trains. The dis- patcher may be expected, in some cases, to give an order for the "clearing" of each signal, thus making him entirely responsible for the blocking of trains and allowing the operators no discretion in the matter. With the second arrangement, where the block wire extends only from one station to the next, the operator alone i8 is responsible for the proper blocking of trains, reporting their arrival and departure to the stations on either side of him and clearing the signal only when the block is clear. There is, of course, with this latter arrangement the usual train dispatcher's wire in each office, but it has nothing to do with the block system and is only used by the operator to notify the dispatcher of the movement of trains. Of the two the latter is much the better plan, the advantages to be gained by using a separate block wire be- tween each two stations being that it places the responsibility upon each operator and that fewer men will be needed, as the dis- patcher will be relieved from the routine work of blocking trains and can devote his time to fixing meeting-points. Each operator is provided with a train register sheet, on which he records the arrival and departure of trains as reported to him and as he reports to others. The sheet is divided into two col- umns by a vertical line, the record of all trains in a given direc- tion being placed on the same side. These sheets are kept on file at each station, but should any question arise as to the acts of an operator, they can be sent for and compared with the sheets from other stations. To prevent operators from making mistakes and giving the wrong signal, there is quite a difference in the means adopted on the various roads. On some roads, where the signals are normally kept at danger, the operator is required, before clearing his signal for an ap- proaching train, to ask the operator at the next station ahead if he can do so, although his train sheet may show that the last train admitted has passed out of the block. He is not allowed to hook or fasten the signal lever in the position corresponding with the "all clear" of the signal, but is required to hold it there so long as it is necessary to keep the signal at "all clear," a method very certain to insure the signal being returned to "danger" as soon as possible. Others depend entirely on the train sheet, assuming that in case of doubt the operator will ask the next operator and find out if the block is clear. Others, again, keep the signal at "danger" only so long as a train is in the block, clearing it as soon as the train has passed the next station. p^ H W O o o KH w o 00 w o ^ |d i I CO A a S ci a < o o i? fc H cu Sc g le-H M CO 2 O M EH o p^ M ^" JI-J O Q jl IS a a. 2- pj M tn CC ! s " rf o 0) |H* s . S" Q PH * p; M en 1 loa s n- a, 'd o p a 8 3 I 5 CX v x3 e l *! l s- o ^ S PI 20 The first is the best method, as two men, one at the beginning and the other at the other end of the block, have to agree before a train is allowed to enter. Besides, keeping the signal normally at danger is an additional safeguard, as it requires the operator to be certain of what he is doing and to put himself on record that the block is clear. In regard to the use of permissive blocking with a telegraphic block system, the method most generally adopted is to put the entire control in the hands Of the train dispatcher and allow the operator to give a "caution" or permissive signal only when au- thorized by him. If the conditions as to weather and track are favorable, permissive blocking is frequently made use of for freight trains. But, between passenger trains, the absolute block is maintained unless exceptionally good reasons present them- selves for doing otherwise. The permissive signal can be given in several ways, as has been said before; but the best plan, I be- lieve, is to require the operator to give a permissive card to the engineer, the same as with a train order stating for what train caution is to be observed. With this card there can be no mis- taking the information given, as might occur with a caution sig- nal, more particularly the three-position signal when such is made by the inclined position of the blade. That a system of signals operated through the means of com- munication afforded by the telegraph instrument is cheap and in every way advantageous, is clearly proved from the fact of its having been so widely adopted by roads that apparently could not afford to spend money on anything not absolutely necessary. But one wreck will very of ten. pay for a good many signals and the few extra operators required to work a block system; so that by drawing on one's imagination as to the size of the \vreck, it is very easy to figure out a great saving to any road. A system of block signals is certainly a much better arrangement than is any practice of flagging trains, but the trainmen must be properly educated as to the extent of the protection afforded by the system and not look for it to do more than it is intended to do. Not that flagging has as yet been abandoned where any sys- tem of signaling is in operation, but it is used merely as a check on operator and enginemen in case they should make a mistake. The benefits to be derived from any system of block signals 21 depend in a great measure on the rules governing their use and also on the extent to which they are observed. A set of rules for operating a telegraphic block signal system gives, first of all, the definition of a block, defines a signal and then states how the different signals are to be read. As these have already been given as have, also, several important points which are usually cov- ered in the rules I will give in a somewhat condensed form only those which have not been previously mentioned and which are essential to the proper working of a telegraphic system of signals : Trains between A and X will be governed in their movements by a block system which is designed to protect trains running in an opposite direction as well as in the same direction. This system will be independent of the general rules governing train movement and the movements directed by special telegraphic orders, and must not be confused with them. The block signal must never be fastened at the "clear" posi- tion, except when the office is closed, but must always be held at that position, when it is desired to clear a train, until the rear car of the train has passed. When there are no train orders, and the block ahead is clear for an approaching train, the signal should be changed to "clear" as soon as and not before the engineer is in sight of it, that the train may enter without reducing speed. At stations where the block signal is used, a red flag by day and a red lantern by night will be attached to the block signal mast to notify trainmen to call for orders, the block signal in addition being kept at danger until the orders are delivered. CHICAGO, MILWAUKEE & ST. PAUL RAILWAY COMPANY. River Division. . . '. 189.. C & E USE PERMISSIVE BLOCK. From to Train , , entered at. . . . M. Operator. 22 Engineers and conductors receiving permissive block card will run with great caution. Where view is obscured they must reduce speed to insure against collision with a train that may be running ahead of them. The responsibility for colliding with trains when permissive signal is given will rest with train receiving and moving under such 'signal. This will in no way relieve conductor and engineer of train stopping within block from flagging. If no markers are displayed on the rear of the train, the operator at the next block station ahead must be notified to give the approaching train a signal that train is broken apart. The block station in the rear must be also notified that the track is blocked until information is received from the conductor that he has all the cars in his train. When a train is on a siding clear of the main line and the markers have been seen, the block may be cleared. In case of failure of the wires, or if, for any reason, the operator cannot get orders for a train, he must give it written notice of the reason the proper signal is not given. A train intending to use a cross-over between block stations must notify the signalman at nearest block station. Train shall not use cross-over until a flagman has been sent out. Signalmen should closely watch each train as it passes, and if anything is noticed that is wrong, must report it to the next sta- tion and have train stopped. The rules governing the use of block signals do not relieve trainmen from observing all other rules relating to the protection of trains. The rules for use where communication is had by telegraph, generally contain a code of signals by which information regard- ing the condition of the block can be quickly transmitted. Where a bell is used the code has to be much more complete, and the number of taps required is often large, as one cannot talk with it as is possible with a telegraph instrument. For this reason a great deal of care should be exercised in the arrangement of the code, so that no serious consequences could result from a mistake in counting the number of taps. The best arrangement of taps is that where a combination of numbers, as, for instance, 2-3-2, is used rather than a consecutive number. 23 While, undoubtedly, there are dangerous situations which may arise with the use of a telegraphic block system from lack of a more complete equipment, it is certainly a great help to the train dispatchers and a protection against collisions. That engineers have run by signals when they were a't danger, that operators have allowed trains to enter blocks when they were not clear, accidents being caused thereby, has only resulted in a closer adherence to the rules on the part of the men and stricter discipline on that of the officers. But the fact still remains, that where a human agent is used he is liable at times to fail, and the greater the pre- cautions taken by mechanical means, and by using two men in place of one, the less likely is it that mistakes will occur and accidents happen. Work in this direction has resulted in the development of a system in which the labor of two men working in conjunction with each other is required to clear a signal and admit a train to the block. This method is called the "Controlled Manual," of which the Sykes system was the first brought into extensive use, and Patenall's improvement of the Sykes, a later development. Other systems have been invented, but, as yet, have only had a limited introduction. The two 'systems mentioned have only been ap- plied to roads having double tracks, but another system has been invented, which, although not yet in extended service, accom- plishes practically the same results on single track. The equipment of a station consists principally of a machine having separate levers for each signal, those for the distant signal, if such are used, being interlocked with the corresponding home signal. Distant signals are shown in the cut, as they are to be found at nearly all stations where this method of operation is used, but it is not to be understood that they are a necessary part of any system. To each of the home signal levers a locking bar and latch are so connected that when the signal is placed in the danger position the latch will fall into a notch cut in the bar and hold the lever in this position. To work the latch, electro-magnets are arranged, with the several parts that comprise the instru- ment, in a suitable box placed on the machine in a position most convenient to the leverman. There are two indicators in the side of the box, one of which, working in connection with the 24 latch, shows whether the instrument is locked or free. As there are two of these cases, one for each track, and as the equipment for each track is separate and exactly alike, that for one track only will be spoken of. Wires are run from the machine at one station to the machine at the next station, and so connected with the electro-magnets working the latch that when the circuit is completed by pressing on a contact piece on one machine the latch of the other is lifted and the lever unlocked. This contact piece, which is lettered P in the cut, is known as a "plunger," and the operation of closing the circuit, thereby unlocking the lever at the next station, is called "plunging." The plunger is constructed mechanically, so that if the operator has once plunged he will not be able to do so again until the signal has been cleared and returned to the danger position. The object of this is to prevent him from letting a second train in the block before the one admitted by him when he plunged has passed his station. To prevent him from clearing his signal and putting it back to danger again, and thus release the plunger, as might easily happen through mistake, a certain portion of the track is made part of an electrical circuit, and arranged so that the circuit between the two instruments made by "plunging" will be broken until a train has passed over the track circuit. If, now, this track circuit be placed at a certain dis- tance beyond the home signal, it is seen at once that the block will be clear before the operator can plunge and again unlock ihe signal at the next station. Working in connection with the plunger and placed just above it, is an indicator which shows the words "clear" or "blocked," for the purpose of indicating whether the operator has or has not plunged. If the operator has plunged, then the instrument at the next station is unlocked and a train can be admitted to the block, so that from his standpoint the track is blocked. If he has not plunged, then no train can be in the block, and, consequently, it must be clear. It must not be for- gotten that the indicator is changed from "blocked" to "clear" when the home signal is returned to the danger position, but that unless the train has actually passed out of the block and over the track circuit the operator cannot unlock the signal circuit at the next station by plunging. A separate wire is strung between 25 -each two stations, and used in connection with an electric bell for transmitting the information necessary for the proper working of the signals. Telegraph instruments can be used, if preferred, but as a bell does not require such close attention and can be understood by anyone, it is the one most generally adopted. In the cut, the levers and signals for a train moving in one direction only are shown. Three block sections with the stations A, B and C are represented, the signals being shown in the proper position for governing the trains which are supposed to be ap- proaching. To make it easier to follow the indications as given by each machine, three positions of the train are shown. The method of operating the signals is as follows : Supposing a train to be in block I, approaching block signal station A, the lever and signals being in the position shown in the cut. A asks B, by ringing the bell, to unlock his lever if block 2 is clear. B, looking at his indicator, sees the word "clear" and plunges, thereby un- locking A's instrument and changing the indicator on his (B's) machine from "clear" to "blocked." The instruments are shown in Fig. i after this action is supposed to have taken place, A's lever being unlocked but with the signals still at danger. A pulls his lever as soon as it is unlocked and lowers the signal admitting the train to block 2. The movement of the lever changes the indicator to again show locked, although the lever is not actually locked until the signal has been returned to the danger position. As soon as the train has passed the home signal, A returns the lever to the danger position and the latch drops into the notch in locking bar, and A is mechanically prevented from again clearing the signal. As the train approaches, B asks C to unlock him, which C does, provided the block is clear. On B's indicator changing from locked to free, he lowers the signal admitting the train to block 3; the indicator changes back to locked, and we have the con- dition of things shown in Fig. 2. When the train has passed, B returns the home signal to danger, unlocking the plunger and making the indicator show "clear" again. He could now plunge and unlock A if the circuit was restored; but this circuit is only restored after the train has passed the track circuit, which, as it is 300 feet beyond the home signal, insures that the block will be clear before another train can be admitted. 26 (D _ 50 CO o 00 27 When the train approaches, C asks D to release him, and we have the condition of things shown in Fig. 3, the operations as above described being again repeated. In Fig. 3 the action of a train on the track circuit is shown. The wheels short-circuit the current, so that the magnet of the relay is demagnetized; the armature dropping makes an electrical contact, which restores the unlocking circuit in the instrument. From this description it is seen that only through the efforts of two men, one at the beginning and the other at the end of the block, can a clear signal be given, and that not until the last train admitted to the block has passed the track circuit can this result be secured. The rules for operating any of the different "Controlled Manual" systems are practically those for a telegraph system, with only such rules added as are made necessary from the con- struction of the machine. One, that is perhaps the most important in this respect, is "In case of failure to get unlocked after oper- ator has plunged, a clearance card must be given for the train to proceed to the next station." Or, the operator, if the rules allow permissive signaling, may signal the train to proceed, using a green flag by day and a green light by night. The bell code of signals is made very large, in order to cover any conditions that may arise. The blocking of trains by this method of operation has proved it to be much better and safer than one where no check is put upon the operator; so much so, that many prominent men believe it to be the best system of those in use to-day. But, with all this, the fact remains that the personal factor is a necessary part in the operation of the plant, and just so long as this is the case will mistakes occur. While men may have the best intentions and strive faithfully to perform their duties, error is an essential part of human action, and, sooner or later, the time will come when some mistake will be made. Any system that will do away with the per- sonal factor and, at the same time, give as reliable indications, must certainly be in the line of progress toward that perfection and absolute security which all systems strive to attain. The ob- jection to an automatic signal that because no man is on watch the signal may be disregarded, is not valid, for all that any of the systems that have been mentioned are designed to do is simply 28 to indicate the condition of the block controlled by the signal. In any case, faith must be placed in the engineer that he will obey such signals. "The warrant for that faith," to quote the words of Mr. Sullivan, general superintendent of the Illinois Central R. R., "is the fact that no engineman of sound mind will knowingly run into danger. A man will give all he hath for his life the pledge of the engineman is the highest that can be given." Many automatic signal systems have been invented, some of which are in use and giving good satisfaction; others, again, have been tried, and, it is hoped, forever relegated to the scrap pile. The automatic mechanical systems have, as yet, had only a limited introduction. The staff system, while extensively used abroad, has been put in service on but one road in this country, and that for the operation of only one block. The most successful systems, and those generally alluded to when an automatic system is spoken of, are those which depend upon electricity for the controlling agent, whether or not it is the force actually used to work the signal. Of these, two systems may be taken as representing more clearly than any others the different lines on which automatic signals have been developed. One is the Westing-house electro-pneumatic, where compressed air is made to work the signal, its action being controlled by electricity ; the other is the Hall, where electricity is the only force used. The equipment of the electro-pneumatic system more nearly resembles that of a telegraphic or a controlled manual system than anjr of the others/as it gives the indications by means of the ordinary sema- phore blade. To the engineer the two systems are alike, except he knows that the electro-pneumatic is automatic, and therefore shows exactly the state of the block and not what the operator represents it to be. A current of electricity run through the rails and energizing a magnet, which is short-circuited by a train, or even a pair of wheels, in the block, is the agent depended upon for the proper working of the system. This magnet, by means of a second and more powerful current of electricity and the action of compressed air, changes the signal to indicate the condition of the block. If the current passes from one rail to the other without going through the magnet, as it will do when the rails are connected by a pair of wheels, the signal is made to indicate danger; if the 2 9 current of electricity goes through the magnet, the track must be clear and the signal is held in the "all clear" position. It is seen that if such a system is made reliable and does not get out of order easily, it is an ideal one, having the double ad- vantage of being automatic and using the position signal. The testimony of the superintendent of a road where sixteen miles of track have been equipped and successfully operated for some length of time, shows that the apparatus is reliable and practicable. He says that "there is a failure to give a correct indication only once in 250,000 times, and that the error then is always on the side of safety;" that "without this system the traffic could not be handled on the same number of tracks, owing to the time it would take operators to go through the necessary motions with either a telegraphic or controlled manual system ;" that "no dispatchers are needed, as the trains follow each other irrespective of orders, being governed entirely by the signals." The Hall automatic electric signal has come into more general use than the one just spoken of, and is no doubt familiar to most railroad men. This signal differs considerably from those of the semaphore type, in that its indications are given by color and not by position. The mechanism of the signal is placed in a large box having a circular glass center. Behind this glass a red disc is shown for danger; raising the disc out of sight and showing a white background is the means employed to indicate "all clear." The signal is operated by a current of electricity controlled by a track circuit As long as the current is flowing through the mag- net, the disc is held up and all is clear; when the circuit is broken, by a train entering the block, the disc falls and indicates danger. It is thus seen that there is practically no difference in the way the two systems are operated, so far as results are concerned, if the difference in the way the indications are given is ignored. Of course, every man is entitled to his own opinion, and while many think there are great objections to giving up the semaphore type and relying entirely on a color signal, others think that the ad- vantages of a position signal are offset by the increased cost of installation and maintenance, that of the electro-pneumatic being greater than that of the Hall. Certain it is that both of these sys- tems are coming more into use every day, as the prejudice against 30 an automatic signal is gradually being done away with. They are also making a record, both in the matter of expense and the ex- peditious handling of trains, that other systems cannot ap- proach. The method pursued in the 'operation of an automatic electric signal is very simple; the indication of the signal being positive, a train finding one at danger, stops; when the signal clears, it proceeds. As the blocks are generally short and trains can be run as close together as it is safe to run them, the blocking of trains is absolute and no permissive blocking is provided for. Ap- parently, the only rule necessary is "to obey the signal," but in practice it is found that they, like everything else about a railroad, occasionally get out of order and give a false indication ; that is, they indicate danger when the block is clear. If some provision was not made in the rules to cover such cases, it would result in tying up the road until the signal was repaired. Practice on the different roads varies, in the rule adopted for the guidance of trainmen when a signal is found at danger. A reason for this difference is found in the character of the country through which the road runs, the grades and curves, as well as the general diffi- culties of operation. If the road is easily operated, the grades light and the country open, the general practice is for a train to stop, when the signal is found at danger, from two to three minutes, and then to proceed as under a caution signal. If the next signal is found at "clear," the train proceeds under the clear signal, report- ing the block that was out of order. The practice where a road passes through numerous tunnels, over high trestles and heavy grades, is for the train to stop at the signal for five minutes, at the same time sending on a flagman who precedes the train all the way to the next signal. The writer recently rode on an engine over a division that was equipped with an automatic electrical signal, and had a good opportunity for watching its performance. There were twenty- four tunnels in the 100 miles of road, and it was certainly a great satisfaction to pass a signal showing "all clear" before the train entered any one of them. The trainmen on that division say they "do not see how they ever got along without the signals, and if the company were to do away with them, a good many of 'the boys' would want to 'quit the business/ " From my experience with the automatic electrical signal, I think it is the system of all others for a road to adopt, for it will show at all times if there is a train or a part of a train in the block, it has no operators to make mistakes, it will always indicate dan- ger when anything goes wrong with the apparatus, and it will fulfill all the requirements that a signal alone can be expected to fulfill. A railroad superintendent, speaking of the benefits to be de- rived from an automatic signal, thus aptly describes the difference between a telegraphic and an automatic electric system. He says that "the telegraphic block sometimes goes to sleep, sometimes gets drunk, sometimes becomes insane, and almost always lies when in trouble. The automatic block is a mechanism that has neither the ability to go to sleep, get drunk, become insane, nor to lie. It speaks for itself." The automatic mechanical systems go a step further than do either of the electric systems mentioned, in that they attempt to stop the train by opening a valve and setting the air brakes, if the engineer disregards the signal and runs by it. It is, perhaps, needless to say that nothing better could be desired in the way of a signal system than one that will give correct indications and stop a train if the signal is disregarded, provided the apparatus is made practicable and durable. But the facts are, that to make the apparatus work successfully the blocks have to be very short, too short for a road running fifty-car trains, and that where trains are run at high speed the life of the apparatus is limited. The mechanical systems, however, have been very successfully ap- plied to elevated roads, where the speed is low and the blocks as well as the trains short. The staff system, of which, I suppose, nearly everyone has heard, is quite a new thing in this country, and, so far, is giving a very satisfactory performance. Two machines are provided, one being placed in the station at each end of the block. There are twenty-one staffs in the two machines, removing one of which from either machine locks both, so that no more staffs can be taken out. Put the staff back in either machine, again making a total of twenty-one staffs, and both of the machines will be un- locked and a staff can be removed from either. Permissive block- ing is accomplished by providing six tablets or tickets, which tab- lets are unlocked and removed from the machine by using one of the staffs, called a "pefrfiissive staff," as a key. Each one of these tablets can be given to an engineer in place of the regular staff, so that it is possible to let seven trains follow each other into the block. The staff system is, ordinarily, worked as an absolute block, the tablets, or permissive blocking, being used for freight trains, and then only when conditions are favorable. There is only one rule to be observed in the operation of this system, and that is that "no engineer must enter the block unless he has a tablet or staff with him." If the engineer is given a tablet, or the permissive staff, caution must be observed, as there may be other trains in the block that are running in the same direction. Before giving the details of construction of the various systems, let me call attention to the three methods of operating block signal systems, about which this article has been written, and which it is well to bear in mind : ist. The telegraphic method, where operators can at will clear the signal, a telegraph line being used as the means of com- munication between the two stations. 2d. The controlled manual, where the work of two men, one at the beginning and the other at the end of the block, is necessary to clear the signal, and where a train having been admitted to the block, the signal cannot again be cleared until the block is clear. 3d. The automatic, where the signals, either by position or by color, indicate the actual condition of the block. Where the sig- nals are entirely automatic, and in case of failure will assume the danger position. CHAPTER III. CONSTRUCTION THE TELEGRAPHIC SYSTEMS. The telegraphic systems derive the name from the method used in conveying information from one station to the next re- garding the position of trains and the state of the block. The equipment of a station consists of: First A wire and the necessary telegraph instruments, or if so desired, bells may be provided, if a suitable code is arranged chat will cover all the conditions likely to present themselves in the blocking of trains. Second Of a signal by which information can be conveyed to engineers and trainmen of the condition of the block, and whether or not they have the right to proceed. Third Of a lever, if such is used, and the necessary connec- tions for the proper working of the signal. The construction of a semaphore signal has already been ex- plained, but as a road may block trains by using any one of several different designs, it is necessary that a description be given of the ones most generally used. -.. \Vhile most of the fixed signals are used for the blocking of trains, they are often used solely as a train order signal (commonly called by, trainmen "order boards"), so that while there may be a difference in the information con- veyed, there is no difference in the construction of the signals used for the two purposes. The simplest form of a signal is a flag stuck in the edge of the platform, or placed in some more conspicuous position, where trainmen will be most likely to see it. A very good arrangement, as shown in Fig. i, is to bolt a simple bracket, in which the flag staff may be placed, to the out- side of the station building. Trainmen will soon learn its loca- tion and will then know just where to look for a signal in case one should be put out for them. A little hook on top of the casting 34 near the end serves as a catch to hold the lantern, when, at night, one is used. The "Swift Train Order Signal," as shown in Fig. 2, is some- thing of an advance from a simple flag, and is really a much better arrangement, as it is placed in a more conspicuous position, and, from its being made of sheet iron, cannot be blown about by every wind, but is always seen to the best advantage. Its construction is very simple, consisting of an oval-shaped piece of sheet iron, riveted at its center to a shaft. The turning of this shaft, and with it the sheet, through a quarter of a circle, by means of a bell crank and a lever placed at the operator's office, is the method used in giving the different indications. The night indications are made by a lamp placed on the top of the shaft, which is made to extend up through the framework supporting the signal. The lamp is the same as an ordinary switch lamp, having two of the 35 lenses red and the other two white. These are arranged so that the light, as seen from an approaching train, from either direction, will show red or white, and give the same indication as is given by the board. If the board is set parallel with the track, it will not be visible to an approaching train, and is, therefore, understood to mean that there are "no orders," or that the line is "clear." If the board SWIFT'S TRAIN ORDER SIGNAL,. is set at right angles with the track, it will be plainly visible, and the indication is made to "call for orders," or "danger, track- blocked." Hooks are provided in the operator's office to hold the lever in the position in which it is placed. These hooks are painted red and white, corresponding with the indications made by the signal, to always remind the operator of the position in which the signal has been placed. The signal shown in Fig. 3 is one used extensively on ib" 36 E. T., V. & Ga., and other roads, and although of a semaphore type, the indications being made by the position of the blade, it is of a radically different construction. A cast-iron framework SIGNAL USED ON E. T., VA. & GA. R. R. or bracket bolted to the outside of the station building, carries at its outer end two shafts, which are set at right angles with each other, and are provided with the necessary levers and cranks to turn them through a quarter of a circle. A blade bolted to the 37 horizontal shaft and painted red, serves, when in a horizontal posi- tion, to give the danger indication ; a counterbalance weight being provided on the opposite side of this shaft to make the blade assume the danger position, should any of the parts become dis- connected. The lower end of the vertical shaft is provided with a casting, to which are bolted two blades, in the manner shown in the cut. These blades are painted white and are spaced a sufficient dis- tance apart to allow the red blade to pass between them, when it is lowered from the horizontal position. This blade, if at right angles with the track, will, of course, be seen by an approaching train, and being in the vertical position would indicate "all clear." If parallel with the track, only the narrow edge of each would be presented, and no indication would be given, as they could not be seen. To indicate "danger," the arrangement of the cranks attached to the two shafts is such, that if the red blade is in the horizontal position, the white blades are parallel with the track, and the red blade only will be seen. The "all clear" signal is given by pull- ing the lever and turning the white blades through a quarter of a circle, bringing them at right angles to the track, and on both sides of the red blade, which is lowered from the horizontal posi- tion, thus leaving only the white blades visible, and consequently giving the "all clear" signal or "safety" indication. The lamp, supported by the vertical shaft, is the same as that used for the Swift signal, and gives the indications by color that correspond with the indications made by the blades. This signal, it will be seen, is one in which the indications are given by color, as well as position, and beyond the fact that it is impossible to designate with the arrangement, as shown, the particular direction in which it is desired to hold trains, the indi- cations as made are clear and unmistakable. The great objec- tion, however, to a construction of this kind one that is of great force in a northern climate is that in bad weather the blades are very apt to become clogged with snow, or else frozen to- gether, in which case there is a possibility of a wrong indication being given, and a certainty that the signal could not be worked. With all the signals that have so far been described, it is im- possible to designate by the signal the direction in which it is 38 desired to block trains, so that to prevent mistakes, every train which is at a station where the danger signal is displayed must assume that it is intended for it and be governed accordingly. Taking up those signals which are designed to give separate indi- cations for trains running in opposite directions, that of Gravit's Railway Signal, sometimes nicknamed "the bootjack," may be said to represent the first step in this direction. This signal is shown in Fig. 4, and is of very peculiar con- struction. It is in general use on the Lake Shore & Michigan GRAVIT'S RAILWAY SIGNAL. Southern road. The two blades, which are fixed at an angle of 90 deg. with each other, are so mounted on a shaft that they can be turned through a complete circle. A lever placed in the operator's office serves, by means of a chain, and up and down rod which also acts as a weight, and a rack and pinion, to turn the shaft and with it the blades to any of the four positions it is neces- sary for them to take. On the lever stand there are four notches 39 or positions for the lever, with lettered plates at each, indicating to the operator the position of the signal blades, as "All blocked," "Clear for west-bound trains." The lamp case seen in the cut below the signal blades, is fitted with the necessary colored lenses, a lamp being raised or lowered behind them, so that the light will show through the lenses, giving a color indication in each direc- tion that will correspond with the indications made by the blades. Openings in the case on the station side are also provided with colored glasses, so that the operator can at all times see that the correct indication is given by the lamp. With this signal it is im- possible to indicate either "danger" or "safety" in both directions and have the blades occupy the usual positions. This is certainly a great objection, as it becomes necessary to have the blades at different times occupy different positions when indicating the same thing; danger being indicated by a horizontal position of the blade in one case and an inclined position above the center in another; safety, likewise, being indicated by a vertical position and also an inclined position below the center. The Mozier three-position semaphore signal, shown in Fig. 5, is a signal that is somewhat of a departure from the ordinary semaphore signal, not only in its construction, but in the manner in which the several indications are given. It is in general use on the Erie road, and is reported as giving very great satisfaction. It is designed to give the three indications danger, caution and safety with a single blade, but instead of making the cautionary indication by the usual position below the center, it is made by raising the blade to an inclined position above the center. This position, as well as the general arrangement of levers, etc., is clearly shown in the cut. In the construction of the signal, two chains or wires are run from a lever placed in the operator's office to a pulley fastened on the pole, and from there to the signal casting, one being used to pull the blade to "safety" and the other to pull it to the "danger" position ; a very good arrangement, for as both motions are posi- tive, it can be depended upon that the signals will occupy posi- tions corresponding with the positions of the levers. A weight sliding in a vertical plane on two roller bearings is suspended by means of a chain from two pins on opposite sides of the center of the arm-plate casting, for the purpose of making the signal as- 40 sume the danger position if any of the parts should break or be- come disconnected a very necessary thing", as has already been pointed out. The arm-plate casting is made to hold two glasses, one red MOZIER THREE-POSITION SEMAPHORE SIGNAL, AS USED ON N. Y., L. E. & W. R. R. and the other green, that one being brought in front of the lamp which, by its color, will give an indication corresponding with the indication made by the blade. With the arrangement, as shown in the cut, the lamp is raised by a windlass and chain to the top of the mast, which is represented as being made of iron pipe, but any form of pole will answer just as well, if some means be provided by which the lamp can be put in place. The construction of the signal is a good one, although I believe that solid connections to the arm-plate casting give better results, and are certainly safer than any chain or wire as is used in the present case. The objections to giving a cautionary indication by any signal have already been noted, but as the operation of SIGNAL, USED ON CENT. R. R. OF GA. each road is a problem in itself, to be dealt with by men holding very different opinions on such matters, it is not to be wondered at that the practice should be very different on the different roads in such an important branch of railroad operation as signaling. A semaphore signal that is very extensively used in the South is shown in Fig. 6. It is of a somewhat cheaper form than the ordinary semaphore, and is operated by means of a cable passed 42 around a circular rim cast on the arm plate. The arm-plate cast- ing is made heavier than the blade, so as to carry the signal back to the danger position when the lever is released. It will be noticed that the red lens is carried in an arm projecting above the center and not in the counterbalance part of the arm-plate casting; a lens being used instead of an ordinary red glass so as to con- centrate the rays of light, as the lamp used is an ordinary hand lantern. The construction of this signal is very light, and for severe climates is not a good one, as snow or ice is very apt to make it stick in the "all clear" position. The signal shown in Fig. 7 is that of the three-position signal used on the Pennsylvania road. A very noticeable feature in the design of this signal is that the blade, when in a truly vertical posi- tion, projects from the side of the pole so as to be plainly visible and give a positive indication that a clear signal is intended. The arm-plate casting is made to hold two glasses, a red and a green, as the usual practice on this road is to allow the operator to give a cautionary signal (the inclined position of the blade) whenever it is desired to block trains permissive! y. The arrangement as used on the St. Paul road is shown in Fig. 8, and is one, I believe, that fulfills all the requirements of a good signal, and at the same time is simple in construction and of low cost. It will be noticed that the ends of the blades are pointed, a .practice that the St. Paul road has been the first to adopt; the reason for doing this being the desirability of making some dis- tinction between a block signal and a signal used at an inter- locking plant. Undoubtedly some such distinction must be con- sidered advisable, when it is remembered that with a signal used at an interlocked crossing the engineer must stop his train at the signal if at danger, the derail not allowing him to run by it. When, with a block signal, he is allowed to pass it, if it is necessary to do so, and have the train stop in front of the station. Certainly it is not consistent practice to allow him to run by a signal in one in- stance and require him to stop at the signal in another, if there is no way in which he can distinguish one from the other. What objection can there be to pointing a block signal blade, if in any way such pointing helps to denote the character of the indication given? BLOCK SIGNAL USED ON PENN. R. R. 44 BLOCK SIGNAL USED ON C., M. & ST. P. 45 In the equipment of a station, only those parts have been shown in the drawings which are necessary to work a signal placed :n front of or near the station. When a distant signal is used in connection w r ith the home block signal, it is necessary to use a lever stand of much heavier construction than any that have been shown, as it requires about all the strength the average man possesses to clear a signal placed a distance of 1,500 feet or over. The levers also must be so interlocked that the distant R0IO , ft= v^ *i> F H 7 F k |i *v\\ j II OF THE UNIVEBSITY signal cannot be cleared until after the home signal, and in returning the signals to the danger position, that of the distant signal must be moved first. A lever stand provided .with the necessary locking and which is of simple construction, is shown in Fig. 9, the parts by which the locking is accomplished being shown in plan view in Figs. 10 and n. The long bars, having a notch in one side, are called "locking bars," one being pro- vided for each lever. The cross piece having its end tapered to fit in the notches cut in the bar, is called a "locking dog," and is made of such a length as will allow it to fit in be- tween the two locking bars, provided one end be placed in 4 6 one of the notches, as shown in Fig. 10. It will be seen in this figure, where both levers are supposed to be in the normal position that is, with both signals at danger that from the position of the locking dog, the home signal lever is the only one that can be moved. If this lever be pulled over and the signal cleared, the bar will be drawn back, bringing the notch in that bar opposite the one in the other bar. This now makes it possible to pull over the distant signal lever, as the locking dog is forced into the notch in the other bar, as shown in Fig. n, locking the home signal lever in the reversed position, until the distant signal lever is once more returned to the normal position. In locating the signals at any station, several things have to be taken into consideration, which are regulated to a great extent by the system and the number of signals used. Where the two signals are placed on one pole, the location of the pole does not fix the exact spot at which trains must stop, as it frequently hap- pens that trains have to pass the signal when at danger, in making a stop in front of the station. An arrangement of this kind has but one thing to recommend it, and that is its cheapness, while there are several grave objections. As has already been stated, a train must often run by the signal when at danger, to make a stop or to do any switching at the station. At night, if they have so run by the signal, the engineer and trainmen cannot see if it has been cleared, and the conductor has to walk to the other side of the signal to make sure that it has been cleared for his train to enter the block. That the operator from his office can see the light showing through the red glass of but one signal, and would not be likely to detect it should the glass of the other signal get broken. The best practice in locating such a signal at any station, is to put the signal on the same side of the track as the station, and at either, end of the station building, where the operator can see the signal lamp from his office window. The object of such a loca- tion is to have the signal where it can be seen by a conductor without his having to look through or over a train to see it, and also where the operator can see that the light is burning properly, and that at least one glass is in its place. Of course, if the view is such that on account of a curve, buildings, or what is more likely, a water-tank, the signal so located would not be seen, it 47 would have to be placed on the opposite side of the track. Where this will not overcome the difficulty, the best plan is to put the signal on a bracket pole, as is shown in Fig. 12, bringing the blade very nearly over the center of the track. A better arrangement, if a road can go to the necessary ex- pense, is to have each of the two signals on a separate pole, placed A BRACKET POI,E. on the right-hand side of the track as viewed from an approaching train, and at a sufficient distance past the station to allow the train to make the stop in front of the station without having to pass the signal. The operator would then be able to see each signal light, and would be likely to notice it were either glass broken. This is a very important point, when it is borne in mind that a signal is changed from "danger" to one indicating "safety," whenever this happens; and that it happens very often is not to be wondered at when it is seen that a glass may be broken by the jarring occasioned by letting the blade return to a horizontal 48 position with too much force, from strains due to improper setting of the glass in the arm-plate casting, or by someone throwing a stone or shooting the glass out, as will often happen near large cities. In regard to the location of the distant signal, no stated dis- tance can be said to answer for all situations, as the grades and the speed at which trains are run have to be considered before deciding at what distance from the home signal such a signal should be placed. Ordinarily the distance is 1,200 feet, but as the distant signal is the governing one, as regards the speed of the train, it would seem that a greater distance than this is to be pre- ferred, as in so short a distance a heavy freight or fast passenger train cannot be brought to a stop, should they pass the distant signal at schedule speed. While it is possible to work a signal a distance of 3,000 feet from the home signal, it is not advisable to go beyond 2,500 feet, owing to the difficulty of keeping the wires properly adjusted and the labor required to clear the signal. The change in the length of such a wire, due to expansion and con- traction, is considerable; so that unless very carefully looked after and adjusted, the blade will not be brought to the proper position, and the engineer will be in doubt as to what the signal indicates. Of course, the objection to such long wires would, to a great extent, be done away with if a good compensator might be had to automatically take up the changes in the length of the wire, so that the proper working of the signal would not be affected. At the present time, only one wire compensator, that invented by Mitchell & Stevens, can be said to be a practical success, and were it not for its cost, it would certainly come into more general use. Its construction is very plainly shown in Fig. 13. The ex- pansion and contraction of the wires are taken up by a sliding frame, the wires being kept taut under constant tension, by means of a weight. The objection to the use of a weight is that if the wire that pulls the signal back to the danger position were to break, the weight would pull the signal to the "clear" position, and thus give a wrong indication. To overcome this objection, the two wires are attached to a loose lever, as shown in the cut, which would be pulled away from the cranks if either wire were to break, thus leaving the signal free to return to the danger position by the force of gravity. 49 There is one question in connection with a telegraphic system of signals, regarding which very different views are held by dif- ferent superintendents, but about which I think there should be no doubt as to which is the best practice. It is this: Shall the block signal be also used as an order signal, or shall there be two signals at a station for trainmen to observe one the block signal and the other the train order signal? The argument in favor of using the two signals is that an operator, on receiving a train DOWN PULL WIRE ' UPANO DOWN ROD TO 6IGNAL "~ PULL WIRE FiC.13. MITCHEU/S & STEVENS' COMPENSATOR NORMAL POSITION. order, has to put the train order signal at danger before the "O. K." is given by the train dispatcher, and is therefore certain to stop a train for which he has an order; that with only the block signal, which is kept at danger unless cleared for a train, there is nothing to depend upon for the delivery of the order, except the operator's memory, and that on hearing the whistle for the signal, he may forget that he has an order to deliver and give the train a clear signal, particularly if he had fallen asleep and had just waked up. The argument in favor of making the block signal answer for both purposes, and which I believe to be the stronger of the two, 50 is that the engineer is more likely to take the indication of the more conspicuous signal and forget to look at. the other one, than is the operator to forget that he has an order for a train. That this is true, and that an engineer will sometimes fail to look at the order signal while taking the indication of the block signal, the writer is convinced of, from the fact that two such instances have come within his knowledge. CHAPTER IV. CONSTRUCTION THE CONTROLLED MANUAL SYSTEMS. With these systems, as their name implies, the labor of two men, one controlling that of the other, is required to clear the signal for a train to enter the block. The work of "clearing" the signal is done by the man at the entrance of the block, the same as with a telegraphic system, but the controlling power, or the actual permission to so clear the signal, is given only by the man at the end of the block. The arguments used in favor of this method of operation are, that where two men are required to work in this way in connec- tion with each other, they are less likely to make mistakes, each acting as a check on the other ; that placing the control of the signal in the hands of the man at the end of the block is a much more reliable and certain plan of operation, as he will know whether a train admitted had actually passed out of the block; that by making it impossible, by means of the track circuit, for the signal to be again cleared until after the train admitted has passed out of the block, all chance of a mistake being made by either operator is eliminated, and the indications given by the signal can be relied upon as showing the actual condition of the block. The arrangement of the several parts of a Controlled Manual system is somewhat more complicated than are those of the tele- graphic systems, and while a description of the instruments suffi- cient to explain the operation of such a system has already been given in a previous article, the special parts peculiar to each system have yet to be described. Of the three systems using this method of operation, that known as the Sykes system will be the first considered, not only as it was the first to be invented, but because it has come into more general use than either of the other two. This system 52 may be said to consist of a machine having the necessary levers for operating the signals, of the Sykes. lock instrument, and of an interlocking relay, the latter being used in connection with the track circuit to prevent an operator from again releasing a signal until the train for which the signal had been released has passed out of the block. A machine having the two levers for operating the signals governing trains moving in opposite directions, and the corres- ponding lock instruments, is shown in Figs, i and 2. The prin- cipal parts of the instrument, the lock bolt, the lock bar, the lock rod and the plunger rod, are also shown in the figures, the levers being in the normal position with the signals at danger. To explain the operation of the instrument, the operator, when he plunges, releases, by means of an electric circuit, the lock rod of the instrument in the next station, allowing it to fall by its own weight and lift the lock bolt from the lock bar, re- leasing the lever and allowing the signal to be cleared. As the indicator is attached to the lock rod, the indication would be changed to show the word "free" as soon as the lock rod dropped, showing to the operator that the lever had been unlocked. Clear- ing the signal, forces the lock bar forward, which, by means of the inclined plane and roller, raises the lock rod to its normal position, where it is held until again released by the operator at the next station. Raising the lock rod, changes the indicator back to show the word "locked," although the lock bolt does not fall into the hole in the lock bar until the lever is returned to the normal position. The construction of the instrument whereby the lock rod is released and allowed to drop when the operator at the next sta- tion plunges, is shown in Fig. 3. The magnet M attracts the armature A, thereby raising it and slightly turning balance lever L on its center C ; this releases trip T, which is then free to turn on its center, allowing the lock rod to drop by its own weight, unlocking the signal lever, as has already been explained. Fig. 4 shows the construction of the instrument, whereby an electric circuit is completed when an operator plunges. Pressing in the plunger P raises the cross-bar X, which, by means of the contact springs 'D and E, completes two circuits one through contact D, Fig. 3, to the magnet of the next instrument, releasing 53 54 the lock rod, and the other at E t Fig. 4, a circuit through the interlocking relay. This contact at E causes the relay to break the circuit just completed through the magnet of the instrument at the next station, making it impossible to again complete the circuit at D by plunging until the armature of the relay has been restored to its original position by a train passing onto and off of the track circuit. The plunger when released is forced out of its original posi- tion by the action of the spring S, a dash pot Q being provided to retard this action, so that the electric contacts made at D and E will not be of too short a period of time. The arrangement by which an operator is mechanically pre- vented from again plunging until the signal has been cleared and returned to danger, is also shown in Fig. 4. A trip rod F is pro- vided, having attached to its side a pawl piece G and sliding block B, and carrying on its upper end a plate /, on which are painted the words "blocked" and "clear." Attached to the plunger rod (not shown in the cut) is a pin K, which, by means of the pawl piece G, supports the trip rod in the position shown in the cut. The plunger is provided with a side piece N on the side next the trip rod, so that when the plunger is pressed in, this side piece will strike the pawl piece G, pushing it off the pin K and allowing the trip rod to drop, the sliding block B striking on the top of the side piece N, which is then underneath it. When the trip rod falls, and with it the indicator, it brings the word "blocked" in front of the opening in the case, indicating to the operator that he has admitted a train to the block, and cannot plunge again until the train has passed the station and the instrument has been re- stored to its normal position. On releasing the plunger, the sliding block B drops still further on the projection of the side piece N, and mechanically prevents the plunger from again being pressed in. The instrument is restored to its normal position by pulling the signal to clear, which the operator will do as soon as the train which had been admitted to the block by his plunging, approaches the station. Clearing the signal, draws the plunger rod to the bottom of its stroke (Fig. 2), catching the pin K (Fig. 4) under the pawl piece G, so that when the signal is returned to the danger position after the passage of the train, the plunger rod is raised 55 with it and the plunger unlocked. When the trip rod is raised, the sliding block B is raised with it out of the way of the side piece N, being held in the same relative position by the pressure of the pawl piece G. The raising of the trip rod changes the indicator to again show "clear," indicating to the operator that it is pos- sible for him to plunge once more. A plan view of the instrument, the parts by which the locking of the plunger is accomplished being shown in section, is shown in Fig. 5. To explain how the locking circuit is broken and restored through the combined action of the plunger, the interlocking re- lay and the track circuit, reference must be had to Fig. 6, in which Sta.. B all the parts made use of in this operation are shown. The relay and track circuit shown are those belonging to station B at the end of the block, the releasing magnet being the one belonging to the instrument of station A at the entrance of the block. Sepa- rate views of the contact points D and E are shown to make it easier to follow the action that takes place when the plunger is pressed in and crossbar X raised. The construction of the bar X is such that the contact at D is made on the up stroke and that at E is made on the down stroke, the projecting point that engages with the spring E being ar- ranged to turn on the bar as a center and press against the spring 56 only on the down stroke of the bar. The relay points being shown in the normal position, it is seen that the lower magnet is energized by the current which passes through the two rails forming part of the track circuit, and the armature is held up. The circuit of the upper magnet being broken at E, its armature is down and makes a contact at 0. When the plunger is pressed in and the crossbar raised, a con- tact is made between the points at D, completing the electric cir- cuit through the main battery, the magnet M and the points O and P, energizing the releasing magnet M, and unlocking the signal lever at station A. On the crossbar X being forced down by the action of the spring S 1 (Fig. 4), contact is made, as already explained, at the point E, momentarily completing the circuit, energizing the upper magnet, and thereby raising the armature and breaking the releasing circuit at 0. On the circuit being broken at E, the armature falls, but now strikes on the hook on the upper end of the armature belonging to the lower magnet, and is held up instead of making a contact at 0, and it is impos- sible to complete the releasing circuit as long as it is in this position. On a train passing out of the block and on the track circuit, the current passes from one rail through the wheels of the train to the other rail, and the lower magnet of the relay is demag- netized and the armature drops, breaking the releasing circuit at P, and letting the armature of the upper relay fall, again making a contact at 0. When the train passes off the track circuit, the current again flows through the lower magnet, energizing it and raising the armature, making a contact at P and restoring the relay to its normal position. From this it is seen that the interlocking relay in connection with a track circuit, makes it impossible for an operator to plunge and again release the instrument at the next station, until the train for which he had previously plunged had passed on and en- tirely off the track circuit, and therefore out of the block. To prevent an operator from leaving his signal in the clear position, and not returning it to danger after the passage of a train, as he should do, an instrument has been designed called an ''electric slot," which automatically sets the signal at clanger and compels the operator to return the lever to the normal position 57 before he can again gain control of the signal. On his re- storing the lever to the normal position, it is, of course, locked until released by the operator at the next station, and thus the purpose for which the slot was designed is accomplished. The action of the electric slot is obtained through the use of an electro- magnet, energized by a current from a battery passing through the two rails of an insulated section of track. Tlie armature of this magnet acts as a latch, and makes the connection between the signal and the lever whenever the magnet is energized. When the armature falls, as it will do when the magnet is short-circuited by a pair of wheels on the insulated section of track, the connec- tion between the lever and signal is broken, and the latter goes to danger by the force of gravity. The Sykes system is in use on the New York Central & Hud- son River Railroad, between Poughkeepsie and Buffalo; on the New York, Lake Erie & Western, between Jersey City and Turner's, and on a portion of the New York, New Haven & Hart- ford Railroad. In the practical working of the Sykes instrument, several de- fects were discovered, which, in the endeavor to perfect the sys- tem, resulted in a set of requirements being drawn up by Mr. C. H. Platt, of the N. Y. C. & H. R. R. R., which a properly designed controlled manual system should fulfill. These requirements were met in the machine designed by Mr. Patenall, electrician of the Johnson Railroad Signal Co., and which is now known as Patenall's block signal instrument. The requirements of the new machine, and the objections to the Sykes instrument which it was designed to overcome, are : First That gravity must be overcome by the action of the electric current in unlocking the signal levers, instead of allowing it to assist, as it now does, in withdrawing the lock rod from the locked position. This means nothing more than that a failure of the apparatus should lock the signal lever at danger, and not un- lock it, as now happens with the Sykes lock, if the latch holding the lock rod in place should slip or be jarred loose, allowing the operator to clear the signal whether the block was occupied or not. Second That the magnets for unlocking the signal levers must not be in the electric circuit except at the moment of actual use, to prevent an accidental release by the crossing of wires, lightning or other causes. This requiring an intentional setting of the instrument before it can be plunged to and released by the next succeeding station. Third That no interlocking relay must be used, the contacts all to be made in the instrument, where it is impossible for them to be tampered with or changed. Fourth That the signal levers be perfectly free after having been once unlocked, so that the operator can change the indica- tion of the signal as often as he desires, and not to have the lever locked whenever the signal is returned to the danger position, as is the case with the Sykes. The construction of the instrument is shown in Fig. 7, in which P is the plunger and D and E the contact points by which the releasing circuit is completed, when the plunger is pressed in. M is the releasing magnet, which locks the sliding bar vS whenever the latch A is dropped into either of the two notches in the bar 5\ This sliding bar, when it is drawn out by the hand latch H to its full extent, engages, by means of the pin K, with the rocker C, lifting the lock rod R and unlocking the signal lever by freeing the locking bar or tappet T, as is shown in dotted lines in Fig. 10. This sliding bar S, by means of a second crank / (shown in dotted lines in Fig. 7), lowers, when pulled out, the vertical bar B f which, in turn, moves the lever L, bringing the inner end into contact with the plate N. This inner end is made in the form of a jaw and presses on both sides of the plate N, making a metallic connection between the plates fastened to each side. The plate N is made of insulating material having three metal strips, two on one side and one on the other,, to which are attached the wires from the two circuits used in operating the instrument. The wire from the releasing magnet is attached to the one strip, the line wire from the instrument at the next station being attached to the lower and the wire from the track circuit to the upper strip on the other side. There are two indicators, one working in connec- tion with the plunger Pand showing the words "train on" whenever the operator has plunged; the other working in connection with the slide 5" by means of the lug G, the rod / and the indicator plate O, and showing the w r ord "locked" or "free," according to the actual state of the signal lever. 59 The operation of the instrument may be explained as follows: Supposing a train to have entered block I and to be approaching station B. B asks C to release him, at the same time drawing out the hand latch H and pressing it down between the two lugs on the bracket F, to hold it in that position. This pulls out the slide vS' and raises the inner end of the lever L, so that a metallic con- nection is made between the strip on one side of the plate TV and the lower strip on the other, putting the releasing magnet in the circuit with the line wire and the instrument at the next station. 6o When C plunges, a contact is made between the points of his instrument at D E, as is shown in Fig. 9, sending a current of elec- tricity through the magnet M of B's instrument, raising the arma- ture and with it the latch A, unlocking the slide S, which B now 6i draws out, raising the lock rod R and unlocking the signal lever, as is shown in dotted lines in Fig. 10. When the plunger is pressed in, the pawl V is forced off from the lug Q and the slide X drops, carrying with it the indicator Y. On releasing the plunger, the slide drops still further, making it impossible to again plunge until the slide has been raised to its former position. The words "train on" have now been brought in front of the opening in the case, indicating to the operator that the track is blocked. When the slide 5* was pulled out to its extreme position, and the bar B dropped, the inner end of the lever L was raised, break- ing the connection with the line wire, and connecting the magnet M with the track circuit relay. The breaking of the circuit demag- netized the magnet M and the armature falls, dropping the latch A into the slot in the slide S, locking it in this position until it is again raised by the magnet M. When the slide 5 1 is pulled out, the lug G engages with the roller U, raising the rod and changing the indicator O to show "free," as the lever is now unlocked. When the train passes out of the block and over the track circuit, it demagnetizes the track circuit relay, the armature of which falls and completes a circuit through the magnet M, raising the latch and allowing the operator to shove in the slide 5 1 . This lowers the lock rod R, locking the signal lever at T, Fig. 10, breaks the connection at the plate N, raises the bar B and with it the slide X, unlocking the plunger and making it possible for the operator to plunge once again. The indicators are changed back to their normal position, showing to the operator that the machine is locked and that no train is in the block. An electric slot on the signal pole is provided, which restores the signal to danger on the passage of a train, to make it impossible for the operator to hold the signal in the "all clear" position, and admit a second train to the block. This system is in use on the N. Y. C. & H. R. R. R., between New York and Poughkeepsie; on the N. Y., N. H. & H. R. R., between New Haven and Providence, and on the Long Island Railroad. The equipment on the N. Y. C. & H. R. R. R., through the tunnels in New York city, is, perhaps, the most complete of its 62 kind in the world, and one that, Considering the business that is handled, can hardly be equaled in the amount of protection afforded by any other system. As has been previously stated, neither of these systems the Sykes nor the Patenall are applicable to single-track roads, from the fact that a train on passing over the releasing section would release the instruments at the station in front as well as in the rear, and thus make it possible to admit a train to the block and bring about a head-end collision. A controlled manual system that is applicable to single as well as double-track roads, and one that has been in service for about two years and has given good results, has just been patented by Messrs. Fry & Basford. As it is of low cost, thus bringing it within the reach of many roads not able to afford the more costly apparatus of either of the other two systems, a short description of the system would be advisable. The apparatus made use of may be briefly stated to be: An electro-magnet for controlling the signal lever; a polarized relay and pole-changing switch to energize the electro-magnet at the next station and unlock the signal lever; a track circuit, and a track instrument with the necessary relays to restore the releasing circuit when a train passes out of the block; an electric slot for putting the signal at danger immediately a train has passed by the signal. As the drawing showing the several relays and the different circuits is somewhat difficult to read, I will not give it here, but will say that the apparatus is inclosed in a neat box with the handles of the two switches that are worked by the operator pro- jecting from one side. But one wire is used in this system, the telegraph instruments being cut out whenever it is desired to unlock the signal. The line wire is also grounded whenever the signal lever is pulled over and the signal cleared, so that the operator is cut off from com- municating with the next station and is forced to keep the lever in the normal position as much of the time as possible. The lock for controlling the signal lever is shown in Fig. n, it being made to work with the ordinary levers used in operating the signals by the telegraphic method. L is the signal lever; ^ a sliding bar, with notches cut in it as shown and attached to the 6.3 lever L; A is the armature of the magnet M, the end being made to fit in the notches cut in the bar S, to prevent the lever from being pulled over unless the armature is raised ; M is the magnet for raising the armature and releasing the lever when the proper connections are made and the electric circuit completed between the two stations. The operation of this system is as follows : Supposing a train to be approaching station A, and that A has asked B to release his signal, the block being clear, B shifts his pole-changing switch and reverses the current sent out from his batter v to the main line. At the same time, A shifts his switch connecting the polarized relay with the main line, thus closing a local circuit through the releasing magnet M and unlocking the signal lever. Pulling the signal lever to "clear," breaks the main-line circuit, dropping the armature of a magnet in the instrument and making it impossible for the signal to be released a second time should the operator at B leave his pole-changing switch in the reversed position. On shifting the pole-changing switch back to its nor- mal position, the circuit is broken between the battery and the ground, and positively held open until the train which has entered the block at A clears the track circuit and, consequently, the block 6 4 at B. In this way the admission of a second train at A is prevented while the block is occupied. To prevent a train on passing out of the block from restoring the circuit to stations on either side, and make it possible to unlock the lever at the next station ahead, as well as the station in the rear, two circuits are made use of, by which, unless they are passed over and completed in a certain succession, the releasing circuit is not restored and, in conse- quence, it is impossible to clear the signal. One of these circuits is completed by the track instrument, the other by a track circuit which is placed on the station siding, as well as on the main line, so that a train pulling into the siding would restore the circuits the same as if it had passed out of the block on the main line. The two circuits completed by this means are so arranged that- the circuit made through the rails works a relay which closes the circuit completed by the track instrument, so that unless a train passes over the track instalment before it passes over the track circuit, no action is produced on the relay of the second circuit, and the releasing circuit to the next station is not restored. From this it is seen that the system affords the same protection to trains on a single track that is given by the other systems, which are applicable only to double-track roads. Where roads have been equipped with a controlled manual system, the matter of expense has not cut such a figure as with the more simple apparatus, and the equipment has usually been more complete and all that good practice would require. As a rule, the instruments are placed in station buildings, but where these are too far apart, special cabins or towers are erected for the purpose. If placed at a station, it is usual to put the signals for each track on the right-hand side, and at such a distance beyond the station as will allow the trains to stop at the platform without having to pass the signal. Where there is no station and a tower is used, it is usual to put the two blades on a pole placed im- mediately in front of it, as shown in Fig. 12, as the connections are shorter and the cost is less. If there are several, tracks to be governed and there is no room to put the pole next the track which the signal controls, it is necessary to make use of a bracket pole having separate masts for each track, or else a bridge may be erected spanning all the tracks, the signal being put im- mediately over the track which it controls. The use of distant signals is, of course, optional, but the general practice is to put them in wherever the view is obscured, or the conditions such that the home signal cannot be seen until the train is quite near to it. With a controlled manual system, it is possible to lock elec- trically all the switches and the levers of any interlocked crossing machine that may be in the block, so that when a clear signal is given it not only means that the block is clear, but that all the switches are set for the main line, and the derails and signals of any interlocked crossing are set for the train to proceed. This locking electrically is performed by fixing an electro-magnet on each switch and lever, so that it is impossible to open them unless the magnet is energized. This is done from the tower where the block signal machine is placed, and is so arranged that when the electric switch is set to release a switch in the block, the releasing circuit between the two stations is broken, and the operator at 66 one station cannot plunge and release the signal at the next station. The automatic torpedo signal an auxiliary signal that works in connection with the home signal has of late come very much into use with the controlled manual systems. The machine is bolted to the rails, and is so arranged that when the semaphore signal is put in the danger position, a torpedo is moved forward from the magazine and placed where it will be exploded by the first wheel that passes over it. When the signal is cleared the torpedo is withdrawn from the rail and a train passes by without exploding it. It is thus seen that such an instrument is of great value in such places as a tunnel, or where the signal lights are apt to be ob- scured by smoke, or are in any way difficult to see, as it furnishes an audible signal about which there can be no mistake, and one which compels the engineer's attention. If exploded, it furnishes incontrovertible evidence that the engineer did not obey the signal but ran by it. With the controlled manual systems, it is, of course, impos- sible to use permissive blocking, as such use does away entirely with the protection afforded by the automatic features of the system. For, if two trains are admitted to the block, the first one passing out would release the signal at the entrance of the block, and allow a careless operator to admit a third train before the second train had cleared the block. This is undoubtedly a great objection to such a system, and one that will stand very much in the way of its general adoption by many roads. Its field of usefulness is thus limited to sections of roads, or divisions, where, on account of the difficulties of operation, or the number of trains run, it is not safe to make use of permissive blocking. CHAPTER V. CONSTRUCTION THE AUTOMATIC ELECTRIC SYSTEMS. The automatic systems may be divided into two classes, the automatic electric and the automatic mechanical, the distinction being made in the power used to operate the several systems. It is the purpose of this article to consider the automatic elec- tric systems only; but before doing so, it would be well to say a few words about the electric battery and the relays that are used with these systems, for it is upon the arrangement and con- struction of these two things, more than anything else, that an electric system depends for its proper working. The battery used is the one commonly used for telegraphing, and is what is known as the gravity battery, being made by placing a piece of copper and a piece of zinc in a glass jar and surrounding them with a solu- tion of sulphate of copper. Chemical action takes place between the sulphate of copper and the zinc, resulting in the formation of an electric current flowing from the copper to the zinc, if the two are connected by a wire to complete the circuit. If the wire in w T hich the current of electricity is flowing is wound around a piece of iron, it will cause the iron to become magnetized in proportion to the amount of current and the num- ber of turns of the wire. If the iron is soft or annealed, it will lose its magnetism when the circuit is broken. A piece of steel or hard iron will retain a part of its magnetism and become a permanent magnet. It is this property of a piece of soft iron to become magnetized by a current of electricity and then to lose its magnetism when the circuit is broken, that is made use of in the construction of the relay to transform the energy of the electric current flowing through the wire into the motion of the relay armature. For, by arranging a piece of soft iron in front of the poles of the magnet, it will be attracted when the magnet is ener- gized and pulled away, either by a spring or gravity, when the 68 circuit is broken and the iron loses its magnetism, a certain motion being thus imparted to the piece of iron each time the circuit is completed or broken. The piece of iron is called an armature, and the motion imparted to it is the one made use of in doing any mechanical work by electricity, whether it be to make or break electric circuits, to move valves, to turn a pulley, or any of the thousand and one instances that we find in every-day use. Electricity is susceptible of measurement, the same as if it was a piece of solid matter having size and weight, and while science is yet unable to say that it is anything more than a force, its laws and properties are well understood. The electrical units of meas- urement are the "ampere," the "volt" and the "ohm" the ampere measuring the amount of current that is flowing in a given circuit, the volt being the tension or pressure existing between the two poles of the battery, and the ohm the resistance offered by the wire, or other parts of the circuit, to the passage of the current. The amount of electricity flowing through a given resistance is directly proportional to the pressure; the lower the resistance, the greater will be the amount of electricity flowing w T ith a given pressure ; the greater the resistance, the greater must the pressure be to send a given current of electricity through that resistance. If there are two channels through which a current of electricity can flow, the amount flowing through each will be inversely pro- portional to their resistance. A gravity cell, such as is used to work the different relays, gives a current of one-quarter of an ampere, at a pressure or electro-motive force of .9 or about one volt, and has an internal resistance of four ohms to the passage of a current of electricity. If the cells are arranged in series that is, with the copper of one attached to the zinc of the next the electro-motive force or voltage is increased one volt for each cell, and the amount of current remains at ^ampere; if the cells are in parallel, with all the coppers connected to one wire and all the zincs to the other wire, the electro-motive force remains the same, while the amperes are increased | ampere for each cell. Having these tacts and figures, we are now in a position to take up and consider the arrangement of a track circuit, the relay, and the action which takes place when a train runs on the track circuit; for, as the track -circuit is the controlling agent of the best of the automatic electric systems, it is necessary that a thor- 69 ough understanding be had of the part that it plays in the suc- cessful working of such a system. In equipping a road with an automatic electric system, the track is divided up into blocks of any desired length, a signal of the pattern adopted being placed at the entrance of each block. The rails of each block are made into a track circuit by insulating the rails of one from those of the next block, by means of a fiber end piece of the same size as the rail, and by using wooden splices in place of the iron ones, as shown in Fig. 3. It has been found impossible, in practice, to make the length of any one circuit longer than 3,000 feet, owing to the increase of the resistance of the circuit, causing in wet weather such a loss of electricity, by leakage across the track from one rail to the other, that enough electricity does not flow through the relay to magnetize it, producing the same result that a pair of wheels would if it were in the block. In consequence of this there may be several track circuits or sections in a block, each one being in- sulated and separate from the next, but so arranged, by means of relays, that the same results are obtained as with a single-track circuit the length of the block. Each rail in the section is joined with the one next it by a bond wire, as is shown in Fig. 2, for the purpose of making a continuous circuit from one end of the section to the other, the contacts made by the angle bars being imperfect on account of wear, rust or the loosening of bolts. There are two bond wires to each joint, being usually placed on opposite sides of the rail; so that if, from any cause, one should get broken, the circuit would be maintained through the other one. At the end of each section furthest away from the signal, a battery of two cells is located, the two poles of which are respectively joined by wire to the two lines of rail; the manner of attaching- a wire to a rail being shown in Fig. I, the connection between the two wires being soldered. 70 The cells, for protection, as well as to prevent their freezing in cold weather, are placed in a box or battery well, shown in Fig. 4, and buried in the ground. At the signal end of the section a relay is placed, being connected by wires to the two lines of rails, and in this way establishing a complete circuit from the battery to one rail, from the rail to the relay, from the relay to the other rail, and from that rail back to the battery. Owing to the fact that water makes the earth a fairly good conductor, it is necessary that the resistance of the circuit be kept as low as possible, so that even under adverse conditions there will be less resistance through the relay than across the track, and the electricity will follow the rails and energize the relay, making it work properly in all kinds of weather. To keep the amount of electricity that leaks across from one rail to the other as small as possible, the two cells are arranged in parallel, the amount of electricity being in- creased thereby, while the voltage or pressure remains the same as with one cell. This reducing of the pressure reduces the power of the current to flow through a certain resistance, so that where there would be a large leakage of electricity across from one rail to the other, if the cells were in series, there would be but very little 'if they were in parallel, the current flowing through the relay instead and energizing it. In practice it has been found that the resistance of the relay should very nearly equal that of the external resistance of a circuit; but as the resistance of a track circuit is only half an ohm, it is practically nothing, and so the relays are made of the same resistance as a cell, or four ohms. The relay used in connection with a track circuit, from the nature of the work, must be a very sensitive one and able to work with the least possible current. The magnets should be com- paratively large, the armature light and well proportioned, and the parts well fitted without lost motion, so that a maximum of pressure for a given current will be exerted on the platinum points and a good contact made. The working parts of the relay should be enclosed in a dust-proof box or case, so that no foreign sub- stance can get between the points and thus prevent a contact. Assuming that everything has been properly arranged, and that a current of electricity is flowing through the relay, ener- gizing the magnet and attracting the armature; if now a train, or even a pair of wheels, should enter the section, a short circuit would be formed from one rail to the other, cutting out the relay, as there is practically no resistance through the wheels and axles the current from the battery flowing to the rail, from that rail through the wheels to the other rail, and back to the,battery. When the current is cut out in this way, the relay loses its magnetism and the armature drops, separating the contact points on the end of the armature, and thereby breaking the current pass- ing through them. This other or second circuit flowing through the contact points is a more powerful one than that used for the track circuit, and is the one made use of to operate the signal, whether this is done by the application of electro-magnetism to the parts of the signal, or by controlling the action of compressed air, or the force of gravity, any one of which may be the power actually used to work the signal. The battery cells in the second, or signal circuit, are arranged in series, as the resistance of the relays and the wires connecting the signals is much greater than it is with a track circuit. The number of cells varies with the number of signals to be operated by the current and the length of the wire connecting 72 the signals, or, in other words, according to the resistance. If the resistance is low, few cells will be required; if high, it means that there is more work to be done, and a greater number will have to be used. The number used varies all the way from four to twelve cells for each circuit, eight being about the average number required for blocks one-half mile long. The cells are either placed in battery wells or in small houses built especially for the purpose. These are generally sunk in the ground to pre- vent the cells freezing in cold weather. The length of the blocks varies on almost every road; but as the expense of maintenance of an electric signal is small, the blocks are usually made much shorter than where operators have to be employed to work the signals. Then, again, as the length of the block regulates the distance apart which it is necessary to keep trains the shorter the blocks, the greater being the number of trains that can be run in a given time there is every advan- tage to be had in making them as short as the amount of money the road can afford to spend in this direction will allow. An excellent feature of the track-circuit system is the ease with which switches and side tracks can be connected with the signal, and, by setting the signal at danger, stop a train if the main line is not clear. A switch is protected by running either the track or the secondary circuits through a circuit breaker, worked from the points of the switch. If the switch is set for the main line, the circuit is not interrupted, and the signal shows "All Clear." If, however, the switch is open, the circuit is broken, and the signal will indicate danger. By wiring up the ends of the side track and making it form a part of the track circuit, any train standing on that portion of the track will keep the signal at danger the same as if it were on the main line; so that whenever a train enters a side track in the block, the signal remains at danger until the train has entered the siding and cleared the main line. A track circuit also affords protection against a broken rail, for if this should happen, the circuit through the rail would be broken, from a separation of the ends of the rail, and the signal would be set at danger. Everything that has so far been considered is common to all 73 the automatic track-circuit systems, and may be said to be the ground work upon which they depend for their proper working, so that it is from this point that they diverge to the several devices which constitute the different systems. Taking up the Westinghouse pneumatic system, it will be re- membered that this system makes use of the ordinary sema- phore to give the different indications of the state of the block, the work of moving the signal being performed by means of com- pressed air controlled by a valve, operated by an electro-magnet; the compressed air being supplied from a central pumping station, at a pressure of sixty pounds, and in as dry a state as possible, to prevent clogging the apparatus from the condensation of water. The general arrangement of the apparatus showing the con- nections to the signal blade are shown in Fig. 5, the photograph being made of a signal on a road where green is used for the all- clear signal. It will be noticed that the signal casting is made to hold two glasses, one red and the other green, that one being brought in front of the lamp which will give an indication cor- responding with the position of the blade. As is customary where this system is in use, there are two blades on the pole the upper a home signal, painted red, and governing the block immediately ahead ; the lower, a cautionary signal, painted green, and working in connection with the home signal of the next succeeding block. Each signal is operated by a piston working in a cylinder which is three inches in diameter, the piston rod being so connected to the balance lever as to move the signal to the safety position, when air is admitted from the supply pipe to the cylinder; the weight of the signal casting and that on the balance lever, if such is necessary, causing the signal to return to the danger position when the air is released. From this it is seen that although the normal position of the signal is at safety, not at danger, the fact of its being at safety is a sure indication that everything is all right; for if anything should happen to the air supply, or any of the parts become disconnected, or either of the circuits get broken, the signal would immediately assume the danger position, and remain there until the defect was remedied. The details of the cylinder, the valve and the electro-magnet are clearly shown in Fig 6, the parts being shown in a position corresponding with the danger position of the signal. M is the magnet, A the armature, and V a valve worked by the armature. P is a piston working in the cylin- der C, S the pipe by which the air is supplied; E, F passages for admitting air to the cylinder ; X a valve worked by the spring Z, to shut off the supply of air whenever the current passing through the magnet M is broken ; and Y the exhaust opening into the atmosphere by which the compressed air is allowed to escape. THE ELECTRO-PNEUMATIC SIGNAL. B is a circuit breaker, worked by a pin on the side of the piston rod, being so arranged as to break the circuit operating the pre- cautionary signal whenever the home signal assumes the danger position. The operation of the signal may be explained as follows : Sup- posing the signal to be in the danger position, and that the signal 75 circuit, or the one energizing the magnet, had just been closed by the track-circuit relay. The current would then flow through the magnet M, causing the armature A to be attracted. This presses the valve X from^its seat, at the same time seating the valve F, closing the exhaust and allowing air from the supply pipe ^ to flow along the passages E and F into the cylinder behind the piston P, forcing it through the cylinder the length of its stroke and clearing the signal. As the piston was pressed down, the spring of the circuit breaker would be released, the circuit closed, and, if the next succeeding block was clear, the cautionary signal would assume the "All Clear" or "Safety" position. A diagram representing a piece of track, divided into several block sections, with the necessary signals and the several electric circuits, is shown in Fig. 7. The upper signal, with the pointed end, on each pole, is the home signal ; the lower one, with the end notched, is the cautionary signal, showing the condition of the next succeeding block ahead. An engine is supposed to be in Block 2, the signals of the several blocks being shown in the posi- tion they would assume in consequence of this block being occu- pied. In Block i we find that, as there is no train or part of a train in this block, the current from the battery is flowing through the rails of the block and energizing the relay, the armature being held up and the signal circuit closed. This energizes the magnet of the home-signal cylinder, and, by attracting the armature, opens the valve and admits air to the cylinder, clearing the signal. As the lower, or cautionary signal, must indicate the condition of the next succeeding block the signal circuit operating the mag- net must be controlled by the track-circuit relay of that block, and, as the relay is short-circuited by the engine in that block, the signal circuit is broken and the signal is shown at danger. This is clearly shown by tracing out the two circuits of Block 2. The armature of the relay is down, breaking the signal circuit at the point E, the current from the main battery being cut off from the magnets operating the home signal of Block 2 and the cautionary signal at the entrance of Block I the signals, in consequence, indicating danger. It will be noticed that the circuit operating the cautionary signal is made to pass through the circuit breaker worked by the home signal placed on the same pole, so that if the home signal 7 6 -1 I / THE HALI, SIGNAL. FIG. 9. 77 should for any cause assume the danger position, the circuit of the cautionary signal would be broken and that signal would also go to danger. The obvious reason for this is that as the home signal is the controlling signal, when it indicates "Dan- ger, stop!" the indication given by the other signal must, of course, be the same, or there would be at the entrance of the block two signals, one indicating "Danger, stop!" and the other "Safety, go ahead!" Breaking the signal circuit at this point does not cause both of the signals operated by the main battery of Block 3 to assume the danger position, as the two magnets operated by this current are in parallel and not in series ; the cur- rent flowing from the battery through the contact points of the track-circuit relay, after which it is divided, one-half flowing through the magnet of the home signal, and the other half to the magnet of the cautionary signal. The circuit from each signal back to the battery is completed, either by a common return wire or by connections made with the ground, as is represented in the cut. The signals at the entrance of Block 3 are shown at safety, indicating that not only is Block 3 clear, but that Block 4 is clear also. If we suppose the train to have moved up into Block 3, both signals governing the entrance of that block would go to danger, the home signal at the entrance of Block 2 and the cautionary signal of Block i changing to safety, while the cautionary signal of Block 2 would remain at danger. From this description it is seen that a train is at all times protected by a stop signal immediately behind it, and a caution- ary signal which is the distance of an entire block behind that one, to give warning to any following trains that a train is in the next succeeding block. This system has been in successful operation for a number of years, and is at present in use, to a greater or less extent, on the Pennsylvania, the Central of New Jersey, the Boston & Maine, the Chicago & Northwestern and other roads. Criticisms that have been made against the system are that the cost of installation is large, and that there is a pos- sibility of the signal giving a wrong indication, from water freez- ing around the air valve in winter, due to a condensation of moisture upon the valve and in the passages when air was re- 78 leased from the cylinder. The manufacturers claim that this latter difficulty has been overcome, by providing better means of drying the air at the pumping station, by draining the reser- voirs regularly of water, and by putting the cylinder inside of an iron pipe, which pipe is made, to answer for the signal pole, so that the parts are protected and will not become clogged with snow or ice. Passing on now to the Hall electric signal, we find, as has been previously stated, that it differs very widely from the ordinary semaphore signal, in that the indications are made by different colors and not by the position of a signal blade. The form and general appearance of this signal is shown in Fig. 8, the signal being represented as indicating safety. The view shown is one taken on the Chicago & Northwestern, the signal being put in between the two tracks, wherever possible, as the trains of that road run on the left-hand track. As will be seen by refer- ence to the figure, the working parts are inclosed in a large box placed on the top of an iron pole, a large glass-covered opening being made in the box, behind which a colored disk is raised or lowered to give the different indications. A white glass ex- posed in the back of the case whenever the disk is raised, does away with the necessity of bringing a white disk in front of the opening for the safety indication, and serves also to illuminate the disk when the same is brought in front of the opening. The case is painted black for the purpose of making a strong contrast with the red of the disk or the white of the background, and in this way make the indication more distinct and visible at a much greater distance than it would be otherwise. To distinguish a cautionary signal from a home signal, the case of the .former is painted white, instead of black, so as to form a strong contrast with the disk, which is painted green in place of red, green being the color used to indicate caution, from its use as such for a night signal. Cautionary signals of this pattern are used in the same manner as with the semaphore signal, the two cases being ar- ranged as shown in Fig. 9, the home signal being the one placed on top, or above the cautionary signal. Two interior views of the signal case are shown in Fig. 10, the signal being represented in one case as at danger, and in the other at safetv. It will be seen that the construction of this 79 signal is exceedingly simple, consisting of a magnet supported on a suitable frame, which also acts as a bearing plate for the armature; of an armature made of special design, which, it is claimed, will lift a heavier weight with the same current than will anv other form of armature; the two aluminum wire frames FIG. ii. SIDE VIEW. fastened to the armature, over the larger of which colored silk is stretched to give the danger indication, a piece of colored glass being put in the other, which is brought in front of a lamp to give the same indication at night. The large disk being heavier than the small one, the signal moves to danger by the force of 8o gravity. When, however, the magnet is energized, the disks are withdrawn from before the openings by the rotary movement of the armature, and a safety indication is given. The indications at night are made by a lamp placed behind the opening provided in the upper part of the case, as is shown in Fig. n, the clear white light showing through if the signal is in the safety position, and red or green if it is at danger. The circuits that are used with the pneumatic system can be used to operate this signal also, as the signal is held in the safety position by the signal circuit, the same as with the pneumatic. For, when there is a train the block and the track relay is short- circuited, the armature will fall, breaking the signal circuit flow- ing through the signal magnet, and allowing the signal to go to danger by the force of gravity. The cautionary signal being connected with the track-circuit relay of the next succeeding block, in the same way as was shown with the pneumatic system, it will indicate to an approaching train whether that block is occupied or not. The Hall system is in use on the Illinois Central, the Boston & Albany, the Chicago & Northwestern and many other railroads, the total number of signals in use being in the neighborhood of 1,500. Very decided objections have been raised against this form of signal by many experienced railroad men, owing to its de- pending entirely on color, instead of position, for the different indications, they claiming that the impression made on train- men is not a very decided one, and that the signal is not visible at a sufficiently great distance, except under the best conditions of weather, to allow an engineer to stop before passing the signal, should it be at danger. However, these are questions about which almost everyone will have a different opinion, and it must be left for the management of each road to settle in the way they think best. Another objection is that wet snow will stick to the glass, and, by obscuring the s.'gnal, practically put it out of service just when a signal is most needed. Of all the automatic signals, the Hall is the surest in its workings and the easiest to keep in order, the times when it has been known to indicate "Clear" when such was not the fact being few and far between. This has occurred where the 8i case has been defective, allowing water to leak in and saturate the cloth of the large disk. The electro-magnet is, then, not strong enough, when energized, to raise the disk, unless the counterbalance weight is adjusted further out on the arm of the small disk. Should the cloth dry out and the counter- weight not be readjusted, the large disk will be overbalanced and the signal will be held at safety by gravity when the signal circuit is broken, thus giving a wrong indication. Certain it is that one can hardly imagine anything simpler and less likely to get out of order than is this signal, for there is nothing touching the disk to catch or get out of adjustment, the power being applied direct by means of magnetic lines of force; there is no friction in the parts other than from one small shaft, and, by being inclosed in a case, the moving parts are pro- tected from interference from any outside agency. CHAPTER VL CONSTRUCTION THE AUTOMATIC ELEC- TRIC SYSTEMS. Continued. An automatic electric signal, with which, no doubt, the most of us are familiar, is the banner or clock-work signal shown in Figs, i and 2, it being the signal brought out by the Union FIG. i. Switch & Signal Co. to do away with the objections urged against a disk signal, and before the invention of the pneumatic signal. 8 4 It will be seen that the signal is made up of two disks one being circular and painted red for the danger signal, the other oval in shape and painted white for the safety indication. A square plate placed behind the disk and painted black, to make a good background, brings out very strongly by contrast the difference in the form of the two disks. The framework sur- rounding the disks is painted white, and also helps to bring out, FIG. 2. in a way that is very effectual, both the. form and color of the disk- that is seen by an aproaching train. The two disks are fastened to a shaft turned by a weight and clock-work mechanism, the latter being placed in the box seen underneath the disk. The mechanism is alternately released and held by a magnet operated by the signal circuit, or the circuit controlled by the track relay, the disks being turned through a 85 quarter of a circle for each change in the condition of the mag- net, thus alternately presenting one disk and then the other in the direction from which a train would approach. If the block is clear, the circuit is closed and the oval disk is displayed, thus giving a safety indication; when the signal circuit is broken, the armature drops, releasing the mechanism which turns the shaft through a quarter of a circle, and the round disk or danger signal is displayed. An ordinary switch lamp, placed on top of the shaft, is used at night to give indications to correspond with those given by the disks. The clock-work mechanism is shown in Fig. 3, the weight used for turning the signal being attached to a chain wound around the lower shaft, the shafts carrying the disks being placed on the spindle shown at the top of the figure. It will be seen that the armature of the magnet holds, at all times, one or the other of the two flops, which, when they are released, raise a catch hold- ing one of the cross-arms of a shaft geared to the spindle, allowing this shaft to turn through a quarter of a circle, when it is caught 86 by the other catch, the flop being held up by the armature. The weight is wound up by a handle made to fit on the square end of the shaft, seen in the cut, a threaded nut with a projecting pin being fitted to the shaft, and so made as to separate the contact points of two springs through which the signal circuit is run, and set the signal at danger if the weight is permitted to very nearly run down. This is done so that the signal will be left standing at danger when run down, as otherwise the signal might stop at safety and cause an accident by giving an "All Clear" indi- cation when such was not the state of the block. The signal will give some some 600 indications for one winding, or, in other words, it will indicate correctly the condition of the block for 300 trains passing it. This signal is in use on a large number of roads, the Cincin- nati Southern, the C, M. & St. P., and the Providence & Wor- cester having quite a number in service. However, it will, in all probability, never come into very general use, from the fact that there is too great a chance of its getting out of order and giving a wrong safety indication; not that it will very often fail to work, but that when it does so fail, it is almost as liable to give a safety indication as one indicating danger. An automatic signal to be worth anything must be made so that it will very seldom get out of order, as trainmen will lose faith in it if they have to disregard its indications very often. But while everything will get out of order occasionally, par- ticularly if not properly looked after, it does not follow that when a signal does get out of order it should ever show safety. Although the mechanism of this signal seems very simple and well put together, it can very easily become disarranged or get out of adjustment, and show safety when it should not; for if any one of the following things should happen to the mechanism, the signal is just as likely to stop at safety as at danger. Sticking of the armature, from polarization, or the armature binding in the trunions. Jamming of the chain between the weight and the walls of the surrounding pipe. Breakage of the chain, sticking of any of the parts, due to a lack of oil, accumulation of dust, etc. There is one thing about the signal shown in the cut which 8? is quite an improvement over some other signals of the same pat- tern made by this company, and that is in providing a disk to give the safety indication and making the signal one of form as well as of color. Many of the signals have only the one disk, that for indicating danger, in which case it becomes a color signal only; for when the signal is at safety, the edge of the disk is pre- sented, and the Background with this pattern painted white would be the only thing visible. If the background against which the signal was seen happened to be one presenting little contrast with the white of the signal the sky, for instance it is almost impossible to distinguish it until quite close, and while a danger signal would be plainly visible, the safety indication would be very hard to make out. With the signal shown in the cut, the safety indication can be plainly seen, irrespective of the surrounding objects, as the disk, being oval-shaped, would show very plainly, against the black background surrounded by a white ring. An automatic signal that is very interesting, as showing a new development in the application of electricity to the operation of a semaphore signal, by means of a motor, is shown in Fig. 4, the apparatus being the invention of Mr. J. W. Lattig, superintendent of telegraph of the Lehigh Valley Railroad. The semaphore being recognized as the best form of day signal, and practically the standard signal of all the roads in this country, it is the hope of almost everyone that some cheap and efficient method of automatically operating a semaphore signal will be found, to place it within the reach of all; so that, looked at from this standpoint, this signal has much to commend it, and while the number that are in use is small, it is to be hoped that the results ob- tained will be such as to warrant a more extensive trial. As will be seen by reference to Fig. 5, the motor is mounted on the pole below the signal, the signal being cleared by winding up a wire rope attached to the balance lever, the signal blade being pushed by means of an up and down rod into the position indicating safety. The wire rope is made of phosphor bronze, and is wound upon a drum connected to the large gear wheel. This wheel is geared to the armature of the motor, to increase the leverage, and while it. takes a longer time to do the work, it increases very greatly the lifting power. 88 To lock the motor after the signal has once been pulled to safety, and thus economize in the use of electricity, an ingeniously arranged worm gear is made to shunt the electric current from the armature of the motor to a magnet, Fig. 6, placed at one end of the armature shaft, a small circular armature being provided on the end of the shaft, so that when the magnet is energized it will be attracted and held, thus locking the gear wheels and hold- ing the signal in the safety position. Normally, the position of this signal is at danger, the idea being to use the current only when a train is approaching a method of operating an automatic signal known as the "Normal Danger" plan, the several circuits of which will be spoken of later on. When a train approaches the signal, the track circuit relay closes the signal circuit, starting the motor and winding up the wire rope by means of the geared wheel and drum. When the signal has been brought to the safety position, the signal circuit is shunted to the magnet, shown in Fig. 5 above the large gear DETAIL OF LATTIG SIGNAL. wheel, energizing it and attracting the circular armature, lock- ing the armature and drum so that the rope cannot unwind, and holding the signal in the safety position. As soon as the train 90 passes the signal, the signal circuit flowing through the magnet is broken, and the signal allowed to return to the danger position by gravity, the gearing and the motor revolving backwards as the rope unwinds from the drum. The signal has been in service on the Central Railroad of New Jersey for over a year, and is reported as giving satisfactory re- sults. As the motor is well boxed, there seems to be little chance MOTOR FOR LATTIG SIGNAL. of anything going wrong with it and causing it to give a safety indication when the block is not clear. The chance of this hap- pening from snow or ice clogging the blade is very small, as the signal being normally at danger, it would be held in that position and not at safety, should sufficient ice collect upon it to prevent its working. The battery, however, required is large, being eight to twelve cells of the Edison Leland type, to each signal, a fact that will 9 1 most likely prevent the signal coming into any very general use. These cells are intended to be used on open circuit work, or inter- mittently, and would soon run down if the current was used con- tinuously. They give a much more powerful current than the ordinary gravity cell, and are about the best cells to be had for this purpose on the market. The Kinsman block signal system is one that it would be well to speak of, not so much for what has been done by the system, as from the several discussions that have taken place and the very generally expressed desire to have some definite information in regard to it. This is really not a signal system in any sense of the word, but one designed to stop a train by shutting off the steam and applying the air brakes without the Fift 7- LOCOMOTIVE EQUIPMENT OF THE KINSMAN BLOCK SYSTEM CO? help of the engineer, if there is a train in the next succeeding block, or the second block ahead of the engine. It is, of course, perfectly possible to work a signal in connection with the electric circuits used, but this is not put forward as being in any way a necessary part of the system. The circuits are the same as those already spoken of in con- nection with the Westinghouse pneumatic signals, but with this difference, that the signal circuit instead of opening a valve in the signal cylinder, it trips a handle on the engine, opening an air valve and admitting air to a cylinder which shuts off the throttle and applies the brakes on the train. A general view of the arrangement, as applied to an engine, is shown in Fig. 7, the special parts belonging to the system 92 being the cylinder for closing the throttle, the magnet for open- ing the air valve when energized by a current of electricity, and the wire brushes attached to the equalizer of the engine truck, to pick up the current from a pair of guard rails placed on each side of the track, they being connected to the signal circuit battery, when the track circuit relay is demagnetized. The circuits are arranged the same as for a home and distant signal, except that the current, in each case, is conducted to the guard rails instead of to the signals. As there were two signals, so will there be two sets of guard rails that are energized behind a train and which will stop a following train, should it pass over either of them. When an application of air has been made, the engineer does not lose control of the train, for the magnet is energized moment- arily only, as the engine passes over the guard rails, and the engineer can pull up the handle, which will again be held by the armature of the magnet. This will release his air, and, by shutting the throttle lever home, making- the latch in the air cylinder catch and make the connection between the lever and the throttle valve, he can again open the valve and admit steam to the cylin- ders. Of course, when the engineer does this and goes ahead, he must run with caution, expecting to find a train in the next block, or possibly a switch wrong, as causing the application of the air on his train. When doing so, he assumes all responsi- bility for the proper management of his train, and would be blamed for any accident that might result from his having gone ahead. The track circuit and relays used with this system are ar- ranged the same as with the signal systems, except that the signal circuit is completed instead of broken, when the track circuit relay is demagnetized, the armature making a back contact in- stead of a front one. As the normal condition of the armature is up, the normal condition of the signal circuit is broken; so that while the track circuit is continuous, the signal circuit is an open one, and is completed only when the armature is down and an engine is passing over the guard rails. This plan of having the normal condition of the signal circuit an open one is a very bad arrangement, and is very much like making the safet\ indication the normal position of a semaphore signal, for with an 93 open circuit the danger application is the one that is positive, and not the one indicating safety. In other words, if anything should happen to this circuit and it should fail to give an application, a safety indication is made, and not one that will indicate danger; the circuit fails to safety instead of to danger. To be safe, the sys- tem should be normally clear and fail to danger a fault which it is manifestly impossible to overcome, from the fact that the apparatus is placed on the engine, where a continuous current through the magnet cannot be used. It is to be regretted that in the most extensive application made of this system, the relays used in connection with the track circuit were of a sbmewhat experimental form and one not well adapted to the work to be performed, or the number of failures, as gathered from the report read before the American Society of Civil Engineers, would not have been so large. In -that report, it is stated that no failure of the system had occurred where the system had failed to indicate danger when it should have done so, and while this is probably true, there were several instances where it would have failed in this way had it so happened that one train followed another into the block. There are many things which may happen to the signal circuit and to the several parts of the apparatus, any one of which will cause the system to fail to make an application, and in this way give a safety indication when danger exists. These are : Sticking of the armature of the track circuit re.lay, from polarization, or from the armature binding on the trunnions; giving out of the spring of the application valve ; breaking of the wires at any part of the circuit, or loosening of the binding screws, breaking the connection; broken jar, allowing the liquid to escape; grounding of the current, the voltage being high; a piece of paper, dust or any non-conducting substance getting between the contact points of the relay; brushes failing to make a contact; waste catching in the teeth of the gear wheel, preventing the air valve from being turned; valve sticking, from lack of proper oiling. In describing the circuits spoken of in connection with the several automatic electric systems, it has been supposed that the signals have been applied only to roads having double tracks, the trains on a given track all running in the same direction. It has also been taken for granted that the blocks were short 94 enough to make it desirable to have a distant signal the length of an entire block away from the home signal. In many cases it is found that in looking over the ground preparatory to installing an automatic system, that this latter arrangement, owing to the length of the block, the consequent expense and the possible slow- ing up of trains, from their having to run the distance of almost two blocks apart, is not advisable; in which case it is customary to do away with a cautionary signal and make use of what is known as the overlapping section. To use an overlapping sec- tion is to practically extend the block past the signal into the block ahead, the two signals standing at danger so long as a train remains on the overlapping section that is, that a signal shall not return to the safety position until the train has not only entered the next block ahead, but shall have passed the signal the Fig. 8 Locomotive JEr.^ir.eering distance of the overlapping section, the section being of such a length as will give ample distance in which a following train could be brought to a stop, should it have approached the home signal at speed and found it at danger. In Fig. 8 an arrangement of overlap track circuits is shown, the overlap being of any length desired. vS is the signal placed at the entrance of each block, R the relay and B the battery for the main track circuit, T being the relay and C the battery for the overlap section ; A is the main battery operating the signal. It will be noticed that the relay for the overlap section is a double-point relay, the armature being provided with two contact points and making and breaking two separate circuits whenever the armature is attracted and released by the magnet. By tracing out the several circuits, it is seen that a train passing the signal and entering the overlap section puts the signal immediately behind it at danger, and keeps the signal at the entrance of the block it has just left at danger also. 95 for if the track circuit relay T is demagnetized, it will break the signal circuit to signal S-i at H and to signal S-2 at M, and set them both at danger. When the train passes from the overlap section to the main track circuit, the relay R is demagnetized, breaking the signal circuit at K, keeping signal S-2 at danger, while signal S-i goes to safety, in consequence of the circuit being restored at the point H, the armature being attracted by the magnet of the relay T, which became energized as soon as the train passed off the overlap section. Thus, there will always be two signals at danger behind a train as long as it is on the overlapping section, and only one signal Fig. 9 Zocomotiae Eiiyineenny at danger when it has passed off the circuit on to the main track circuit of the block. All the circuits that have been considered, so far, are those for use only on a double-track road, but when a single-track road is to be equipped other circuits must be used, as. the conditions are found to be more complicated and very different, for not only must a train protect itself by setting the signals in its rear, but it must set the signals ahead of it at danger, also, for trains running in the opposite direction. Such a system of circuits is shown in Fig. 9, being those used by the jftall Signal Co. in several in- stallations which they have made. This arrangement of circuits provides what is called by trSem "preliminary blocking sections," in which the second opposing signal in advance is set at danger by a train before it passes into the overlap section of the signal. Without this preliminary section it is possible for opposing 9 6 trains to enter the block at or about the same time, in which case each engineer would suppose that the signal had been set at danger by his train, and a collision would most probably result. With the use of the preliminary section this is impossible, as an engineer passing any signal would know that the track was clear for the block ahead and the preliminary section, also. If a train should enter the preliminary section at the same time that a train running in the opposite direction passed the signal of the block in advance, the signal at the entrance of the block in the face of the train on the preliminary section would be set at danger and that train would be stopped; the train in the block finding the next signal at danger would also come to a stop, and a col- lision would be prevented. By referring to the figure it will be seen that a train entering the preliminary section N, would put the signal D at danger by breaking the signal circuit at the contact G-i of the track circuit relay R-i. On the train passing signal A and entering the track circuit 0, the signal A would be put at danger by the relay M-i breaking the circuit at the contact points P-i and S-i, the signal D being kept at danger by a circuit breaker X worked by the signal A, so that while the contact is made by the relay at G-i, the signal does not return to safety. When the train enters the preliminary section Q, the relay R-2 is demagnetized, breaking the circuit of signal F at G-2 and that of signal A at H-2, keeping the signals A and D at danger, while the signal F, the next one in advance, is put at danger, also. To show more clearly the effect of the preliminary section, let it be supposed that a train should enter the preliminary section N just after a train running in the opposite direction has passed the signal D; when the train at D passed the signal, it put the signals A and C at danger by short-circuiting the relay, the signal D being set at danger by the circuit breaker worked by signal A. The train having passed signal D at safety, would, of course, proceed on its way, not knowing that a train was running against it in the opposite direction. Not so with the train on the pre- liminary section, for although it would put the signal B at danger as soon as it entered the section, it would find signal A imme- diately ahead of it standing at danger and would come to a stop, The train running from signal D to signal B, finding signal B at danger, would also come to a stop, the length of the overlap sec- 97 tion, or the distance between the signals A and B, being the dis- tance between the two trains. An arrangement of circuits that is being extensively adver- tised by one of the signal companies is one known as the "Normal Danger" plan, the signals being normally at danger until the ap- proach of a train, say, to within 2,000 feet of the signal, when the signal will indicate safety, provided the block is clear. Claims are made for this arrangement that it effects a great saving in battery power, the signal circuit being closed only when the signal is at safety, and that if the signal is at safety it is positive evidence to the engineer that the signal is in good order and that the track is clear, particularly if he has been able to see the signal change from danger to safety. While these arguments are, without doubt, very good ones, there are several objections to the ar- rangement that with many will considerably outweigh the advan- tages. These are, that very frequently with the long distances in which the signals would be visible to the engineer, he would be inclined to reduce the speed, not knowing if the signal would be cleared for him to proceed ; that if the signal should get out of order in any way, it would very likely not be found out until the approach of a train, thus causing delay, whereas if the signal was normally at clear, track men and others would be able to report the fact immediately the signal was found at danger with- out apparent cause; that track men and battery cleaners would know at all times, when maintaining the signals, that the circuits and batteries were left in good order, the fact of the signals being left at clear being evidence to that effect. Again, defects are more easily detected and repairs made, the signal going to safety and indicating that everything is all right, immediately the trouble has been remedied. An arrangement of circuits that is in use on a good many roads, but one that is fast being done away with, owing to the superior advantages of a track circuit, is what is known as a wire circuit system, the signals being set at danger by means of a track instrument placed at the entrance of the block, and being cleared by a similar instrument placed at a certain distance beyond the next succeeding signal. Such an instrument is shown in Fig. 10; the wheels of a passing train striking the end of the lever which projects slightly above the rail, raises the other end and forces 9 8 upward a piston inside the case, making or breaking the signal circuit, whichever it is intended to do, according to the position of the instrument relative to that of the signal. This arrangement of circuits is certainly not to be recom- mended in comparison with a track circuit, since no protection from a following train is afforded should a train break in two, or back up after once passing the clearing instrument. No pro- tection is afforded should a train on a side track not have pulled in far enough to clear the main line. And, again, such an ar- rangement of circuits would fail to give warning in case of a broken rail a failure that is liable to occur at almost any time. In the diagram showing the several circuits, the signal cir- cuits have been represented as extending the entire length of the block, but in actual practice this arrangement is not always used, owing to the increased resistance of the circuit, due to the length of the wires. By making the line wire into a separate circuit from one signal to the next, and using a relay to make and break the circuit of a local battery, the same results are secured and a saving in battery power effected. With an automatic electric system, any switch in the block is very easily protected, the signal at the entrance of the block being put at danger when the switch is opened by attaching a circuit breaker to the points of the switch, so that the current through one rail of the track circuit, which is run through the contact points in the box, is not only broken, but the battery end of the broken circuit is connected with the other rail and an effectual short circuit formed. Bringing the wires of the signal circuit down to the switch box, instead of breaking the wires of the track cir- 99 cuit, while a very effective arrangement, is in many cases a more expensive one, and one that is not necessary, the chance of the others failing to act being very slight. Visible indicators, worked by the signal circuit, should be placed at all switches on the main line, so that if an approaching train has arrived within a certain distance of the signal at the entrance of the block, the indicator will be set at danger, and warn anyone desiring to use the switch that a train is approach- ing. Automatic signals should be located with reference to the character of the road and the number of trains that are run, the idea being to so space the signals as to make the running time through each block about the same. The plan of putting the sig- nals very close together, if there are curves in the track, and very far apart where the signals can be seen, is a very poor arrange- ment, as it will often happen, owing to fog, snow or storms, that the signal cannot be seen until quite close to it, in which case trains following each other would be spaced the distance of the longest block apart, and not that of the shortest. In placing signals at the entrance of a block, it is a very good plan to put the signal a short distance beyond the beginning of the track circuit, so that the signal will go to danger before the engineer passes it, thus letting him see that it is working properly. Where this is done, any engineer stopping because the signal is at danger must be careful not to let his train run into the sec- tion controlling the signal, or the signal will not change to safety when the train keeping the signal at danger passes out of the block. CHAPTER VIL CONSTRUCTION THE AUTOMATIC ME- CHANICAL AND THE STAFF SYSTEMS. With the automatic mechanical systems an attempt is made by mechanical means to automatically put the signal at the en- trance of the block at danger, when a train passes it, and to change the signal back to safety when the train passes out of the block. The necessary impulse, or power, to work the signal is obtained from the wheels of the train, an inclined bar placed against the side of the rail being pressed down by the first wheel passing over it, and the small movement thus obtained, being in- creased by an arrangement of levers, is made to work the ap- paratus. One of these bars, or trips, as they are called, is placed at the entrance of the block to set the signal at danger, and a second trip, placed at the end of the block and connected with the first trip and the signal by a pipe line, is used to change the apparatus and the signal back to the normal position when a train passes over it. Two of these systems have come into a limited amount of use the Rowell-Potter, on the "Alley" Elevated in Chicago, and the Boston, Revere Beach & Lynn in Massachusetts; and the Black system, on the Metropolitan Elevated roads in New York City. The Rowell-Potter system was applied to the Intramural Railway at the World's Fair, but was not a strictly mechanical plant, the release, or restoring of the signal to a normal position, being effected by means of electricity. As applied to the "Alley L," the equipment is nothing more than a station protector, the track for 1,000 feet in front of each station being made into a block, the signal being placed 650 feet from the station. If the blocks were multiplied sufficiently, the entire road would be pro- tected, and trains could then follow each other no closer than the length of a block ; as it is now, several trains may be run between two stations, but no train can pass the signal and not be stopped, IO2 so long as a train is in the station block. The Rowell-Potter system makes a notable departure from the ordinary signal sys- tems, in that provision is made for stopping a train independently of the engineer, should he be negligent, or, from physical causes, be unable to obey the signal. The system is not only automatic, but obedience to the signal is made compulsory, and all chance of a mistake being made is eliminated. The principal parts of the system are: the inclined bars or trips, the safety stop and air valve for applying the air if a danger signal is disregarded, and a latching device .for holding up a weight and allowing the signal to be kept at danger as long as a train is in the block. ROWELS-POTTER SAFETY STOP AND SIGNAL SYSTEM. INCLINED BARS OR STRIPS. The inclined bar, or trip, is shown in Fig. 2, and also at B in Fig. i. One bar is about 6 feet long, the other 4 feet, the point where they are connected being raised i inch above the top of the rail. A train passing over the bars will gradually press them down, owing to their being inclined, and the motion so imparted is transmitted by a connecting rod and crank to a pipe line to work the signal or other parts of the apparatus. The safety stops for applying the air are shown at A and C, Fig. i, being placed at such a distance out from the rail that they will, if raised, strike the roller of the air valve on the engine, opening it and applying the air. Practically, the construction 103 is the same as that of a trip, except that the bars are lighter and are raised at the joint by a crank worked from the pipe line. The air valve is shown in Fig. 3, a cross section of the valve, when open, being shown in Fig. 4. Ordinarily the valve remains closed; but should an engineer attempt to run by a signal, or should a part of the apparatus get out of order, the safety stop bars (A or C, Fig. i) being raised, would strike the roller placed at the bottom of the vertical shaft, pressing it up, which, by means of the cam on the top of the shaft, presses in the valve rod, open- ing the valve and applying the air, the valve being held open by THE AIR VAI,VE. Fig.4 AIR VAI,VE OPEN. the latch dropping down in the slot cut in the rod. The wire fastened to the latch extends up through the running board, so that the engineer can, by pulling up the latch, allow the valve to close and release the brakes before the train has been brought to a full stop. A right and left screw on the shaft allows for ad- justments to be made to keep the roller the standard height above the top of the rail. IO4 The latching device is shown at L, Fig. i, and in detail in Figs. 5 and 6. The object of this device is to hold up the weight (seen at W, Fig. i) after it has been raised by a passing train, allowing the counterweights X and Y, Fig. i, to carry the signal LATCHING DKVICK. to danger and to drop the weight, thereby raising the two weights X and Y, clearing the signal, and restoring the apparatus to its normal position when a train passes over the releasing trip and out of the block. By reference to Fig. 5, it is seen that the lever L carrying the weight is attached at one end to the inclined "bars or trip, and at the other end to the bottom of the rack rod F, Avhich, if raised, will be held up by the pawl piece E, engaging with the teeth cut in the rod F if the rocker R is held by the latch H. If the rocker 7? is not held by the latch H, the rack will pull the rocker down, bringing the pawl E against the projection /, forcing the pawl piece out and allowing the rack rod F to drop. The latch H is worked by the two lugs M and N, on the releasing bar Q, Fig. 6, the releasing bar being operated by a motion plate connected by means of a crank and connecting rod to the pipe line operated by the releasing trip. If the bar is drawn out, the latch will hook the end of the rocker, allowing the pawl to engage with the rack and hold it up when it is raised; if the bar is shoved in, the latch will release the rocker, the weight drawing the rocker down, the pawl piece being drawn against the projection and forced out, releasing the rack and allowing it to drop. The general arrangement of the apparatus is shown in Fig. I, B being the operating trip placed in front of the signal, where a train will pass over it, and, by pressing it down, raise the releasing trip, and also hook the latch of the latching device over the rocker, setting the instrument to hold the weight up when it is raised. Next to the operating trip, and opposite the signal, is the primary safety stop, which, if the signal is in the clear posi- tion, will be depressed, and will allow 7 a train to proceed without stopping it. Next to this the signal trip and latching device is located, the secondary safety stop being placed beyond the pri- mary stop. The two safety stops are connected by a pipe line worked by counterweights, as shown, so that when one is depressed the other is raised, a crank worked by the lever of the signal trip being used to connect the pipe line with the weight and clear the signal when the weight is dropped. At the end of the blocked section a trip, called the releasing trip, is placed, to restore the signal to the clear position when a train passes out of the block. -It is connected to the operating trip by a pipe line, and so arranged that when the operating trip is down, the releasing trip will be raised. When the releasing trip is pressed down it will raise the operating trip, and thus leave the apparatus ready to be worked by the next train. io6 The operation of the system is as follows, supposing the ap- paratus to be in the normal position, with the operating trip raised, the primary safety stopped depressed, and the signal at safety, as is shown iri Fig. 7, and that a train is approaching: When the engine passes over it the operating trip will be depressed, raising the releasing trip and drawing out the releasing bar, latching the rocker of the latching device, so that the weight ROWELL-POTTER AUTOMATIC MECHANICAL SIGNAL AT "SAFETY." when raised will be held up. The signal being in the safety position and the primary stop depressed, the engine passes by without the air valve being opened ; when the engine reaches the signal trip the bars are depressed, raising the weight which now is held up by the latching device, the crank operating the pipe line to the signal being turned by the lever, changing the signal to danger, raising the bars of the primary stop (as is shown in Fig. 8), and depress- ing those of the secondary stop, so that the engine passes on into the block without having the air applied. When the engine has moved beyond the station a distance of 330 feet, it passes over the releasing trip and depresses it, the operating trip in consequence being raised and the releasing bar of the latching device pulled out. This allows the weight to fall and restore the signal to the safety position, the secondary safety stop being raised and the bars of the primary safety stop de- pressed, the apparatus thus being left in the normal position, to allow a second train to enter the block. By connecting the two safety stops with the signal trip in the ROWELS-POTTER AUTOMATIC MECHANICAL SIGNAL AT "DANGER." manner shown, it insures that the brakes would be applied and the train stopped should the signal not be thrown to danger after the passage of a train. It will also enable the engineer to tell whether he has been stopped by a train in the block, or by some defect of the apparatus, due to breakage or to lack of adjustment. Switches are easily protected with this system, by providing a signal and safety stop to be worked from the points of the switch, at such a distance from the switch as will admit of a train being stopped before reaching the danger point. If all the en- gines of a road are equipped with the air valve of the system, it is possible to provide the trainmen with a track instrument in io8 the shape of an inclined plane, which can be placed on the end of the ties to open the valve and compel a train to stop, should the engineer not observe the man signaling him. Claims are also made that the stop is an excellent device to be used in con- nection with any manual system, as it compels obedience on the part of the engineer to the indication of the signal. As installed on the "Alley L," the system has given good satis- faction. While it has been found necessary to strengthen some of the parts, the wear on the apparatus, as a whole, has been very small. While applications, due to a failure of the apparatus, occasionally happen, there have also been a few failures to apply the brakes, when they should have been applied. This has oc- curred through water collecting in the valves, or on the outside, and freezing up the vent holes, preventing the air from escaping when the valve was opened. Should the roller be knocked off, as might easily happen on a surface road, or the inclined bars of the safety stop get broken, no application of the brakes would be made, in which case, should the engineer be depending on the application of the brakes to indicate danger, the consequences might be serious. Black's automatic mechanical system is in use on portions of the elevated roads of New York City, and is a much simpler system than the Rowell-Potter. It is a signal system only, no attempt being made by mechanical means, or otherwise, to com- pel an engineer to obey the signal if it is at danger. The principal parts of the system are the two mechanical trips one the operating trip at the entrance of the block to set the signal at danger, the other a releasing trip to restore the signal to its normal position and the motion plate, by which the motion imparted to the pipe line connecting the two trips is made to work a signal. The signal, a photograph of which is shown in Fig. 9, is of the semaphore type, and is supported on an iron post, the top of the post being made in the form of a box or shield, behind which the blade can be hidden when in the safety position (shown in Fig. 10). The post is painted black, so that the safety indication is given by the absence of the signal, rather than by its vertical position. It will be noticed that three of the jaws on the two up and down rods connecting the "T" crank at the base with the signal, 109 are slotted; this is done to relieve the signal casting of the sudden blows transmitted by the two trips, and allow the signal to be moved to the safety position by the force of gravity. The trips are of the same pattern as those used with the Rowell-Potter system, the wheels of the engine pressing down the inclined bars placed at the side of the rail, thereby turning a shaft connected to the pipe line by a crank, and moving it, the motion being im- BRACK'S AUTOMATIC MECHANICAL SIGNAL AT "DANGER." parted through the medium of a spiral spring. The motion plate, by which the motion imparted to the pipe line is transmitted to the "T" crank of the signal, is shown in Fig. n. The plate P sliding in the frame F, forces the bar A to move from one side to the other, by means of the pin K working in the slot S, the bar A being connected by a rod with the "T" crank at the base of the signal pole. The slot is made somewhat longer than the travel of the pipe line, to allow for changes in the length of the con- no nections, from expansion and contraction, without affecting the movement of the signal. As installed on the New York Elevated, the blocks are of about 1,700 feet in length, the operating trip being placed a short distance beyond the signal, so that the signal will not go to danger until the locomotive has passed it. The releasing section BRACK'S AUTOMATIC MECHANICAL, SIGNAL AT "SAFETY." is carried past the next signal into the next block ahead, thus providing an overlapping section, and insuring that there will always be a signal at danger behind a train. On this road, the blocks are made continuous that is, the entire track is equipped with the device, and not at stations only, as is the case with the Rowell-Potter system on the "Alley L" road in Chicago. While the automatic mechanical systems have undoubtedly Ill been giving satisfaction in the applications that have been made, the argument cannot be made from these installations that such systems would be applicable with any degree of success or dura- bility to the ordinary surface roads of this country. In the equip- ments spoken of, the length of the blocks is between 1,000 and 1,700 feet, and, judging from the strains put upon the apparatus, this is very nearly the limit at which it is possible to operate the mechanism successfully. This length of block, with the long trains that are being run to-day, would furnish no protection, as the signal would be cleared about the time that the rear car MOTION PLATE. passed it. Unless the blocks can be made a mile long, it would be useless to make an application, even if the matter of expense was left out. Again, since the inclined bars are pressed down by the wheels of the train, it follows that the speed at which the train runs will materially affect the blow given the apparatus; and while, with the present equipment, there has been very little trouble from this cause, the speed has been about 40 miles an hour only, so that one cannot say positively what the effect would be at higher speeds. Any equipment to a surface road would be very hard to maintain during the winter, owing to snow and ice clogging up the parts, which, unless they worked freely, would apply the air 112 and stop trains when they should not be stopped. While the objection to the length of the block would be overcome by using an electrical release with the Rowell-Potter system, it is very probable that such an instrument would require too careful ad- justment and have to be of too light a construction to be a prac- tical success, unless the necessary electric current could be ob- tained from a dynamo instead of an electric battery. Such an arrangement was used with the equipment on the Intramural at the World's Fair, and gave very good results, but as a new design has been gotten out by the signal company, it is evident they do not think the arrangement then used would work successfully on a surface road. Regarding the applications of the systems that have been made to the elevated roads, criticisms have been made that it is impossible to run the trains on schedule time if the signals are obeyed, and that, in consequence, they are, on the New York Elevated, disconnected the greater part of the time, or the en- gineers are allowed to run by them when at danger. Any such criticism applies solely to the length of the block, or the distance apart which the signals are placed, and not to defects of the apparatus; for if the signals are further apart than it is desired to run trains, traffic will, of course, be delayed whenever the signals have to be obeyed. Having now explained the construction of the principal block signal systems used in this country, it would be well to more particularly describe the new Electric Train Staff system, which was spoken of as having been installed and in successful operation on the Chicago, Milwaukee & St. Paul Railway. The instrument used is one invented by Messrs. Webb & Thompson, and has been adopted by the London & Northwestern Railway Co., of England, as 'their standard apparatus for blocking trains on single-track lines. The essential feature of this system is that a staff, either of metal or wood, is carried by the engineer as his authority to run his train over a given piece of track, the several staffs provided with this system being locked mechanically in the station instru- ments, so that but one staff can be removed at the same time from either instrument. Two measurements are provided, one being placed at each end of the block or section of track to be H3 controlled by the staffs. A wire connection is made between the two instruments for electrically unlocking the machines and allowing a staff to be withdrawn. Withdrawing a staff from either instrument locks up both instruments of that section, so that no staff can be taken out until the staff withdrawn is replaced in one or the other instrument. A staff may be withdrawn from either instrument and a train started from either end of the block at any time, provided a staff has not been withdrawn and given to some other train that is in the block. To allow trains to be run in the block permissively for it is not always desirable to preserve an absolute block metal tablets of a different form from the staff are provided in a box attached, to the instrument, a special staff, called a permissive staff, being: used as a key to unlock the box. Each one of the six tablets which are provided, may be used to forward a train, the last train taking the permissive staff and any remaining tablets. From this it is seen that when an engineer has a staff with him on his engine, he knows that the block ahead of him is clear,. and that no train running in an opposite direction can enter the block. When he has a tablet with him, instead of the usual staff,, he knows that there may be other trains in the block running in the same direction that he is, but there can be no train running in the opposite direction. A general view of the staff instrument is shown in Fig. 12, am ordinary and a permissive staff being shown in Fig. 13. The principle upon which the machine is constructed is, that when a staff is withdrawn, the electric current energizing one of the mag- nets, which lifts the lock, allowing a staff to be withdrawn, is reversed through the coils of the magnet, neutralizing the effect of the current, and making it impossible to again take out a staff until the currents have been restored to their proper polarity by placing the staff back in one of the instruments. Details of the instrument are shown in Fig. 14, M being the magnet connected by the line wire with the battery of the other instrument through the key ^ (Fig. 15), N the magnet energized by a local battery, the circuit being closed by turning the in- dicator handle H (Fig. 15). T is the armature of the magnets M and N, and is of such a shape as to fit in the slot in the drum D and prevent it from turning, unless the armature is lifted by H 4 the magnets when they are raised. The arm carrying the mag- net is provided with a tail-piece, which is lifted by a projection on the staff whenever an attempt is made to remove a staff. If the magnets are energized by a current of the proper polarity, the armature is held up when the magnet is raised, and the drum WITHDRAWING STAFF FROM WEBB & THOMPSON STAFF INSTRUMENT. D is allowed to turn, the staff being drawn through the segmental slot K to the opening 0, where it can be withdrawn from the instrument. The levers L, placed on the side of the instrument, are made to reverse the electric currents flowing through the magnet N, by changing the contacts from one end to the other whenever a staff is withdrawn from the instrument, and thus not only prevent a second staff being withdrawn, but to reverse the current sent out to the other instrument when the key ,S is pressed, and make it impossible for a staff to be taken out of the other instrument also. THE STAFF. To explain more fully the effect of reversing the current through one of the magnets, it must be noticed that the poles of the two magnets are joined together, forming a single pole piece P. If the iron of the two magnets is magnetized, so that the north pole of one will be opposite the south pole of the other, there will be a mutual attraction, and each will satisfy the mag- netic attraction of the other. If, however, the poles of the two magnets are of the same polarity, it .will have the effect of making the two magnets act as one magnet, the iron projection P between the two coils now becoming the pole piece, and the armature, in DETAILS OF STAFF INSTRUMENT. consequence, will be attracted. From this it is seen that there must be two circuits to work the magnets of each machine one from the main line battery of one instrument, closed by pressing on the key S ; the other an entirely local circuit, closed by turning the indicator handle H of the other instrument. A galvanometer G (Fig. 15), placed in circuit with the main line wire, serves to inform the operator, when he wishes to with- draw a staff, when the operator at the other instrument has pressed down the key and closed the main line circuit. The indicator 7, placed on the left side of the instrument, is used to designate from which instrument a staff has been withdrawn, and also when turned clear around, to break the main line circuit, allowing the galvanometer needle to return to a vertical position, and thus notify the other operator that the staff has been removed and he can release the key. A single-stroke bell, placed on the main line circuit, rings each time that the key of the instrument at the other station is pressed. The staffs (shown in Fig. 13) are made of a piece of iron pipe, name plates of the two stations at the ends of the block being riveted on the end. The rings riveted to the staff serve as pro- jections to fit in between the wings of the drum D (Fig. 14), to raise the several latches, as well as to make it impossible to place a staff taken from the instrument of one block into the instru- ment belonging to the next block. The number of staffs pro- vided for each instrument is usually ten, but as many may be used as will go into one machine. The operation of the instrument is as follows, supposing that n8 A and B represent the stations at the end of the block, and that a train is ready to start from A to B : A asks, by taps of the bell, according to an arranged code of signals, for permission to withdraw a staff. B responds, if there is no train in the block, by pressing on the key 5 (Fig. 15), sending a current through the magnet M and the galvanometer, the needle of which is imme- diately deflected. A turns the indicator handle H to "For Staff/' thereby closing the local circuit, and lifts the staff to the top of the slot, as shown in Fig. 12. The projection on the staff strikes the tail-piece, as before described, lifting the magnets, which, as as the pole piece is now energized, also lifts the armature and unlocks the drum, allowing the staff to be withdrawn from the instrument. Immediately the staff is withdrawn, the operator turns the indicator / (Fig. 15) to "East" or "West Staff Out," as the case may be, breaking the main line circuit, and allowing the galvanometer needle to assume the vertical position and inform the operator at B that a staff has been withdrawn. Withdrawing the staff automatically causes the indicator H to return to the position "For Bell," breaking the local circuit and preventing the battery from being run down. The staff is then given to the conductor, who gives it to the engineer, to be kept on the engine. On the arrival of the train at B, the staff .is taken to the office and given to the operator, who places it in his staff instrument, and notifies A, by taps on the bell, that the staff has been placed in the instrument. Another staff can then be withdrawn from either instrument, and a train sent in either direction through the block that is desired. To block trains permissively, six tablets are provided, a pro- jection or lug being provided on the end of the permissive staff (shown at G, Fig. 13), by which the box E (Fig. 15), on the face of the instrument, in which the tablets are placed, may be un- locked and the tablets removed, as shown. As soon as any of the tablets are removed from the box, a catch-piece C is dropped into the opening 0, so that the large ring W on the permissive staff will not admit of the staff being put back in the instrument, and as the permissive staff has been withdrawn, both instruments are locked, and no other staff can be withdrawn. When it is desired to send one or more trains into the block, the permissive staff is withdrawn from the instrument the same as any other staff, and is used to unlock the box containing the six tablets. A tablet is given to each train that it is desired to send forward, the last train being given the staff and any remain- ing tablets; for, unless all the tablets have been placed in the tablet box of one instrument or the other, it is impossible to put the permissive staff in either instrument, and, until this is done, A STAFF LOCK. the section would be blocked, as no other staff can be with- drawn. A staff lock, shown in Fig. 16, is provided at all switches in the block. Projections on the staff, seen at P, Fig. 13, serve as a key with which to unlock the switch. To unlock the switch, the end of the staff is placed in the lock and turned, when, if the catch is pressed down, the switch may be thrown by pulling over the lever, as seen in Fig. 17. When the staff is turned in the lock and the switch opened, it is impossible to withdraw the staff, so that a trainman not only has to leave the switch set for the main line, 120 but is compelled to lock it before removing the staff. As no switches can be opened without the staff, and there can be but one staff taken out of the two instruments at the same time, an engineer knows when he has the staff that all the switches in the block are set for the main line and that they are locked. There have been several staff instruments of different patterns perfected and put in use in England, but the one shown here is very simple and durable, and not likely to get out of order. The battery power required is large, but this is due more to the high resistance to which the magnets are wound than to the particular design of the machine The system is in use on the C, M. & St. P. Railway, between SWITCH UNLOCKED AND THROWN. the stations, Savanna, in Illinois, and Sabula Junction, in Iowa a section of track that is not only very crooked, but in which there is a very long bridge over the Mississippi River, making it highly important that every precaution be taken against acci- dents, not only from rear end collisions, but from dispatchers and others making mistakes, and allowing two trains, traveling in opposite directions, to try and use that piece of track at the same time. The system has given the greatest satisfaction since it has been installed, not only in regard to the safety with which the traffic can be handled, but the facility with which trains can be dispatched. To quote the words of Mr. C. A. Goodnow, super- 121 intendent of the C, M. & St. P. Railway, in a paper read before the Western Railway Club: "I cite these few examples of the many complications that must necessarily arise in the handling of traffic on single track, to illustrate the facility with which the staff system does its train dispatching; its possibilities in connection with the movement fe cc i *r 8 1 -> aj V 1 --. 8 o m* JL3 5 :- 0) p^ Q > u O s *V ^> 2 00 O> a o u_ ^ / >o 00 r i h ^ (CM 2 b 2 12 r- s _ 'co^^oo"* ^e ^ D t " * ^ c f T> * 05 ui * OC r* 162 in whether there is or is not a probability that they will ever be needed. It is to be noticed that dwarf signal No. 3, while governing back-up movements in two directions, either on the main line or on the cross-over, is a single-blade signal, and does not in- dicate to trainmen the route they are to take, but only that the route is clear for them to back up. The reason for this being that in practice it has been found to be the best plan, as the move- ments are made slowly, and the towerman has, in any event, to know the movement it is desired to make before throwing the levers, for the matter to be left entirely in his hands for him to throw the proper switches before clearing the signal. Again, a single-blade dwarf signal can be put in between tracks in many places where there is not room in which to put one having two blades. The saving in original cost of the plant is also quite an item, particularly if there are very many dwarf signals required. The combinations of the levers by which the derails are closed and the signals cleared for a train to proceed over the crossing, are very simple with the plant under consideration, the derails of any one track being first closed, the home signal next, and, last of all, the distant signal. When it is desired to use the cross-over, the engineer will run by sufficiently to clear signal 4, when the towerman, putting levers I, 2 and 6 back to normal, will reverse lever 8 and pull 4. Reversing lever 4 clears the signal, and the switch being thrown, the train is switched over to the other main line. With plants having any more levers than those necessary to operate a simple crossing, it is necessary to provide in the tower, for the guidance of the signalman, a "combination sheet," show- ing the order in which the different levers should be pulled to clear the different routes. This order is made necessary by the locking of the levers, those which are free to move being first reversed, then those which are released by the reversal of the first lever, and so on, until all the signals for the route desired have been closed. Although it would appear to anyone a very easy matter to find out which are the proper levers to reverse to give a clear route, such is not the case, as has been shown in many in- stances, where, from some cause or other, it has become neces- sary to put new men in the tower who do not know the proper levers, or when, as in case of accident to the signalman, the trainmen have had to work the levers to let their train over the crossing. A large size map, with the switches and derails plainly num- bered, as well as a combination sheet in which the levers to be pulled to set the different movements are shown in large type, should be framed and hung up in every tower. The wisdom of so doing will be evident to anyone who has ever tried, unaided, to clear the signals for a train at even a small crossing. CHAPTER X. FOR JUNCTION POINTS AND DRAWBRIDGES CONSTRUCTION OF THE IMPROVED SAXBY & FARMER INTERLOCK- ING MACHINE. Junction' points, like crossings, unless protected, are dan- gerous places for trains to run by without first making a stop, not only because the switch may be wrong, but because a train may be approaching on the other track that will also want to use the junction switch. By interlocking the switches, provid- ing signals to govern the different routes, and by putting in de- rails to enforce obedience on the part of the trainmen to the indications of the signals, there will be no necessity for any train's stopping, unless it be to register or to get orders, as it is prac- tically impossible for the signalman to make a mistake and set the switches and clear the signals, so as to permit two trains to come into collision with each other. In the preceding chapter the signals necessary to govern a simple switch or junction point of two tracks were shown, and while the same general plan of signaling is followed in all such installations, the particular arrangement of the signals will vary with each location. In Fig. i a junction of a single with a double track road is shown, cross-over switches being provided to allow trains to run on to the single track from either of the other two. It will be noticed that two signals are provided for all facing-point switches, or those where it is possible for a train running in the normal direction of traffic, either to keep to the main line or to take the switch. In these cases the upper blade governs the high-speed route, or is the one that is "cleared" when the switch i6 7 is set for the main line; when the switch is set for the cross- over, the lower blade is the one that governs. The junction switch, or the one numbered 6, is used both as a facing and a trailing point; but as it is normally a trailing point, it must be so signaled, a single-blade dwarf signal being used to govern trains going on to the branch line, while the two signals numbered 10, which govern the switch in the normal direction of traffic, are of the standard semaphore pattern. As the two signals which govern the switch in the normal direction of traffic should never be cleared at the same time, they can both be connected to and worked by the same lever in the machine, a selector being used to connect the proper signal with the lever, the same as if they were both on the same pole. It is for this reason that they both are numbered 10. The only derails necessary are those on the branch and on the main line, just before reaching the junction switch, where, if a train should run by the signal, it might run into another train, should it happen to be using the switch at that time. Derails are not necessary at any other point, as there are no other switches upon which a train might come into collision with some other train, should the signal be run by when at danger. Any train running by signals Nos. 2 or 7 when at danger would only be going in the normal direction of traffic, so that no collision, which the inter- locking was designed to prevent, could happen. Interlocking signals are not intended to prevent collisions between two trains running in the same direction, as that is the province of a block signal, and not an interlocking plant. The locking required to make the operation of the plant safe is somewhat more complicated than any of those which have been given, as there are a greater number of levers that have to be locked when the signals governing the different routes have been cleared. For instance, lever 2, when reversed, must lock lever 5 both ways, normal or reversed, as lever 2 clears the signals for either route. It must also lock signal 3 reversed when cross-over switch 5 is reversed, to make it safe for a train to run on the other track. For, as lever 3, when reversed, locks the lever con- trolling the route on the main or the branch line in the proper position for a train to proceed in either direction, lever 2, when 1 68 reversed, practically locks all the levers of the route indicated, making it safe for a train to proceed through the limits of the interlocking. The distant-signal lever No. i, when reversed, locks lever 2 reversed with lever 5 normal. This prevents the clearing of the distant signal when signal 2 is cleared for the cross-over or the slow-speed route. The combinations of levers to be reversed to give clear signals for the principal routes, or the "Combination Sheet," as it is called, are as follows: Main track, east bound, 2, i. Main track, west bound, 7, 9, 10, 12. East bound, main to branch, 6, 5, 3, 2. Branch line to west bound main, 6, 9, 10, n. The signals and derails necessary to make the use of a junction of two double-track lines safe, so that trains need not stop unless the signals have been cleared for some other train, are shown in Fig. 2. The signals required are much the same as those shown in Fig. i, but it is necessary to use one more derail, so as to protect trains that use the cross-over and back over on to the west bound main line. A dwarf signal (No. 14) has also to be put in to govern trains making this movement. In working out the locking of the levers necessary to make the operation of this plant safe, there are several points to which attention should be called. As it would be unsafe for a train to use the cross-over switch if it were possible for the signalman at the same time to close any one of the derails, lever 9, when reversed, must lock levers 6, 8 and TO in the normal position. Lever 6, when reversed, must lock lever 7, either in the normal or the reversed positions, to prevent that lever being thrown after 6 is reversed. Lever 6 must also lock lever 10 normal, when lever 7 is reversed to make it impossible for two trains to come together where the two tracks cross each other. It is also to be noticed that the signal levers in all cases, when reversed, lock the levers of the signals governing the opposing route in the normal position, so that it is impossible for the operator to give clear signals to trains running in opposite directions, that would come into collision with each other. 169 CO 00 CO < p id i o 0> w 00 VIIX (M ^ ^^ V-X ^ T> (B) O s~^ *s r r Ul tc, UI DC ^ < < CM) (Jo) 9 00 1 a (/> 0. (0 ^ r- CO S "' < a M co r co f E 6S)(Si , LP- 1Z > : CTC * iv ujaUosq Ion 1- Xi S =fc ^ ^x>| kn >1 Uod-loo) FIG. u 3. LOCOMOTIVE ENGINEERING. N, Y. DOG SHEET, NATIONAL, MACHINE. ing may be arranged in a very small frame space, and at the same time may be gotten at easily. The special locking used by the National Company is a very simple and easily arranged contrivance, consisting simply of a square block held by a frame or "cage," fastened on the outside of the tappet, as shown in Fig. 4, special dogs being used, which not only fit in the tappet in the ordinary manner, but project over the bar and bear against the block. If the tappet is in a position where the block is held in between the two they will practically act as one, while if the block is withdrawn, both dogs will be free to move. In this way one lever is made to lock another lever only when a third lever is in a certain position, normal or re- versed, whichever is desired. By making cages to hold several blocks, and by using dogs of special form, any special locking, no matter how complicated, can be easily arranged. 18; The claims made for this style of machine are "that as the rocker is made with either single or double tappet connections, single or double locking can be used, placed on either side of the frame with equal convenience and accessibility, thus doubling the capacity, the symmetrical construction of which secures the great- est economy of room and minimum cost of maintenance, for the reason that more locking can be carried in less space and with less cost than in any other machine a feature of special importance in all installations ; that the locking frame is so designed that the locking is open to the unobstructed, view of every part, and thus FIG. 4. SPECIAL LOCKING. not only facilitates free inspection but permits the placing of new locking (for additional levers), or the introduction of special locking, with the greatest convenience and economy." Few or no objections can be brought against this style of ma- chine, save in the construction of the locking, as the tappet bars are connected in a very simple and direct manner. It will be noticed that with this kind of locking the dogs extend from one end of the machine to the other, instead of across the locking bars, as is the case with the improved Saxby & Farmer locking. Making the bar long in this way means that it will require more force to drive it, and as the bar is driven by sliding friction be- tween the end of the dog and the notch cut in the tappet, it is very apt to cause wear. Again, as the dogs all lie in a vertical plane, any wear will take place on one side only, and will in time affect the locking, possibly necessitating its renewal much sooner than with the other stvle. 1 88 The Johnson machine is illustrated in Fig. 5, a detail draw- ing having been shown in the article on Controlled-Manual Block Systems, the Patenall lock instruments having been applied to one of these machines. This machine differs from the one made by FIG. 5. THE JOHNSON INTERLOCKING MACHINE. the National Co., in placing the rocker on a bracket fastened to and moving with the lever, instead of suspending it from the main frame. Practically the results are the same, as motion is im- parted to the tappet both before and after the movement of the lever, as in the other machine. Special locking is provided for by a loose dog fastened to the tappet, which acts in the same manner as the block used with the National machine. Instead, however, of fastening this to the i8 9 surface of the tappet, the bar is cut away to allow the dog to work in the same plane with the other locking dogs, a tie-piece being used to join the two parts of the tappet bar together. This construction is objectionable and is not as good as the National arrangement, owing to the chance of the bar becoming disconnected and throwing the locking out of service. With a machine having the rocker situated as it is on this one, the locking can be placed only on the front part of the frame, where it will be behind the pipe connections to the levers, and very hard to get at to make any changes or repairs. While the improved Saxby & Farmer locking may take up more room than the Stevens locking, there is little doubt that it can be repaired more easily, owing to its being in a horizontal position, or that it will last longer, there being less friction between the locking bars and the dogs. However, both types of machines are strongly made, and will need but little repairs or attention be- yond an occasional oiling, and whatever cleaning is necessary to keep every part clean and free from gum, for if the locking is not kept clean it is apt to stick and give, trouble. Next to the machine, the most important part used in the construction of an interlocking plant are the movements by which the switch points are opened and closed and also locked, whenever the lever to which they are connected is moved. By far the simplest arrangement that can be used is to make a direct connection from the lever to the head rod of the switch, the same as if the connection was made from an ordinary switch- stand. This would provide for throwing the switch, but would not lock it, and as there is considerable spring in pipe lines used to make the connection to the lever, some form of lock with all facing-point switches is absolutely necessary to prevent the points from moving under a passing train. Locks are easily made by attaching a rod to the points of the switch and providing a plunger or lock pin which can be entered into holes drilled in the rod, thus locking it, arid with it the points of the switch. There are two forms of these locks which, as they are generally used at facing switches, are called "facing- point locks," one being placed inside of the line of rails, and shown in Figs. 6 and 7, and the other placed outside of the rails and shown in Fig. 8. As will be noticed in the one shown in Figs. 6 and 7, the two points of the switch are connected together by a front rod, which answers for a lock rod by passing it through the plunger casting bolted to the tie, holes being drilled in the rod for the plunger to enter and lock the switch, either for the main line or for the side track. With the form shown in Fig. 8, the lock rod is attached to INSIDE-CONNECTED FACING-POINT LOCK. the front rod midway between the two points, and is made long enough to pass through the plunger casting which is placed outside of the rails, holes being drilled in the bar to lock the rod and with it the points of the switch. Of the two methods, the second is the one ordinarily used and is much the best, as the casting is out of the way and not liable to be caught and torn out by any loose brake beam, or other defective part, on a passing train. However, the inside lock is the only one in use in England, the claim being made for it that it is the strongest, and that by the lock being made a part of the switch it has got to be kept in working order. Using locks such as are here shown to lock the switch, neces- sitates using a separate lever to work the lock, which, in a great many instances, is a very expensive arrangement. Where such LOCOMOTIVE eNQINEERINQ. N, Y. CONNECTIONS TO INSIDE-CONNECTED FACING-POINT LOCK. LOCOMOTIVE ENQINEERINQ. N. Y. OUTSIDE-CONNECTED FACING-POINT LOCK. are used, the general plan followed in numbering the levers and arranging the locking is the same as that which has already been described, except that the levers of the signals, when reversed, must lock the lever of the switch lock reversed in all cases, whether the switch is in the normal or the reversed position, for the reason that a switch should always be locked before it is used by a train, irrespective of the way the switch is set the order in which the levers would have to be reversed to set up any route for a train being: First, the switch or derail lever; second, the locking lever; third, the home-signal lever, and, last of all, the lever of the distant signal. Owing to the cost of providing two levers to move and lock a switch, and the time required to pull the second lever, many attempts have been made to perfect a machine which would ac- complish this by the movement of one lever. The result has been the invention of the switch-and-lock movement, which can be said to be a practical success, the principal feature of the design being the preliminary action of unlocking the switch before the points are moved, and then locking them again after the movement has taken place, very much the same as the preliminary latch locking acts with the levers 'of the interlocking machine. A general plan of a switch-and-lock movement, showing its construction and the connections to the switch, is shown in Fig. 9, a detail drawing of the movement being shown in Fig 10. A is the base casting, having a guide at one end, through which the lock bar L is made to pass. A slide bar R is connected by a pipe line to a lever in the tower, and is moved a distance of 8J inches whenever the lever is reversed or returned to the normal position. Riveted to the slide bar are two lock pins P, which are made to enter holes drilled in the lock bar, and lock the bar with the base casting, so that it is impossible to move the switch until the slide bar is moved and the lock pin with- drawn. Working on one side of the slide bar is a switch crank C, which bears against an operating or driving pin D, fitted between the upper and lower bars of the slide bar R, and is connected to the switch rod by a short connecting rod. The switch being shown in the normal position, if, now, the man in the tower should reverse the lever, the movements that would be made would be as follows: The slide bar would be moved, withdrawing the lock pin from lock bar, leaving the switch points free to be moved, the operating pin sliding along the straight surface of the switch crank. After the switch had been unlocked, the operating pin would strike the arm of the 193 switch crank C, causing it to revolve and carry with it the switch and lock bar, the throw of the crank being regulated so SWITCH-AND-L,OCK MOVEMENT AND CONNECTIONS TO SWITCH. FIG. 10 (o o o LOCOMOTIVE ENGINEERING, N. Y. DETAILS OF SWITCH-AND-LOCK MOVEMENT. that when the switch had completed its movement the operating pin would slide along the other arm of the crank, entering the 194 lock pin in the other hole in the lock bar, locking the bar and the switch in its new position. Theoretically the results obtained are the same as when separate levers are used to work and lock the switch, but in practice it is found that the use of the switch-and-lock move- ment has some limitations beyond which it cannot be said to be good practice to go. The weak point in the movement is that with the short motion available there is not travel enough in which to make the locking 1 of the switch certain, owing to the spring in the connections when the switch is any great distance from the tower. As with the locking of the levers in the tower, it is taken for granted that if the lever is pulled over for its entire distance the switch has been thrown, and it is therefore possible to reverse other levers which depend for safety on the fact that the switch is properly set and locked. Any plant cannot be considered safe where the switch-and-lock movement is such a distance from the tower that it would be possible to reverse and latch the lever, owing to the spring in the connections when the lock pin had not entered the lock bar, as would be the case if some- thing should get in between the points of the switch and prevent it from closing. If a bolt lock was put in the connection to the signal, it would be impossible to clear the signal if the switch was not properly closed; but as it is not practicable to put a lock on the signal for every switch which the signal may be used to govern, dependence must be placed upon the switch being closed properly when the lever is pulled over. No definite distance can be laid down beyond which it would be unsafe to go, as the number of cranks and any curves which may be in the pipe line will materially affect the amount of spring and lost motion in the connections. Then, again, the number of movements that are placed on a lever will also make a difference in the chances of failure, owing to the greater power required to work the switches. My experience has been that a switch-and-lock movement for distances up to 500 feet is a safe and easy method of working a switch, but beyond that I would advise a facing-point lock, although the latter cannot be said to be a perfectly safe arrange- 195 ment. Instances can be named where serious wrecks have occurred from a failure of a facing-point lock, owing to a break in the switch connection allowing the lever to be reversed with- out closing the switch, the plunger entering the same hole in the lock bar when its lever was reversed, and locking the switch in the same or open position. On the Pennsylvania a staggered lock is being tried to prevent an accident of this kind happening, the staggered lock being made with two lock bars and two plungers of different shapes, which fit in different-shaped holes cut in the bars; so that unless the switch has been moved, the plunger cannot enter the hole in the lock bar and the lever cannot be reversed, and so the signal cannot be cleared. More than two switch-and-lock movements should not be put on one lever, as the power required to work them would be more than it is safe to put upon a pipe line. While it may work well for a time, the wear upon the parts would be so great that very soon there would not be throw enough at the movement to cause the plunger to enter the hole in the lock bar and lock the switch. Again, the power required to work three movements is often more than one man is able to exert, and by getting another man to help pull over the lever there is great danger of doubling up the pipe should anything prevent one of the movements from closing, the lever being reversed while the switches would be open. More especially is this likely to happen in winter, when the cranks, the movements and the switch points are apt to be clogged with snow and ice. A very important part of the apparatus that is designed to work in connection with a switch-and-lock movement or a facing- point lock, and which, as yet, has not been spoken of, is a detector bar or slide, which prevents any movement of a switch or lock being made so long as any wheel of a train may be on the bar. The necessity of providing some such arrangement as this is ap- parent, as without it, it would be possible for a signalman to throw a switch under a moving train not that he may do so intention- ally, but to prevent his pulling the lever by mistake, thinking, perhaps, that the train had cleared the switch. The construction of a detector bar is shown in Fig. n, and the arrangement by which connection is made to the crank that moves the switch can be seen in Fig. 9. A is the bar which is 196 held against the side of the rail by the link K, the link being pivoted on a casting C or fulcrum point called a clip, which is fastened to the rail as shown. Lugs B are provided on the clip, which limit the movement of the bar and prevent it from ever being allowed to get too much out of adjustment. When an attempt is made to throw the switch, the first move- ment of the crank moves the connecting or driving rod of the detector bar and causes the bar to move through the arc of a circle, with the pin of the clip as a center and the link as a radius. This brings the bar to the top of the rail, raising it to about one inch above when the bar is in the center, or has been moved through one-half of its stroke. If a wheel was on the rail at the LOCOMOTIVE ENGINEERING. IV. V. DETAILS OF DETECTOR BAR. time the attempt was made to change the switch, the bar would be raised, striking against the outer edge of the tread of the wheel, which projects about an inch over the side of the rail, and no movement of the switch and lock movement could be made, and therefore the switch could not be changed. As the bars are made 45 feet long, and are placed alongside or ahead of the switch, so that the wheels should first pass over the bar before getting to the switch, it is impossible to change the switch under any part of the train, as this length is greater than the distance between any two pairs of wheels of any car or engine, or of any two cars that may be coupled together. The general arrangement of a switch-and-lock movement and the connection to the detector bar are shown in Fig. 12, the photo- I 9 7 graph helping to make the drawing more easily understood. When a facing-point lock is used, in place of a switch-and-lock GENERAL ARRANGEMENT OF SWITCH-AND-LOCK MOVEMENT AND CON- NECTIONS TO DETECTOR BAR. SIDE VIEW OF SWITCH-AND-LOCK MOVEMENT AND CONNECTIONS TO SWITCH. movement, the detector bar is connected to the lever that works the lock, to make it impossible to unlock the switch while a I 9 8 train is passing over it. If there is no room in which the bar may be placed ahead of the switch, it is usual to use two bars, one on each outer rail, thereby insuring that it will be impossible to change the switch under a train when either track is in use. As the detector bar makes it impossible to change the switch so long as a pair of wheels are on the rails over the bar, it is necessary for an engineer to run by the switch sufficiently to clear the bar, if he wishes the switch to be changed before going in the opposite direction. To have to do this seems to trainmen a useless pre- caution, and they often blame the signalman for making them run further than is apparently necessary, whereas the man is power- less to make any change until the bar is cleared. Trainmen should never stop their trains with any of the wheels over the bar, for if they do so, they have not cleared the interlocking, and until they do so clear it the switches cannot be changed or a route set for anv other train. CHAPTER XII. DETAILS OF CONSTRUCTION. There are many points regarding the form and particular arrangement of the several parts used in working the switches and signals at an interlocking plant, all of which are of great interest to signal engineers, or to those whose duties require them to keep such parts in repair, but to the average railroad man, who perhaps will never have occasion to use any knowl- edge that he may have gained relative to the "details of con- struction," a discussion of these questions cannot be expected to be so interesting or as useful. With those, however, who are daily required to read and act upon the indications of the signals, a knowledge of the con- struction will help them to put greater faith in the indications and to better understand the precautions taken to insure that a signal will not give a wrong indication. The construction of the different machines and the method of moving and locking a switch having been explained, the parts next to be considered are those used in making the connections, or the means whereby the motion of the levers is transmitted to the switches and signals. Those for the switches and locks are made of i-inch iron pipe, while for connections to signals they are generally made of No. 9 galvanized steel wire. Pipe is used, as giving for its weight and section the greatest stiffness, this quality being of the first im- portance from the fact that but a single connection is used be- tween the lever and a switch, the power being applied as a pull for motion in one direction and a push in the other. If the pipe was used in tension only, but few supports would be needed, but to transmit any force by compression, it must be kept in a straight line, and to do this it must be firmly supported at short and regular intervals. These supports are called "pipe carriers," and are made of several different patterns, those of the anti- 200 friction type being the best and the ones now used in all new work. They are made as may be seen in Fig. i, so as to allow the wheel which carries the pipe to roll on a center pin in a slot cast in the frame, all friction being done away with when the pipe is moved, except that due to rolling, which is very small. There being no rubbing friction, no oil is required with this style of carrier, and it is possible to box them in, no covers being needed. The tendency of the pipe being to bend in any direction ANTI-FRICTION PIPE CARRIER. when power is applied, a small roller is fixed in the top part of the frame of the carrier, to hold the pipe in the groove in the lower wheel. As the lower wheel has to carry the weight of the pipe, the bearing seldom comes against the upper wheel, so that the latter is usually made in the form of a sleeve slipped over a bolt, which answers for a shaft upon which it may turn. With the Evans carrier (seen in Fig. 2), the bearings of this upper wheel are also made anti-frictional, a rod being slipped through the top part of the frame to tie the several parts together, which, in this pattern, have to be made in separate pieces, to allow the wheels to be put in their proper places. The carrier, for this reason, is somewhat more troublesome to put in, two lag screws being needed for each section of the frame, and it is 2OI also more troublesome to work with, owing to the difficulty of removing any of the wheels, as is so often desirable in repairing or putting in any new parts, while the advantages gained by using the anti-friction roller on top seem to be very slight. Pipe carriers are usually spaced 8 feet apart and are fastened to foundations which are buried in the ground. On some roads this distance is reduced to 7 feet, but for ordinary straight work there does not seem to be any need of going to the extra expense involved in putting them closer together. For 500 feet of pipe THE EVANS DOUBLE ANTI-FRICTION PIPE CARRIER. line, the spring of the pipe under a load of two switch-and-lock movements will be nearly one inch, or for 9 inches travel at the lever there will be but 8 inches at the movement. This spring has to be provided for in putting in the work, by making one of the crank arms at the movement longer than the other, so that the 8f inches of travel required for the full stroke of the movement will be obtained. In running the pipe lines, they should always be made as straight as possible, and never curved if it can be avoided. Where this has been done, owing to curves in the track, the carriers should be put closer together, to prevent the pipe from springing or bending when the load is put upon it. In curving out to run around a side track, when the angle is not sharp enough to require the ordinary form of crank, radial or ordinary cranks should be used, as shown in Fig. 3, the pipe all being run in straight lines. In joining the several lengths of p>pe 202 to make a long line, they are screwed together with an ordinary screw coupling such as is furnished with the pipe, an iron plug about 4 inches long, having holes for rivets drilled in each end, being put inside and riveted to each pipe, as shown in Fig. 4. Riveting the pipe in this way takes a great deal of the strain from the threads, and will even hold the pipe together should the threads work loose. Jaws for making the connections to cranks or switch-and-lock movements are fastened to the pipe by an ordinary coupling, the threaded end being turned down to go inside the pipe and be riveted to it in the same manner as a plug. As this method is patented and controlled by one of the signal companies, other ways of accomplishing the same end are to be found in use, but none that are on the market at the present time can be said to be satisfactory, except to weld the pipe to the solid end of the jaw. The connection between the lever and the switch being prac- tically a solid connection, expansion and contraction have to be provided for to keep it of the same length, or else the movement might not be completed when the lever was reversed. The sim- plest way of doing this is to put a rocker in the middle of the pipe line, so that the motion or direction of travel in one half will be 20 3 in the opposite direction from what it is in the other. When one half of the pipe line expanded with an increase of temperature, the other would have expanded the same, the rocker being turned on its center to a new position, while the total length of the con- nection from the lever to the movement would not be changed. Putting a rocker in the pipe line to reverse the motion is a very awkward and expensive arrangement, on account of the space required and the difficulty of running the pipe where there are more than two lines to be compensated. A PLUGGED PIPE JOINT. A device that accomplishes all that is required in this direc- tion and in a very neat way, too is shown in Fig. 5, it being what is known as a "lazy jack" or compensator. As will be seen, it is made by arranging two cranks with a short connecting rod between them, on a base casting, one of the cranks being of a very obtuse angle and practically a rocker, and the other of a very acute angle, to change the direction of the thrust of the pipe line and make it as near a right angle as possible. In this way the pipe is continued in a straight line, while its motion is reversed and expansion and contraction provided for. But owing to the 204 angles at which the two cranks are set with each other, the throw or travel of the pipe line is limited, the cranks, if turned too far, getting too near the dead center when it would not be possible to move them either way. For this reason one compensator is usually provided for every 500 feet of pipe, although distances of 700 feet are not uncommon. For distances less than 100 feet no compensator is needed. Where the pipe line is under 500 feet in length, the com- A "LAZY JACK" COMPENSATOR. pensator is put in the middle of this distance, or 250 feet from either end, and so, when two or more are used, they are each put in the middle of the length of pipe which they are to compensate. This appears simple enough, but workmen, unless watched, will divide the distance, spacing the compensators evenly between the levers and the movements, thus leaving a certain part of the line unprovided for. For example, if the distance to be compensated is 900 feet, the compensators should be put one-quarter of this distance, or 225 feet from each end, instead of 300 feet, as would seem at first sight to be the proper distance. 205 If it should be necessary to place two cranks quite close to- gether in any pipe line, they may be connected up so as to reverse the motion, and thus take the place of a compensator. To arrange them in this way will be found very convenient in running pipe lines around side tracks that have been put in after a plant is in operation, and when it is desired, on account of the con- nections to the switches, to keep them in the same relative posi- tion with reference to the track. This arrangement is clearly shown in Fig. 6, where the cranks are arranged to compensate and the lines outside of the side track are brought under and ex- tend on parallel with the main line in the same numerical order. The foundations for the cranks, compensators and other parts of the apparatus subjected to very heavy strain are made of heavy oak lumber, dovetailed and braced, the whole being buried in the ground to the proper depth, tamped and finally concreted, this being made of one part cement to two of sand and three parts of crushed stone, such as will go through a 2-inch mesh. The con- crete should be put in at least 12 inches deep, and for 12 inches around each side of the foundation. The cement used should be of the best quality, on account of the strains that are put upon the foundation, and while it is not necessary to^e^iouthe expense 206 of getting the imported article, none but the best of domestic manufacture should be used. At the present day the standard practice of connecting the home signal with the lever in the tower is to use two lines of wire, bolt locks, to prevent the signals being cleared unless the switch is closed, being connected to the down-pull wire, or the one that pulls the signal to clear. On several roads, however, a decided step in advance has been taken by making this connection of pipe instead of wire, thereby doing away with the necessity of constant adjustment, and the possibility of giving a wrong signal from this cause or from the breakage of a wire. That the consequences would be serious were a wrong indi- cation to be given by the home signal, no one will doubt, and while the cost of making the connection of pipe is considerably more than where two lines of wire are used, the additional safety secured in the operation of the plant most certainly warrants its use. When the home signals are wire-connected, a bolt lock should be put in the connection to the distant signal as well as to the home signal. This is to make it impossible to clear the distant signal, as well as the home signal, unless the switch or derail has been properly closed, for if a bolt lock is not used, there is nothing to prevent the signals being cleared if the wire to the home signal should break between the bolt lock and the lever, the operator being then able to reverse the home-signal lever and pull over the lever of the distant signal. Should the wires to the home signal be caught, or should the derail not be closed and the bolt lock hold the wire, it is possible, from the spring in the wire, for an operator to reverse the home-signal lever and then give a clear distant signal when the home signal had not been cleared. Connecting the home signal with pipe admits of the use of a strong and well-designed selector, and one that will act as a bolt lock as well as a selector. This selector is shown in Fig. 7, and, as will be seen, consists of two slide bars working in grooves in a cast base piece, and which are connected to the two signals to be operated, a third slide being provided which is connected to the signal lever, and so arranged by means of a crossbar that it may be made to slide in either of -the grooves occupied by the other two bars, and, by shoving them out, clear the signals to which 207 they are connected. The crossbar usually is connected .to the lock bar of the switch, the operating slide being brought behind the slide which will clear the proper signal, to govern the switch the way it is set. Where not convenient to connect the selector directly to the lock bar of the switch, the crossbar may be worked by a motion plate operated by the pipe line working the switch, or else a direct connection, consisting of a crank and short con- nection, may be used. When the crossbar is attached directly to the lock rod, the selector is made to act as a bolt lock, for, as the operating slide works in a notch cut in the crossbar when TO SIONAU LEVER SELECTOR FOR PIPE CONNECTIONS. the slide has been pushed ahead to clear the signal, the crossbar will be locked, and with it the switch. Where the connections are of wire there are several different forms of selectors that may be used, the design of each being made upon the same general plan. The connections being of wire, the signals must all be cleared by a pull instead of a push, each signal being connected to a long hook which can be caught by a sliding plate connected to the lever in the tower, and drawn back when the lever is reversed, instead of being pushed forward, as is the case with the pipe-connected selector. Differences in design are found in the several devices, whereby the movement of the pipe lines to operate the switch is auto- 208 matically made to select the hook connected to the proper signaL Of these the one in most common use is where selection is made by means of a driving bar carrying dogs, which are set to throw every hook out of engagment with the sliding plate except the one to be operated, the driving bar being driven by means of a motion plate worked by the pipe line connected to the switch. Another way, and one that is more certain in its action, is to make the switch connection turn a shaft by means of an escape- ment crank, cam lugs on the shaft being provided to raise all but the proper hook to be operated. With all of these forms of selectors, dependence is placed for their proper operation upon the assumption that the proper hook will drop and be caught by the sliding plate when the dogs or cams are set by the movement of the switch connection, and that the adjustment of the wire will be such as to bring the point of the hook in the proper position to be caught. As the adjustment of the hooks is often bad, owing to changes in the length of the wire and they can easily be prevented, by snow, ice, dirt, or by a poor adjustment of the dogs on the driving bar, from being caught it will be seen that the arrangement is not a very good one. To be relied upon, it should be positive in its action, and this can only be obtained by using pipe connections instead of wire. A device that is sometimes used in place of a derailing switch at short sidings, or where it is desirable to allow cars to stand as near the main line as safety will permit, is shown in Fig. 8, and is known as a "scotch block." It is connected to a lever in the ma- chine, so that it can be raised on top of the rail in the position shown when it is desired to block the line, or else lowered out of the way when the track is to be used. As may be supposed, it answers the purpose very well, derailing any car or engine that attempts to pass it, but as in so doing it is apt to cause some damage to the trucks or brake gear, the device should not be used, unless there are very good reasons for not putting in the ordinary form of derail. In the design of the tower, simplicity and cheapness, as well as adaptability for the purpose for which it is constructed, are the principal points to which attention may be called. As will be seen by Fig. 9, the tower is made two stories in 209 A "SCOTCH" BLOCK. C. M. & ST. P. TOWER. 2IO height, to allow the operator to have an unobstructed view, as far as possible, of the tracks protected by the interlocking, the stairs to give access to the operating room being put on the out- side of the building. A substantial frame work is put in, on which to place the machine, and heavy oak ''lead-out" timbers are placed on top of the foundation on which to holt the rocker shafts, or the ROCKER SHAFT LEAD-OUT. cranks used to run the different connections out from the tower. The arrangement for this purpose which is the easiest to put in but the most expensive to use, is a rocker shaft lead-out as shown in Fig. 10. These should be made with the arms welded on in- stead of being fitted on a hexagonal shaft and fastened by a set screw, as is the case with the one shown, for the reason that the cranks are not accurately fitted as they should be and cannot be held tight by the set screw. As a little lost motion at the center 211 is very much increased at the end of the crank, the travel of the pipe line being reduced by just that amount, it will be seen that this arrangement is not as good as when the arms are welded on, for, owing to the levers being interlocked, all lost motion in the connection is to be avoided. By a judicious use of ordinary and box cranks, many lead- A Box CRANK LEAD-OUT. outs put in with rocker shafts could have been put in much more cheaply and yet with practically the same results. A box crank is shown in Fig. n, and consists of an arrangement whereby any number of cranks are made to work on a single base, and make it possible to turn at a right angle, and in a very small space, a large number of pipe lines. Towers should always be made somewhat larger than is neces- 212 sary to hold the machine to be used, to allow for possible additions to the plant, the expense of so doing being very little in compari- son with what it will cost to enlarge one already built. One very prominent road builds its smallest towers large enough to take a 24-lever machine, thereby providing for the A SPECIAI, TOWER FOR Use IN CITIES. future and also give ample room in which to get at all the parts. For towers that have to be located in places where there is but little room between tracks, for instance or where, as with the new electric street railway interlocking in cities, the tower has to be put upon the sidewalk, some form of iron framework has to be used, as shown in Fig. 12, the pipe connections being boxed in 213 (Fig. 13) to prevent anyone except the proper persons from get- ting at or tampering with them. The location of the tower is an important point and should be carefully considered before a decision regarding it is made. In a general way it may be said that it should be placed at some LEAD-OUT CONNECTIONS FOR SPECIAL TOWER. central point, where the best view of the tracks is to be had, this point being almost always on the outside of the curve. If there is but little choice in this respect, the tower should be located where the straightest, and therefore the simplest, connections can be made to the switches and signals. Among recent additions and improvements to the tower may 2I 4 be mentioned a ladder, to allow of easy access to the roof in case of fire, ?.s is shown in Fig. 9, and the casing surrounding the stove, SIGNAI, TOWER STOVE AND CASING. shown in Fig. 14, which has proved itself to be just what was wanted, as nearly all the heat is carried to the operating room, which is kept warm, while no coal is wasted heating the lower one. 215 It is customary in climates where there is much snow or ice in winter to box the pipe lines, as well as the switch-and-lock movements, to prevent their being clogged by ice forming on the pipe and carriers. The boxing should be made of 2-inch material, the side pieces being 16 feet long, to bring the joints over the pipe- carrier foundations. Where the wire lines to the distant signals are run in cities, or in yards where people are likely to be tripped up by them, they also should be boxed. This boxing should also be made of 2-inch material, pieces 6 inches wide being used for the sides and 8 inches for the top. This is not the usual prac- tice, i -inch stuff being most generally used; but as the 2-inch will outlast the other almost double and will need but little attention or repairs, there can be no question but that it is true economy to use it. In the operation of an interlocking plant, a safety device known as "electric locking" of the levers is now being introduced more than ever before, as the advantages to be had by its use are becoming more generally known. With it the levers of the ma- I chine are electrically locked, so that the operator, after once clear- ing the signal for a train to proceed, is unable to move the levers controlling the derails and switches until the train has cleared the limits of the interlocking. The lock itself is a very simple con- trivance, consisting of an electro magnet, supported on a suitable frame and bolted to the locking brackets of the interlocking ma- | chine, as shown in Fig. 15, so that the armature when down will engage with a lug on the locking bar and prevent it from moving, and when raised by the attraction of the magnet will leave the bar free. A heavy casing is provided to inclose the magnet, so that when locked by the padlock provided for the purpose, the operator will be prevented from getting at the armature and releasing the lever before the train has cleared the interlocking. To drop the locks and lock up the machine, a circuit breaker is attached to all the signal levers, so that whenever a signal is cleared, the circuit through the lock is broken, de-energizing the magnets and dropping the armatures. To prevent the circuit from being restored when the signal is returned to the normal position, which should be done by the time that the last car of a train has passed it, a track circuit is made use of to energize a relay, the armature of which, when down, breaks the circuit 2l6 through the locks in the same manner as the circuit breaker on the signal lever. The circuits made use of, and the manner of connecting them up, are shown in Fig. 16,, T being the track circuit through the rails and energizing the magnet E; F, the locking circuit through EI/ECTRIC LOCKS FOR INTERLOCKING MACHINES. the locks L, the circuit breakers B, the contact points D of the track relay E, the magnet M and the contact points P of that magnet; R, the releasing circuits energizing the magnets N, the armatures of which make a back contact when the magnets are de-energized, and complete the circuit F through the magnet M (but not through the locks), if the track-circuit relay is energized and the circuit breakers closed. The operation of these locks, and the effect on the circuits when a train passes through the in- terlocking, is as follows: The locks being normally held up, the switches are set and signals cleared for the train to proceed. Clearing any of the signals breaks the circuit F at the point B, dropping the locks and locking the derail lever reversed, so that it cannot be changed, at the same time dropping the armature of the magnet M, breaking the circuit at the point P, and prevent- ing the locks from being raised when the signal is returned to the normal position. When the train passes on to the track circuit, it de-energizes the magnet, breaking the circuit F at the point LucunwtU-e nyincermg FIG. 16.. ELECTRIC LOCKING CIRCUITS. D, so that the circuit cannot be restored and the locks lifted, if all the wheels of the train have not passed out of the interlocking. When the train reaches the releasing section, the magnet N is de-energized, the armature falling and completing the circuit F through the magnet M as soon as the circuit is restored at the point D and circuit breakers X. Energizing the magnet M raises the armature and completes the circuit through the point P, so that when the train passes off the releasing section, although the circuit F is broken at the contact point of the magnet N, it is maintained through the contact point P. The locks, however, are not raised until the circuit is broken at the magnet N, for although the circuit was completed through the locks and the point P 218 when the magnet M was energized, it was also complete through the contact points of the magnet N, and as there is less resistance through these than through the locks, most of the current would flow that way, and the locks would not be raised until the con- tact at that point was broken. If the signals have not been restored to danger before the train pases off the releasing section, the circuit F will not be com- pleted through the magnet M and the machine will remain locked up. An arrangement of circuits applicable to a simple crossing, in which the locking circuit is done away with, is shown in Fig. 17. With this arrangement the track circuit inside the derails is made the releasing section and no locking circuit is used, the action being the same as in the previous arrangement, with the magnet E left out. The cost of putting on this arrangement, complete, is not more than $150, and if proper care be used in the installa- tion, it will cost but little to maintain and will seldom get out of order. In case of accidents, or a failure of the circuits to release the locks, a switch is provided by which the operator can close and release the locks, by breaking a glass inserted in the cover of the box in which the switch is placed. Inclosing the switch in this way is done to put a check upon the operator, to prevent him from throwing the locking out of service without good and sufficient reasons. The circuits here shown are the simplest and best for this purpose that have ever been designed, and have given most excellent results since they were put in service. They are much superior to the mechanical or interlocked relays that are used by the signal companies, as the action is positive, there being no trouble with sticking of the armatures, as with the latter, nor is it possible for the operator at any time to release the locks by jarring the relays. The advantages to be derived from locking the levers in this way are : That the operator must return the signal to danger after the passage of every train, before it leaves the limit of the interlocking, thus compelling him to keep the signals at darker and not clear any one of them until on the approach of a train; that as long as any part of a train is within the interlocking, 219 no change can be made in any of the switches that would lead to a collision or derailment an advantage that is very great, considering the number of times that switches are run through by operators throwing the switch in front of a train, thinking EI/ECTRIC LOCKING CIRCUITS APPLICABLE TO A SIMPLE CROSSING. that it has passed out of the interlocking; that having once cleared the signal for a train, it is impossible for the operator to open the derail or change switches so as to cause a derail- ment, although he is perfectly free to return the signal to danger at any time. 220 That this latter is a very desirable feature, is shown by the large number of accidents that have happened by operators taking away the signals from one train and giving them to another, the train from which the signals were taken being derailed, owing to their having approached the crossing ex- pecting to be allowed to proceed. One instance can be named where the operator cleared the signal for an approaching freight and then went to sleep. Hearing the whistle of a train on the other road, and forgetting that he had just cleared the signals for a freight, he changed the levers, opening the derails just in front of the engine. The engineer of the freight having observed the signal at clear, was approaching the crossing at a fair rate of speed, which, as the train was a heavy one, was sufficient to shove the engine and four cars completely over the crossing, entirely blocking it. Had the train for which the signals had been changed not been a passenger, and a light one at that, a collision would have happened, it having come to a stop within a car length only of the other train. Then, again, with the electrical locking, there can be no ques- tion as to which train the signal had been given, and the claim so often made by engineers when they get into trouble, that the signals were taken away from them, will not hold with a plant so equipped. To those who fail to see the advantage of this and the great help that it gives in enforcing discipline, I would refer them to the superintendents who are able to send their engineers to plants so equipped, with the request that they come back and tell them if they were able to change the signals and switches, as they claim was done to them, and that if they could do so they would be given full pay for the time they were off. 'CHAPTER XIIL THE WESTINGHOUSE ELECTRO-PNEUMATIC AND THE GIBBS ELECTRIC STREET RAILWAY SYSTEMS. Many attempts have been and are now being made to operate the switches and signals of an interlocking plant by some power other than that of a human being, invention having passed successively from the first pneumatic machine put in service in 1876, to a hydraulic machine used in 1880, to a combination of these or a hydro-pneumatic machine in 1884 and to the electro- pneumatic in 1891. Of these, the electro-pneumatic is the only one that can be considered a success, the few plants in this country that are now operated by means of air, water or elec- tricity having as yet not been in service long enough to demon- strate that they are anything more than an experiment. As the several inventions along the lines named have been made principally by Mr. Geo. Westinghouse, Jr., the system that is in use to-day is one that bears his name. With the electro- pneumatic system the power to move a switch or clear a signal is obtained by the use of compressed air, the action of this power being controlled by a valve worked by an electro-magnet, the electric current to energize the magnet being in turn controlled by an interlocking machine having switches by which the circuits through the several electro-magnets can be completed. As the connections to the switch and the signal instruments are of pipe for carrying the compressed air, and insulated wire for conducting the electric current, there is no complication of parts, as with a mechanical plant, the only apparatus used outside of the tower being such as is required at the switch or signal. For the same reason there is no limit to the distance that the different movements can be placed from the tower, other than what it would be safe for a man in the tower to operate. A 222 description of the valve and cylinder for operating a semaphore signal has already been given in the article on "Automatic Electric Block Signals," that appeared in the May issue of "Locomotive Engineering," the same arrangement being used for an interlocking plant. To briefly describe the action of the valve: When the lever of the interlocking machine which works the electric switch is turned, the circuit through the magnet of the signal instrument is closed, energizing the magnet, causing it to attract the arma- ture and open the valve admitting air to the cylinder, at the same time closing the exhaust passage by which the compressed air is permitted to escape. As soon as the compressed air is ad- mitted to the cylinder, the piston is forced through the length of its stroke, the movement being transmitted by means of a balance lever and connecting rod to the signal, which is moved to the position indicating safety. When the operating lever is returned to its normal position, and the circuit broken, the magnet loses its power, allowing the valve to be pressed up by the force of a spring, the passage by which air is admitted to the cylinder being closed and the one releasing the air being opened, the piston in consequence being pressed up and the signal returned to danger by the force of gravity. The arrangement used for operating a switch is shown in Fig. i, and consists of the switch-and-lock movement used with a mechanical plant and a cylinder, the piston of which is made to work the movement the same as if it were connected to the lever of an interlocking machine. The connections to the switch points, the lock bar and the detector bar being made in the same way as with a mechanically operated plant, the same cer- tainty of action and protection is afforded as if it were operated by mechanical connections instead of by compressed air. The construction of the valve and cylinder by which the necessary movement of the piston is obtained, is shown in Fig. 2. The valve used is of the ordinary slide-valve pattern, passages to the cylinder and to the exhaust being arranged in the same way as on a locomotive. To work the valve, two small pistons are fastened to each end of a yoke, which is made to fit over the valve, the small cylinders in which these pistons work being known as valve cylinders. Admission of air to the valve cylinder 223 to move the piston, and with it the valve, is controlled by a magnet of the same construction as the signal magnet, air being admitted behind the piston when the magnet is energized by the electric current sent out from the interlocking machine. As a check upon the performance of the admission valve, a lock pin is provided which fits in a socket in the back of the valve and locks it in either of the two positions it should occupy. This locking pin is riveted to a piston which is held down by means of a coiled spring placed on the opposite side ELECTRO-PNEUMATIC SWITCH AND LOCK MOVEMENT. from the valve. To lift the lock pin and release the valve so that it can be moved, a magnet is provided, the armature of which, when attracted, is made to open an exhaust passage to the atmos- phere and allow the air on top of the lock piston to escape. The pressure of the air on the other side of the piston over- comes the pressure of the spring, raising the piston and with it the lock pin, thus leaving the valve free to be moved. When the current through the lock magnet ceases, the armature is released, closing the exhaust passage and allowing air to ac- cumulate on that side of the piston, once more restoring the 224 equilibrium, when the pressure of the spring will force the piston back to its normal position and lock the valve. The operation of the valve admitting air to the cylinder, by which motion is imparted to the piston in the cylinder and the switch-and-lock movement worked, is as follows, the lever of the interlocking machine being in the normal position with the circuit closed through one of the valve magnets: With the first DETAIL, OF VAI/VE AND CYLINDER. movement of the lever, an electric circuit is formed through the lock magnet, which, when energized, opens the lock exhaust and permits the pressure of the air to raise the piston, and with it the lock pin from its seat. A further movement of the lever breaks the circuit through one of the valve magnets, permitting the air to escape from the cylinder controlled by that valve, and energizes the other valve magnet, admitting air to the valve 225 cylinder. This forces the valve piston, and with it the valve, to the other position in which air is admitted to one end of the cylinder and opened to the exhaust at the other, the compressed air forcing the piston through the cylinder and performing the movements desired. When the movement of the lever is completed, the circuit through the lock magnet is broken and the magnet de-energized : the armature, being released, closes the exhaust passage, al- lowing the coiled spring to force the piston back, seating the pin upon the other side of the valve and locking it, until the whole process is repeated in the other direction. In this way the valve is locked open, to one end of the cylinder or the other, at all times, the pressure being kept upon the piston to prevent any accidental movement of the switch. The construction of the interlocking machine is very different from that of the machines used in operating a me- chanical plant, as the levers, instead of transmitting the mechan- ical force necessary to work the different parts, have only to change the several electric switches controlling the movement that it is desired to make. The general appearance of the ma- chine is shown in Fig. 3, the upper row of levers, called the switch levers, being used to make and break the circuits that connect the main battery with the switch valves, the lower levers being called signal levers, as they are used to make and break the circuits to the signal valves. Each of these levers is attached to a shaft carrying a rubber roller, on which are fastened small brass strips with which the connections are made between the two springs bearing against the roller, when the roller is turned to the proper position, the circuit between the two springs which are the two poles of the circuit being thus completed. By arranging these strips so that the several circuits are com- pleted in the proper order, the movement desired is made by merely turning the lever. So also, if all the levers are not in the proper position to safely perform the movement, contacts would not be made on the roller of the lever improperly set and no current would be sent out from the machine, the levers being thus electrically interlocked. To prevent the levers controlling conflicting routes from being reversed (if such it may be called) at the same time, 226 locking bars driven by mitre gears on the shaft are arranged, by which it is made impossible to turn the levers that clear any two routes that may lead to a collision. The locking is of the im- proved Saxby & Farmer type, arranged in a manner similar to that shown with the mechanical machine. As a still further check, there is placed at each signal and switch movement a circuit breaker, from which wires are run to magnets placed on the machine, the armatures of which are ELECTRO-PNEUMATIC INTERLOCKING MACHINE. made to operate latches controlling the levers, so as to delay the completion of the movement until after the signal or switch has completed its movement, the object of this being to prevent any other lever from being moved until the switch or signal that is being operated has completed its movement; for until the lever has been moved to its extreme position, the locking will not permit any other lever to be moved. The electric switch bv which the indication is sent to the machine, that the move- 227 ment of the switch has been completed, is to be seen placed on top of the switch-and-lock movement, a photograph showing its construction being shown in Fig. 4. It is thus seen that a double check is had upon all the move- ments of the levers, and that the possibility of a mistake being made by an ignorant or careless operator is well guarded against. And, to quote from a recent technical paper, "the sequence of movements, by virtue of which a clear signal cannot be given INDICATION SWITCH Box. until the route has been prepared for it by setting the switches in their proper position, is absolutely secured by the order in which the several electrical circuits are closed." With this system, owing to the amount of work the magnet has to perform, the current has to be of a comparatively high voltage, and as the current is used at all times, whether a move- ment is being made or not, it has been found necessary, or rather more economical, to generate the electricity by means of a dynamo, more especially as a power plant has to be pro- vided to furnish the compressH air. It is also customary to 228 use storage batteries, keeping them charged, so that in case of a shut-down of the dynamo, the batteries will furnish the current to operate the plant. At small plants where the current is generated simply for the interlocking, it is usual to run the dynamo only during the day the storage batteries that were charged when the dynamo was running, furnishing the necessary current at night. In the operation of an electro-pneumatic plant, the facility with which the different movements are made reduces, to a great extent, the number of men required to do the work. While in most cases one man is- required to move the levers, another as a train director, and other men have to be em- ployed to run the engine, the number required at a mechanical plant having one hundred or more levers, is so large that the cost of operation of the electro-pneumatic is not any greater and, in many cases, is much less. More especially is this the case where, as at terminals, a plant for generating electricity has been put in for other purposes, the only expense then being the cost of the additional power expended to provide the current used to work the plant and for furnishing compressed air. Owing to there being no mechanical connections from the machine in the tower to each of the different movements, this system lends itself most readily to all applications \vhere there are many and complicated sets of switches, sharp curves, or to places where it would be almost impossible to operate a me- chanically connected plant. The connections from the tower being of wire only, the tower can always be placed in the most advantageous position for controlling the movements of trains, the arrangement shown in Fig. 5 being a good example of what it is possible to accomplish in this direction. At locations such as are found at large terminals or freight yards, where the number of movements to be made is large and the number of switches and signals to be worked is very great, the dispatch and safety with which every movement can be made warrants the use of such a system, even if the first cost is greater. At all such locations it is usual to equip the tower with such electrical devices as will aid the train director in his work, enabling him to keep himself informed as to the movements of trains, the condition of the tracks and the position 229 of the signals. The view shown in Fig. 6 well illustrates this, being a photograph of the instruments in a tower on the Pennsylvania road, and kindly furnished by the Union Switch & Signal Co. The several instruments shown comprise train describers, by which information is signaled as to what train is coming; TOWER HOUSE FOR ELECTRO PNEUMATIC INTERLOCKING MACHINE. disk and semaphore indicators to inform the operator when a track is occupied ; drop annunciators to give information regard- ing the starting of trains on certain tracks; telegraph instru- ments, telephones and electric bells. In addition to these a miniature model of the tracks operated is provided as a part of each interlocking machine, the switches being movable and mechanically connected to the levers, so that the model will accurately represent the position of the switches on the ground. 230 This is a great help in the operation of the machine, as the operator can see at a glance the position of all the switches, and not have to look at each lever to see what position it is in. Incandescent electric lamps can also be used with great success, instead of oil, at an electro-pneumatic plant; for while ELECTRICAL ANNUNCIATOR AND INDICATOR INSTRUMENTS IN P. R. R. TOWER. the cost will be about the same, the service will be very much better. The cost of an electro-pneumatic plant for small installations is greatly in excess of one mechanically connected, but when the number of levers is large, or the tracks very complicated, the cost will, at most, be the same and may be somewhat less. As there is hardly any limit to the number of movements that 231 can be put upon a single lever, a very much smaller machine can be used, the cost in consequence being very much reduced. As an instance of this kind, at the Stewart avenue plant, Chicago, there are but 48 working levers in the machine, oc- cupying a space 5 x 24 feet, to work 84 signals, 37 switches, 22 double slips and 22 movable frogs; while with a mechanical machine, according to American practice, 187 working levers, occupying a space 14 x 77 feet and to English practice, 243 levers, occupying a space 17x93 feet would be required. If the plant has been properly installed, the repairs will be no greater than with a mechanical plant and it will be much easier to keep in working order. In winter, trouble will be experienced from freezing up of the valves and pipes, unless proper precautions are taken to get rid of the water of con- densation formed in compressing the air. The contacts at the different movements must also be kept free from snow and ice, or else a contact will not be made, and it will not be possible to entirely reverse the lever. An interlocking device which has come into use in the last two years, for the protection of a crossing of an electric road with a steam or with another electric road, and which promises, with the rapid increase in the use of electricity for street railroads, and their extension consequent upon the same, to be quite ex-, tensively introduced, has been patented by Mr. George Gibbs, Mechanical Engineer of the Chicago, Milwaukee & St. Paul Railway. At any street crossing, when a change is made to electricity as a motive power, the danger of using the crossing is very much enhanced, owing to the increased speed with which the the cars approach the crossing, their great weight and con- sequent inertia, and the liability of failure of their source of power, due to a shut-down at the power house, blowing out of fuses, or jumping off of the trolley wheel when the car is on the crossing. "There seems to be no good reason," to use the words of Mr. Charles Hansel, C. K, "why the statute of the State of Illinois, which requires all new crossings of steam rail- roads at the same level to be protected by a suitable system of signals, derails, etc., should not include the crossings at grade of all railroads or railways which carry human freight, for it 232 cannot be regarded as unreasonable that street railways should comply with the requirements of public safety in the same manner and in the same measure as is required of steam roads. "In cities where such grade crossings occur, the police regula- tions generally require the railroads to keep a flagman posted at the crossing to signal traffic. This practice gives but a small measure of protection, while the charge to the railroads for operating is the same as if this flagman had physical control of the crossing. Considering the subject from a financial point of view only, it appears that if we can construct a system of signaling to be controlled by a single man in a tower overlooking the cross- ing, with mechanism so arranged as to make it impossible for trains or cars on the steam railroad and the electric railway to reach the crossing at the same time, we have invested well; for with such a device the danger of crossing is eliminated, and the operation of either line is the same as regards safety as if no such crossing existed, and both roads are relieved of the constant danger to life and property and consequent payment of damages. The saving in rates of insurance to the electric line where such protection is provided, is alone sufficient to pay the fixed charges on the investment." The principal features of the apparatus used by Mr. Gibbs are the cutting off of the current from an insulated section of the trolley line, making it impossible to move cars when it is desired to block the line, and in the use of a derailing device to be operated in paved streets, to prevent cars from rolling on or over the crossing, when the current is cut off from the trolley. These devices are connected to and operated by any of the usual forms of interlocking machines, the derailing device or "scotch block" being connected to the same lever that works the switch and cuts the current off from the trolley wire. A general plan of the arrangement, as applied to a double- track crossing of a steam and electric road, is shown in Fig. 7, the trolley lines in which sections about 800 to 1,000 feet, according to grade, are insulated, being shown in the center of the tracks of the electric road. These insulated sections extend to within about 50 feet of the crossing, but never over it, and in this way make it impossible for the current to be taken away from a car at a place where it would be liable to be hit by a train on the other 233 road. A feed wire connects the insulated section with the electric switches on the machine, which, in turn, are connected to the supply wire from the power house, the insulated section, when the switch is closed by the reversal of the lever of the interlocking machine, becoming a "live" wire and delivering current to any car that may be in the section. The construction of the scotch block is shown in Fig. 8. It PLAN SHOWING WIRING, GIBBS ELECTRIC STREET WAY INTERLOCKING. consists of a strong cast-iron box, which is placed outside of the rails, a groove being cast in the box in which an iron block 2x4 inches in size is raised or lowered whenever the cam lever is moved by the lever of the interlocking machine to which it is con- nected. When the block is down the top is flush with the top of the box, but when it has been raised, owing to the line of motion being inclined to the axis of the rail, it will project out from the box and over the head of the rail. The block is raised and low- ered by means of a roller and pin working in the slot of the cam, 2.34 but when in either of its extreme positions a solid bearing almost the size of the block is made on the cam and directly in line with the axis of the cam and the cam shaft. In this way the thrust of a wheel, whether of a car or heavy wagon, will be taken by the shaft and not transmitted to the pipe line and lever. The block, when raised, projects about 4 inches above the top of the rail, sufficient, it has been found, to either derail the car or to practi- cally bring it to a stop. The block does not project entirely over the head of the rail, and so does not interfere with wagons or other vehicles which are continually running in the flangeway of the street road. This is an important feature in its design and one that has a great deal to do with the success of its application. FIG. 8. LOCOMOTIVE e/va//veewi/va. N. Y. DETAILS OF " SCOTCH" BLOCK. In Fig. 9 are shown the electric switches and the connections to the levers of the machines by which they are worked. The levers to which they are connected are the ones that work the scotch blocks, the current being always shut off from the insulated section of trolley wire when the block is raised. Placed on each side of the two electric switches, relays are to be seen through which the current to each insulated trolley section is made to pass. Either of these relays, when energized by a car in the insulated section, is made to complete a local circuit and work a bell to inform the operator when a car has entered the section. A small electric switch worked by the signal lever breaks this local circuit and stops the bell as soon as the signal is cleared, the operator's attention, in this way, being more forcibly attracted to the fact that the signal had not been cleared. It will be noticed, by examining the figure, that a large three- 235 way reversible switch, to which the supply wire is run before being connected to the two cut-out switches, is placed on the frame-work immediately below the switches. This is used to cut out the switches and relays, in case any accidental short circuit FIG. 9. ELECTRIC CUT-OUT SWITCHES AND CONNECTIONS TO IN- TERLOCKING MACHINE. should be made, or any of the parts get broken, the trolley wire being in this way energized so as not to block the electric road. An additional feature which experience has demonstrated to be of great value in the safe operation of the plant, and one which has also been patented by Mr. Gibbs, is an arrangement of locks and releasing circuits whereby the levers of the machine are locked up after the signal has once been cleared for a car on the 236 electric road until after the car has passed over the crossing. The circuits used are shown in Fig. 10, the locks, relays and circuit breakers being connected in the same manner as was shown in a previous article, where electric locking was applied to the levers controlling a simple crossing, except that when the releasing relay (the upper one shown in the figure) is energized the current is made to pass through the locks, releasing them as soon as the circuit is completed instead of around the locks in the shunt H -Ml' I' I* LOCOMOTIVE ettaiHee*iNo. N, fe ELECTRIC LOCKING CIRCUITS FOR ONE TRACK ONLY OF A DOUBLE TRACK ELECTRIC RAILWAY INTERLOCKING. A, supply wire ; B, wire to insulated section of trolley ; C, releasing relay ; /?, cir- cuit breakers ; E, locking relay ; F, hand releasing switch ; G, locks ; H, battery. circuit, when they would not release the levers until after the releasing circuit had been broken. To obtain a current with which to energize the releasing relay, a short section of the trolley wire on the further end of the crossing is insulated in the same manner as the^ cut-out section and is connected to the supply wire through the coils of the releasing relay. When the car has passed over the crossing and the trolley runs on the insulated section, the relay is energized, lifting its armature and completing the circuit through the locks, lifting 237 them and also restoring the circuit through the locking relay, causing the circuit to be maintained and the locks held up after r-mm LOCOMOTIVE ENGINEERING. N . Y. RELAY FOR HEAVY CURRENTS USED BY MOTOR CARS. A, core wound with two stands No. 6 double cotton -covered magnet wire wrapped with Okonite taping ^ inch thick and paper well shellacked. the car has passed off the releasing section. As the current which passes through the releasing relay is the one used to propel the car, it is at times a very heavy one, amounting to seventy-five SIGNALS FOR STREET CARS. GIBBS ELECTRIC STREET RAILWAY INTERLOCKING. SCOTCH BLOCK. 239 amperes at 500 volts, or about 50 horse-power. A relay to carry such a current as this is constructed of very different proportions from the relays ordinarily used for the currents generated by a battery, and, for this reason, its design may be of interest to those who are not familiar with such things. This relay is shown in Fig. n, the core pieces being i inch in diameter and 4 inches long, arranged in the usual horseshoe form and wound with a single layer of two No. 6 cotton-covered magnet wires, which are soldered together where they enter the binding posts. The iron core pieces are covered with -J-inch insulating material to prevent the current, the voltage of which is very greatly increased during thunder storms, from breaking through the insulation, and by reaching the locking circuits, burn- ing out the locks and relays. To prevent, if possible, any damage being done in case the insulation should break down, the wires of the locking circuits are connected to the relay with a piece of J-ampere fuse wire, 6 inches long, which will burn out and prevent the current from grounding through the locks. This system is now in use at twenty crossings of electric with steam railroads in the city of Chicago and one in the State of New York, and is destined to have a greater number of applica- tions as soon as the benefits to be derived from its use become more generally known. It is practically, in its workings, as safe an appliance as the interlocking used at a crossing of two steam roads, and insures that the cars of the electric road will have the same protection afforded them in using the crossing that is given to the trains on the other road. CHAPTER XIV. AGREEMENTS, CONTRACTS, SPECIFICA- TIONS, INSTALLATION AND REPAIRS. In the early days of railroading, before many roads were built and while the rights of each were in process of develop- ment, it appeared a just and reasonable thing for one road to grant any road the privilege of crossing its tracks whenever such a thing was desired it being understood that the road making the crossing would be responsible for the maintenance of the same. As business increased and the demand for other tracks arose, the question immediately presented itself as to who should pay for the additional crossings made necessary thereby, each road very generously desiring to make the other road stand the entire cost. This led to the roads interested drawing up an agreement, in which it was clearly stated who should maintain the crossing and what proportion of the cost each road should bear when any new work was to be put in. The expediency of so doing having been established, any roa,d desiring to cross the tracks of another was forced to sign an agreement whereby they were to put in and forever maintain the crossing at their own expense, and, furthermore, would put in and maintain any crossings made necessary by the laying of new tracks by the road first built. The protection afforded at a crossing by the use of an interlocking plant having, in comparatively late years only, attracted the attention of railroad officers, no mention was made in the earlier crossing agreements of any appliances to make the use of a crossing safe, a flagman to signal trains being all that was thought necessary at even the busiest crossings. In these cases, when it was desired to install an interlocking plant, a new agreement had to be drawn up, covering the proportion of the original cost which each road should bear, and what 242 proportion each should pay of the cost of operation and main- tenance, one road or the other taking charge of the installation, operation and maintenance and billing against the other for their proportion of the cost. At ' the present day, when one road desires to cross the tracks of another, the requirement is generally made that an interlocking plant shall be put in and maintained by the road wishing the crossing, as well as putting in and keeping up the crossing frogs, ties, etc., whether or not the business over either of the tracks will be such as to warrant the additional expense. But as laws have been passed in several of the States requiring all new crossings to be protected, and as it is in most cases only a question of a few years before an interlocking would be needed, it is certainly good policy for any road in drawing up an agree- ment allowing another road to cross its tracks, to include an interlocking plant in its requirements, stating what form of apparatus will be required. Such an agreement should prac- tically be an "iron clad" affair, covering every point, of which the following list comprised the main items called for in an agree- ment lately made and which are given here for the information of those who may not be familiar with what may be required. First. That the party of the first part, or the road first built, shall have the use of the tracks now owned and operated by it without any material impairment of their usefulness or safety by the party of the second part, or the road desiring to put in the crossing. Second. That all crossings which it is the desire of the first party to construct, maintain and operate over the tracks of the second party shall be furnished and properly put in by the party of the second part. Third. That the second party will furnish all the materials for and construct and put in all crossing frogs, crossing signals, gates, targets and other fixtures, according to plans and specifi- cations furnished by the first party. All parts to be forever maintained and kept in good repair at the sole cost of the second party. And in case of failure to promptly furnish and put in or keep in repair any of these parts, the party of the first part may do the work and bill against the party of the second part for the full amount so expended. Any damages resulting from 243 defective condition of said parts being paid for by the party of the second part, saving the party of the first part harmless there- from. Fourth. In the passage of the respective trains of the parties interested over the crossing, the passenger trains of the party of the first part shall have preference over the passenger trains of the party of the second part, and in like manner the freight trains' of the party of the first part shall have preference ; but in all cases passenger trains shall have preference over freight trains of either road. Fifth. If at any time a difference of opinion between said parties shall arise, the question in dispute shall be referred to a board of arbitration consisting of three competent disinterested parties, one to be chosen by each of the parties to the agreement, and the two so chosen to choose a third. That written notice shall be given of the time and place of the meeting, and that at the time and place appointed they shall proceed summarily to hear and dispose of the matter in dispute, the determination of such board of arbitration being absolutely final and conclusive upon the parties interested. Sixth. That within ninety days from the date of these presents the party of the second part will provide said crossing with an interlocking plant, with pipe home signal connections, electric locking, and annunciators, and if at any time a device satisfactory to the chief engineer of the party of the first part be manu- factured for the purpose of giving a continuous rail over the crossing on the line having the right of way over the same, it shall be put in and connected with the interlocking plant. The specification, locking, dog-sheet and general plan of the inter- locking shall be submitted to and approved by the party of the first part before the contract shall be executed. That the party of the second part shall bear the entire cost of such interlocking plant, operating and maintaining it at their own expense. That any additional tracks laid by either party are to be connected to the interlocking and maintained at the sole cost of the party of the second part, and that the party of the first part may take charge of and maintain and operate the plant, if the operation and maintenance by the party of the second part be not satis- factory to the party of the first part, the party of the second part 244 paying all bills upon presentation for the amount expended in maintaining and operating said interlocking plant. While the above conditions comprise the most important points generally considered in an agreement between two steam roads, there are several other points w r hich have to be included when the agreement is entered into by parties controlling a steam railroad with those controlling a line operated by elec- tricity. Briefly, these are, that all overhead wires shall be main- tained at a height of twenty-three feet above the tracks of the steam road, conductors to be arranged for the return electric current, so as to prevent, as far as possible, leakage from its tracks, that will affect the operation of electricallly controlled railway signals, telegraph or telephone wires. That the over- head electric wires shall be so arranged as not to interfere in any way with the operation of the street gates, which the steam road is compelled by city ordinance to maintain at such points of crossing. The agreement having been signed by the parties interested, the next step is to prepare the plans and specifications for the interlocking and submit them to the signal companies for their bids. The plans submitted by the railroad company, when drawn up by the signal engineer, usually consist of a plan of the tracks with all the derails, switches and signals required, properly shown and numbered, the location of each and the distance from the crossing being clearly stated. Lines repre- senting the connections to the switches and signals are also drawn, as well as lines to show where any heavy boxing is to be put in and which it is desired to have the signal company include in their bid. Where no signal experts are employed by the railroad company, a plan of the tracks is usually submitted to the signal companies with a request that they draw up a plan showing the proper signals to be used and how the several con- nections should be run, their bid being made upon the plan so drawn up. Their bids are usually made upon specifications gotten up by them for the kind of apparatus which they manufacture, it being understood that the work is to be put in to the satisfaction of the railroad company contracting for the same. These speci- fications usually state the kind of machine, the number of 245 levers and the number of switches and signals to be operated. They also give the size of the tower and of what the signal connections shall be made, practice in this regard being different on the different roads. The switch connections are given in detail, the size of pipe, method of fastening and means f compensation being also given. The distances apart that the pipe carriers and wire pulleys are to be spaced is stated, as well as the sizes of the different foundations and the thickness of lumber to be used in the boxing. The part of the work that the railroad company is expected to do is plainly stated, comprising, for the most part, the track work in preparing the switches, derails and movable frogs ready to be connected. All preliminary grading necessary to be done and proper drainage wherever required. To furnish broken stone, sand and cement for concreting the heavier foundations, and to provide permits for building the towers and for digging across streets when necessary in cities. Railroad companies having a signal department usually submit with the plans to the signal company for their bid, any specifications in regard to details that they may wish to have followed when the work is put in. These generally relate to standards of the railroad not called for by the specifications of the signal company, or where some apparatus is to be used of a different design from that manufactured and furnished by the company bidding on the work. . The bids having been submitted and a selection made, the contracts are signed and the work commenced by the signal company, the railroad furnishing the material and doing the work that it was agreed in the contract they should do. The signal expert appointed by the railroad to inspect the work goes over the ground with the signal company's foreman and gives the exact location of the tower, the derails, signals and where the pipe lines shall pass under the tracks. The different ways of doing work are discussed and an understanding arrived at as to what will be considered good work and what the railroad will require, it being specified in the contract that the work shall be done to the satisfaction of the railroad company. As the work proceeds, points in regard to construction will, from time to time, come up which the signal inspector will be 246 expected to settle, to prevent a possible rejection on his part when the plant has been completed. Inspections will have to be made quite frequently to see that the work is being properly done and that the standards of the railroad are being followed. When railroads have a regularly organized signal department, they often install a plant with their own men, buying the material from one of the signal companies, as their experience has been that the work will, at least, be better done, even if it should cost no less. After the plant has been completed and before it is put in service, each lever should be connected to the switch or signal it is to throw, and a trial made to see that every part works properly. To do this without interrupting traffic only one switch, or the switches to be worked by one lever, should be connected at the same time, those to the arm plate casting of the signal being left connected, as until the blade is bolted up no indication is made. Should the work prove satisfactory, a day is appointed on which the plant will be put in service, this being a day or so before the signal company will have finished the boxing, painting and other work that can be completed after the plant is ready to be connected up and put in service. Notice of the fact is given to all the railroad companies interested, so that proper bulletins can be issued notifying train- men that the interlocking will be put in service at such a time and that they must be careful to obey the signals in running over the crossing. In connecting up the plant the signal blades should first be bolted on, this being done so that all the blades will be put up at the time appointed for the plant to go into service. After the time has passed, but not before, the derails may be connected up, the trainmen then being responsible for the consequences should they allow their train to run past a signal when at danger. With States that require an inspection to be made by the railroad commissioners, or some one appointed by them, before a permit will be granted allowing trains to proceed over a crossing without stopping, an appointment with the Commission must be made, when they will inspect the plant, blue prints of the general plan, locking and dog sheets being sent in for exam- 247 ination and approval before the inspection is made. Should the work have been done as they think it ought to be to make the operation of the plant safe, a day is usually named in which the plant may be put in service, the permit when received reading as having been issued on that day. Should the locking of the levers not meet their approval, or some part be left undone, in their judgment essential to the safe operation of the plant, a permit is refused, and, when the changes or additions desired have been completed, another inspection has to be made. For this reason, when a Railroad Commission has to approve the plans and inspect the interlocking, it is a very good scheme to submit the plans to them before the work is proceeded with, so that no delay will be caused by any objection on their part. If some other interlocking plant is already in service on that division, a bulletin notice is all that need be issued to trainmen, informing them that the interlocking will be put in service. But if it is the first plant to be installed, a set of rules should be issued to the men and an examination made to see that they know how to read the signals and understand what the consequences will be if they disregard them. These rules should describe a signal and explain what the different indications mean; they should make it clear as to which signal will be used to govern a given track, and what will be the several duties of the train- men when using the tracks protected by the interlocking. After a plant has been accepted by the railroad and put in operation, they are then responsible for its condition and will have to see that it is kept in good working order, repairmen and inspectors being kept for this purpose. The organization of the force differs on almost every road, each one believing that theirs is the best and that the others are either not keeping up the plants as they should be, or else are doing it at a very much greater cost. Several roads employ men as inspectors who are responsible for the proper condition of the plants within a certain district, other men being employed to do the work and report to them. Each repairman is given so many plants to take care of, he being expected to go and look them over as often as possible, making any adjustments or slight repairs that may be needed. Should 248 there be no repairs to make, he is to put in his time cleaning up the plant, or, at any rate, looking for work. This way of doing things is, in the long run, a somewhat ex- pensive one, as in most instances there is no work for the repair- man to do, or rather, none but what in the majority of cases the signalman can look after. The man's time has to be charged to repairs, whether there is any work to be done or not. In case of any large amount of repairs being needed, additional help has to be employed, as one man is not able to do very much by himself, and while the claim may be made that by keeping a man on the ground almost all the time the plant will not run down suffi- ciently to need any very general repairs, experience and reason will, I think, prove that this is not so, and that no matter how well this man may attend to his duties, it is not practicable for him to keep the plant in first-class condition, or, in other words, from wearing out. Another plan is to divide the road into districts, making each one so large that the repairman is able to get around only occasion- ally, making inspections and light repairs only and having help furnished him when any large amount of work is to be done. At some central point an additional force is employed, which is used in putting in any new work, making general repairs, or any repairs needed in case of accident. These men should all be experienced workmen, one or two of them being capable of taking charge of or putting their hands to any work that may turn up. They should be able to do a good job in pipe fitting, machine work, blacksmithing, carpentry, cleaning batteries and adjusting relays, and while such men are not to be found every day, if the signal engineer will see to it that only good men are employed and that they are given a chance to learn, they will become very efficient and be able to do good work in any of the lines enumerated. Of course, at very large plants, one, or perhaps two repairmen will be needed at all times, their presence being made necessary more as a protection in case of accident, to keep the road from being blocked, than because of the amount of work there is to do. Another plan, again, is to have practically no regular repair force, but only such men as are needed for new work and general repairs. On the road following this plan, the signal engineer has the hiring of one of the operators, who will be responsible for 249 the maintenance of the plant and must make all necessary repairs and adjustments while attending to his other duties. This man must have had sufficient experience, before being appointed to a position, to enable him to make any ordinary repairs, and being in a position to reap the benefits from keeping the plant in good condition, is very apt to do so, provided he can find the time. And right here is where the greatest objection to the plan can be made, for, if the road at that point is a very busy one, then the man will certainly not be able to leave the tower long enough to do anything more than to change the adjustment of the con- nections, or to clean and oil the different parts when necessary. Of the three ways spoken of, the last is much the cheapest, but if the chances of accident and possible delay to trains where this plan is followed be taken into consideration, the conclusion must be arrived at that it is economizing in the matter of labor at the expense of safety. Where a road has in service a number of block signals, interlocking plants and switch signals, sufficient to keep a regular repair force employed, such as is outlined in the second plan, it will be found by far the best and easiest method of having any new work done or repairs quickly made. The cost will be but little, if any, more than with the last plan outlined above, and the equipment will be kept in better condition. In regard to the railroad employing their own men and putting in their own work, I think that, as a rule, this is not a good plan, for unless the railroad has as good foremen as the signal companies employ, and they seldom have, the work will be more cheaply done by the signal company. The claim that the work will be much better done when put in by the railroad will not count for much, if a competent inspector is appointed to supervise the in- stallation, as the signal companies are certainly willing and try to put in the work to the entire satisfaction of the parties having the work done. Their reputation is at stake, and while they may not do as good work when no one is sent to supervise the installation, they can hardly be blamed for this, as they have not to take care of a plant after it is finished, and by doing new work only, do not find out the w r eak points in their work. Unless a railroad will employ the services of a good foreman, it will, I venture to say, cost them more in putting in an interlock- ing plant than if the work was done by one of the signal com- 250 panics. The chances of a man's doing the work wrong and having to go over it a second time are so great, and the work is so often done in the country, where the men have of necessity to be left very much to themselves, that unless the man in charge not only knows how to do the work, but is capable of properly handling men, the work will cost very much more than was expected. As a matter of protection, in case of accident or break- age of the apparatus, every signalman should be taught how to disconnect and spike a derail or switch, to make adjustments, or any repairs that may be needed to make it safe for trains to pass through the interlocking. A catalogue of the apparatus manu- factured by the signal company installing the plant will be found an excellent book to put in each tower, as the signalmen will then be able to make themselves familiar with the correct name of each FIG. i. part, and, in case of accident, to give page and order number of any new pieces that may be wanted, in this way insuring that the right parts will be sent him. To enable the signalman to make temporary repairs, he must be provided with a set of tools, of which the following is a list of those commonly furnished, the articles needed in furnishing a tower when first put in service being also given: One machinist's hammer, one spike maul, one hand axe, one hand saw, one claw- bar, one 12-inch combination monkey wrench (Fig. i), one socket wrench for f-inch bolts, one pair lo-inch Button's pliers, two cold chisels, one fine file, six sheets emery cloth, one long-neck oil can, one short-neck oil can, one squirt can, one white hand lantern, one red hand lantern, two red flags, one telegraph table with drawer, one office chair and cushion, one bracket lamp, one cor- rugated rubber mat the length of the machine, one Seth Thomas eight-day clock, six fire buckets, six hand grenades, one water bucket, one tin drinking cup, one broom, one mop, one coal hod, shovel and poker. The number of fire buckets and hand grenades in this list may be surprising, as one would hardly think there would be much danger of a tower catching fire when as isolated as they usually are. But the fact is they catch fire very easily, from matches care- FIG. 2. \ 1 D-D B A jj s ill u. LOCOMOTIVE ENGINEERING. N. Y. COAL AND OIL HOUSE, C., M. & ST. P. RY. A, Coal-room, floor plank laid on cinders ; B, Oil-room, earth floor ; C, Zinc-lined shelf. lessly thrown away, from blazing grass and from many other causes, and, from the fact of their being isolated, generally burn down entirely after once catching fire, the interlocking machine being destroyed, causing delay and great inconvenience in the handling of trains. To reduce the chance of fire no lamps should be allowed in the tower, nor should any greasy waste be allowed to lie around or accumulate. The lamps should be cleaned in the day time, in a house provided for the purpose the design shown in Fig. 2 being that of the coal and oil house supplied to all of the interlocking plants of the Chicago, Milwaukee & St. Paul 252 railway the house being placed not less than one hundred feet from the tower. For the guidance of the signalman, the following set of rules, signed by the general superintendents of the roads interested, is framed and put in all the towers of the St. Paul Company, the man in charge being held strictly accountable for the proper observance of every one of them: I Trainmen are instructed to obey the directions and signals of the signalman; you will, therefore, see that trains are passed without delay or stoppage when it is known to be safe to do so. 2. Precedence on conflicting routes will be determined by time cards. A delayed train must not be given the line running on the time of an opposing regular train. 3. The normal position of signals is at danger; of derailing switches, open; of levers, thrown ahead, where they must always remain when no train movements are being made'. Each signal in succession must be thrown to its normal position as soon as rear of train has passed it. 4. When train gives notice of its approach, set switches and signals for the desired route; but be sure that no obstacles exist on route before setting signal for it. 5. When a signal has once been given for any train, should it be necessary to change the position of signals or switches, the signal may be changed to danger, but the switches must not be changed nor the signal given to another train on an opposing route until the train which first had the signals has come to a full stop. 6. No switching which requires blocking the main track must be allowed within five minutes of the time of any regular train. 7. Levers used in switching must be returned to their normal position as soon as switching is completed. 8. In case of derailment at switch, disconnect switch and take detector bar off by removing clips from rail. Take great care to protect interlocking from unnecessary damage while replacing cars or locomotives. Report all cases of failure of trainmen or others to observe proper precautions to prevent unnecessary damage, as you will be held accountable for the same. Pass no trains after the derailment until all parts liable to damage have been examined and steps taken to protect trains. Be sure that track is safe before allowing trains to pass over. 253 9. In case of accident, notify division superintendent and me- chanical engineer at once by telegraph and call section foreman. 10. In case of accident to switch, disconnect it, set it for the main line and spike it. Use a flagman to protect all disconnected switches. 11. Should it be impossible from derangement to throw sig- nals when switches are closed train must come to full stop; signalman may then flag it past home signal after protecting all conflicting routes by placing their signals at danger. 12. Never move a switch lever when a train covers the switch or detector bar. 13. During freezing weather move all levers frequently and keep apparatus free from snow and ice. 14. You are not allowed to and must see that no one but an authorized workman makes any change in the apparatus or lock- ing, except under written order from the division superintendent or mechanical engineer. 15. Do not handle the apparatus roughly; pull the levers with a steady movement, being especially careful to move the signals without injurious jerk. 1 6. Keep all switches, locks and detector bars clear of cinders, ballast, sand, etc., and keep apparatus oiled. 17. Report at once any disregard of rule forbidding use of sand by engines with number of engine, train, etc. 1 8. Inspect all switches, signals and lighted lamps carefully as often as the weather or other indications may require, reporting every case of trouble. 19. Lamps must be handled carefully; keep them clean and in order, as per special circular of instructions; light and place them in position at the proper time. 20. Daily reports must be filled out according to instructions and sent regularly to division superintendent and mechanical en- gineer. These reports must be full and intelligible, giving exact character and location of trouble or defects in plant and pattern number or correct number of broken part. 21. Allow no one to enter tower whose duties do not require him there, without a written permit from the division superintendent or mechanical engineer. The daily reports alluded to are made on manifold paper of 254 the form as given herewith, a copy being made out each day by each of the signalmen and sent to the proper officers. Form S i. CHICAGO, MILWAUKEE & ST. PAUL RAILWAY CO. DAILY SIGNAL REPORT. Tower. Time M 189, To.. Is apparatus in good working order ? What trains have been delayed ? Note carefully any defects, looseness or breakage of apparatus or trouble with lamps, and state what steps you have taken to repair. Give train and engine numbers in cases of disobedience of signal rules. Remarks (Thirteen blank lines.) .Signalman. These reports are of great assistance in keeping track of the workings of the different plants and of any accidents or delays which may have happened. It also permits of a record being kept of almost everything that takes place at each of the towers, and by summarizing them comparisons can be made of the performance of each. The following list is a somewhat condensed summary of these reports for forty-one interlocking plants for the month of July, 1895, from which some idea can be formed of what has occurred during the month, and has been mentioned in the reports : Damage by accident, 2. Failure through neglect of operator, 5. Train off track at derail, I. Ran through open derail, trailing point, I. Defects of apparatus; lack of adjustment, 6. Other defects, 2. Track relays out of adjustment, I. Failures of circuits from other causes, 2. Large repairs completed, 8. Number of times at work making repairs, 88. 255 Inspections made by repairmen, 147. Trains delayed; defects of apparatus, Trains delayed by other trains, 32. . 6 . O >> , I i i CO' 1 PM* ^ ^^> CO _j u_r **-*" *> 'c3 CO ^ ^ 1 1 & Q '3 4-> a 'e5 S 3 CJ PQ -M a 0) . a 2 pS o o CO a* vn * * "^ Wi t^ K^ 1 P^ ti | o CO PM* o . O 3 ' < p4 t-5 K ^ H M J ON - H 1 2 ^ 1-1 ** M o r/> n ,J< . 1 O 8 5 r 3 8 5 S CH 2 ^ W H H ^ I CO ^> Q) |i Maintenance. "p o 3 c i ^ M W . c? in K> ^ W 9 : 1 PH (^ J CQ | O f PM CO PH* ^ CO d <% 4-> Cfl g 6 0*