-NRLF 
 
 B 14 5Ei4 3fl7 
 
1^^^^ 
 
 r\a/F <UJ 09 '9QQ "ON 
 
 Sssssss^s 
 
AMERICAN PRACTICE 
 
 IN 
 
 BLOCK SIGNALING 
 
 WITH DESCRIPTIONS AND DRAWINGS OF 
 THE DIFFERENT SYSTEMS IN USE ON 
 RAILROADS IN THE UNITED STATES, 
 
 1891: 
 THE RAILROAD GAZETTE 
 
 NEW YORK. 
 
' 
 
 - 
 
 
CONTENTS. 
 
 Introduction, 
 
 The Simple Block System, 
 
 The Sykes System, 
 
 Single-Track Blocking, 
 
 Automatic Clockwork Track-Circuit Signals, 
 
 The Eloctro-Pneumatic Track-Circuit System, 
 
 The Hall Signal, 
 
 Black's Mechanical Block Signal, 
 
 -A- IF* IE 3 E 3ST ID I ZXI (Illustrated descriptions). 
 
 The Sykes Block Signal Apparatus, 
 
 Automatic Signals on the Boston & Albany, 
 
 The Westinghouse System of Pnu'ematic Interlocking, 
 
 Automatic Block Signals on the Pennsylvania, 
 
 Electric Apparatus for Automatic Block Signals, - 
 
 Automatic Signals on the Fitchburg Railroad, 
 
 The H&ll Block Signal, 
 
 Hall Block Signals on the New York Central, 
 
 Black's Automatic Block Signal, 
 
 Illuminated Semaphore Signals, 
 
 The Koyl Parabolic Semaphore, 
 
 The Stewart-Hall Train-Order Signal, 
 
*"' 
 
AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 INTRODUCTION. 
 
 The pages which follow are a reprint of articles which appeared in the Railroad 
 Gazette in 1890 and previous years. It was the purpose to give in those articles a short 
 description of the methods of block signaling now in use in this country, summarizing 
 the salient features of each system. It was not within the scheme of the articles to ex- 
 haustively discuss or describe principles, methods, or appliances, but to give to one 
 comparatively new to the subject a fairly good grasp of it in general and in detail, and 
 to indicate to him the appliances which he might profitably investigate should he wish 
 for fuller or more accurate knowledge. All of the matter up to page 22 is substantially 
 a reprint of articles published in 1890, although a few alterations have been made to 
 bring the information up to date (1891). The descriptions of apparatus in the appendix 
 were published at various times, as indicated by the dates attached to each. 
 
 There are practically two systems of block signaling, of which we may call the first the 
 simple block system. Assuming that every railroad is properly supplied with telegraph 
 offices, this may be said to be merely a code of rules by which the attendant at a given 
 station exhibits a signal to hold all trains moving in a certain direction until he receives 
 word by electrical communication that the last preceding train has reached another 
 station and gone out of the section of track intended to be protected. This is an absolute 
 block system. When the regulations permit a second train to be sent over the road 
 before the first one has gone out of the block section, but under instructions to run 
 slowly, the term per missive blocking is used. 
 
 The signals by which trains are admitted to the sections under a block system may be 
 electrically connected with each other and locked by the Sykes system, so that when an 
 operator, after admitting a train to a section, puts his signal to danger, the lever of the 
 signal is automatically locked, so that the signal cannot again be lowered (placed in 
 the safety position) for another train until the train first mentioned has actually reached 
 the next station and operated an electromagnet. When signals are fitted with these 
 locks, permissive blocking is impracticable, unless (1) the operator disconnects the wires 
 by which the Sykes locks control the signal levers, or (2) the signal is left in the danger 
 
2 INTRODUCTION. 
 
 position and enginemen are instructed by flag, lamp, or hand signal to proceed regard- 
 less of the "regular signal. In either case the safeguard provided by the locks becomes 
 inoperative, as, when two trains are in a section, the first passing out releases the signal 
 so that a third may be admitted, while the second may be indefinitely detained within 
 the section. 
 
 The second system is the automatic. In this no attendant is provided, but each 
 signal stands ordinarily at "all clear" to admit a train to its section. The train on enter- 
 ing sets the signal at danger by the operation of an electric circuit, actuated by the 
 passage of the wheels, and resets the signal at " clear" when it emerges from the section. 
 The first automatic signal system was the original Hall, which was used in Massachusetts 
 and Connecticut about 1871. In this the electric communication from one station to 
 another was by means of a line wire strung upon poles. Some eight or nine years later the 
 track-circuit system was introduced on the Fitchburg road by the Union Signal Company 
 In this the electric circuit is conducted from one end of a block section to the other 
 through the rails of the track ; and the proper working of the system depends upon the 
 integrity of this circuit. The presence in the section of even a single pair of iron wheels 
 connected together by an iron axle allows the passage of the electric current from one 
 rail to the other, and withdraws the force that holds the signal at "clear." If a train 
 breaks apart, the exit of a portion of it from the block section does not clear the signal at 
 the entrance, as is the case with a simple line-wire system. 
 
 The most common form of track circuit signal is that in which the signal consists of 
 a disc operated by clockwork, the latter being controlled by the electric current. In the 
 pneumatic track-circuit system the signal consists of a semaphore operated by compressed 
 air, which is controlled by the electric circuit. 
 
THE SIMPLE BLOCK SYSTEM. 
 
 The most extensive block system in this country is that of the Pennsylvania Rail- 
 road, which is substantially the same as that in use on the great majority of the railroads 
 of Great Britain, though there are numerous differences in the detail of operation. The 
 Pennsylvania uses the simple block system (without Sykes locks). When the system 
 was introduced the number of trains had already grown so large that it became necessary, 
 in order to accommodate them, to establish stations especially for signaling, the regular 
 stations being too far apart. These intermediate stations are generally two-story build- 
 ings, and are termed "towers" ; and these buildings, being characteristic of the system, 
 have come to be regarded by many people as an essential part of it; but, in point of fact, many 
 of the block-signal operators are located in ordinary station offices. On those sections of 
 the Pennsylvania where trains are most frequent the block sections are from one to 
 two miles long. Near large terminal stations the intervals are in many cases considera- 
 bly less than a mile. Regular telegraph stations are used wherever possible, but the 
 larger stations have to have two telegraph offices, one for block signaling and one for 
 ordinary business. Special stations are established between the regular stations at such 
 points as will best divide the space and maintain an approximately uniform length of 
 block. On portions of the road where trains are less frequent the sections are made 
 longer, in some cases four or five miles. Each station has a fixed signal. This con- 
 sists of a semaphore with a single light, which shows red when the arm of the semaphore 
 is at "danger" (horizontal) and white when the arm of the semaphore is dropped, to 
 indicate ' all clear." There is a separate semaphore arm for each track, but the eastbound 
 and westbound arms are generally placed on one post, and a single lamp answers for both. 
 The older form of signal was a disc, but these are being gradually displaced by the sema- 
 phore, which is now standard. The electrical apparatus consists simply of a Morse tele- 
 graph line, with the usual instruments. On the passage of a train the operator places 
 the signal at danger to stop following trains, and reports the time to the station which the 
 train last passed and to that toward which it is proceeding. 
 
 Summary of Pennsylvania Joules. A block section is called a "block." Trains 
 will be governed absolutely by fixed signals, and will not observe the time-space rule. 
 The old form signal has a green disc to indicate caution. With semaphores a position 
 midway between horizontal and the nearly vertical position is employed to indicate cau- 
 
4 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 tion. These cautionary indications are of course used only in permissive blocking. The 
 signal normally stands at danger. After being changed for a train, it must be returned 
 to danger as soon as the whole of the train has passed the signal. A train mu*t not be 
 backed after stopping at a block station. In case of failure of wires, or the operator for 
 any reason cannot get orders for a train, he must give it written notice of the reason for 
 the proper signal not being displayed. Trains must not be admitted to a block section 
 under the permissive system to follow a passenger train, and a passenger train must not 
 be admitted to follow permissively any train until it is first stopped and notified that 
 there is a train ahead. Exceptions are, however, made to the last paragraph. A train 
 intending to use a crossover between the block stations must notify the operator. A 
 train must not be reported as having passed until the rear end has got 300 ft. beyond the 
 signal. Trainmen are not relieved from observing all ordinary rules in regard to the pro- 
 tection of their train. 
 
 The customary place for the signal is immediately opposite the telegraph office. 
 Where this is at a passenger station, inconvenience sometimes results from the fact that 
 the train held by the signal is not in a convenient position for discharging and loading 
 passengers. To provide against this the station should be equipped with two signals, a 
 " home" and a " starting, " the former to stop trains before they reach the station, and 
 the latter to hold trains which are standing at the station. By this means, if the block 
 section in advance is occupied, a train may be safely admitted to the station while yet it 
 is kept under control, so that it cannot leave without the permission of the operator. If 
 a train is detained at the station, the following train need not be held back at the en- 
 entrance of the next preceding block, one, two, or more miles away, but may be allowed 
 to come up to the home signal, whence it can proceed, without delay, to the platform as 
 soon as the one in advance has made way for it. 
 
 A danger in signaling is the possibility of trains entering the main track from a siding 
 or at a crossover track midway of the section without the knowledge of the operater at 
 the entrance to the section. To provide against this, all switches connecting with such 
 sidings or crossovers should be under the control of the operator. This may be effected 
 by an electromagnetic lock, so arranged that it cannot be released except from the oper- 
 ator' s office, the latter being connected with the switch by wire ; or an ordinary switch 
 may be locked by a special key, which must be obtained from the block-signal operator. 
 Neither of these systems is used to any extent in the United States. The regulations of 
 the Pennsylvania, as noted above, require simply that a conductor intending to turn a 
 switch between two block stations for the purpose of using another main track than that 
 on which he belongs must notify the block operator beforehand, and get his acknowledg- 
 ment, with authority to so use the track. In view of the difficulty of controlling these 
 outlying switches satisfactorily, an essential point in preparing a road for operation under 
 a block system is the lengthening and alteration of side tracks so that as many as pos- 
 
THE SIMPLE BLOCK SYSTEM. 5 
 
 sible of such tracks shall connect with the main track at a station that is, between two 
 signals, a home and a starting signal, which are controlled and handled by the same 
 operator. 
 
 A train must of course never pass a block signal until its indication is absolutely 
 known. When there is a fog or a driving snowstorm, or the signal is obscured by steam 
 from a locomotive or any other cause, a fast train approaching a block station must be 
 slackened in order to permit the engineer to make sure of the indication of the signal 
 before it is too late for him to stop. The annoyance from numerous delays to fast trains 
 from this cause has led to the introduction of cautionary signals, erected at a distance 
 from the home signal, and indicating the position of the latter. Home and distant sig- 
 nals generally differ in color, and the end of the distant signal blade is notched. Custom- 
 arily the positive signal blade is painted red and the cautionary green. The light on the 
 cautionary signal is made to indicate green for caution and white for "all clear." The 
 cautionary signals are erected at from 1,000 to 2,000 feet from the home signal. If an 
 approaching engineman finds one of them in the "all clear" position, he knows that 
 the home signal has been pulled to "all clear," and that he need not expect to be 
 stopped at that signal. This distant (cautionary) signal must of course be interlocked 
 with the home (positive) signal so that it can never by mistake be pulled to safety until 
 the positive signal has actually been so pulled. 
 
 Illuminated blades, which are extensively used for switch signals in yards (for move- 
 ments other than those of fast trains on main tracks), have been used to a limited extent 
 for fast-route signals, and are equally applicable to the block system. The term "illumi- 
 nated blade" means a blade in connection with which a lamp (hidden from the engineer) 
 is attached to the post is such a position that it throws light directly on the face of the 
 blade. The engineman can thus see its position at night the same as in the daytime, and 
 a signal lamp is unnecessary. Illuminated blades are prescribed for all new work on 
 the Pennsylvania lines west of Pittsburgh, and the standard color of blades there is 
 yellow. By this means the color indication can be entirely discarded. 
 
 The painting of arms yellow or some color which has not by common custom received 
 some definite significance is a step in the right direction, as there is an inconsistency in 
 painting the face of a semaphore arm white or red, because it must indicate when hori- 
 zontal, the opposite of white, and when down the opposite of red. Green being gener- 
 ally used for caution gives a wrong indication when the arm is either up or down. 
 
 The chief fault found with illuminated blades is the difficulty of making them visibl e 
 at the proper distance. To overcome this the Union Switch & Signal Co. has introduced 
 a blade carrying a corrugated reflecting surface of brilliant, non-corrosive metal. Koyl's 
 parabolic semaphore, manufactured by the National Switch & Signal Co., consists of a 
 semaphore arm made on the lines of a section of a parabola, so as to more efficiently re- 
 flect the rays of light in parallel lines. Both these signals may or may not be arranged 
 
6 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 so as to throw a red light upon the blade when it is in the horizontal position, and a 
 white light when it is pulled down. These devices are familiar to those who have followed 
 the progress of the art in the columns of the Railroad Gazette, and are described in de- 
 tail in the appendix. 
 
 Cost of Maintenance. The principal item of cost is of course the wages of opera- 
 tors and inspectors, to which is to be added the maintenance of buildings, with fuel and 
 lights, where a building is erected especially for this service. At stations where 
 operators have no switches to attend to, and no other work of any kind, they work 12 
 hours each per day, seven days in the week. The pay of these men is from $45 to $55 
 per month. Where the duties are more complex the pay is higher, and, where a con- 
 siderable number of interlocking switches is operated, the working time for each man is 
 eight hours daily. The men at these important towers are paid, on the Pennsylvania, 
 from $50 to $70 per month. On the New York, Lake Erie & Western some of the 
 operators, who work 12 hours daily, alternate weekly between day and night work. 
 The duties of the inspectors are light, so far as 'simple block stations are concerned. 
 Their work is chiefly in connection with interlocking towers, which at all important 
 stations are operated, as intimated above, by the same man who attends to the block 
 signaling, and the time spent inspecting simple block stations is treated as a secondary 
 matter. For more detailed estimates of expense the reader is referred to a subsequent 
 chapter on the Sykes system. As in the system here described the cost of inspection is 
 but a very small fraction of the total expense, and as the latter must depend upon con- 
 ditions which must be calculated in each case by itself, farther consideration of the sub- 
 ject here is unnecessary. A recent estimate on a prominent road showed that 100 miles 
 of its line could be worked under the block system, with block sections four miles long, 
 by the erection of only three towers between regular stations. When it is considered 
 that trains running at 40 miles an hour and on 10-minute intervals are 6f miles apart, 
 the possibilities of the block system will be readily recognized. 
 
 The West Shore, the Chicago, Burlington & Quincy and the New York, Lake Erie 
 & Western are the principal roads, outside those controlled by the Pennsylvania, which 
 use the block system as above described. Others use it on very short sections of road or 
 for only a portion of the trains. It is scarcely necessary to say that this system has 
 given full satisfaction wherever used. No officer on a road using it ever thought of 
 abolishing it or of diminishing its use in any way. Operators have admitted a train to a 
 section when it was not clear, and engineers have disregarded danger signals, causing 
 collisions ; but the excellent record on the Pennsylvania, where the system has been in 
 operation on 500 miles of double track for over ten years, shows that these defects of 
 discipline are not to be regarded as incurable. The first step toward abating them by 
 mechanical means is the adoption of the Sykes system, which we shall next consider. 
 
THE SYKES SYSTEM.* 
 
 The most common lorm of electric locking, as an additional safeguard to be 
 used in connection with the block system is that known as the Sykes system. In 
 fact, this is the only apparatus of the kind yet put in use in this country. It 
 is in use on the New York, New Haven & Hartford, the New York, Lake Erie & 
 Western and the New York Central & Hudson River. The latter company has 
 only about 18 miles of road equipped with this apparatus (all in New York City); but as 
 the system has here been in use longer than on either of the other roads named, and las 
 to meet the most trying conditions, we shall base our description on the information given 
 by officers of this road. The New York Central allows no permissive blocking whatever 
 where the Sykes instruments are in use. 
 
 The apparatus consists essentially in a series of electromagnets so connected with 
 the levers by which the operator moves the outdoor signals that the operator at the out- 
 going end of a block section controls the lever by which the operator at the incoming 
 end admits the trains. Thus after A sends a train to and puts his signal at danger, 
 he is unable to again pull the signal to "all clear " until B unlocks his (A's} lever, and 
 B of course refuses to do this until the train has arrived and passed out of the section. 
 To provide against a possible mistake by B, who might prematurely unlock A' a lever, 
 there is also an automatic arrangement by which A's lever, after having been put 
 through the motions to admit a train, cannot be unlocked until the train itself actually 
 passes out of the section. This is secured by running an electric circuit, which controls 
 A's lever, through two or three rail lengths of the track at a point just beyond B. The 
 circuit goes from the battery to one rail of the insulated section of track, thence by line 
 wire to A's signal, which it holds locked at danger by energizing an electromagnet. On 
 the passage of a pair of wheels over these insulated rails, the circuit is led through the 
 wheels and axles from one rail to the other and thence back to the battery without going 
 to the electromagnet at the distant station, thus demagnetizing that instrument and 
 allowing the signal to be again operated. 
 
 It will be seen that where trains are run permissively that is, where a second train 
 passes A before the first one has passed B the automatic feature of this system becomes 
 useless, as the first train will release A's signal while the second train is still in the sec 
 
 * Drawings and detailed description are given on page 22. 
 
8 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 tion, and the apparatus will then afford no protection against a careless operator admitting 
 a third train to the section before the second has cleared it. 
 
 One section of seven miles of double track, on which there are seven stations, was 
 equipped by the New York Central in 1882. The cost of this, including two interlocking 
 machines for handling three switches and six switches, respectively, was $8,300, of which 
 $2,050 was for the seven cabins and 25 semaphores. The cost of operating these seven 
 stations is, per year : 
 
 Salaries of cabin men $9,480 
 
 Salary of electrician 1,000 
 
 Four men attending signal lights 1,920 
 
 Cost of repairs (estimated) 1,000 
 
 .. $13,400 
 
 This estimate would be^o-Mgh.for a system comprising many stations, as the salary 
 and wages account wouj<i a be' distributed ^ve.r*j more stations. On another road the cost 
 of maintaining 24 statiprif using Sykes loyks.was estimated at $3,043 per year, divided as 
 follows: \ . x V ''.;* *\ 
 
 ^ \kl-. . * 
 
 Battery supplies .*.- ..'.*.... $ 153 
 
 Battery men and inspector of electrical apparatus 1,440 
 
 Ordinary inspector (occupied partly with other duties) 300 
 
 Miscellaneous repairs, materials, paint, etc 150 
 
 Deterioration (estimated) 1,000 
 
 $3,043 
 
 This of course does not include the wages of the operators, who in this case have 
 at many of the stations other duties to perform. 
 
 If we add the salaries of the operators to this estimate, the average cost per station 
 per year will be, in round numbers : 
 
 Two operators (day and night), at $55 each per month $1,320.00 
 
 Battery supplies ..., 6.50 
 
 Inspectors 72.50 
 
 Miscellaneous repairs 6.50 
 
 Deterioration 41.50 
 
 $1,447.00 
 
 This takes no account of lamp lighters. Where there are no distant signals the 
 operator can generally light the lamps himself. In both the cases cited there are but few 
 distant signals. Where these are near regular stations the work of attending to the lamps 
 can be economically devolved upon the man who attends to the ordinary switch lamps. 
 The expense chargeable to block signaling for this service is to be added to the total above 
 given ($1,447), while on the other hand this sum can be diminished by an amount equal 
 to such portion of the operators' salaries as can be fairly charged to ordinary station 
 work. 
 
 The Sykes instruments, like most other electromagnetic devices, require constant and 
 
THE SYKES SYSTEM. 9 
 
 careful inspection. One inspector has told us that from a careful record he found his 
 instruments to fail once in 20 days. He did not give the number of trains, but, as the 
 causes of the failures were mostly on account of inadequate inspection or too infrequent 
 renewal of battery, he had no doubt that by the increase of his force of inspectors he 
 could reduce these failures to practically nil. On another section, where he was able to 
 provide sufficient men to look more carefully after the apparatus, a number of signals 
 had been worked under a very heavy traffic for more than a year without a single failure. 
 
 k . * I 
 
SINGLE-TRACK BLOCKING. 
 
 Within the past seven years a number of railroads have introduced the block system 
 on single-track lines, modifying the regulations somewhat to conform to the different 
 conditions. This system, as used on the Canadian Pacific, was described in the Railroad 
 Gazette of Dec 2, ] 887. Its operation on a division of the Chicago, Milwaukee & St. 
 Paul was described June 22, 1888, and the same system, with a novel arrangement of tele- 
 graph wires and sounders, as used on the Chicago & Council Bluffs division of that road, 
 was described in the issue of Jan. 17, 1890. An account of the system as employed on the 
 Wabash was given Feb. 8, 1889. The salient feature of these applications is the moderate 
 cost, which has been reduced to a minimum. There are several cases in the United States 
 in which the trains both ways on a single-track road aggregate more than the trains one 
 way on most double-tracks; but naturally a single-track road has fewer trains than a double- 
 track line, and in nearly or quite every case the improved system has been put into effect 
 wholly by the employment of regular station operators, no towers being established, and 
 the only important additional expense being the employment of night operators at some 
 stations where otherwise the office would be kept open only during the daytime. There is 
 no practical difficulty in making this system as effective as the regular double-track block, 
 but most or all of the companies using it have put it in as a necessity, and have not felt 
 able to incur the expense of new signals, carefully located for purposes of block signaling, 
 of distant signals and of protection for sidetracks and other facing- point switches between 
 stations. They have treated the block system essentially as an adjunct of the train- 
 dispatching system. While by no means perfect, there can be no question that the sys- 
 tem is a valuable safeguard against rear collisions. Assuming that conductors and 
 engineers are properly disciplined so as not to depend upon the system for protection 
 against dangers which it does not pretend to cover, its value as a substitute for the uncer- 
 tain flagging system, especially in cold and stormy weather, is undoubted. 
 
 The rules under which this system is operated vary considerably on the different 
 roads named, as will be seen by reference to the accounts cited above. One of the most 
 valuable rules in connection with this plan of working is that which requires inferior- 
 class trains to time themselves so as to be wholly out of a block section before the time 
 at which a superior-class train is due to enter it. Another is that which provides that 
 when a train is to take a sidetrack at a station the operator must not open the block 
 
SINGLE-TRACK BLOCKING. 11 
 
 for another train until this one has completely cleared the main track. All the roads 
 named, we believe, employ the ordinary train-order signal for stopping and starting 
 trains, though the Chicago, Milwaukee & St. Paul has a semaphore for block signaling 
 and one of the old-style disc signals for train orders. The use of two danger signals at 
 the same station would seem to be of questionable expediency, as locomotive runners 
 would probably be as likely to overlook one of the signals as an operator would be to 
 make a mistake in using one signal for 1rwo purposes. A number of roads which use the 
 block system on single-track allow train dispatchers discretion in suspending rules dur- 
 ing clear weather and on portions of the road free from curves, when traffic can be has- 
 tened thereby; and these rules are used temporarily during fog or severe snow storms on 
 a good many miles where the companies have not yet seen their way clear to incur the 
 necessary expense of their constant operation. 
 
 On the Canadian division of the Michigan Central passenger trains are kept one or 
 more stations apart by the regular train-order system, the dispatcher giving a special 
 written order to the operator in regular form for each operation; and this system is ex- 
 tended to freight and other trains during fogs and snow storms. 
 
 Cost of Operation. As intimated above, there is generally no special item of ex- 
 pense connected with this system, except the employment of additional operators. The 
 Wabash employed three at $45 each per month on a section of 20 miles. The Chicago, 
 Milwaukee & St. Paul added six on a section of 130 miles. On this 130 miles the cost of 
 new semaphores and a line wire with electromagnets (sounders) of a special form was 
 $30 per office, these averaging four miles apart. 
 
 A recent letter from a Wabash officer gives details of operation of the system on a 
 20-mile section of that road which were not fully explained in the article above referred 
 to. We quote a paragraph: 
 
 "When a train passes a station, the operator reports the time, and adds, ' signal out.' 
 This report is watched for by the operator at the station the train previously passed, 
 and he then responds 'signal in,' releasing the block. The dispatchers overhear these 
 reports, and know always that the proper responses are being made, but the operators 
 are so trained that it is not necessary to call them to release the block after the train 
 has passed the next station. There has not been a single occasion when it was found 
 necessary to suspend the operation of this system in order to avoid delays. It is 
 true that some delays to trains occurred by reason of the use of the system, but we have 
 had no accidents, although it is a very busy piece of the road; as high as 60 trains a 
 day being moved over the 20 miles of single track between Decatur and Bement. I know 
 of no instance where an operator has made a mistake, or permitted a train to go into 
 a section when it was dangerous to do so." 
 
 This correspondent refers to delays. These cannot be accurately compared, as be- 
 tween a time-interval and a distance-interval system of spacing trains ; but, as every one 
 
12 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 knows, the efficiency of a block -system as a means of running trains close together de- 
 pends wholly upon the length of the block sections. On the Wabash they are about four 
 miles long. With stations half a mile apart, eight times as many trains could be run, and 
 with perfect safety, at any speed, provided a caution signal were erected at a sufficient 
 distance from each block signal to allow a train to be got "under control" before it 
 reached the block signal. 
 
 Although the Canadian Pacific was the first road whose single-track block system 
 was brought prominently before the railroad public, the eastern division of the Lake 
 Shore & Michigan Southern seems to have been the pioneer in this respect, and we there- 
 fore add here an extract from a letter from one of the officers of that road : 
 
 "We have used for the last fifteen years positive block on passenger trains following 
 each other, only allowing one train between the same stations at the same time. We only 
 use positive block on freight trains during foggy and stormy weather. The positive block 
 on passenger trains is solely in the hands of the telegraph operators, but, at the same 
 time, the train dispatchers oversee the work, and see that it is done properly. The posi- 
 tive block on freight trains is in the hands of the train dispatcher, to whom the state of 
 the weather is given everj 7 hour, and on whose judgment the necessity for the use of 
 positive block depends. If a fog or snow storm should suddenly come up, the operators 
 immediately notify the train dispatcher, whether it is the regular time to report or not. 
 If the train dispatcher directs the use of positive block for freight trains, then it is wholly 
 in the hands of the telegraph operators, under the same conditions as the passenger trains. 
 The same rules also apply to trains on double track." 
 
AUTOMATIC CLOCKWORK TRACK-CIRCUIT SIGNALS.* 
 
 The most common form of automatic block signal in this country is the Union Switch 
 & Signal Company's " Union system." This system is in use on the Boston & Albany, 
 Old Colony, New York, Providence & Boston and a number of other roads. The same 
 apparatus is used on these and many other roads as a station signal ; but these applica- 
 tions should not be confounded with the block system, as in nearly every case there is 
 only one signal at a station, and the regulations under which it is used make it a 
 cautionary signal. 
 
 The arrangement of the apparatus in this system is, briefly, as follows : The battery 
 is placed underground (or in any position where it is protected from freezing) at the out- 
 going end of a section, and from this the electric current is conducted, through one of 
 the rails of the track, to the signal at the incoming end of the section; thence, after pass- 
 ing through the signal relay, it returns to the battery through the opposite rail. At 
 switches the current is led through a circuit breaker which is opened whenever a switch 
 rail is moved so as to break the main track. The circuit is also led through the rails of 
 side tracks for a short distance, so that a train entering a side track is protected, the 
 same as when on the main track, until all its cars are fully clear of the main line. The 
 signal itself is a disc fixed to a vertical spindle, with which it is made to turn one-quar- 
 ter of a revolution every time the circuit is opened or closed. The turning is effected by 
 clockwork, actuated by a weight, which has to be periodically wound up. A lamp fixed 
 to the upper end of the spindle gives the same indications as the disc. The electric cur- 
 rent flowing through the rails does not operate the disc directly, but by means of a relay 
 opens and closes a more powerful local circuit which starts the clockwork. The opera- 
 tion of the signal is simple, the presence of a pair of wheels on the track in any portion 
 of the block section serving to devitalize the electromagnet which holds the signal at " all 
 clear." This condition continues as long as the train or any portion of it is in the 
 section. 
 
 The signal post is placed about 100 or 150 feet within the block section, so that an 
 engineman on approaching it and finding it "all clear" may see it change to danger (for 
 the protection of his own train). If it fails to move from white to red, he knows that 
 
 * Drawings and detailed description are given on page 30. 
 
14 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 some part of the apparatus is out of order, and he must then assume that there is danger 
 ahead. 
 
 By a proper combination of relays and by the use of a section of line wire (on poles) 
 the signal for each block section may be so arranged that it will not be restored to "all 
 clear" until the train has passed several hundred feet beyond the outgoing end of the 
 section, thus providing, in effect, a distant signal. 
 
 The roads using these signals give, in general, very favorable reports of their be- 
 havior. Reports made to railroad commissioners and other official bodies have given 
 detailed records of the number of stops occasioned by the signals, subdividing the lists 
 into those caused by a preceding train in the section, switch misplaced, broken rail, 
 failure of battery or instruments, carelessness of custodians or inspectors and other con- 
 tingencies; but the records available show such widely divergent results that they cannot 
 be regarded as a proper basis upon which to form an opinion. Many of the failures and 
 delays were the direct result of inexperience or lack of proper supervision. The practical 
 question, however, with regard to stops caused by imperfect operation of any automatic 
 signal, is, how numerous and how serious will be the delays to trains ? And the remedy 
 for these faults lies in constant and careful inspection and great care in securing the 
 very best material and workmanship. The difficulties in this direction have not been 
 entirely overcome, as is evident from the changes in the methods and means of taking 
 care of signals ; but that the results are in general satisfactory is evidenced by the erec- 
 tion of new signals by those roads which have tried them the most carefully. The Boston 
 & Albany is now completing the equipment of a whole division of 54 miles, and other 
 New England roads use them largely and with satisfaction. The Union Switch & Signal 
 Co., however, regards its electro-pneumatic system (with semaphores) as so greatly 
 superior to the clockwork system that it takes no special pains to spread the use of the 
 latter. 
 
 The possibility that track circuit signals may indicate safety when danger exists, 
 which is a vital question with all automatic signals, will be discussed hereafter in con- 
 nection with the pneumatic system. 
 
 A point in favor of the clockwork signals, which is regarded as important by some 
 experts, lies in the fact that the day signal is a disc and not a semaphore. All automatic 
 systems must make some provision for releasing a train when the signal, through some 
 accident or mishap, stops it while the track is in fact clear. The common way of doing this 
 is to require trains to come to a full stop, and then, after one, two, three, or any prescribed 
 number of minutes, to proceed under control through the section. To do this the engineer 
 must pass the signal while it stands at danger. If such unnecessary stops occur fre- 
 quently, engineers, becoming habituated to passing signals standing at danger, will be 
 liable to carelessly pass danger signals in yards and at other points where discipline 
 imperatively requires that a train shall never pass a signal showing danger. But with all 
 
AUTOMATIC CLOCKWORK TRACK-CIRCUIT SIGNALS. 15 
 
 such yard signals made in semaphore form, and all automatic block signals made in disc 
 form, it is argued that the engineer need not fall into the careless habit mentioned, 
 because the differing forms will always be a guide to him in deciding their different 
 degrees of importance. 
 
 One of the roads using this system has given us the following memoranda concerning 
 cost of erection and maintenance. Sixteen signals, covering 16 blocks in which were 36 
 switches, cost $8, 076.54. This is divided as follows: 
 
 16 signals at $300 $4,800.00 
 
 36 switch connections at $40 1,440 00 
 
 Labor, signals, average $96.78 each 1,548.54 
 
 Labor, switches, at $8 each '. 288.00 
 
 $8,076.54 
 
 On one section containing 83 signals, all within a territory 11 miles in length, the 
 expenses for maintenance for one year were $7,062.67. Of this, material and supplies are 
 charged $2,831.85, and labor and superintendence $4,230.82. This makes a total of $85.09 
 per signal per year. On another section 160 signals cost for one year $86 91 per signal. 
 This last included important renewal work (10 miles of new poles and a large amount of 
 underground wire). 
 
 On another road the expense of maintenance was estimated at $90 per signal per year, 
 the employment of the inspectors on other work for a considerable portion of their time 
 making a more accurate estimate impossible. 
 
THE ELECTRO-PNEUMATIC TRACK-CIRCUIT SYSTEM.* 
 
 The main difference between the apparatus of these signals and that of the clock- 
 work signals is in the motive power and the form of the signal. In the pneumatic system 
 the signal, which is a semaphore, is cleared by compressed air and goes to danger by 
 gravity. The compressed air is supplied to the various signals by a pipe running along- 
 side the track, a stationary steam engine or other power for compressing air being 
 located at a central point where it can supply a number of signals. Pneumatic signals 
 were used on the West Shore road as early as 1883, but the system there was abandoned 
 principally because the pipes for conveying compressed air, and other details, were not 
 properly maintained. When the time came for renewing the apparatus the prosperity 
 of the railroad company had declined and no action was taken. 
 
 The electro-pneumatic semaphore has also been in use on the Fitchburgh road since 
 1883, and its operation there was described in the Railroad Gazette of June 15, 1888. 
 
 The latest form of electro-pneumatic signal was described in the Railroad Gazette 
 of Dec. 21, 1888, and Aug. 23 and Sept. 6, 1889. These improved signals are in use on 
 the Pennsylvania road east of Pittsburgh, where six miles of four-track line have been 
 operated for about two years, and a portion of it since 1884. The officers of the road say 
 that these signals have given them no trouble whatever. Seven miles of the four-track 
 line of the New York Central & Hudson River, between 138th street and Woodlawn, 
 New York City, are being equipped with this system, and it was put in use last year on 
 seven miles of the Central of New Jersey. 
 
 This system, as operated at the places named, embraces important features not yet 
 employed in any other automatic signals, and the plan is in many respects an ideal one. 
 The signal is a semaphore. It is placed exactly at the entrance of a block, and therefore 
 does not turn to danger in the face of the engineer. There is a distant signal for each 
 home signal. Each block section is worked with a single circuit, and the distant signal 
 is controlled by a wire circuit upon poles. 
 
 It has been regarded as an important principle in automatic signaling that the loco- 
 motive engineer should witness the operation of the signal as he passed it, thus having 
 constant evidence that the signal is in working order. The location of the signal precisely 
 
 * Drawings and detailed description are given on pages 45, 46 and 49. 
 
THE ELECTRO-PNEUMATIC TRACK-CIRCUIT SYSTEM. 17 
 
 at the entrance of the section, as is the case with the pneumatic signal on the Pennsylvania, 
 makes this inconvenient, if not impossible, as it is hardly to be regarded as expedient, even 
 if it were practicable, to require an engineer to turn around and look at a signal after he 
 has passed it. By the enforcement of adequate discipline the brakemen of a train could, 
 however, be utilized as monitors for this purpose ; and we understand that something of 
 this kind is done on the Pennsylvania. This is, however, a question of practice. There is 
 no reason why the electro-pneumatic signals or any automatic signals cannot be placed 
 in advance of the entrance to the seection if desired. 
 
 The track circuit system is an ideal automatic block signal system. It promises more 
 complete protection than is possible with any other. The electro-pneumatic system as 
 now in operation on the Pennsylvania, New York Central and Central of New Jersey is 
 the highest development yet reached by the track circuit ; but any track-circuit system 
 requires not only perfect construction, but the most careful maintenance. Of course the 
 most serious failure of any automatic signal is when it indicates safety when it should 
 show danger. We have heard of a few such failures of the electro- pneumatic signals. 
 We would not say that these false indications have come from unavoidable defects in 
 the system. On The contrary, there are no stronger advocates of the electro-pneumatic 
 system than those who have used it most. 
 
 If an automatic signal stands at safety, and fails to change when the train passes it, 
 the engineer must, according to the rules, proceed cautiously ; but by this reduction of 
 speed the train loses time and the next following train may soon be too close upon it. If, 
 on the arrival of this second train, the condition of the apparatus has undergone a change, 
 so that the signal changes from safety to danger in the proper manner, the engineer, 
 observing this movement, proceeds at full speed. A train should, therefore, be protected 
 by hand signal whenever it enters a section, unless it is known that the signal indicates 
 danger behind it. 
 
 The cost of erecting these signals, with the necessary apparatus for compressing and 
 conveying the air, is rather high as compared with other systems, but the use of com- 
 pressed air in place of manual power for the operation of switches and signals in yards, 
 which is often highly advantageous, affords a means of dividing the expense. The latest 
 statistics we have concerning the cost of maintenance of pneumatic block signals are 
 those reported for the ten miles on the Fitchburg road in the article above referred to. The 
 cost per signal per yea.r there was $133, or approximately 5u per cent, more than that of 
 maintaining clockwork signals. 
 
THE HALL SIGNAL.* 
 
 The Hall wire-circuit automatic block signal is the oldest automatic signal used to 
 any extent in this country, and the instruments put up by Mr. Thomas S. Hall on 16 
 miles of the Eastern Eailroad (now the Boston & Maine) in 1871 are still in use ; but the 
 company has recently been reorganized, after several years of inactivity, and the signal, 
 as now offered, may be regarded as substantially a new device, the improvements over 
 the old pattern being radical in many respects. The apparatus and some of its applica- 
 tions were described in the Railroad Gazette, June 18, September 12 and December 5, 
 1890. The signals are used extensively on the Hartford Division of the New York, New 
 Haven & Hartford and on the Boston & Worcester Division of the Boston & Albany 
 The latter road in 1890 equipped 38 miles of double track with continuous blocks. The 
 New York Central & Hudson River has recently equipped 8 miles of double track (de- 
 scribed in the Railroad Gazette, December 5, 1890) and the Michigan Central has also 15 
 miles of double track so protected. 
 
 The Hall Company makes no use of the track circuit, claiming that the adjustment of 
 battery and of instruments is such a delicate operation that it cannot be made to give 
 satisfactory service. The connection from one signal to another is by line wire upon 
 poles, and passing trains actuate the signals by means of levers placed at right angles to 
 the rails in such a position that the wheels depress one arm, and, by the elevation of the 
 other, open or close an electric circuit. The signal is a disc of very light weight, in- 
 closed within a wooden case in which is a circular opening covered with glass. This disc 
 is attached to the armature of an electromagnet in such a way that when the magnet is 
 energized the disc is held up out of sight, and on the cessation of the current falls by 
 gravity in front of the glass covered opening. It is colored red, and the surrounding 
 surface of the case is black, so that when visible it indicates danger. At night a light 
 is placed back of the opening Jn the case, and indicates safety or danger according 
 as the disc is held up or is dropped, the disc being of red silk and translucent. The 
 signal magnet in this system is operated ^directly by the primary circuit, no relay or 
 other device for multiplying power' being necessary. The current whose continuity 
 is necessary to maintain the signal, at "'all clear" is, by the entrance of a train to the 
 section, broken at two places, arid one of these contacts is so broken that it cannot be 
 
 -- Drawings and detailed description are given on pages 53-58. 
 
THE HALL SIGNAL. 19 
 
 restored except by the action of the track instrument at the outgoing end of the 
 section. 
 
 The merits claimed for the Hall signal over the track-circuit system are : 
 
 All operations are by positive making and breaking of metallic connections. 
 " Shunts." by which a small portion of a current still flows through the instrument in- 
 tended to be devitalized, are not used. With this apparatus the amount of battery 
 power may be variable within wide limits, and extreme delicacy of adjustment of magnet 
 armatures is not important. The number of electromagnets is reduced to a minimum 
 and the labor of care and inspection is correspondingly less. With these advantages, 
 supplemented by good design and workmanship, the number of unnecessary stops is 
 claimed to be largely reduced. In fact, the company shows records of signals which 
 have run a year without causing a single unnecessary stop. It is claimed that such a 
 thing as showing safety while danger exists has never been known, and is, in fact, 
 impossible. 
 
 The Hall wire-circuit system does not indicate the presence of a detached car in the 
 block section. The company admits this point, but claims that the superior reliability, 
 with reduced cost of the system, is of enough value to outweigh it, a-nd also that the 
 use of air brakes on freight trains, and the exercise of good discipline otherwise, practi- 
 cally provide for such a contingency. The prices charged are about the same as those 
 asked for the Union clock-work signals. The company also furnishes a combination 
 of track instruments, interlocking instruments and relays by means of which, when two 
 or more trains enter a section of, say, one mile in length, the trains being a half mile 
 apart, each train will, on passing a track instrument 2,000 feet after entering the section, 
 break the circuit controlling the signal in such a way that only the last train can restore 
 it to "all clear " ; in other words, a train passing out of a section cannot clear the signal 
 at the entrance of the section if a following train has already got 2,000 feet within it. 
 
 The cost of operation of the Hall signals cannot be stated with accuracy, because the 
 form now being used most extensively has not been in operation long enough to afford a 
 good basis for estimating. The maintenance of a number of the signals on the New 
 York, New Haven & Hartford has cost about $65 per year, but the battery power of the 
 later patterns is greatly reduced, and the cost of material and of labor is correspond- 
 ingly less. 
 
BLACK'S MECHANICAL BLOCK SIGNAL.* 
 
 This apparatus is used on the Manhattan Elevated (New York City) and other pas- 
 senger lines in the vicinity of that city. The Manhattan has a series of 32 continuous 
 block sections in operation. The instrument consists of a small semaphore, the post 
 being cast iron, which is set to danger by a lever actuated by the wheels of a passing 
 train. The same signal is connected by a gas-pipe rod, extending along the road, with 
 another track instrument at the outgoing end of the section, and the train, on striking 
 this, resets the signal at safety. The maximum length of block operated is about 1,700 
 ft., and at this distance the signals have worked with great success for two or three years. 
 It will be understood that the moderate speed and uniform length of the trains on these 
 city railroads permit the use of a signal whose capabilities for longer blocks or for more 
 varied service are limited. 
 
 SUMMARY. 
 
 To briefly summarize the salient features of the different systems of block signaling, 
 it may be said that the simplest form, that which we have termed "single-track block- 
 ing," and which is used on various Western roads, is cheap, readily adapted to nearly 
 all roads of light traffic and generally to a road of any traffic, and is an absolute necessity 
 to any road which would avoid the well-known difficulties connected with protecting 
 trains from rear collisions by flags, lamps and torpedoes. The drawbacks to the use of 
 this system are the liability of officers and trainmen to expect more of it than it can, in 
 the nature of things, perform. Where there is only one signal at a station, while there 
 may be two or a half dozen switches, and where the distance between stations is so long 
 that trains must sometimes be allowed to follow each other within the same block, the 
 block system is not responsible for and cannot be charged with the various dangerous 
 contingencies that may arise. In fact, the system can in such cases be used only a part 
 of the time. The danger is that when or where it is not in use the men are liable to 
 assume, perhaps unconsciously, that it is in use, instead of adopting other precautions 
 such as would be taken had the block system never been heard of. 
 
 The block system in use on the Pennsylvania is simply a more careful and system- 
 atic application of this same principle. The only reason why scores of roads do not 
 imitate the Pennsylvania seems to be simply a lack of the courage requisite to incur a 
 considerable expenditure for first cost, and a failure to appreciate the fact that short 
 
 * Drawings and detailed description are given on page 62. 
 
BLACK S MECHANICAL BLOCK SIGNAL. 21 
 
 sections are not essential to the utility of the system. Officers too readily assume that a 
 large expenditure for operators (to give their exclusive attention to block signaling) is 
 indispensable, whereas the system in many cases could.be advantageously adopted with- 
 out going to that expense. 
 
 The Sykes system is admitted by the best judges among those who have used it to 
 be valuable, but its cost both for introduction and maintenance is rather high. Its use 
 is still limited to the busiest lines, even in England. Its value is largely or wholly 
 neutralized if permissive blocking is practiced, and very few roads have progressed to 
 the point where they wholly forbid that practice. 
 
 The special merit of automatic signals lies in the possibility of reducing the run- 
 ning expenses 90 per cent., more or less, below the cost of a man-operated system. As- 
 suming that such imperfections as they are now burdened with can be eliminated, and 
 especially that they can be made perfectly self -detecting, the question concerning their 
 general availability is, will the unnecessary stops caused by the failure of apparatus and 
 extraneous accidents delay traffic to such an extent as to compel a resort to the regular 
 block system (with operators at each station) ? Some English experts claim that this 
 will be the ultimate outcome. It is to be remembered, however, that they judge largely 
 by experience on the most crowded lines, and that by reason of the low rates of wages in 
 that country the question of operating expense does not assume the importance that at- 
 taches to it here. Granting this point, it still remains true that many hundred miles of 
 road in this country now carry a volume of traffic sufficiently large to demand a block 
 system of some sort, while, at the same time, they will not for many years be used by a 
 sufficient number of trains to make five-minute delays intolerable. 
 
 It may therefore be said that there are three classes of roads in this country which 
 afford fields for three different kinds of block signaling respectively : 
 
 1. Those parts of large roads which lie in the vicinity of the principal cities, where 
 yards are located very close together, and where the traffic is very heavy these need 
 man-operated signals, with short blocks. 
 
 2. Many double-track roads of less importance need the block system, but cannot 
 afford to establish stations as close together as is necessary to run trains at short in- 
 tervals. This is the field for automatic signals. 
 
 3. The roads of thin traffic and with long stretches between stations cannot afford 
 special block operators, and cannot afford even the few hundred dollars per block neces- 
 sary to establish an automatic system. These should block by means of their regular 
 station operators. This is especially true of roads where, by reason of heavy grades, 
 prevalence of fogs, or other conditions, the speed of trains cannot be maintained at a 
 reasonably uniform rate, and of those located in northern climates where flagging in 
 winter is dangerous both to the men and the trains. 
 
APPENDIX. 
 
 THE SYKES BLOCK-SIGNAL APPARATUS. 
 
 [October 3, 1890.] 
 
 What is known as the " Sykes System" is the application to an ordinary manual 
 block system of certain electrical and mechanical devices which insure that the signal 
 governing the entrance to a given block cannot be cleared until the preceding train has 
 passed out of it and the operator at the end of the block has given his consent. 
 
 Under the practice of The Union Switch & Signal Co., which has proprietary control 
 of the Sykes system for this country, these results are secured by the use of a Sykes lock 
 instrument, an interlocking relay, and a short* insulated section of track, with proper 
 metallic circuits connecting these ; also a bell wire or telegraph line for communicat- 
 ing between adjacent block stations. 
 
 The Sykes lock instrument is placed in the operator's office immediately over the 
 lever by which he controls his signal. The interlocking relay is located in any conven- 
 ient place, usually in a closet. The insulated section of track is located at the entrance 
 of the block, and is usually two rails in length (about 60 ft.). The bell- wire push-but- 
 'tons are placed near the signal lever and the Sykes instrument. 
 
 The connection and relative operation of the signal lever and the Sykes lock instru- 
 ment are illustrated by Figs. 1 and 2, which show two operating levers and two lock in- 
 struments, the regular equipment of a block station working blocks in both directions. 
 
 The important connections are the lock bolt, the lock bar, the lock rod, and the 
 plunger rod. Normally, the signal lever being home and the signal at danger, the lock 
 bolt is entered in a hole in the lock bar, and both plunger rod and lock rod are in their 
 extreme upward positions, displaying the words "Clear" and "Locked" (Fig. 1), which 
 indicate to the operator that the plunger is free to be worked (and thus unlock a signal 
 lever at an adjacent block station), but that his own signal lever is locked and cannot be 
 moved. 
 
 The operation in practice is as follows, everything being normal levers home, signals 
 at danger, and tracks unoccupied : If the operator desires to allow a train to enter one of 
 
THE SYKES BLOCK-SIGNAL APPARATUS. 
 
 Fig. 1. Levers and Lock. F g. 2. 
 
24 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 the blocks which his signals control, he notifies the operator next in advance (by bell 
 wire or telegraph line); the advance operator, if everything is all right, responds by 
 u plunging" on one of his instruments. That has the effect of releasing the lock rod at 
 the first station; the lock rod falls by its own weight, and in so doing withdraws the 
 lock bolt from the lock bar. The operator then pulls over his lever and clears his signal; 
 this movement forces the lock bar forward, and, through the action of the inclined plane 
 and roller, also forces the lock rod upward to its normal position, where it is automatically 
 held until again released by the operator at the station in advance. This upward move- 
 ment of the lock rod leaves the lock bolt free to be sprung into the hole in the lock bar, 
 when the lever is again returned to its normal position. 
 
 The only function of the Sykes system so far alluded to is that by which the operator, 
 on request of an adjacent operator, may "plunge" and thus release the latter' s signal 
 lever. The additional and important function of the combined apparatus is to prevent an 
 operator from plunging a second time until the train for which the preceding operator 
 desired to clear his signal has passed into, through, and out of the block in question. 
 This result is secured by the combined action of the Sykes instrument, the interlocking 
 relay, and the insulated section of track. 
 
 The Sykes instrument is shown in position and in connection with the operating 
 signal levers in Figs. 1 and 2, as above referred to. It is shown in detail in Figs. 3, 4, 5, 
 6 and 7. 
 
 Fig. 3 exposes those parts which are employed in unlocking the signal lever. When 
 an adjacent operator plunges he simply passes a current through the electromagnets 21 
 of his neighbor's instrument, thus attracting and raising armature 22, and imparting a 
 slight rotation to balanced lever 23 about its centre 24. This releases trip 25 (which then 
 is free to rotate about its centre 26) and permits the lock rod 35 to drop by its own weight, 
 and thus unlock the operator's signal lever as explained in connection with fig. 2. 
 
 Fig. 5 best illustrates the action and results of plunging. When plunger 27 is push- 
 ed in, cross-bar 28 (figs. 6 and 6) is raised, breaking one circuit and completing another 
 by means of springs at 29 (figs. 3 and 6). When plunger 27 is forced out to its original 
 position by the action of spring 30, an electrical contact is effected, by springs, at 31 
 (figs. 5 and 6). One end of cross-bar 28 is free to rotate, in one direction only, about 90 
 degrees, but is restored to its original position by a small contained spiral spring ; thus, 
 in fig. 5, the small projection on left of cross-bar 28 causes it to revolve on the upward 
 stjoke, and no contact at 31 is effected. But on the downward stroke the same projection 
 presses the flat springs to the left and effects the desired contact. In order that this con- 
 tact may not be too brief (an electromagnet at adjacent station is thereby charged) the 
 downward stroke of cross-bar 28 is retarded by dash pot 32, which has a small vent hole 
 below. 
 
 In plunging there is also an important mechanical interaction between plunger rod 33 
 
THE SYKES BLOCK-SIGNAL APPARATUS. 
 
 BLOCKED 
 
 --:i 
 
 fclJ 
 
 1' 
 
 - 
 
 I 
 
 Fig. 4. 
 
 \ 
 
 1 
 
 Fig. 6 Section on A B. SYKES SYSTEM DETAILS. Fig 7 Section on E F. 
 
AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 and trip rod 34 (figs. 4, 5 and 7). Plunger rod 33 has attached to it a pin 3o (figs. 4 and 
 5. Plunger 27 has attached to it a side piece 37 (figs. 4, 5 and 7). Trip rod 34 has attached 
 to it a pawl piece 38 and a sliding block 89 (figs. 4, 5 and 7). Normally, as shown in 
 fig. 2, the plunger rod is in its extreme upper position. In this position, as shown in 
 fig. 5, the plunger rod 33, by means of pin 36, supports pawl piece 38 and the trip rod 
 34, thus displaying the word "Clear" (fig. 4) to the operator, which signifies that his 
 plunger is free or clear, and may be operated in response to request from adjacent oper- 
 ator. 
 
 When the plunger 27 is pushed in, the side piece 37 forces pawl piece 33 off from 
 pin 36, and trip rod 34 drops ; at the same time sliding block 39, previously sup- 
 ported by the pressure of pawl piece 38, drops until it rests on side piece 37, which is 
 tnen under it. When the plunger comes out, on its return stroke, sliding block 39 falls 
 still farther on to the extension or foot of side piece 37, preventing the plunger 27 from 
 
 being again forced in. When the trip rod 34 
 dropped the word "Blocked" was displayed 
 instead of the word "Clear," signifying to the 
 operator that he could not plunge again 
 until certain conditions had been complied 
 with. To release the plunger, the trip rod must 
 be lifted back to its normal position. This can 
 be effected only by reversing the operator's 
 signal lever and again putting it in the for- 
 ward or home position. When the signal lever 
 is reversed, the plunger rod is drawn to the 
 bottom of its stroke (see fig. 2). In its down- 
 ward stroke the pin 36 on plunger rod 33 (figs. 
 4 and 5) forces aside pawl piece 38, which 
 then snaps back in normal position above pin 
 36, resting on and supported by it. When 
 the signal lever is returned to its normal or 
 home position, and the plunger rod is forced 
 upward, it lifts with it trip rod 34 and restores 
 Fig. 5 -Section on L N. it to i is normal position ; the pressure of 
 
 pawl pieces 38 against sliding block 39 keeps them in the same relative position as when 
 the trip rod was down, and in this way the sliding block 39 is lifted up out of the way 
 and the plunger is ready to be operated again, as is indicated by the word " Clear," which 
 is again displayed. It thus appears that when an operator plunges he is mechanically 
 prevented from plunging a second time until he reverses his lever and again restores it to 
 its normal or home position. 
 
THE SYKES BLOCK-SIGNAL APPARATUS. 
 
 27 
 
 It will no v be shown that when an operator plunges and releases the lever of the 
 operator next to the rear, the electrical circuit thus utilized is automatically broken, and 
 cannot be made complete again until the train for which the preceding operator 
 desired to give a clear signal has passed over the intervening block. The automatic action 
 
 Fig. 8. 
 
 Fig. 9. 
 
 Arrangement of Circuits. 
 NOTE. 50 is the lock-wire ; the wire below is the common return for lock and bell circuits. 
 
 THE SYKES SYSTEM. 
 
 of breaking and restoring the lock circuit may be understood by reference to figs. 8 and 
 9, which indicate the relation, connection and mutual interaction of the Sykes instrument, 
 the interlocking relay and the insulated section of track. Two tracks are shown in the 
 plan, but only such instruments and circuits are shown as are necessary for the control of 
 trains in both directions, between two adjacent block stations 8 and 9 For a continuous 
 system two sets of instruments at each station are required. Two different views of the 
 top of the Sykes instrument are given in each, fig. 8 and fig. 9, so as to better display the 
 contacts and circuits which are broken and established by the up stroke and the down 
 
28 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 stroke of the cross-bar 28 which takes its motion from the plunger 27, as has been pre- 
 viously explained. The interlocking relays are shown in their normal condition, the upper 
 magnets dead (owing to the break in the circuit at 31), and their armatures down ; the 
 lower magnets charged from the track battery, and their armatures up. 
 
 Suppose operator 8 desires to clear his signal to allow a train to pass 8 towards 9; he 
 uses his bell wire or Morse instrument to request operator 9 to release his lever. If the 
 road is clear and everything normal, operator 9 responds by plunging; this forces cross- 
 bar 28 upward, and the spring 30, restrained by the dash pot, slowly forces it down 
 again. Kef erring to fig. 9, the upward stroke of cross-bar 28 (lower view) breaks the 
 contact between spring 29 and the right-hand contact screw, and effects a contact be- 
 tween spring 29 and the left-hand contact screw; this completes circuit 29, 40, 41, 66, 42, 
 62, 43, 44, 45, 46, 47, 29, 48, 49, 50, 61, and charges the electromagnets at station 8, re- 
 leasing the operator's lever at that point and permitting him to clear his signal for the 
 entrance of the train into the block, which extends from station 8 to station 9. The 
 action of tripping the lock rod is instantaneous, and requires only a momentary charging 
 of the electromagnets. Referring to the upper view in fig. 9, it is understood that that 
 end of cross-bar 28 rotates during the upward stroke, and effects the contact 31 only on 
 its downward stroke ; this contact 31 has the effect of breaking the circuit previously 
 established. Contact 31 completes circuit 31, 52, 54, 53 ; electromagnet 54 thus charged 
 attracts its armature, which is automatically hooked up by the spring hook on leg of 
 armature of electromagnet 55 ; this breaks the original circuit at 56, and operator 9 can- 
 not again unlock a lever for operator 8 until the train has passed over, and off of, the in- 
 sulated section of track 57, 58. When no train is on the insulated section, circuit 59, 57, 
 61, 55, 60, 58 is complete, electromagnet 55 is charged, its armature held up and con- 
 tact 62 kept good. When a train passes over insulated section 57, 58, the first axle 
 establishes a short circuit with battery 59 and devitalizes electromagnet 55 ; its armature 
 drops, destroying contact 62, but allowing armature 54 to drop also, thus restoring con- 
 tact 56, which was previously broken. Contact 62 cannot be restored, however, as long 
 as the short circuit at the insulated section of track exists. Each axle of the train is a 
 possible short-circuit path, so that the entire train must have passed over and off the in- 
 sulated section before the short-circuiting ceases ; then the original circuit 59, 57, 61, 55, 
 60, 58 is re-established, electromagnet 55 is charged, its armature attracted, and contact 
 62 restored and maintained ; and the original lock circuit by which operator 9 plunged 
 and released a lever in station 8 is again made complete. Operator 9 had to reverse his 
 signal lever in order to admit the train to the insulated section of track, and in putting 
 his lever home or into its normal position he unlocked his own plunger, so he could again 
 plunge for 8. If the train had passed completely over the insulated section of track, the 
 lock circuit would have been restored and his plunging for 8 would be effective, and he 
 would release his lever for another train. 
 
THE SYKES BLOCK-SIGNAL APPARATUS. 29 
 
 In the case of the first block station in a series it is necessary to introduce some de- 
 vice by which each train entering that block under a clear signal will automatically set 
 that signal to danger again. This is accomplished by the use of what is known as an 
 " electric slot " in the connections between the operator's lever and the signal blade. 
 This electric slot involves an electromagnet normally charged and taking current from 
 a battery through the two rails of an insulated section of track : when the current is 
 passing through the electromagnet the connection between the operating lever and the 
 signal blade is preserved complete, and the operator has full control over his signal. 
 When a train reaches the insulated section of track, however, the first and all other 
 axles establish a short circuit ; the electromagnet is thus discharged, the connection 
 broken and the signal set to danger by the action of its counter- weight. The operator 
 cannot again get control of his signal except by putting his signal lever normal ; when 
 that has been done and the circuit restored by the complete passage of the train, the con- 
 nection between lever and signal blade is automatically restored and the operator again 
 has control of his lever, subject, of course, to the action of the operator next in advance, 
 who may plunge and unlock him if all the conditions are favorable. 
 
 It has been shown : First, that an operator by plunging may unlock the signal lever 
 of the preceding operator. Second, that in so doing his plunger is automatically locked 
 up, and can be unlocked only by reversing his lever and putting it home again (in other 
 words, giving a clear signal and then the danger signal). Third, that the lock circuit 
 utilized by the inward stroke of the plunger is automatically broken during the outward 
 stroke of the plunger, and can be restored and made complete only by the passage of the 
 train onto, over and off of the insulated section of track. 
 
 The resultant effect of the combined apparatus under these several conditions insures 
 that "the signal governing the entrance to a given block cannot be cleared until the last 
 train which received a clear signal to enter that block has passed oat of it, and the 
 operator at the end of the block has given his consent." 
 
AUTOMATIC SIGNALS ON THE BOSTON & ALBANY. 
 
 [June 24, 1887. Written by G. W. Blodgett.] 
 
 This road has had a larger experience with electric signals than almost any other in 
 the country. A large portion of the road is now equipped with such signals, and addi- 
 tions to the plant are constantly being made. A brief description of the apparatus em- 
 ployed will be followed by some account of its practical working and the results obtained. 
 
 The road has double track throughout its whole length, except the first ten miles, 
 where there are four tracks. This short distance and some other detached portions of 
 the road have continuous overlapping blocks. The remaining applications (with a single 
 exception) are "station blocks" so called; that is, there is a signal each side of the 
 station about one-half mile distant, which is connected with every switch in the track to 
 which the signal belongs, and the function of which is to protect a train while standing 
 at a station or switching. As traffic increases, or for other reasons it becomes advisable 
 so to do, these applications can be made parts of a system of continuous blocks with no 
 other change than simply overlapping the sections. 
 
 For the first mile from Boston all trains are of the same class and run on the same 
 tracks. After that they diverge, express passenger and freight trains running westward 
 on track No. 1 and eastward on track No. 2, while suburban passenger trains travel 
 westward on track No. 3 and eastward on No. 4 for 10 miles, to the end of the four-track 
 section. A large part of the suburban trains go no farther, and beyond this point all 
 trains again run on two tracks. 
 
 In the first mile the signals are about -fc mile apart, then % mile for two miles more 
 (which includes a large yard for outgoing and another for incoming freight trains, be- 
 sides two important junctions), and then about a mile apart as far as the continuous 
 blocks extend. 
 
 The larger part of the road is equipped with the rail circuit clockwork signals of the 
 Union Switch and Signal Co. The first applications of this signal were made in l8b2, 
 when 6 blocks were put up as an experiment. 
 
 BLOCK SIGNALS WITH BAIL CIRCUITS. 
 
 The road is divided into sections of varying lengths according to the amount of traffic 
 or local circumstances, as above mentioned, each of which sections has a signal at the 
 
AUTOMATIC SIGNALS ON THE BOSTON & ALBANY. 
 
 31 
 
 ntrol/ing Si&ial. 
 
 * '^j~ t ' " - .- 
 
 ytSS^ginning (operated by clockwork and a 
 
 position indicat- 
 train enters the sec- 
 poijn,, and restored to that 
 the train leaves the 
 
 section. Each section is electrically in- 
 sulated from those before and behind it. 
 At the end farthest from the signal is an 
 electric battery giving a constant current, 
 which is connected to the track, one pole 
 to each rail ; at the signal end of the sec- 
 tion the coils of a relay are connected to 
 the rails in like manner, so that there is a- 
 constant flow of electricity from the battery 
 through one line of rails to the relay, 
 thence through its coils to the other line of 
 rails, thence by these rails back to the bat- 
 tery. This relay controls the local circuit 
 (so called) of another battery connected to 
 an electromagnet in the signal, which 
 governs the motion of the clockwork 
 operating it. 
 
 Fig. 1 shows the track clear, relay It 
 holding the local circuit through 8 closed. 
 
 The successful operation of signals by 
 means of rail circuits requires that the 
 electromotive force (or pressure, as it may 
 be considered) of the battery connected to 
 the rail shall be so small that there will be 
 only slight leakage from the rails even in 
 wet weather. It is found advisable, there- 
 fore, to include in the track circui only the 
 coils of a relay of low resistance, and make 
 this relay open and close the circuit for 
 another battery of any size desired, which 
 shall operate the signal mechanism. There 
 are then two electric circuits for each 
 signal ; one through the rails control- 
 ling another through the signal. Both 
 
32 
 
 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 circuits are normally closed ; that is, there is a constant flow of electricity through the 
 whole length of each. The clockwork is so constructed that when the current is passing 
 through the magnet the signal is held in the position of "all clear" or "safety." When 
 this is interrupted, the signal makes one-fourth of a revolution to the position indicating 
 "danger" or "stop." ' 
 
 When a train enters the section the current in the rails takes the path of least resist. 
 ance, through the wheels of the train instead of the relay magnet. The armature of this 
 magnet then falls off and opens the circuit of the battery controlling the signal, and this 
 takes the " danger" position as Jong as the section is occupied. When the rear of the 
 train passes out of the section the circuits are again closed and the signal shows clear ; 
 but if so much as a pair of wheels remains in any portion of the section the signal will 
 continue to show danger until the obstruction is removed. The same effect is produced 
 whenever, for any cause, either circuit is interrupted if, for instance, the battery fails, 
 
 iMile 
 
 Signal S 'goes to Danger at A Signal S z goes fo Danger afC. 
 
 u n n U"AH clear" D // a // // "A II clear >, />dofnext 
 
 over/ap. 
 
 or the wires are broken, or the clockwork is run down, the signal shows " danger." If a 
 rail in the track is broken, and the parts separated by so much as ^o of an inch, the 
 signal will be at danger. Many instances of this kind have occurred, and not a few where 
 the indications was most timely. Indeed, the use of automatic signals has often dis- 
 covered broken rails which might have remained in the track a long time without such 
 displacement of parts as would have rendered them liable to detection by the ordinary 
 methods of inspection. 
 
 To make perfect electrical connection between the rails, a wire extends past each 
 joint, the ends of which are connected to the two rails by a tight-fitting pin in a hole 
 drilled in the flange of the rail. While rails are new and fish-plates tightly screwed up, 
 this is not absolutely needed, but as soon as they begin to rust there is trouble if the 
 rails be not connected by the wires. Signals have worked for several months with an 
 
AUTOMATIC SIGNALS ON THE BOSTON & ALBANY. 33 
 
 unwired track, but ordinarily they will do so only for a few weeks, even if rails and fish- 
 plates be perfectly new. 
 
 Figs. 2 and 3 show the circuit breakers connected with each switch, and the wire con- 
 nections by which the rails of side tracks are included in the track circuit, for the pur- 
 pose of keeping signals at "danger" until trains entering the side track are fully clear of 
 the main line. In fig. 3 points X and Fare connected by the curved flat brass which is 
 held against them by a spring, and the two rails are thus electrically connected, the same 
 as when a pair of wheels is upon them. In fig. 2 the switch rail T having been with- 
 drawn from the main- track rail, has pushed the brass connections away from X and J, 
 breaking the connection between the opposite rails.* 
 
 When the block signals are continuous that is, with no spaces between the sections 
 the safety of trains folio wing each other at short intervals is very greatly increased by 
 making the sections overlap each other. This causes a signal to remain at danger until 
 the train has passed a certain distance (usually about 1,000 ft.) beyond the next signal. 
 While the train is running this short distance there are two red signals behind it, one at 
 the beginning of the section where the train is, and the other at the beginning of the pre- 
 ceding section. 
 
 The arrangement of circuits which accomplishes this is shown in fig. 4, which 
 assumes the block to be one mile in length. The principal track circuit for the signal 
 S 1 passes through the armature of a relay, J? 1 ; the coils of this relay are in a wire circuit 
 connected with the battery J5 2 , which is controlled by a relay, -S 2 , placed at the end of the 
 overlap. The coils of this last relay are connected to the rails of the overlap. It will 
 be seen that a train on any portion of this short section will operate the relay R z and 
 consequently -S 1 , and set both signals. Hence, so long as an engineman does not pass a 
 red signal he can never approach a preceding train nearer than the length of the overlap. f 
 
 Signals are placed a short distance (usually about 200 ft.) beyond the beginning of 
 the section, in Border that an engineer may see the signal operate for his train. Should 
 it fail to do so, he is to stop, the same as for a danger signal, and proceed only as the 
 way is known to be clear. The engineer of every train stopped by a signal must with- 
 out delay report the stop and the cause if known (on blank cards provided for the pur- 
 pose) as, for instance, a preceding train in section or an open switch. If the cause be 
 not apparent to the engineer, he simply reports " cause not known," and it is put in the 
 hands of a repairman to investigate. When the latter has ascertained the cause (for 
 instance, a broken rail, failure of battery, derangement of some part of the apparatus or 
 other cause not at first apparent) he returns the card with his explanation indorsed 
 
 * In figs. 2 and 3, B is the battery at one end of the block section, and B he relay controlling the signal at the 
 other end. The switch S (or any number of switches) may be at any point between these two. 
 
 t When a train is on the section C D relay 72 2 is demagnetized, thus opening both the circuits through battery 
 jB 2 ; S 2 and S 1 then both show " danger.' 1 When a train is in the section beyond D, relay B 3 is demagnetized, hold- 
 ing S' 2 to " danger." Signal S* stands 200 ft. from C and 800 ft. from D. 
 
34 AMERICAN PRACTICE IX BLOCK SIGNALING. 
 
 thereon. If lie cannot find out the cause, he returns the card with that statement, and 
 it is usually never ascertained. There is a small fraction of one per cent, of such stops 
 at signals. Jt is quite certain that some of these are, due to previous trains, open 
 switches or other legitimate causes, but in the absence of positive proof they are not so 
 classified. Sometimes employes needlessly cause stops of trains at signals, and to save 
 themselves the consequences carefully conceal the fact, which is not always afterwards 
 discovered, and when this is the case such stops have to be reported " cause unknown." 
 A careful record is kept of all stops and their causes, and every month a debit and credit 
 account is made up of the operation of the signals on each division of the road, which 
 shows at a glance what proportion of stops is due to neglect of employes, defective 
 apparatus, unavoidable causes, etc., as well as all legitimate stops. 
 
 The only stops credited to the system are those due to (1) previous trains in section, 
 
 (2) open switches, (3) broken rails, (4) repairing track, (5) [sometimes] using single 
 track, (6) cars left on turnouts too near the main track. Lost motion in switches, broken 
 track wires, Of any other failure of the track circuit is usually charged to the neglect of 
 trackmen ; those due to failure of batteries, corrosion of apparatus, and certain other 
 derangements to neglect of signalmen, so that the blame may be placed where it belongs. 
 Employes are held to a strict account for all avoidable stops caused by them, and the 
 ratio has been reduced to one surprisingly small. To the debit side of the account is 
 charged all such stops as are caused by defective construction of any part of the appa 
 ratus. The number of these has heretofore been unreasonably large. First-class 
 mechanical construction costs but little, if any, more than such as would not pass inspec- 
 tion in any good machine-shop, and gives immeasurably better satisfaction in service. 
 There remain a certain number of stops due to "unknown" causes, and certain stops due 
 to climatic conditions, unavoidable accidents to the apparatus, derailments, lightning, 
 etc., which are grouped by themselves under the head of "accidental." Longer experi- 
 ence will doubtless suggest ways in which the number of these may be diminished. 
 
 The severe tests of actual service under all varieties of climate and temperature show 
 that the perfect railroad signal has not yet been invented. In each system certain deficien- 
 cies, or failing cases, must be provided against in order that the signal may work regu- 
 larly or be used with safety. 
 
 The most dangerous error an automatic signal can make is to show clear when a train 
 is in the section. The Union signal is, perhaps, more free than any other from such fail- 
 ures, but they are by no means unknown. The cases which have come under my own 
 observation have been due to (1) a failure of the track circuit relay to drop its armature 
 when the current was shunted out of the magnet ; (2) too much battery on the rail circuit; 
 
 (3) crossed wires between the signal and overlap relay ; (4) failure of the signal magnet to 
 release the clockwork. when the circuit was opened ; or (5) the sticking of some mechanical 
 part of the apparatus which should have moved freely. Of these the first is by far the 
 
AUTOMATIC SIGNALS ON THE BOSTON & ALBANY. 35 
 
 most common, except in ice and sleet storms ; like the fourth, ii is usually due to fixed 
 magnetism in the cores or armature of the relay, and could be prevented by the use of 
 better iron in their construction. The second cause is the fault of the signalman, and the 
 third may also be. This last may be remedied by a different arrangement of circuits, 
 which the Boston & Albany will adopt in all new work. The fifth may, or may not, be the 
 signalman's fault. A rainstorm sharply followed by freezing weather wiil stick every 
 signal in an hour in the position it happens to be at the time. A heavy fall of damp snow 
 will sometimes (but rarely) do the same thing. 
 
 Another failing case of bad repute is when the signal stands clear with a switch open 
 This usually shows a faulty connection in the switch-box. There is no way (with the ar- 
 rangement of circuits shown in figs. 2 and 3) to know beforehand whether opening the 
 switch will set the signal. A far safer connection is shown in figs. 5 and 5a, where the 
 current in the rails is made constantly to pass through the switch -box, when the switch 
 is on the main line. The switch-box must be in good order or the current cannot pass. 
 All the switch connections on the Boston & Albany are now being changed to this 
 style. * 
 
 A multitude of causes may make a signal stand at "danger" when no train is in the 
 section or switch open. Any derangement of the apparatus (except the special ones enum- 
 erated above) or interruption of the rail circuit by displacement of the track or other- 
 wise will do this. Stops thus caused are principally a matter of annoyance and expense. 
 They do not introduce an element of danger, except that, if very frequent, they would 
 tend to make enginemen careless of the indications of the signal when it did warn of ex- 
 isting danger. Though there may be a considerable number of such stops in a month on 
 some divisions of the road, it is found, when account is taken of the number of trains run- 
 ning, that the ratio of failures to number of operations is very small. 
 
 The cost of operating each Union signal, including superintendence, was, during the 
 year ending Oct. 1, 1886, about $75.69, or $(3.31 per month. 
 
 There are roads equipped with Union signals which claim to have fewer unneces- 
 sary s'ops per signal than the Boston & Albany, and to run their signals at less expense, 
 but they have for the most part no overlapping sections (which would very greatly 
 complicate their application); their trains run at longer intervals, and in some cases the 
 account of stops and their causes is not so carefully kept. 
 
 * The current in the rail A B, when the main track is unbroken, must normally pass through the points X Y'> 
 when the switch is moved, the connection between X and Fis broken and the opposite rails connected (as by a pair 
 of wheels) through Fand Z. 
 
THE WESTINGHOUSE SYSTEM OF PNEUMATIC INTERLOCKING. 
 
 [December 21, 1888.] 
 
 Before entering into a description of the operation of this system, it will be neces. 
 sary to describe the construction and operation of its several parts. There are a steam- 
 generating boiler, an air compressor, and a condensing tank through which the air must 
 pass before entering the main air pipe. This deprives the air of any moisture which it 
 may have had originally, or collected in passing through the heated cylinder of the com- 
 pressor, and prevents its collecting in the valves or cylinders where it might interfere 
 with their operation. 
 
 Each signal blade is connected directly to a pneumatic cylinder, the pressure to 
 which is controlled by a small valve actuated by an electro-magnet, which in turn is con- 
 trolled by the operator in the cabin. The air supply to each of these cylinders is taken 
 from a cylindrical tank at the bottom of the post, all of which are connected directly to 
 the main air-pipe; consequently, all signals have, at all times, the full pressure of the 
 compressed air, right at their cylinder valves. The control of this pressure by the elec- 
 tric valve and the valves by the operator will be treated later. 
 
 From this same air pipe pressure is conducted to the switch valves, where it is stored 
 in a reservoir, which forms the valve support, and is provided with a cap or plug with 
 three ports formed in it, and a D-valve seated over them, exactly as is done in a steam 
 engine. Encasing this D-valve and its ports (see fig. 4, A] is a hollow cap fastened to 
 the reservoir and connected with, it so that the full pressure from the reservoir is at all 
 times in it, and consequently on top of the D-valve, holding it seated. One of these 
 ports connects directly to the open air; this is the centre one, while the right and left 
 ones connect each to one end of the cylinder operating the switch. This D-valve is so 
 constructed that it is impossible to admit pressure to one of the ports before having con- 
 nected the other with the exhaust. Jt is, therefore, very evident that it is impossible to 
 have pressure on both sides at one time, and also that the full air pressure is always 
 holding the switch in the position last moved to. With this description, it will be clear 
 how the pressure can be changed to one end or the other of the switch cylinder. 
 
 The switch movement (fig. 5) consists of a long cylinder, 5 in. in diameter, provided 
 with two flanges for securing it to the ties, and two studs or trunnions on the opposite 
 side forming pivots for an arm operating the lock and detector bar of the switch, a piston 
 
THE WESTINGHOUSE SYSTEM OF PNEUMATIC INTERLOCKING. 37 
 
 composed of a plunger packed at each end and formed into a rack between, engaging 
 into a pinion which rotates about f of a revolution to each movement of the rack piston. 
 This pinion is keyed fast to a shaft on which a crank is formed, and turns with the pinion. 
 To this crank the operating rod of the switch is connected, and also a link joining it to 
 the rod, already mentioned, operating the lock and detector bar. It will be noticed that 
 this crank stands beyond the centre line of its axis continued through the centre of the 
 switch connection, and that it might move a corresponding distance to the right of this 
 centre line before giving any appreciable motion to the switch itself, on account of the 
 small arc thus described. It is the peculiar arrangement of this crank that renders this 
 movement so simple in effecting the motion of the detector bar and the preliminary un- 
 locking of the switch and a final motion of the bar and locking of the switch after it has 
 been moved. By reference to the cut it will be clearly seen that the movement, when 
 normal, holds the switch locked in one of its two positions and the detector bar below 
 rail level. Also that the first motion to take place is the simultaneous raising of the bar 
 and unlocking of the switch. The lock bolt thus operated is of sufficient length to have 
 been fully withdrawn from the hole in the lock rod of the switch before the motion of 
 the crank is imparted to the rod moving the switch. It is also, for the sake of simplicity, 
 allowed to travel still farther from the lock rod during half the motion of the switch, 
 when it again approaches the rod, and by the time it ar.rives at the bar again the switch 
 must have moved so as to bring the second hole in the lock rod opposite the pin before 
 it will become locked, and indicate it in the cabin in a way to be yet described. On the 
 casting forming a guide for the lock rod, directly in front of the locking pin, is placed a 
 circuit-controlling device (fig. 5, A), which, when the lock pin has entered the lock rod 
 of the switch, holds the circuit open, and when the pin is withdrawn permits it to be- 
 come closed. The function of this device will be described later. 
 
 Having described the construction of the switch valve, we will explain the manner 
 of operating it. On each side of the hollow chamber or cap encasing the D- valve, (fig. 4, 
 B) are two small cylinders containing pistons, and a stem from each extending through 
 a stuffing-box into this cap or chamber and resting one against each end of the D-valve. 
 Connected to the heads of these two cylinders are two small pipes which run directly into 
 the cabin and to the machine, where they run to the ports of a three-way cock operated 
 by the switch lever (fig. 3, A). This cock is identical in operation with the D-valve, in 
 that but one port can have pressure on it at a time, the only difference in its construc- 
 tion being that its seat is cylindrical, or rather, conical, instead of flat on a horizontal 
 plane, as in the D- valve. It is evident that pressure must be on one or the other of 
 these small cylinders, (fig. 4, B), at all times. It is also evident that since the D-valve is 
 set between these pistons, any motion of them will be imparted to the valve also, and 
 that the pressure on each piston acts against the other one through the valve, thus mak- 
 ing the two pistons and the D-valve act as a solid plunger in a single cylinder. The 
 
AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 way connected with the piston stems, but simply guided between 
 rdlcfaing compensation for wear on the seat and under face of valve. It is 
 fco throw two or more switches by the same lever, as is the case with a cross- 
 with movable frogs. In such cases, unless they be too far apart, but one 
 , and each one of the two or three switch cylinders is connected directly to it in 
 the same manner as is done with a single one. It is perhaps necessary to explain now the 
 necessity of this valve being interposed between the cock at the machine and the switch 
 cylinder, since it will be evident that the operation would be the same if the pipes from the 
 machine went directly to the- switch cylinder. While this is the case, still a serious 
 feature in this arrangement prevents its adoption. Owing to the long distance it is often- 
 times found convenient to operate switches from the cabin with this system, and the 
 consequent long line of pipe necessary to be filled with air and exhausted at every move- 
 ment of the switch, it is found not only more economical in saving air, but very much 
 more efficient in operation to place this valve as close to the switch or switches operated 
 as possible, and fill the small pipes leading from the small cylinders of the valve to the 
 machine with water in summer and chloride of calcium, alcohol or some other non- 
 freezing liquid in winter. When this is done and the air from the machine cock is 
 admitted on top of it, it instantly acts against the pistons of the small cylinders of the 
 switch valve, fig. 4, (7, since the liquid will not compress, but acts as a solid rod. This 
 would not be the case were the air used alone, since it would require some time to com- 
 press to the pressure necessary to move the valve, and waste a corresponding amount by 
 connecting the opposite side to the exhaust. In order to compensate for loss of liquid 
 by evaporation or leak, an automatic filler is attached to all hydraulic pipes, fig. 9, which, 
 normally, is opened with all of them not having pressure on them, and automatically 
 closed from them, by means of a check valve, when the pressure is admitted on top of 
 the water in them. This insures a full supply of liquid in these pipes at all times, and 
 consequently a quick action of the switch valve. 
 
 The cylinder operating the signal will now be described. As before stated, this cylin- 
 der, fig. 8, has the pressure right at the valve controlling its admission to it. This is 
 also, controlled by an electromagnet, the circuit of which is controlled by the operator 
 tb rough the machine. The piston of this cylinder is connected with the blade either 
 directly or through a balance lever, fig. 10, and in its normal condition is in the upper 
 end of the cylinder, being held" there by the counter- weight blade or balance lever. In' 
 this position of the piston the blade is in the horizonital or danger position, and can only 
 be moved from that position by the admission of air on top of the piston, thus depressing 
 it sufficient to give the blade the proper angle (60) indicating safety or caution, accord- 
 ing to the nature of the signal. This is accomplished by a small pin valve, fig. 8, B, 
 which normally holds the pressure closed from the cylinder, and the cylinder open to 
 the exhaust. When operated by the electromagnet becoming charged from a current 
 
THE WE3TINGHOUSE SYSTEM OF PNEUMATIC INTERLOCKING 
 
 PLAN OF SWITCH & CONNECTIONS 
 
 SWITCH MOVEMENT. 
 
 THE WESTINGHOUSE SYSTEM OF PNEUMATIC INTERLOCKING. 
 
40 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 sent through it by the operator, the reverse condition takes place; i. e., the pressure is 
 admitted to the cylinder on top of the piston, and the exhaust is closed completely. The 
 pressure thus confined depresses the piston and operates the blade. The instant the cur- 
 rent is broken in the magnet the armature is released and the air again unseats the valve 
 closing the exhaust, and again cuts off the pressure, thus allowing the signal to return to 
 danger. This cylinder is also provided with a circuit breaker controlling the current to 
 an electric lock to the lever operating it, fig. 8, A. The construction of this lock will be 
 described with the machine, later. This circuit is closed only when the signal is in its 
 danger position, and open at all other times ; and since the lock releases the lever only 
 when the current is on it, it is evident that the lever is unlocked only when the signal is 
 in the danger position. Consequently when the signal is cleared the lever operating it is 
 automatically locked; and should the signal fail to go to danger after the circuit has been 
 broken by the lever controlling it, that lever will re main locked electrically, and hold all 
 switches locked mechanically over which that signal gives right of way, until it does 
 return to danger. 
 
 The small pot or drilling signal, fig. 6, consists*simply of one of the same cylinders as 
 are used to operate the semaphore signals, placed horizontally in a cast-iron box or case 
 and connected to an arm keyed fast to a vertical shaft to which the signal target and lamp 
 are secured. When operated, the cylinder turns this shaft one-quarter of a revolution, 
 thus changing the target or light. The opposite side of this arm is extended, and con- 
 nected to a long spiral spring, which returns the signal to danger when cylinder is dis- 
 charged. 
 
 When it is desired to operate indicators in connection with the signal, a device is pro- 
 vided in a well-covered box, shown to the right of fig. 10, fastened directly under the signal 
 blade and operated by it. This apparatus is provided with a pair of electromagnets for each 
 indicator rod, and a simple means of throwing one or the other of these rods into engage- 
 ment with the signal, by them, so that it will be opened rigidly thereby. The number or 
 letter (see fig. 7) displayed when the signal is cleared indicates to what track the switches 
 are set. This system of signaling is of advantage in yards where a great deal of drilling 
 is done, on account of its simplicity in construction and operation, the small number of 
 lamps employed and the ease with which they can be read. When the signal is at 
 danger the indicators are obscured by a screen which hangs in front of them. 
 
 It is necessary here to explain that all levers controlling signals (fig. 2, A} when 
 thrown out of their normal (vertical) position, i. e., to the right or left, effect the locking 
 of switches during the first part of their stroke, and close the circuit on the signal at the 
 end of the stroke. After the electric locking takes place, when a signal has been cleared 
 by the signal lever being thrown completely to the right or left, it is possible to throw the 
 lever sufficiently far normal again to break the circuit to the signal, but not far enough 
 to release the locking to the switches; in this way the signal must go to danger before the 
 
THE WESTINGHOUSE SYSTEM OF PNEUMATIC INTERLOCKING. 41 
 
 switch levers can be released. The great advantage of this lies in the fact that should 
 a signal stick at safety, it indirectly locks all switches which require shifting in order to 
 set a signal for a route conflicting with it. It is the refore impossible to give two con- 
 flicting signals at any time by mistake or improper working of the signals. 
 
 The interlocking between levers of the machine is confined to that between switch 
 and signal levers only, and never (unless ordered so) between switch levers themselves. 
 Signal levers are interlocked between each other through the switch levers, as will be 
 described next. 
 
 Figs. 1, 2 and 3 will make clear the general construction of the machine, and it is 
 only necessary to explain that the framing is cast iron, the levers, valves, locks, etc, 
 brass, and the top plate of hard rubber, as are the rollers lying horizontally over it. 
 Each switch lever consists of a small brass lever keyed at the centre to a steel shaft which 
 extends through a bearing formed in the front of the machine, a hard rubber roller lying 
 horizontally over the top plate, and terminates in the three-way cock in the rear of the 
 machine, with which it is fastened rigidly. The upper end of this lever is provided with 
 a rubber handle, and the lever end extends down far enough to just clear a dog or latch 
 (fig. 1, A) pivoted loosely under the machine, and extending through its front directly 
 under the lever. These latches perform the locking of the switch levers by the signal 
 levers. In the normal condition of all signal levers, all of these latches lie in a notch 
 cut in the locking bars (fig. 2, It) in front of the machine, and offer no obstruction to the 
 movement of the switch levers ; but the instant a signal lever is moved from its normal 
 position, the latches of all switches affected by it will be raised so as to cause the ends of 
 the switch levers to strike them, and prevent them being moved far enough to open the 
 valves operating the switches. 
 
 The rubber rollers referred to as forming part of the switch and signal lever spindles 
 are cast rigidly thereto, and provided with a series of metallic strips or collars (fig. 3, B] 
 extending part way round them, their ends terminating each in one of the six slots cut 
 the full length of the roller parallel with its axis. These strips are not all put on in the 
 same relative position with the centre line of the operating lever, but are staggered, so as 
 to either make or break their contact with the upright ends of the strips (fig. 3, (7) on the 
 rubber plate running parallel with and directly under them, when the roller is rotated 
 by movement of the switch lever. To one end of each of these strips on the rubber plate 
 the controlling wires to the various signals run, and the other ends are joined together 
 and run to one common battery supplying all signals. The other pole of this battery is 
 connected to the main air pipe, which is used as a common return for all circuits. The 
 breaks in each one of these strips are controlled by the levers operating switches over 
 which the signal thus controlled gives right of way, and also by one or more signal levers, 
 as the interlocking may require. It will be very apparent that, before the current to any 
 signal can be established, all breaks in the strip carrying its current must be closed by 
 
42 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 the bands or collars on the rollers making contact between them. This is done by 
 placing the levers in a position to properly set all switches for that signal. It will be also 
 evident that in, having moved a switch lever to close the strip for one signal, it will at 
 the same time break the circuit at a strip controlling another signal, requiring that 
 switch in its original position. In this way a very simple and effectual interlocking 
 between signals is accomplished. 
 
 The strips on the rollers are so arranged that they make contact between the upright 
 strips only when the levers are in their extreme positions, one way or the other. In 
 order to be able to move a switch lever from one side, to the extreme stroke on the other 
 side, and thereby close the circuit for another signal requiring that switch reversed, with- 
 out any certain knowledge that the switch has moved, a device is attached to the rear 
 end of the roller, which consists of a quadrant (fig. 3, D) secured to the roller by a- set 
 screw, and having cut through it above the roller a radial slot through which projects 
 horizontally a peculiarly shaped locking lever, provided with a small latch or tongue 
 pivoted near its centre, and capable of a horizontal movement right and left on its pivot, 
 but held in a central position, parallel with the lever on which it is pivoted, by a flat 
 spring on each side. This latch, like the lever, extends through the slot in the quadrant 
 and ends flush with the end of the lever. The lever with the latch thus arranged is 
 pivoted in a suitable bracket fast to the machine, and connected at its far end with the 
 armature of an electro-magnet (fig. 3, _E r ), the circuit to which is controlled directly by 
 the lock pin of the switch movement. This circuit is normally open, i. e., when the 
 switch is locked, and closed during its operation. The armature of the magnet, there- 
 fore, normally hangs by gravity away from the magnet and keeps the end of the lever, 
 projecting through the quadrant, elevated, so that a small steel pin in the centre of the 
 upper inside slot of this quadrant, when the switch lever is thrown beyond the vertical 
 position, strikes the latch or tongue and carries its free end with it as far as its con- 
 struction will permit ; the lever then will have been moved sufficiently far to have oper- 
 ated the valve, and consequently moved the switch, but not far enough to have made the 
 contact between the strips controlling the signal. Before this is possible the switch must 
 have been unlocked, moved, and then locked in the other position. The unlocking of 
 the switch closes the circuit on the magnet, which becoming charged depresses the end 
 of the lever projecting through the quadrant, into a recessed portion of the radial slot, 
 holding the switch lever still locked thereby. At the same time, the small latch or tongue 
 being thrown below the small pin which had carried it out of its central position, flies 
 back, under the pin, into its central position on the other side of the pin. The latch and 
 lever assume this position as long as the switch remains unlocked, but on being locked in 
 the position moved to by the lever being reversed, the circuit is broken on the magnet 
 and the quadrant end of the locking lever is raised from the recess in the quadrant and 
 the lever thus unlocked is free to be moved to the end of its stroke, when the signal cir- 
 
THE WESTINGHOUSE SYSTEM OF PNEUMATIC INTERLOCKING. 
 
 43 
 
 cuit becomes closed. Thus the closing of the signal circuit depends directly on, not only 
 the movement of the switch, but the locking of it after it has moved. The electric lock- 
 ing of signal levers is effected by a similar, but simpler device, which consists of an elec- 
 tromagnet whose armature, like that of the indication magnet, is connected to a hori- 
 zontal lever, fig. 3, F, pivoted in its centre, and its far end projecting through a locking 
 quadrant fast to the roller, and engaging in such a manner as to lock it from being moved 
 out of its centre position, if normal, or from being put normal, and thus release the 
 
 Fig. 3. 
 THE WESTINGHOUSE SYSTEM OF PNEUMATIC INTERLOCKING. 
 
 switches locked by it, if out of its centre position, when the current to the magnet is 
 broken through the circuit breaker of the signal controlled by that lever being at 
 
 safety. , ,, 
 
 Directly above the machine is placed a miniature model of the tracks t operated, fig. 
 3 G and small movable switches thereon are connected directly to the rubber roller, so 
 that 'after the indication from the switches is received, and the roller turned as far as 
 possible, these small switches assume the position of the corresponding onps on the 
 ground In this way the operator can, at a glance, see the condition of his tracks at .any 
 
44 
 
 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 time. In order to prevent an operator by mistake throwing his signal back to danger, 
 and then his signal lever normal, and finally a switch, thus released, under or in front of 
 a passing train, an interlocking relay, fig. 3, H, is included in the controlling signal and 
 locking circuits, which, after the operator gives the signal, places the lever thus locked 
 out of his control, in so much that while he has the power to throw the signal to danger 
 at any time, it is not in his power to throw it normal and release the switches until the 
 train has passed over the route set for it and cleared the last point of danger. 
 
 When within about one mile of the crossing, junction, yard, or of whatever the tracks 
 interlocked consist, an approaching train automatically drops an annunciator on the rear 
 of the track model, so as to display through an opening in the model board, fig. 3, /, a 
 number or letter designating the track on which the train is approaching. At the same 
 time, a bell begins to ring, and continues to do so until the train has passed over the short 
 insulated section provided for that purpose, fig. 3, J. These drops are restored to their 
 normal (obscure) position by a blast of air controlled by a small valve, fig. 3, K, in the 
 front board of the machine, by the operator. 
 
 The advantages this system possesses over all others are numerous. Space required 
 is limited, thus reducing size, and therefore cost of towers. The work is light, conse- 
 quently female operators can be employed, thus reducing expenses. It gives great 
 facilities for special locks. Large yards can be worked from one tower, as distance is of 
 little object, switches half a mile away working as well as those close to the tower. There 
 is no danger of signals being left at safety, as the machine remains locked until the signal 
 lias returned to danger. Any number of switches can, if necessary, be worked from one 
 lever. 
 
 The following table shows the plants now in service: 
 
 Pneumatic Interlocking Systems in Operation. 
 
 Levers. 
 
 Location. 
 
 Railroads. 
 
 Put in service. 
 
 6 
 
 Bound Brook, N. J 
 
 Phil. & Read, and L. V 
 
 1884. 
 1884. 
 1884. 
 1884. 
 
 1884. 
 1884. 
 1884. 
 1885. 
 1885. 
 1885. 
 1888. 
 May 6, 1888. 
 Aug. 19, 1888. 
 
 6 
 
 Brightwood Ind 
 
 I. V. andC.,C., C. &I. 
 
 12m 
 12 
 6 
 
 12 ol 
 14 
 24 
 10 
 6 
 24 n 
 24 
 *24 
 
 !W St] 
 
 d stv 
 wst; 
 
 rle 
 
 W^ilkinsburg Pa 
 
 Pennsylvania 
 
 
 East Liberty Pa . 
 
 Pen nsylvania 
 
 
 Valparaiso Ind 
 
 Chic. & Grand Trunk, P., F. W. & C. and N. Y. 
 C & St. L 
 
 e 
 
 Stock Yards Chicago 
 
 C B & Q and Stock Yard R. R 
 
 
 Erie, Pa 
 Oakland Cal 
 
 L. S. & M. S. and Erie & Pitts 
 
 S. Pac. R. R. Tower No. 1 
 
 
 
 
 " ' " No.2 
 
 n 
 
 " " " No.3 
 
 le 
 
 K 
 
 " "for " No. 1, shipped Dec. 8 
 
 
 17th st Pitts. Yd 
 
 Pennsylvania , 
 
 
 14th st. " 
 
 
 
 
 
 *At the 14th St. Pittsburgh Yard tower, the highest number of movements in 24 hours is 1,500, and the highest 
 number of movements in one hour is 86. The machine is operated by one man. 
 
AUTOMATIC BLOCK SIGNALS ON THE PENNSYLVANIA. 
 
 [August 23, 1889.] 
 
 The Union Switch & Signal Company's 
 rail-circuit system of block signaling is 
 now in use on six miles of the four-track 
 road of the Pennsylvania main line east 
 of Pittsburgh. The system has been in 
 use between Wilkinsburg and East Lib- 
 erty, about 2 miles, for 5 years, with a 
 high degree of success, and it is this that 
 has led the company to equip the addi- 
 tional 3f, miles. This latter portion of the 
 road, from East Libery westward, is shown 
 in the sketch given herewith, and the sig- 
 nals upon it were put in operation on 
 August 8. 
 
 As will be observed from the drawing, 
 the signals are here arranged on the 
 original plan as shown in models ex- 
 hibited several years ago ; that is, the 
 blocks are short, averaging half a mile in 
 length, and each stop or home signal has 
 a distant or caution signal, which caution 
 signal is placed upon the same post with 
 the stop signal of the preceding block sec- 
 tion. This simplifies the system for the 
 engineer, reducing the number of local- 
 ities that he must watch for, and also obvi- 
 ously simplifies the arrangement of the 
 posts and apparatus. The distance between 
 each caution signal and its corresponding 
 
46 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 stop signal is ample for bringing a train to a stop, and with caution signals for every 
 block section on the open road the engineman may run at full speed on the most obscure 
 portions of the road or in the densest fog, with the assurance that the signals will always 
 give him at least the length of one block in which to stop, after he has passed the post. 
 
 Instead of "banner" or "gridiron" signals operated by clock work (the best known 
 form, and the one in use on the Boston & Albany, Old Colony and other New England 
 roads), the semaphore is used, and the power for moving the arms is compressed air. The 
 apparatus is the same in its general features as that used to operate semaphores in the 
 pneumatic interlocking system, described in a previous chapter (pages 37-44). A pipe ex- 
 tends along the roadway the whole six miles from Pittsburgh to Wilkinsburg, furnishing 
 air not only for the block signals, but also for the interlocking apparatus at four towers 
 The distant signals are connected to the stop signals by a line wire on poles. The signals 
 are arranged to go to danger after the engine of a train passes them. A train stopped 
 by a home signal may, after waiting two minutes, proceed with caution, expecting to 
 find a train on the block aheap or switches not properly set. 
 
ELECTRIC APPARATUS FOR AUTOMATIC BLOCK SIGNALS. 
 
 [September G, 1889.] 
 
 Following is a plan showing the arrangement of the batteries, electromagnets and 
 wires used in operating the signals described in the preceding chapter. The diagram 
 shows three sections of four-track road, and the signals are arranged in the natural order 
 as related to the tracks. That is, taking post / for example, and looking in the direc- 
 tion the train is moving, the two right-hand arms are for track F 1, and stand at danger 
 to protect the train on the section. The left-hand arms are for track F 2, and show 
 all clear, the section being unoccupied. The normal position of arms is horizontal- 
 danger as on a failure of battery or other apparatus they automatically assume that 
 position ; but their usual position is down safety as at all times when the track is 
 
 Switch box 
 
 -M, 
 
 Su/ilch box 
 
 JBf 
 
 ARRANGEMENT OF ELECTRIC CIRCUITS FOR THE UNION SWITCH & SIGNAL COMPANY'S PNEUMATIC 
 
 AUTOMATIC BLOCK SIGNALS. 
 
 clear the power which works the arms is in operation to hold them down. The cross- 
 over from track 8 to track 4 in section C is shown in position to lead trains from the 
 passenger to the freight track ; this, it will be observed, breaks the circuits in both 
 main tracks and throws to danger all the arms on post c, as well as the distant signals 
 for the same section, which are on post b. The method of connecting the circuit 
 breaker, located in the switchbox, was illustrated in the fiailroad Gazette of June 24 
 1887. (See page 31.) 
 
 The arrangement of the electric circuits can best be understood by following out the 
 connections for a single section, say B\ At the east end of this section is the battery, 
 
48 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 of which one pole is connected to each rail. The track being unoccupied, the current 
 flows through the rails to relay y, closing it. This closes the circuit through the local 
 battery, and causes the electromagnet in the pneumatic signal cylinder B* to actuate 
 the air valve therein and pull down the arm IP on post b. It will be understood that 
 the cylinders A\ B*, (7 4 , D\ shown in the engraving on an enlarged scale, are located 
 upon the post with their respective arms. 
 
 In the diagram the pneumatic cylinders and the semaphore arms are lettered and 
 numbered to correspond with the track section to which they belong. The circuit 
 through pneumatic cylinder B^ is carried by a line wire L on poles to the distant signal 
 for the same section, located on post a, and the current operates to pull down that signal 
 also. All the circuits through the pneumatic cylinders terminate in the ground (6r). If 
 now, a train enters section .A 4 from the west it opens relay X, and, through cylinder 
 J. 4 , throws to danger the home semaphore A 4 on post a. Strictly speaking, this need 
 not be allowed to affect arm B* on post a, as that arm gives an indication for section 
 B* and not for section A* ; but to simplify the indications for the engineer and to 
 obviate even an appearance of inconsistency, the opening of the circuit through cylinder 
 A* is made to open that through the cylinder shown to the left of it, which actuates 
 distant signal B^ on post a just referred to. This opening is accomplished by a circuit 
 breaker at the left-hand end of cylinder A\ by which the current from wire L is turned 
 to the ground before it reaches cylinder B*. Thus a train on a section always keeps 
 horizontal both of the arms that are immediately behind it. Of the two arms on any 
 one post, an engineer may find the home arm down and the distant arm up, as is shown 
 on post e for track 1 ; bu-t he will never find the home arm up and the distant arm 
 down, although such a combination will involve no danger if the indication of the home 
 arm is obeyed. 
 
AUTOMATIC SIGNALS ON THE FITCHBURG RAILROAD. 
 
 [June 15, 1888.] 
 
 The Fitchburg was the first to adopt the Union rail-circuit signals for any except 
 experimental purposes. Ten miles of this road were equipped with the old-style instru- 
 ments of this system about 1879, and gradual extensions have been made since that time. 
 The arrangement of batteries, relays, etc., for these signals is exactly like that described 
 in the article on "Automatic Signals on the Boston & Albany" in the Railroad Gazette 
 for June 24, 1887 ; that is, at the end of the section farthest from the signal is placed a 
 battery having its poles connected, one to each rail, while at the end nearest the signal 
 is a relay with its coils connected in like manner, one pole to each rail. This relay opens 
 and closes the circuit of a local battery, which governs the movements of the signal. The 
 only difference between the two systems is in the form of the signal which is here a 
 semaphore arm and the motive power, which is compressed air moving a piston in a 
 closed cylinder. Each signal is placed about 200 feet from the beginning of the section, 
 in order that the engineers may see them operate, and is provided with an overlapping 
 circuit of about 1,000 feet in length, as described for the system referred to. 
 
 In 1883 twelve miles of the eastbound track on a 60-foot grade from As~hburnham to 
 Fitchburg were equipped with electro-pneumatic semaphores by the Union Switch & Sig- 
 nal Company. The sections are about a mile in length. The general appearance of the 
 signal is shown by fig. 1. At the top of an iron post about 24 ft. high is placed the sema- 
 phore arm, which moves about 60 deg. in a vertical plane in the ordinary manner. The 
 arm itself, however, instead of being connected to a distant lever, is attached to a rod 
 about 3 ft. long which ends in a yoke or stirrup. In the yoke is a box containing an 
 electromagnet and a closed cylinder, fixed to the iron post, within which works a piston 
 actuated by compressed air. The section of this cylinder is shown in fig. 2. The air is 
 supplied through a feedpipe P. The valve admitting it to the cylinder is controlled by 
 an electromagnet which is so arranged that when the current circulates in the coils the 
 armature is attracted and the valve is held open, admitting the air to the cylinder. This 
 drives the piston before it to the bottom and brings down the blade. When for any 
 cause the current is interrupted, the valve closes, the exhaust is opened, and the air es- 
 capes; a counter-weight brings the signal to danger. 
 
50 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 Below and in front of the piston rod is seen an electric circuit-closer attached to a small 
 spindle, independent of the piston rod, but which is operated by the piston itself, when 
 its own movement toward the bottom of the cylinder is nearly completed. This is so 
 adjusted as to keep the auxiliary circuit which it governs closed whenever the signal is 
 $IT'el%aTj- Conveying the indication of this fact to any desired distance; it may also be 
 a'rranged-.to^ing a bell or give other warning whenever the signal shows danger. The 
 ipo^jetnent'of the circuit- closing device is made to depend on that of the piston in such a 
 manner .!%<t the completed movement of the piston [and so of the signal] must take place 
 Jt>ef dfe 4he auxiliary circuit will be closed (or opened, as the case may be). On the Fitch- 
 buig tliis is made use of only in the case of a few special signals, to ring a bell at certain 
 switches from which the signal cannot be seen or to warn switchmen when a train has 
 entered the section. 
 
 The compressed air at 60 Ibs. pressure is supplied by a pump located at Fitchburg. 
 This pump is automatic and similar in principle to that employed for the Westinghouse 
 air brake. The air passes tirst through a coil of pipe to cool it, thence into a reservoir, 
 with which is connected a blow-off cock to remove the condensed water. The air from 
 this reservoir passes into a 1-in. pipe running between the tracks about 1| ft. below the 
 surface. At bridges, etc., where it comes to the surface, expansion and contraction are 
 provided for by a long bend or a round turn in the pipe. At each signal a J-in. branch 
 is connected with the main pipe. This leads to a reservoir at the bottom of the post, 
 which holds air enough to operate the signal about a dozen times. There is a stop- cock 
 in this branch, which can be closed should occasion require. Stop-cocks are also in- 
 serted in the main pipe about every half-mile for the purpose of locating and confining 
 any trouble with the main pipe to a small section, so that only one or at most a few sig- 
 nals need be affected. From the reservoir at the bottom of the post a -in. pipe runs to 
 the air cylinder operating the signal, and when the valve is open the pressure brings the 
 signal clear. The electromagnet is in the circuit of a local battery, which is controlled 
 by a relay, the coils of which are connected to the rails as in the clockwork signals. 
 When a train enters the section this relay opens the local circuit, the armature of the 
 signal magnet falls off and closes the valve leading to the cylinder and opens an 
 exhaust. The air escapes and the signal arm takes a horizontal position until the section is 
 clear. The same thing happens if the batteries fail or a wire or rail breaks. Duplicate 
 pumping apparatuses located at the other end of the grade (Ashburnham), so that in case 
 of accident or repairs to the pump at Fitchburg, or a break in the main pipe, the signals 
 can be worked from the other end as far as the break, for an indefinite time. The air in 
 the auxiliary reservoir at the signal posts is also sufficient for a number of operations. 
 
 It is not necessary to run the pump for these signals more than three or four hours 
 per day, and as much during the night. The regular work of the engineer is to run 
 hoisting machinery for a coal dump ; in addition to this he runs the pumps for the sig- 
 
AUTOMATIC SIGNALS ON THE FITCHBURG RAILROAD. 
 
 51 
 
 nals occasionally until the air pressure reaches 601bs.; then he stops until it falls to about 
 401bs., when he pumps again. 
 
 The number of operations of these sig- 
 nals for two months in 1887 was 48,487, or 
 about 795 per day. The number of failures 
 in the same period was 133, or one failure to 
 365 operations; 22 disk signals during the 
 same pe'iod made 69,844 operations, and 
 there were 54 failures, or one in 1,293 opera- 
 tions. The cost of 12 electromagnet sema- 
 phores with the necessary pumping appar- 
 atus, etc., was $11,126, or $927.17 per signal. 
 The cost of 14 disk signals erected in 1887 
 for another railroad was $6,971, or $497.94 per 
 signal. 
 
 The cost of maintenance of the electro- 
 pneumatic signals is about $133.33 per signal 
 per year; that of clockwork disk signals about 
 $75 per year each signal. 
 
 The application of these signals in a con- 
 siderably more complicated form was made 
 on 13 miles of the West Shore road in 1884. 
 The same system was applied to short sec- 
 tions of the Pennsylvania and some other 
 roads about the same time. In this arrange- 
 ment each block signal was provided with 
 a distant signal at the beginning of the pre- 
 ceding section, so that there were on each 
 post two signals, the upper of which was 
 painted red and referred to the section begin- 
 ning at the signal, and the lower, which was 
 green and cut with a dovetailed end, belonged 
 to the section second in advance of where 
 the signal stood. Each block signal was con- 
 nected with its distant signal by a wire cir- 
 cuit, so that the latter reproduced all the move 
 
 ments Of the former. Besides this, each Electro-Pneumatic Block Signal, Fitchburg Railroad. 
 
 red signal when in the danger position was caused to close a shunt circuit around the 
 magnet of the green signal on the same post, so that the latter also stood at danger as long as 
 
 Fig, 2. 
 
52 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 a train was in the section. There were thus three signals at danger behind a train, two at 
 the beginning of the section where the train was, and one at the beginning of the pre- 
 ceding section. They were lighted at night by lamps which showed red when the signals 
 were in the stop position, and green for all clear. If a train found the two arms on any 
 post standing at danger, or two red lights at night, it would indicate that the section be- 
 ginning at the post was obstructed, while the upper arm inclined and the lower one 
 horizontal (or at night a green light with a red one below it) would mean that the section 
 immediately in advance was clear, but a train or other obstruction must be looked for 
 on the second section ahead. 
 
 This was the first extensive application of electro-pneumatic semaphores put up by 
 the Union Switch & Signal Company, and they were then without that knowledge of the 
 proper methods of construction and operation which has been since gained by experience. 
 Partly from this cause, and partly, it is claimed, because the road was new and the 
 settlement and displacement of the roadbed caused many leaks in the pipes, this system 
 never gave complete satisfaction, and has now been discontinued, the semaphores being 
 worked by signalmen and without distant signals. 
 
THE HALL BLOCK SIGNAL. 
 
 {September 12, 1890.] 
 
 The illustrations herewith show the apparatus used in Hall's electric automatic block 
 system. The Hall is a wire circuit system, the circuit being normally closed. The signal 
 is a circular disk of silk stretched upon an aluminum ring and inclosed within a wooden 
 case with a glass-covered opening. The front of this case being painted dark and show- 
 ing some 10 square feet of surface to an approaching engineman, the signal is a con- 
 spicuous object. Fig. 1 shows the arrangement of wires and electromagnets for operat- 
 ing a simple block-signal circuit. At the entrance of the section is located the "block " 
 track instrument, O S, the operation of which sets the signal at danger. The similar 
 instrument at the other end, S, is called the "clear" track instrument, its function 
 
 o-s. 
 
 THE HALL AUTOMATIC BLOCK SIGNAL-DIAGRAM OF ELECTRIC CIRCUITS Fig. 1. 
 
 being to restore the signal to the safety position. These two instruments are alike in 
 principle and construction, except that the clear instrument stands normally open, while 
 the block instrument stands normally closed. The "clear" track instrument is located 
 1,500 or 2,000 feet beyond the end of the section, so that the longest train will be wholly 
 clear of the section before the foremost wheel touches it, though the circuits are so 
 arranged that the signal does not go to the safety position until the whole of a passing 
 train goes over the instrument. 
 
 It is the relay and JTthe battery. They may be located at any point within the 
 block. D is the signal disk, described more fully in connection with figs. 2 and 3. The 
 circuit is normally closed, and signal D is held in the position shown (safety), by the 
 force of the electromagnet, the circuit being completed from the battery X through wire 
 1, track instrument C S, wires 2 and 3, electromagnet S, wire 4, contact point p, wire 5, 
 electromagnet r, wire 6, to battery. A train in entering the section opens this circuit,, 
 
54 
 
 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 the first wheel of the train breaking the contract between the spring and its anvil C S; 
 electromagnets r and s are demagnetized, signal D falls to a position in front of the 
 glass-covered opening (to danger), and the contact at p is broken. After the whole of 
 
 Hall Signal Fig. 2. 
 
 Hall Signal Inst-ument Fig. 3. 
 
 the train has passed over the "block" instrument, the contact between the spring and 
 its anvil will be restored ; but as the circuit is now broken at p, the signal will remain 
 down (at danger) until the points atp are again brought in contact ; that is, until the 
 
THE HALL BLOCK SIGNAL. 55 
 
 train, in passing out of the section, completes a circuit that shall energize electromagnet 
 r. This is accomplished by the closing of the spring of the "clear " track instrument O 
 #, which completes a circuit from battery X through wire 7, spring and anvil at 8 
 wire 8, electromagnet r, wire 6, to the battery. The contact at^> is now closed, and the 
 signal circuit is complete, but the signal will remain at danger until the train has entirely 
 cleared the "clear " track instrument, from the fact that as long as the spring at S is in 
 contact with its anvil two circuits are completed, one through the clear-track instrument 
 and the relay magnet and the other through the block track instrument, signal magnet 
 
 Hall Track Instrument Fig. 4. 
 
 and relay magnet. This divides the battery power and leaves S too weak to lift the 
 disk. 
 
 Fig. 2 is a view of the case which contains the signal instrument. A white reflector 
 in or behind the case is exposed through the glass-covered aperture as long as the red 
 disk is held out of sight. The front of the case being dark, safety is thus shown by a 
 white disk in the midst of a dark ground. The falling of the red disk before the glass 
 produces the danger signal. At night a lamp is placed between the reflector and the 
 disk, so as to illuminate the latter when it is down and to show clear (white) when the 
 disk is held up. A distant signal is of substantially the same construction, its disk being 
 made of green silk instead of red. 
 
56 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 Fig. 3 shows the construction of the signal instrument (Z>, fig. 1); L and the corre. 
 spending arm below it are the two wings of an armature which revolves on the shaft W 
 between the prolonged cores S and T of the electromagnet, and to which are attached the 
 disk D and its counter- weight rod C. The disk being heavier than its counter- weight, the 
 signal moves to danger by gravity. When the electromagnet is energized the disk is 
 drawn up out of sight by the rotary movement of the armature. 
 
 Figs. 4 and. 5 show the track instrument, fig. 5 being an enlarged view of the top 
 plate. The lever Z/, upon being depressed by the wheels of a passing train, forces up the 
 piston JSj moving in an air chamber D, and communicates motion to the key lever A 
 (fig. 5) of the circuit-closing apparatus. The upper and lower ends of the air chamber 
 are connected with each other by a port X, so arranged that when the piston is forced 
 upward a portion of the air above the piston is forced out through the port X and open- 
 ing Y, which is placed a little below the top of the cylinder. When the piston has risen. 
 
 Top Plate Track Instrument Fig. 5. 
 
 high enough to cover the opening T, the communication with the lower end of the cylin- 
 der is cut off and the air remaining in the upper part is confined and constitutes a cushion, 
 preventing the piston rod from being thrown forcibly up against the top cap. The piston 
 rod extends up through the cylinder head, or top plate, as shown in fig. 5. Upon being 
 actuated by the lever L its beveled top engages the roller of the swinging arm A, which 
 forces the spring B to a contact with its anvil (7, thus completing a circuit between 1 and 
 2, the wire connections. When the piston has been raised by the action of a passing 
 train, the air forced out by it is driven through the port Xand enters the air chamber below 
 the piston ; so that when it falls back the air so introduced retards it in its fall, thus pre- 
 
THE HALL BLOCK SYSTEM. 57 
 
 venting injurious shocks. JR, is a valve for regulating this air pressure. The lower end of 
 the piston rod moves in a closed chamber J3, in which the end of the track lever works. 
 This opening is closed by means of movable plates F, fixed on the lever and working 
 against the edges of the opening. The lever is confined between two rubber springs Cf 
 and H, which are so compressed that any weight less than that imposed by the pressure 
 of an ordinary car wheel fails to operate the piston. 
 
 The Hall company provides a modification of this system for permissive blocking 
 whereby a second train entering a section before the first has cleared it cuts out the elec- 
 tric circuit from the signal behind it, so that the signal can be cleared only by the last 
 train of the series. This method is used on the New York Central & Hudson River, and 
 is more fully described on a succeeding page. 
 
 The Hall signals have shown some remarkable records. For example, one of the 
 earliest signals (of the latest form), located near Wellesley Hills, Mass., on the Boston & 
 Albany, worked 20 months without a fault. Going into service May 30, 1888, and being 
 used as a positive block signal, it has never got out of order, caused an unnecessary stop, 
 or shown safety when danger existed, thus making a perfect mechanical record. [Jan. 
 24, 1890.] In consequence of its satisfactory operation, the Boston & Albany equipped 
 the entire line from Riverside to Worcester, 33 miles, double track, with the system, the 
 sections being overlapped. 
 
 The New York, New Haven & Hartford uses the system even more extensively than 
 the Boston & Albany, and now employs in regular block service 92 signals, protecting all 
 station yards (also the bridge over the Connecticut River at Windsor Locks, where the 
 two main tracks are intervolved), dangerous points and switches on the main line be- 
 tween New Haven and Springfield, 64 miles. A sample of the records on this road is that 
 given by the supervisor of signals on the Hartford division, to the effect that all the sig- 
 nals in his charge (92) (October, 1890) had worked for 38 days without an unnecessary 
 stop or a complaint of any kind from trainmen. 
 
 For single-track working the company provides electric interlocking apparatus, the 
 instruments being made on the principle of those used for the Hall highway-crossing sig- 
 nal. By a simple arrangement a train entering a section sets a signal at danger in the 
 rear, and at the same time the one at the other end of the section is locked in the danger 
 position so as to stop trains from the opposite direction. 
 
HALL BLOCK SIGNALS ON THE NEW YORK CENTRAL. 
 
 {December 5, 1890.] 
 
 The automatic electric block system of the Hall Signal Company, which has been in 
 use for two or three years on the Boston & Albany and New York, New Haven & Hart- 
 ford roads, is now in use on an eight-mile section of the New York Central & Hudson River 
 (double track) near Peekskill, N. Y. This application is in various respects more com- 
 plete than either of the others mentioned, and we print herewith a diagram showing the 
 arrangement of the signals and giving a general idea of the way in which the road has 
 been equipped. The portion of line blocked extends from Oscawana on the south to 
 near Roa Hook on the north, 8 miles, and there are six blocks on the northbound 
 track and seven on the southbound. Each block has a home and a distant signal. All 
 switches within the sections are equipped with a circuit-breaker, so that whenever they 
 are moved off the main track they open the circuit and set the signal for that section at 
 "danger." Each distant signal works simultaneously with its home signal. 
 
 As will be seen by the diagram, the piece of road blocked is quite crooked, the 
 Peekskill station especially being in an obscure location. There are short tunnels near 
 signals 100 and 101 (Oscawana and Crugers), and there is a drawbridge in sections 111 
 and 112. There is an ascending grade going south from Peekskill where heavy freight 
 Trains are liable to lose time, making block signals specially needful. At Peekskill 
 station there is a grade crossing from which the view is very short, and in connection 
 with the block system the signal company has put in bells, which seasonably warn the 
 gate-tender at this crossing of the approach of trains. 
 
 The diagram can be easily read if the meaning of the four principal letters is re- 
 membered. These are: H, home signal ; D, distant signal; B, "block" track instru- 
 ment, by which a train sets a signal- at danger; C, "clearing" track instrument, by 
 which a train restores a signal to the clear position. Thus for block No. 100, D 100 is 
 the distant and H 100 the home signal ; when the engine of a train passes B 100 it sets 
 the signal at danger, and when it passes C 100 the signal is restored. 
 
 Track instrument B 111 also sets a-ringing bells W and X\ and B 108 starts bells 
 Y and Z ; these bells are silenced by the passage of trains over the track instruments 
 near them. The track instrument for H 107 starts bell , and B 104 starts bell A. 
 
HALL BLOCK SIGNALS ON THE NEW YORK CENTRAL. 
 
 59 
 
 -CH2 
 
 MONTROSE STATION, 
 
 CRUCERS ST 
 HI01 
 
 HALL AUTOMATIC BLOCK SIGNALS ON THE NEW YORK 
 CENTRAL & HUDSON RIVER RAILROAD. 
 
60 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 These bells are stopped in the same manner as the others. At S, south of C 109, is a 
 private siding. The drawbridge is interlocked not only with signals 111, 112, but also 
 with 113. 
 
 These bells, as also those in the gate-tender's cabin, are operated on the same prin- 
 ciple as that employed in the automatic highway-crossing signal made by this company, 
 which was illustrated in the Railroad Gazette of March 28, 1890. Bells A, It, W, X, J, 
 Z, as will be seen by reference to the diagram, ring continuously from the time a train 
 approaches within a certain distance until it reaches the bell ; and this ringing warns 
 the switch-tender not to disturb the main track in the face of the approaching train. 
 
 When a train goes into a siding to wait for another train to pass it, the circuit, which 
 would ordinarily be closed by the passage of a train over the clearing track instrument 
 for that section, is closed by the conductor or trainmen by means of a circuit-closer fixed 
 near the switch. The switches of crossover tracks are equipped in the same manner. A 
 southbound conductor arriving at Peekskill, for instance, and wishing to go to the freight 
 sidings on the east side of the main line, goes to the switch for the crossover track, and 
 if he hears no warning from bell Z, turns it ; this sets H 108 at danger. When he has 
 returned to his own track and set the switch for the main track he uses the hand instru- 
 ment (circuit-closer) to restore H 108 to safety, 
 
 S 110 is a special signal to indicate the position of the switch connecting with the 
 siding north of it. A number of passenger trains whose trips end at Peekskill run in 
 upon this side track, and the signal is introduced so that they need not run slowly all the 
 way from signal 108. 
 
 All these block sections are equipped with auxiliary circuits, by means of which 
 trains can be run under the permissive block system, and the track instruments which 
 appear in the diagram without letters are parts of this apparatus. The operation of the 
 permissive circuits can be explained in a general way as follows: A train enters, say, 
 section 104 ; sets H 104 to danger, as the engine passes it, by operating B 104 ; at the same 
 time D 104 goes to caution and warns the following train to slacken its speed. This second 
 train, finding H 104 showing danger, stops two, three or five minutes (as the rule may be), 
 and may then proceed cautiously. When the engine of the first train reaches C 102, this 
 track instrument (whose office is primarily to restore H 102 to the clear position) so 
 arranges the circuits that when the second train passes B 104 (if it does so before the 
 first train goes out of the section) the pressure of its wheels on the lever of B 104 will pre- 
 vent H 104 being cleared by C 104 when the first train passes the latter. If, however, no 
 train is closely following, the arrangement of circuits produced by C 102 is changed when 
 the first train reaches C 104, and the original condition restored, so that absolute blocking 
 may be resumed. 
 
 The apparatus for this method of working is not so complicated as it would seem, and 
 it has fulfilled its- office perfectly during the two months it has been in use here. The 
 
HALL BLOCK SIGNALS ON THE NEW YORK CENTRAL. 61 
 
 officers of the road say that the whole system has worked very satisfactorily. The only 
 unnecessary stops have been caused by trainmen forgetting to clear signals after entering 
 side tracks, and by improper ad j ustment of instruments, due to an inexperienced inspector. 
 This last trouble occurred only three times, and these were all during the first week of 
 operation ; since that time the apparatus has worked perfectly. [Dec. 5, 1890.] 
 
BLACK'S AUTOMATIC BLOCK SIGNAL. 
 
 [January 24, 1890.] 
 
 This apparatus, which has beeen in use for two or three years on the elevated roads 
 of New York and elsewhere, is simple and strong in construction and has been found 
 very efficient in operation. The essential parts are shown in the illustrations herewith. 
 
 Fig. 1 shows a plan and side elevation of the track instrument. This consists of a 
 
 Fig 4, 
 
 n 
 
 \ L 
 
 J 
 
 \ 
 
 ^r 
 
 
 
 
 
 -=J- 
 
 gsTO 
 
 WMM 
 
 iWTO 
 
 ^=- 
 
 
 W^ 
 
 
 
 
 trn 
 
 
 
 Fig. 1. 
 
 Fig. 3. 
 BLACK'S AUTOMATIC BLOCK SIGNAL. 
 
 lever placed just outside of the rail, which is depressed by the tread of the wheel pass- 
 ing over it. This operates a rocking shaft, which is connected with the ground connec- 
 tion leading to the signal. As will be seen, this connection from the rocking shaft to the 
 ground connection is made by means of a rod having on one end of it a strong spiral 
 spring, the object of which is to prevent severe shocks to the various parts. 
 
BLACK'S AUTOMATIC BLOCK SIGNAL. 
 
 Fig. 2 shows the standard form of semaphore signal as used with this system on the 
 Manhattan Elevated. The post, which is of iron, supports a cast-iron shield, behind 
 which the blade is hidden when in the "all clear" position. This shield is painted black, 
 so that safety is indicated by the absence of the arm rather than by its perpendicular 
 position. The motion plate, by which motion is transmitted from the line of ground con- 
 nection to the bell crank at the foot of the signal posts, is shown in fig. 4. It will be seen 
 that this motion plate provides for a certain amount of change in 
 the length of the connecttions from expansion and contraction, 
 or slack. By the use of this device the labor and care necessary S 
 to keep the rods adjusted has been reduced to a minimum. On 
 the Manhattan one man attends to the inspection of 54 signals. 
 
 Fig. 3 is a diagram, not made to scale, showing in a general 
 way the operation of the apparatus. A A are levers operating 
 the connection to signal A. A train moving from right to left 
 has passed the signal and has set it to danger by means of the 
 track instrument A, near the signal. When the engine arrives 
 at the first lever beyond signal B it sets signal B to danger, 
 and immediately after, at the lever marked A, sets signal A at 
 clear. This is repeated throughout the system. 
 
 The length of the blocks on the New York Elevated 
 through which this system has been operated is about 1,700 
 feet. At the regular speed of trains on the Manhattan this 
 distance is traversed in about 47 seconds, which is as close to- 
 gether as trains can be run in ordinary service, this amount of 
 time being generally required at the terminal stations to detach 
 the engine and attach a fresh one at the other end of the train. 
 Discharging the passengers and taking in the new load also 
 requires nearly a minute, although it is often done more 
 quickly than that. 
 
 The trains on the Manhattan lines where this signal is 
 used are of nearly uniform length ; the invariable maximum 
 is five cars and an engine. Four-car trains are run a good deal, and empty engines make 
 occasional trips, but the variation is immaterial from a signaling standpoint. The signals 
 and track instruments are therefore made to accommodate the five-car trains. The 
 engine sets each signal to danger when the rear car has passed 25 feet beyond the signal, 
 and restores that signal to "all clear" when the rear car has passed the next one ; each 
 is therefore a home signal. No permissive blocking is allowed, but every engineman 
 must be prepared to stop at the signal. 
 
 This apparatus is the invention of Mr. Robert Black, Headmaster, who has been 
 
64 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 connected with the roadway department of the New York elevated roads since they were 
 built. The signals are now made by the Black Automatic Railway Block Signal Co., 
 of 192 Broadway, New York City. They are in more or less use on the Staten Island 
 Rapid Transit road, the Kings County Elevated and the Brooklyn Bridge. The Man- 
 hattan has a series of 32 continuous block sections in operation on the Sixth avenue line 
 above Fifty-ninth street. 
 
ILLUMINATED SEMAPHORE SIGNALS. 
 
 [March 30, 1888.] 
 
 There are many railroad men and experts in signaling who believe that, all things 
 considered, the indications of fixed signals should be by color rather than position. In 
 
 Section on E C D. 
 
 Fig. 1. 
 ILLUMINATED SEMAPHORE SIGNAL. 
 
 Section on A A. 
 
 the evolution of signals the semaphore form has gradually come to be considered the- 
 best Ordinarily this shows danger or caution by day by color and form as well as by 
 position, but the application of this principle to night signals is more difficult and the use- 
 of color alone is still almost universal. Various devices have been employed for illuminat 
 
AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 ing the semaphore blade so as to get a good position signal at night, but they have gener- 
 ally failed from defective illumination. When the lamp is placed in front of the signal it 
 is found that the rays reflected are so diffused that the semaphore is visible at so little 
 
 ILLUMINATED SEMAPHORE SIGNAL. 
 
 distance as to be useless except for yard purposes or other places where trains are moved 
 but slowly. We show two successful semaphores designed to get over these difficulties. 
 
 UNION SWITCH & SIGNAL CO.'S ILLUMINATED SEMAPHOBE. 
 
 This may be used as a purely position signal, or in combination with colored lenses 
 as a combined position and color signal. It is the design of "Messrs. Spicer and Schreuder, 
 and has been quite extensively used. In August, 1889, an order was issued on the Penn- 
 
ILLUMINATED SEMAPHORE SIGNALS. 67 
 
 sylvania lines west of Pittsburgh that illuminated blades should be used for all new work, 
 and this semaphore has been considerably used on those lines. 
 
 In fig. 1 the semaphore is shown as designed to dispense with colored disks, and in 
 fig. 2 it is shown with red and green disks for a three-position signal. Another design is 
 also made with only a red disk. 
 
 The construction of this illuminated semaphore is very clearly shown in the illustra- 
 tions. The reflectors are shown at Jt R', etc. That in the semaphore arm is concave and 
 corrugated, as represented in fig. 1. The reflected light is thrown forward through a plain 
 glass Or, 32 in. long and 2^ in. wide. The divergence of the rays would naturally make 
 the arm of light so shown appear larger than the actual dimensions of the glass. The 
 glass is put in with rubber packing, to diminish the chance of breaking The lamp L is 
 of the same size and power as that used in the ordinary signal of the Union Switch & 
 Signal Co. At B is shown a lense for a back light to let the signalman see that the lamp 
 is burning and that the signal goes to its proper position. 
 
 This is a beautifully distinctive signal, and cannot be confounded with any other 
 lights. 
 
THE KOYL PARABOLIC SEMAPHORE. 
 
 [October 19, 1888.] 
 
 This semaphore is made by the National Switch & Signal Company. It is the inven- 
 tion of Prof. C. Herschel Koyl, formerly of Swarthmore College, now connected with the 
 
 Plan. 
 THE PARABOLIC SEMAPHORE. 
 
 National company. This semaphore has been widely tried in the United States and is 
 the standard form used by the makers of it in their signaling work. 
 
 The reflecting surface and the frame which incloses it form a longitudinal section of 
 
THE KOYL PARABOLIC SEMAPHORE. 69 
 
 a paraboloid. This is mounted in the ordinary casting which supports the semaphore 
 arm, and rotates about the axis in the post, which is also the axis of the paraboloid, in 
 the focus of which the lamp is placed. To a front view the semaphore presents the 
 dimensions of 5 feet 6 inches in length, from the axis to the end of the blade, and 11 
 inches in width, and has a reflector 4 inches wide along the centre, from end to end. 
 
 A reflecting surface of the form specified has the property of making parallel all rays 
 which fall upon it from the focus, the consequence being that there is sent along the 
 track by this means a continuous band or beam of light of definite dimensions. The 
 specification of the patent also says, "in view of the fact that this band or beam should 
 be of such dimensions that at the conventional distance at which the signal comes under 
 the observation of the engineer it shall be wide enough to extend across the track to 
 which it is appropriated, and deep or broad enough to about cover the observation por- 
 tion of the cab of the engine, it will be found desirable to slightly modify both the longi- 
 tudinal and transverse curvature of the paraboloidal section which forms the reflecting 
 surface, so that at such distance there will be sufficient divergence of the rays to produce 
 a beam of the dimensions indicated. To accomplish this result, however, the modifica- 
 tion in form of the reflector need be very slight, and it remains to all intents and pur- 
 poses of the contour of the section of a paraboloid." 
 
 The object has been to concentrate upon the track all the light from the arm, and 
 then to diverge it only so much as the necessities of the case may require, the method 
 adopted for diverging it being to slightly corrugate the reflector at right angles to the 
 four directions in which the divergence is required. 
 
 The result is a semaphore which presents the same appearance by day and by night, 
 which combines the excellencies of a color signal with those of a position signal. The 
 upper half of the lamp being red or green and the lower part clear, the arm when hori- 
 zontal appears red or green, and when dropped white. 
 
 The reflecting surface is now made of aluminum, which is lighter than glass, tough 
 and does not tarnish. 
 
THE STEWART-HALL TRAIN ORDER SIGNAL. 
 
 [January 23, 1891.] 
 
 The diagram printed herewith shows the arrangement of wires and electromagnets 
 used in a train order signal recently patented by Robert Stewart, Superintendent of 
 Telegraph of the Central of New Jersey, and W. P. Hall, President of the Hall Signal 
 Company. The special features of the signal are, 1, that it works entirely by electricity, 
 so that it can be placed in the most convenient location, and not necessarily in front of 
 the office ; and, 2, a provision against letting a train pass while there are orders for it, 
 through the operator's forgetful ness. As is well known, the last mentioned purpose is 
 
 THE STEWART-HALL ELECTRIC TRAIN ORDER SIGNAL. 
 Arrangement of Instruments and Wires. 
 
 appreciated by practical superintendents and dispatchers, and various devices have been 
 heretofore invented for accomplishing it. The convenience of locating the signal itself 
 some distance beyond the station, so that a train can be held for orders, while at the same 
 time permitting it to stop where passengers or freight can be discharged and the tank 
 filled with water, is obvious. This, of course, requires two signals, one for movements 
 in each direction. The out-door signal may be either the Hall disk signal as used in the 
 block system of this company or the semaphore shown below. 
 
 The electromagnet $, in fig. 1, represents the signal instrument, the presence of the 
 electric current in the coils holding the signal in the safety position as shown. D rep- 
 resents the "drop," which is mounted in a cabinet fixed at some place in the office 
 
THE STEWART-HALL TRAIN ORDER SIGNAL. 
 
 71 
 
 where the operator must leave his desk in order to reach it. It will be seen that when- 
 ever the circuit is opened the armature, in falling away, opens its own circuit in such a 
 way that it can be again closed only by lifting the armature of D 
 by hand. As will be seen in fig. 2, this armature carries a tablet 
 lettered so .as to warn the operator of the fact that a train order is 
 on hand. The opening of the circuit permits this tablet to fall 
 from its hidden position to a stop which leaves it in front of a 
 glass-covered opening in the case. 
 
 On receiving an order for a train the operator sets the signal 
 at danger by opening the switch C Y , which is located on his desk. 
 If now, when the train arrives, he should forget that he is to 
 stop it, and should attempt to clear the signal by closing the 
 key (7, he would find that the signal remained at danger, and, on 
 going to the cabinet D, to close the circuit by lifting the armature, 
 would be reminded by the tablet of the presence of the train 
 order. The lifting of the armature is accomplished by means of 
 a large push button or plunger, and this button can be arranged 
 with a clip to receive a folded order, so that the circuit cannot 
 conveniently be closed without first picking up the order. 
 
 Two of these signals have been erected at Somerville, 14. J., 
 on the Central of New Jersey, one for westbound and one for 
 eastbound trains, and additional applications are now being made 
 on this and other roads. 
 
 THE HALL ILLUMINATED ELECTRIC SEMAPHORE. 
 
 In connection with the train order signal described above, 
 the Hall company uses the new form of semaphore shown in the 
 accompanying cut. The mechanism for operating this signal is 
 precisely the same as that used in the disk signal. In the disk 
 signal the counterweight is lighter than the signal, but in the 
 semaphore it is made heavier so that the arm shall assume the 
 horizontal position in case of failure of the circuit or operator. 
 The cut shows the signal as actually used at Somerville, the 1 
 case being a modification of the pattern used for disk signal ; TheHall Electric Semaphore. 
 but the company is making plans for a case with a glass-covered 
 
 opening of a different shape. With a rectangular opening of the right proportions the 
 ordinary semaphore can be quite closely imitated, the side of the case answering for 
 
72 AMERICAN PRACTICE IN BLOCK SIGNALING. 
 
 the post. The angle shown by the dotted line in the cut does not indicate the limit 
 of the power of the magnet. 
 
 The opening in the case, which furnishes a light background for the dark blade, is 
 covered with transparent glass in front, and with ground or painted glass at the back. 
 The lamp for illuminating the signal at night is placed 18 inches or more away from the 
 case, and is fitted with a reflector shaped to diffuse the rays of light equally over the 
 whole surface of the glass. The blade is made of silk or cloth stretched on a hollow 
 wire, the same as the Hall disc, and is therefore sufficiently translucent to show its color 
 at night. It can be made of any desired color that contrasts sufficiently with white. 
 
W.LLIAM P. HALL, President. A. W. HALL, General Manager. 
 
 W. S. GILMORE, Treasurer. S. MARSH YOUNG, General Agent. 
 
 THE HALL SIGNAL COMPANY 
 
 5O Broadway, New York. 34O The Rookery, Chicago. 
 
 MANUFACTURERS OF 
 
 All Kinds of Electric Signaling Apparatus 
 
 FOR RAILROADS. 
 
 The Hall Automatic Electric Block Signal System, 
 
 The Hall Highway-Crossing Alarm Bell, 
 
 The Stewart- Hall Electric Train-Order Signal, 
 
 The Hall System of Interlocking Signals for Grade Crossings of One 
 
 Railroad with Another. 
 
 ELECTRIC DISTANT SIGNALS FOR SWITCHES. 
 
 The Hall apparatus, described in the foregoing pages of this volume, is, as the title of the book implies, prin- 
 cipally that used for block signaling. The chapter devoted to the Hall system in the first part of the book (a 
 reprint of an article in the Railroad Gazette) will be found to be couched in the conservative, cautious and con- 
 densed language usually found in an editorial, and it was, moreover, written in connection with articles about 
 other systems of signaling, which circumstance led the writer to treat some points in a manner not conducive to 
 the best understanding of this particular system. The descriptions of the Hall apparatus in the appendix like- 
 wise suffered from editorial limitations both in text and illustration, and do not give an adequate idea of the 
 completeness of our system or of its adaptability to all the demands of perfect railroad service. While finding no 
 fault with the publishers for this condition of things, we have concluded that the best use we can make of these 
 advertising pages, will be to lay before the reader a few additional facts, which, by supplementing the illustrated 
 descriptions and other matter, will give him a more complete idea of what we can do and are doing. These 
 facts, with some reference to the evidence that we can show to those unacquainted with our systems to prove 
 our claims, will be found on the 
 
 NEXT PAGE. 
 
 THE HALL HIGHWAY CROSSING SIGNAL 
 is described on page 3. 
 
THE HALL SIGNAL COMPANY. 
 
 [Advertisement. ] 
 
 THE HALL ELECTRIC AUTOMATIC BLOCK-SIGNAL SYSTEM. 
 
 By the above term is meant the Hall system with a 
 wire upon poles and the Hall disk signal, as shown in 
 the descriptions in this book.* To profitably read those 
 descriptions the reader should first fix in his mind 
 some of the essential conditions of block signaling. 
 
 The main features demanded in an automatic block 
 signal are well known. They are : that it shall 
 promptly, on the passage of a train, change from 
 safety to danger, and thus protect the train against a 
 rear collision until it shall have passed beyond another 
 signal further on; that the signal shall instantly go to 
 danger whenever a switch is turned to the side track, 
 and that it shall perform these operations unerringly ; 
 that the changing from danger back to safety after the 
 train moves out of the block section shall be done 
 promptly, and that failures to do this, and other fail- 
 ures (which, while not dangerous, cause delays to 
 trains and other annoyances) shall be reduced to a 
 satisfactory minimum. These points are set forth at 
 length in our illustrated catalogue, which is already 
 familiar to most railroad operating officers. Ihe 
 Hall signal meets all th.se conditions. The vital 
 question is, can trams be run close together with 
 safety, regularity and convenience. The first 
 trouble, where trains follow each other closely, is 
 that the detention of one by a signal that fails to 
 operate detains a number of others. This empha- 
 sizes the demand for perfect apparatus and con- 
 stant, intelligent care. The satisfactory way in which 
 the Hall signals meet this test is beyond comparison 
 with the showing made by any other automatic signal. 
 The good records from a few roade, shown in our illus- 
 trated catalogue published a year ago, are now re- 
 peated on a number of others, and in localities where 
 the service is, if possible, more exacting. These rec- 
 ords will be given in detail to railroad officers inter- 
 ested. As most interested readers know, it is next to 
 impossible to maintain gJod discipline among engi- 
 neers where they are frequently stopped by signals un- 
 necessarilythat is, where there is some fault in the 
 battery or connections, or something else that has re- 
 sulted from carelessness; and a signal which thus 
 causes annoyance is, in some respects, as undesirable 
 as one with worse faults. Our records will show that 
 the number of these unnecessary stops can be, and is, 
 kept very much smaller with the Hall than with any 
 
 other automatic signal. There is no mystery about 
 this. The explanation is found in the greater simplicity 
 of the apparatus and the superior mecnanical construc- 
 tion of everything connected \\ita the system. Bat- 
 teries can be kept up by ordinarily intelligent men. 
 The detection of bad connections, wrong adjustments, 
 and other troubles is not such an intricate job as to re- 
 quire an expert with elaborate training. We insist on 
 good men, and require faithful service, but we do not 
 ask impossibilities of them. A chief trouble with 
 track-circuit signals is that the faults of the system, of 
 the apparatus, and of the men, when all combined 
 make up such an aggregate of delays, annoyances and 
 dangers that they are intolerable. The system de- 
 mands delicate battery power which must be very 
 carefully adjusted; and even then wet weather will 
 often baffle the best attempts at adjustment of the in- 
 struments. A vital feature of the system is the shunt, 
 by which the current, when the signal is to be set at 
 danger, is not wholly, but only partially, withdrawn 
 from the signal magnet. This is a feature which prob- 
 ably can never be made to work with the simplicity 
 and certainty of a good wire circuit system in which 
 the circuit is positively and totally broken every time 
 a train enters the section. As to construction of appa- 
 ratus the Hall Signal Company only ask a compari- 
 son of their goods with those of any other manufacturer. 
 The opinions of railroad officers who have used our in- 
 struments and others side by side will be freely shown. 
 As to negligence of inspectors, the Hall Signal Com- 
 pany expects human fallibility, and, as above inti- 
 mated, only demands good, intelligent men, managed 
 with reasonable discipline. All we have to say is that 
 such men have succeeded in regularly making satisfac- 
 tory records. The company's standing offer to guaran- 
 tee the maintenance of signals forfivt years at a fixed 
 low rate per year should be sufficient confirmation of 
 this; but we do not ask roads using our signals to put 
 up with a mere guarantee, for we f ally recognize that 
 a money forfeit would be poor compensation for an un- 
 satisfactory signal system; and inspection of signals in 
 satisfactory operation on prominent roads is earnestly 
 invited. We desire railroad officers to convince them- 
 selves of the merits of our system. The complaint 
 about the deceptions of selling agents shall not be 
 justly applied to the Hall signals. 
 
 * We own the best patents on rail circuit systems and contract to put in such systems with apparatus possessing all the 
 mechanical perfection which has so long characterized the work of our shops. This is now practicable in consequence of the 
 expiration of patents which formerly kept this company or rather' its predecessor out of this field. We also make semaphore 
 signals, both inclosed in acase with a glass-covered opening, and uninclosed; but the present chapter is devoted to the Hall sys- 
 tem, as it is now most largely used- 
 
[Advertisement.} 
 
 THE HALL SIGNAL COMPANY. 
 
 AUTOMATIC HIGH W AY-CROSSING SIGNALS. 
 
 The heavy expense involved in keeping watchmen 
 at road crossings makes a reliable automatic apparatus 
 for warning travelers an important desideratum ; but 
 most superintendents have 
 had their faith in auto- 
 matic bells seriously weak- 
 ened, if not destroyed, by 
 the very poor service given 
 by the devices heretofore 
 made and sold. The Hall 
 bells give perfect satisfac- 
 tion. This is a sweeping 
 statement, but it is based on 
 the testimony of intelligent 
 and conservative railroad 
 officers. The apparatus is 
 simple and of the best de- 
 sign, and carefully made. 
 The first cost is somewhat 
 higher than that of other 
 makes, but the freedom 
 from annoyances soon over- 
 balances that, while a bell 
 which fails to ring when a 
 train is approaching, and 
 persists in ringing when a 
 train is not coming, is dear 
 at any price. The Hall bell 
 is operated by the simple 
 and reliable interlocking 
 magnets used in our single- 
 track block signals (de- 
 scribed on a succeeding 
 page), these in turn being 
 actuated by the Hall track 
 instruments. The bell be- 
 gins to ring when the train 
 passes the track instrument, 
 and continues to ring until 
 the train reaches the cross- 
 ing, whether the interven- 
 ing time be short or long. 
 By a simple arrangement 
 of interlocking instruments 
 and track instruments, the bell is made to ring by the 
 approach of a train from either direction on a single 
 track, the track instrument which would cause a 
 
 AUTOMATIC CROSSING SIGNAL. 
 The gong is covered by the circular wire screen. 
 
 false signal by being operated by a train moving 
 away from the crossing instead of toward it, 
 being automatically cut out. 
 
 The experience of this 
 company has shown that 
 an open circuit Is the 
 most desirable in all re- 
 spects for the operation of 
 a signal of this kind. The 
 primary reason for using 
 a closed circuit that the 
 failure of a battery or 
 wire shall automatically re- 
 veal itself is found to be 
 practically valueless in a 
 highway - crossing signal ; 
 and, in view of the reduc- 
 tion in first cost of wires, 
 instruments and battery, 
 and of expense for battery 
 material and maintenance, 
 as .well as the labor of 
 taking care of the signals, 
 the open-circuit system is 
 far preferable. The records 
 of the operation of these 
 signals on the many 
 railroads using them are 
 such as to substantiate 
 these statements. The bell 
 as used for highway cross- 
 ing signals is also used 
 with much satisfaction at 
 crossings where gates, with 
 an attendant, are main- 
 tained. The certainty of 
 its operations make it a re- 
 liable indicator by which 
 the gate-tender may be noti- 
 fied so as to always close 
 the gates in proper season 
 before the arrival of a train; 
 and where travelers on 
 the highway are familiar 
 with the locality the bell will serve as a warn- 
 ing to them at night when the attendant is off 
 duty. 
 
THE HALL SIGNAL COMPANY. 
 
 [Advertisement.] 
 
 ELECTRIC INTERLOCKING INSTRUMENT. 
 
 The highway crossing bell is operated by means of 
 the interlocking instrument shown on the opposite 
 page. When approaching trains strike the first track 
 instrument they lock the advancing and retreating 
 springs in one position, which sets the bell ringing, and 
 when the wheels strike the track instrument at the 
 crossing the springs are instantly thrown to the other 
 position, cutting off the current from the bell. 
 
 This instrument consists principally of a main elec- 
 tro-magnet A and its armature, and an unlocking 
 electro-magnet B and its armature, the latter consti- 
 tuting a locking lever. To the armature of A is fast- 
 ened a horizontal bar which, by means of tappets, 
 actuates the advancing spring C and the retreating 
 spring 7). When the main electro-magnet is momen- 
 tarily energizedjind its armature attracted, contact is 
 
 spring C and its anvil c, 
 
 while the contact between retreating spring D and its 
 anvil d is broken. The armature of B instantly drops, 
 locking the armature of A and retaining the springs in 
 this position. When B is momentarily energized and 
 its armature attracted, the armature of A is unlocked 
 and contact is made between retreating spring D and 
 its anvil d, while the contact between advancing 
 spring C and its anvil c is broken. Thus, when the 
 advancing spring is open the retreating spring must be 
 closed, and when the retreating spring is open the ad- 
 vancing spring must be closed. Upon this principle is 
 based the Hall System of Electric Interlocking. 
 
 These instruments are also used in single-track block 
 signaling, in signals for the protection of grade cross- 
 ings (of one railroad with another), and in signals for 
 the protection of single or double track junctions. 
 
[Advertisement. ] 
 
 THE HALL SIGNAL COMPANY. 
 
 INTERLOCKING INSTRUMENT. 
 
THE HALL SIGNAL COMPANY. 
 
 [Advertisement ] 
 
 THE HALL TRACK INSTRUMENT. 
 
 This instrument, shown in the illustrated description 
 of the Hall block signal, is of such general adaptability 
 that it is used for any and every situation where it is 
 desired to have a train open or close an electric circuit. 
 This is the only device of the kind that has ever given 
 satisfactory service for any length of time. It has 
 been in use for many years (though recently improved) 
 
 and has withstood the hardest service with exceedingly 
 small expense for care and repairs. The combination 
 of rubber and air cushions for absorbing all shocks 
 gives it great durability and prevents all objectionable 
 noise, a feature that will be appreciated by those 
 familiar with the clanging produced by some kinds of 
 track levers. 
 
 DISCS AND INCLOSED SIGNALS. 
 
 The Hall disc signal has been objected to in certain 
 quarters because it lacks some of the distinctive fea- 
 tures of the ordinary out-door semaphore ; but there 
 are advantages in its present form which each year's 
 use makes more apparent. One of the first points 
 mentioned in favor of semaphores is that they can be 
 seen great distances; but this advantage is often neu- 
 tralized by locating the signal where it has a very poor 
 background, while in other cases the post is made so 
 high (with a view to rendering it visible at a long dis- 
 tance) that engineers, keeping their eyes on the signal, 
 overlook a caboose directly in front of them and run 
 into it. This has occurred repeatedly. The Hall sig- 
 nal case provides a uniform background for every 
 signal. This makes even a small disc more effective 
 than a large semaphore, as the background that is, 
 the outline of the case attracts the eye as quickly as 
 would a semaphore. The signal is generally only 
 about 12 ft. above the ground, and therefore in the 
 line easily and naturally followed by the engineer's 
 eye. Men will not overlook a red tail light while look- 
 ing for the signal. The main object in placing signals 
 on tall poles to give engineers warning at a point a 
 long distance in advance of the stopping place is 
 better secured by the erection of an auxiliary (distant) 
 signal. In foggy weather, or other obscure conditions, 
 a vey tall signal is no better than one of moderate 
 height ; while in clear weather there is no necessity 
 for it at all except to get a good background, and that 
 the Hall signal has in any position. As will be seen 
 by the description of the Hall electric semaphore, we 
 combine the advantages of a semaphore with those 
 of an inclosed signal (with a good background) 
 when desired. 
 
 It may seem like a commonplace argument to men- 
 
 tion here the superior economy found in maintaining 
 and operating inclosed signals, as compared with any 
 automatic signal exposed to wind, snow and rain ; but 
 when one takes a broad view of the matter this is 
 really a more important point than it seems, for it 
 reaches into all the future years that the signal may 
 be used. Indeed, the persistent success of our inclosed 
 signals, in spite of the disadvantages (after all, mostly 
 theoretical) they have been believed to labor under, 
 has been largely owing to this favorable feature. If 
 an automatic signal is to work outdoors in spite of the 
 elements, it must have a large reserve power ; where a 
 force of 10 Ibs. is ordinarily needed, 50 Ibs. or 100 Ibs. 
 must be provided to allow for variation in weight, 
 wind pressure and other uncertain factors. Now, this 
 necessity for additional power is responsible for most 
 of the inevitable complications and added expense. 
 The task of keeping in order a Hall inclosed signal 
 may be compared to that imposed upon an operator in 
 charge of a simple Morse telegraph instrument, which 
 is almost nil as far as skill and time are concerned ; 
 while every outdoor automatic signal thus far tried 
 has be*en made to give 'passable service only by being 
 subjected to constant experiments, changes of plan 
 and of apparatus, and other expedients ; and after 10 
 years' trial the prospect of a satisfactory outcome is 
 no better than at the start. In presenting this aspect 
 of the case we may possibly seem to be " retarding the 
 progress of science "; but our labors to provide for rail- 
 roads (1) just such signal apparatus as the service needs, 
 and (2) such as will do what it seems to do, will attest 
 to all conversant with the facts that we are not un- 
 progressive. It is only by constantly considering both 
 of the above factors that any real progress can be 
 attained. 
 
THIS BOOK IS DUE ON THE LAST DATE 
 
 _ -._ . STAMEED BELOW 
 L ^% I ^^^ 
 
 l\ltW^AAl|ff[NE OF 25 CENTS 
 
 WrLLTBE^sWffsED FOR FAILURE TO RETURN 
 THIS BOOK ON THE DATE DUE. THE PENALTY 
 WILL INCREASE TO SO CENTS ON THE FOURTH 
 DAY AND TO $I.OO ON THE SEVENTH DAY 
 OVERDUE. 
 
 MAO Q 1941 M 
 
 SENT ON ILL 
 
 ninn O iCF*ti iii 
 
 
 P 
 
 JUL 2 3 W7 
 
 
 U.C. BERKELEY 
 
 1 180rf8tLV 
 
 
 
 
 . C 4 A% 
 
 ^ft%ftflir'ti\ w 
 
 
 ^0">* 
 
 
 
 
 
 
 
 
 M tLil'ftOfW^ 
 
 
 O Jvl va W 
 
 
 >_. ^/ h D 
 
 
 rvit-N"* *^ 
 
 r- 1OC^ 
 
 
 JUL 5 OTJ 
 
 
 
 
 W ^ 2 9 W 
 
 
 KC. C1R. APR i 8 1981 
 
 MAO A A tA^M L 
 
 
 WAK 2 f9$# 4 
 
 ) 
 
 r 
 
 LD 21"100m-7,'39(402s 
 
YD 17819 
 
 THE UNIVERSITY OF CALIFORNIA LIBRARY