Story - Modernizing the NYC's Air Line Subdivision for High Speed Freight Service - 1950's

Compiled and edited by Dale Berry

The Michigan Central Railroad ( MC) purchased the western half of the Air Line immediately after it was built in 1870 by the Michigan Air Line Railroad. This part of the Air Line connected Jackson with Niles, by way of Homer and Three Rivers and was a single track line, except for a short portion from Jackson west to Haires, which paralleled the Lake Shore & Michigan Southern's (LS&MS) Fort Wayne Branch. At some point after the LS&MS and MC came under common ownership, both lines were apparently used by both roads on this short stretch.

By the time the Air Line was built, Jackson was quickly becoming the new division point for the Michigan Central and the old division point at Marshall was being minimized. Niles was also becoming a major terminal of the Michigan Central, with MC trains being classified there for Chicago and points west. During its heyday, Niles boasted of classification yards and humps in both directions, as well as repair shops and a large icing facility.

A map of the Michigan Central/ NYC Air Line Branch. The dashed line between Three Rivers Jct. and Elkhart is part of the Kalamazoo Branch and Old Road which was included in the CTC plan in 1958. The western part of the branch between Three Rivers and Niles was removed in 1933. [Map by Dale Berry]


The Michigan Central main line west of Jackson was a busy single track route until around 1900. The Air Line became a second "track" for the MC from Jackson to Niles, with westbound freight trains using the main line via Battle Creek and Kalamazoo, and many eastbound freight trains using the Air Line from Niles to Jackson. (The line east from Jackson to Detroit was already double tracked by this time).  Most passenger trains continued to use the main line as population centers ( Battle Creek and Kalamazoo) were located along this route. The Air Line had minimal passenger trains throughout it's existence because towns long the line were small and infrequent.

The Air Line (not to be confused with the LS&MS/ NYC Airline between Toledo and Elkhart) was a busy branch line, with stations along the route manned by operator/agents who assisted dispatchers in controlling the movement of trains by controlling blocks and issuing train orders to conductors and engineers. The line was "dark", in that it had no automatic block signals. Dispatching records from October 29, 1908 show seven (7) westbound trains and fifteen (15) eastbound trains on the Air Line on that particular day. These included two 3-car passenger trains in each direction. Freight trains were 40-80 cars and averaged about 1,500 tons. The record indicates that there were 16 open block stations along the route, including those at Jackson and Niles.

The short section of the LS&MS Fort Wayne branch (which paralleled the Air Line near Jackson) had five (5) trains in each direction from Jackson to Haires in 1908.

A 1911 Michigan Central employee timetable shows passing sidings at most stations on the Air Line. They were typically over one mile in length, with the largest at 10,857 feet in Three Rivers. The length of passing tracks give some indication of the importance of a branch. Many other branch lines (ex. the Fort Wayne branch, had much shorter passing tracks). Yard limits on the Air Line were limited to Jackson (from Jackson Jct. to Haires interlocking tower) and Niles (from Niles east to the east switch of the Air Line Bluff Track).

The only signals on the line were station block signals (lamps lit by oil, usually protruding out on a pole from the depot itself). The exception to this was at various interlocking towers along the route as well as their approach signals. Interlockers were located at OD Tower (crossing the LS&MS in Jackson), Homer (crossing the LS&MS Lansing Branch and the DT&M), Fairfax (crossing of the MC Goshen Branch), Wasepi (crossing the GR&I), VE Tower at Three Rivers (crossing the LS&MS Kalamazoo Branch) and Cassopolis (crossing of the GTW). Some junctions may have also been interlocked along the line (at Haires, for example).

The Air Line Branch was part of MC's Middle Division, later called the Michigan Division. The line was dispatched by the MC dispatcher in Jackson, who also controlled the LS&MS from Jackson to Haires (double track with the Air Line) and other lines. Until the late 1950's, the dispatcher was located at the Michigan Central (later NYC) division offices on Page Avenue at Elm Street in Jackson, near what is known as  Jackson Junction. The dispatcher communicated with stations along the Air Line by telephone, using track-side pole lines. Open line telephones (similar to a large "party line") were located at each station, and at wayside phone booths at various points along the railroad. Train operations were conducted by timetable and train order until Centralized Traffic Control (CTC) was installed in 1958.

The Air Line also had a significant mid-way locomotive facility just east of  Clarendon, with a concrete coaling tower, cinder pit, sand house and a small shop. A standpipe off a water tower at the St. Joseph River also provided water for steam locomotives using the line. (The concrete coaling tower continues to exist today on private property, though the railroad has long been removed). There were also track pans at Clarendon, Wasepi and Vandalia.

This is the Michigan Central Air Line at Wasepi, looking west in 1915.  Wasepi is the crossing of the Air Line and the Grand Rapids & Indiana ( PRR) line from Fort Wayne to Grand Rapids.  The Wasepi interlocking tower, and the GR&I (left) and MC (right) depots can be seen in the distance. The MC wreck train has pulled into the GR&I interchange track. [Charles Milliken photo, Doug Leffler collection]


The part of the branch extending from the west end of the Three Rivers Yard to Niles was removed during the depression in 1933, as a way to reduce cost. By that time, the now double-tracked MC main line was apparently doing the job by itself and there was only limited local traffic and few towns on this western portion of the Air Line. As many as three interlocking plants were eliminated in this abandonment, as well as agent/operators at depots, and section crews to maintain the track. In 1940, the double track between Haires and Jackson was also discontinued, though it may have been used for a time as a storage track.

In the early 1950's, the New York Central began a modernization project in an effort to move trains faster at less cost. The Niles terminal was downgraded substantially, with operations moved to an expanded and modernized Robert Young Yard at Elkhart, Indiana. Now that Michigan freight needed to move to Elkhart, the Air Line became a vital, direct link to the city, bypassing the towns of Battle Creek and Kalamazoo. The railroad dramatically improved the Air Line for its new mission. The line was renewed for 60 m.p.h. freight operation and two-mile long controlled sidings were installed near Spring Arbor (CP 2 & 3), Clarendon (CP 4 & 5), Sherwood (CP 6 & 7), Wasepi (CP 8 & 9). A new cut-off was installed in the southeast quadrant of Three Rivers (at a new location known as Three Rivers Jct.), allowing westbound trains rapid entry to the Kalamazoo Branch. Additional sidings and junctions were also remotely controlled at OD in Jackson, Haires and White Pigeon. Three sidings were controlled between White Pigeon and "B" Tower in Elkhart, where the route crossed the former Lake Shore main line and entered Elkhart Yard.

Other improvements in the route from Three Rivers to Elkhart were also made. Don Meints notes that there were two substantial pieces of construction. The longest was a 1.7 mile segment from the east side to the south side of Three Rivers to avoid the cramped wye curve in downtown Three Rivers. The old connection with the Kalamazoo Branch east of Three Rivers was retained for some years, but eventually taken out when it had no use. The second was 0.7 miles of a new wye track at White Pigeon. The old wye curved from the US-12 crossing toward the White Pigeon depot. The new wye went straight across US-12 and then curved to a connection with the Old Road at US-131 crossing, which was given the name of White Pigeon Jct.

A two-mile siding also was put in at Vistula, Indiana. Near Vistula, a sink hole came to life during the rebuilding. "Trackmen were stationed there to watch the sink hole until it was finally stabilized -- in a year or so, as I recall", Don says. He also noted that the waiting room of the Constantine depot was converted into a trainman's bunk room, complete with men's room and shower, to be used by the third brakeman that worked between Elkhart and Constantine as required by Indiana law at the time. To allow the brakeman to get on and off the engine, the train had to slow to a crawl, then work "full throttle" to get up the grade from the depot curve.

Before the centralized traffic control (CTC) equipment could be installed, tested and placed into service, the NYC again began moving most of their freight trains to the Air Line. Don notes that this freight moved for a while by manual block system rules and the NYC placed an operator and block station at the east end of each controlled siding, including Three Rivers Jct., the name given to the junction of the new cut off from the Air Line to the Kalamazoo Branch.  The tower at Wasepi was temporarily reopened on all three shifts, with the tricks being shared by PRR and NYC men.  As the CTC/TCS installation proceeded westward from Jackson, these operator jobs were abolished.

Jackson West dispatcher Frank Curtis, at the CTC panel which controlled the Air Line branch from Jackson to "B" interlocking in Elkhart, Indiana, in 1963. The angled crossing on the right is "OD" interlocking in Jackson. The perpendicular crossing in the center of the photo is Wasepi. Three controlled sidings are in between. The two toggle switches below each control point are for maintainer call and snow melters (see article). The two lever switches below the toggle switches controlled signals and switches and the button on the bottom sent the changes to the control point. [Doug Leffler photo]


The CTC dispatcher's panel was installed in the Jackson station. The railroad was in the process of moving the Jackson East and West Dispatchers to the depot, from their division office near Jackson Jct. Rahn Stokes, son of a NYC dispatcher, said that their family moved to Jackson in 1958, when the NYC closed the Bay City train dispatcher's office. The Michigan Division dispatchers were located in the old Terminal Building on Page Avenue at the west end of Jackson Yard. "The NYC had plans to close this building and move the Division offices to the depot," Rahn said. "The CTC machine was installed at the depot from the start. They moved the train dispatcher for the Middle Division (Jackson to East End, at the east end of the Niles yards) to the depot by himself for a while and he also dispatched the Elkhart Subdivision (commonly called the 'Air Line') using the CTC machine. Others were moved to the depot shortly thereafter.

The CTC machine operated under railroad Rule 261, according to Stokes. "The system only needed one code line. Pulse Length Modulation was the method used to send information back and forth between the 'stations' and the dispatcher's control machine." Similar to binary computer code, the modulation sent consisted of long and short pulses over the wire. "A finite number of bits were at the beginning of each transmission and were used to identify the station being addressed. Every station was continually 'listening' and would drop off as soon as the station address code did not match. Eventually, there would be only one station remaining, the addressee. The remaining bits were sent in the same order every time and if  bit was long it meant the same thing, 'present' or a '1'."

Stokes also noted that several other features could be controlled by the CTC panel. "Maintainer Call, or 'MC' turned on and off a rotating light on the control station equipment bungalow, which was used to tell the signal maintainer to call the dispatcher by telephone. Or, if a train was being held at the control point (CP), "a member of the train could be summoned to call."

"The 'SM' switch on the panel was used to turn the snow melters on and off.  Other switches cleared the signals - for east or west - and operated the turnout (switch) either to reverse (for the siding) or normal. The controls were lined up in a vertical column beneath each CP on the panel. The relays in the CTC machine would send out the code. The station addressed would conform the status of everything at the CP to the corresponding status bits of the control signal."

"After the station had completed conforming, it would send an 'Indication Signal' back to the CTC machine. The CTC machine would then update the display and if a route had been cleared, it would display the red - route locked - light.  If anything was unable to be done - for example, if the turnout couldn't be thrown and locked - the out-of-correspondence light would remain on and the route would not lock".


New York Central train DC-1 heads west at CP-2 near Spring Arbor in 1963.  [Doug Leffler]

"The track speed on the Air Line was now 60 m.p.h.," Stokes continued, "and the speed through the controlled siding turnouts was 30 m.p.h. (Medium Speed), which was the maximum speed for taking or leaving the siding.  The distant signals were two-arm (searchlight) signals, with the home signals being three-arm signals.  On the single track between the control points, automatic block signals were used.  The field logic controlled the aspect of the signal which was displayed at the CP.  The signal engineers designed the system to display certain aspects for certain situations.  The dispatcher could only clear the signal and the logic at the CP would determine what aspect to give."

The Air Line was now a modern, high-speed freight route and it handled some of the New York Central's hottest trains between Elkhart and Detroit, and beyond.  Hot shot freights like LS-3 and van trains (westbound) and perishables with NY-2 and NY-4 (eastbound) pounded the rail.  Other trains were moved out of their way by the dispatcher who was sitting in front of the controls in the Jackson depot.

Local industries continued to be served along the line. A Saturday-only Jackson-Centerville turn (out of Jackson and back) handled local traffic on the old Air Line. Constantine was served by the Kalamazoo-Constantine turn six days a week. Local freights on the Old Road served the towns in Indiana. Don Meints notes that through freights were not allowed to pick up or set out cars, except for bad order cars that developed enroute.

Under New York Central ownership, the Air Line served as an important freight line into the Penn Central era, and later into Conrail. Traffic on the line declined as Michigan downsized as a manufacturing state, and as Conrail rationalized its rail lines. The Air Line was removed west of the former OD Tower (Jackson) in 1982. Remaining east-west through freight moved to Conrail's other "airline", between Toledo and Elkhart. Freight traffic in Michigan that remains uses the former Michigan Central/ NYC main line from Detroit to Elkhart, by way of BO Tower in Kalamazoo.

The modern Air Line division existed from about 1958 to 1980 - only twenty two years - but it was exciting while it lasted. The new Air Line was an example of 1950's Traffic Control System technology which improved train capacity on single track through the use of a responsive CTC installation. CTC and TCS also eliminated hundreds of operator jobs in stations and interlocking towers throughout Michigan. It set the stage for microprocessor-based train control which is used today on major rail lines left in Michigan.

Today, the former 60 m.p.h. Michigan Air Line is a bike trail from Jackson to Concord, Michigan.


[Editor's Note:  Dale Berry, Tim Lab, Doug Leffler, Don Meints, Jerry Pinkepank, Jim Sinclair, Al Smith and Rahn Stokes contributed to this article.]


From Rahn Stokes: In Doug’s photo of Mr. Curtis (above), I can tell from looking at the panel what is likely happening at that moment. All the MCs (Maintainer Call) and SMs (Snow Melters) are off. Jackson is on the right, Elkhart on the left. The signal at CP OD (Jackson) is cleared west and the turnout is lined from the westbound Airline (normal) to the single track. There is a train occupying the two blocks immediately to the west of CP-9 (Wasepi West) and is lined east through Wasepi and Sherwood (CP-9. CP-8, CP-7, CP-6). Mr. Curtis is holding the signal at CP-5 ( Clarendon West) and the signal at CP-2 (Spring Arbor) until he can determine where to make the meet between the eastbound train just west of CP-9 and the westbound train that is not out of Jackson yet. It is going to be at either Clarendon or Spring Arbor.

From Al Smith: The system, although informally called "CTC", was formally labeled "TCS" or "Traffic Control System" by the New York Central.

From Jerry Pinkepank: For decades in North America there were just two principal manufacturers of railroad block signal and interlocking equipment, Union Switch & Signal of Swissvale, Pa (a Pittsburgh suburb) and General Railway Signal of Rochester, New York. Centralized Traffic Control, or CTC, was developed by GRS and first installed between Toledo and Berwick, Ohio, in 1927.  Union Switch and Signal was closely following this development and produced their own version in 1928, which was installed in Michigan on the Pere Marquette between Mount Morris and Bridgeport, Michigan, although they probably had to steer clear of the GRS patents or obtain a license. Mr. Sedgwick North Wright, a GRS signal engineer, is known as “the father of CTC”, and had over 90 patents in his name including critical ones for CTC such as the APB circuit (“Absolute Permissive Block”). There does not appear to have been a trade name registered for Centralized Traffic Control (GRS at first called it “the General Railway Signal Company’s dispatching system”) and very soon  both GRS and US&S were calling their respective systems Centralized Traffic Control. If there was any attempt by either firm to register and defend Centralized Traffic Control or CTC as a trade name, that did not happen. Both of these systems were relay-based, and were unable to handle high volumes of traffic. TCS (“Traffic Control System”) was GRS’s trade name applied to an upgraded version of CTC using transistors instead of relays. GRS developed TCS in 1955 to handle the NYC’s reduction of four track main lines to two tracks. Solid state electronics soon replaced relays in most railroad signaling and US&S soon had their own version, so the distinctive TCS name was eventually dropped. NYC used TCS for a time to designate all CTC, including the earlier relay-based installations, for the sake of simplification of terminology. When John Kenefick went from NYC to Union Pacific in 1968 he applied the TCS name to UP installation as well, but UP soon returned to calling it CTC.

I had direct experience with the problems that pre-solid state CTC machines had in  keeping up with heavy traffic. They were set up so that outgoing commands had priority over incoming indications. When I went to Alliance, Nebraska in 1979 as Assistant Superintendent for Burlington Northern, I had the dispatcher’s office reporting to me, and due to the Powder River Basin coal buildup we were experiencing traffic levels far beyond what had been imagined when the 1950 machine operating Alliance to Ravenna, Nebraska was installed. It routinely ran 20 minutes behind in terms of indications coming back from the field. Dispatchers just bore that in mind and adjusted accordingly, something that amazed me. I regularly sat behind Gary Melton (now treasurer of the ATDA in Cleveland) when he was a 29 year old dispatcher on that district, and watched the operation unfold. At one point I asked him if he had made a particular move to set up a fleet meet, and he just sighed and said, “I don’t think that far ahead!” He was actually dispatching from the train sheet (which was manual, not generated by the machine) and not the CTC board. I think he was adjusting OS times based on known lags in the indications at particular traffic levels.  All this by way of saying that what GRS had to do to create TCS was a significant upgrade to the relay logic in the earlier machines.

Michigan Central

Air Line Division

Station Listing




from Jackson

 Jct. with the Main Line  

 Jackson Junction

     Jct. with Main Line


 OD Interlocking Tower

     Crossing of the LSMS


 Fort Wayne Switch (Haires)

     Jct. with the Ft. Wayne and Jackson Br.

 Snyder's 8.21
 Spring Arbor 10.27
 CP 2 11.0
 CP 3 13.0
 Concord 14.54
 Pulaski 18.73


     Crossing of the LSMS Lansing Br.

     Crossing of the DT&M

 Clarendon 26.84
 CP 4 27.0
 CP 5 29.0
  Tekonsha 33.44
 Burlington 37.42
 Union City 41.61
 Sherwood 48.46
 CP 6 45.20
 CP 7 47.20
 Colon 53.74


     Crossing of the Goshen Branch



     Crossing of the PRR/GR&I

 CP 8 60.70
 CP 9 62.70
 Centerville 64.34

 Three Rivers Junction

     Jct. with Wye to Elkhart

 Three Rivers - (VE Tower) 69.53
 Fabius 74.24
 Corey's 76.83
 Jones 78.87
 Newburg 80.31
 Vandalia 84.80
 Sandy Beach 87.37
 Forest Hall 88.89

 AP Tower

Crossing of the GTW



 Dailey 94.19
 Barron Lake 99.37


     Jct. with the Main Line

 Jct. with the Main Line  


In Doug Leffler’s photo of CP 2 at Spring Arbor, the thick, black code line can be seen coiling into the TCS hut from the pole line. In later years the code line was often buried, or in some cases an alternate technology (“ElectroCode”) was used to send the commands and  indications through the rail, or microwave links are used. All of the CTC installed on the Burlington Northern in the Powder River Basin and adjoining areas receiving the radically increased traffic used microwave links to carry the code.

From Jim Sinclair: This is evolving into a very interesting thread and as a signal maintainer with almost 30 years seniority, I feel I should jump in on this discussion. CTC was General Railway Signal's trade name, and TCS was Union Switch and Signal's trade name. To the best of my knowledge, the NYC used GRS signals and equipment exclusively. Thus, the CTC between Jackson and Elkhart was GRS and Doug's excellent photo of the board in Jackson proves that. A question was raised about whether the Air Line's code line was buried. This route had a very modern pole line that I presume was built when the line was upgraded and the CTC was installed. I find it hard to believe that such a fine (modern) pole line did not include the code line. During my brief signaling career with Conrail (1976-77), I bid on an open maintainer's job in Homer, but it went to a higher seniority man. Had I been awarded that job, I would be able to say for sure whether the code line was buried or in the air. Conventional wisdom tells me it's unusual the code line was buried while all the other signal and communication circuits were in the air. On the other hand, I remember reading an article about the Seaboard burying a "one wire" code line, so it's very possible you're right on the money about the Air Line. I could go on, but there's an ice storm coming in later this evening, so it's back to bed for a little more sleep. Thankfully, I have no pole line left and our code line is really "in the air" via microwave and radio frequency!  Commercial power outages will be my primary problem if it gets severe.

From Rahn Stokes: You are absolutely correct about GRS and the NYC. The reason I say the code line was buried - and you might very well be correct on that also - was an incident one winter. I was still a crew caller at the time and was called by the Superintendent, no less, to see if I would drive a relief crew to Wasepi, because the other crew had died [exceeded their allowed hours of service]. It was in the late evening and I was sleeping. After I got off the phone my mother told me that Jackson was in a State of Emergency due to the fact that the temperature was about 55 degrees, the snow had mostly melted, and was flooding Jackson. When I got to the depot to pick up the superintendent's car, I found the code line had been cut somehow.  My dad was working as the Train Dispatcher and the Elkhart Subdivision was being run by train order and was snarled. The Signal Maintainer was there and had been pleading with a contractor to hire him and his equipment to repair the line. This was during the Penn Central era and the guy had not been paid for the last time he did something for the railroad and wasn't about to do anything else until he was paid. Dirt roads become ice roads when snow melts rapidly, but we made it, they cut some kind of deal with the contractor and the code line was later fixed, and I cut my college class went home and back to bed!

From Jerry Pinkepank: The NYC contracted with General Railway Signal to do the 4 main to 2 mains work (in Ohio and Indiana), and GRS called the resulting system TCS. When John Kenefick left NYC for UP after the PC merger, he began having the same thing installed on the UP, where there was no track reduction issue but a serious need to increase line capacity, and it was called TCS on the UP as well, for several years. The initial CTC installation on the T&OC between Toledo and Berwick in 1927 was done by GRS and they called it Centralized Traffic Control, although perhaps not immediately, as I will explain  in a moment. So GRS used both names, and used the TCS name for the more sophisticated systems that could handle heavy traffic volumes. The TCS name seems to have dropped out of use, probably because today’s solid state systems with CRT displays have overtaken its distinctive features. Apparently GRS failed to trademark the Centralized Traffic Control name or deliberately let it become generic, because when US&S made their first CTC installation (on the PM between Mount Morris and Bridgeport, Michigan) in 1928, they used the same name for it. Because the two installations were less than a year apart, it looks like both GRS and US&S were working on the idea at the same time, and perhaps the term was  already in use to describe a concept signal engineers were already working toward. The excellent little book “Search for Safety,” which was a US&S publication, credits GRS’s Sedgwick North Wright as “the father of CTC” but says on p. 184 that the initial installation was called "the 'General Railway Signal  Company’s dispatching system’, …later generically renamed Centralized Traffic Control”. I doubt that GRS would have used something so awkward as “the General Railway Signal Company’s dispatching system” as a trade name—Wright’s earlier, and related, inventions of the NX interlocking and the APB track circuit (the latter particularly essential to CTC) both immediately had snappy names that could be used in advertising. (APB stands for Absolute Permissive Block, which sounds contradictory until its concept of head block signals is understood—in CTC, those are the signals at the control points).

The conversion to CTC on the Pere Marquette proceeded slowly, as it did on most roads. I know the main line through Lansing was not converted until 1952. Except where line capacity issues dictated the need, the roads usually based the installation on savings in operator positions, and often there would be other reasons for retaining operators, such as when they were required at an interlocking plant, or at a point where trains originated and needed their clearance and orders (orders were still needed for slow orders, to protect track work in progress (Form B), provide warning of temporary footing or clearance hazards, etc.).  Annual budgets were limited.  The 1950 CTC installation on the CB&Q between Alliance and Ravenna, NE that I mentioned earlier dealt with the budget issue by being “single-ended”—one end of the siding had a dual control switch and normal CTC home signal configuration for a control point. The other end had a spring switch and a “take siding” indicator on the mast of the home signal for the direction approaching the siding. Dispatchers would normally have the train approaching the spring switch end hold the main line.

From Tim Lab: The TCS was officially the term that the NYC used for traffic movement over the Elkhart-Jackson Subdivision. It was officially changed in Employee Timetable #5 Dated Sunday April 26, 1959.  In Employee Timetable #3 Dated Sunday April 27, 1958 it was listed as MBS or Manual Block System as they had not yet completed the work on the line.

From Don Meints: About TCS: The 1956 New York Central operating rules included rules 550 thru 562 governed the Traffic Control System. The book also defined that system as “a block-signal system under which train movements are authorized by block signals whose indications supersede the superiority of trains for both opposing and following movements on the same track.”  In other words, trains ran as the signals directed, and without written orders. The system was not, however, “paperless” since all trains were recorded on the dispatcher’s train sheet for crew hours and train consists, but intermediate station times were recorded on the paper printout and not on the train sheet. So, for the NYC, the operating system was “TCS” regardless of manufacturer or electronic system used. 

The Michigan Division of the New York Central (and the LS&MS earlier) covered the lines between Toledo and Elkhart (both the Old Road and the Air Line through Bryan and Kendallville) and the Grand Rapids and G&M Branches, with the superintendent at Elkhart, Indiana.  Since Mr. Collingwood’s article discusses only operations on two or more tracks, the first use was on the Toledo-Bryan-Elkhart line specifically.  Sometime between 1916 and 1920 the above Air Line was taken from the Michigan Division and added to the Toledo Division.  After that change, the Michigan Division included only the Old Road, and the Grand Rapids, G&M, Fayette, and Monroe Branches.  The other lines (Jackson, Lansing, Ypsilanti, Fort Wayne, and part of the DT&M) were under the Lansing Division superintendent located at Hillsdale.  The Lansing Division was later merged into the Toledo Division but after the Grand Rapids and Ypsilanti Branches were transferred to the Michigan Central’s Michigan Division.  (Since the MC had a Grand Rapids Branch, the White Pigeon-Grand Rapids line became the Kalamazoo Branch.) 

In Michigan, the Michigan Central had four divisions: Detroit, Middle, West, and Northern.  It was around 1920 that the Northern and Middle were combined into what was named the Michigan Division.  The Air Line between Jackson and Niles was part of the Middle and then Michigan Division and remained so named until about 1958 when the Jackson-Elkhart line was rebuilt. The part of the Air Line between Jackson and Three Rivers was then renamed the Jackson-Elkhart Subdivision. To my recollection this was the first use of the term “subdivision” on the MC. When it was put in use the White Pigeon-Elkhart segment was also transferred to the Michigan Division in Jackson. With the opening of Young yard in Elkhart, that part of the MC West Division from Porter to Kensington was transferred to the NYC Western Division, and the West Division from Niles to Porter was made part of and dispatched by the Michigan Division in Jackson.

From Rahn Stokes: The systems evolved from TCS/CTC to CATDF ( Conrail's Computer Assisted Train Dispatching Facility in Dearborn) to TrainTrack ( NS's intermediate step from CATDF to Unified Train Control System, which the last I knew they were still testing). The CATD was proprietary to Conrail. An individual who contributed (I'm not sure if he designed it or not) to the project lived in Toledo, and would not move, so Conrail made the Dispatch and Yard  System's headquarters in Dearborn (for him). From the DYS they monitored all six of Conrail's CATDFs throughout the country. If any display needed to be changed because of a new switch or something, they changed it from the DYS in Dearborn. An interesting bit of information about this system is that while it appeared modern when looking at it from the CATDF, which it was, the circuitry, signals, relays, turnout motors etc., were very old and interfaces had to be developed for them until they had to be replaced. The interfaces allowed modern technology to work seamlessly with that of yesteryear. TrainTrack uses the Windows Operating System, and before you fall over in a dead faint, Windows is not part of the vital circuits.

From Jerry Pinkepank: We need to be a little careful about generalizing the rulebook terminology to all purposes.  The rulebook (and timetables) would not distinguish between CTC and TCS, just adopting one term for all the Rule 261 situations, which is sensible, but from a signal engineering standpoint, that did not, for example,  thereby upgrade Toledo-Berwick (which was still in operation until 1961 with the original machine at Fostoria) to TCS from a technological viewpoint.

I just noticed on p. 189 of “Search for Safety” what I think is a reference to TCS without naming it:  “In the early 1950’s relay coding systems reached the high point in design. Still there existed a need to handle larger installations with heavy traffic—what was needed was a reliable high-speed system. To satisfy the situation, a solid state coding system using transistor circuitry was developed by W. D. Hailes of General Railway Signal Company in 1955. Similar developments at Union Switch & Signal Company were pioneered by Dr. P.N. Bossart, T.J. Blocker, W.H. Moore and E.R. Callender.”  Reading between the lines, GRS developed this for NYC and then US&S made a competitive response (This is a US&S publication). The 1955 date raises an interesting question. Did Alfred Perlman request and get action that rapidly, or was the work actually begun under his predecessors?

Something that may not be widely appreciated about CTC/TCS is the huge increase in traffic handling capacity that it caused on existing double track when installed there. This always involved new crossovers with facing points, replacing the low-speed trailing point crossovers which were conventional on ABS or dark double track, which were intended to prevent a picked point from throwing a train into opposing traffic. The facing point crossovers were all at interlockings where the mechanism (often including point locks) insured against that kind of accident.  With CTC/TCS, every control point became an interlocking, the points held firmly by the dual control switch machine, and this included the crossovers. The immediate effect was that overtakes were much more efficient to handle, and on a line with a wide divergence of train speeds, this had a dramatic effect on capacity.  It meant that in some cases you could actually increase capacity when going from double to single. I am sure this was the case for my 1987 CTC plan on the ex-Milwaukee between Milwaukee and the Twin Cities, (when I was AVP Operations Analysis at Soo Line 1986-1989),  which was implemented shortly after that. The disparity between the speeds of the Amtrak, intermodal, merchandise and bulk trains was so great that delays from overtakes were a constant problem, but became easy after the single-tracking. In the case of NYC’s TCS reducing from four tracks to two, this is the principal reason it was practical to do that. The four tracks were set up to handle 35mph freight trains on the freight mains and 80mph passenger trains and 50 mph “high car” freight trains on the passenger mains.  With frequent high speed facing point crossovers (NYC’s long-point No. 20 crossovers were designed for 50mph diverging speeds), the problem of speed diversity and frequent overtakes was minimized.

From Rahn Stokes: The CPs on the Elkhart Subdivision all used commercial power. They had battery backup and when the power went off at a CP the dispatcher knew it because the station light above the CP, which would light when the CTC machine was receiving a message from that station, would operate in reverse. It would stay on all of the time and go off when a signal was being received. The code line only handled the CTC instructions. The intermediate signals were on the pole line as well as the other communications lines. The level of detail on the Elkhart subdivision was a lot better than that on CATDF or TrainTrack. CTC displayed the individual blocks of the intermediate signals. The new displays only show CP line ups and whether the block between CPs is occupied.

Shortly after the Elkhart Subdivision reopened for service, an interesting thing happened at Wasepi where the PRR crossed the Air Line. The diamond was controlled from the CTC at Jackson for a while, then eventually it became a first come first served automatic interlocking. My dad (a train dispatcher in Jackson) told me that he saw a train on the PRR run a red signal and cross the diamond with out stopping. Nothing was lined up so there was no collision. Apparently at this early stage of operation the PRR crews were not accustomed to traffic on the Air Line. The NYC had stopped its directional running from Niles over the Air Line, so there was not much traffic on the NYC at Wasepi until the line was improved.

From Doug Leffler: One little bit of information that Jim Tarbell told me years ago was that the local Jackson shop forces used the old pre-1958 rebuilt Air Line for breaking in steam locomotives after they were rebuilt.



     1.  Meints, Graydon M., Michigan Railroad Lines Volume I, © 2005 Michigan State University Press. p. 295-296.

     2.  Meints, Graydon M., Michigan Railroad Lines Volume I, © 2005 Michigan State University Press. Note 39 on p. 360.

     3.  Brignano, Mary and McCullough, Hax, The Search For Safety: A History of Railroad Signals and the People Who Made Them, commissioned and published by the Union Switch & Signal Division of American Standard, Swissvale, PA, 1981.  The company has since become WABCO Signals.

Latest Story Update:  12/31/2008