About pairs of numbers for east- and west-bound trains

Amtrak Unlimited Discussion Forum

Help Support Amtrak Unlimited Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Siegmund

Lead Service Attendant
Joined
Nov 19, 2018
Messages
470
Location
northwestern Montana
Kind of a weird and specific question here, but one I've been curious about for decades, and never stumbled across an answer to.

Amtrak's tendency has been to use pairs of consecutive numbers for each route. Normal enough, most railroads did. But the western long-distance routes use odd-then-even pairs, as do the Silvers and the Crescent (1-2 Sunset, 19-20 Crescent, 97-98 Silver Meteor)... while the other eastern long-distance routes use even-then-odd pairs (58-59 City of New Orleans, 40-41 for the old Broadway limited, 66-67 Night Owl, 52-53 for the Floridian and later Auto-Train despite the other florida trains doing it the other way around) as do almost all the short haul routes everywhere (downeaster 680 through 689; wolverines 350 to 355; capitol corridor starting from 520.)

Why would they do that? It offended my inner sense of order when I was a teen and it still does.

Perusing historical timetables, even-number-lower seems to be primarily a Pennsylvania-ism, though there are a handful of other example elsewhere in the country (there was a seaboard-L&N joint train from Miami to New Orleans that was 38/39 contrary to Seaboard's normal practice elsewhere, for instance.) It doesn't look like Penn Central inflicted the Pennsylvania method on the surviving NYC trains in 1968 or 1969.

Amtrak used it in the northeast as soon as it assigned its own numbers (including the Empire Service), but only gradually moved to it in the midwest. The early timetables show it used for Milwaukee and Detroit trains but not for St. Louis, Carbondale, or Quincy. I had a hypothesis that it mattered whether the inbound or outbound train departed earlier in the morning - but that doesn't seem to work either. By the sometime in the late 70s it seems to have become systematized.

Anybody have any insight into the history behind it?
 
Interesting...I had wondered about that, too...but never did figure it out, either. One possibility is that they wanted paired trains to start with the same digit...as in 40 and 41, rather than 39 and 40... Or, maybe there was an unused number, since there were not always the same number of trains in each direction, in multiple train routes, and rather than skip the number, they used it for the next pair...
Not sure if it was for those reasons, or just someone's whim....🤔
 
Interesting...I had wondered about that, too...but never did figure it out, either. One possibility is that they wanted paired trains to start with the same digit...as in 40 and 41, rather than 39 and 40... Or, maybe there was an unused number, since there were not always the same number of trains in each direction, in multiple train routes, and rather than skip the number, they used it for the next pair...
Not sure if it was for those reasons, or just someone's whim....🤔
You mean like the Crescent being 19 and 20?

Guess that leaves out that theory.

I personally thing an Amtrak executive decided but couldn't remember all the numbers that followed each other so he used the ones he could remember. :)
 
Last edited:
You men like the Crescent being 19 and 20?

Guess that leaves out that theory.

I personally thing an Amtrak executive decided but couldn't remember all the numbers that followed each other so he used the ones he could remember. :)
I wasn't saying they were all like that, just the ones that were...besides, The Crescent became an Amtrak train, many years after the rest of them.
But your theory is as good as any...:)
 
Some were inherited from the previous railroads and never changed IIRC - the Coast Starlight comes to mind with its non-sequential numbers. I suspect Amtrak had a system at one point. Airlines have had weird numbering protocols as well over the years: AA varied even numbers with odd to indicate direction of flight, although rarely were the numbers sequential on the same route. (Frequent flyers even knew which direction meal service was starting based on odd or even flight numbers.)
 
I always thought (and it seems to be my experience, at least in the Northeast) that the southbound and westbound trains had odd numbers and the northbound and eastbound trains had even numbers. Thus NER 65/67 is the overnight from Boston to Newport News, whereas 66 is the northbound train. Westbound Capitol our of DC is #29, eastbound out of Chicago is #30. Southbound Silver Meteor is #97, the Northbound is #98. And so forth. There may be some exceptions, but I don't remember riding on any of them.
 
Some were inherited from the previous railroads and never changed IIRC - the Coast Starlight comes to mind with its non-sequential numbers.

The Coast Starlight did inherit an SP weirdness that trains were odd toward San Francisco and even away from it, rather than odd south/west and even north/east. In the early 70s timetables it was shown as 11 Seattle-Oakland, 12 Oakland-LA, 13 LA-Oakland, 14 Oakland-Seattle. (Did it "inherit" the old Portland-Oakland Cascade's numbers 11 and 12? Maybe, sort of...)

I suspect Amtrak had a system at one point.

I am hoping someone who knows what the system was will wander into this thread someday :)

My personal theory is that it started with inheriting the Broadway and National Limited and NEC and Keystone service from the Pennsylvania, and that at some point several years later the NEC method got applied to all of the other short-haul routes... but I am not actually recalling a time when NEC and Midwest and Western short-haul were in the same department, though I do recall NEC, short-haul, and long-distance as 3 departments.

It just seems so random to (for instance) use 30-31 for the National Limited and 32-33 for the Shenandoah, and then turn around and use 29-30 for the Capitol Limited - or to use 48-49 for the Lake Shore and 63-64 for the Niagara Rainbow (and later Maple Leaf) at the same time.
 
I have been wondering how the trains got their numbers also. For example, how is it that the Sunset Limited is 1 and 2, the SWC is 3 and 4 etc.? I had been operating on the assumption that perhaps the numbers reflect the order they joined Amtrak, but I really don't know. Do others?
 
I have been wondering how the trains got their numbers also. For example, how is it that the Sunset Limited is 1 and 2, the SWC is 3 and 4 etc.? I had been operating on the assumption that perhaps the numbers reflect the order they joined Amtrak, but I really don't know. Do others?
A major number rationalization took place in the early '70s. Unfortunately so far I have been unable to locate the thread that discussed that in this forum. Next I am going to forage in a few other fora.
 
I have been wondering how the trains got their numbers also. For example, how is it that the Sunset Limited is 1 and 2, the SWC is 3 and 4 etc.? I had been operating on the assumption that perhaps the numbers reflect the order they joined Amtrak, but I really don't know. Do others?
The Transcontinental trains do have a numbering pattern, the Sunset Limited is 1 and 2, The Chief is 3 and 4, The Zephyr is 5 and 6, and the Empire Builder is 7 and 8. The Sunset Limited did keep its SP numbers, I'm not sure how many other trains in the Amtrak system kept their numbers. Like the Starlight, its predecessor trains were 98 and 99 (Coast Daylight), 75 and 76 (The Lark), 11 and 12 (The Streamlined Cascade), 17 and 18 (Cascade pre streamlining), 9 and 10 (Shasta Daylight) and 7 and 8 (Shasta). So the Starlight has at least 1 of the numbers from one of its predecessor trains.

It would be nice for Amtrak to rationalize its numbering system, the problem with that is there aren't really enough trains for that to matter. Its confusing that many trains on the West Coast are 5xx or 7xx, but they don't tend to run near enough to each other for that to cause confusion.
 
The logic of the numbers of the Transcon (Midwest to West Coast) trains has always been obvious to me: they're numbered in sequence from south to north: Sunset is 1/2, then go north to the SW Chief which is 3/4, north to the CZ for 5/6, and then further north to the EB for 7/8. 9/10 was initially not used but did get used for the North Coast Hiawatha although that was between 5/6 and 7/8 geographically. Why the CS is 11/14 has been explained many times but in short 12 and 13 were the hidden numbers from when each train was first westbound and then after Oakland eastbound back in a day when the trains operated with timetable authority and odd or even numbers were important.

It's also been obvious to me that Amtrak initially took the PC numbering system and built the rest of the system around it. NE Corridor number series (although maybe not specific trains) were mostly unchanged: WAS-NYP in the low 100s, BOS trains in the high 100s, NYP-PHL "clockers" in the 200s, New Haven - Hartford - Springfield shuttles (back then with through cars from the Corridor trains) were in the 400s, and PHL-HAR trains in the 600s. Find a PC schedule from around 1970 and you'll see they're all there. So why the other corridors ended up in the 300s, 500s, and 700s would be because PC didn't use them for corridors.

As for why the western trains ended up numbered in odd/even pairs while many of the eastern trains ended up even/odd, who knows. Perhaps different people doing it; perhaps with eastern bias, they expected more trains to be added in the east so left room by initially just using x0/x1 for the east coast - Chicago trains (30/31 for the NYP-KCY National Limited, 40/41 for the Broadway Limited, and 50/51 for what would become the Cardinal). Doing it that way, if a second frequency was added in that market (which perhaps they were thinking would happen if Amtrak was successful - plan for success, not failure), it could then have the x2/x3 number and fit in the scheme (which you could argue happened in 1975 when the LSL was added via a different route but still a second NY-CHI train and was given 48/49 so at least then, 4x meant NY-Chicago.

Another example, early on, of using the same series for the same market as trains were added was the Montreal services: 60/61 for WAS-NYP-Montreal via Springfield, 68/69 for NYG-Montreal via Albany. But as trains were dropped and others added, and then numbers became tight, some of the logic became lost and we ended up with inconsistencies such as the overnight WAS-BOS being 66/67 (65 as a weekend alternate schedule for 67 came later). At one time, the 6x series was fully used: 60/61 for the WAS-Montreal, 62/65 was at one time a NYG-SYR train (providing a late train from NYG to SYR and an early morning train from SYR rather than today having to wait for the first through train from NFL), 63/64 I believe was initially the NYG-Detroit train before more recently going to Toronto, and then 66/67 and 68/69 as mentioned above.

But at some point, particularly as corridors became successful, numbers became scarce and then sometimes it became just find an available pair of numbers, particularly if a <100 pair was desired. Originally, the NYG-ALB trains were mostly 7x until they outgrew it and most of the Empire Service trains were moved to 2xx which was now available as the Clockers had been discontinued.

But let's face it, if you knew in 1971 what the 2021 system would look like (ignore the Covid changes), you might well have developed the initial numbering system quite differently. But of course they didn't so they did what made sense back then.
 
When the 'Maple Leaf' was inaugurated......each train had three numbers. The New York bound train departed west from Toronto to Hamilton (39 miles) so carried an odd number #97. At Hamilton it swung around the end of Lake Ontario........now eastbound to Niagara Falls and carried even number #96. At Niagara Falls it became Amtrak #64.

The westbound Maple Leaf from New York was Amtrak #63.....VIA #99 at Niagara Falls then eastbound VIA #98 at Hamilton to Toronto.

http://www.timetables.org/full.php?group=19810426&item=0026
NF.jpg
 
The logic of the numbers of the Transcon (Midwest to West Coast) trains has always been obvious to me: they're numbered in sequence from south to north: Sunset is 1/2, then go north to the SW Chief which is 3/4, north to the CZ for 5/6, and then further north to the EB for 7/8. 9/10 was initially not used but did get used for the North Coast Hiawatha although that was between 5/6 and 7/8 geographically. Why the CS is 11/14 has been explained many times but in short 12 and 13 were the hidden numbers from when each train was first westbound and then after Oakland eastbound back in a day when the trains operated with timetable authority and odd or even numbers were important.

It's also been obvious to me that Amtrak initially took the PC numbering system and built the rest of the system around it. NE Corridor number series (although maybe not specific trains) were mostly unchanged: WAS-NYP in the low 100s, BOS trains in the high 100s, NYP-PHL "clockers" in the 200s, New Haven - Hartford - Springfield shuttles (back then with through cars from the Corridor trains) were in the 400s, and PHL-HAR trains in the 600s. Find a PC schedule from around 1970 and you'll see they're all there. So why the other corridors ended up in the 300s, 500s, and 700s would be because PC didn't use them for corridors.

As for why the western trains ended up numbered in odd/even pairs while many of the eastern trains ended up even/odd, who knows. Perhaps different people doing it; perhaps with eastern bias, they expected more trains to be added in the east so left room by initially just using x0/x1 for the east coast - Chicago trains (30/31 for the NYP-KCY National Limited, 40/41 for the Broadway Limited, and 50/51 for what would become the Cardinal). Doing it that way, if a second frequency was added in that market (which perhaps they were thinking would happen if Amtrak was successful - plan for success, not failure), it could then have the x2/x3 number and fit in the scheme (which you could argue happened in 1975 when the LSL was added via a different route but still a second NY-CHI train and was given 48/49 so at least then, 4x meant NY-Chicago.

Another example, early on, of using the same series for the same market as trains were added was the Montreal services: 60/61 for WAS-NYP-Montreal via Springfield, 68/69 for NYG-Montreal via Albany. But as trains were dropped and others added, and then numbers became tight, some of the logic became lost and we ended up with inconsistencies such as the overnight WAS-BOS being 66/67 (65 as a weekend alternate schedule for 67 came later). At one time, the 6x series was fully used: 60/61 for the WAS-Montreal, 62/65 was at one time a NYG-SYR train (providing a late train from NYG to SYR and an early morning train from SYR rather than today having to wait for the first through train from NFL), 63/64 I believe was initially the NYG-Detroit train before more recently going to Toronto, and then 66/67 and 68/69 as mentioned above.

But at some point, particularly as corridors became successful, numbers became scarce and then sometimes it became just find an available pair of numbers, particularly if a <100 pair was desired. Originally, the NYG-ALB trains were mostly 7x until they outgrew it and most of the Empire Service trains were moved to 2xx which was now available as the Clockers had been discontinued.

But let's face it, if you knew in 1971 what the 2021 system would look like (ignore the Covid changes), you might well have developed the initial numbering system quite differently. But of course they didn't so they did what made sense back then.
Excellent analysis...probably the best explanation given here, so far...
 
When the 'Maple Leaf' was inaugurated......each train had three numbers. The New York bound train departed west from Toronto to Hamilton (39 miles) so carried an odd number #97. At Hamilton it swung around the end of Lake Ontario........now eastbound to Niagara Falls and carried even number #96. At Niagara Falls it became Amtrak #64.

The westbound Maple Leaf from New York was Amtrak #63.....VIA #99 at Niagara Falls then eastbound VIA #98 at Hamilton to Toronto.

http://www.timetables.org/full.php?group=19810426&item=0026
View attachment 20549
Prior to the Amtrak/VIA Maple Leaf, the Penn Central-TH&B-Canadian Pacific "Dayliner" RDC between Buffalo and Toronto also did similar...
From Buffalo to Welland, it was PC #371, from Welland to Hamilton, it was TH&B #371, and from Hamilton to Toronto, it was CP #322. Going the other way, its numbers were 321-376-376.

https://timetableworld.com/ttw-viewer?token=e25f29cd-61ce-47a4-b8c8-ae5eaae2ed6f
Just wanted to add, that the connecting New York to Buffalo PC #'s were 61 westward, and 62 eastward...
 
Prior to the Amtrak/VIA Maple Leaf, the Penn Central-TH&B-Canadian Pacific "Dayliner" RDC between Buffalo and Toronto also did similar...
From Buffalo to Welland, it was PC #371, from Welland to Hamilton, it was TH&B #371, and from Hamilton to Toronto, it was CP #322. Going the other way, its numbers were 321-376-376.

Although they used CP RDC Equipment.......they were Penn Central and later Conrail trains between Buffalo and Welland ON for nearly 10 years after Amtrak was formed. There was no Amtrak involvement other than the New York connection in Buffalo.
 
Although they used CP RDC Equipment.......they were Penn Central and later Conrail trains between Buffalo and Welland ON for nearly 10 years after Amtrak was formed. There was no Amtrak involvement other than the New York connection in Buffalo.
Right...I guess I wasn't clear about that. And during the period before the Amtrak Lake Shore ran, that RDC was the only way to go west from Buffalo, without backtracking all the way to NYC...
 
Timetable 1971 05 has trains #11 and 12 running from Seattle to San Diego. The 1971 11 timetable split things up and named them. Coast Starlight was 11 and 14, and was listed separately as Seattle to Portland. The combo of Coast Starlight and Coast Daylight was numbered 11-12 southbound, 13-14 northbound, running from Seattle to San Diego. The Coast Daylight was listed separately as 12 and 13, running from San Francisco to San Diego. Also, the LA to San Diego run was split out as the San Diegans, numbered in the 700s. The Coast Daylight ceased as a separate train in May of 1974, leaving the Coast Starlight, Seattle to LA, as 11 and 14, with the LA to San Diego run as multiple San Diegans. Yeah, I've been spending a lot of time with the early schedules.
 
Some were inherited from the previous railroads and never changed IIRC - the Coast Starlight comes to mind with its non-sequential numbers. I suspect Amtrak had a system at one point. Airlines have had weird numbering protocols as well over the years: AA varied even numbers with odd to indicate direction of flight, although rarely were the numbers sequential on the same route. (Frequent flyers even knew which direction meal service was starting based on odd or even flight numbers.)
The Starlight's numbers were not really inherited from SP, though they did conform to SP's operating practices. The train that became the Starlight was basically a run through combination of SP trains 98 and 99, the Coast Starlight, and 11 and 12, the Cascade. It initially ran under those numbers 98/99 Oakland-LA, 11/12 Oakland Portland until Amtrak, in concert with the railroads, renumbered everything through 1971. I am not sure when 98/99, 11/12 got renumbered. I know renumbering the Super Chief from its Santa Fe numbers 17&18 to 3&4 didn't happen until about December that year.

Trains were still operating under timetable and train order rules in 1971with the employee timetable itself giving authority to occupy track unless overridden by train order. Therefore, train numbers had to conform to individual railroad conventions and operating practices.

In SP's case, everything was either eastbound or westbound, SP had no "railroad" north or south. The direction was determined whether a train was operating toward San Francisco (westbound with odd train numbers) or away from San Francisco (eastbound with even train numbers). So a train that operated through the Bay Area could not have one, single train number, since it would change "direction" there. So, in conformance with SP operating practice, Amtrak's Starlight was jointly assigned train numbers by SP and Amtrak, becoming 11 west (south) bound between Portland and Oakland and 12 east (south) bound between Oakland and LA, 13 west (north) bound between LA and Oakland and 14 east (north) bound between Oakland and Portland.

12/13 only appeared sporadically and early in Amtrak's public timetables. IIRC mainly when operation north of Oakland was triweekly, the train was named the Coast Daylight/Starlight and it was 12/13 on the Oakland days and 11-12/13-14 on the Seattle days. But for the most part, the train just showed as 11/14 in Amtrak's public timetables. 12 and 13 were an SP operational thing and were in their employee timetables, not Amtrak's public ones. The short lived Spirit of California had similar split numbers since it went to/from Sacramento and SP still was still using timetable authority when the "Medflyer" was running.

11 was the only number that was the same as SP's antecedent train. That was by happenstance, not design.

SP, like most railroads, stopped using timetable authority in the 80s, so train numbers have no operating significance or authority now. Amtrak could renumber the Starlight as 11/12 if they wanted to. UP would not care (plus UP uses all 4 compass directions).
 
Last edited:
Kind of a weird and specific question here, but one I've been curious about for decades, and never stumbled across an answer to.

Amtrak's tendency has been to use pairs of consecutive numbers for each route. Normal enough, most railroads did. But the western long-distance routes use odd-then-even pairs, as do the Silvers and the Crescent (1-2 Sunset, 19-20 Crescent, 97-98 Silver Meteor)... while the other eastern long-distance routes use even-then-odd pairs (58-59 City of New Orleans, 40-41 for the old Broadway limited, 66-67 Night Owl, 52-53 for the Floridian and later Auto-Train despite the other florida trains doing it the other way around) as do almost all the short haul routes everywhere (downeaster 680 through 689; wolverines 350 to 355; capitol corridor starting from 520.)

Why would they do that? It offended my inner sense of order when I was a teen and it still does.
The organizing principal seems to be organizing the core eastern trains ("core" meaning overnight or the longest daytime trains) around multiples of 10: #s 20, 30, 50, 80 and 90 are all in use; 40 and 60 were (Broadway and Montrealer). The reverse direction numbers then end with the corresponding "1" unless that was taken.

The only outlier that is not consistent with that theory is the westbound CL being # 29 when there is no # 31 anywhere in the system and has not been since before the CL split off from being a section of the Broadway (when it was 440/441, IIRC) and became its own entity. Maybe they wanted to leave 31 available for another outgrowth of the Sunset or TE to be consistent with the 1-21 / 2-22 pattern? (Although I think the TE's old Houston section # was 521/522 and not 31/32.)
 
The organizing principal seems to be organizing the core eastern trains ("core" meaning overnight or the longest daytime trains) around multiples of 10: #s 20, 30, 50, 80 and 90 are all in use; 40 and 60 were (Broadway and Montrealer). The reverse direction numbers then end with the corresponding "1" unless that was taken.

The only outlier that is not consistent with that theory is the westbound CL being # 29 when there is no # 31 anywhere in the system and has not been since before the CL split off from being a section of the Broadway (when it was 440/441, IIRC) and became its own entity. Maybe they wanted to leave 31 available for another outgrowth of the Sunset or TE to be consistent with the 1-21 / 2-22 pattern? (Although I think the TE's old Houston section # was 521/522 and not 31/32.)
Hmm?
19/20
29/30
48/49
97/98
91/92

I’m not seeing a 0 and 1 pairing.
 
Hmm?
19/20
29/30
48/49
97/98
91/92

I’m not seeing a 0 and 1 pairing.
40/41 (in the past)
50/51
60/61 (in the past)

The others:
20, Crescent, 21 is taken by the TE so the opposite direction is 19
30, CL, this is the one that doesn't fit my theory, as I acknowledged
80, Carolinian, 81 was taken by the Tampa section of the Silver Star when it was split so the opposite direction is 79
90, Palmetto, 91 is taken by the Silver Star (originally the Miami section before it was combined) so the opposite direction is 89

The others aren't 10s so I wasn't addressing those.

I think the plan may have been to make all the "core" East Coast - Chicago trains as multiples of 10 (and I meant to mention that in my other post but forgot, so apologies for that). The Lake Shore (48/49) wasn't part of Original Amtrak in 1971, and the Silver Services don't touch Chicago so they would fall outside of this. Then I suspect Amtrak kept the 10s idea on the other longer-distance East Coast services that developed, like the Cardinal and Montrealer, and the Crescent, which was not Original Amtrak either or maybe that would have been 20/21 and the TE would have been numbered something different?

Hey this is just a theory! If someone has a better one, or if the original Amtrak Numbering Officer is here, I am glad to be corrected as to what the thinking was / is. Or we should just acknowledge it is completely random, end the speculation and ask for the thread to be closed.
 
Last edited:
40/41 (in the past)
50/51
60/61 (in the past)

The others:
20, Crescent, 21 is taken by the TE so the opposite direction is 19
30, CL, this is the one that doesn't fit my theory, as I acknowledged
80, Carolinian, 81 was taken by the Tampa section of the Silver Star when it was split so the opposite direction is 79
90, Palmetto, 91 is taken by the Silver Star (originally the Miami section before it was combined) so the opposite direction is 89

The others aren't 10s so I wasn't addressing those.

I think the plan may have been to make all the "core" East Coast - Chicago trains as multiples of 10 (and I meant to mention that in my other post but forgot, so apologies for that). The Lake Shore (48/49) wasn't part of Original Amtrak in 1971, and the Silver Services don't touch Chicago so they would fall outside of this. Then I suspect Amtrak kept the 10s idea on the other longer-distance East Coast services that developed, like the Cardinal and Montrealer, and the Crescent, which was not Original Amtrak either or maybe that would have been 20/21 and the TE would have been numbered something different?

Hey this is just a theory! If someone has a better one, or if the original Amtrak Numbering Officer is here, I am glad to be corrected as to what the thinking was / is. Or we should just acknowledge it is completely random, end the speculation and ask for the thread to be closed.
If we closed threads because their content was somewhat speculative, to exaggerate a bit, there would be just a few open threads left 😬
 
Back
Top