Even (or odd) numbers only


Mark Mathu
 

Back in the pre-World war II days, the railroad I'm researching
(GB&W) used even numbers only for boxcars, and odd numbers only for
flatcars and gondolas. I've seen similar numbering on other Midwest
railroads... what was the reason for numbering cars like that?

--
Mark Mathu
Whitefish Bay, Wis.
The Green Bay Route: http://www.greenbayroute.com/


Ian Cranstone
 

On 14-Dec-04, at 7:00 AM, Mark Mathu wrote:

Back in the pre-World war II days, the railroad I'm researching
(GB&W) used even numbers only for boxcars, and odd numbers only for
flatcars and gondolas. I've seen similar numbering on other Midwest
railroads... what was the reason for numbering cars like that?
As I recall the Canadian Northern and Canadian Pacific prior to 1913 did the same thing (even number for roofed cars, odd numbers for open cars). My theory is that it was a simple check digit -- if you saw an even number on a switchlist or some such document for a coal gondola, you'd know something was wrong.

Ian Cranstone
Osgoode, Ontario, Canada
lamontc@...


Richard Hendrickson
 

On 14-Dec-04, at 7:00 AM, Mark Mathu wrote:

Back in the pre-World war II days, the railroad I'm researching
(GB&W) used even numbers only for boxcars, and odd numbers only for
flatcars and gondolas. I've seen similar numbering on other Midwest
railroads... what was the reason for numbering cars like that?
and Ian Cranstone responded:

As I recall the Canadian Northern and Canadian Pacific prior to 1913
did the same thing (even number for roofed cars, odd numbers for open
cars). My theory is that it was a simple check digit -- if you saw an
even number on a switchlist or some such document for a coal gondola,
you'd know something was wrong.
That's always been my understanding as well, Ian. The practice was fairly
widespread in the early 20th century and some RRs, notably C&NW and GB&W,
continued it through the '50s and into the '60s. Since the C&NW, at least,
had a large freight car fleet numbered in that fashion, changing the
numbering system would have been a major headache.

Richard H. Hendrickson
Ashland, Oregon 97520


Tim O'Connor
 

My theory is that it was a simple check digit -- if you saw an
even number on a switchlist or some such document for a coal gondola,
you'd know something was wrong.
That's always been my understanding as well, Ian.
There's only one problem with this theory. The C&NW had some
car series with even numbers only... It seems to me that just
as likely an explanation is that it provided a neat way to
provide for future expansion without introducing a new number
series. I know C&NW did "backfill" this way with a large group
of PS-1's that were interleaved with earlier orders of the
same type of car.

The only railroad I've heard of that implemented a reliable
"check digit" scheme is KCS. A check digit is a mathematical
function that takes the other digits as input, such that if
ANY of the other digits is wrong, then the check digit also
will be wrong. (Error detection and even correction is very
common in computers because storage & communication devices
are inherently error prone.)

Tim O.


Ian Cranstone
 

On 14-Dec-04, at 4:25 PM, Tim O'Connor wrote:


My theory is that it was a simple check digit -- if you saw an
even number on a switchlist or some such document for a coal gondola,
you'd know something was wrong.
That's always been my understanding as well, Ian.
There's only one problem with this theory. The C&NW had some
car series with even numbers only... It seems to me that just
as likely an explanation is that it provided a neat way to
provide for future expansion without introducing a new number
series. I know C&NW did "backfill" this way with a large group
of PS-1's that were interleaved with earlier orders of the
same type of car.
This probably just means C&NW did this for some other reason. Only some of the Canadian roads opted into this odd-even numbering thing --
after mentioning Canadian Northern (with their fleet absorbed into Canadian National during the 1920s) and Canadian Pacific through about 1913, it also occurs to me that the Temiskaming & Northern Ontario (later Ontario Northland) also did this through the 1920s, and the Quebec Central also observed this practice in the first half of the 20th century. They all followed the same even number for a house car, and odd number for an open car practice.

CP also "backfilled" at least one series (295000-series auto cars as I recall) when they discontinued the odd-even numbering practice about 1913 with new odd-number cars interleaved into the existing even-number cars), although other existing series were just left the way they were. New series were numbered consecutively.

The only railroad I've heard of that implemented a reliable
"check digit" scheme is KCS. A check digit is a mathematical
function that takes the other digits as input, such that if
ANY of the other digits is wrong, then the check digit also
will be wrong. (Error detection and even correction is very
common in computers because storage & communication devices
are inherently error prone.)
Somewhere around here I have the KCS check digit formula (which strikes me as very useful if you're modelling a six-digit KCS car), as well as the ACI check sum formula (Iess necessary unless you're modelling ACI labels in a rather large scale!). Of course check sums aren't perfect either, but they do reduce the odds of a mistake considerably.

Ian Cranstone
Osgoode, Ontario, Canada
lamontc@...


Mark Mathu
 

Tim O'Connor wrote:

My theory is that it was a simple check digit -- if you saw an
even number on a switchlist or some such document for a coal
gondola, you'd know something was wrong.
That's always been my understanding as well, Ian.
There's only one problem with this theory. The C&NW had some car
series with even numbers only... It seems to me that just as
likely an explanation is that it provided a neat way to provide
for future expansion without introducing a new number series. I
know C&NW did "backfill" this way with a large group of PS-1's
that were interleaved with earlier orders of the same type of car.

I don't think I follow that line of reasoning...
For the sake of illustration, if a railroad had an even-number only
car series 1000-1098 (50 cars) so that they could later expand the
series without introducing a new number series by adding 1001-1099
(50 cars) -- couldn't they just as easily numbered the first series
1000-1049, and the later series 1050-1099 and not introduce a new
number series?

- Mark Mathu
Whitefish Bay, Wis.


Tim O'Connor
 

Mark, read your question aloud to yourself and I think you'll
have your answer.

For the sake of illustration, if a railroad had an even-number only
car series 1000-1098 (50 cars) so that they could later expand the
series without introducing a new number series by adding 1001-1099
(50 cars) -- couldn't they just as easily numbered the first series
1000-1049, and the later series 1050-1099 and not introduce a new
number series?

- Mark Mathu
Whitefish Bay, Wis.


Mark Mathu
 

Richard Hendrickson wrote:
The practice was fairly widespread in the early 20th century and
some RRs, notably C&NW and GB&W, continued it through the '50s and
into the '60s.
In the case of the GB&W, at least a few of the even-numbers-only
boxcars made it into the early 1970s. They seemed to have used it
exclusively before the 1940s.

The GB&W had a "dry spell" in the late 1930s and 1940s when no new
equipment was purchased, but starting with their first new orders of
boxcars and hoppers in 1950 they switched to using all numbers in
the car series.


Brian Paul Ehni <behni@...>
 

Here's a thought: Perhaps they did it to make it look like the railroad had
more equipment than it actually did. Anyone looking at the records would
know, but a casual observer might think different.
--
Thanks!

Brian Ehni

From: Mark Mathu <mark@...>
Reply-To: <STMFC@...>
Date: Wed, 15 Dec 2004 18:46:30 -0000
To: <STMFC@...>
Subject: [STMFC] Re: Even (or odd) numbers only



Richard Hendrickson wrote:
The practice was fairly widespread in the early 20th century and
some RRs, notably C&NW and GB&W, continued it through the '50s and
into the '60s.
In the case of the GB&W, at least a few of the even-numbers-only
boxcars made it into the early 1970s. They seemed to have used it
exclusively before the 1940s.

The GB&W had a "dry spell" in the late 1930s and 1940s when no new
equipment was purchased, but starting with their first new orders of
boxcars and hoppers in 1950 they switched to using all numbers in
the car series.







Yahoo! Groups Links







Greg Martin
 

Richard Hendrickson wrote:

"The practice was fairly widespread in the early 20th century and some RRs, notably C&NW and GB&W, continued it through the '50s and into the '60s."

Ricahrd and all,

I am thinking the answer is going to be something very simple and I am awaiting an answer. I posed the question to a friend (and fellow modeler) that works for the Greenbrier Companies (Gunderson's car accounting divison) and he didn't have the immediate answer but he said that some current leasing companies still use the practice today. He said he was going to call "a couple old heads" and get us the answer we seek... I backed my bet by phoning a friend in the BNSF Car application department, but he is out of the office until Friday. He has been with the RR since 1964 in the same capacity...

We shall see...

Greg Martin