Baldwin had two sets of numbers while building a locomotive. Baldwin built the locomotive by its class number based on its wheel arrangement and its cylinder size, and assigned its builders plate number at then end of the construction when they were actually bolted to the locomotive.

It is a known fact that Baldwin builders plate numbers are often out of order compared to the class numbers, and also as to the final assigned road number for a locomotive.

We all use the railroads road number to distinguish between locomotives, but my question is, when does that road number take effect and it become that number?

I have been doing some research into the Pershing 2-8-0's built by Baldwin for the army during WW1. The army originally assigned them their road numbers sequentially startiong with 1. However, along the way, some of the locomotives were diverted for other service and their numbers changed. I had always thought that this probably happened after they were test fired and delivered, but the Baldwin records are very clear that these engines, although built as USA XXX were renumbered to USA YYY BEFORE test firing and boiler certificates were issued, and in some cases a later locomotive was built as USA XXX, so that everything was all mixed up.

Thus my question. If the engine was ordered as XXX but inspected, test fired and a boiler certificate issued as road number YYY, is it really YYY or is it XXX renumbered before delivery to YYY? And then if there is another engine ordered as XXX is it first XXX or second XXX?

This has actual impact in today's real world so I would like to hear from those of you with an opinion (and I expect several). I have mine also, but I still want to hear others chime in on this.

Some people don't like it but no system is perfect. For my two cents you get if the loco was painted as XXX then repainted as YYY before it left the plant this should be mentioned in a foot note. If it was a paperwork change before the number was ever put on the loco forget about it.

Here are a couple of other examples of number/owner confusion. A Portland builder of logging machinery assigned a number when an order was placed. If the order was canceled they would go back and reuse the number up to a year later. American Hoist & Derrick built a batch of cranes in the early 1950s. They were ordered by the Navy but all left the plant lettered for the Army. The AH&D builder's list shows all of them as US Navy.

A few things to remember;

A railroad assigned the locomotive numbers. For example, when UP ordered the 800 class 4-8-4 they told ALCO that the road numbers would be 800-814. Once UP placed an order, ALCO would assign an order number. Once the order number was assigned and after drawings were complete and agreed on, ALCO assigned the locomotive beds (frames), boilers and tenders to respective departments. The first bed out (frame) was assigned the first number in the order and was also assigned the first road number. The first boiler out, generally (but not always) went to the first bed and so on and so on. However, the first locomotive of a series, out of the shop, may not have been the first one assigned. For example Number 801 may have left the shop before 800. It really is not that complicated and it had nothing to do with the order in which they were finished and painted. It has more to do with assigning numbers to major components at time of completion based on an order number and quantity.

Once a locomotive was finished, UP received a file that bed number &&&&& with boiler !!!!! And tender @@@@@, all of which make up delivered unit number ?????? Known to UP as FEF-1 No. 810. All drawings and notes were contained in each numbered package.

Best I could do on an IPad.

I have another 2 cents. I feel as if they left the factory in one scheme their origin number would be XXX. Each originally left the factory in a certain number and scheme, but were changed when a railroad eventually received them. For example, The EMD 103 certainly didn't stay the 103 its entire life. It became a CNO&TP engine numbered 6100A. Here's the link about its little known service life, and coincidentally the actual scheme for the engine in service became the footnote http://www.oil-electric.com/2011/03/ft- ... t.html?m=1

My impression from my research on WWII BLW locomotives is that the builders numbers at BLW were of little significance to the factory. I'm basing this only on the USATC S160 Consolidations, particularly the ones built by BLW during WWII. This group of 2210 locomotives were built by ALCO, Lima, and Baldwin. The Army assigned road numbers in blocks to each manufacture. I believe I read that some of the USATC road numbers were recycled from the unbuilt locomotives from a cancelled order of USATC 2-8-2s.

The card index for our S160 covers an order of 180 locomotives. 120 were standard gauge and another 60 were 5" 6" gauge for India. The locomotives are identified by their class and serial number within the class. For ours, that's 280 19S 809 (wheel configuration, steam, 19" cylinders and serial number). The index meticulously lists every part by drawing number, quantity used, weight, and material. When different locomotives within the order are configured differently or when there was a design change made in mid-order, the index listed which drawings apply to which locomotives. For example, all the parts for sixty locomotives numbered 719 to 778 (the Indian locomotives) and related to the gauge are different than they are for the standard gauge locomotives.

Many of the parts such as valve gear and rods for our locomotive are marked "280 19S 809." Oddly, the number stamped into the cab was for a different locomotive number within the same order. After moving layers of grease and paint, we also found that the cylinder block had a serial number of a different locomotive in white paint. Neither of the mismatched parts matched the serial number for another Alaska Railroad S160, so the swap was apparently made at Baldwin at the time of manufacture. I have not found any use to the road number or builders number in the BLW card index.

Specifications for four earlier Alaska Railroad locomotives of this design were prepared by BLW for the ARR and ran about 12 pages long. They only identified the locomotives by class and serial number, not by road number or builder's number. However, the cover letter listed both BLW serial number and USATC road number. (The builder's photo attached to the specifications is of USATC #1702, now Great Smokey Mountains RR #1702, one of the first S160 built by Baldwin.)

A Lima card index for the same locomotive design listed road numbers, but only on the cover. Lima kept track of design changes with a record of change sheet showing the date a drawing was replaced and in the index they did not track changes in design by serial number. Drawings for different configurations were identified by the configuration, i.e., oil fired, coal fired, or screw reverse.

I can see why the builder wouldn't care what the road number was. That can be changed in a short while. It sounds like Baldwin didn't give a damn about keeping the parts sorted by individual locomotives since they were building a huge number of identical machines. Getting them out the door as fast as possible was the goal. I remembor someone in England getting one of the Baldwin WWII 2-8-0s from Poland. It had been through the shop several times and had Baldwin, Lima and Alco parts.

The builders did not give a scrap about the road numbers. Road numbers are only a means for the operating road to keep track of their own units/classes of locomotives.

Our culture of railfans tend to think that road numbers are everything, they really are not. It is noted that the FRA requires changes of road numbers to be filed with the FRA, but beyond that....

The USRA units are a little different and should be looked at differently. In the case of USRA, they issued the order numbers and specifications to the builders and indeed, the name of the game was units "out the door", no matter who built them, how or in what order. In this case, it is true, the builders did not care too much about order numbers and what part went with what unit. Each part had a part number Identifier, not so much an individual part number. There is a difference. For example, a main rod could be part 500. It could also say 500-1 (meaning it is a main rod 500 and that it is the first (1) built, meaning it is 500-1)...but who cares what unit it goes with.

In the case of the USRA units, they may have been painted/numbered "out the door".....

I thought Baldwin stamped an "E" number on a lot of parts to represent the order number. 1522 had some parts stamped with the E number and the builder's number was stamped on the steam dome and cast into the builder's plates. 59134 in 1522's case.
I don't know how the other builders did it.

Somebody has to choose a number that represents the asset, and it's hardly ever the road number. This is for ownership and tax purposes as much as anything.

During the 1860s neither C.P. Huntington (who was dealing with the builders) or the various builders became interested in the road numbers/names until it was time to paint the locomotives. We find letters from the builders to Huntington asking what number he wanted on them. In other words, the CP locomotive were generally numbered in the order in which the builders asked for the number.

Huntington kept track of the numbers. But then Crocker bought four engines locally in California and assigned the next available numbers, not knowing that Huntington had already assigned those numbers. When eight locomotives Crocker did not know were coming showed up, they had a real mess to straighten out. Huntington tried to correct the problem by skipping ahead in his numbering. But, he skipped too far, leaving two numbers blank. These were finally filled in a year later. But, it was exactly one year later, so some roster historians have assumed the invoice dates were typos, not recognizing that there really were two 1868 locos numbered in with a bunch of 1867 locos. They had telegraph, but for the most part dealt with these issues by mail--which took a month each way.

At one point Crocker suggested dropping the names, which were being assigned to the locomotives along with numbers. Coming up with names was proving to be a problem. But then nine locomotives were landed at Sacramento within a week. These all came as kits, with pieces crated. The crates had been shuffled on the dock in New York and in loading, and were shuffled again in unloading at San Francisco and in shipment to Sacramento, and no one knew which parts went with which locomotives. The crates were apparently marked with both part and locomotive numbers, so they really didn't help much. Thenceforth all locomotives had their names painted on each crate and part, and the company continued to use names for this reason until the end of shipment by sea in 1869.

When the CP acquired the Western Pacific locomotives, they assigned them letters to distinguish them as locomotives of both companies were operating on the same roads.

One other odd situation on the CP was the use of the term “rebuilt” in various company published rosters. As it turns out, these “rebuilt” locomotives had nothing but the number of the earlier locomotive of the same number. In the case of the “rebuilt” CP number 3 (of 1872), not only did it have a different builder (Rogers) than the original no. 3 (Danforth), the original locomotive still existed (as Southern Pacific no. 1) and still exists today, long after the “rebuilt” no. 3 was itself scrapped. “Rebuilt” was apparently an accounting device to charge the new locomotive to operating rather than capital expense.

DJS

Ours is a Baldwin locomotive but has ALCO cast into the driver journal boxes. It probably came from the factory that way. I've read that an S160 in the UK has the left frame from one manufacturer and right frame from another.

The S160s were shipped to Europe with a complement of spare parts, X parts were shipped for every XX locomotives. I have a copy of the allowed spares list somewhere in my files. Spares included cylinder block halves and rods. Both are interchangeable from side to side.

I forgot to mention that the classification scheme that gives 280 19S was implemented in about 1940. The earlier BLW scheme was more cryptic.