SR classes

[JimC]

On 17/06/2022 at 21:42, Compound2632 said:

 

That's interesting. Holcroft's book was published in 1962, a third of a century after the events under discussion, so was written when he was in his seventies, presumably. There are some points where he is perhaps remembering imperfectly:

Holcroft tells us that he kept diaries in his Southern days.

[JimC]

On 19/06/2022 at 16:37, rogerzilla said:

The Swindon system makes little sense either, since they started classes with random numbers (e.g. 4073) and filled in gaps so that the last class (1500) has a really low number, while the class built shortly before it is 9400.

Not the least confusing part of Swindon numbers is that they had at least 4 systems (counting grouping), and the later ones were just superimposed on earlier. So originally locomotives were numbered sequentially,  then under Dean there were ranges attached to types, which covered up to about 4000 with successive lots of a given type allocated sequentially, and then Churchward introduced the second digit scheme, which covered wheel arrangement/type rather than class. ie 4 cylinder 4-6-0 4000, 5000, 6000, 7000. 2 cyl 4-6-0 29, 49, 59, 69 etc (the Hawksworth Counties were originally going to be 99, but it was changed, allegedly because the identity of the new class was leaked to the press). Then at the grouping rather than give a whole new range, which might have made more sense, they used spare numbers under 2000.

In the case of 4073, it seems likely that the Castles were originally considered as an expansion of the Star Class, which finished at 4072. There were a number of other cases where what might be called a sub class started where the previous left off. 

 

The 15s and 16s were using number ranges from the Dean scheme which had been vacated by scrapped pre grouping 0-6-0Ts.
 

The trouble with really rational number schemes is that they tend not to survive 40 or 50 years of new classes unless you start with really large numbers as per TOPS, so either renumberings or confusion, take your pick.

[JimC]

On 20/06/2022 at 04:28, kevinlms said:

Yes I know it represents the first of the Pacific's so a 4-6-2, but why not 2C11, which puts the wheels in order, then class number?

2C11 would be either the 11th 4-6-0 or the first 4-6-2. Having the letter at the end provides a delimiter. 

[Wickham Green too]

11 minutes ago, JimC said:

...... unless you start with really large numbers as per TOPS, ......

Even TOPS has re-issued two class designations that I know of - even though the associated numbers were never carried on the locos first time round.

Edited June 22, 2022 by Wickham Green too

[kevinlms]

1 hour ago, JimC said:

2C11 would be either the 11th 4-6-0 or the first 4-6-2. Having the letter at the end provides a delimiter. 

Shows what a truly awful system it was then! The French (from which Bulleid apparently based it on) didn't have that problem.

At least BR didn't try to use it.

[SRman]

3 hours ago, JimC said:

... snipped ...
 

The trouble with really rational number schemes is that they tend not to survive 40 or 50 years of new classes unless you start with really large numbers as per TOPS, so either renumberings or confusion, take your pick.

Even TOPS didn't get it entirely right in earlier days. I'm thinking of classes 24 and 25 here, where they were numbered sequentially into their TOPS classes without regard to the sub-classes (24/0, 24/1, 25/0, 25/1, 25/2 and 25/3). There were one or two other diesel classes affected similarly before BR got its act together and separated those sub-classes.

The other place where TOPS was initially very messy was the DMU classes, where each vehicle type (powered or trailer) was given a separate class and/or sub-class. Thankfully they tidied that up later, so class 101 motors and trailers were all classed as 101, for example.

[rogerzilla]

The one weirdness of TOPS is that class numbers have to start at 001 rather than 000.  This meant it wasn't a straight conversion from D and E numbers to TOPS. Hence D9000 became 55022 and D8000 became 20050.

[Compound2632]

2 hours ago, rogerzilla said:

The one weirdness of TOPS is that class numbers have to start at 001 rather than 000.  This meant it wasn't a straight conversion from D and E numbers to TOPS. Hence D9000 became 55022 and D8000 became 20050.

 

Not at all weirdness. 87 001 was the first member of Class 87 and the highest 3-digit number tells you how many are in the class, until you start mucking about with re-numbering sub-classes. It's the idea that the first member of the 47xx class is 00 that is weird!

[Wickham Green too]

6 minutes ago, Compound2632 said:

... It's the idea that the first member of the 47xx class is 00 that is weird!

In which case all the B.R. standard steam plus D/E prefix classes, most G.W.R. classes and plenty of others were 'weird' ........ I think the word's a little harsh when that was pretty well the norm.

[Compound2632]

17 minutes ago, Wickham Green too said:

I think the word's a little harsh when that was pretty well the norm.

 

Hey, it was the other fellow who used the word first! My only crime is criticising the Great Western. But I admit the Midland fell into this weirdness trap at the great re-numbering, 4-4-0s starting at 300 and resuming at 700, 4-2-2s at 600, etc. and then the 2000 Class 0-6-4Ts, and the LMS carried on the same. But for some strange reason, the oldest 2-4-0 was not given the number 0, which would have been the logical application of the system.

Edited June 23, 2022 by Compound2632

[Oldddudders]

Well before TOPS appeared, ISTR BR had fallen out of love with X000 numbers. I think E5000 became E5024, and maybe other classes were similarly changed?

[kevinlms]

13 minutes ago, Compound2632 said:

 

Hey, it was the other fellow who used the word first! My only crime is criticising the Great Western. But I admit the Midland fell into this weirdness trap at the great re-numbering, 4-4-0s starting at 300 and resuming at 700, 4-2-2s at 600, etc. and then the 2000 Class 0-6-4Ts, and the LMS carried on the same. But for some strange reason, the oldest 2-4-0 was not given the number 0, which would have been the logical application of the system.

No railway started at 0 to my knowledge, but plenty started numbering at 1.

 

The famous Gresley A4's as built, carried entirely illogical numbers (to me anyway), with 4 blocks of numbers ranging from 2509 - 4903, until 1946/47, when the 34 locos were renumbered 1-34 (but not apparently in the same order!)

BR merely added 60000 to those last LNER, which was entirely logical to me.

 

Fact is block numbering of locos should ALWAYS start at 1, 101, 1001, 5001 etc. That way it doesn't require mental juggling to work out that 5127, isn't the 127th loco in the series, but the 128th! That is assuming that there is no gaps.

[Wickham Green too]

Ah - but gaps will always appear ..... sometimes pretty quickly : D6502 - the third of the class - only lasted about four years, for instance.

 

I'm sure I've come across an industrial loco numbered 0 .... might be in the Bagnall book ................... and, wandering off topic a little, there are at least two stations with platforms 0 to confuse the unwary ( pretty sure nobody's numbered a loco 9¾ anyway ).

 

[Nearholmer]

36 minutes ago, kevinlms said:

No railway started at 0 to my knowledge, but plenty started numbering at 1.

There were a few No.0 industrial locos. If I can find an example easily, I will point to a photo.

 

My favourite numbering scheme is to name the locos Primus, Secundus, Tertius, etc, which again was used by a few industrial concerns. It’s a scheme that would get mighty cumbersome, with five yard long nameplates, on a railway with lots of engines though!

 

 

 

 

Edited June 23, 2022 by Nearholmer

[JimC]

2 hours ago, Compound2632 said:

It's the idea that the first member of the 47xx class is 00 that is weird!

 

GWR number series started with 1 before Churchward's revised scheme, but the downside of that was that the 100th locomotive in the 27xx series of 0-6-0 tanks was 2800, then 2-8-0s started at 2801. Everything's a compromise! Later on 0-6-0T 2800 was renumbered, I forget to what, and the 2-8-0 prototype no 97 renumbered 2800.

[rogerzilla]

The Western Region pulled the old GWR trick of cast numberplates, meaning the Westerns went into TOPS as Class 52 but were never renumbered.  So D1000 stayed, at least in the flesh, as D1000 (or 1000, as some "D"s were painted over to hide them).

Edited June 23, 2022 by rogerzilla

[Artless Bodger]

4 hours ago, Nearholmer said:

There were a few No.0 industrial locos. If I can find an example easily, I will point to a photo.

 

My favourite numbering scheme is to name the locos Primus, Secundus, Tertius, etc, which again was used by a few industrial concerns. It’s a scheme that would get mighty cumbersome, with five yard long nameplates, on a railway with lots of engines though!

 

 

 

 

Number and name in one - saves the amount of metal for plates and sure to find approval by some of our classically educated ruling classes, not sure what the latin for something like the 'one thousand and third' might be?

 

I think it was the Saxon railway adminstration that gave its loco classes Roman numeral designations, at least we haven't had numbering systems using them, nor computer based numbers using binary, though wasn't there a scheme suggested to apply bar code numbers to locos so they could be read by trackside scanners (probably an April Fool spoof iirc)?

[Artless Bodger]

I have not really understood the concept of the duplicate list - renumbering old locos to allow newer ones to fit in number blocks etc I can understand, but given the duplicate list locos were apparently still in use, for service stock work if not revenue service, then how did that fit in with being considered replaced on an accountancy 1 for 1 basis? And weren't some duplicates later renumbered back into stock? Were locos on the duplicate list not repaired or maintained, run until sold or fell apart?

 

 

[Compound2632]

33 minutes ago, Artless Bodger said:

I have not really understood the concept of the duplicate list - renumbering old locos to allow newer ones to fit in number blocks etc I can understand, but given the duplicate list locos were apparently still in use, for service stock work if not revenue service, then how did that fit in with being considered replaced on an accountancy 1 for 1 basis? And weren't some duplicates later renumbered back into stock? Were locos on the duplicate list not repaired or maintained, run until sold or fell apart?

 

On the Midland and the LNWR, duplicate list locomotives were treated for operational purposes exactly the same as any other locomotives. (The same goes for duplicate carriage and wagon stock.) The duplicate list was an accounting device - rolling stock on the duplicate list had ceased to be capital assets, having been renewed, but remained in service. How that worked from an accounting point of view I don't understand, but evidently it did! They weren't included in the returns of working stock made to the shareholders' half-yearly meetings and to the Board of Trade, until there was a change in the statutory requirements with the Railway Companies (Accounts and Returns) Act 1911; so for the year ending 31 December 1913 onwards, all rolling stock was reported. But they were not regarded as having zero market value. There are a couple of surviving valuations of carriage and wagon stock for the Midland Railway, from 1904 and 1905, which give the market value for the ordinary stock and the duplicate stock; I believe these were prepared in support of legal disputes over the company's rateable value. 

 

In the Midland's great renumbering of 1907, locomotives in the duplicate stock were simply numbered in the same series with all the rest, without, so far as I am aware, ceasing to be duplicate stock. I think this may indicate that there was a shift from considering locomotive numbering as an accounting convenience to considering it an operational convenience. 

 

But we really need an historian of 19th century business accounting to explain this to us.

[JimC]

I suspect to really understand 19thC accounting practices like the duplicate list you really need to be a 19thC accountant. The way I saw it explained was that it rapidly became obvious that, what with lead times, economies of scale etc, that it was quite impossible to schedule things so the replacement locomotive arrived the day its predecessor fell off its perch. In addition to that it was obviously vital  that the replacements always arrived before the predecessors died, else nothing to run the services.  So there was often life left in the old engines when the new ones arrived, and thus the duplicate list, because using up the remaining mileage until the old locomotives were ready to scrap was an obvious cost saving. One presumes that the duplicate list would indeed be run until they dropped, maybe allowing minor repairs, maybe not I suppose.

 

All well and good, but things were often not even as simple as that. Supposing the line decides halfway through the renewal programme that actually they want to expand the fleet and re-designate some of the replacements as extra locomotives. Then in order to keep things tidy the ones with most life that were scheduled to be replaced come off the duplicate list and go back to the main list for a period of time until another replacement arrives.

 

Presumably in the end, though, it was decided that the renumbering and so on was a waste of time. No-one on the running side needed to know the capital status of the locomotives, so it could be dealt with by an entry in a ledger rather than a number on the side of the locomotive.

[Nearholmer]

Once a thing was “written off”, so no longer on the capital account, it’s value (however residual and small) didn’t have to be totted-up, and I wonder whether there were tax advantages in keeping the value of the capital assets as low as possible, although I doubt it because barely anything was taxed in C19th compared with now.
 

Might it have mattered in terms of dividends paid? But then, I think they were paid against subscribed capital, rather than currently valued capital, so probably not.

 

Was it advantageous in the operation of the company to take things off the capital account, thereby allowing further spending of capital without exceeding the subscribed amount? If so, how the heck did the affect of inflation get factored-in?

 

We need a really impenetrable article in ‘Backtrack’ to explain all this!

[Compound2632]

46 minutes ago, JimC said:

it was quite impossible to schedule things so the replacement locomotive arrived the day its predecessor fell off its perch. In addition to that it was obviously vital  that the replacements always arrived before the predecessors died, else nothing to run the services. 

 

Not locomotives in this instance, but the same procedure for carriages and wagons:

 

MR Carriage and Wagon Committee minute No. 3299, 16 September 1897

 

Renewal of Stock

               Read Board Minute No. 6859, as follows:-

"Resolved

               That the Carriage and Wagon Superintendent lay before his Committee at the March meeting in each year a statement of each description of vehicles falling in for renewal in the year following (i.e. from 1st January to 31st December) showing the:

               Age of the stock,

               Probable number which will be entirely broken up,

               Probable number which may remain for some time as duplicate stock.

               That such statement be at once sent by the Carriage and Wagon Committee to the Traffic Committee to approve the type of the vehicles to be constructed to replace those falling in for renewal.

               That as soon as the Carriage and Wagon Committee have received from the Traffic Committee its approval, it shall authorise the Superintendent to at once commence with the work.

               Renewals should as far as possible be in advance and no stock taken out of traffic until it can be replaced by renewed stock. This plan strictly adhered to will save the Traffic Department much inconvenience and expense of hauling empty vehicles about the line and often loss of traffic from shortness of stock."

               Mr Clayton pointed out certain difficulties that would be in the way of carrying out the instructions contained in the above minute, and he was requested to make a report on the subject.

 

[TNA RAIL 491/257]

 

Subsequent discussion in the minutes is concerned with the carriage stock; it's not clear that the business about liaising with the Traffic Committee was carried through for the wagon stock. in any case for some years previously the number of wagons broken up and wagons built as renewals hadn't balanced over any short time period (month or half-year).

[kevinlms]

2 hours ago, Artless Bodger said:

Number and name in one - saves the amount of metal for plates and sure to find approval by some of our classically educated ruling classes, not sure what the latin for something like the 'one thousand and third' might be?

 

I think it was the Saxon railway adminstration that gave its loco classes Roman numeral designations, at least we haven't had numbering systems using them, nor computer based numbers using binary, though wasn't there a scheme suggested to apply bar code numbers to locos so they could be read by trackside scanners (probably an April Fool spoof iirc)?

The bar codes were a real thing in the US and indeed one of several pioneers. Poor maintenance (graffiti?), brought the system undone.

 

https://en.wikipedia.org/wiki/KarTrak

[pH]

7 hours ago, kevinlms said:

Fact is block numbering of locos should ALWAYS start at 1, 101, 1001, 5001 etc. That way it doesn't require mental juggling to work out that 5127, isn't the 127th loco in the series, but the 128th! That is assuming that there is no gaps.

From an operating point of view, it doesn’t matter where in a series a loco is - 1st, 50th, 127th, 842nd … The number is just a label to identify the particular loco. A crew will be allocated loco “5127” for their work, not the “127th loco in the 5 class”. A ‘label’ series of “5AAA”, “5AAB”, “5AAC” etc. would work just as well for identifying individual locos.

 

[Compound2632]

Just now, pH said:

A ‘label’ series of “5AAA”, “5AAB”, “5AAC” etc. would work just as well for identifying individual locos.

 

Or even some more memorable names!