Pre-grouping loco numbering

[Taunton]

It is apparent looking at the numbering of pre-grouping locos that there was often no concept of numbering a class consecutively. Many railways had started off incrementing numbers as they built more early locos, but as these were progressively and randomly withdrawn, new locos were given numbers in the old series, commonly the lowest number available. After a few generations of this, the numbers would be a real scatter.

Also common was after 20 or 30 years, locos would be renumbered, sometimes by just putting an A suffix to the number, and a new one built would be given the number just vacated. Sometimes called the "Duplicate List". This could reach an extent that such a new loco, after its own 30 years, would be renumbered with an A suffix, and the previous A would still be around and have its suffix changed again to B. There were also renumberings within the sequence for no apparent reason. Large and small, new and old locos would have adjacent numbers.

I have seen described that it was an accounting feature, new locos charged to "revenue" would be given numbers within the existing sequence, while mechanically-identical new locos of a class charged to "capital" would be given numbers at the end. While the accountant is quite at liberty to do this charging, why it impacts on the actual running number affixed to the loco is not apparent. It is almost like there was a desperation not to "waste numbers".

The GWR seems to have been the first to make a break from this and ensure new locos were numbered in blocks, but others carried on as before.

Anyone know why it was done like that?

 

[30907]

Could you charge a new loco to revenue? I've read of rebuilds being deemed revenue expenditure, as if it was a repair - was it some of the less successful LSW 4-6-0s that were "rebuilt" even though only the driving wheel centres were original (think that's not to be taken literally).
Duplicate list locos presumably had no book value and therefore were a bonus to the operators?
Random numbering seems to have been phased out by a number of companies pre grouping, maybe because locos were being built in bigger batches, but systematic renumbering a la GW/LNE was less common.

 

[krus_aragon]

I suppose much of this comes down to organisation size, inertia, and necessity.

The smallest of railways might have little use for numbers on their locomotives, and just refer to them by name. If your workshop and staff know every locomotive they work with, a numbering system might have little practical use (other than giving the appearance of being a larger, more important concern). Even as you get larger, if you have few types of locomotive then the differences between them may be self-evident.

As a case in point, the LNWR's Northern Division at Crewe. For over a decade between 1845 and 1858 they built roughly four hundred locomotives. The 2-2-2 passenger locomotives came in 6ft or 7ft driving wheel variations, whereas the 2-4-0 goods locomotives were all much the same, the only key difference being whether they used an indirect or direct variation of the Stephenson valve gear. O S Nock's listings of these locomotives by build year illustrates their practice of using the lowest number available: It's clear in the early days when they were replacing a withdrawn locomotive and when they were expanding the fleet.

2-2-2 Singles
1845: 49 Columbine, 19 Princess, 73 Prince, 74 Deva, 75 Apollo, 76 Albion, 77 Mersey
1846: 79 Belted Will, 80 Dalemain, 84 Saddleback, 85 Ingestre, 86 Sandon, 4 Hecate, 14 Witch, 16 Lynx, 25 Harpy, 29 Alaric, 33 Erebus, 44 Harlequin, 54 Medusa, 91 Knowsley, 92 Trentham, 93 Premier, 94 Helvellyn, 95 Hydra, 96 Polyphemus
...

2-4-0 Goods
1845: 78 Lonsdale, 81 Greystocke, 83 Windermere, 87 Eden, 82 Wordsworth
1846: 88 Lune, 89 Bela, 90 Lowther
...

Then we get to the issue of corporate inertia. The LNWR stuck to this numbering approach throughout their history: Nock states that only the Jubilee and Alfred the Great classes happened to end up with consecutive numbers. As a company the LNWR had a reputation for conservatism and frugality: this was the company that felt 40mph was a prefectly reasonable speed for an express train, and prioritised economy and shareholders' dividends over speed. Why, then, should they contemplate spending money on fitting new numberplates to their locomotives, if there was little obvious gain?

Looking at a glance over at other railways, I note that the GWR implemented a numbering system in 1875, and then replacend it with a new system in 1912. The Midland introduced their system in 1907, and I don't see any evidence that any of the LNER's constituents had a system. So I conclude that the idea of systematic numbering and the benefits thereof only became really self-evident around the turn of the century. (Post-grouping, with many disparate sets of identically numbered locomotives, there was an evident need to do something.)

With regard to the GWR's 1875 renumbering, I think they had an opportunity that few other large railways had. After losing the Gauge Wars, the 1870s saw the GWR converting much of their railway to mixed or narrow gauge. As part of this, they'd be withdrawing broad gauge locomotives and building new narrow gauge locomotives to replace them. (Apparently the broad gauge locomotives were never numbered, but narrow gauge ones had been.) Given that they were overhauling their locomotive base, it was the ideal opportunity to decide to impose a numbering system. The other large companies didn't have the same sort of event, so there was never a natural time to consider renumbering.

 

[etr221]

All very complicated, and dependent on wide variety of factors - including the relative influence of the accountants, operators and engineers in such matters; and the value placed on a 'logical' (re-)numbering versus the confusion of changing numbers (is loco X the same as it was last year, or was it Y?). And then different railways, different managers, different priorities...

Even in BR days, there some discontinuous numbering within classes - thinking in particular of the D8xx Warships, amongst the modernisation plan diesels. It was only with TOPS in the 1970s that a comprehensive numbering system including a separate class number was introduced in Britain. In my understanding, the first railway to use it was the kkStB (Austria), who introduced it in 1904.

Could you charge a new loco to revenue? I've read of rebuilds being deemed revenue expenditure, as if it was a repair - was it some of the less successful LSW 4-6-0s that were "rebuilt" even though only the driving wheel centres were original (think that's not to be taken literally).

What do you mean by a 'new' loco? When a railway was built, it would have some new locos - which would be charged to capital. At some later date, it would build some new locos for a new line, or other capital enhancement, and these would also be charged to capital, as part of the general increase in capital value. But at the same time - and as part of the same batch (order) - there would be more locos built to replace some (or all) of the original locos which had reached the end of their nominal lives - my thought is that these would be paid for from revenue (funds built up for the purpose), and thus charged thereto. But would be part of the capital stock, and so - on many railways - would have been given the numbers of the old (replaced) locos. And considered 'replacement' rather than 'new' - at least from the accountants point of view. But the ideosyncracies of pre-nationalistion railway accounting are beyond me.

Duplicate list locos presumably had no book value and therefore were a bonus to the operators?

The replaced locos would either have been scrapped, or - if still of use - retained on the duplicate list. So, yes, those locos placed on the duplicate list would have no (or minimal, i.e. scrap) value, as a 'bonus' for the railway.

 

[randyrippley]

Don't forget you have an interesting philosophical question as to "what is a new locomotive?"
If you have a new class of say 100 locos then yes, they're all new and part of that capital budget. But lets assume that at the same time there's an overprovision for servicing spares. Spare boilers / frames / wheelsets. Maybe 10% (for arguments sake). What that really means is you have the parts for 110 locos, but the costs are only attributed to 100. If business increases, some of the parts pool could be used to assemble more "free" locos. But you can't assign them to capital as thats already been done on the original build. And the parts used could have been recycled from the parts pool, so the "new" locos aren't new even though the number has increased. Meaning you can't even depreciate then properly. Far easier to pay for then through revenue and account for them separately.
I have a feeling that under some circumstances the auditors would take a dim view if you tried to effectively create new capital assets for free