Mileage datum points

[rodent279]

Time for another of my "piffling trivia that no one cares about" questions.

 

Just out of interest, when a new railway was built, that diverged from an existing railway, where would be the datum point for milestones? From the junction, I guess, as in the case below, where a milestone is shown as "1 mile from Mangotsfield Junction", but where exactly at Mangotsfield Junction? The actual point of divergence? In which case, is it the switch blades of the points, or the frogs/crossings? Or would it be some other point, like the signal box?

Or was no one really that bothered, and a rough approximation made, from some arbitrary datum?

 

https://maps.nls.uk/geo/explore/side-by-side/#zoom=18&lat=51.46731&lon=-2.48181&layers=178&right=BingHyb

 

(In the map above, it can only be from Mangotsfield North Junction, as South Junction is too close).

[Trog]

I suspect that different railways and engineers did it differently. Fronts joints, toes or where the branch plain line starts all have there attractions. It does not really matter as a couple of redesigns of the junction later and the branch will end up starting a little above or below zero. Or you could just split the mileage of the existing line and use it on both the main and branch, so both share a common zero point.

[Compound2632]

Mangotsfield North Junction is the point of junction with the main Birmingham-Bristol main line, so it's natural that that should be the zero for the Bath branch mileage. In instances where the branch mileage was measured independently of the main line mileage, there would be a zero milepost marking the point of junction - so the position of that post would be, by definition, the point of junction. This was evidently the case at the 1880s date of the map you link to. Around the turn of the century, the Midland carried out a great re-measuring of its network from a zero point at St Pancras Passenger Station, resulting in replacement of mileposts on many lines; along with this, a set of rather beautiful maps, the "Distance Diagrams". The eighth edition of sheet 53 shows Mangotsfield North Junction at 214 mi 14 ch and Mangotsfield North Signal Box at 214 mi 13 ch by the shortest route from St Pancras but also shows that the actual milepost mileage was measured from the zero point of the "West Road", Derby London Road Junction - 124 mi 46 ch for the junction and 124 mi 45 ch for the box.

 

There's a nice essay on Midland Railway mileposts by John Gough here; a discussion in the context of the Settle & Carlisle line here; and a more general SRS article here. From reading these, I think that the location of the zero post would be determined by the point of junction laid down in the Act of Parliament for the new line: "A railway commencing at a junction with the Great North & South Railway's Ayebury to Beaton line 2 miles and 17 chains north of Ayebury station" or some such.

[TheSignalEngineer]

On 16/06/2021 at 18:29, Trog said:

you could just split the mileage of the existing line and use it on both the main and branch, so both share a common zero point.

That was done in some places, although not consistently on the same company. 

The Midland mileages to Birmingham come from Derby and continue from the same datum between Kingsbury and Water Orton via the old line through Whitacre or new (1908) direct line with a mismatch at Water Orton. The New Line mileage of 33m22c at Water Orton East jumps to 34m43c as measured via Whitacre.  

Landor Street to Kings Norton and beyond to Yate is even more complex. The old line is easy, continuing from Landor Street via Camp Hill to Kings Norton and Barnt Green. Via New street the  mileage goes as far as 41m51c at Proof House Junction then changes to the original London and Birmingham Railway mileage of 112m51c to the centre of New Street Station at 112m73c. There it changes to the Stour Valley Railway mileage of 0m05c. Why there and not the original Stour Valley datum I have no idea. Outside New Street PSB at 0m16c Stour Valley it resumes the Midland mileage at 42m35c onto the West Suburban Railway. At Kings Norton 48m02c via New Street reverts to 46m77c via Camp Hill.

 

Midland practice on the Derby - Bristol line at least seems to be that if you diverge when running from Derby the mileage continues and if a line converges a new mileage starts back from the junction e.g the Nuneaton line starts from zero at Whitacre but the Walsall line runs continuously with the Derby mileage from Water Orton West Junction.

[rodent279]

24 minutes ago, TheSignalEngineer said:

That was done in some places, although not consistently on the same company. 

The Midland mileages to Birmingham come from Derby and continue from the same datum between Kingsbury and Water Orton via the old line through Whitacre or new (1908) direct line with a mismatch at Water Orton. The New Line mileage of 33m22c at Water Orton East jumps to 34m43c as measured via Whitacre.  

Landor Street to Kings Norton and beyond to Yate is even more complex. The old line is easy, continuing from Landor Street via Camp Hill to Kings Norton and Barnt Green. Via New street the  mileage goes as far as 41m51c at Proof House Junction then changes to the original London and Birmingham Railway mileage of 112m51c to the centre of New Street Station at 112m73c. There it changes to the Stour Valley Railway mileage of 0m05c. Why there and not the original Stour Valley datum I have no idea. Outside New Street PSB at 0m16c Stour Valley it resumes the Midland mileage at 42m35c onto the West Suburban Railway. At Kings Norton 48m02c via New Street reverts to 46m77c via Camp Hill.

 

Midland practice on the Derby - Bristol line at least seems to be that if you diverge when running from Derby the mileage continues and if a line converges a new mileage starts back from the junction e.g the Nuneaton line starts from zero at Whitacre but the Walsall line runs continuously with the Derby mileage from Water Orton West Junction.

So, theoretically at least, the divergence from the Birmingham-Bristol line at Mangotsfield should carry on counting from Derby, via Camp Hill. And in fact, if the surviving mileposts near Bitton station are to be believed, they do, the nearest to Bitton station is something like 128.25, I'll have to check. Maybe the map referenced is showing the mileposts as originally planned?

[Compound2632]

7 hours ago, rodent279 said:

So, theoretically at least, the divergence from the Birmingham-Bristol line at Mangotsfield should carry on counting from Derby, via Camp Hill. And in fact, if the surviving mileposts near Bitton station are to be believed, they do, the nearest to Bitton station is something like 128.25, I'll have to check. Maybe the map referenced is showing the mileposts as originally planned?

 

The existing mileposts give the mileage from Derby London Road Junction in fact, not just in theory - as @TheSignalEngineer describes - having been installed after the 1899/1900 re-measurement (hence following the old line through Whitacre, the cut-off being of later date). Your map extract, being of 1880s date, shows the original actual mileposts in place before the remeasurement - the OS surveyors would only record what was actually there, not what might merely be planned. 

 

According to the Distance Diagram, Bitton Station is at 128 mi 34 ch by this milepost mileage so the 128¼ milepost should be just 14 ch (128 m) north of the station datum (the centre-line of the platforms?) or perhaps more usefully, 5 ch (100 m) north of the site of the signalbox in 1918.

 

7 hours ago, TheSignalEngineer said:

Landor Street to Kings Norton and beyond to Yate is even more complex. The old line is easy, continuing from Landor Street via Camp Hill to Kings Norton and Barnt Green. Via New street the  mileage goes as far as 41m51c at Proof House Junction then changes to the original London and Birmingham Railway mileage of 112m51c to the centre of New Street Station at 112m73c. There it changes to the Stour Valley Railway mileage of 0m05c. Why there and not the original Stour Valley datum I have no idea. Outside New Street PSB at 0m16c Stour Valley it resumes the Midland mileage at 42m35c onto the West Suburban Railway. At Kings Norton 48m02c via New Street reverts to 46m77c via Camp Hill.

 

The Distance Diagram shows the milepost mileage as continuous through New Street station and onto the West Suburban as far as Lifford and Kings Norton Junctions. Between Derby Junction and the West Suburban Junction (now marked by the PSB) was LNWR territory, so naturally their mileposts apply on the ground. (Was there some subsequent adjustment to bring the Midland mileage up to Proof House Junction?) 

 

I wonder if the centre line of New Street Station shifted 5 chains to the west when the station was enlarged in 1885? That could account for the Stour Valley oddity. The mileage from London Euston presumably uses the original London & Birmingham zero but isn't London & Birmingham mileage, New Street not opening until nine years after that company ceased to exist, the terminus until then being Curzon Street - but you knew that. The LNWR seems to have been altogether less fastidious about consolidating its mileage measurements.

Edited June 20, 2021 by Compound2632

[rodent279]

1 hour ago, Compound2632 said:

According to the Distance Diagram, Bitton Station is at 128 mi 34 ch by this milepost mileage so the 128¼ milepost should be just 14 ch (128 m) north of the station datum (the centre-line of the platforms?) or perhaps more usefully, 5 ch (100 m) north of the site of the signalbox in 1918.

 

That figures. Can't say I've actually seen the 128 ¼ milepost at Bitton, but it's next door neighbour further down the line is still in place, just south of the station, though the former Up line is now a cycle path at this point.

 

[Michael Hodgson]

I remember reading that one of the big companies moved all its mileposts because they had measured from the London terminus and the buffer stops there had been moved! 

 

I don't think the OP's suggestion that perhaps nobody was really bothered is really fair.  The mileage was extremely important for billing purposes on freight and for calculating fares in the days of Parliamentary trains (1d per mile), so it affected the company's profits.  A signal box could not have been used as a datum in the early days as the railways didn't have signal boxes everywhere until block working became compulsory under the Regulation of Railways Act 1889.  And as Compound2632 says, mileages were specified in the original enabling Acts under which the railways were built.

[TheSignalEngineer]

2 hours ago, Compound2632 said:

The Distance Diagram shows the milepost mileage as continuous through New Street station and onto the West Suburban as far as Lifford and Kings Norton Junctions. Between Derby Junction and the West Suburban Junction (now marked by the PSB) was LNWR territory, so naturally their mileposts apply on the ground. (Was there some subsequent adjustment to bring the Midland mileage up to Proof House Junction?) 

The 1913? Distance Diagram shows the junction with the LNWR at Grand Junction being 41m22c from London Road Junction. The thick line is then dotted and the lines shown thin with L&NW against them up to the West Suburban junction at 42m38c from London Road Junction. 

 

Grand Junction replaced Derby Junction and Gloucester Junction ( at the east end of Exchange Sidings) when the 'Blue Hole' was built taking the Midland under the LNW at Landor Street. This remodelling came about due to the congestion as traffic increased culminating with the Derby Junction Collision in May 1892.

https://newspapers.library.wales/view/3231063/3231066/46/rev george evans babell pensarn carmarthen

The use of Midland mileage up to Proof House dates from the remodelling which took place under WCRM in 2000.

[Edwin_m]

Nothing I've seen contradicts the Midland mileages following the principles set out in the link above, once later changes such as the Whitacre cut-off are taken into account.  This incidentially disproves the enthusiast myth that all Midland lines were miled from (or Up towards) Derby.  

 

In more recent years there has been a great reluctance to re-number or re-position mileposts, because so many operating and engineering documents rely on them.  It's better to have a system that's confusing but consistent, chopping and changing mileages to reflect later closures and the occasional opening, than to have someone relying on an out-of-date document or knowledge with possibly disastrous consequences.  

[Compound2632]

I do have a question about Midland mileposts - and I think mileposts in general: when were they painted yellow? The Midland colour scheme was white with black numerals and edging but I have no information on later painting schemes.

[Michael Hodgson]

Useful reference site ....

 

http://www.railsigns.uk/info/milepost1/milepost1.html

http://www.railsigns.uk/sect28page1/sect28page1.html

[TheSignalEngineer]

Yellow painting only came in during the BR period. When exactly I can't remember but they were still white when I joined in 1966.

[Compound2632]

It's confusing to the uninitiated that heading in the down direction, the milepost numbers go up, while going up, they go down!

[Trog]

1 hour ago, Compound2632 said:

It's confusing to the uninitiated that heading in the down direction, the milepost numbers go up, while going up, they go down!

Although on the Great Central up mileage and up direction are the same.

 

Also as I think that as the mileposts are specified in the Acts of Parliament authorising the construction of each line, you strictly need parliamentary permission to remove or alter them.  

Edited June 20, 2021 by Trog

[Compound2632]

1 hour ago, Trog said:

Although on the Great Central up mileage and up direction are the same.

 

Presumably the Great Central main line zero was Manchester London Road.

 

1 hour ago, Trog said:

Also as I think that as the mileposts are specified in the Acts of Parliament authorising the construction of each line, you strictly need parliamentary permission to remove or alter them.  

 

I don't think that can be right; I'm sure the Midland's remeasurement and re-mileposting didn't have an Act of Parliament.

Edited June 20, 2021 by Compound2632

[PenrithBeacon]

30 minutes ago, Compound2632 said:

 

Presumably the Great Central main line zero was Manchester London Road.

...

Yes and the London extension lines still carry those numbers counted from London Road even though there isn't a direct connection any more.

[Edwin_m]

 

7 hours ago, TheSignalEngineer said:

Yellow painting only came in during the BR period. When exactly I can't remember but they were still white when I joined in 1966.

One of the links above suggests early 60s but from childhood recollections I would say about 1980.  They are certainly a lot more conspicuous painted yellow.  

5 hours ago, Trog said:

Although on the Great Central up mileage and up direction are the same.

 

Also as I think that as the mileposts are specified in the Acts of Parliament authorising the construction of each line, you strictly need parliamentary permission to remove or alter them.  

The relevant text is also quote in a linked article, suggesting that posts are required every quarter of a mile but there's nothing there that prevents them being changed.  Might cause a problem in future if they decide to replace them with metric measurements for ETCS (which works in metric so has speed restrictions in km/h) or any other reason.  

[PenrithBeacon]

The WCML was made up of a number of independent companies until the formation of the LNWR in 1846 and the railway still carries their milepost numbering system to this day. An exception, there might be others, appears to be the M&B as the LNW placed a milepost numbered 189 miles or so in London Road Station. This might have been done because the LNWR didn't want to count down from Crewe to Manchester, the M&B started at Manchester and counted up to Crewe. It might still be there, perhaps Railtrack have left it alone. 

The odd thing of course, is that that on the other side of Piccadilly, the former MS&L (GCR) lines start at 0 from the buffer stops on the northern side of the station and get well up into the 180s (IIRC) at Marylebone.

[HGR]

1 hour ago, Edwin_m said:

 

One of the links above suggests early 60s but from childhood recollections I would say about 1980.  They are certainly a lot more conspicuous painted yellow.  

The relevant text is also quote in a linked article, suggesting that posts are required every quarter of a mile but there's nothing there that prevents them being changed.  Might cause a problem in future if they decide to replace them with metric measurements for ETCS (which works in metric so has speed restrictions in km/h) or any other reason.  

The ECML gained a fresh set of yellow mileposts (the whole numbers initially) in 1979. The numbering remained unchanged as was prior to that date, the only difference was the yellow ones were now provided on both sides of the line. The earlier white MPs remained in place but the former NER ones (York northwards to the border) were only on the DOWN side of the line, as was normal NER practice. The new yellow ones were variously up to a few feet away from the originals - not sure if this simply reflected more accurate means of measuring their true position.

 

One thing that was evident for example between York and Darlington was that there was an offset of up to two chains between the position of some locations along the route as quoted in the Sectional Appendix and the measurements taken from the nearest quarter milepost. In most cases the S.A. was +2 ch. beyond the MP positions, but not consistently in every case.

 

[Edwin_m]

1 hour ago, PenrithBeacon said:

The WCML was made up of a number of independent companies until the formation of the LNWR in 1846 and the railway still carries their milepost numbering system to this day. An exception, there might be others, appears to be the M&B as the LNW placed a milepost numbered 189 miles or so in London Road Station. This might have been done because the LNWR didn't want to count down from Crewe to Manchester, the M&B started at Manchester and counted up to Crewe. It might still be there, perhaps Railtrack have left it alone. 

The odd thing of course, is that that on the other side of Piccadilly, the former MS&L (GCR) lines start at 0 from the buffer stops on the northern side of the station and get well up into the 180s (IIRC) at Marylebone.

The mileage to Manchester and Liverpool is continuous from a zero at Euston (via Crewe), with Piccadilly being 188 miles 67 chains according to a Quail map (not latest edition, but these things rarely change).  The same mileage continues to Golborne Junction where it changes to a zero a short distance away at at Newton-le-Willows junction, the original starting point for the northward branch off the Liverpool and Manchester.  This series continues to Preston where it re-starts from zero and does the same at Lancaster, so this part of the WCML does indeed reflect its early history.  It re-starts at zero at Carlisle, which is the zero point for most of the Caledonian. 

[Compound2632]

11 minutes ago, Edwin_m said:

It re-starts at zero at Carlisle, which is the zero point for most of the Caledonian. 

 

Presumably the mileposts between Kinnaber and Aberdeen still give the distance from Carlisle via Perth, despite the loss of the old main line between Stanley and Kinnaber?

[Compound2632]

16 minutes ago, Edwin_m said:

The mileage to Manchester and Liverpool is continuous from a zero at Euston (via Crewe), with Piccadilly being 188 miles 67 chains according to a Quail map (not latest edition, but these things rarely change).  The same mileage continues to Golborne Junction where it changes to a zero a short distance away at at Newton-le-Willows junction, the original starting point for the northward branch off the Liverpool and Manchester.  This series continues to Preston where it re-starts from zero and does the same at Lancaster, so this part of the WCML does indeed reflect its early history.  

 

A consequence of the Midland re-mileposting was that it gave the company the distinction of the highest-numbered milepost in the country, on the Dentonholme goods lines - 308 miles or thereabouts - but later eclipsed by a Great Western re-mileposting giving 326¼ miles outside Penzance - via Bristol, though.

[Edwin_m]

1 hour ago, Compound2632 said:

 

Presumably the mileposts between Kinnaber and Aberdeen still give the distance from Carlisle via Perth, despite the loss of the old main line between Stanley and Kinnaber?

Yes, the mileage changes at the former junction from 33 miles 26 chains from Dundee Tay Bridge, to 203 miles 11 chains from Carlisle via Perth and Forfar.  One of many cases where an apparently random change gives a clue to railway history.  

 

I noticed on a visit to Ireland that north of Bray there is a modern change of mileage sign that even gives the origin points, where the mileage from Pearse via Dun Laoghaire gives way to the mileage from Harcourt Street.  

[keefer]

1 hour ago, Compound2632 said:

 

Presumably the mileposts between Kinnaber and Aberdeen still give the distance from Carlisle via Perth, despite the loss of the old main line between Stanley and Kinnaber?

 

24 minutes ago, Edwin_m said:

Yes, the mileage changes at the former junction from 33 miles 26 chains from Dundee Tay Bridge, to 203 miles 11 chains from Carlisle via Perth and Forfar.  One of many cases where an apparently random change gives a clue to railway history.  

 

I noticed on a visit to Ireland that north of Bray there is a modern change of mileage sign that even gives the origin points, where the mileage from Pearse via Dun Laoghaire gives way to the mileage from Harcourt Street.  

 

A few years ago I noticed a 'Carlisle 225' MP at the north end of the Down platform at Stonehaven.

On looking in an old Sec. Appendix, the northbound ECML had:

0 MP at (in) Edinburgh Waverley up to Dundee.

0 MP at Dundee up to (the former site of) Kinnaber Junction.

North of Kinnaber, MP measured from Carlisle - as Compound says, via Perth and Forfar (the extinct Strathmore route)