Menu Close

Project developments – railway company records

In this post, long-term volunteer Philip James thinks about some of the practical issues in project work, particularly under Covid. This helps us both appreciate what it is project volunteers are doing, and how they do it – important if we’re to be transparent about how the data is created and the choices made along the way. This is prompted by the move from dealing with state-produced, printed accident investigation reports to the hand-written, company-produced records (held by The National Archives).

This is a great example of collaboration across institutions, as thanks to colleagues at The National Archives, we were able to share images of the railway company accident records with the NRM team. This meant we were able to keep the NRM supplied with records during Covid – they’d finished the state records and asked for more to do. These records, combined with the work of volunteers at The National Archives, will be coming into the project in due course. As ever, our thanks to all involved!



It is now more than two years since my first guest post and a lot has changed. The project has moved on to other sources of accident data, including hand-written reports. In the wider world, COVID-19 has profoundly affected life and for me, my father has died and I am now the sole carer for my elderly mother. Until the death of my father, I was having a relatively good and productive lock-down but life is different now.


Technology used

When working from my own home, I found it necessary to have two monitors, one for the project spreadsheet and one for the data file I was analysing. I now have to spend much more of my time away from home and have had to adapt to a different way of working. A laptop computer enables me to work from my mother’s home and a separate large monitor enables me to spread the display over two screens replicating the configuration in my own home.

A possible issue will be maintaining a good working posture as in my own home I have appropriate furniture for using IT equipment. In my mother’s home, I have to improvise. I expect this will be an issue for many people having to work from home during the pandemic. It will be interesting to see what problems people have with posture and eyesight in future.

Until now, I have mostly used Microsoft Office products such as Excel, Word and Outlook. On my laptop, I decided to try using Open Source tools and therefore the opportunity arises to make comparisons. I have used Apache Open Office and Neat Office, applications that seek to replicate the functionality and characteristics of Microsoft Office. In general, the open source tools seem to be good but there are detail differences and in some cases functionality may be deficient although not bad in the context of what is essentially a free product. On the email side, Outlook is definitely far superior to Mail for Windows 10 although the latter is manageable.

Internet connectivity has been surprisingly good despite having to rely on a wireless connection to a nearby hotspot rather than the cable broadband connection to which I am accustomed. As always, regular backups of data are vital and now I have the added task of keeping two different computers synchronised. Fortunately external disk drives and memory sticks simplify this task although care is needed.


Project experience and issues

Compared with the typewritten accident reports I have worked on before, many of the hand-written reports are hard to read and lack most of the detail present in the reports prepared by the state Railway Inspectors. In particular, they give no idea of how an accident happened, sometimes even if it was fatal. It may be possible to make some credible assumptions based on the limited information given but in general, it can only be said that accidents took place on railway property.

While this is a valid assessment of the hand-written reports I have worked on, some batches do have more detail and clearer handwriting. The presence or absence of detail probably reflects the contemporary policy for what should be recorded, a policy that might differ from one company to another.

As before, an internet search can help to confirm the names and spellings of accident locations and the counties in which they are located. It is possible that our project leads will decide to standardise on such things as spelling and county names to be used. (Editor’s note: Yes to this! We do this, so far as possible.) They may also have to look at the names used to describe occupations or job functions (Editor’s note: this is more of a challenge). A possible approach is to use a standardised name or spelling to aid reference while retaining the original for historical perspective.

The spelling example I like to quote is ‘wagon’ or ‘waggon’. The former is used in the recent handwritten reports while the latter is invariably used in the typewritten reports prepared by railway inspectors. There is information about it online. In summary, ‘wagon and waggon are different spellings of the same word meaning, among other things, a sturdy four-wheeled vehicle for transporting things. Waggon was preferred in British English until a century ago, and it still appears occasionally, but it is fast becoming archaic. In this century, the shorter one is preferred in all main varieties of English.’ (Editor’s note: where possible we’ve gone for ‘wagon’, but consistency is a challenge. The crucial thing is that in the notes accompanying our data we’ve noted the issue.)

Personally, I have a high regard for the state Railway Inspectors and hence a bias to their way of working. Their reports were seen as something of a benchmark for accident reporting and the Health and Safety Executive received some criticism when it took over the Railway Inspectorate and made changes. That said, natural language is continually evolving and the spellings we choose to use are a choice with consistency being of most importance.

About twenty years ago, I attended a speed-reading course run by Katherine Redway of Kathryn Redway Associates. A statistic I picked up is that the English language contains about five times as many words as it did in Shakespeare’s time. Presumably many of the words have been imported from other languages, e.g. concorde, robot and schadenfreude.

County boundaries have changed over the years, many in recent decades but some much longer ago. In particular, some towns or cities may be a county in their own right. (Editor’s note: here we’ve asked the teams to use the historic county – and in some cases, country – as at the time of the accident.)

Many of the reports made use of abbreviations and some of these took time to work out. Sometimes a ‘eureka’ moment would lead to one or more becoming clear. In general, I have worked through the accident reports line by line but for some of the reports, in particular those that had the least detail, it was quicker to work on a column by column basis. In one batch, I noticed that there were, perhaps predictably, relatively few entries over the Christmas period. This helped me complete the batch sooner than expected.


Interesting Case

In the context of the accidents themselves, no single interesting cases have emerged, mainly because the data I have seen is relatively terse. One accident aroused my interest because it occurred in the Pembrokeshire village of Kilgetty. I had wondered where this name came from. It was used to refer to a form of disk encryption I encountered on computers, typically laptops about 20 years ago. Perhaps one of the software designers came from Kilgetty or the project had some tenuous association with it.

This also reminded me of the UK code name, Granby, for the operation relating to the first Gulf War. It was the name of a hotel close to an important Ministry of Defence building. Unlike some US code names, their UK equivalents don’t seem to bear any obvious link to the operation in question; probably a good thing but it could mean Kilgetty was a name chosen at random.


Accident Locations and Companies

The primary company involved in an accident was the same for each accident in a batch but spotting other companies involved in a tersely-worded report was harder. Sometimes a clue lies in the location of the accident. The railway companies were territorial but did ‘invade’ each other’s territories in search of business. They might also have running powers over other companies’ tracks to reach sidings or goods yards well within those companies’ territories. Identifying these sidings and goods yards is harder. Fortunately, there are surviving maps and track diagrams annotated with the names of the companies represented.

Railway Clearing House diagram of the Millway Docks area, showing the railway companies involved.
Millwall Docks (Harrow Lane) and Poplar courtesy of Wikimedia. An example of a map of the East and West India Docks which were served by the London and Blackwall Railway (LBR) and the North London Line (via Victoria Park and Bow) but with sidings owned by other companies present.

The LBR became part of the Eastern Counties Railway (ECR) and later Great Eastern Railway while the North London Railway was once called the East and West India Docks and Birmingham Junction Railway. Today, the goods sidings have gone but the approach lines are part of the Docklands Light Railway (DLR) and the DLR Poplar Depot occupies the site of the former Harrow Lane Sidings.

Larger companies had a practise of acquiring smaller ones. A case in point, the London, Tilbury and Southend Railway (LTSR), my local line, was acquired by the Midland Railway (MR) so locations such as Tilbury can quite legitimately be shown as MR for the period of the accident reports we are dealing with (pre-1923).

In its early years, the majority of directors on the LTSR Board were appointed by the London and Blackwall Railway and the Eastern Counties Railway, a legacy of how the line came into existence. This led to some problems for the LTSR particularly when the ECR took over the LBR and had designs on taking over the LTSR as well. They probably had their sights on the docks at Tilbury.

Fenchurch Street Station was built to serve the LBR and later served the ECR as well. Today it is only served by the LTSR or its successors C2C / Essex Thame-side. The LTSR proper did not start until trains reached Forest Gate Junction via Stratford, the route via Plaistow being built later. Until that point, trains ran over lines built for the LBR and ECR hence their directors role.

The Barking to Gospel Oak (GOBLIN) line was the means by which the MR linked the LTSR to the rest of its network. This route is composed of the Tottenham and Hampstead and Tottenham and Forest Gate railways.

Map of the Limehouse area.
Limehouse Curve.
Courtesy of Disused

My daily journey to work took me past Limehouse (then called Stepney East and in earlier times, just Stepney) and to the east of the station was the Limehouse Curve forming a triangular junction with what is now part of the Docklands Railway. This would have been an access route for freight trains to Poplar. The Limehouse Curve viaduct was extant but disused for many years but was largely demolished about twenty years ago. Flats or offices now occupy the alignment but the junction locations at either end are apparent and the former route easy to see.

The diagram also shows the location of the original Limehouse Station opened by the LBR but not reinstated by the DLR. Some of the brick arches forming the viaduct were originally fitted out as houses. Such accommodation was not popular but the windows showing their location were visible for decades.

Philip James

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.