Menu Close

Allocating accident investigations

In this post, project volunteer Brian Grainger, with the NRM team, raises some questions about the practicalities of the Board of Trade accident inspectors’ work. His previous post, here, delved into who the inspectors were. It’s therefore nice to welcome Brian back to the blog.  When Brian put this together he was working on the latest batch of accident reports, covering 1900-1910. By virtue of seeing so many reports – an indictment of the dangers of the railway industry at this time – he started thinking about which accidents were investigated and why. These are important questions, as they shape the records we have left today (given only around 3% of cases were investigated by the state inspectors).

If you have any thoughts on these questions, or a suggestion for a guest post, we’d be very keen to hear from you – do please get in touch.


In the course of work on the Railway Work, Life & Death project, some questions came to mind:

  • Were fatal accidents investigated and reported on more quickly than other accidents?
  • Did each inspector undertake inquiries in specific areas or relating to specific railway companies?
  • How were accidents allocated to and reported by each inspector?

In an attempt to have at least an initial view on these questions an analysis of the information in three quarterly reports (1901 Q4 B, 1903 Q1 B and 1905 Q1 B covering 160 individual accidents) has been attempted. It should be stressed that this is from a very small selection of data and over a short time period. It is quite possible that in later years the procedures may well have changed.


Were fatal accidents investigated and reported on more quickly than other accidents?

In Appendix B reports of this period there are three dates quoted:

  • The date of the accident
  • The date of the order instigating the inquiry
  • The date of the report

These dates give some intriguing intervals between the dates, for example how long was the interval between the accident and the report date, between the accident and the order date and between the order date and the report. Unfortunately there is generally no indication as to the date of the inquiry itself.

Looking at all accidents over the three quarters examined the average time between the accident and the report dates is 45 days with a range between 20 and 104 days. Reading these two accident reports there does not appear to be any obvious reason why they were so long or so short.

The average time between the accident and the order dates is 24 days (with a range of 5 to 74 days) and between the order and the report date is 20 days (with a range of 4 to 67 days).

These average times show a strong consistency over the five year period suggesting that the procedures were fairly consistent.

Looking now at fatal accidents only (total 46) the average time between the accident and report date is 48 days (with a range of 22 to 104 days), the average time between the accident and order dates is 25 days (with a range of 5 to 74 days) and between the order and report dates is 24 days (with a range of 4 and 67 days).

It can be clearly seen that there is little or no difference in the average time intervals between the all accidents and the fatal figures and indeed in both the time intervals between the accident to order dates and the order to report dates are slightly longer. This would suggest that all accidents were treated similarly from the point of view how and when they were investigated and reported on.


Did each inspector undertake inquiries in specific areas or relating to specific railway companies?

This is by no means clear from this small sample of data. In the three quarters looked at there were only 2 inspectors mentioned, JH Armytage and John PS Main (although there may be other inspectors during this 5 year period in other quarters). Because of this it is difficult to come to any clear conclusion on this question.

In these quarters Armytage conducted 89 and Main 71 inquiries. Whilst there are some patterns, for example only Main conducted inquiries in the east and south west of England and Armytage carried out nearly twice the number in Lancashire and Yorkshire, elsewhere it is less clear. Both Armytage and Main conducted enquiries in Ireland.

Looking at railway companies the picture is, probably predictably due to their geographic nature, very similar. Only in one instance, the London and North Western Railway, is there any real difference between the inspectors with Armytage conducting 11 and Main 3 inquiries.

At this stage and with limited data it is difficult to know whether inspectors undertook inquiries in specific areas or relating to specific railway companies.


How were accidents allocated to each inspector?

In the three quarters examined inquiries were often ordered in batches of up to 9 on a particular date and only 30 were ordered as not part of a batch. The batches in 1901 and 1903 tended to be on any day of the week although in 1905 this appears to predominantly on a Thursday and it is clear by that time the preference was to order inquiries in batches with only one inquiry requested as not part of a batch of 2 or more.

It is also clear that in many cases the batches were allocated to a single inspector irrespective of the accident location. For example on 19/03/1903 a batch of 7 inquiries were allocated to John PS Main for accidents ranging from Devon to Aberdeenshire; this suggests the accident location was not necessarily a factor in allocating inquiries. Conversely on 08/04/1903 a batch of 5 inquiries were allocated to JH Armytage but in this case all were in Ireland, clearing indicating that at least in some instances location was a consideration.

The batch of 9 mentioned above was ordered on 04/03/1903 with 6 allocated to Armytage (from Kent to Lanarkshire) and 3 to Main (from Middlesex to Devon).

Turning to the Report dates it is clear that reports were also dated in batches with one instance of 8 reports all dated the same day. In this particular case this included the 5 inquiries by Armytage in Ireland ordered on the same day plus 3 by Main ranging from Devon to Yorkshire. A batch of 6 reports by Armytage dated 23/03/1903 was ordered on 2 dates, 04/03/1903 and 11/03/1903. Five of these were in Scotland and the remainder in Kent. Another batch of 6 reports dated 13/03/1905 were ordered on 16/02/1905, 22/02/1905 and 02/03/1905 these all being in Cheshire, Lancashire and Yorkshire.

It seems there is strong correlation between order and report dates and in some case also to location.

In looking at the report dates from each inspector it is possible to tentatively suggest itineraries in travelling to each accident location. Obviously at this stage it can only be speculative particularly as we do not know the inquiry dates but looking at a combination of both order and report dates it might be possible to follow an inspector’s travels around Britain.

As an example we can look at inquiries by Main covering a 3 month period in 1905. In late February he was in Norfolk and Suffolk, then 2 weeks later reporting on 2 accidents in Devon and in a further 2 weeks investigating accidents in London. A fortnight later he went to Cheshire for a single accident and another fortnight later was in London and Sussex. The next week he went to Gloucestershire for 2 inquiries and then seemingly over the next 2 weeks to an accident in Glamorgan and 2 accidents in Dorset and finally a single accident in London. Clearly this is speculative but does suggest a possible itinerary for John PS Main. It should be noted that his workload does appear to be less than that of JH Armytage; over the same period Armytage reported on 32 accidents against Main’s 16 accidents.

There is a suggestion that if an accident occurred in a particular area where an inspector was due to visit this was added to that itinerary at the last minute.



On the first question, there is no evidence that fatal accidents were investigated and reported on any more quickly than other accidents.

Regarding inspectors reporting on accidents in particular area or for specific companies the evidence is not clear from the limited information looked at here. There certainly is some correlation between inspectors and accident locations but the number of instances where this is contradicted does not from this small sample suggest that this was the general case.

The allocation of inquiries to inspectors does appear to be somewhat arbitrary but again with only two inspectors in the sample this is more than likely the only viable method. However there is clear evidence that at least some attempt is made to group inquiries and inspectors. A good example are the reports by Main in 1905, all of them being south of a line from the Wash to the Severn.

Reports were clearly issued in batches suggesting those in the same batch were undertaken as a single itinerary.


Brian Grainger

An Architect (although now retired) with a lifelong interest in canals and railways, Brian has since 2016 been a home volunteer transcriber working on the Railway Work, Life & Death Project. Also a keen genealogist, he is currently researching the Nesmyth family tree.


1 Comment

  1. TW

    “This would suggest that all accidents were treated similarly from the point of view how and when they were investigated and reported on.”

    Is this meant to say that though they weren’t investigated quicker, fatal accidents were more likely to be selected for investigation? That’s what my own calcuations have lead me to believe.

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.