Allocation of Regimental Service Numbers

[TimPowell186]

Could anyone advise how regimental service numbers were allocated. I am interested in 24272 Private Frank Pearman of the 5th OBLI whose M.I.C. does not give an entry into theatre date. I assume these were given out chronilogically as they begin as 5 digit numbers such as 10130 James Edward Lewis Hinton who arrived May 1915 and ended up as 6 digit numbers as high as 267146 Private Fred Turner.

Can a date of entry be inferred by the service number?

[Sepoy]

The best source of information on this subject can be found on the Long Long Trail in the following articles

http://www.1914-1918.net/renumbering20.htm

and
http://www.1914-1918.net/renumbering.htm

I know that some people with interest in specific Battalions have used the surviving Pension/Service records to work out rough attestation/enlistment dates, but I do not think it is a precise science.

Sepoy

[LEUZEWOOD]

I've always found the following site useful to determine rough enlistment dates, but on first inspection can't see the 5th OBLI on there.

http://armyservicenumbers.blogspot.co.uk/p/index.html

[saw119]

An interesting discussion as I have done this for one of my men for which no service record has survived. It can only ever give you a rough date of enlistment but I decided it was better than nothing despite my misgivings. The Army Service Numbers website is exceptionally useful in some instances, for example I've checked it's reliability on Victorian service numbers and found it to be very accurate, but nobody has definitively proven that service numbers were allocated to the same pattern in the Great War.

[Guest]

Could anyone advise how regimental service numbers were allocated. I am interested in 24272 Private Frank Pearman of the 5th OBLI whose M.I.C. does not give an entry into theatre date. I assume these were given out chronilogically as they begin as 5 digit numbers such as 10130 James Edward Lewis Hinton who arrived May 1915 and ended up as 6 digit numbers as high as 267146 Private Fred Turner.

Can a date of entry be inferred by the service number?

The numbering system was theoretically quite simple but in reality became rather complex. As a starting point, generally speaking, numbers were allocated chronologically, however in cases where a unit had a number of recruiting stations there is some evidence that blocks of numbers were issued. There are also cases where blocks or runs of numbers stop abruptly and restart with the numbers between seemingly unused.

If your aim is to find the enlistment date for your man, the best approach is by 'bracketing' using the surviving service records and pension records. As most records were destroyed this can be a rather laborious task. If you have access to Ancestry.co.uk try searching for Service Records and Pension Records for Army numbers either side of 24272 with Oxford* in the Regiment box. This will capture both Oxford, Oxfordshire, I have done this for a few Yeomanry units and typically one finds a surviving Service record within 20-30 numbers either side. In the case of the Infantry this may be a larger number. Once you have a bracket with your man in the middle this should allow you to get dates within which he enlisted.

To do this just enter Oxford* in the box marked Regiment and hit search. It will generate 2780 returns which you can surf over the "View Record" and boxes will pop up with the numbers - it will take about an hour to quickly find the closest numbered records

A quick glance shows

22610 enlisted on 17th Sep 1915 (1,662 numbers before 24272)

25097 enlisted on 1st Nov 1915 (825 numbers after 24272)

Which suggest that there is a good chance he enlisted between these dates. i.e. you have a bracket of 43 days. A bit more searching would narrow this down. Almost certainly second half of October 1915.....but be mindful that if block of numbers were allocated to different battalions within the same Regiment there might be 'later' numbers with 'earlier' dates and vice versa. To increase certainty it is worth trawling all 2780 records and again worth cross checking the numbers against MICs as some might show which Bn they served in....

There is no established record of how the service records or pension records were stored and consequently the pattern of records that survived is quite random. You might be lucky and find lots of records with Army numbers close to your man, or it may be a harder task. Good luck with your search.

MG

[Graham Stewart]

An interesting discussion as I have done this for one of my men for which no service record has survived. It can only ever give you a rough date of enlistment but I decided it was better than nothing despite my misgivings. The Army Service Numbers website is exceptionally useful in some instances, for example I've checked it's reliability on Victorian service numbers and found it to be very accurate, but nobody has definitively proven that service numbers were allocated to the same pattern in the Great War.

Don't know about that - these two lads didn't do too badly regarding the Infantry, which included WWI numbering;- "Regimental and Army Numbers of the British Line Infantry Soldier from AD. 1800 to 2008" By David Langley & Graham Stewart

[Graham Stewart]

The numbering system was theoretically quite simple but in reality became rather complex. As a starting point, generally speaking, numbers were allocated chronologically, however in cases where a unit had a number of recruiting stations there is some evidence that blocks of numbers were issued. There are also cases where blocks or runs of numbers stop abruptly and restart with the numbers between seemingly unused.

If your aim is to find the enlistment date for your man, the best approach is by 'bracketing' using the surviving service records and pension records. As most records were destroyed this can be a rather laborious task. If you have access to Ancestry.co.uk try searching for Service Records and Pension Records for Army numbers either side of 24272 with Oxford* in the Regiment box. This will capture both Oxford, Oxfordshire, I have done this for a few Yeomanry units and typically one finds a surviving Service record within 20-30 numbers either side. In the case of the Infantry this man be a larger number. Once you have a bracket with your man in the middle this should allow you to get dates within which he enlisted.

To do this just enter Oxford* in the box marked Regiment and hit search. It will generate 2780 returns which you can surf over the "View Record" and boxes will pop up with the numbers - it will take about an hour to quickly find the closest numbered records

A quick glance shows

22610 enlisted on 17th Sep 1915 (1,662 numbers before 24272)

25097 enlisted on 1st Nov 1915 (825 numbers after 24272)

Which suggest that there is a good chance he enlisted between these dates. i.e. you have a bracket of 43 days. A bit more searching would narrow this down. Almost certainly second half of October 1915.....but be mindful that if block of numbers were allocated to different battalions within the same Regiment there might be 'later' numbers with 'earlier' dates and vice versa. To increase certainty it is worth trawling all 2780 records and again worth cross checking the numbers against MICs as some might show which Bn they served in....

There is no established record of how the service records or pension records were stored and consequently the pattern of records that survived is quite random. You might be lucky and find lots of records with Army numbers close to your man, or it may be a harder task. Good luck with your search.

MG

Edit the 5th Bn OBLI was raised in

While I agree with your method, it does only work when no transfers are involved and this is where it falls down and from my own studies once you go into late 1915 and on into 1916 numbering allocations within Regiments, you can forget enlistment dating altogether - there are just too many transfers to prove anything.

[saw119]

While I agree with your method, it does only work when no transfers are involved and this is where it falls down and from my own studies once you go into late 1915 and on into 1916 numbering allocations within Regiments, you can forget enlistment dating altogether - there are just too many transfers to prove anything.

This is exactly the situation I was talking about when I mentioned the WWI numbers in the post quoted above and what pitfalls are encountered. Best estimate is my man enlisted sometime late 1915/early 1916 which makes it quite difficult to be certain.

[James]

Another resource is to use the SWB lists which will give date of join (which could be with an entirely different unit as noted) I have found quite a lot of success combining all the methods discribed (and resources). By building up a database the wildcards are quickly highlighted. You can therefore get a good aproximation of when an individual joined a unit but not necessarily when they first enlisted.

[Guest]

While I agree with your method, it does only work when no transfers are involved and this is where it falls down and from my own studies once you go into late 1915 and on into 1916 numbering allocations within Regiments, you can forget enlistment dating altogether - there are just too many transfers to prove anything.

Noted , and I would certainly agree that as the War progressed it became very complex and in many cases impossible to unravel. I assume you mean when trying to establish when a man joined the 5th OBLI (for example) if he was transferred in from another regiment and assuming his Service Record does not exist.

The approach I have used focused on the Service Records which in turn relies on the attestation forms....which I assume is only signed once when the man enlists, regardless of a subsequent transfer - please correct me if I am mistaken on this. In the example above the dates came from the attestation documents, signed and witnessed and dated twice, usually with the second date appearing one day after the recruit signed his attestation form.

Even if a man subsequently transfers to a different unit, I assume that his original attestation form will remain in his Service Record - also showing the original unit.. If I am right on this, then using the bracketing method with Service Records should in theory establish when a man (who was enlisting for the first time) joined. More than happy to be shot down on this as my experience is limited to the TF and the Yeomanry in particular where the samples are significantly smaller than a line infantry Regiment.

In my experience the main flaw in my approach is that blocks allocated to different recruiting stations sometimes creates 'lags' whereby No 24600 (for example) sometimes attest later than 24601 (for example) simply becasuse the recruiting station using the block up to 24600 experienced a greater surge in recruits.

Given that half of all recruits joined before Jan 1916 (i.e those that volunteered rather than those that were conscripted) I guess there is roughly a 50% chance of establishing the month a man joined up, particularly if he was one of the New Army volunteers.

As mentioned in my initial post, in theory it was simple but in reality it was complex.

Where can I buy your book Graham? It sound very interesting...and I can't find it on Amazon.

MG

P.S. I found that in some TF units blocks sometimes also had an alphabetical run, so for example if on a particular day 40 men had been recruited, they were numbered in a way that corresponded to their surnames in alphabetical order. There is also a geographic concentration of runs of numbers, reflecting (I believe) the geographic allocation of numbers to different stations. I realise this does not necessarily apply in this case but it does illustrate some of the quirkiness of the numbering system and perhaps provide some approaches a researcher might attempt to take in order to unravel an individual's service.

[Muerrisch]

Please sir, please sir, the other half reporting for duty.

The booklet was printed locally, spiral bound. I have a few copies left, £13 to include postage UK.

Alternatively the Military Historical Society published it as an epic 5 -parter in its Bulletin.

Graham and I were awarded to Lummis Cup for our efforts.

BUT SEE ALSO THE MASSIVE STUDY BY PAUL NIXON AT

http://armyservicenumbers.blogspot.co.uk/

This is the result of a huge effort and a prime source ............ in our defence, Graham and I covered the whole history from year dot,whereas Paul limited the period but

WOW! What a reference source.

[Guest]

Yes please. Where do I sent the money?

MG

The Massive study by Paul Nixon at...???

[jay dubaya]

Link above in post 3

[Muerrisch]

Martin G your box is full! Try a PM to me.

[Guest]

Martin G your box is full! Try a PM to me.

It's all the hate mail... Box now cleared. MG

[ss002d6252]

I found that in some TF units blocks sometimes also had an alphabetical run, so for example if on a particular day 40 men had been recruited, they were numbered in a way that corresponded to their surnames in alphabetical order. There is also a geographic concentration of runs of numbers, reflecting (I believe) the geographic allocation of numbers to different stations. I realise this does not necessarily apply in this case but it does illustrate some of the quirkiness of the numbering system and perhaps provide some approaches a researcher might attempt to take in order to unravel an individual's service.

I found this with the 6th DLI and the pattern seemed to change on a daily basis.

Craig

[Graham Stewart]

Question for you numbering pundits;-

What single factor ensures that a mans 'date of enlistment' changes from those around him - excluding transfers? For example you have a block of 100 personnel all enlisting on the 24th November 1914 and yet one in that group has an enlistment date "14th February 1915".

[Guest]

Question for you numbering pundits;-

What single factor ensures that a mans 'date of enlistment' changes from those around him - excluding transfers? For example you have a block of 100 personnel all enlisting on the 24th November 1914 and yet one in that group has an enlistment date "14th February 1915".

What is the source of the dates? I.e. do we have all 100 attestation forms or are the dates provided in a secondary document(s) that involved a transcription of the original information. If it was the latter I would say clerical error.

If the original docs existed for all 100 I would say a re-used number. i.e a number allocated, then the man is medically discharged [Edit: or discovered to be too young when his mother turned up looking for him...] before he could start training and the number re-used.

MG

[ss002d6252]

I was wondering about the re-use of a number.

Craig

[kenf48]

Fascinating discussion though this is, to go back to the original post and query there are a couple of erroneous assumptions therein in that six digit numbers apply to TF soldiers renumbered in 1917 so they cannot be considered chronological in that sense i.e. Private Fred Turner was renumbered in the series allocated to the Buckinghamshire Battalions (his original number was 5398); and secondly you cannot infer a date of entry into theatre from the number with any certainty, only the date of attestation/enlistment as there are so many variables, a few of which have been touched on above.

The CWGC notes Frank Pearman was with the 5th Bn when he was killed in action, as the Battalion went to France in May 1915 he was clearly a replacement who entered theatre after the 1st January 1916, whether or not that was before or after the Battle of the Somme is speculation. His age may offer some clue, if he was attested in late 1915 he almost certainly enlisted under the Derby Scheme and would have been placed in a 'group'

See http://www.1914-1918.net/derbyscheme.html

Frank was born in 1885 (1911 Census) and was married so would have been placed in Group 36 with a mobilisation date of 29 May 1916 (this could be further deferred but is unlikely to have been earlier) then 12 weeks training so I'd guess he was a replacement for Somme losses but as has been stressed above it's by no means an exact science.

The war diary, including the attack in which he was killed has been transcribed and is here http://www.lightbobs.com/5-oxf--bucks-li-1916-1917.html

As an illustration of how random it can be 25097 cited by Martin above originally enlisted in the Worcestershire Regiment and did not go overseas (to India) until October 1917 and his OBLI number was 30432 allocated when he transferred to the 1st Bn In India in January 1918 (more Ancestry indexing!).

Another example 26745 (another Derby Scheme man) was attested 11 November 1915, was mobilised on the 9th September 1916 to the OBLI and then transferred to the Rifle Brigade the following month and went overseas with them in December 1916.

Ken

[Graham Stewart]

When looking at your Databases and you find an 'odd date' of enlistment in an otherwise tight date grouping, you'll find that the one single factor is 'Desertion'. All previous service is 'forfeit' and your 'date of enlistment' begins when you rejoin your unit(if caught). This is means that many enlistment dates on the SWB Rolls are 'oddballs' in enlistment groupings, not falling into place on your Databases and the date on the Attestation Form is invalid.

You can be transferred, renumbered and returned to your unit of origin, be renumbered again or given back your old number, but your date of enlistment will not change.

You can be transferred to the Class 'B' Reserve(Derby Men) and then be re-called to active service, but your date of enlistment will not change.

[Guest]

All very interesting.

Here's an anomaly:

I have a database of 1,752 men who served in the Derbyshire Yeomanry. Of these men I have 869 men with 4-digit numbers AND the 5-digit numbers, however the sequence of number order in the 4-digit numbered men does not translate smoothly to the 5-digit numbers.

The 5-digit numbers have two sequences...one starting with 75000 and the other starting with 76900 and they appear to be rather mixed up among the men when listed by their original 4-digit number. Having studied the data it seems very clear that on renumbering, men in the 1/1st Derbyshire Yeomanry were given a number from the 75000 sequence, all of them in order corresponding with their number order of the old 4-digit series. The men on the 2/1st and 3/1st were given numbers from the 76900 series, again in perfect sequence. The only way I managed to spot the underlying driver is that I managed to get (and annotate) the men who had the 1914-15 Star from the medal roll. There are just a small handful of men who were clearly 1/1st in 1914-1915 (served overseas, MIC evidence, medal roll evidence) who ended up in the 2/1st or 3/1st by mid 1917 and subsequently received a 'late' 5-digit number.

Given the Yeomanry were much smaller units, with only 3 lines it was probably much easier to spot, but it does highlight just one of the many anomalies. I assume it was driven by administrative convenience; the 1/1st being in Salonika on the date of renumbering and the 2/1st and 3/1st being based in Blighty.

The main point here is that pre renumbering sequence and post renumbering sequence overall had a low 'regimental' correlation but a high 'battalion' correlation' on the date of the renumbering. Something to watch out for.

MG

[TimPowell186]

Thanks very much everyone

[Muerrisch]

I was wondering about the re-use of a number.

Craig

Absolutely forbidden.

Which is not to say impossible of course.

However [there is always a however]

Although the regular battalions of a regiment shared a common series beginning at 0001 in 1881, it is always the case that the Special Reserve battalion had a different series, and each of the TF battalions had its own. As an example RWF had 6 series on the go in 1914 so duplications existed, and could look like re-use if one of the pair [or worse!] was killed.

[saw119]

Thanks very much everyone

I'm sure it's much clearer know!