Jump to content
Great War Forum

Remembered Today:

1911 Census Demographics - Regular Battalions


Recommended Posts

On 11/6/2016 at 14:04, MBrockway said:

Definitely a big improvement.  Google maps is OK so far as it goes, but it is flawed when you try and use it for more sophisticated stuff.  I also find it very slow to refresh/reload on moving around, especially since the last upgrade.  Much better to use better open source maps and GIS API's ... and without all the sinister tracking that Google is prone to!

 

I'm sure you'll find ways of embedding QGIS charts into webpages that are just as easy as creating a new Google Maps map.

 

One caveat about the gazetteer tools - be careful about simple word matching - your 'Kettering' could get hits other than the Kettering, Northants., that you expect and it's very difficult to spot a dot that's wrong on the map in a large dataset.  Are you able to select one of the dots/dot clusters and 'drill down' to see the source data?

 

How about trapping it with code in your Excel for detecting multiple matches?

 

You need to get it right, because such powerful maps will quickly get propagated across cyberspace and future researchers will use them as cast iron sources!

 

Mark

 

 

You make some very valid points. I am already encountering dozens of names that don't find a match such as Swansea which was a bit puzzling. Then I discover that the GIS source data has Swansea / Abertawe ...so I than to then duplicate the entry in the Master and create one entry for Swansea and one for Abertawe. (It is Govt data of course) Times a hundred for the anglicised Welsh town.  Malvern did not get a hit either and has to be split into its various parts. West Hartlepool has no entry so I had to create it from scratch. A surprising omission.. Initially it is a slog but each iteration gets us there slowly. Essentially I am tweaking the reference data to better match my own stylistic benchmarks to Stoke -on-Trent become Stoke on Trent (I have a hyphen phobia) and then matches immediately. 

 

The existence of the same place name in different parts of the country is of course a challenge. There are 38 places in the UK called Church End and five of these are in Bedfordshire. I have used the excel function to highlight the duplicates and have it as a checklist. before the final run. The Census data usually included the County as well so with a little tinkering with the formula I might be able to write a simple formula that require both town and county to 'match'. However, we then enter the problems of boundary changes. A few towns are alleged to have been in two different counties by the 1911 clerks. Big ones. Like Birmingham which some clerks believed was either in Warwickshire, Worcestershire or Staffordshire. Really. So I have to 'clean' the 1911 Census data returns first. It is OK for larger towns (and there is only one Birmingham in the UK data) but if a small town was the victim of a boundary change between 1911 and 1916 it might not show up. It ends up as N/A on the spreadsheet (over a hundred or around 10% of the sample of the 4th Worcesters). Of these half are overseas or Ireland (OS database does not reference any Irish town with the same  framework.... so as soon as one problem is solved, it creates another.... I am losing probably 15-20 names simply because I can not identify the town or village, even by Ctrl F searching the reference data for parts of the word. Given the OS data has 48,000 places, it is fairly comprehensive and if I drop say 1-2% it doesn't really deplete the main aim which is to show geographic dispersion of the origins of each battalion. 

 

Despite this I am confident that there will be a 99% solution. I may have to manually check 5% or around 50 names per Battalion. Often the original entry is a phonetic spelling: Ackney for example or an overseas location (Colonies usually)

 

I am acutely aware of crap-in, crap-out, so I am doing this meticulously from the start so that I don't import early mistakes into multiple sets of subsequent data.  

 

My aims are to attach the maps to e-books of the diaries with medal rolls. Either a link or preferably an embedded image that can be zoomed-in. I think that there are many more potential applications, however the 1911 Census data for India will generate at least 50,000 names and birthplaces for the Infantry alone. I will do a Cavalry Regiment as I have long been interested to see if their recruiting was Town or Country.

 

MG

Edited by Guest
Link to post
Share on other sites
  • Admin
12 minutes ago, QGE said:

 

You make some very valid points. I am already encountering dozens of names that don't find a match such as Swansea which was a bit puling. Then I discover that the GIS source data has Swansea / Abertawe ...so I than to then duplicate the entry in the Master and create one entry for Swansea and one for Abertawe. (It is Govt data of course) Times a hundred for the anglicised Welsh town.  Malvern did not get a hit either and has to be split into its various parts. West Hartlepool has no entry so I had to create it from scratch. A surprising omission.. Initially it is a slog but each iteration gets us there slowly. Essentially I am tweaking the reference data to better match my own stylistic benchmarks to Stoke -on-Trent become Stoke on Trent (I have a hyphen phobia) and then matches immediately. 

 

The existence of the same place name in different parts of the country is of course a challenge. There are 38 places in the UK called Church End and five of these are in Bedfordshire. I have used the excel function to highlight the duplicates and have it as a checklist. before the final run. The Census data usually included the County as well so with a little tinkering with the formula I might be able to write a simple formula that require both town and county to 'match'. However, we then enter the problems of boundary changes. A few towns are alleged to have been in two different counties by the 1911 clerks. Big ones. Like Birmingham which some clerks believed was either in Warwickshire, Worcestershire or Staffordshire. Really. So I have to 'clean' the 1911 Census data returns first. It is OK for larger towns (and there is only one Birmingham in the UK data) but if a small town was the victim of a boundary change between 1911 and 1916 it might not show up. It ends up as N/A on the spreadsheet (over a hundred or around 10% of the sample of the 4th Worcesters). Of these half are overseas or Ireland (OS database does not reference any Irish town with the same  framework.... so as soon as one problem is solved, it creates another.... I am losing probably 15-20 names simply because I can not identify the town or village, even by Ctrl F searching the reference data for parts of the word. Given the OS data has 48,000 places, it is fairly comprehensive and if I drop say 1-2% it doesn't really deplete the main aim which is to show geographic dispersion of the origins of each battalion. 

 

Despite this I am confident that there will be a 99% solution. I may have to manually check 5% or around 50 names per Battalion. Often the original entry is a phonetic spelling: Ackney for example or an overseas location (Colonies usually)

 

I am acutely aware of crap-in, crap-out, so I am doing this meticulously from the start so that I don't import early mistakes into multiple sets of subsequent data.  

 

My aims are to attach the maps to e-books of the diaries with medal rolls. Either a link or preferably an embedded image that can be zoomed-in. I think that there are many more potential applications, however the 1911 Census data for India will generate at least 50,000 names and birthplaces for the Infantry alone. I will do a Cavalry Regiment as I have long been interested to see if their recruiting was Town or Country.

 

MG

I'm looking at some mapping for work purposes (local authorities and council tax data) and the data is far from perfect (in fact I'm stepping back slightly to look at whether it's worthwhile using the data set - which is a government one).

 

Slow and steady and the result will eventually be obtained.


Crag

Link to post
Share on other sites

How confident are you that the Indian census returns include all the men in the battalion and not just those living in Barracks?  Presumably most of the officers lived in billets off Base, possibly some of the SNCO's and perhaps some of the married enlisted men.

 

This is certainly an issue with the Census returns for UK Barracks/Depots as I was considering using those to build a 1911 database of the Rifles battalions on Home Service and found too many were missing from the Institutional Census Returns.  Found several in the return for a nearby Asylum!

Link to post
Share on other sites

I would expect/ hope that regiments officially recruited nationally, such as the two Rifles, would produce recruiting loci which were well correlated with population. In other words, lots of recruits from big cities, not many from Shanklin or Achiltibuie.

 

What  might be very revealing is the correlation of recruiting with county/city etc of the county/city regiments. As a prime example, the RWF were known as the Brummagem Fusiliers for good reason until the Great War altered their demographic in favour of Wales.

 

I believe, but cannot prove, that the 1911 India census returns were for the entire unit. One quick way to reassure would be to compare establishment with actuals. The caveat might be the dreaded furlough, when soldiers were allowed walkabout throughout India and to be rationed by units along the way. And of course the officers took Blighty leave. I rather fear that the criteria for the census in India might be a bit difficult to establish. We need to find the equivalent of the Decree of Caesar Augustus.

Link to post
Share on other sites

Obstacles and difficulties are there to be overcome and are what make a decent project worthwhile!  :thumbsup:

Link to post
Share on other sites
On 11/6/2016 at 17:59, MBrockway said:

How confident are you that the Indian census returns include all the men in the battalion and not just those living in Barracks?  Presumably most of the officers lived in billets off Base, possibly some of the SNCO's and perhaps some of the married enlisted men.

 

This is certainly an issue with the Census returns for UK Barracks/Depots as I was considering using those to build a 1911 database of the Rifles battalions on Home Service and found too many were missing from the Institutional Census Returns.  Found several in the return for a nearby Asylum!

 

The 1911 Military Census for India and the Colonies is a complete shambles. Then Ancestry managed to make it worse. I have indexed it. Twice. Thanks to Ancestry's rather inexplicable tendency to remove all the cover pages, and summary pages, meaning about 3 days hard yards of indexing (that Ancestry itself had not done) became irrelevant in an instant. I recently updated it and posted on the GWF. The deletion of the cover pages and summary pages was an act of sheer madness. Whoever in the Govt allows this to happen needs to be fired. They have no idea about how to manage an archive of national importance and have allowed a commercial company exclusive access who have systematically done their very worst to provide easy access. It is a shambles. 

 

The challenge with the Census is that units often had detachments elsewhere in India, often Company size. It is a case of trying to find a needle in a very large haystack as the files have no logical order. A 'missing' Company might take up 3-4 pages of over 4,000 pages of randomly distributed units. We know the theoretical Peace Establishment for India in 1911 was 950 ORs, [Edit] 940 Rank & File however drafts had a large impact on the actual numbers. In addition India was a tough climate and many were hospitalised, so one needs to also trawl the other locations to find the invalids. It is hard yards.

 

4th Bn KRRC. I have traced 927 and I am definitely a Company short (Remember it was an 8-Company structure)

4th Bn Worcesters. I have over 1,000 and I am extremely confident I have tracked all but those in odd jobs or invalids. Clearly a draft had arrived and the returnees had not yet departed. 

 

The 1991 Census in India recorded everyone who spent the night in a specific location on 2nd Apr 1911. If a man was doing duty beyond the barracks, that locations such as a sanitorium would record his presence. Unfortunately one cannot search by unit. All the wives and children are recorded (as we would expect) in fine detail. 

 

The Officers all head each Company list. 

 

I would be interested in seeing the link for UK military establishments to do some research. MG

Edited by Guest
Link to post
Share on other sites

The front page of instructions for the census of 1911 of the British army in India specifically provides for all ranks on the unit Establishment to be included, and has provision for annotating those absent on census night. Should be 100% therefore.

Link to post
Share on other sites
8 minutes ago, Muerrisch said:

The front page of instructions for the census of 1911 of the British army in India specifically provides for all ranks on the unit Establishment to be included, and has provision for annotating those absent on census night. Should be 100% therefore.

 

They are there, the challenge is finding them. On first pass I usually get 850 of the expected 950. There is usually a Company of 100 at some dusty outpost drinkin' gin and beer while quartered safe out 'ere ... but when it came to slaughter, they did their work on water.

 

It's not just Inja's sunny climes that we have data for ...we also have data for every outpost from Burma to Tsingtao

 

MG

 

Right Flank Rear

 

Edited by Guest
Link to post
Share on other sites

Ohhhhhhh!

 

So the problem was interpretation of "unit establishment", whereby a detached company was apparently treated as a unit, and either rendered its own return or was included under a cantonment or barracks or garrison roll. Not good.

 

That way madness lies.

Link to post
Share on other sites
5 minutes ago, Muerrisch said:

Ohhhhhhh!

 

So the problem was interpretation of "unit establishment", whereby a detached company was apparently treated as a unit, and either rendered its own return or was included under a cantonment or barracks or garrison roll. Not good.

 

That way madness lies.

 I may be wrong, but I have spent hundreds of hours poring over the files. My assessment is that returns were done on a Company basis. This is logical as they were the lowest level of administration. Within a static battalion the returns are 99% Company by Company, headed by the Officers then WOs and SNCOs then Rank & File. There is little variation. Out-lier Companies on detachment made their own returns, however no-one appears to have considered that 105 year later someone might want to peruse the data consolidated by battalion. How inconsiderate. 

 

Occasionally one sees a gem, such as the Grass Farm and its establishment. I can only assume it was a farm aimed at generating better forage for the horses rather than cannabis. 

 

MG

 

Link to post
Share on other sites
53 minutes ago, QGE said:

I would be interested in seeing the link for UK military establishments to do some research. MG

 

Sorry Martin - I no longer have Ancestry access to get it back for you :(

 

I remember it took some very devious delving using the summary & cover pages to locate it,  If Ancestry have scrubbed these from the UK data too, then we could be in trouble.

 

I 100% echo your comments about mishandling of a key national research asset too!

 

I have similar issues with the new IWM search engine.  Considerably harder now to find items in the collection,

 

All seems to be driven by dumbing down the interfaces so smartphone touchscreens can use them - grrrr :angry2:

Link to post
Share on other sites

The Summary Books are in a separate file (they were originally at the beginning of each set of data but have been separated) and typically consist of just 3 pages: Cover Page, Instructions Page and a Summary sheet with numbers. They were particularly useful in that the Cover Page is distinctive and was easy to spot on the filmstrip view in Ancestry, so hopping from one file to the next was easy. This is no longer the case. 

 

Examples below show the sequences taken randomly from the middle of the Summary files. The three infantry battalions all showing more than 1,000 all ranks. Given Peace Establishment for India was 950 [edit] 1,031 All Ranks, this suggests to me that the Census captured all the men and possibly recently arrived. The 'missing' will be in hospital or on detachments and captured in that institution's Census. The alternate explanation is attached ASC, RAMC etc. Regardless of the detail it seems that all the men are counted as we would expect. 

 

 

2nd Bn Royal West Kent Regt. Roberts Barracks Peshawar 2nd Apr 1911.

Book I ......300 + 70 dependents (wives and children)

Book II......300

Book III.....300

Book IV.....116

Total........1,016 + 70 dependents 

 

2nd Bn North Staffordshire Regt, Peshawar NWP 2nd Apr 1911

Book I.....297 + 80 dependents

Book II....300

Book III....299

Book IV...206

Total.....1,102 + 80 dependents

 

...and is followed by...

1st Bn Royal Munster Fusiliers  9th Apr 1911

B & F Companies 1st Bn Royal Munster Fusiliers

Book I.....249

D Coy 1st Bn Royal Munster Fusiliers

Book I....135

A & E Companies 1st Bn Royal Munster Fusiliers

Book 1...273

1st Bn Royal Munster Fusiliers (No Company stated. By process of elimination must contain C and G Coys)

Book 1.....277

H Company 1st Bn Royal Munster Fusiliers

Book 1......143 + 31 dependents

Total......1,077 + 31 dependents

 

followed by

 

7th Div Staff Office Meerut....... total 15

Meerut Cav Bde Staff Office......total 3

 

 

 

 

Edited by Guest
Link to post
Share on other sites

While looking at material relating to the East Surrey Regiment held at the Surrey History Centre I came across the following relating to the 1911 Census in the 1st Battalions Order (reference: ESR/2/2/23).

Bootneck

 

1st Battalion East Surrey Regiment Part I Orders Monday 28 March 1911

 

253. CENSUS RETURNS.                Forms H of the Census Return have been issued to all concerned to-day. They should be returned to Orderly Room on 3-4-11 containing the particulars of every Officer, N.C.O. and man belonging to, or attached to each Coy. (excluding men in Hospital) who sleeps in Barracks on Sunday night 2-3rd April 1911.

            The heading and 1st column should be left blank and the form prepared for signature of Commanding Officer.

 

1st Battalion East Surrey Regiment Part I Orders Saturday 1 April 1911

 

269. CENSUS 1911.                          All Wesleyans troops are to Census returns as “Methodists”, that being the designated recognised by the Census Commissioners in Ireland.

                                    Hd. Qtrs. No. 13049 (A.1.)

                        (Irish Command Order No. 383 dated 31-3-11.)

Edited by bootneck
Link to post
Share on other sites

1st Bn Border Regiment, based in Maymyo, Burma in 1911. Returns show 1130. The birthplaces of 1,017 can be identified to locations in Great Britain. The map below shows the distribution. Note where more than one man is born in the same town, the points overlap. This shows the Towns and villages that the men came from. There are 396 unique locations spread across the country. There are an additional 24 men born oversaes and 9 born in modern Eire who are not shown. There were over three times as many men born in London than in Cumberland. MG

 

1st Bn Border Regiment 1911 Census: Birth Places of All Ranks. 

QGIS 1st Border.JPG

Edited by Guest
Link to post
Share on other sites

4th Bn Worcestershire Regt distribution compared to 1st Bn Border Regt

 

Interestingly very few recruits from Scotland, Wales and Ireland. The Worcester's distribution shows a distinct line along the English/Welsh border. I suspect English regiments were not allowed to actively recruit outside England and these men were expatriate Celts who were recruited in England. MG

QGIS Comp.jpg

Edited by Guest
Link to post
Share on other sites

10th Hussars Indian Census 1911 - Place of Birth

 

601 individulas of the 683 names. Some were born overseas and the remainder did not specify a Town or it is illegible/untraceable. 

QGIS 10th H Census 1911.JPG

Link to post
Share on other sites
6 hours ago, QGE said:

The Summary Books are in a separate file (they were originally at the beginning of each set of data but have been separated) and typically consist of just 3 pages: Cover Page, Instructions Page and a Summary sheet with numbers. They were particularly useful in that the Cover Page is distinctive and was easy to spot on the filmstrip view in Ancestry, so hopping from one file to the next was easy. This is no longer the case. 

 

Examples below show the sequences taken randomly from the middle of the Summary files. The three infantry battalions all showing more than 1,000 all ranks. Given Peace Establishment for India was 950, this suggests to me that the Census captured all the men and possibly recently arrived. The alternate explanation is attached ASC, RAMC etc. Regardless of the detail it seems that all the men are counted as we would expect. 

 

 

2nd Bn Royal West Kent Regt. Roberts Barracks Peshawar 2nd Apr 1911.

Book I ......300 + 70 dependents (wives and children)

Book II......300

Book III.....300

Book IV.....116

Total........1,016 + 70 dependents 

 

2nd Bn North Staffordshire Regt, Peshawar NWP 2nd Apr 1911

Book I.....297 + 80 dependents

Book II....300

Book III....299

Book IV...206

Total.....1,102 + 80 dependents

 

...and is followed by...

1st Bn Royal Munster Fusiliers  9th Apr 1911

B & F Companies 1st Bn Royal Munster Fusiliers

Book I.....249

D Coy 1st Bn Royal Munster Fusiliers

Book I....135

A & E Companies 1st Bn Royal Munster Fusiliers

Book 1...273

1st Bn Royal Munster Fusiliers (No Company stated. By process of elimination must contain C and G Coys)

Book 1.....277

H Company 1st Bn Royal Munster Fusiliers

Book 1......143 + 31 dependents

Total......1,077 + 31 dependents

 

followed by

 

7th Div Staff Office Meerut....... total 15

Meerut Cav Bde Staff Office......total 3

 

 

 

 

 

Martin, I am a little sceptical regarding incoming drafts accounting for the substantial number of men over Peace Esatblishment. I agree that it is the obvious explanation but it fails to pass the litmus test of Treasury meanness.

 

A modern analogy is the Church of England and interrugnums [interregma? interregmi?] There are simply not enough priests to go round, so whenever a Vicar leaves a parish, the hole is not filled for, typically, a year or so. Our local church twice thus inside 20 years. Quite a saving.

 

These extra men have to be paid, fed, and watered and indeed clothed. There were strict demarcations of fiscal responsibility between GB and India, even though both batted for the same side.

 

Thus far, I cannot come up with a better idea than yours but I am wrestling with it.

 

Link to post
Share on other sites
2 minutes ago, Muerrisch said:

 

Martin, I am a little sceptical regarding incoming drafts accounting for the substantial number of men over Peace Esatblishment. I agree that it is the obvious explanation but it fails to pass the litmus test of Treasury meanness.

 

A modern analogy is the Church of England and interrugnums [interregma? interregmi?] There are simply not enough priests to go round, so whenever a Vicar leaves a parish, the hole is not filled for, typically, a year or so. Our local church twice thus inside 20 years. Quite a saving.

 

These extra men have to be paid, fed, and watered and indeed clothed. There were strict demarcations of fiscal responsibility between GB and India, even though both batted for the same side.

 

Thus far, I cannot come up with a better idea than yours but I am wrestling with it.

 

 Supernumerary to establishment?

 

I will double check the RMF just to reassure myself... MG

Link to post
Share on other sites

WIKI says [and I do not disagree]:

 

The Indian Trooping season generally began with troop ships leaving England in September, and ended with the last ships leaving India in March. This pattern was probably established once troop ships no longer sailed around the Cape of Good Hope and started using the "Overland Route", and then the Suez Canal after its opening in 1869.

The reasons for a restricted period were to travel in the cooler months so that

  • troops were not travelling during the hot summer months in unventilated ships, particularly in the Red Sea and the Gulf of Aden, when conditions could become dangerous.
  • unacclimatised troops from Britain were not travelling from the ports of Bombay or Karachi to their cantonments during the heat of an Indian summer.
In 1916, when normal procedures were disrupted due to the First World War, the "Karachi troop train incident" of the 5th June, 1916, resulted in the death of nineteen Territorial Troops due to heat stroke on a troop train between Karachi and Lahore.

Initially troops changed ships at Suez, so there were different ships on the routes England to Suez, and Suez to India, but subsequently (and by 1886) ships sailed a round trip from England to India, approximately three weeks in each direction.

Each season generally there were only two of the twelve or so voyages which called at Aden on the way out to India and three on the way back. The extra one coming from India was needed to effect the annual relief of the British infantry battalion in Aden.[1]

EDIT: I believe that the census was 2nd April? If so, surely the time-expired troops would have been on the Feb/March troopships on the way Home?

 

 

 

Edited by Muerrisch
Link to post
Share on other sites

Second sweep I get 1073 which is only 4 off what the summary pages state. 

 

Aug 1914 Returns show 1st Bn RMF still in Burma. I note the Establishment is for 940 "Rank & File" and 1,031 "All Ranks" so my mistake with 950. :

 

..........................Of......WOs...Sgts...Dmr/Bglr....R&F.......Total All Ranks

Establishment.....28.......2........45.........16..........940..........1,031

Strength.............23........2........45.........16..........989..........1,075

 

Here is a list of the Battalions in India and their total strength in Aug 1914

India Bn Establishment.jpg

Link to post
Share on other sites

Just for Derek - the Black Watch

 

988 plots of a potential 1050. Of these 924 hail from Scotland (88%) 126 from outside Scotland of whom 88 were born in England. Please note places with more than one man are still represented as single dots. There are 286 unique locations.

 

 

QGIS Black Watch Census 1911.JPG

Edited by Guest
Link to post
Share on other sites
3 hours ago, Derek Black said:

Excellent!

 

Thank you Martin

 

Cheers,

Derek.

 And since you were so nice about it, here is a closeup of Scotland and the borders as far south as Newcastle. If you would lik ethe hard data PM or email me. 

 

QGIS Black Watch Census 1911 2.JPG

Edited by Guest
Link to post
Share on other sites

A further refinement of the 1st Bn Borders 1911 Census data on place of birth. The larger the number born in a location, the larger the points become and the more intense in colour. I have applied a 50% transparency which allows the smaller underlying dots to still be viewed through the larger ones at discrete levels of zoom. 

 

QGIS 1st Border Size scaled.JPG

QGIS 1st Border Size scaled 2.JPG

Edited by Guest
typo/spelling
Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...