St Brides Beach
St Brides Beach


The first generally useful census to be held in the UK was in 1841 and it was repeated at 10 years intervals (with the exception of 1941). The volume and accuracy of the recorded data has tended to increase but the essence has remained the same - to try and identify every single person and their location at census time. This objective was achieved by the completion of an enumeration return for predefined Enumeration districts. Needless to say there were exceptions from the norm created mainly by non-domestic habitations (typical examples include vessels, military units and workhouses). 

To a large extent any particular return is only as good as the Enumerator employed to complete the documentation - some took great care and wrote legibly others less so. In Wales written and verbal bilinguality is a significant issue as English speakers lived in places which retained Welsh placenames (and to a lesser extent vice versa).  

One of the original reasons for the creation of the Cenquest 1891 census database was the disappointing result obtained when searching for known individuals. Searching for a Welsh person is difficult enough due to the limited number of names in use and hence the proliferation of similarly named individuals. This in turn means that ancillary data such as places of residence or birth are often of critical importance in the identification of an individual.  So often the failure to obtain a "hit" was clearly the result of inadequacy in the Welsh language by the enumerator or transcriber or both. The other area of concern was the lack of any checks on the internal consistency of the data. 

Cenquest has tried to reduce those sources of error. Transcription has been undertaken by staff with local knowledge and bilingual skills. Wherever possible the data has been subjected to internal consistency checks and any queries generated have been re-checked against the source materials.

The aim has been to provide information of a consistently high standard that is not slavishly dependent on the individual enumerator or transcriber. Much has been written on preserving the "purity" of the original record by transcribing "as is". Our approach is to suggest corrections or modifications where this will assist a family history researcher and to footnote any change made to the original record. We have permitted the addition of extra information in the form of honorifics, placename derivations and map references. These additions are ours and may be wrong, inaccurate or incomplete but are made in good faith to assist in the identification of people and places. In all cases we encourage family history researchers to support their identifications with a copy of the original information. We can provide this either as a hardcopy print forwarded by ordinary mail or as a scanned digital image attached to an email. Please note that the quality of the image is variable due to both legibility and preservation of the original.

 Data and facilities 

The primary dataset is a transcription of the census enumeration books which are available to Cenquest in the form of photographic materials courtesy of The National Archives.

Essentially this is a list of all places (properties) within a census enumeration district together with details of address and occupants. Schedule numbers (schedules) were allocated on an incremental sequential basis for each household and/or family. Within a schedule the members of the household were listed in order of relationship to the most "senior" or Head. A typical schedule might consist of father and mother together with their children in age order followed by servants. In other words an individual schedule may include distant relatives, visitors and workers besides the core family members.

Special enumeration records were often completed in respect of vessels, military and naval establishments together with pauper Institutions and Trinity House lights. The enumeration was usually undertaken by a senior official or his delegate. However small groups of people in these categories may also appear within the main text of an enumeration book but more often at the end as an Addendum. These "households" often contain a mix of family and organizational relationships. 

As far as possible we have tried to clarify the existence of these groups of people by generating artificial enumeration districts (900 series) and schedule numbers (where necessary) so as to organize the data in a meaningful manner (we have given relevant reference numbers and noted the source of any extractions).

Where an enumerator has omitted a schedule number we have generated a decimal fraction - so for example if 2 schedules have been omitted between schedules 7 and 8 then we have generated 7.1 and 7.2.

Enumeration districts.

The Enumeration districts define an area whose boundaries were well defined in order to avoid duplication or omissions in the recording process. Enumeration districts were grouped into sub-Registration districts which in turn provided the Registration districts of each County. Needless to say a degree of "administrative convenience" (and confusion) arises at county boundaries with some villages split across a number of counties and some parts of counties recorded in the returns of another. (Perhaps one of the more notorious is the inclusion of Cardigan the county town of Cardiganshire in the returns for Pembrokeshire, but there are many other examples. (It should also be noted that the counties and county boundaries of a century ago are not necessarily those of today). The Enumeration districts are often described in some considerable detail and we have tried to enhance this written record with visual additions. Where possible we have used contemporary material - but this has not been possible in all cases. (Should you have any material that you wish to share with fellow researchers please contact us).


Although the majority of people recorded in the census lived in the county of their birth, there were those who were born in other parts of Wales, Britain and indeed from all around the globe. The following examples illuminate:

 Mary Anne Ladd (8 years old) - living in Moylegrove but born in the Welsh colony of Patagonia (Southern Argentina).

 Charles S Sturt (a retired general of the Indian Army) - living in Newport but born in Sydney, Australia.

 George F Graham (a retired mariner) - living in St Dogmells but born in Montreal, Quebec, Canada.

 Charles Morris (a young tailors apprentice) - living in Cardigan town was born in Scranton, Luzerne County, Pennsylvania, USA.

 Mary A Harwood (a scholar) - living in St Mary Pembroke Dock was born in San Francisco, California, USA.

 Emma Eliza Wemys (a 65 year old married lady) living in Blaenporth, Cardiganshire was born in Port Elizabeth, Natal, which is now part of the Republic of South Africa.

 It is of course of crucial importance to your family history research to try and obtain the correct identification of the place of birth of your ancestor in order to obtain a Birth certificate - another verifiable documentation link in your researches.

There are many potential sources of confusion but especially so in Wales where the issues of Welsh and alternative spelling arise in abundance. (Alternative spellings are those that by their sheer frequency of occurrence need to be recognised even if strictly speaking they may be spelling errors).

 Further complications can arise from abbreviation (Saint, St, St., Sain and Sant), punctuation and concatenation (or any combination thereof).


For example the following:

SanclÍr, St Clears, St. Clears, Saint Clears (Carmarthenshire)

Rowen, Ro-wen, Y Ro-wen, Y Rowen (Caernarvonshire).

Duplicated placenames

Newport in Pembrokeshire (Welsh placename Trefdraeth)

Newport in Monmouthshire (Welsh placename Casnewydd - a modern artifact)

Note: Newport also exists in other areas such as the Isle of Wight and Cambridgeshire.   

Similar placenames 

St Dogwells (OS grid ref. SM9727) Welsh place name Llantydewi

St Dogmells (OS grid ref. SN1645) Welsh place name Llandudoch and with alternative spellings of St Dogmaels, St Dogmels.  

Unresolved placenames

For example St Mary(s), Pembrokeshire could refer to

St Mary (in-Liberty) in Tenby 

or St Mary Pembroke (Dock).

Bilingual placenames

It may well be that a researcher has access to documents or tradition that recall only the lesser used (and possibly abandoned) placename resulting in a failure to locate or identify a place that would have been recognised by previous generations.

For example Ambleston in Pembrokeshire is easily located and identified - but Treamlod the Welsh placename spelling of this village may elude researchers without local knowledge or relevant maps.

Similarly Hwlffordd (Haverfordwest) may not be immediately obvious.

Wdig (Goodwick) may leave both past and present generations baffled!

There is no doubt that errors have been introduced by the processes of enumeration and transcription. Cenquest has tried to minimise these by internal validity checks. Where a Pembrokeshire placename was not recognised by the Gazetteer then the source material has been reinspected and if necessary, further searches made of our placename sources in order to try and resolve the issue (many of the more elusive places have been Farms or Hamlets). For example Trecifft in Cardigan is not a village in Cardiganshire but an abandoned farm near to Cardigan town.

It has not been possible to eliminate all "unknowns" even within our target counties - we welcome any contributions that may be forthcoming from private researchers. 

Welsh language.

Clearly this is not the place for a detailed discussion but it is worth noting a few salient points about Welsh.

The adjective normally follows the noun in Welsh, in contrast to English where it precedes the noun. Hence Whitemill in English is Felinwen or "Millwhite" in Welsh.

The first letter of a Welsh word may be mutated depending upon the (historical) context according to one of three mutations:

Treiglad Meddal (soft mutation)

Treiglad Trywnol (nasal mutation)

Treiglad Llaes (aspirate mutation)

In practice this means that if the source is within a Welsh text, then a placename, for example may start off somewhat differently than if it were written in isolation.

Original Leading Letter  B C D G Ll M P Rh T
Soft mutation    F G Dd   L F B R D
Nasal mutation M Ngh N Ng     Mh   Nh
Aspirate mutation   Ch         Ph   Th

For example:


but Pembrokeshire=Shire Pembroke=Sir Benfro 

 Welsh Personal Names.

 Welsh names took the general form A son of B son of C etc.

The Welsh for "son of" is "ap" and this may be preserved in the modern derivatives by combination or concatenation.

 For example Morgan son of Rhys would be

Morgan ap Rhys

Or in modern forms

Morgan Pryce

Morgan Preece

Morgan Price

Morgan Rice

Morgan Rees

Morgan Reece

Morgan Rhys

As an aside it may be noted that there was a tendency (tradition may be too strong a word) for children to be named after or in memory of, their grandparents. Indeed in some families to not do so would have been taken as a form of rebuff or insult. The typical rule would have been to name the first son after the paternal grandfather (fatherís father) and the first daughter after the maternal grandmother (motherís mother) the next son and daughter would then be named after the maternal grandfather and paternal grandmother respectively. 

The Welsh names list is fairly limited and hence repetitive; the Enumerators took advantage of this to abbreviate many entries.

Typical examples are:

Wm      (William)

Geo      (George)

Hy        (Henry)

Dd       (David)

Ja         (James)

Jo         (John)

Jno       (Johnathan)

Thos     (Thomas)

Archbd (Archibald)

 My      (Mary)

Please do not take the above examples as definitive - the My above may actually have been that particular Enumerators abbreviation for May. More important is that the researcher makes appropriate use of wildcard/joker characters when completing his search parameters. In other words it is unlikely that David Thomas William Jones is enumerated in that form - far more likely to be David Thos W Jones or similar. 



The Cenquest database is easily accessible on an intuitive basis but we do wish to emphasise the power and utility of using "*" the wildcard or joker character in order to maximize the probability of achieving a successful "hit".

The main reason for this in respect of personal names is that the Enumerator may not have recorded your ancestors name in a completely formal manner (in comparison with certificates of Birth, Marriage or Death).

 For example

James David Thomas may have been enumerated as 

James Dd Thomas or

James D. Thomas

Please note that wherever possible Cenquest has omitted punctuation (especially applies to placenames) in order to simplify searches.

Entering * Thomas will return approx 6870 returns

Entering James* Thomas will return approx 250 hits (note the search is not case sensitive)

Entering *James* Thomas will return around 267 hits including David James Thomas

Entering James David Thomas returns a single hit (from Dinas) but misses two possibles - (James D Thomas - one in Pembroke Dock and one in Llandilo)

Similarly when searching for female ancestors and unsure of the date of marriage or re-marriage (i.e. name change) then simply enter the surname/family name as *

For example when looking for somebody whose name may be Mary Anne Williams

Mary* * will give about 8692 hits



Mary Ann* * will return some 1069 hits

Mary Anne *will return around 257 hits

Mary Anne Williams will give around 10 hits (and some of them are clearly under marriageable age)


 and after pressing the (deep) view button for the person


 and after pressing the (deep) view button for the "family"


Please note that there are in fact 4 people in the family but you need to scroll down the screen to see her 5 year old son David George Williams who is a Welsh speaker.


Click here to get Adobe Reader

Right click and select 'save target as' to download the User Notes. (pdf)