Wales Surname Distribution (National Institute)

From FamilySearch Wiki
Jump to navigation Jump to search
National Institute for Genealogical StudiesNational Institute for Genealogical Studies.gif

The original content for this article was contributed by The National Institute for Genealogical Studies in June 2012. It is an excerpt from their course English - Understanding Names in Genealogy  by Dr. Penelope Christensen. The Institute offers over 200 comprehensive genealogy courses for a fee ($).

A Study in Welsh Surname Distribution[edit | edit source]

The Surname - Wales[edit | edit source]

Although there are relatively few surnames in Wales, different areas of Wales have distinctive groupings of these surnames. This fact has been demonstrated by Rowlands and Rowlands (The Surnames of Wales for Family Historians and Others, 1996), using all surnames from the marriage registers from 1813-1837. Their technique assists the researcher in pinpointing the exact area when the only knowledge is that ‘he was from Wales.’  Even the name Jones is comparatively scarce in Pembrokeshire and some parts of Glamorgan, but extremely common in parts of North Wales.

The surnames were analyzed by the 89 old administrative subdivisions of the counties called hundreds, roughly 90 percent of those marrying in each hundred giving their residence as within that hundred. 253 surnames were ranked into six categories (shown below), with several thousands of others occurring in numbers too small to be useful.

Results for Single Surnames

* Incidence percentage relates to total population in all of Wales except where noted.

*Group and Description
Surnames: Alphabetical Order
I. Found throughout Wales. Over 2.5%
Davies, Evans, Griffiths, Hughes, Jones, Lewis, Morgan, Roberts, Thomas, Williams
II. Only occasionally absent. 0.7-2.5%
Edwards, James, Lloyd, Morris, Owen, Parry, Phillips, Powell, Price, Rees, Richards
III. Strong regional presence and absent in some areas. Over 0.3%
Bevan, Bowen, David, Ellis, Harris, Hopkins, Howells, Humphreys, Jenkins, John, Llewelyn, Pritchard, Pugh, Rowlands, Watkins
IV. Strong local or regional presence (over 0.7%) but lower overall incidence
Allen, Arthur, Ashton, Austin, Bennett, Beynon, Breese, Brown, Daniel, Edmunds, Foulkes, Francies, George, Harry, Havard, Herbert, Matthews, Prosser, Prothero, Rogers, Smith, Stephens, Vaughan
V. Purely local, but significant, presence in only one hundred.
57 surnames
VI. Purely local but small concentrations
137 surnames

The Rowlands present maps showing the distribution of 26 surnames where the striking individuality of each is apparent, and a glossary of over 250 Welsh surnames includes derivations, meanings and incidence.

Overall 13.84 percent of the population were named Jones. There is a concentration in north Wales, especially in Penllyn hundred, Merionethshire (over 30 percent) but in parts of the south it is as low as 1 percent.
With an average incidence of 7.09 percent, there is a definite concentration in Cardiganshire, particularly Moyddin hundred.
There is a significant presence only in Glamorgan and the Usk area of Monmouthshire.
Whilst almost absent in north and central Wales, the name has greatest concentration in Pembrokeshire and Glamorgan.

These results are interesting in themselves and may assist in defining likely areas from which those having unusual surnames emanate. However for the large number of folk having common Welsh names a more sophisticated approach is necessary. This has been done by statistically combining the computed incidences of the names to come up with an area where they all occur. The Rowlands have propounded an hypothesis.

Hypothesis: If a connected group of Welsh people (giving at least two surnames) move away from their place of origin, it should be possible to predict that place of origin using a knowledge of the incidence and distribution of all names across Wales.

The Rowlands feel that their survey is relevant for families over the period 1780-1880, and thus would be of great use to many whose ancestors emigrated during that era. In 80 percent of the tests of their hypotheses the place of origin has been successfully narrowed to 10 hundreds, out of a total of 89. It is clear that the more surnames associated in the group, then the more precise will be the estimate of their place of origin. The Rowlands’ offer a limited service of this type to researchers stuck with the ‘they came from Wales’ problem. Write with an SAE or two IRC’s to:

John and Sheila Rowlands
P.O. Box 37, Aberystwyth, Ceredigion, Wales SY23 2WL

Once such an analysis has been carried out, the researcher could then concentrate their efforts on finding the family group in the pinpointed area, in such records as the census prior to emigration, estate records as well as church and chapel registers.

The Surname—Father Link[edit | edit source]

Studies in human genetics show that the Y chromosome is inherited through the male line, as are most surnames. Theoretically, if everything were perfect in society, the surname would indicate the biological paternity. We know that this is not always the case. A previous section on surname changes indicated the reasons for breaks in the continuous descent of a surname through the male line. In addition Brook (Does It Run in the Family? Links between Genealogy and Genetics, Journal of One-Name Studies Vol 7 #6, pages 6-11) has indicated that 3-14 percent of children are not the biological offspring of the mother’s husband.

For those intent on absolute veracity of the biological line modern research is beginning to provide suitable tests, for example Y chromosome and mitochondrial DNA testing. These services are likely to become more widespread and affordable. Individual studies, such as the proof of USA President Jefferson fathering an illegitimate child by his slave (Foster et al.), and the Sykes surname study (Sykes and Irven) have been widely discussed. At the time of this writing a huge DNA data base is being amassed by researchers from Brigham Young University. Genealogists are being invited to submit a complete four-generation pedigree plus a sample of blood at designated workshops all over North America.

Lasker et al. have pursued research in the DNA/surname area and have formed surname distribution maps which could also, therefore be considered as gene maps. Members of the Guild of One-Name Studies participated in producing an Atlas of British Surnames (Lasker and Mascie-Taylor).

No Surname[edit | edit source]

Sometimes you find people with no surname, usually because of an error of omission or illegibility by the clerk, or because the document has suffered the ravages of time and the surname has been lost to damp, mice, or what have you. In parish registers, provided the family has stayed put, it may be possible to go through and do family reconstructions for, say, all the families headed by a John and Mary, to find the one most likely to be having a child at this time. This could then provide one piece of negative proof for the linkage; two others are needed but at least you now have a theory to test against other evidence.


Information in this Wiki page is excerpted from the online course English - Understanding Names in Genealogy offered by The National Institute for Genealogical Studies. To learn more about this course or other courses available from the Institute, see our website. We can be contacted at

We welcome updates and additions to this Wiki page.