It's a very satisfying feeling when you break down a brick wall... even if all this one took was 14 years of effort and finally and belatedly, the application of the FAN principle.

I learned my great-grandfather's name early in my research. Following the trail laid down by my mother's birth certificate and her father's birth certificate led me to the names of her paternal grandparents: Thomas Jones and Agnes James.  It was easy enough to find their marriage in 1893 in Merthyr Tydfil, and so learn their fathers' names: Thomas Jones (again) and William James; and to find them in the censuses in 1901 and 1911 to get approximate ages and places of birth.

The James line was easily traced back to rural Pembrokeshire, but looking for the birth of a Thomas Jones son of Thomas Jones born in about 1870 in Dowlais, Merthyr Tydfil, South Wales was... well, frankly a nightmare. The more so, as Thomas Jones the elder was deceased by the time of his son's marriage, and I couldn't definitely identify Thomas Jones the younger in any of the censuses prior to his marriage.

I put the problem on the back burner after 2 years of trying and returned to it in 2011, when information from the the 1939 register became available, thanks to the efforts of Guy Etchells and others. It was hideously expensive, and it wasn't online -- you had to write to the NHS Information Centre and wait about a month -- but I thought it might be worthwhile as I knew my mother had been living with her grandparents at the start of World War 2. So I wrote a cheque and waited.

It was worth every penny. Not only did I get the birth date of Thomas Jones the younger (5 March 1870) but I also got the date of his death (15 June 1946).1 With his birth date, and the assistance of Merthyr Tydfil Register Office2 (who confirmed that there was only one Thomas Jones born in their registration district on that date) I finally got his birth certificate and both parents' names: Thomas Jones and Elizabeth Hughes. Progress!

But not very much progress.  As he was born in 1870, and I now had both parents' names, I should easily find him in the 1871 census.  But... No. No Thomas Jones of the right age born anywhere in Glamorgan with a father Thomas and a mother Elizabeth. And no candidate in 1881 either.

Well, that's stretching the truth a little. There was an Elizabeth Jones (widow, aged 28) with a son Thomas (aged 1, born Dowlais) in the home of her grandfather William Harries in Aberdare in 1871. But the location wasn't likely... And there was a widow Elizabeth Jones of the same age in Merthyr Tydfil in 1881 with a son Thomas aged 12.  Both tempting candidates, but there wasn't enough evidence to convince me. So I moved on to easier pickings elsewhere in my tree.

Until the end of last year, when I had some time on my hands and decided to give the problem another go (and do the things I should have done earlier, except I didn't know so much then.)

First: find a census entry for Thomas Jones the younger in 1891 (2 years before his marriage.)  At his marriage, he gave his address as 10 Rees Street, so I tried looking there (obvious, I know -- I have no excuses except inexperience).  He wasn't there; a Daniel Davies was living there with a wife Elizabeth and 4 children. However, next door at number 9 was lodging a Thomas Jones of the right age and possible (imprecise) birthplace.  Wrong occupation, but occupations change a lot.  Hmmm.

One of the witnesses at Thomas Jones' marriage was a Daniel Davies. What are the odds that this wasn't the same Daniel Davies?  I'd say they were low.  So I decided to investigate Daniel Davies and his family.

Ten years earlier, in 1881, Daniel had been living with a wife Eleanor. Perhaps Elizabeth and Eleanor were the same people and the result of a census enumerators error, although they didn't share a birthplace or an approximate year of birth.  I looked for the death of an Eleanor Davies and found a single possibility in 1883 (right district, right age at death). So I hyopthesised that Daniel remarried an Elizabeth between 1883 and 1891, but there were too many possibilities and no way to eliminate them without ordering certificates, which I wasn't prepared to do... yet.

By 1901, Daniel and Elizabeth had moved to an address that was quite close to where my mother was born, and 10 years later they were living one or possibly two doors away from Thomas Jones and his wife Agnes.  Thomas and Agnes had all their children with them except one: 10 year-old Mary Jones. I knew Mary was alive in 1946 when she registered the death of Thomas Jones.

Daniel and Elizabeth had a 10-year old Mary Jones (a 'niece') living with them. And the 1911 census helpfully told me they had been married 21 years, so gave an approximate date of 1889/1890.

The only possible marriage of a Daniel Davies to an Elizabeth in that time period was a marriage to: Elizabeth Jones. I sent for the certificate, faintly annoyed that the PDF 'certificate' service currently available from the GRO only covers births and deaths, but quietly confident that I'd finally made a breakthrough.

Wrong.  Elizabeth Jones was a widow, but her father was Daniel Williams (not Hughes). Back to the drawing board.

It was now time to investigate the possible Elizabeth Jones that I had found in the 1871 census. The census record said that she was the grand-daughter of the head of household William Harries but the ages didn't stack up and she couldn't possibly be a grand-daughter. Perhaps a daughter, by his current marriage or a previous marriage of himself or his wife Jane? Looking at the pattern of children's names and ages, it seemed that Jane's maiden name was Matthews and the couple had married in 1846.  So Elizabeth was unlikely to be a daughter of the couple (she was born about 1843) but she might be a step-daughter of one or the other -- William and Jane were both quite old for their marriage to have been their first.

If Elizabeth's maiden name was Hughes, she was unlikley to be the daughter of William Harries so I focussed on his wife Jane Matthews. There was no previous marriage for her that I could locate (and she had married aa a Matthews in 1846) so this theory didn't fit well at all.  I was beginning to think that this Elizabeth was a red herring.

Still, I needed to do more work to eliminate her from consideration. Given her age, I calculated the likely possible year of her marriage as 1860. There were (of course) lots of possibilities for the marriage of an Elizabeth to a Jones in South Wales between 1860 and 1871, but my eye was caught by the marriage of Thomas Jones to Elizabeth Hughes in 1862, and I sent for that certificate. (There were quite a few marriages of Thomas Jones to Elizabeth Hughes in the 30 years before Thomas the younger was born, but this finally looked promising).

More waiting.

When the certificate arrived, I discovered that this Elizabeth was the daughter of Hugh Hughes (deceased) and one of the witnesses at the marriage had been a Daniel Williams... And when I went looking for Elizabeth Hughes in the 1861 census in the area given as her address (Cwmbargoed), I found Eliza Hughes of the right age to match both this '1871' Elizabeth and the one that married Daniel Davies, living as the step-daughter of Daniel Williams with a sister Margaret Hughes and a mother Mary.  In 1851, Mary, Margaret and Eliza are living in Cwmbargoed with the head of household Hugh Hughes. There's a death record for Hugh Hughes in 1853.

Now the pieces were falling into place.

Elizabeth Hughes (daughter of Hugh Hughes and Mary) was born around 1844. Her mother was widowed in 1853 and remarried to Daniel Williams before 1861. In 1862, Elizabeth married Thomas Jones with Daniel Williams as a witness, naming Hugh Hughes as her father.

Some time between 1862 and 1890 (when Elizabeth remarried to Daniel Davies and named Daniel Williams as her father), her first husband Thomas Jones died.  In 1891, Thomas was living next door to Elizabeth and Daniel Davies, and they continued to be neighbours until at least the 1911 census.

I had identified Elizabeth Hughes, and had a father's name for her first husband Thomas, but I still wasn't certain that she was the Elizabeth Hughes in Aberdare in 1871.

Until I went looking for her mother's maiden name.


The same as Jane Matthews, wife of William Harries.

Jane was born about 1815 in Llanon, Carmarthenshire.

Mary was born about 1821 in Llanon, Carmarthenshire.

I sent for their marriage certificates. The father of both of them was Thomas Matthews, a farmer.

So in 1871, the widowed Elizabeth Hughes was staying with her one year old son in the household of her aunt Jane.

What about the widowed Elizabeth Jones in Merthyr Tydfil in 1881? Well, she was living at a house opposite the address given by Elizabeth Jones the widow who married Daniel Davies.

If anyone felt the ground shaking just after the New Year, that was me doing a happy dance.  I'd been waiting for it 14 years, so I'd definitely earned it.

1Although the 1939 Register is now online, it doesn't include the date of death as that was, I think, the result of a personalised look-up to see what information the NHS was allowed to provide.  Also, it wasn't necessary to provide proof of death to get an individuals' records opened -- the NHS checked whether they could disclose the data. For example, the letter I received from the NHS about my father's household has more people documented than the corresponding FindMyPast image.  Sometimes it's one step forwards and one step back when it comes to the provision of online records...

2 Local Register Offices are usually more helpful than the GRO at Southport, when it comes to reference checking (a service the GRO are supposedly no longer able to provide).

Jan 10, 2018 By ColeValleyGirl

Add new comment