This post is about identifying the father of my mothers paternal grandfather, my second great grandfather – what was his name? (52Ancestors#12).
I previously wrote about her grandfather Edward Roberts in 2018. The circumstances surrounding his birth have always been a mystery and whilst we knew his mother was married twice, we believed her second husband was his father, another Edward Roberts, even though the couple didn’t marry until about ten years after Edward juniors birth. For more information read my previous post here.
A miracle occurred during Covid as I once again revisited the question of Edwards birth. For over 15 years I have been searching for his birth record. I’ve purchased many certificates over the years for Edward Roberts’ and Edward Bakers in England and India. Low and behold a search of the GRO index in 2021 revealed an Edward Roberts Baker, born at Westerham Kent in 1869! Could this be my Edward? I couldn’t order it fast enough and the week I had to wait for it to arrive was agonising.
Good things come to those who wait as they say – I can’t help but be reminded of the old saying – was it ‘the butcher, the baker or the candlestick maker’!
It was definitely him, the mother stated as Annie Baker formerly Laundon. We had always thought his birthdate was 6 August 1868, but it was only slightly out, being recorded as 7 August 1869. The address of London Road Westerham, consistent with where Edward senior was living in 1871 and where the whole family resided by 1881. Why wasn’t it on the index for all these years, I had definitely checked it many, many times? I can only surmise that the recent digitisation program found it was missing from the index and corrected the error.
There were a few questions however. Ann named the father as her first husband Thomas BAKER, was he really Edwards father? Ann also used the name of Roberts as Edwards middle name – was this to suggest that her second husband Edward senior was the biological father, perhaps they had been together before 1869? Ann was still married to Thomas at the time, so did she name him as Edwards father for convention or respectability?
Where were they all in early December 1868? Based on what we can glean from the records Ann and her husband Thomas Baker were living in Westerham Kent by 1865 when their son Eldred was born. It is unclear when Thomas and Ann split up, but they are clearly living apart by 1871. Certainly, from the records it seems both Thomas and Edward were in the right location around the time of Edwards birth to be his father but which one was it?
|Year||Ann Laundon b1835||Thomas Baker b1823||Edward Roberts b1841|
|1857||Married Thomas Baker||Married Ann Laundon|
|1861||Army Barracks, Hampshire||Army Barracks, Hampshire||Westerham, Kent – lodger|
|1865||Eldred Baker born Westerham, Kent||Westerham, Kent (children 1858-1865)|
|1869||Edward Roberts born London Road, |
|Named as father of Edward Roberts Baker|
|1871||Croydon, Surrey, Mrs Annie Roberts. |
With sons Thomas, Eldred and Edward aged 3.
|Canterbury, Kent – Visitor to Barnes family.|
Listed as (Croydon) Army pensioner.
With daughter Alma.
|Westerham, Kent – lodger|
Lodge lane, near High Street
|1878||Married Edward Roberts at Croydon.|
Widow, residing at Norwood.
|Married Annie Baker at Croydon.|
Bachelor, residing at Norwood.
|1881||London Road, Westerham, Kent||Died before 1880 (Marriage of Alma)||London Road, Westerham, Kent|
The DNA story
I have been actively researching my mothers family since the early 2000’s and had been convinced from the records that Edward was a child of Edward Roberts senior and not Thomas Baker. Discovering the birth certificate finally in 2021 stating Thomas Baker was the father put a spanner in the works. I had researched both the lines of Thomas Baker and Edward Roberts senior. DNA has been a focus of my research since 2010, so what was it telling me?
We have many matches going back several generations confirming that Ann Laundon is our ancestor and clearly Edwards mother, but the paternal side is quite different.
Thomas Bakers family were also from Kent and nearby Dorset. Over many years I have scrutinised the extended family of his parents and compared them to the DNA test results of my mother, three of her siblings and a descendant of their fathers brother. We had no matches to anyone in this family. I revisited my research many times and concluded that the reason for this was that Thomas was probably not his father and that Edward Roberts senior seemed more likely.
Edwards Roberts’ family were from Gloucestershire. His father John Roberts was connected to a prodigious DAY line that according to a Family Search tree connects to Royalty! The maternal Tweed line was not as extensive but additional generations had been identified.
For several years it has concerned me that we had only identified one DNA match on this line, 15cMs at AncestryDNA. It was a sixth cousin relationship to my mother and her siblings, to their 5th great grandparents ancestors John BAYLIS and Mary CHURN.
In over a decade of DNA research no other matches have been found on either the Roberts or Tweed lines. At Rootstech 2022, the ‘find my relatives’ tool told me I had over 700 cousins at the event related to me on the Day line – with so many interested in genealogy surely many of them would have tested, why didn’t I have any matches? With no other matches to these ancestors at any DNA site (and we have tested at all of them) and without being about to view the segment data, I remained skeptical that the match was an indication of a true genetic connection.
Was my research wrong? Or, was it that the right people just hadn’t tested yet? Were there insufficient descendants?
- Edwards Roberts Baker b1869 only had three half siblings, all on his mothers side;
- Edwards Roberts Baker b1869 only had 2 sons;
- Edward Arthur Roberts b1893 (my grandfather, known as Ted) had 7 children and from DNA testing we have about 95% of his DNA;
- James George Roberts b1895 (my great uncle, known as Jimmy) had 3 children but only one had descendants;
- Edward Roberts senior and Thomas Baker were both one of seven children, surely there should be other descendants we should match?
For the last several years I have pondered whether Edwards father was not Thomas Baker or Edward Roberts, but perhaps someone else entirely?
A breakthrough came earlier this year when I managed to test an additional three second cousins on this line. We now had four grandchildren of James Roberts tested (the brother of my grandfather). Whilst Ted and Jimmy had the same parents they can inherit different segments of DNA from each of them due to a process called recombination. I was hopeful that Jimmy had inherited different segments to his brother that might give us more clues to the identity of their paternal grandfather. Combining the tests of all descendants we now have approximately 60% of Edward Roberts Bakers DNA – hopefully this will give us more clues for our research.
Over many years I have concentrated on trying to identify the origins of Mums other problematic paternal great grandfather Arthur George Courtney. I have a separate research blog for him – you can view the Research Summary here. As part of that process I have been laboriously identifying clusters and triangulated groups of DNA matches as part of the puzzle in putting the pieces back together to identify him. It was that work that helped me in also finding out more about Edward.
I waited with anticipation for the DNA results of my four second cousins at AncestryDNA.
When the results first came through, I must admit I was a little disappointed as there weren’t too many new matches that I didn’t know already. I have been extensively interrogating my mothers results at AncestryDNA and the other DNA sites for many years now.
However one family stood out as requiring further attention, the family of Edmund and Philadelphia DYE from Kent in England. We previously had a number of DNA matches to this family, but I had not been able to connect them to to my tree.
My cousins also matched this family, but even more closely than us!
DNA Matches are shown in lime green.
My mother and her brother had reasonably good sized matches with Cousin A at 50cMs and 82cMs, all three belonging to the same generation, but they only shared smaller amounts with the other testers – about 27-30cMs. This group had intrigued me for some time and I had wondered if they might fit into George’s genealogy somehow. Matches to Cousin A to the four second cousins (who are a generation more distant) were varied, one shared 90cMs, two approx 50cMs and the last only 15cMs.
It was matches to Cousin C that really caught my attention. My mother only shared 27cMs, yet matches to my four second cousins were quite different, they were 110cMs, 102cMs, 65cMs and 50cMs. Being a generation more distant it was unusual for them to share so much more DNA than my mother, but that is the magic of recombination.
It was these matches that spurred me on to do more research and in particular, to focus on the Edmund Dye and Charlotte Horsnell couple. I investigated the family of Charlotte Horsnell, it was quite tricky but eventually I discovered that Charlottes grandparents were James HODSON and Mary FOX.
These Hodson Fox names were quite familiar to me as I had already investigated another AncestryDNA cluster and a triangulated group on Chromosome 6, whose genealogy both led back to this couple.
As the couple were from Leicestershire I had felt that perhaps they might be distant relations on Edward Roberts’ mothers side, as Ann LAUNDON was born in Leicestershire. Was this telling me something different?
I needed to revisit my previous assumptions.
Again, DNA Matches are shown in lime green.
Given we had DNA matches back to Edmund Dye’s parents, as well as to Charlotte Horsnell’s grandparents, it seemed likely that our connection was via one of the children of this couple.
Studying the family of Edmund and Charlotte Dye more closely I found they may have had 4 or 5 children, with only one having descendants.
- Annie Dye was born in 1843 and married John Wilkins Usher in 1870 when she was 27 years of age. She died after a long illness aged 39 and does not appear to have had any children;
- Henry Dye’s birth was registered on 31 Oct 1844 at Sundridge with a birth date of 22 Sep1844. There are no further records for him. I have concluded that Henry is actually the same person as Edmund Dye;
- Edmund Dye appears in census records stated as being born about 1845 but unlike all the other children there is no birth registration record for him. There is a baptism on 5 Jan 1845 at Sundridge for Edmund. It seems likely that his birth was probably registered as Henry and he was later baptised as Edmund, being the eldest son. Edmund died aged 36 of consumption, unmarried, with no known issue;
- George Dye was born in 1848 and died in 1860, aged 12, no issue;
- Edith Charlotte Dye was born in 1864 and was known as ‘Lottie’. She married Edward Everden and had at least 2 children;
- There is a significant gap between the birth of George and Lottie, but there do not appear to be any other children. Edmund senior died in 1888 and only four children are mentioned in his obituary. By the time of Charlottes death in 1899 her estate passed to her only surviving child Edith Charlotte Everden.
As I studied the family I suddenly realised that they lived in Westerham Kent, where Edward Roberts was born in 1869. Staring me in the face was Edmund Dye who seemed exactly the right age to be the father of Edward Roberts – 24 at that time and in the right location! Based on the known genealogy he seems likely to be the lone suspect to be Edward’s father (except perhaps for Henry?)!
Was this my new hypothesis?
Life in Westerham, Kent
There is no evidence that Edward Roberts senior ever lived with Ann and her children until some time after 1871 and maybe even as late as 1878 when they married (presumably after Thomas Baker died). Ann confuses the issue even more by calling herself Mrs Roberts in the 1871 census and giving the child his name. What had happened to cause Thomas and Ann to split? Perhaps Edward senior took pity on Ann and decided to look after her. In 1871 he was lodging in Westerham, at Lodge Lane, close to the High Street and worked in the seed trade. Later in life, he clearly took on the role of stepfather to Ann’s children and whilst they did not have any other children together they remained married until his death in 1902.
Westerham is a town and civil parish in the Sevenoaks District of Kent, England. It is located 3.4 miles east of Oxted and 6 miles west of Sevenoaks, adjacent to the Kent border with both Greater London and Surrey. The Dye family were prominent tradespeople and lived nearby at Warren Farm Sundridge. In 1861 Edmund b1791 (m Philadelphia Yowell) was listed as managing the farm of 52 acres employing 5 men and 1 boy. In the same year his son Edmund b 1817 (m Charlotte Horsnell) was noted as being a master wheelwright and smith, managing premises in the High Street at Westerham also employing five men, a boy, plus his son Edmund b1845 aged 16 was also working there as a wheelwright.
The ironmongers shop in the High Street is in very close proximity to London Road where Edward Roberts was born in 1869 and where he continued to live until about 1885. By the time Edward emigrated to Australia in 1910 he was described as having the ability to grow anything, a fine horseman and a fair blacksmith. Where did he learn these skills? Clearly his stepfather was responsible for passing on his horticultural skills, Edward had worked as a gardener when living in England and spent some time in the army where he may have learnt to ride. Living in Westerham for most of his early life, the Roberts’ must have been aquainted with the Dye family and I can’t help but wonder whether somehow young Edward may have acquired his blacksmith skills there?
The hypothesis and further testing
The DNA evidence suggests Edward Roberts is descended from a child of Edmund and Charlotte Dye, however the evidence is not overwhelming. Edmund Dye 1844-1880 seems the most likely candidate however as he didn’t have any other children, we have no descendants we can ask to undertake further testing.
Edward Roberts’ father would have been born some time before 1850, possibly earlier. Edmund Dye is the right age and in the right location. The DNA evidence is reasonable given the timeframe with two AncestryDNA clusters and currently six ‘confirmed’ triangulated groups on chromosomes 1, 4, 6 and 21. Four of these groups connect back to the Dye/Yowell couple and two to the Hodson/Fox couple. Hopefully over time more evidence will emerge.
I was lucky enough to be able to obtain photographs of the Dye’s to compare to my family. The top row are our potential Dye ancestors and the bottom row my family, three photos of Edward Roberts b1869 and the last two his sons, Edward (Ted my grandfather) and his younger brother James (Jimmy). In my opinion, there seems to be a strong family resemblance – in particular the resemblance between Lottie and Jimmy seems noteworthy.
So are we safe in concluding that this our line? It seems probable, but I would like more genetic evidence. I’d probably say it is now my current working hypothesis.
I’d appreciate your thoughts!
Y-DNA and a possible Danish connection?
Y-DNA testing could help to confirm that our male line is the same as this Dye family. I have approached many descendants of the Dye family but have not yet identified anyone suitable and willing to test. If you are a male descendant of Edmund Dye b1791 (m Philadelphia Yowell) carrying the Dye surname, please contact me if you are willing to test to help us to confirm/refute our hypothesis.
Two descendants of Edward Roberts Baker have already undertaken Y-DNA testing but to date there are no matches via STR testing and only one Big-Y match, with the Big-Y match carrying the surname of Taylor. This Taylor line extends to colonial times in the US but does eventually get back to Suffolk in England by the 1600’s. Big Y testing has identified our haplogroup to be R-BY23391 (R-BY160942 in the block tree).
What is so unusual about our Y-DNA signature that there are no matches even at the lower STR levels?
Autosomal results suggest that we may have connections to a Dye family in the US, that extends back to Denmark, with name variations of Dey and Duyts.
These are all only AncestryDNA matches with no segment data, not triangulated groups, so they could be just co-incidence. It is unusual for autosomal testing to go back this far.
However one wonders whether this may explain the lack of Y-DNA matches. How prevalent is DNA testing in Denmark?
Could we be connected via another Danish emigrant who came to England?
Mum’s latest ethnicity suggests 7% from that region, so are we on the right track?
Do you know more about the families mentioned in this post? If you are connected to any of them (particularly if you have DNA tested, or are willing to test) I would love to hear from you. It’s the power of DNA that can help us breakthrough our brick walls!
Featured Image: Westerham c1831, W.H. Ireland’s View of Kent.
Postscript: View my presentation to the Society of Australian Genealogists on You Tube.