DNA results not matching part of my tree

[PaulLB]

I have a problem with one line of my tree that I would be grateful for any advice in working out how to move forward.

One of my maternal Great Great Grandfathers should be Samuel Fairfield born 1842, who married Emma Lewsley born 1841. They were born and lived in Hull. Samuel’s father should be Frederick Fairfield born about 1809 in Hull according to the 1851 census, although no birth record exists anywhere in England for him.

I have done an Ancestry DNA test and have a second cousin through one of Samuel’s sons. I have also found matches through Emma Lewsley. I have found no matches through Frederick Fairfield, despite corresponding with 5 people who have DNA tested through Ancestry who should be my 4th cousins. I have one match (8cM) through Samuel, who should be my 3rd cousin once removed. I have contacted two people who should be related to me through Samuel and Emma, one of whom replied to say they hadn’t done a DNA test and the other didn’t reply. I have offered both of these people to buy them a test but neither replied to the offer.

I have 6 shared matches with my second cousin of 24-59cM and have tried contacting them. Two replied but had very little detail about their ancestry. The ones who didn’t reply have no published trees.

I suspect that Samuel might not be my relation. I have 9% Scandinavian ancestry according to Ancestry, all through my mother’s side but have found no link to Scandinavia anywhere in my tree. Given Hull was a busy port, I wondered whether Emma might have had a dalliance with a Scandinavian seaman! Any advice on what to do next would be much appreciated.

 

[peter]

The table in the DNA Masterclass shows that 4th cousins will only share DNA 71% of the time, so you can't rule out the possibility that Frederick Fairfield was Samuel's biological father. However you also seem to be questioning whether you are descended from Samuel. You say you have a match "through Samuel" but how do you know it isn't through his wife?

 

[PaulLB]

The table in the DNA Masterclass shows that 4th cousins will only share DNA 71% of the time, so you can't rule out the possibility that Frederick Fairfield was Samuel's biological father. However you also seem to be questioning whether you are descended from Samuel. You say you have a match "through Samuel" but how do you know it isn't through his wife?

You are right, I meant to say descended through Emma and Samuel and given the low DNA match it could well be that we are half 3rd cousin once removed. I calculate that there is a 0.02% chance that I am related to the 5 people who have DNA tested where we should be related through Frederick, based on 71% chance of sharing DNA with your 4th cousins, so it does look unlikely that the tree is correct.

 

[Pauline]

I calculate that there is a 0.02% chance that I am related to the 5 people who have DNA tested where we should be related through Frederick, based on 71% chance of sharing DNA with your 4th cousins, so it does look unlikely that the tree is correct.

Trouble is, it’s never that simple. If, as it perfectly possible, you yourself have inherited very little DNA from Frederick, then you are very unlikely to have matches with other descendants of his. I think you need more than simply not having a match with these 4th cousins before assuming your tree is wrong.

 

[peter]

Trouble is, it’s never that simple. If, as it perfectly possible, you yourself have inherited very little DNA from Frederick, then you are very unlikely to have matches with other descendants of his. I think you need more than simply not having a match with these 4th cousins before assuming your tree is wrong.

Absolutely correct. It would help if someone from an earlier generation who - on paper - is also descended from Samuel could test.

 

[PaulLB]

My second cousin is also not a match to any of Frederick’s descendants but I realise it still could be possible that my great grandfather Fairfield didn’t inherit much DNA from his grandfather Frederick.

 

[peter]

My second cousin is also not a match to any of Frederick’s descendants but I realise it still could be possible that my great grandfather Fairfield didn’t inherit much DNA from his grandfather Frederick.

Do you have access to your cousin's matches, or are you relying on Shared Matches? If you do have access to his matches then you can focus in on the matches you share (not just your 'shared matches') - they will all be connected through your great-grandparents, though not in equal proportions.

 

[PaulLB]

Unfortunately, I haven’t been able to get a reply from her. She is quite old and hasn’t logged on to Ancestry for a long time. I found out that she didn’t share DNA with Frederick’s relatives from the relatives rather than her. It’s quite frustrating that I can't connect to her or any of the people we share DNA with but I guess that is pretty normal for Ancestry as most of the people who do DNA tests don't do it because of an interest in family history.

 

[peter]

Unfortunately, I haven’t been able to get a reply from her. She is quite old and hasn’t logged on to Ancestry for a long time. I found out that she didn’t share DNA with Frederick’s relatives from the relatives rather than her. It’s quite frustrating that I can't connect to her or any of the people we share DNA with but I guess that is pretty normal for Ancestry as most of the people who do DNA tests don't do it because of an interest in family history.

It sounds as if you know who she is, so it's probably worth trying to contact her some other way.

 

[JimP]

My DNA test had a similar result. My great-grandfather, Charles Boyle, was an only child, but he had 12 first cousins, and I had traced 21 children of those first cousins. One was a nun, but most had 2 or 3 children. Ancestry gave me zero matches. One the other hand, I had a cluster of matches at the 2d-3rd cousin level whom I could not place. Working from their trees, I identified a family that lived near where my grandmother was born, and a son in that family who was the likely source of the sperm.

I reached out to several matches from that family, and none of them responded. The only two that I heard from contacted me first, because I showed up as a close match whom they could not place.

Unfortunately, the putative father had only one legitimate child, a daughter, and as far as I can determine she never had children. So there is no descendant of his that I can test to confirm my theory.

 

[peter]

So there is no descendant of his that I can test to confirm my theory.

You would only need to test a descendant if there are other plausible candidates. If there were no brothers, or if you have been able to eliminate the brothers because the amounts DNA you share with their descendants are incompatible with them being the father, then there may well be only one candidate.

Of course, all this depends upon whether you have stress-tested the hypothesis - this involves constructing the family tree as you believe it should be, then confirming that every DNA match with the extended family fits perfectly by comparing the shared DNA against the ranges in the coloured chart in the Masterclass (be careful to distinguish between half-cousins and full cousins). If even one doesn't fit, or if several are right on the boundary, you'll need to re-evaluate your assumptions.

A common error is to assume that the father must have been single at the time of conception - if he was already married that would be a perfect explanation for why he didn't marry your great-great grandmother. Was she still single, or was she married - and if so, how long had she been married?

 

[JimP]

Peter,
I have faithfully followed the instructions in the Masterclass. All the matches who have trees, or whom I can place because they appear on other trees, fit within the predicted ranges.

My suspect, Lewis, had 3 brothers who lived to adulthood, two of whom don't seem to have any descendants who have tested on Ancestry, so I can't rule him out. The other brother would have been 13 years old at the time of my grandmother's conception, but his descendants are all just under the averages for half-great-aunt and half first cousins once removed, and well above the averages (but within range) for full 1C2R and 2C1R.

I also had to consider the father of Lewis, Stephen, as he continued to sire children, by his second wife, 12 years after my grandmother was born. My relationships with identified ancestors of Stephen's siblings, who would be 3rd cousins or 3rd cousins once removed, are, with one exception, below average. In this scenario, the descendants of Lewis' siblings, now classified as half 1C1R and half 2C, all fall well above the average.

My g-g-grandmother was married at the time, and had an older daughter (presumably) with her husband. They separated sometime between my grandmother's birth in 1915 and the 1920 Census, when Charles was living with his widower father in Wilmington, Delaware, and Myrtle was boarding in Marcus Hook, Pennsylvania and working in a mill. Lewis did not marry until after he returned from fighting in France during the war; both his brothers were younger. Lewis' oldest brother Jacob (for whose descendants I have no identified DNA matches) was married before 1914. So I have leaned toward the single one, who was also the one more close in age to Myrtle.

Based on the DNA matches, the father is either Lewis or Jacob. But without a direct descendant of one of them to test against, I cannot be 100% certain.

 

[peter]

JimP. you have talked about shared DNA being above average or below average, but you haven't said that it was out of range - so I'm not sure you should be eliminating Stephen from consideration, based on what you have told us. Of course, if Stephen is the father you won't have any matches with cousins of his wife (ie the mother of Lewis and Jacob). It's easy to check, but it needs doing.