New Ancestry Shared Match Tool Solves Mystery

I may have already solved a mystery with the new Ancestry Pro Tools shared matches feature. But I’m not done solving it, really.

I have a small cluster of DNA matches that stubbornly could not be identified. The new tool tells me that the oldest in that cluster, I’ll call him R, is a 2nd cousin (Ancestry’s guesstimate) to someone I have already identified and placed in my tree with lots of paper trail evidence. I’ll call her J.

That should mean that R and J share a set of great-grandparents. Both have trees that include that level, and they do not have any matches or ones that look like they are close. R has one great-grandparent from Norway, the rest from Sweden, and I know he is related via my Norwegian ancestors, so I have been focused on the brick wall that is his great-grandmother, who I’ll call M.

Then I did a rough sketch of a family tree for R and J, and looked at the four grandparents of each — if R and J are second cousins, then they should have grandparents who are siblings. I listed all four grandparents for each. And I noticed something I had not noticed before: J’s grandmother L has a sister named Mary Carolina Johnson, and R’s grandmother is named Mary K. Johnson. Carolina is often spelled Karolina in Scandinavia, and spelling differences like that can be very fluid. But it’s not as simple as just saying “Aha! They match!” I need to have paper trail evidence.

On all the trees that have R’s grandmother Mary K. Johnson, she is listed as coming from Sweden, and her cemetery marker has August 3, 1898 as her date of birth. Mary Carolina Johnson was born on August 3, 1895 in St. Paul Minnesota. The year is different, but that the month and day are the same is very suspicious. I had previously tried finding descendants for Mary Carolina, and had failed, marking her record “brick wall.” So it is still possible they are the same person, and that for some reason Mary K. took four years off her age and stuck with that story.

Unfortunately, here’s a major problem: In my tree for Mary K. , her parents are Frank Johnson and Tilda Maria Jonsson, both born in Sweden. In my tree for Mary Carolina, she has parents Martin Johnson and Marie Stein, and Martin is the son of my great-grandfather’s brother. Can I just assume that Mary K.’s parents are incorrect? Again, I check the paper trail.

I have found Mary Carolina in the 1920 census with her parents Martin and Marie. I note that one tree that has Mary K. in it, has her listed in the 1920 census with parents Frank and Mary (which is likely Tilda Maria, Americanizing her name as her husband obviously has). That would mean they cannot be the same person. Furthermore, the Mary K. record is in a smaller town north of Saint Paul, and Mary Carolina is in Saint Paul — and Mary K. goes on to marry and live in that smaller town, so that census record should make more sense.

So then I check out that 1920 census record listed as a source for Mary K. in someone’s tree. There is no daughter Mary born about 1898-1899 in that census or in the 1900 census living with father Frank and mother Mary/Tilda Marie! That is NOT valid source for Mary K. having those parents, and in fact makes it unlikely that those are her parents.

Now I am thoroughly convinced that Mary K. and Mary Carolina are the same person, but I still need more positive paper trail evidence. I don’t see a marriage record for Mary and the man listed in birth records of her children as the father, though that would still leave some doubt. So I am going to change this in my tree, and add a “hypothesis” tag.