Sometimes the best DNA discoveries are the ones that we did not intend to make.
Years ago I was able to show, using the records that I could find after completing a more than exhaustive search, that the Florence Ellen Butler who married Ira Sargent in 1870 in Union County, Iowa, was the daughter of Benjamin Butler who lived in that county in the same year. Florence evaporated in the 1880s and, given that she was born in Missouri or Iowa in the mid-1850s, the chance of a birth record was virtually zero. There was going to be no record clearly stating who her parents were.
Until I can locate a death certificate for Florence (and there is no guarantee of that given when she likely died), this may be all the evidence I am able to gather.
Benjamin Butler was born around 1820 in New York State or Canada. He and his family lived in Canada, Michigan, Iowa, and Missouri between the 1840s and his death in Vernon County, Missouri, in the early 1880s. No children are mentioned in his probate (other than the one his wife had shortly before he died) and he left no will. Benjamin’s first wife, and presumably Florence’s mother, was a woman named Margaret Stephens. She apparently died in the 1850s.
Benjamin, Margaret, and their children were living in St. Clair County, Michigan, in 1850. The very next household is headed by an Alonzo Stephens and contains a man named Rufus Stephens. Alonzo was of an age to be Margaret’s sibling and Rufus was old enough to be her father. The places of birth as given in the 1850 census for the Butler and Stephens children were consistent with the families moving from Canada to Michigan at about the same time. The proximity of the families in 1850, the common last name of Stephens, and a similar time of migration suggested the families were connected.
Alonzo is known to be a son of Rufus Stephens. But Rufus Stephens left no records easily connecting him to Margaret (Stephens) Butler.
Then a descendant of Alonzo showed up as being a DNA match for me on FamilyTreeDNA. The number of centimorgans that we share (43) is consistent with our paper genealogical relationship if Rufus is Margaret (Butler) Stephens’ father and Margaret and Alonzo were brother and sister. The test results were consistent with me being Rufus’ descendant.
None of the shared matches I have with this testee share significantly more centimorgans of DNA with me than the testee does. This was frustrating as it could have confirmed which branch of my family the match was through. Given the distance of this relationship it is not unusual that none of my closer relatives appeared on a list of matches with this person.
But it’s always possible that the testee and I are related in another way and it has nothing to do with Rufus. Possible, but not very probable given the shared matches we have and where my other families lived. It is also possible the testee and I are related through the Stephens family, but that the connection is further back than Rufus and that Rufus is Margaret’s uncle or her cousin. The DNA isn’t precise in predicting the specific relationship when it is this far back.
The most likely and most probable scenario is that Rufus is Margaret’s father.
But I need to write up my evidence and my conclusion.
And I need to look at my shared matches with this descendant of Rufus in more detail and see if we share matches on any other DNA sites.