Both of my Smith great-grandparents have been a pain, genealogically speaking. I spent years thinking that Fannie's maiden name was Cotton, then I thought it was Mansell, then I found out that it was probably neither. Her paternal line is a mystery because I don't know who her father was--and may never know. I have a few more clues with her maternal line, enough that I may someday figure out who her grandparents were. At least, through the years, there has been some movement on her side of the family.
Not so with Stephen. I know that his mother was named Mary E. and that from 1850 on, she was the single parent of Stephen, his sister Cynthia, and his brothers John, Alexander Jackson, Minor Jefferson, and Moses Calvin. I know a little about Stephen's siblings and their future lives. I long suspected that Mary's maiden name was Williams, and that finally was confirmed by the death certificate of her son Alexander Jackson. I know that Stephen's father was probably named John Smith. And that's it. I don't even know for sure where John and Mary came from before they came to Alabama. In fact, I don't have solid evidence that John ever lived in Alabama.
|1850 Coffee Co. AL census|
Mary E. Smith and children: Synthia A., John, Jackson A.,
Minor J., and Stephen A.
Then in the last month it seems like every genealogical clue I've been given is about the Smiths. First, I got an email from a Perkins descendant. In my post, "Will the Real Ancestor Please Stand Up?" I talked about the fact that my brother's y-DNA test results show matches, not to Smiths, but to men named Banks and Perkins. (Not only did John Smith have to have a difficult paper trail, he had to have difficult DNA, too.) She related that she had also corresponded with a Banks descendant; all three of us were looking for a connection among our Smith, Perkins, and Banks ancestors.
James Washington Perkins, the oldest known ancestor of the Perkins descendants, was born in Georgia and died in Texas. My Smith family, as far as I know, never lived in those states. The Banks descendant that matches my brother on 37 markers was from Bulloch County, Georgia, where his family has lived for 8 generations. After having their y-DNA tested and results compared, several Banks descendants believe their oldest known ancestor is Charles Banks of Edgefield County, South Carolina.
John Smith was supposed to have been born in Virginia, but I always wondered where he met his wife Mary. On several censuses Mary E. Williams Smith reported her place of birth as South Carolina. A little research on the Perkins family revealed that they also had a connection to South Carolina. The grandparents of James Washington Perkins, James and Rebecca (Corley) Smelley, were from Edgefield County, South Carolina. So I might not have found a person that ties the Banks, Perkins, and Smith families together, but I might have found a place. At the very least, I thought I could narrow my search for Williams and Smith families to Edgefield County, South Carolina, where there are several to choose from.
Next, out of the blue, I got a message on Ancestry.com from a new match who turned out to be a 2nd cousin on my Smith side. His grandmother, Barbara Smith, was the sister of my grandfather Weaver. I asked him to upload his results to Gedmatch, which he did in short order, and I now have a Smith family member to compare matches against. It's also really interesting to compare our DNA on Gedmatch. He is now my largest match, next to my brother, with a whopping 312.8 cM's. He has even more in common with my brother--389.6 cM's.
Using the Triangulation feature on Gedmatch, I found a list of people that match both my Smith cousin and me. I contacted a few of them and heard from one--let's call him Mr. F. I actually realized later that he had once contacted me but we couldn't find the ancestor we had in common. This time I had just researched the Perkins family, so my email to him asked if he had an ancestor by that name. He did; his Ann Perkins (or Parkins), born 1800, married William Orr. Their daughter Rachel Orr was his great-great-grandmother. I compared Mr. F. and several of his relatives on the FTDNA Chromosome Browser and found that they all matched in a certain segment of Chromosome 7. Then I switched over to Gedmatch and compared my Smith cousin to them. (He tested on Ancestry.com so is available on Gedmatch but not FTDNA.) And guess what? He lines up in a great big segment at the same place on Chromosome 7 as Mr. F. and his relatives.
Then within a week--I am serious--I got another message on Ancestry.com, this time from a lady who just wanted to help me with my great-grandmother Fannie's maiden name. I explained that even though Fannie listed Mansil as her last name on her marriage license, she couldn't really be the daughter of John Mansell. I was curious if she was a relative, so I asked why she had been looking at my tree.
|Fannie and Stephen's marriage license|
It turns out that she also has Smith ancestors in Coffee Co., AL. We determined that we really are not related--her family has had y-DNA testing done and they really are descended from a known Smith. But what she told me next kindof rocked my world. Her ancestors in Coffee Co., who were next-door neighbors of my Mary E. Williams Smith, were Prescotts and Donaldsons--and they moved there from Edgefield Co., SC, along with several other families, including the Williamses.
|1860 Coffee Co. census|
Mary E. living near Prescotts, Williamses, and Donaldsons
In light of all this new information, I am completely revisiting what I think I know about John A. Smith. The death certificate of Andrew Jackson Smith, Stephen's brother, lists his parents as John Smith and Mary E. Williams. So at least as far as Andrew Jackson knew, his father's name was John Smith. I think the erroneous information starts with a marriage license issued in Chesterfield, Virginia, in 1828 for a John A. Smith and an Elizabeth Williams. I think people assumed that this was our John Smith and that he came from Virginia. A lot of people on Ancestry.com, and even genealogists in my own family, have listed his birthdate as 1805, but it couldn't be if he is the John Smith in Chesterfield, VA, who was listed on the 1820 census. He would only have been 15 years old.
In 1840 there is a John A. Smith in Pike Co., AL, whose census information has been attached to my John Smith. But again, he couldn't be my John Smith. He has way too many children, and I hate to say, I didn't even question this, but--the Smiths didn't live in Pike Co. in 1840. They didn't move there until sometime between 1860 and 1880. He also is supposed to be buried in the same place as Mary--in Pike Co.--but again he wasn't living in Pike Co. around 1850 when he was supposed to have died.
For this reason, one researcher on Ancestry.com thinks that John Smith, the husband of Mary E. and father of Stephen, is the completely different, and much older John W. Smith, who is listed on the 1850 Census Mortality Schedule with a death date of February 1850 at the age of 87. She has compared the census records of two John A. Smiths in Coffee Co. and concluded that neither of them could be the father of my Smith family. I don't know if I agree with all her conclusions, but she has certainly given me something to think about. It just makes sense to question everything when years of research have gotten you no closer to an answer. What do I know? John Smith may have really been born a Perkins.