Using X-DNA to Identify an Unknown Ancestor: Nancy Gray and James S. Harris

Recently I was able to confirm a long-held ancestral hypothesis through the use of X-DNA, the DNA of the X-chromosome. X-DNA is one of the more obscure elements in the DNA genealogist’s toolkit, which I’ve seldom used, and even more seldom used correctly. This article will detail my methods and argument for how I’ve arrived at the conclusion that Nancy Harris (born about 1824 in Alabama), the wife of James S. Harris, was by birth a Gray, and James’s cousin.

Background: Nancy Harris

From the beginning of my genealogical research, Nancy Harris, the wife of James S. Harris, has been an unknown. She was my third great-grandmother, my grandmother Sarah Cook Richardson‘s great-grandmother, and frustratingly, we knew the maiden name of James S. Harris’s mother, Priscilla Gray, but not the maiden name of his wife, Nancy.

Nancy Harris was born about 1824 in Alabama, some fourteen years younger than her husband James, who was born about 1810 in Georgia. They both have been a rather obscure couple, from lack of records and loss of any known burial site. Like so many others, they both last appeared on the 1880 census, and then by 1900, were gone without a trace.

To me as a historian, 1880 seems like not that long ago, only a hundred years before I was born. I’ve handled many documents that were much older, seen many tombstones from long before. But 1880 is nonetheless just outside living memory. Both my grandmother’s Harris grandparents died in the decade before she was born; and James and Nancy Harris died in the decade or so before her mother, Ada Priscilla (Harris) Cook, was born.

Maddeningly, James and Nancy Harris somehow managed to escape having their marriage recorded. By the birthdates of their children, we can guess that they would have been married about 1843 or 1844 in Alabama — their oldest son, Alonzo, was born about 1844. Though the Harrises were from Georgia, James S. Harris appears in voter registration records in Tallassee, Alabama, in 1836, and his father Hardy Harris is documented in Alabama as early as 1826, receiving a U.S. land patent for land he had possibly occupied as early as 1821. With Nancy having been born in Alabama about 1824 — which her census records all consistently state — there is little doubt that they would have been married somewhere in the vicinity of Autauga or Coosa counties in Alabama. Yet no marriage record has been discovered anywhere.

The Red Herring

Lo and behold, to fill the void where Nancy’s missing marriage record and unknown maiden name should go, Ancestry’s little green-leaf hints discovered a specious record that looks, on its surface, to fit, but upon even a moment’s consideration, does not:

James Harris to Sarah N. Andrews, 26 Jul 1827, Hancock County, Georgia.

“Georgia, County Marriages, 1785-1950”, database, FamilySearch (https://www.familysearch.org/ark:/61903/1:1:KXJ2-FQK : Sat Oct 21 14:56:57 UTC 2023), Entry for James Harris and Sarah N. Andrews, 26 Jul 1827.

Nancy Harris would only have been about three years old in 1827; James only about seventeen. As shown above, Nancy was born in Alabama about 1824, and James was certainly in Alabama by 1826. James regularly used the middle initial “S” in his official records, the few that we have. This record clearly does not apply to this couple. Yet the vast majority of trees on Ancestry connect this “hint,” identifying Nancy as “Sarah Nancy Andrews” without even a second thought, and even connecting her to parents. This has been a major stumbling block to uncovering Nancy’s true identity.

James S. Harris’s True First Marriage

I suspected for a long time that Nancy Harris might not be James S. Harris’s first wife, on account of the wide age difference. Recently, I discovered records of James S. Harris’s true first marriage — not to “Sarah N. Andrews,” but to Sarah Ann Bull:

James S. Harris to Sarah Ann Bull, 28 Dec 1838, Talladega County, Alabama.

“Alabama County Marriages, 1809-1950”, database, FamilySearch (https://www.familysearch.org/ark:/61903/1:1:29FL-8FF : Fri Oct 20 04:36:39 UTC 2023), Entry for James S Harriss and Sarah Ann Bull, 28 Dec 1838.

In the past, I had probably seen and passed over this marriage, since at least in my mind, I consider Talladega County part of “North Alabama,” even though it is, to actually look at a map, only about 7 miles from where Hardy Harris was living in Coosa County in 1850. Hatchet Creek, near which Hardy Harris’s land was, actually flowed into Talladega County, the part that is now in Clay County (created in 1866).

We had been wondering who William H. Harris (1840–1864) was, the child who lived with Hardy and Priscilla Harris on the 1850 and 1860 censuses. It was my dad who discovered the probate record in Autauga County:

The State of Alabama
Autauga County

Know all men by these presents that we James S. Harris and Jesse Gray are held firmly bound unto Henly Brown Judge of the County Orphans Court … in the sum of three thousand dollars … dated this 19th day of February 1844.

The condition of the above obligation is such that … the above bound James S. Harris has been duly appointed guardian of all and singular the good and chattels, rights and credits of William H. Harris and Sarah Harris, minor heirs of Sarah Ann Harris, deceased, who was the heir of William C. Bull, deceased.

“Alabama Probate Records, 1809-1985.” Images. FamilySearch. https://www.familysearch.org/ark:/61903/3:1:33S7-9B3P-SH6 : 17 November 2023. County courthouses, Alabama. Autauga County, Alabama, Bonds Book F (1842-1869), 29.

William C. Bull was the father of Sarah Ann (Bull) Harris. Sarah’s children, William and Sarah, were due to inherit from their grandfather’s estate, so James was appointed guardian to represent their interests in court. So James married Sarah in December 1838; William was born 27 Mar 1840. James appeared on the 1840 census in Talladega County, living near the Bulls:

James S. Harris
2 white males between ages 30 and 40.
1 white male under age 5.
2 white females between ages 20 and 30.

“United States Census, 1840”, database, FamilySearch (https://www.familysearch.org/ark:/61903/1:1:XHB4-5QR : Tue Oct 03 23:18:14 UTC 2023), Entry for James S Harris, 1840.

The daughter Sarah was probably born about 1841–1842. Sarah Ann the wife probably died about 1842–1843. And then in Feb 1844, James entered into this bond for the guardianship of his children. The fact that he received the bond in Autauga County suggests that Sarah Ann had died there; that James had probably moved back to Autauga County between 1840 and 1844.

The Gray Hypothesis

I’ve known for many years that James S. Harris’s mother, Priscilla, had the maiden name of Gray. But for most of that time, I knew nothing more about the Gray family. It wasn’t until we tested my grandmother’s DNA on Ancestry.com, and I got in touch with Gail Bice Crosthwait, a cousin on both Gray and Oates, that I discovered the connection to George Gray (born circa 1759 in South Carolina).

I also discovered very quickly that Grandmother had a lot of Gray DNA. It seemed to be the most prominent element of her maternal side.

The first time the idea coalesced that Nancy, James S. Harris’s wife, might be a Gray, was in examining James and Nancy’s daughter Naomi (Harris) Akins (1849–1940). She lived to be ninety years old, outliving two husbands and all of her siblings. But at the end of her life, it wasn’t members of the Harris family who took her in, but members of the Gray family.

For the 1910 census, as a widow, age fifty-seven, Naomi was living with James W. and Martha R. (Gray) Wingard. In 1930, at age seventy-eight, she was still living with the Wingards. On both censuses, her relation was given as “cousin.”

When Naomi died in 1940, it was still Grays she was living with. She died at the home of W. A. Gray, “her cousin.”

Funeral Services Held At Coosa River Church For Mrs. Naomi Akins

Mrs. Naomi Akins died Thursday morning, January 18, at the home of her cousin, W. A. Gray, of the Coosa River community.

Mrs. Akins, a lifelong resident of Elmore County, was born July 20, 1849. She was the daughter of Mr. and Mrs. William Harris.

Funeral services were held at 2 p. m Friday, January 19, at Coosa River Church, the Rev. R. F. Hyde officiating.

Burial was in the Coosa River cemetery, McMorris-Harden in charge. Surviving are her cousin, W. A. Gray, and several nieces.

The Wetumpka Herald, Thu 25 Jan 1940

This seemed odd to me. Naomi’s connection to the Gray family was not exactly immediate. Her grandmother was a Gray. These people were the children and grandchildren of Jeremiah M. Gray, son of John Gray, who was the brother of Priscilla (Gray) Harris. So Naomi would be at best a second cousin.

Maybe there was no one else to take her in. But she did have Harris nieces and nephews living. Even if these Grays took her in as a charity case, why call her a “cousin”? In most cases, lacking a close family connection, such a person would have been called a “boarder”. Why adopt her as a close relative — unless she was in fact a close relative?

What if Nancy were a Gray? Somehow closer to these Grays than it appeared? It would explain both this “cousin” connection and Grandmother’s extraordinary amount of Gray DNA.

In terms of the life of James S. Harris, it makes sense. His wife had just died and he had two young children to take care of. He needed to marry quickly. So he moved back home and married the closest woman available, his cousin. Jesse Gray, James’s uncle, was helping him deal with the legal issues surrounding his wife’s death (see above). What if he were involved in the more practical issue of finding a wife?

Alas, the documents yielded no more evidence. Nancy Harris was not listed among the heirs of John Gray, who had a will. Or among the heirs of any other Gray who had a will. Neither Naomi, nor any of her siblings, had their mother’s maiden name given in their death certificates. In fact, by the time Naomi died, they couldn’t even remember her father’s name. But that would bolster the supposition, at least, that her father was not the source of her connection to the Grays.

The Weight of DNA

As mentioned above, it became apparent very quickly from my grandmother’s AncestryDNA results that she had a lot of Gray — more than she should.

To cite just a few examples of this:

• Sarah Cook Richardson to S.B., 135 cM on 5 segments.
  George Gray > Jesse Gray > William Mullins Gray.
  Paper-trail relationship: 4th Cousin 1x Removed.
  Expected cM Range: 0–117 cM (28 cM average).
• Sarah Cook Richardson to A.L., 105 cM on 5 segments.
  George Gray > Jesse Gray > Narcissa Gray Tennyson.
  Paper-trail relationship: 4th Cousin.
  Expected cM range: 0–127 cM (35 cM average).
• Sarah Cook Richardson to B.A.S., 70 cM on 5 segments.
  George Gray > John Gray > Jeremiah Mullins Gray.
  Paper-trail relationship: 4th Cousin 2x Removed.
  Expected cM range: 0–109 cM (22 cM average).
  

In all these cases, her match was clearly far above average for the documented relationship; in at least one, the first, it was completely out of range of observed values. But basing an argument for a closer-than-we-thought relationship to the Grays was complicated by the fact that everybody in Autauga County is kin to everybody else. In that tight-knit community of Liberty or Pine Flat where Grandmother’s family has lived for almost 200 years, descendants today may have more than one family connection to each other, and may or may not have properly documented it in their trees. It was difficult to find, for certain, Grays who weren’t also kin to the Cooks or Oates.

Earlier this year, I began an effort in earnest to fully map my grandparents’ chromosomes, beginning with a visual phasing project on my dad and his two brothers. I hoped that this would help me untangle the Grays from everybody else. I wanted to confirm, among other things, that Grandmother’s paternal side wasn’t somehow related to her maternal side, and that we didn’t have Gray on the paternal side or Oates on the maternal side where we weren’t expecting it. I also hoped to make some progress in determining if Grandmother had Gray from more than one source.

I made a lot of progress. And though I’m not finished, I started to notice several things:

• The great majority of Grandmother’s Gray matches were to descendants of Jesse Gray or John Gray — three or four times as many to these two as to other children of George Gray, and two or three times as strong in terms of centiMorgan range.
• Nearly every DNA segment that could be identified with Eli Harris appeared to have Gray matches in common. Only matches that appeared to have deep Harris (ancestors of Hardy Harris) has no obvious Gray matches.
• Finally, I began to see matches to a Mullins family, intersecting Gray matches, with some evidence of triangulation. That is, multiple matches to Mullins, occurring on the same segment, or in common with other people descended from Mullins.

The Mullins Connection

I had long seen the name Mullins listed in Gray family trees as a middle name — John Gray having a son, Jeremiah M. Gray, who was supposedly Jeremiah Mullins Gray, and Jesse Gray having a son William Mullins Gray. I noticed in the trees of several of Grandmother’s matches, common descent from a Jeremiah Mullins, who died in Jones County, Georgia, about 1836 — the area of Jones County, Baldwin County, and Hancock County where the Grays and Harrises were said to have come from. Jeremiah Mullins Gray — Jeremiah Mullins — it set off alarm bells. On a hunch, I immediately went to look for a will for this Jeremiah Mullins.

I found a will. And in it, I found exactly what I was hoping to find: evidence that Jeremiah Mullins was the father-in-law of John Gray — hence the naming of John’s son Jeremiah Mullins Gray. More than that, I discovered that Jeremiah Mullins had two daughters who married Grays:

Item 2nd. It is my wish and desire that all the ballance of my estate both real & personal be divided into seven equal distributive shares, viz., between my sons Levi Mullins and Pleasant J. Mullins and my daughters Phoeba Burgy, Nancy Gray, Martha Gray, and Elizabeth Lester, allowing one distributive share to the children of my deceased daughter Clary Ann Herrington to be divided among them in common …

“Georgia Probate Records, 1742-1990,” database with images, FamilySearch (https://familysearch.org/ark:/61903/3:1:3QS7-993B-YRDB : 20 May 2014), Jones > Wills 1809-1864 vol A-D > image 164 of 387; citing Jones County, Georgia, probate court.

It Comes Together

So John Gray married a Mullins — who else? Jesse, of course. Jesse also named children with “Mullins” as a middle name — and this would also explain the strong DNA connections between the descendants of John and Jesse: they were double cousins, sharing both sets of their grandparents and so having twice as much DNA in common as normal first cousins.

According to the research of Gail Bice Crosthwait, John Gray married Nancy Mullins and Jesse Gray married Martha Mullins. These marriages probably would have taken place in Hancock County, Georgia, between 1802 and 1806, and just slipped between the cracks of what marriage records are extant: the earliest extant for either Hancock County or Baldwin County (created 1803) being 1806. Jones County was created in December 1807.

The strongest DNA matches being to descendants of John Gray, and the “cousins” who claimed Naomi Harris being grandchildren of John, the obvious inclination was to attach Nancy Harris to John Gray. But John Gray died in Elmore County, Alabama, about 1870, and had a will that names his children.

So that leaves Jesse. Could Nancy have been a daughter of Jesse Gray? Jessie did not have a will. For the 1850 census, only his three youngest children were living in his household:

• Jesse M. Gray, born about 1829.
• Abbada (or Alzada) Gray, born about 1832.
• Pleasant M. Gray, born about 1838.

Jesse’s wife, Martha Mullins, had apparently passed away, perhaps recently; but she was still having children into the 1830s. So a daughter born about 1824 would certainly be feasible.

Other online trees, including Gail’s, and known DNA matches, added two older sons for Jesse Gray:

• William Mullins Gray, born about 1808.
• Cecil Gray, born about 1815.

No other daughters were listed in anybody’s trees. Did Jesse have any other unknown daughters?

According to the 1840 census, he did:

Jesse Gray
1 free white males aged 50-59 [Jesse Gray]
1 free white males aged 20-29 [Cecil Gray]
1 free white male under 5 [Pleasant M. Gray]
2 free white males aged 10-14 [Jesse M. Gray]
1 free white female aged 40-49 [Martha (Mullins) Gray]
1 free white female aged 15-19 [Nancy Gray (?)]
1 free white female under 5 [Alzada Gray]

1840 Federal Census, Autauga County, Alabama. Page 18. Household of Jesse Gray.

Could this be Nancy? The pieces appeared to be coming together. But how could I prove it?

Gray Cousins, Revisited

So if Nancy Harris were a daughter of Jesse Gray and Martha Mullins, how would that affect the earlier situation for Naomi Harris as a Gray “cousin”? The Grays who took Naomi in, Martha Rebecca (Gray) Wingard and William A. Gray, were children and grandchildren of Jeremiah Mullins Gray, who was the son of John Gray and Nancy Mullins. But as discussed earlier, John Gray had a will which named his children, and Nancy Harris was not one of them.

But if Nancy were a daughter of Jesse Gray and Martha Mullins, she would still be a closer cousin to the children of Jeremiah M. Gray than she would have been otherwise. Since the children of John Gray and Nancy Mullins, and Jesse Gray and Martha Mullins, were double first cousins — since both of their parents were siblings — they probably would have considered themselves almost as close as siblings too; so by extension, they might have considered their double second cousins almost as close as first cousins. And Naomi Harris would be a be a double second cousin “plus” — since she was not only descended from the Gray and Mullins siblings, but also another Gray sibling, Priscilla.

The X-DNA Smoking Gun

X-DNA — the DNA of the X-chromosome — is an unusual animal. In many respects, it behaves similarly to autosomal DNA, but as the female sex chromosome, it has its own unique inheritance pattern:

• Men inherit X-DNA (i.e. the X-chromosome) from only their mothers (receiving the Y-chromosome from their fathers).
• Women inherit X-DNA from both their parents — one X-chromosome from their mother and one from their father.
• Men inherit a recombination of both their mother’s X-chromosomes.
• Women inherit a recombination of their mother’s X-chromosomes, and their father’s whole X-chromosome (since he only has one).

This means that:

• Men only inherit X-DNA from two of their four grandparents (their mother’s parents).
• Women only inherit X-DNA from three of their four grandparents (their mother’s parents, and their father’s mother).
• X-DNA recombines 1/3 less often than autosomal DNA — so it has the probability of containing much older ancestral connections than autosomal DNA typically does.

Because of its unique inheritance pattern, X-DNA has the possibility both of matching distant ancestral connections, and of excluding certain relationships from consideration as possibilities.

By excluding certain relationships, I mean this: if I have an unknown match on the X-chromosome, I can reason that it must have come from, for me as a man, my maternal side; and for my mother, it must have come from either her mothers parents, or her father’s mother.

This means, for the Harris-Gray family, that my Grandmother Sarah SHOULD NOT have Gray on her X-chromosome. Because Gray comes from her mother, Ada Priscilla Harris, and for Ada, it comes from her father, Eli Harris — for Eli, he ONLY could have inherited X-DNA from his mother. Our only Gray connection, as far as is documented, is through Eli’s father, James S. Harris, and James’s mother, Priscilla (Gray) Harris.

In fact, almost as soon as I looked, I discovered what appeared to be Gray matches on Grandmother’s X-chromosome. I quickly found two or three of them, intersecting on the same segment.

• J.M. – Descends from Jesse Gray > Cecil Gray > Missouri Ann (Gray) Price. Also descends from Hannah (Oates) Bazzell > Hannah (Bazzell) Cook > Fadra Melissa (Cook) Price — both of which could be valid paths for X-DNA, so this could be either Gray DNA or Oates DNA.
• B.D. – Descends from John Gray > Jeremiah Mullins Gray > Mary Elizabeth (Gray) Herring.
• J.B. – Descends from Jesse Gray > Jesse Madison Gray > Narcissa (Gray) Tennison.

But wait. As soon as I exulted in this find, I realized there was a problem: All three of these matches traversed two males; a man can’t inherit X-DNA from his father, so neither Cecil Gray, not Jeremiah Mullins Gray, nor Jesse Madison Gray could have inherited Gray X-DNA. This couldn’t be Gray X-DNA after all.

But men do inherit X-DNA from their mothers. And all three of these men — Cecil Gray, Jeremiah Mullins Gray, and Jesse Madison Gray — had the same grandparents on their maternal side. Their mothers were sisters: not Gray, but Mullins.

In all three of these cases, there is a clear path of X-DNA inheritance from the tester to either Nancy or Martha Mullins — who were full sisters. Though this doesn’t offer certainty, it is a significant triangulation.

Finally, I found a fourth intersecting match, from a close cousin with Grandmother, making clear that this DNA does come from Nancy Harris.

Conclusion

So I find the evidence convincing that Nancy Harris, wife of James S. Harris, was by birth Nancy Gray, the daughter of Jesse Gray and Martha (Mullins) Gray, and the granddaughter of George Gray and Jeremiah Mullins. To summarize this argument:

• Common Assumptions about Nancy Harris Are Wrong: Nancy Harris is not “Sarah N. Andrews”, and this 1827 marriage record does not fit. James S. Harris married (1) Sarah Ann Bull, who died about 1843. Nancy Harris, fourteen years younger than James, was his second wife.
• Appearance of Closer than Expected Relationship to Gray-Mullins Family: Naomi (Harris) Akins, daughter of James S. and Nancy Harris, was called a “cousin” of the Gray-Mullins family, and taken in by them at the end of her life, which would appear unusual if she were only a second cousin.
• Too Much Gray Autosomal DNA: Sarah Ruth (Cook) Richardson, great-granddaughter of Nancy Harris, has too much Gray autosomal DNA, and too strong autosomal matches to Gray cousins, to only be connected to the Gray family by Priscilla (Gray) Harris. Another, closer Gray relationship seems to be indicated.
• Mullins Autosomal DNA: Autosomal DNA matches to descendants of Jeremiah Mullins, father of Nancy (Mullins) Gray and Martha (Mullins) Gray, as well as to other Georgia Mullins families. We know of no other path by which Grandmother would have Mullins DNA, than through the wife of one of the Grays.
• X-DNA Matches to Mullins: Sarah (Cook) Richardson should not have Gray matches on the X-chromosome; but I find three triangulated X-DNA matches to Nancy (Mullins) Gray or Martha (Mullins) Gray, with fourth intersecting close match to Nancy Harris descendant.

This proof is strongly circumstantial. The paper trail “seems to fit” — what we find in records is consistent with Nancy Harris being a Gray — but there is as yet no positive documentary evidence of the fact. The DNA evidence likewise fits, and supported by triangulation, appears quite strong. While I am confident enough in my conclusion to publish it, I would like to see more evidence:

• More evidence of autosomal triangulation on Mullins family.
• Additional X-DNA matches to Gray or Mullins.
• More solid documentary evidence.

I appreciate any comments or input.