(Minor revisions and simplification of a story originally published on 16 February 2023)
BACKGROUND EXPLANATION
My great grandfather Charles Francis Edwards was born in Keokuk, Iowa on 22 February 1869. He claimed he was orphaned at a young age. His Montana marriage license to wed Mary Jane Wescott in 1896 indicated that his parents were Louis Edwards and Martha Hoover.
Investigations undertaken on my behalf by Alice Veen of Prairie Roots Genealogy in Iowa led to the establishment of his mother and her ancestry. She was Barbara (not Martha) Hoover (c.1832-1890). Barbara married three times, first in 1855 to a Hoover cousin named William. After William’s untimely death in 1858, she married Louis Edwards in Illinois in 1861. Louis soon left to take part in the American Civil War, became very ill and came home to die in Illinois on 18 February 1866, a full 3 years before the birth of Barbara’s son Charles. Barbara married her third husband, George Payton, in What Cheer, Keokuk, Iowa in 1873. Clearly, Charles was born between marriages for his mother.
Who then was Charles’s father?
Traditional genealogical records proved useless for determining this well-guarded family secret. Recent developments in genetic genealogy were the only hope for finding a father for Charles; his living descendants would contain valuable information in their genetic makeup.
Y CHROMOSOME ANALYSIS
The Y chromosome is passed down virtually unchanged from father to son down the line, enabling determination of male inheritance quite clearly. This usually means following the family surname back up the male line, only broken where the supposed father was not the actual biological father of the son. Charles Edwards certainly did NOT inherit the Edwards Y chromosome. His Y chromosome would have come from his unknown father who somehow encountered Barbara in the spring of 1868 when she became pregnant with Charles. Of course, both Charles and his unknown father (“UF”) have long since passed away so traditional paternity testing is out of the question.
The next best thing is to have a male child of Charles submit a DNA sample for determining his Y chromosome haplogroup and locating any close living matches who have also tested and share similar Y chromosome DNA. The two sons of Charles are also deceased, but both had sons who would have inherited that same Y chromosome from UF. One of them graciously agreed to be tested through FamilytreeDNA at the 67 marker level. (To protect his privacy, I am going to call him “YEdwards”.)
When YEdwards’ results came back with Haplogroup R-M269 (as of April 2025 classified as R-P312), not surprisingly there were no matches with the Edwards surname. Although this haplogroup is common in Central Europe, Spain, Portugal, France and the British Isles, there were just 2 matches (here called “LS” and “RN”, again to protect their privacy) leading to 2 surnames: Saum(s) and Nagle/Nagell.
Match Name | Genetic distance | Most distant male ancestor | Haplogroup | Final SNP |
LS | 3 | (Saum) | R-M269 | - |
RN | 7 | Hans Nagell b1615 Germany | R-FGC8158 (he has done a more advanced test leading to greater differentiation) | R-FGC8158 |
When the results were first posted, LS had a couple generations of his tree online. From LS’s parents’ names, I was able to construct his family tree going back an additional four generations to Adam Saum (1785-1855) and his wife Margaret Miller (1792-1856) who had a family of some 14 children including no fewer than 11 sons! Later, I determined two additional generations to Nicholas Saum, German immigrant to Saumsville, Virginia, in the 1700s. Nicholas had numerous children by 3 wives. This was not going to be simple.
RN did not post a tree, but he did list his most distant known male ancestor as Hans Nagell of Germany.
The Genetic Distance is indicative of how much mutation has occurred in the Y chromosome; the lower the number, the closer the relationship. FamilytreeDNA gives tips for the likelihood that two men share a common male ancestor within a certain number of generations.
Comparing RN to YEdwards, with a Genetic Distance of 7, the likelihood of a shared male ancestor looks like this for Nagle:
Within 4 generations | 3.74% |
Within 8 generations | 31.56% |
Within 12 generations | 67.02% |
Within 16 generations | 88.1% |
Within 20 generations | 96.49% |
Within 24 generations | 99.1% |
Comparing LS to YEdwards, with a Genetic Distance of 3, the table looks like this for Saum:
Within 4 generations | 44.6% |
Within 8 generations | 84.25% |
Within 12 generations | 96.64% |
Within 16 generations | 99.38% |
Within 20 generations | 99.9% |
As of April 2025 the FamilytreeDNA website is using a time chart showing that LS and YEdwards likely share a common ancestor between 1300 and 1850 CE, while for RN it would be 400-1600 CE. Clearly, it is more likely that UF for Charles Edwards will be found more recently in the Saum family tree of LS.
Next step was to build the Saum family tree from that determined for LS, starting with the most distant male ancestor found for him, namely Nicholas Saum. 6 generations separate LS from Nicholas Saum. This could be a promising lead and well worth the time and effort to build out the extended family tree of Nicholas Saum and his descendants.
The Y chromosome test has given us a possible surname for the UF for Charles Edwards, but on its own cannot give us the name of a specific man.
AUTOSOMAL DNA ANALYSIS
The autosomal (family finder) test is another genetic test that can determine ancestry based on the other 22 (non-sex-determining) sets of chromosomes.
We all inherit 50% of our DNA from our mothers and 50% from our fathers (one chromosome from each parent of each of the 22 chromosomes in the nucleii of all our cells), about 25% from each grandparent, 12.5% from great grandparents, etc. The exact amount varies because of recombination of the chromosomes. As you go back more and more generations, the amount diminishes and some ancestors are not represented at all in a person’s genetic makeup. The more family members who have their autosomal DNA tested, the more of the family’s overall genetic background becomes apparent.
My mother Elinor had her autosomal DNA tested in her later years. She is one generation closer to the UF for Charles Edwards than I am. Charles was my mother’s grandfather so UF would be her great grandfather and we might expect that about 12.5% of her DNA came from him (although that could vary). Similarly, YEdwards is of the same generation and would also expect about 12.5% of his autosomal DNA (as well as 100% of his Y chromosome) to have come from UF. Since I have access to the DNA results for both YEdwards and Elinor, these are the best results for me to use to analyze their matches who come from within the Saum tree that I have built. (I subsequently obtained access to the Ancestry DNA results of another of Elinor and YEdwards’ first cousins (LM) and will discuss LM’s results later.)
Because comparisons can be made only to DNA matches who have tested at the same company, I took advantage of every opportunity, when granted permission by the original test taker, to upload raw DNA data from the testing companies to other sites such as GEDMATCH and My Heritage. I have also done additional tests myself with Ancestry DNA and 23&Me (recently withdrawn) and have uploaded my own raw DNA results to Living DNA. Every site has yielded more Saum connections.
Finally, some patterns began to emerge that gave me an additional clue. I had a lot of matches that were to the Henderson family, which was otherwise unknown to me. (I had a well built-out tree on all known branches of my tree other than the mystery line for UF.) Frequently, the Hendersons also fit into the Saum tree through descendants of a marriage between George Adam Saum and Susannah Henderson. Sometimes matches were just to Saums and sometimes just to Hendersons, leading me to investigate this family further since we could be matching our DNA matches through either the Saum or Henderson line. But, for our family, it seemed we had to be descending from both, most likely from a marriage of the two. I did exhaustive research to ensure there were no other Saum-Henderson marriages that could also be the source of my family having inherited DNA from both those families. There was one, but it was for a male Henderson to a female Saum, which would not have given the Y chromosome results obtained for YEdwards.
The one marriage between a male Saum and female Henderson is the one of George Adam Saum and Susannah Henderson whose children were all born within a few years of Barbara (c. 1832).
George Adam Saum (b. Saumsville, VA 1797, d. Highland, OH 1836) married in 1819 to Susannah Henderson (b. 1805, d. 1855) – 7 children, including 3 daughters and 4 sons:
1. Charles Frederick Saum (b. 1821, Ohio, d. 1886 Lawrence, Douglas, KS)
2. Jacob Saum (b. 1822, d. 1861 (after being shot by neighbours for being an abolitionist and Lincoln supporter – several years before Charles was conceived)
3. George Thompson Saum (1824-d. 1912 Kellogg, Jasper, Iowa)
4. Stephen Adam Saum (1834-d. 1927 in FL but had farmed most of his life Kellogg, Jasper, Iowa)
Conducting thorough research on the three possibilities from this family for UF first led me to focus on the third and fourth sons since both spent most of their lives in Iowa, which is likely where Barbara was living when she conceived Charles. A map shows that the distance between Kellogg and What Cheer where Barbara’s parents were living was about 50 miles, not exactly next door but certainly closer than the 280 miles to Lawrence, Kansas where oldest brother Charles lived. These people moved around more than we might think and certainly maintained connections with family members who might not live nearby. As will be seen later, the connection was most likely through Barbara's sister who lived near two of the Saum brothers.
Google Earth image of Iowa and Kansas showing locations of Barbara’s parents and the Saum brothers
THREE BROTHERS POSSIBLE AS CHARLES'S FATHER
1. Charles Frederick Saum (1821-1888)
Charles Frederick Saum was born in Dodsonville, Highland, Ohio in 1821. He had one daughter born in 1844 to his first wife Mary Elizabeth Shawver before her death in 1846. His next marriage was in 1847 to Mary Bardell; the couple had nine children born between 1849 and 1869.
Threatened along with his next-younger brother Jacob N. Saum in a poster in 1861 for being an abolitionist supporter of Abraham Lincoln, he survived when Jacob was shot later that year by neighbors. This provided the impetus for Charles to make a hasty departure to Lawrence, Douglas, Kansas with Lane’s army in 1861. He lived in the area until 1878 when he moved to Mt. Ayr, Osborne, Kansas. He was a farmer, carpenter, and notary public. He wrote several articles for newspapers about farming and about his journey to Mt. Ayr. He was interested in politics and served as a delegate in 1880.
An obituary for his daughter Emma in 1910 contains more information about him as described in another story.
Of the three brothers, Charles Frederick Saum lived the farthest distance from where widowed Barbara Hoover Edwards was living in the spring of 1868 when she became pregnant (somewhere in Iowa, possibly at What Cheer with her parents which is where she was living at the time of the next census in 1870). Still, the similarity in his name to Charles Francis Edwards (and the ambiguous DNA match results to be discussed later) keep him well in contention as a potential father.
2. George Thompson Saum (1824-1912)
Born in Dodsonville, Highland, Ohio 4 January 1824, George married first wife Sarah Roush in 1847 and, after her death, Marieta Boatright in 1855; she died in 1867. He purchased land and moved to Buena Vista Twp., Jasper County, Iowa about 1854-55. He married his third wife Ella Wilson 2 March 1869. He had a total of 9 children (2 each from first and third marriages and five from the second).
His land in Buena Vista Township, Jasper County Iowa was located at section 1, Twp.79 Rge.18 W5M, and he had nearby land in sections 5 and 6 of Twp.79, Rge.17 W5M in Richland Township. The History of Jasper County gives his landholdings as 1000 acres. This land was just south of Kellogg, Iowa which was also his postal address at the time. Kellogg was a new town in 1865, being laid out in anticipation of the Mississippi and Missouri Railroad being built through the area. The population never exceeded 800 people.
The next township north of Buena Vista is Kellogg Township, which happened to be the home of Barbara Hoover’s sister Susan Anna Hoover Bleakney. The Bleakney land was located just north of the town of Kellogg at section 14, Twp.80 Rge.18 W5M. When Barbara gave birth to her twin daughters Minnie and Grace in 1865 or 1866, she did so at the Bleakney home. One can assume Barbara could have been a regular visitor or perhaps was even living at the Bleakney home at the time she became pregnant in the spring of 1868. She would have been very near the home of the widowed George Thompson Saum at the time. The distance is only about 3 ½ miles. The two would definitely have known each other in this small rural community. However, if George was aware of her pregnancy and the birth of a son Charles at her parents’ home at What Cheer, Iowa, on 22 February 1869, it would seem a real slap in Barbara’s face for him to have married Ella Wilson only about a week later!
Google Earth map showing locations of George Thompson Saum property, Saums cemetery and the Clark and Susan Anna Bleakney home near Kellogg, Iowa about 3 to 4 miles away
George Thompson Saum died 19 September 1912 when his car was hit by a train while he was driving from Waukee to Des Moines. (His stone is in the Saums Cemetery located on his original section of land as shown above.)
His obituary indicates that “Mr. Saum was not a church man nor a member of any secret order however he was a man of very pronounced opinions even in politics and the democratic party to which he belonged. He was a careful business man and regarded his word as his best asset and was the kind of a man whose word was as good as his bond.” He does not sound like the sort of man who would abandon a woman pregnant by him to her own devices - assuming, of course, that he ever knew of her situation.
3. Stephen Adam Saum (1834-1927)
Like the other brothers, he was born in Dodsonville, Highland, Ohio. In 1853 he was living in Knox, Illinois where he married Rachel Ellis 21 December 1856. By 1860 he had joined his older brother George Thompson Saum in Jasper County, Iowa. Like George, Stephen had land in the vicinity of Kellogg and would have been neighbors to Barbara’s sister Susan Anna and her husband Clark Bleakney.
Stephen joined the Union army 3 August 1862 where he served as a Private in Company K of the Iowa 28th Infantry. By 1864 he had been promoted to full 6th corporal. He mustered out 31 July 1865. By the time of the 1870 census he was living in Buena Vista Township, Jasper County, putting him equally well situated as brother George to have had a relationship with Barbara Hoover Edwards in the spring of 1868.
Stephen was a manufacturer of pumps and a businessman in Kellogg Township, often acting as a travelling salesman (presumably on behalf of his company Saum, David & Co.) He died in Florida in 1927.
Stephen and Ella had three children, two of whom did not survive childhood. Stephen’s medical records indicate that he contracted mumps in 1863 during his Civil War service, leading to atrophy of his left testicle. He did not father any additional children with his wife after the Civil War, making him the least likely of the three brothers to have fathered Charles Edwards.
ANALYSIS OF AUTOSOMAL MATCHES TO THE SAUM FAMILY
Several or my family members have had DNA tests done with a variety of the popular testing companies: Familytree DNA, My Heritage and Ancestry DNA. Some of these results have also been uploaded to GEDMATCH. I have tried to take advantage of every clue leading to the family that Charles Edwards’ father came from. The tools on the various sites vary significantly in terms of the information provided and the tools available to narrow down matches.
Except for Ancestry DNA, the testing sites provide chromosome information for where a test taker and a match share DNA on a particular chromosome and the amount of such match expressed in centimorgans (cM). This enabled me to make a spreadsheet for overlap areas on chromosomes where the same DNA was shared by one or more matches, meaning that it was likely inherited from the same ancestor some distance up all their trees.
Portion of Spreadsheet for Elinor’s DNA matches by Chromosome - portion C6 | |
test co. | tester | match | | chromosome | start position | end position | cM |
FTDNA | ELINOR | Richard | | 6 | 88.5 | 102.3 | 12.2 |
MH | ELINOR | Norma | | 6 | 88.5 | 105 | 15.3 |
FTDNA | ELINOR | Oneil | | 6 | 88.5 | 106.3 | 16.7 |
FTDNA | ELINOR | Rachel | | 6 | 88.5 | 108.5 | 20 |
MH | ELINOR | Marvin | | 6 | 88.5 | 109 | 21 |
FTDNA | ELINOR | Scott | | 6 | 89.1 | 104.9 | 13.8 |
FTDNA | ELINOR | ayla | | 6 | 90.4 | 108.5 | 17 |
FTDNA | ELINOR | John Joseph | 6 | 91.7 | 125.1 | 28.8 |
FTDNA | ELINOR | Scott (5) | | 6 | 93 | 109.1 | 15 |
FTDNA | ELINOR | Murray | 6 | 98.8 | 110 | 12 |
Ancestry | ELINOR | BL | Gedmatch | 6 | 99.9 | 125.7 | 22.4 |
Ancestry | ELINOR | Norton | Gedmatch | 6 | 107.3 | 125.7 | 14.9 |
FTDNA | ELINOR | Vancil | | 6 | 108.1 | 125.5 | 13.7 |
FTDNA | ELINOR | Elizabeth | | 6 | 110.9 | 125.1 | 11.2 |
| | | | | | | |
The above section of my spreadsheet shows a portion of my mother’s chromosome #6 and shows several matches in the area from 88.5 to 125.7 (these numbers have been truncated for ease of use.) These people all relate through the Saum family tree. The amount of shared DNA is not large but this is just for one segment of one chromosome, varying from a low of 11.2 for the match called Elizabeth to a high of 28.8 for a match named John Joseph. Not everyone posts a family tree on their DNA sites, but fortunately John Joseph had an extensive tree that enabled me to see that he was indeed a part of the Saum/Henderson family as a descendant of oldest brother Charles Frederick Saum.
The Shared cM Tool is a well-researched and generally accepted tool developed by Blaine Bettinger and accessible on the DNA Painter website. It is used for ascertaining the potential relationships that two test takers could have based on the number of cM of DNA that they share. Using the total 45 cM shared by Elinor and John Joseph over all their chromosomes gives this result:
Blaine T. Bettingerwww.thegeneticgenealogist.com
CC 4.0 Attribution License
YEdwards also matched John Joseph on FTDNA and shared 52 cM with him. Relationship possibilities according to this amount are shown in this chart:
Blaine T. Bettinger
www.thegeneticgenealogist.com
CC 4.0 Attribution License
John Joseph (a descendant of Charles Frederick Saum), Elinor and YEdwards would be either half second cousins once removed (Half 2C1R) (if Charles Frederick Saum is UF) or third cousins (3C) (if one of the other two brothers is UF).
The foregoing analysis is just an example of the methods that can be used. Using a variety of tools enabled me to find dozens of Saum/Henderson matches for various of my family members and to add all these “new” family members to my own tree so that I could determine potential relationships. Consistently, the Shared cM tool indicated that these matches could arise if any of the three Saum brothers was indeed Charles Edwards’ UF.
My Heritage and Ancestry DNA both added tools for suggesting how a tester and a match might share ancestors. For this to work, the tester must have a genealogy family tree associated with their DNA results. Adding a hypothetical UF where you think he might belong in your tree is a necessity. For my trees, I had added Charles Frederick Saum as the hypothetical father for Charles Edwards.
Having uploaded my mother Elinor’s family tree along with her raw DNA data from FamilytreeDNA to My Heritage enabled me to view their tools called “Theory of Family Relativity” and “Smart Matches”. One of her Saum/Henderson matches found this way on My Heritage was L.P., a descendant of Charles Frederick Saum. L.P. would be either a half 2nd Cousin once removed or a Third Cousin of Elinor’s depending on which of the brothers is the UF of Charles Edwards. (It might not immediately be obvious, but a half 2nd cousin and a 3rd cousin are equivalent.) Once again, the Shared cM Tool indicates these are quite probable relationships for this amount of shared DNA (40.6cM). (There are many other DNA matches on My Heritage for Elinor to members of the extended Saum/Henderson family, most often through the two older brothers.)
Turning now to the tools at Ancestry, their “ThruLines” tool is one of the most useful in that it sets out possible trees showing exactly how all the DNA tested cousins relate to a shared ancestor.
I have access on Ancestry to DNA results for three family members of my own generation plus one for a younger generation and was able to see dozens of Saum/Henderson DNA matches with exactly how we might relate. But testing an older generation is even better as the relationship to UF is one generation closer and therefore more of UF’s DNA is likely still in evidence. Fortunately, another first cousin of my mother’s, L.M., has granted me access to his DNA on Ancestry. L.M.’s ThruLines to the Saum/Henderson families have proven very useful. To give you an idea of the number of matches that show up for L.M.’s DNA and how certain this makes me that I have found the right family for UF:
· Of the 8 great grandparents of L.M., hypothetical Charles Frederick Saum has the most DNA matches for L.M. (excluding descendants of Charles Edwards himself) – 87 compared to just 3 for Barbara Hoover who we know to have been Charles Edwards’ mother!
· Of the sixteen 2nd great grandparents of L.M., the parental couple for the three brothers (George Adam Saum and Susannah Henderson) again have significantly more than the other 2nd great grandparents. This Saum/Henderson couple have a total of 102 DNA matches to L.M. compared to just 16 for Barbara Hoover’s parents (again excluding those of us who descend from Charles Edwards himself).
· The pattern continues back for the next couple of generations with significant DNA matches for all branches going back from the Saum/Henderson families.
L.M.’s closest matches have shared DNA with him in amounts up to 112 cM that he shares with “Linda”, a descendant of Charles Frederick Saum. She would be either a half second cousin twice removed (half 2C2R) or a third cousin once removed (3C1R) to L.M., again falling well within the range of possible relationships according to Blaine Bettinger’s Shared cM Tool:
Blaine T. Bettinger
www.thegeneticgenealogist.com
CC 4.0 Attribution License
from the DNA Painter Website
"What are the Odds?" Tool
A tool developed by Jonny Perl available on his DNA Painter Website is called “What are the Odds?” (WATO) located at https://dnapainter.com/tools/probability. This tool allows you to upload your family tree and include DNA matches in that tree along with the amount of shared DNA. A question is asked, in my case “Who is the father of Charles Edwards?”, the test taker’s name is added and then hypotheses are posed or suggested by the website. If the odds come back as “0”, the hypothesis is an impossibility. Anything from “1” up is possible, with the highest number being deemed the most likely. Even Jonny Perl will admit that just because one hypothesis is deemed most likely does not necessarily make it the right answer.
So, what happens when I pose various hypotheses for UF based on different test takers’ DNA matches? I will use only the oldest generation of testers available to me – namely Elinor, YEdwards and L.M.
1. YEdwards as the Tester
- Charles Frederick Saum most likely
- George Thompson Saum second most likely
- Abraham Henderson third (but can be eliminated as not fitting with Y chromosome Saum result)
- Asa Henderson (a descendant of the other Saum/Henderson marriage mentioned earlier, but also can be eliminated as not fitting with the Y chromosome Saum result)
- Stephen Adam Saum possible, score of 1, remains a possibility
2. L.M. as the Tester
- George Thompson Saum most likely
- Charles Frederick Saum second most likely
- Stephen Adam Saum a possibility again with a score of just 1
3. Elinor as the Tester
- Charles Frederick Saum most likely
- George Thompson Saum second most likely
- Stephen Adam Saum third most likely
The foregoing results are therefore inconclusive but it does seem that both Charles Frederick and George Thompson Saum are the likeliest candidates as father for Charles Edwards but Stephen Adam Saum remains a possibility.
CONCLUSIONS
In my opinion, the least likely father for Charles is youngest brother Stephen Adam Saum because of his damaged testicle after suffering from mumps during the Civil War. As between the two older brothers, it is really a toss-up (George Thompson Saum for his geographic proximity to Barbara’s sister’s Bleakney family or Charles Frederick Saum for the similarity in name and the huge number of DNA matches among his descendants to our family).
One possible way to pinpoint one of the brothers would be by chromosome analysis to determine if any specific DNA segment(s) could have been passed to us by only one of the brothers, but I don't have enough data (nor perhaps enough patience!) to make this a viable option. We may never be able to identify the father with certainty. However, it is certain that the Saum family is our family; we are related to these people.
I welcome any additional information that may come to light working to either prove or disprove my theory. I remain hopeful that as testing increases and additional tools become available, one or other of the brothers may be claimed with certainty.
Reference DNA Sites:
Ancestry: https://www.ancestry.com
DNA Painter website: https://dnapainter.com
My Heritage: https://www.myheritage.com/
Familytree DNA: https://www.familytreedna.com/
GEDMATCH: https://www.gedmatch.com/
(Sources for genealogical data provided for family members can be obtained through the author.)
.png)
