The following is a summary of several years and tons of notes on finding the father of John Newton Mix (aka John Newton Henderson). My grandmother, Anna Henderson was sent to foster care with her two siblings after her mother Susan McDevitt Henderson died in 1898 in Philadelphia. She was later sent to the middle of Pennsylvania where she was adopted. She had a baptismal certificate so we knew the names of her birth parents. After hiring a genealogist who lived in Philadelphia, and combing through countless online records, I was able to find John Newton Henderson's mother, but his father was a mystery.

Background

John N Mix
(1862-1917)

John shows up first as John N Mix in 1870. He later went in the navy as an Apprentice at age 17, but navy life didn't sit well with him and he went AWOL. For the time after the period, he went by his mother maiden name, Henderson. Oddly, he joined the navy later under the name Newton Henderson, same birthday, in Philadelphia, and again went AWOL while in Europe. A few months later, he was listed as a stowaway on a ship returning the US under the name John N Mix, brickmaker from Philadelphia (he had been John Newton Henderson, brickmaker, at the time he left). He never shows up as a Mix again until his death (when his sister provided the information for his death certificate, and listed his mother as Elizabeth Henderson, father unknown).

I spent years looking for his father. I tracked every Mix or Meeks (alternate spelling) that came through Philadelphia. He was born during the start of the Civil War, and many people came through the city. At this time, I believed his sister Mary Ella Mix had the same father (but that was wrong). Here is a list of some of the people I had considered to candidates for the Mix father, but never saw anything compelling.

• Oliver P Meeks/Mix (shoemaker lived nearby, from Maryland - he was the only resident Mix in Philly but he was married)
• Edward Meeks, young rich guy who started a book business in Philadelphia
• Hiram B Meeks, came through during the Civil War from upstate PA, and was in Philadelphia at the right time
• George H Mix, the criminal from Connecticut who was arrested in Philadelphia
• Horace B Mix from NY, came through Philadelphia during the war, left a wife here who divorced him
• William B Mix, a ship officer from Connecticut during the war who stopped in Philly at the right time
• John W Mix from upstate PA, attended law school in Philadelphia at the right time

grandma's baptismal certificate

While these people all came through the area, there was nothing that made any one more compelling than another. The census records show some discrepancies on when John N was born, so it's not clear if Dec 25 1862 was his actual birthday (or just a date he used that he could remember) so he might have been born a little earlier in the year. I have lists of all the young Mix/Meeks families in the surrounding areas with their details (not shown!) who have mostly been ruled out. For several years, I would think one looked compelling, but in each case, it was just someone with the right name passing through the same big city. Nothing actually connected any of these people to my family other than proximity.

I also considered that his father was not even really named Mix. His mother switched to that name about the time of John's birth and his sister also went by Mix. Was it just made up? There were other men who lived in the same apartment as the Henderson family according to directories and I checked them out as well. Oddly in the 2nd 1870 census, Elizabeth shows up with the last name Kelly, like her new daughter, but later she went back to Mix (yet her daughter remained a Kelly). So it wasn't clear if Elizabeth actually had a wedding ceremony with Mr Mix, or considered it "common law" or adopted child's father's surname for social reasons. She was too quick to adopt the name Kelly, it seemed.

I stalled making progress for a couple years but occasionally checked to see if new records showed anything.

Maybe DNA?

Later, when ancestry.com offered the autosomal DNA testing, I took the test but mainly to see my ethnicity. I did check for the name Mix but didn't see anything obvious and went back to thinking that it was just made up. At one point I noticed that I had DNA representatives for everyone *but* Mr. Mix. So I started sorting out all the near matches and working back those that didn't fit into the family. I even started working back the trees of close matches of people who had incomplete trees (I've gotten good at working back trees pretty quickly). One day in working out a close match's incomplete tree I noticed a Mix, but his Mix family was from Indiana! I continued to research people and started finding a pattern - these new matches were matches to each other, and to the same people in SE Indiana.

I then got my 2nd cousin Ann and my sister to do DNA tests, and I saw the same pattern expanding. There were three families that were converging in Switzerland County, Indiana - Mix, Fallis, and Wallick. The Wallick connection was strong, but didn't match any known people in these families. Then I noticed that Nathaniel Mix and Philip Wallick had both lived in Bullskin Township, Fayette County, PA, before these families appeared together in Indiana. I realized that Nathaniel Mix (who came originally from New York) had been here in PA when he was of marrying age, and that an extended group from this area made the trek to SE Indiana at the same time. I started to suspect his wife was a Wallick.

I eventually found his wife's death in a newspaper, so I knew her first name (it seems nobody on the internet new his first wife's name. Then I found that Philip Wallick had a daughter with the same name Catharine baptized in PA (!). Lastly, I found a death record for Nathaniel's son Lyman W Mix, Jr - the W was for Wallick! It was common in that era to use the wife's surname for a child's middle name. OK, so now I used *my* DNA to solve Nathaniel Mix's wife's previously unknown maiden name.

Nathaniel had two sons who married Fallis sisters, daughters of Dr Samuel Fallis (who was a consistent DNA match for all three of us). The problem was to sort out who could be the father of John N Mix/Henderson. From a DNA point-of-view, all three boys have almost the same DNA - there isn't a way to show a connection to someone to sort this out. Looking at the three boys, brothers Lyman Jr and Samuel, and their cousin George, it seemed they had different opportunities to be in Philadelphia. The brothers were wrapped up in the Civil War very early, but George was free. The only issue with George was that he joined the local militia in July 1862 and was married to Catharine Starkey in October of 1862. He would have had to have left Elizabeth in Philadelphia and gotten back to his home area to quickly court Catharine before getting married. This seemed at the time, the least problem. I ordered the Civil War muster rolls and pension records for all three Mix boys.

I did consider that there might be some other "missing Mix" person who left home early on and was a son of Henry H Mix or Lyman W Mix, Sr. But there are accounts of how many children Henry had and these match up. Also, Lyman W Mix, Sr died young and the estate records only mention three children (and two boys). I had even considered that children of Catharine Mix who married George Fallis (siblings of the Mix families!) may have lied and used their mother's maiden name while in the East during the Civil War, but I found nothing. I wanted to rule out all other possibilities. Following is my analysis of the three Mix boys.

example snippets of my DNA spreadsheet trackers for Mix, Wallick, and Fallis families showing group color codings

George Mix?

I next paid for DNA testing for a descendent of George Mix (Delores) and a descendent of a brother of George (Barbara). Barbara was older and two less generations (so 4x the Mix DNA!). The results were disappointing. Delores, who should have been a 3rd cousin didn't match my sister, Ann, or me. She did match Barb, but Barb was a disappointing match to us as well (ancestry predicted only 4th cousin at 30cM to Ann). Well, I stuck with this theory for too long because I did have an example of a 3rd cousin not matching and assumed this was just a bad luck of the draw.

I also got shared access to a few other Mix, Fallis and Wallick relatives, which continued to reinforce the original conclusion that this was the correct family, but didn't sort out which of the three boys was the correct one. I created multiple spreadsheets to track matches and analyze the clusters of people with common ancestors.

Over the next year, I turned my attention to building a complete tree of the Jacob Conrad descendents. This was a bigger project with more accounts, more analysis, and I was able to predict and then verify many new Conrad relatives. In short, I was getting better at sifting through the patterns of matches and know which are more likely just noise and due to chance and best ignored.

Then I came back to the "Mr. Mix" project and gave it a new checkup. It quickly became obvious that I spent too much time looking at George Mix and ignoring the negative feedback from the DNA. While I did have a case where a 3rd cousin didn't show up in a match list, there were no cases where several close cousins didn't match. That is, with my sister, Ann, and me all having different DNA, *one* of us should have matched Delores if we were third cousins! And the weak link to Barbara reinforced that this was probably the wrong guy. It was time to cast a wider net and spend more time on the other candidates.

Lyman Mix?

I went back and looked at Mix brothers Lyman Jr and Samuel. Lyman was looking promising as his service brought him all the way to Maryland. He was wintering over in an Indiana Cavalry unit at the time he'd need to be courting Elizabeth Henderson, at least briefly, in early 1862. The only problem would be getting them together.

I had considered that Elizabeth Henderson might have been the one to go visiting the troops (and that's still possible). Maybe she left home to be a "camp follower." Let's consider all possibilities - she did have four children with four different fathers. But, she was living at home in 1860 and later with her parents with her kids in 1870. She wasn't homeless. She also didn't have relatives serving in the military. If she wanted to be close to the troops for whatever reason, there were tons of them coming through Philadelphia, staying in camps there, and then shipping out. Also, John N consistently says he was born in Philadelphia.

When I studied the ability of an young enlisted man to leave his unit and travel to Philadelphia from southern Maryland in 1862, it turns out it wasn't that easy. Officers could more easily take leave, but the enlisted men needed a furlough, and those were difficult to get and it was believed to be hard to get the young troops back. Better to just keep them in camp and keep drilling them. Lyman Mix was promoted to corporal in 1862, so it's unlikely he had recently gone AWOL.

Samuel Mix?

Lastly I found myself taking a detailed look at Samuel. In his service records, he had dysentery in his unit in late 1861 and was sent to the hospital in Indianapolis. But fortunately for him, his uncle Milton Spencer, who owned a store in Indianapolis and whom Samuel had resided with and worked for in 1860, came by to volunteer to nurse Samuel back to health so the Army wouldn't have to. Samuel was recuperating with his uncle for six weeks (!) before returning to his unit.

I had originally pictured Samuel lying in bed for six weeks, as his civil war pension seems to read. Then he got up and went back to his old unit, fit as a fiddle. But the more I read these pension applications, I came to see that the purpose of the write ups was to get a pension, and some of them seem a bit exaggerated. With Samuel living in Indianapolis, he was living within blocks of a railway station that could take him right to Philadelphia. He was recuperating for six weeks! It's possible that Uncle Milton might have even sent Samuel on business errands to Philadelphia for his store. So between the two available brothers, it was beginning to look like Samuel was actually the best candidate when it comes to having the opportunity.

Appearances

I had made notes of the physical descriptions of John Newton Mix and the three father candidates before, and the only thing that stood out was that Lyman Mix had black hair while others had brown hair. But since they all had the same starting DNA, I didn't make too much of that. There could be recessive genes in action and all the cousins had the same gene pool available. But now I decided to look at the descriptions of Samuel's kids. These descriptions show up in military service records and draft registrations. This made the case for Samuel Mix stronger still.

Philip, s/o George

It's worth noting that the Samuel Mix family tends to have shorter height, fair/ruddy complexion, and lighter hair, like John Newton Mix. Lyman, Jr was taller, had a dark complexion (even when older as a businessman), and black hair. These are all largely heritable attributes.

John N Mix, left; Lester H Mix, s/o Samuel Mix, right (suspected half-brother)

Back to DNA

Lyman Mix Jr

Lyman W Mix, Sr had one daughter who seems to have a few descendants alive now (it would be great to get one tested!). Samuel Mix had four sons live to adulthood but none had any kids. They mainly got married late in life and it's not clear if any had issues conceiving. Samuel's wife was declared insane after his death and moved around among her sons. Anyways, it was a good time to go back check the DNA matches. There just are not a lot of Mix descendants out there. In many cases with the accounts shared with me, the closest Mix matches to these other people (outside of immediate family) are my sister, Ann, and me. But one thing that I hadn't really looked at is that Lyman Mix Sr had died young and his widow remarried and had several children (half-siblings to Lyman Jr and Samuel). If I haven't been seeing many Mix matches, what about these people? I went back and filled in most of the descendants of the Roberts children (the half siblings). I had noticed a couple common DNA matches before but hadn't given them any special weight before. If I'm descended from this side of the Mix family, there should at least be more matches to these people who have the same Fallis daughter in common.

Yup, as I started tracing back matches-of-matches to those that were matches through the Roberts kids, I found that there were *many* of them. At this point, there are a couple tight clusters of people matching us and each other who are all descended from Susanna Fallis (Mix/Roberts), the first wife of Lyman Mix Senior. At this time there are 19 of them, ranging from 4th to 5th cousins. As suspected, there many more relatives to found if I indirectly focused on the Fallis family (in order to infer my Mix connection).

It's worth pointing out that a substantial number of DNA matches are for transient people who moved out of the area early on. Unlike the valleys where my parents grew up where the same people have lived for almost 200 years, this part of Indiana is an area where many families stopped on their way west, so there is not a lot "everybody is related to everybody several times over" among most descendants that I analyzed. I've only seen one tree that had any appreciable overlap. I wouldn't use this type of DNA groupings to make strong statements about the relationships to families in the valleys where my parents grew up.

Conclusion

Well, I've changed my mind too many times on this to think this is done, but this documents my reasoning to-date. Omitted are the piles of notes, pension papers, etc, mostly about the wrong people, but also spreadsheets showing the expected patterns. But this explains how I came to use DNA to assign the father of John Newton Mix. At this time, I feel pretty confident that Samuel Mix is my missing person. Here is my subjective score for each candidate.

The lessons I learned in this DNA project are:

• It's largely impossible to create new genealogical information by looking at matches from a single DNA account, if the ancestor is earlier than your grandparents.
• I originally put too much weight on the DNA magnitudes in the 10-45cM range. That is, I assumed that all bigger numbers were closer relatives and that is not true. The variation in cM levels is too big to make strong conclusions from differences in cM magnitude.
• If you look for name matches across enough trees, you will always find some that overlap. But unless they are sharing DNA as a group, it's just a random occurance and not very interesting.
• You will need to create mini-family-trees locally (on your PC) for the most interesting matches yourself. You can not depend on making progress by only using the family trees that others have posted.
• In order to identify matches, it's best to find all the current descendents of the people of interest, at least through 1900. That way when you see accounts with small trees, you'll be able to see the connection quickly.
• It's most important to look at the Shared Matches as they relate to other shared matches and not just matches in isolation.
• When looking at groups of matches:
• look at all of the accounts that you have (either that you own or via user guest access)
• look at the "Shared Matches" of each of these people
• do not give too much importance to "Shared Matches" of your "Shared Matches" unless they circle back into your "group"
• pay as much attention to members of your "groups" that you've identified a relationship as to those that are still mysteries
• realize that some members of your group are there by accident and don't necessarily share a common ancestor, but the more they overlap other members of the group, the more likely that they do
• realize that some members of your group will have a common ancestor back further generations than others. The cM numbers will hint at the answer, but it's often misleading
• the match cM level is influenced by both the number of generations from you to the comman ancestor as the other person to the common ancestor. That is, someone who has 2-3 less generations separating them from that ancestor will look closer compared to other matches to the same person.
• expect that you will always have a lot matches for which that you'll never find a relationship - if progress stalls, move on to another match

If anyone has questions or would like to see more details, let me know.

Tim Conrad
11/10/18

Links

• John Newton Mix on ancestry.com