More Crann DNA

In my previous Blog, I looked at creating genealogical trees using proposed Crann DNA matches. The matches I looked at were primarily at AncestryDNA. The advantage of Ancestry is that if the matches have trees, they are easy to find. Since that Blog, I had a question from Molly in my Newfoundland Dicks DNA Study Group. Molly would like to know if the DNA shows that she is descended from the Cranns. Molly has two lines of Dicks ancestry. One of them leads back to an early Crann connection.

Molly (or Marilyn) and her brother Howie are on a Dicks/Crann line on the right. They are also on a Dicks/Joyce line. My wife’s mom Joan and Joan’s half Aunt Esther are on another Dicks line. In my previous two Blogs, I got around a lot of the non-Crann common ancestors by finding a Crann descendant who moved from England to New Zealand. This tended to isolate matches to just Cranns and made things simpler. In the chart above, a lot of these people are related to each other in multiple ways due to living in isolated areas. Also Joan, Marilyn, Howie and Forrest did not test at AncestryDNA. Fortunately, they are all at Gedmatch.

Here was the match list at AncestryDNA:

These were presumed to have Crann ancestors. I know that Esther, Heather and Ninky are listed at Gedmatch. There are also others.

People Who Match One or Both of Two Kits

Getmatch has a utility where you put in two matching people and others show up that match both those people. This is sort of what AncestryDNA does with their matching feature. I did that for Esther and Heather. From that list, I found some people that match Heather and also match Esther.

Wayne and Marjorie

Wayne and Marjorie show up first on this list. Marjorie is probably the M.R. listed at AncestryDNA with the large tree. Wayne and Marjorie are siblings. If I’m interpreting Marjorie’s tree correctly, I get this:

This adds another New Zealand line in green to compare with the Newfoundlers. Also on that list of common matches are my mother in law, Joan and Molly who asked about her Crann connections. Forrest, who was listed as a Crann/Dicks descendant was not an obvious match to the New Zealand DNA testers. However, she did show a tree match to John Crann. Based on that, I’ll add Forrest, Molly and Howie:

Forrest also had that John Crann’s wife was Elizabeth, so I added that in. Under this tree, it would not be surprising for Forrest to not have a DNA match as Forrest and Heather would be 4th cousins twice removed. Next, I’ll add my mother in law and her half Aunt Esther:

Note that I just corrected Wayne and Marjorie and moved them up one step after getting in touch with Marjorie’s daughter Donna. In all subsequent diagrams, they should be shown as here.

My next step is to take all the tested people in the Tree above and compare their DNA in a spreadsheet, to look for Triangulation Groups (TGs).

Chromosome 10 TG

Here Heather, Wayne, Marjorie, Molly and Howie match each other. Joan and Esther match each other but not the rest of the group, so they are not in the same TG.

Based on the above, this appears to show a common ancestor of Crann for Marilyn. It’s a little surprising as Marilyn and Howie are 6th cousins to Wayne and Marjorie.

TG11: Heather, Esther, and Joan

I suppose Esther and Joan do not want to miss out on this TG which seems to point to Crann in Netherbury, Dorset, England:

This shows that the DNA that Joan and Esther match with Heather got to them somehow. That path had to be through the Upshall or Dicks wife (or both).

TG22: Esther, Heather, Wayne and Marjorie

Here, the match between Marjorie and Wayne are not highlighted as siblings are usually counted as one person in a TG.

It is interesting that after 250 years, the DNA still points to the Dorset, England home of the Cranns via Newfoundland and New Zealand.

These were only a few TGs. I only picked the DNA matches where there were pretty good New Zealand trees. There are probably other New Zealand DNA tested people that triangulate with Newfoundlanders who descend from the Crann family of Dorset.

If Heather wanted to map her Crann DNA based on these four testers, it would look like this:

 

A Dorset, Newfoundland and New Zealand Connection by DNA: Part 2

In my previous Blog I created a proposed tree based on AncestryDNA matches, Gedmatch matches and family trees. First, I created a more solid Crann tree by patching together existing trees. This would be the backbone of the study as there were pretty clear relationships. It looked like this:

One family had Newfoundland roots (Matson). Heather had New Zealand roots. They both matched by DNA and both had an ancestor in Netherbury, Dorset, England. Based on that tree, I added two other trees based on matching DNA:

One was another Crann tree. The other had no know Crann ancestors but a likely Crann DNA match. I felt comfortable doing that for a few reasons. The first reason was the AncestryDNA matches of the people in the proposed tree:

The 2nd and 3rd columns above showed how each person matched my wife’s great Aunt Esther and Heather. Further, Matson was related to Terence. This created a sort of circle. This is my interpretation of how Ancestry does their circles. The fact that there was this circular matching is in my opinion like what many do with Triangulation of DNA matches. This tends to insure that there is a common ancestor. This could also insure that the match of a match is not going far afield.

The other reason why Heather is an important match is that she is from New Zealand. I assume that her Crann ancestors went directly there. That means that I wouldn’t have to take into consideration Newfoundland intermarriages when considering DNA matches with Heather. In other words, I could assume that Heather was related on one line only. Or at least it would be more likely.

The  Elsie Connection

In the last Blog, I looked at the Terrence – Matson connection. They matched each other by DNA. They also matched Esther and Heather at AncestryDNA. In the same way, I would like to look at the Elsie connection. I mentioned in my last Blog that Elsie had 4 people in her tree. That was not totally right. She has 3 people and one is listed twice. I’ll ignore her grandmother as it is the same person she has down as her mother. Perhaps it was at that point that she gave up on her Ancestry Tree.

As before, I create a new tree for Elsie at Ancestry. I called it the Chafe/Hann Tree because at this time, we know of no Cranns in her ancestry. The problem with that is that we will need to build out both sides of Elsie’s ancestry. As I worked back her ancestry I looked at the Ancestry leaf hints. One hint surprised me as it was the first time I’d come across an Upshall in my genetic research. The fact that an Upshall popped up unannounced while I was chasing a probably Crann DNA lead seemed significant to me.

I already knew that there was at least one Hann family living in Harbour Buffet where some of my wife’s ancestor came from. From the Newfoundland’s Grand Banks web site, I find this family in the 1921 census for Little Harbour East:

This gives a month and year for each person’s birth and tells us where they were born. Little Harbour East is not far from Harbour Buffet. Actually, it is even closer than I had now that I have some good information from Devon Griffin:

Unlike the other wrong Little Harbour East I had, the right one is across the bay from Harbour Buffet. Here is the marriage record:

It looks like Jessie was quite young. I wonder who Malinda was. Esther’s middle name is Alinda. A little over a month after this, it looks like the two witnesses wed:

Now it looks like the groom for the previous wedding was a witness. Hmm…

George Upshall

From an Ancestry tree, I did get that Jesse (or Jessie) Upshall’s father was George Upshall. Of course, I don’t see that Ancestry Tree at the moment. I had trouble finding George at Ancestry also, but I appear to have found him at FamilySearch. There, he is shown as marrying in 1896. This only works if this was a second marriage as his proposed daughter Jessie was born about 1890.

This shows that the marriage took place in Little Harbour East and that George was a widower. Both the groom and bride were living in Little Harbour East at the time of the wedding.

Another tree gives George as the father of Melinda Upshall. That leaves me with this tree:

Based on this, I’d like to make a guess as to a new proposed Crann/Upshall Tree:

A New Guess for a Crann/Upshall Tree

This tree supposes that a daughter of John Crann b. 1791 married an Upshall. That Upshall then had at least two sons. One was Henry b. 1841 and one was George b. 1857 shown in purple above. Henry and George could have carried down that Crann DNA to Esther and Elsie. I took out the arrow going from John Crann b. 1791 to the Elizabeth that married Christopher Dicks in red above. However, it now occurs to me that it would be possible that that arrow could still be there as there could be a Crann daughter in both slots – on the Upshall and Dicks side.

Let’s look at my AncestryDNA relationship chart again:

Ancestry thinks by the DNA that Elsie and Esther should be 3rd cousins. My chart has them as 3rd cousins, once removed. Ancestry has Elsie and Heather as 4th cousins. I show them as 5th cousins by the chart. The problem with what I did was that I didn’t follow the Chafe and Hann lines up to eliminate other possible Crann connections. However, I think that my chart gives a plausible solution to the DNA matches. It is satisfying to be able to propose some possible relationships based on logical assumptions after so many years of dealing with genealogical records that just don’t seem to exist in many cases.

Summary and Conclusions

  • This method works well with larger DNA matches
  • I started with a large match where there appeared to be a known common ancestor.
  • Based on that match and known ancestor, I developed trees based on other common DNA matches and common ancestries.
  • This method was helped by a non-Newfoundland match. This resulted in narrowing down the search to one surname.
  • Problems could result if I didn’t get the right surname to begin with
  • Other problems could result by not eliminating other possible genealogical connections
  • I drew a proposed tree to make sure the proposed relationships make sense time-wise. The tree also makes sure the proposed genealogical relationships match the ones proposed by the DNA relationships

 

 

 

 

 

 

 

A Dorset, Newfoundland and New Zealand Connection by DNA

Heather first contacted me through AncestryDNA. This was in December 2015. She lived in New Zealand and matched my wife’s great Aunt Esther. What I know about Esther was that all her ancestors came by way of Newfoundland and her line came from Harbour Buffet. I took a look at Heather’s Ancestry Tree and didn’t see any Newfoundland ancestors. I did see a Crann ancestor she had from Dorset, England. I had heard from someone who thought that Aunt Esther should have some Crann relatives (at least that was my memory – Hann is also a Harbour Buffet name). So I thought that this was interesting. Perhaps we could find an English connection for a Newfoundlander and a New Zealander.

Since that time, I recommended that Heather upload her results to Gedmatch and also join the Newfoundland Gedmatch Facebook group. She did that and I’d like to take a look at the DNA to see what may be found.

Genetic Relatives and Family Trees

First, here is the match between Heather and Esther:

The estimated number of generations to a common ancestor is 3.7. That could mean a few things. However, someone in the 3rd cousin range may be suggested. Here are Heather’s Crann ancestors:

Harold is Heather’s dad, so Samuel Crann would be out 4 generations from Heather.

Here are Esther’s ancestors:

Frederick and Margaret were Esther’s parents. Even at three generations back from Esther, there are a few gaps. Interestingly, looking up the Upshall name on the internet appears to link it to Dorset. The early records of Harbour Buffet mention a Peter and a Thomas Upshall. As a wild stab, I notice these two transcribed Upshall births from the Hazelbury Bryan Parish records in Dorset.

In the original records, Peter is referred to as ‘base-born’. Here is a map of Hazelbury Bryan showing proximity to nearby ports:

Below is Netherbury, where Heather’s ancestor Samuel Crann lived in the early 1800’s. He was a bit to the Southwest of my stab in the dark Upshalls.

Esther and My Mother in Law

My mother in law is Esther’s half niece. She is related through Esther’s paternal side only – barring intermarriage issues.

Here is how Heather matches my mother in law Joan:

This tells us that for these matches, Heather and Joan are matching more on the Upshall side rather than Esther’s Shave side. A slight point of confusion is that Esther has the Dicks family on her father’s and mother’s side.

A Triangulation Group?

It appears that Heather, Esther and Joan are in a Triangulation Group on Chromosome 1. That would mean that Heather, Esther and Joan should have a shared ancestor based on the DNA.

Heather to Esther:

Heather to Joan:

Esther to Joan to complete the triangle:

For some reason, there is a break in the Esther to Joan match. However, clearly there is a triangulation group (TG). That means that at least 2 out of 6 segment matches that Heather has with Esther and Joan are on the Upshall side.

AncestryDNA Shared Matches

So far, I have shown that at least part of the DNA match could be on the Upshall line. I have also shown that some Upshalls lived not too far from some of Heather’s ancestors in Dorset. The next step is to see if there are any Dicks family related to Heather. I have headed up quite a large Newfoundland Dicks family project. If Heather is related to Newfoundland Dicks, perhaps we would have already known that. One easy way to check is to check the Shared Matches at AncestryDNA.

This list shows two 3rd cousins and nine 4th cousins. Even though these are shared matches between Heather and Esther, the relationship shown here is to Esther.  The first person on the list comes up as a probable 3rd cousin to Esther. His name is Terence. When I choose him and then choose his shared matches, I see Heather on Terence’s Shared Match list as a possible 4th cousin. The bad news about Terence’s Ancestry Tree is that there are 4 people on it. The good news is that one of those four, his mother, is listed as a Crann.

Now I feel like I am getting somewhere. I feel like I should be focusing on the Upshall side of Esther’s tree. This, of course, is the side least known:

Esther and heather’s 2nd Shared match

Esther and Heather’s second shared match at AncestryDNA is Elsie. She is a possible 3rd cousin to Esther. Like Terence, Elsie is a possible 4th cousin to Heather. Like Terence she has an Ancestry Tree of 4 people. Her names are Chafe and Hann. Hann is a name that was known to be in Harbour Buffet, Newfoundland where Esther’s parents came from. Elsie also notes some ancestry from Placentia, Newfoundland. Harbour Buffet is in Placentia Bay. The Chafe name sounds familiar to me also.

So far, I have two interesting connections with the first two Shared Matches between Heather and Esther. Too bad my mother in law, Joan, is not also on AncestryDNA. At this point, I could try to work Terence’s and Elsie’s genealogy back up the Crann Line. However, I’ll keep looking at more Shared Matches. Now I’m down to Esther’s 4th cousins.

Fourth Cousin shared matches between Heather and esther

Now we are getting a bit further out. Rather than detailing each Shared Match, I have summarized my findings in a table.

My conclusion from all this is that there is a Crann/Netherbury, Dorset connection between Heather and Esther.

Building a Crann Tree

From here, I could look at the matches of the matches. This could get further out results that I don’t want. So instead, I’ll look to build out Terence’s four person tree. First, I will start a tree with the one Crann that he has in the tree. I find the one Crann in the 1930 Census:

This is helpful as it shows Arline’s parents, their ages, where they were born (Newfoundland) and where their parents were born (also Newfoundland).

Connecting the dots

Hopefully, I’m connecting the right dots. Here is John Crann’s World War II Registration:

This was helpful as it gives his John’s date and place. Here is Trinny Cove about 7 miles across the Bay from Harbour Buffet:

Here is a 1921 Census of Trinny Cove from a Newfoundland Genealogy website:

It looks like there were three houses in Trinny Bay and two of those were Crann houses. The two eldest Cranns were born in Harbor Buffett. I now have two potential sets of parents for John Crann, with my first choice being Richard. I do note that John’s first son was named Richard, so that adds weight to my assumption. In addition, his first daughter was named Julia.

Here’s John and Anastacia’s marriage on 8 Oct 1911, thanks to FamilySearch:

Bell Island is on the NE side of Newfoundland. At about this time I got stuck finding out the parents of John from the bottom up.

Richard crann b. 1856 b. harbour buffet

I thought why not check to see if there is a tree for Richard Crann from the Trinny Cove Census of 1921? I went on to Ancestry and found a familiar sounding Crann. Here was a Samuel Crann. His grandfather was listed as being from Netherbury, Dorset.

The Henry who had all the children was shown as being born in Jean de Baie, Placentia Bay, Newfoundland and dieing at Flat Islands, Newfoundland. Now I have made at least one connection between one Crann who was from Netherbury and moved to Newfoundland.

A Google search helped me find a Crann Families of Newfoundland Web Page:

RICHARD2 CRANN (CRANN1) was born December 1856 in Harbour Buffett, NF, and died December 13, 1941 in Fair Haven, NF.   He married JULIA REID  She was born September 1858 in Little Harbour, NF, and died October 20, 1940 in Fair Haven, NF.

Children of RICHARD CRANN and JULIA REID are:

JOHN BENJAMEN3CRANN

ELIZA CRANN , b.December 1877, Harbour Buffett, NF

ELIZABETH CRANN , b. November 1898, Trinny Cove, NF; m. JOHN PEDDLE; b. Corner Brook NF

What I’ve done:

  • I’ve connected Terrence to Esther and Heather by DNA.
  • I have also connected Terrence’s mom to the Newfoundland Cranns and more specifically a Crann born in Harbour Buffet.
  • I have also found a Crann family showing ancestry to Netherbury, Dorset, England.

What I haven’t done:

  • Haven’t connected the Newfoundland Cranns to Heather’s Cranns in Netherbury
  • Haven’t connected Esther’s family to the Cranns other than by location
  • Haven’t connected Terrence’s ancestor Richard Crann to earlier Cranns

So I’m about halfway there. Let’s look to see if we can connect Heather’s Cranns to this John Crann from Netherbury, Dorset.

John Crann Born 1791

This appears to be the record of John’s baptism in Netherbury:

Here is Heather’s Tree:

Here is a more fleshed out tree I found at Ancestry:

This tree shows that the Newfoundlander John Crann and New Zealander Samuel Crann were brothers. Talk about going different ways! This seems to make Henry Crann b. 1757 the common ancestor of Esther and Heather.

Wild Guessing Time

That’s about it for the research. Now I get to wildly guess and make assumptions. Assumption one is that John Crann’s brothers did not also move to Newfoundland. I have the William above died in England, the first John died young and Samuel went to New Zealand. I’m not sure what happened to the other Robert.

Another Crann Tree

While poking around Ancestry, I found yet another Crann Tree called Henwood. This tree seems to fill in a few blanks:

There are a few things I like about this tree. One thing is that it gives the spread of John Crann’s children from 1817 to 1836. Secondly, it shows a likely tie-in with one of Esther’s matches. Way up above in my Blog I mentioned that Heather, Esther and Terrence had a match with someone named Matson. Part of Matson’s tree looks like this:

It ended with one of Matson’s great grandmothers who was Marina Irene Crann b. 1877. It showed her husband as having died in Harbour Buffet. If I stitch the two trees together at Marina, I get a tree for Matson. While I’m at it, I’ll add Heather in. She’s in green for New Zealand.

Heather and Matson show up at AncestryDNA as 4th cousins with a possible range between 4th and 6th cousins. This tree shows them as 4th cousins, once removed. I’ll now add Terence’s Line and Esther’s Line.

A Proposed Crann Tree

At first, I had Esther’s tree up a generation, but that didn’t give any room for John Crann to have a daughter that could fit into Esther’s tree. Esther has two places where a Crann daughter could fit. One is the wife of an Upshall. The other is Elizabeth, wife of Christopher Dicks. From my diagram above, it looks like Terrence’s missing ancestor would have to be a son of John Crann. Under the above scenario, Esther would be a 4th cousin to Heather and a third cousin once removed to Terrence and Matson.

For a few reasons, I’m favoring the theory that a Crann married an Upshall. I suppose that Catherine Crann who I have in one tree as being born 1821, could have married Upshall in 1840 and had Henry Upshall in 1841.

Summary and Conclusions

  • Comparison of Crann Trees and Crann DNA matches lead to a suggested new Crann Tree
  • By carefully placing the separate trees together, it indicates where the missing Crann in the tree would likely go.
  • If more Crann descendants upload their DNA to Gedmatch, it could help verify these lines through triangulation.
  • It appears that there is more room for analysis. In my Excel spreadsheet of compared matches, there were multiple matches between Ninky, Elsie, Heather and Esther. However, that would make the blog too long.

 

A New Frazer DNA Test on the Stinson Line

It has been a while since there has been a new Frazer test. With the results of Doug’s Aunt Rita in on the Stinson Line, the Frazer DNA Project has roared back into life. Here is the Stinson Line of the Frazers:

I tried to air brush out some of the last names for privacy. Doug and Aunt Rita are on the green line. They share 1,894.6 cM of DNA which is a bit higher than the average posted on the ISOGG Web page:

Note that in comparing those in the chart above, that they may match on the Frazer Line or the Stinson Line. However, in matching those outside this chart, they would be more likely to match on the Frazer Line.

The Numbers, Please

Now it’s time to run the numbers. I found a cool new tool at Gedmatch for doing this. First I added all the Frazers to a group.

Then I just had to choose that group for different analyses. Here is the Autosomal Matrix. This is like ordering the “everything” pizza. This Matrix has ALL the Frazers:

This is also known as the Frazer eye test. I have the Frazer/Stinson group in yellow. These Frazers are only in the Frazer/Stinson Line. Then, there is the green group. They are in the Stinson Line but also in at least one other Frazer Line. There are some good matches where green and yellow intersect, notably with Jane. These are more likely to be Frazer/Stinson Line matches. Where the green intersects with green, it will be difficult to tell if the DNA is from the Frazer/Stinson Line or from another line. The names in white are from my family. I recently had my sister Lori tested. Now there are 5 siblings tested in my family. Paul is a second cousin, once removed. Purple represents the James Line of the Frazers. There is a purple Jonathan and a white Jonathan. Scanning the yellow names from left to right, they match my family a bit more than the James Line Frazers in purple. There are a few exceptions where there are higher matches. This may be due to a match on a collateral line. Or this may be due to the effect that if a match is going to break through a distant relationship it can just as well break through as a somewhat larger match than as a smaller one.

Triangulation Groups

I like to use Triangulation Groups to sort out some of these families. A Triangulation Group points to a particular ancestor which should point to a particular Line of Frazers. The problem in this was alluded to above. That is, what if some Frazers are from more than one line? In order to somewhat get around this, I’ll make special note of the Frazers that are known to be only in the Frazer/Stinson Line. Namely, Rita, Cathy, Ros, Doug and Vivien.

Chromosome 1

In Chromosome 1, I see a Frazer/Stinson  Line Triangulation Group (TG). I previously had my brother Jon in that group, but that appears to have been a mistake:

Here is the what the TG likely looks like:

I say likely, as Jane and Michael also descend from the Richard Frazer Line. However, that would put the common ancestor as Archibald the father of this Archibald Frazer which would be a less likely match.

Chromosome 3

The next TG is also difficult to explain:

Here we have Michael and Rita, both in the Frazer/Stinson Line. They both match Prudence from the somewhat distant James Line of Frazers. I had to double check to find the match between Michael and Rita to finalize the TG. Here is a possible rendering of that TG:

I pulled Michael (pink box) off the Richard Line as he wasn’t put into the Frazer/Stinson Line on this chart to save room. This particular representation forces our attention to the parents of the James (on the right) and Archibald Lines (on the left above). However, a slightly later unknown common collateral line would also be possible. For example, Prudence has a Peyton ancestor. If Michael and Rita also had the father of Prudence’s Peyton ancestor as their ancestor, that would put the top circle one level down. The TG is sure. The interpretation of who the TG represents is not as sure due to holes in the genealogy near the top of the chart.

Chromosome 4

Here is a new TG with Jane, Cathy and Rita:

This TG is important because it is most certainly a Frazer/Stinson Line TG. I had to take down the Gedmatch levels to get the match between Jane and Cathy. If I were to map out the Frazer side grandparents for Rita, Cathy and Jane, they might look something like this:

As you compare Cathy and Jane’s green sections with each other, you can see that the Frazer overlap is relatively small.

Chromosome 5

In Chromosome 5, there is a similar situation with Pat, Cathy and Jane:

Again, I had from my Gedmatch download that Cathy matched Patricia and Cathy matched Jane. What was missing was the Cathy to Jane match. I lowered the Gedmatch thresholds and found a Cathy to Jane match right in the expected area between 73 and 76. Note that the Patricia to Cathy match ended at about 77M. The Jane and Cathy match started about 73.

73 is the crossover for Jane from McBride to Frazer. It is the number that Jane has in common with both her matches. Likewise, Patricia’s crossover from her Frazer grandparent to her Gray grandparent is at 77M. It is the number that she has in common with both her matches. So that tells me something is going on there (i.e. a crossover). The area between 73 and 77 is where Patricia and Jane match.

In reviewing my past work, I see that I had shown a TG with Gladys, Pat and Cathy in this area. To see that, I need to go further up my spreadsheet:

So let’s map out Gladys’ DNA from her grandparents. It looks like her match with Cathy and Patricia tell me that she has Frazer DNA from 54M to 134M. But where is her match with Jane? When I run a one to one match at Gedmatch between Gladys and Jane, I get this:

5 73,514,449 76,680,718 5.3 842

With this information, I’ll draw a revised Chromosome 5 Map:

It was a little tricky to draw this map. I think that what happened was that Patricia and Gladys share the common ancestors: George Frazer and Susannah Price. Notice that Patricia and Gladys have a large match. I think that match is picking up the older crossover in Patricia of George Frazer and Susannah Price at position 77M. Well, you can see I’m still working these things out!

Chromosome 9

Here is a new Frazer/Stinson TG at the end of Chromosome 9:

In it, we have Doug, Rita and Patricia. Interestingly, Ros, Vivien, Gladys and Bill don’t appear to be in this TG. They seem to be busy being related to each other on their non-Frazer sides.

Anything Else, Summary, Conclusions?

  • I found it interesting that Rita matched Prudence. Rita’s match with Prudence was a little larger than any found so far and Rita and Prudence are on the two most distant Frazer Lines of Archibald and James.
  • It is interesting to look at the autosomal matrix for the Frazers as the higher number indicate family groupings. Overlaps in the families where Frazer cousins married cause even higher cMs in the matches.
  • A focus on the Frazer/Stinson TGs helped shore up that line of the Frazers. In cases where a TG could be from one line or the other, the addition of a Frazer/Stinson only Line tester gave more evidence that those TGs were more clearly in the Frazer/Stinson Line.
  • I did some Chromosome mapping based on the TGs. The TGs gave clear indications of crossovers. However, it was not always clear as to which generation we were mapping to as far as specific ancestors.
  • Here is an update of the Frazer TG Matrix:

More Frazer/McPartland DNA and the Mush Move

Since my last post on the Frazer/McPartland connection, one of the McPartland descendants, Charlene, has uploaded her DNA to Gedmatch. Basically, the McPartlands have in their genealogy that they had a Frazer ancestor. This Frazer ancestor is thought to be related to the Frazers that lived in North Roscommon, Ireland. This relationship has been made more sure by the fact that the McPartlands lived near the Frazers and that McPartlands and Frazers have matching DNA.

Here are some of the McPartlands:

I am focusing on the green part as those are the ones that have had their DNA tested and uploaded the results to Gedmatch. The McPartland/Frazer connection is seen at the top where Owen McPartland married Ann Frazer. From this chart, we can also see that Charlene is a 3rd cousin to Karen and Chris.

Charlene’s X Chromosome

When I look at Charlene’s X matches at Gedmatch, I see something very interesting. Her top match is to my sister Heidi:

And here is how Charlene matches her 3rd cousin, Karen:

Notice how close these two matches are. Just to close the loop, here is the huge X match between Heidi and Karen:

The above comparison shows an X triangulation. This, to me, is proof that the three are related.

Here is a possible Frazer McPartland tree which could explain the above X matches:

For this to work well, the Frazer at the top would most likely have two wives. Margaret would have been born from the first wife and from the second wife. The other solution would be to have another generation between Ann and the top Frazer. However, that also introduces problems as the X Chromosome does not travel from father to son. That scenario would require Ann Frazer’s mother to be a Frazer which would mean Ann’s father would also be a Frazer.

Here is a late breaking update on Ann Frazer from a McPartland researcher:

Hi Joel,
Ann was born between 1818 and 1823 (1901 census age 78, and her death registration, also 1901, age 85).
The 1823 date seems likelier, since her last child was born in 1866, and she might well have personally given her age to the census taker, while with the death registration, we’re depending on her son John, with whom she lived, to give the correct date.
All the best,
Sandy
This is good news as it would now not require the Frazer at the top of the tree to have had two wives. If the top Frazer’s wife was born in 1780, she could have had Margaret around 1800 when she was 20 and Ann in 1823 when she was 43. I could narrow that down even a bit further. She could have been born in 1783, had Margaret in 1803 when she was 20 and Ann in 1823 when she was 40.

Who Is the Unknown Frazer? By James Line Genealogy

One way to look at this is through the existing Frazer genealogy. There is the Archibald Line and the James Line. My past assumption has been that this unknown Frazer is in the James Line. But what James Line Frazer would fit the bill? If our genealogy is right, then the sons of James were born to soon to fit the bill of someone born around 1780
JAMES1 FRAZER was born circa 1717 at Aghrafinigan, Ardcarne, Roscommon. He married Katherine Fitzgerald in 1745.
He was a farmer.
Children of James1 Frazer and Katherine Fitzgerald were as follows:

  •       i.   (–?–)2 was born circa 1746.
  •       ii.   ELIZABETH was born circa 1748. She married William Knott.
  •     2.  iii.   ARCHIBALD was born circa 1751. He married Catherine Peyton, daughter of John Peyton and Hannah Wynne, in 1780. He was buried on 13 Aug 1835 at Ardcarne.
  •       iv.   PATRICK was born circa 1755 at Aghrafinigan, Ardcarne, Roscommon. He died in 1831 at Aghnasurn.
  •     3.  v.   MICHAEL was born circa 1764. He married Margaret (–?–). He died on 17 Mar 1837 at Ardcarne.

The children of the above were born too late. The only one who could have fit the bill is John Peyton, son of Archiald But he doesn’t have known Frazer descendants:

JOHN PEYTON3 FRAZER (Archibald2, James1) was born circa 1781. He married Frances Carlton. He died on 22 Nov 1865; aged 84.
Children of John Peyton3 Frazer and Frances Carlton both born at Ardcarne Parish Church, Roscommon, were as follows:

  •       i.   FRANCIS CARLTON4 was baptized on 10 Jun 1824.
  •       ii.   KATHERINE PEYTON was baptized on 20 Dec 1829. She married David Burns, son of Stephen Burns, on 8 Oct 1849.

The Unknown Frazer By Archibald Line Genealogy

By genealogy, it appears that the Archibald Line has more potential for our mysterious link between the Frazers and McPartlands.

Our genealogy has these four Archibald Line brothers born around 1780. That could put any of them as potential candidate to be Ann and/or Margaret’s father.

A Little McPartland Genealogy

Here is the 27 October 1860 Baptismal record that Joanna (a Frazer researcher) found in the Aghanah Catholic Parish record:

This is a record of the birth of Cath. Janam (Jane) [Mc]Partland daughter of Eugene and Ann Frazer. It appears that a Patrick Partland and a Healy were there and that the family lived in Annagh. Annagh is a very popular place name in Ireland. However, the closest Annagh to Dereenagan appears to be here on the shores of Lough Key:

Annagh is in the lower left of the map above. This is a bit confusing as one branch of the Frazer family had an Annagh House (or Lodge) in the Townland of Aughnacloy in County Sligo. To further confuse things, an Ireland Townlands website has Annagh further to the East and North:

It looks like my first choice may have been the better one. Here is a Roman Catholic Parish map of Ballinafad which historically was Aghanagh. This map was taken from the Leitrim-Roscommon Genealogy web page.

Still, I wouldn’t be surprised if there was another explanation! It gets confusing with overlapping parishes. I think that the Civil Parishes are equivalent to the Church of Ireland Parishes. The NLI website has Aughanagh Parish in County Sligo.

Back to the DNA

Above, I established that Heidi, Charlene and Karen all matched on the X Chromosome. It looks like my sister Sharon also matches. Here is how Charlene matches Heidi, Karen, and Sharon on the X Chromosome:

Back to Autosomal dna

Charlene matches Karen autosomally. They are known 3rd cousins. Charlene and Heidi don’t match autosomally. Charlene and my sister Sharon do match on Chromosome 9 from position 22-36M:

Here is Sharon’s Chromosome 9 map showing why she matches Charlene and Heidi and I don’t:

Lighter red is Frazer DNA. Sharon got a full load of that on her Chromosome 9. Heidi and I got nearly all Hartley DNA (darker red) on this Chromosome.

triangulation groups

With the McPartland/Frazer matches, I would like to focus on triangulation groups as they have a good chance of indicating a common ancestor.

Here are Charlene’s matches with my cousin Paul and my sister Sharon on Chromosome 9:

According to my last Blog, Karen and Chris also had matches in the same area. This indicates a common ancestor:

I have the Frazer at the top with a question mark as we are not sure which Frazer this is. It is quite likely to be a correct scenario. I base that on the combination of X Pattern matches and the triangulation. Although the match shows with Heidi, the actual match in this case was with my sister Sharon. Also note that Paul had no X match as he has two male Frazers above him and X does not travel from male to male.

The prudence triangulation group (TG)

Here Charlene matches Prudence, Chris and Karen:

#4 is Betty who is a cousin of Joanna’s family. From my last Blog, Prudence also matched Karen and Chris to complete the triangle. Here is Prudence’s tree.

How can we fit the McPartlands in here? Prudence is on the James line. Earlier in the Blog, I had mentioned John Peyton Frazer as a not likely person of interest. I’ll re-consider him here:

What if John had two other daughters: Margaret and Ann? They would fit in. The problem with this is that I have a baptism of and Ann and a Hannah already in 1823 that may fit the bill. They were born to an Archibald and a James Frazer respectively. There are many possibilities. One would be that the match is through the Peyton side.

The Anne above had a White mother. Perhaps Catherine Peyton had a sister than married a White and had Anne? What if the James above was an unknown James Line descendant? He would fit the bill also.

Here is the simple portrayal of my first scenario:

Here I just mushed the two trees together where this could be seen. In this scenario, The Frazer on the McPartland part would be John Peyton Frazer (or perhaps the James that I know little about). That would make Prudence, Charlene, Karen and Chris 4th cousins, once removed. Under this scenario, Paul and my family fall out from DNA matching as we are a generation or two below the McPartlands.

Joanna’s TG – Chromosome 15

This figure shows Charlene’s matches with Joanna and her sister Janet. After that is Joanna’s cousin Betty. Finally is my cousin Paul. The green section is not a TG as Janet and her sister only count as one in a TG. The yellow section is a TG. Paul does not appear to be in the TG. Is this showing us that Paul’s match with Charlene is on a different line than Joanna’s TG?

Time for my two family mush move (again for illustration purposes only):

This would be a scenario similar to the Prudence TG above (except I forgot to add the ?????). In this scenario the relationship would be 3rd cousin twice removed. This is slightly closer than the 4th cousin once removed Prudence TG scenario.

Summary and Conclusions

  • The addition of Charlene’s McPartland/Frazer DNA to gedmatch has resulted in some interesting comparisons
  • The X match between the McPartlands and my family gives a strong indication of a match along the Frazer line.
  • This X match has also been backed up by a four way Triangulation group between two McPartland families and two Frazer families (my family and my cousin Paul)
  • I used a new (to me) technique called the mush move. This is where in a TG I mush the two trees together.
  • Even thought the connection was not proven, it gives an indication of where the connection likely is.
  • In the comparison between the Prudence TG and the Joanna TG, it shows that Joanna’s higher level of DNA matches are the result of a likely closer relationship with the McPartlands.
  • The mush move technique is helpful in seeing possible links between families as well as possible problems of links between families.
  • In the beginning of the Blog, I was favoring an Archibald link between the McPartlands and the Frazers. At the end of the Blog, I was favoring the James Line. It seems like the existing paper genealogy favors Archibald Line and that the DNA seems to favor a James Line connection. One possible way to reconcile the two would be to have James (wife of Margaret) be the missing James Line Frazer. In that scenario James’ eldest daughter Margaret (my ancestor) would have been named after her mother which would have been traditional.
  • Perhaps more research will bring something to light concerning James Frazer who married Margaret and had Hannah in 1823.

Determining Whether a Match Is Irish Or French Canadian By Visual Phasing

In this Blog I will look at a DNA match that my in-laws have. I would like to know whether the match is Irish or French Canadian. I will use Visual Phasing of my father in law and his two sisters’ DNA match to try to figure that out.

Irish at First Look

Something caught my attention with one of my father in law’s matches at FTDNA. My father in law Richard’s match Ann had this tantalizing detail under her Ancestral Surnames:

White (County Waterford Ireland to New Brunswick Canada)

I had recently found out, with the help of DNA and DNA researchers, that my father in law’s immigrant ancestor had shipped out from Waterford to New Brunswick. I have very few DNA matches for my father in law on this Irish side that I have identified. Most of the matches are French Canadian.

Irish or French Canadian?

At first, I didn’t notice other French Canadian names in Ann’s ancestry. However, after finding out she was listed at Gedmatch and Ancestry, I looked at her Tree and did see some French Canadians.

Visual Phasing

I do have DNA from my father in law Richard and his two sisters Lorraine and Virginia. So perhaps Visual Phasing will give and answer to the question whether the match with Ann is French Canadian or Irish. Ann’s best match to Richard, Lorraine and Virginia is on Chromosome 9:

Lorraine has the largest match above followed by Richard and Virginia. It looks like Richard and Virginia have crossovers at about position 107M.

I have used MS Word for phasing, but it wasn’t the best. PowerPoint worked well, but lately I have preferred using Excel. First I cut and paste the comparison of the my 3 in-laws into Excel.

Then I add the crossover points for the three siblings:

At first I thought that the first crossover belonged to Richard. however, there is a short break in the Lorraine V. Virginia comparison, so that adds an additional first crossover for Virginia. Actually the Virginia/Richard should be Virginia/Lorraine. There are likely 2 close crossovers there. I ignored the last small match between Lorraine and Virginia as there wasn’t anything going on in the comparisons above and below that match. Next I add the locations of the crossovers:

Lorraine and Richard have the largest Fully Identical Region (FIR) shown in green. I map that with the same two colors for Lorraine and Richard:

Lorraine only has two crossovers, so we extend her colors all the way to her left crossover and on the right to her crossover (L):

As Lorraine only had two crossovers, this perhaps explains why she had the largest match with Ann on Chromosome 9. Next, I fill in FIRs and Regions that don’t match (shown as red in the Gedmatch comparisons) with corresponding colors:

Unfortunately, that lead to a bit of a dead end. Instead, I’ll try starting with the Richard and Virginia FIR on the bottom comparison:

This version looks better. Next we choose a Half Identical Region (HIR) shown as yellow above. The longest one starts at position 14 between Lorraine and Virginia. A HIR maps as matching only one color and not matching the other.

Above, I chose for Lorraine and Virginia to match on the green and not match on purple and yellow. That is how the HIR is represented. I can then extend Lorraine’s purple and green to her crossover (L) on the right and fill in more FIRs and non-matching areas:

Now, except for the two ends of Virginia and Richard, I have a four grandparent map represented by four colors. Next, we have to identify the grandparents.

The Pouliot French Canadian Connection

One of my in-laws’ grandparents is a French Canadian Pouliot. Fortunately, my in-laws have a Pouliot cousin named Fred. Fred’s sister has also tested. Here is Fred’s matches with Virginia (78-83.5 and 107-110) and Richard (107-115).

Here is Fred’s sister’s matches with Virginia, Richard, and Lorraine.

Note that Lorraine only has one small match with Fred’s Pouliot sister. This is leading me to believe that the match with Ann is on the Irish side. Can we use these Pouliot matches to identify our blank map above? I think we can. The 2 green matches above are for Virginia and Richard at 17-31M. The only place between 17 and 31 where Fred’s sister could match Virginia and Richard, but not Lorraine is on the yellow. If the match were on the green segments, Fred’s sister would have had to have matched all three siblings at that location. Note that mapping out the smaller matches should also be on the yellow segments.

I should point out that my in-law’s had a father of Irish descent and mother of French Canadian descent. This means that both their paternal grandparents were Irish and both their maternal grandparents were French Canadian. As Pouliot is the maternal grandfather, that sets the maternal side of the map as yellow and purple. That also sets purple as the other maternal grandparent: LeFevre. Further, salmon and green now represent the paternal Irish grandparents.

So Is Ann a French Canadian or Irish Match?

Although I was leaning toward the Irish earlier, I now think that the match is French Canadian. Take another look at the match between Ann and Lorraine, Richard and Virginia:

The pattern looks a lot like the purple LeFevre segments. Lorraine’s larger match is on top. Richard’s green match stops where the purple LeFevre segment stops. Virginia’s smaller blue match starts where the purple Lefevre segment starts again. I’ll put the matches in the same order as Gedmatch to make it easier to see:

If Ann were to have matched on the green paternal grandparent area, there would have have to have been three equal matches in that region shown on the Gedmatch browser.

The fact that Ann did not match with the French Canadian Pouliot grandparent did not mean that she was an Irish match. In this case, it meant that she matched the other French Canadian Grandparent.

Summary and Conclusions

  • Visual Phasing can help map an unknown match to a grandparent.
  • That phasing needs to be in conjunction with at least one known cousin to identify a grandparent.
  • These results help to know where to invest genealogical research time. There is no sense in barking up the wrong tree.

An Updated Z17911 Hartley STR Tree

In my last Blog on the subject, I wrote about a Hartley Z17911 STR Tree. Since that time, I created a broader Z17911 STR Tree. However, that broader tree was not the best idea. Soon after creating that tree, I found out that at least one person in that tree was actually in a new SNP group further downstream from Z17911. This was based on Big Y and SNP testing. Within not too long from creating my tree, the SNP tree as created by Jared Smith went from this:

to this:

The link to Jared’s Website is here.

So, while Goff appeared previously to be in my SNP group, in fact, he was not. He was as far as 4 SNPs away. That means that any closeness in STRs could have been coincidental. When comparing SNPs and STRs, the rule is that SNPs take precedence.

A STR Tree for Hartleys Only

At this point, it seems to make sense to create a Hartley only STR tree. There is still no guarantee that Hartleys that are related to me by STRs will have the same SNP results as me. However, I think that it is more likely than not that they will.

Since my previous Blog, there have been two new Hartley STR testers. I have the results for one of those that tested at 67 STRs and one I don’t have results for yet who tested at 111 STRs. Previously, there was one other Hartley testing at 111 STRs. I have had my STRs tested indirectly through the BigY test. YFull analyzed 500 of my STRs – although some of the results were inconclusive. That means that there are three Hartleys with about 111 STRs tested, but I only have the results for two. I should be able to create a very simple tree from that.

The First Ever Hartley 111 STR Tree

At least I think it is the first. Those in the group I’ll call West Yorkshire Hartley,  and me. My ancestors are from Lancashire, so I’ll be Lancashire Hartley. I think that this will be interesting as I feel that the Lancashire Hartleys predated the Hartleys for West Yorkshire. However, I get the impression that my Hartley YDNA administrator favors an earlier date for the West Yorkshire Hartleys. Here are the differences in 111 STRs between a West Yorkshire Hartley and a Lancashire Hartley:

There are a few interesting things from the numbers above:

  • The 16357 Mode is the SNP above Z17911, so it would be older.
  • STR 449 could be a back mutation. It goes from 32 to 31 and back to 32 for West Yorkshire Hartley.
  • The 455 STR has an orange number above it. That refers to the slowest STR mutation rate. As that is the slowest STR rate and my result is the same as the 455 modes, I infer that my STR test represents the older Hartley version. However, a sample of 2 is not much.
  • I am a GD of 14 from the West Yorkshire Hartley.
  • Both the West Yorkshire and the Lancashire Hartley are a GD of 7 from the Z17911 mode. That would have given us a tie for the oldest STR profile if we hadn’t considered the effect of mutation rates.
The simple 111 STR Hartley tree

This Tree is a bit on the conceptual side. However, it does point out some things:

  • These two Hartleys likely descend from a common Hartley. However, at this stage, we don’t have the 111 STR Mode for that common Hartley.
  • The STR mutations are therefor shown to Z17911 rather than to a common Hartley.
  • As mentioned above, I favor the theory that the West Yorkshire Hartley Line originated in Lancashire. This is partly based on something called the founder effect. That means that due to the large number of Hartleys in the Colne/Trawden area, it is possible that the area was a founding area for the Hartleys. However, the distance between the Lancashire and West Yorkshire Hartleys is not far.
  • I did not include all the STRs for simplicity. The slowest marker is shown in orange.
  • The three last slower moving STRs (540, 445 and 1B07) are in the 111 panel, so will not show up in the 67 STR analysis.
  • I have the year of 1075 (125 years per STR mutation) shown above. This is supposed to represent a difference of 7 GD. However, I don’t know if that date should represent the Hartley Mode or the Z179111 Mode. If the date were to represent the Hartley mode, then that would likely be at the beginning of when Surnames were beginning to come into use.
  • As the overall GD difference between the two Hartleys is 14, I don’t see how the difference to a common Hartley ancestor could be less than 7.
  • There is also the possibility that these two Hartleys had a common ancestor just before the implementation of surnames and that due to this relationship, common area of origin or by coincidence they both took on the Hartley surname

Back to 67 STRs

Let’s keep the above tree in mind as we get down to the six Hartleys with 67 STRs tested. Checking the tree I made in a previous Blog, I see that Lancashire Hartley (me) and West Yorkshire Hartley were at opposite sides of that Tree:

In the above tree, Hartley #2 is the same  as West Yorkshire Hartley.

The New 67 STR Hartley Tree

The Hartley we want to add is believed to have Quaker roots in Lancashire in the 1600’s. He also is taking a Big Y test which is exciting. The results for that exploratory YDNA test will likely show us the first Hartley family SNP. I currently have many private SNPs. However, once the Quaker Hartley tests, his SNPs that are in common with my now private SNPs should become the new Hartley family SNPs. Here are the new Hartley 67 STR results:

  • Due to the fact that there are now 6 Hartley results, this causes there to be a tie in some of the modes. In these cases, shown with a 3 in the bottom row, I used the older values. This ended up in also being the lower values.
  • I chose to make a split on STR 455. This STR has the lowest mutation rate of those in the table. I didn’t think it likely that these last three results would have mutated independently.
  • This split also separates the two Lancashire Hartleys from the two West Yorkshire Hartleys
  • Again, the Lancashire Hartleys tend to be the older group as they are closer to the Hartley mode by one GD (STR difference).
  • For these markers the Z17911 Mode is identical with the Hartley Mode. This suggests that Hartley is an old Surname.  This result agrees with the 111 STR analysis above.

A New 67 STR Hartley Tree

Here is my interpretation of the above data in a tree form:

  • The Hartley Mode results are shown in 2 boxes at the top of the Tree. This is meant to represent a common Hartley signature or the signature of a common Hartley ancestor in the distant past.
  • I split the two branches at the top based on the slow moving STR 455. These two branches appear to represent a Lancashire Hartley Branch and West Yorkshire Hartley Branch
  • On the Lancashire side, Sanchez and Joel are together due to their STR similarities
  • Similarly, Hartley #3 and Bradford West Yorkshire Hartley are together as due to their similarities
  • It appears that the Quaker Hartley’s mutations happened between the Quaker Ancestor and our Hartley tester. However, these mutation would be spread out up to the common Hartley Lancashire ancestor. The same would be true for the Hartley tester with the West Yorkshire ancestor William Hartley. However, his mutations would be spread out up to a common West Yorkshire ancestor under the above scenario.
  • Based on the above point, the Quaker Anc. and Wm. Anc. boxes in the Tree above are not really needed.
  • An early split between these two branches could explain the parallel mutations. For example, Sanchez and W Yorkshire William both have double mutations at location 398b. However, they are shown in different branches and not grouped together. Under my scenario, these two double mutation would have happened independently over a long period of time.
  • Unique mutations are in bold italics.
  • Adding the mutations up the tree gives the GD to the Hartley mode. The double mutations must be counted as two.
  • A rough guess for dating the tree would have the Hartley mode at 1100. The split between Lancashire and West Yorkshire at 1300. The further divisions around 1500. These dates are give or take 100 years or so. The bottom line represents tested Hartleys living today.

Here is the streamlined version of the new Hartley Z17911 Tree with some rough guesses on timeframes:

Summary and Conclusions

  • There would be other ways to draw the 67 STR Hartley Tree. This one seemed most logical to me.
  • The addition of a new Hartley 67 STR tests helped to define a Hartley ancestral mode. It appears to have defined a Lancashire and West Yorkshire branch of Hartleys
  • A pending BigY test should result in one or more Hartley Family SNPs.
  • It is possible that there are unique SNPs for the two Hartley branches shown as coming from Lancashire or West Yorkshire. However, it may take a BigY test from a Hartley from the West Yorkshire Branch to confirm this.