Reading a Parish in Its Descendants’ DNA
How four marriage records from one County Monaghan register were tested against the DNA of more than forty-five descendants: the records and what they leave out, the family-by-family match evidence, the rebuilding of a father’s paternal DNA, the chromosome 2 cluster, the Y-DNA line, and how each finding was weighed.
The Research Question
Were the four couples who married in Donaghmoyne parish between 1841 and 1858, each with a Hamill spouse, members of one family, and if so, how were they related?
The question grew out of the Owen Hamall case study. Owen’s parents, Henry Hamill and Mary McMahon, married in Donaghmoyne in 1841. Among the Hamall family’s DNA matches were descendants of three other Donaghmoyne couples of the same generation, and the same surname runs through all four marriages.
Two levels of answer are possible, and this case study keeps them apart. The first is network level: do these families descend from one interrelated parish population? The second is relationship level: were the four Hamill spouses brothers and a sister, cousins, or more distant kin? DNA at this generational distance can answer the first. The second needs records from before the parish registers begin.
The Four Marriage Records
Four entries, one register, seventeen years. They place four couples in one parish in one generation, and say nothing more.
| Couple | Year | Register ref. | Earliest child in the same register | Settled |
|---|---|---|---|---|
| Henry Hamill & Mary McMahon | 1841 | 91 | Mary, 1847 | Montreal by 1851; Chicago |
| Owen Hamill & Ann King | 1846 | 398 | Mary, 1846 | Wisconsin; Ann in Nebraska by 1880 |
| Charles McKenna & Susan Hamill | 1857 | 811 | Pat, 1858 | Joliet, Illinois (as McCanna) |
| James Hamill & Ann Gartlan | 1858 | 841 | Anne, 1864, and six more to 1878 | Stayed in Dian; children to Montana, Missouri |
What the entries give
Each is known from the rootsireland.ie index to the Donaghmoyne Catholic registers. Each gives the couple, the parish, and the year. The date shows as “1 January,” which is the index’s placeholder for a year with no day recorded; the notes read “Exact date not given. Priest’s name not given.” Parents, witnesses, ages and townlands are not transcribed. Reading the original register pages for those fields is an open task.
The registers survive from about 1840. The parents of these four couples, born roughly 1790 to 1815, married before then. Some of the earliest entries for these families turn up in the neighbouring Inishkeen registers, which begin a few years earlier (baptisms 1836, marriages 1839) but are spotty and difficult to read.
James Hamill of Dian
More than one James Hamill appears in mid-century Monaghan. The James of the 1858 marriage is identified with James Hamill of Dian (1827–1914) by four record types that agree: the marriage; the civil registration of daughter Anne’s birth in 1864, naming the father as “James Hamill, of Dien, Farmer” and the mother as “Anne Hamill, formerly Gartlan”; Griffith’s Valuation of 1861, which lists one James Hamill in Dian; and son Peter Hamill’s 1949 death certificate, which names James Hamill and Ann Gartlan as his parents.
Four couples with one surname in one register is co-residence, not kinship. None of the entries names a parent. Whether any couple is related to any other rests on the DNA, read against these records, and on pre-Famine sources still to be searched.
How the Research Unfolded
Each round of evidence answered the question at one level and showed why it could not reach the next.
Hamall case
One marriage becomes four
Tracing Owen Hamall’s parents to their 1841 Donaghmoyne marriage put the family in a specific parish. Among the family’s matches were descendants of Owen Hamill and Ann King in Wisconsin, Susan Hamill McCanna in Joliet, and James Hamill and Ann Gartlan of Dian, whose marriages sit in the same register.
The Hamall father line, tested
A Big Y-700 test of a male-line Hamall descendant, confirmed by his son, placed the line on haplogroup E-FTD95657. No other man in the Hamill Surname Project shares it.
Four families, measured
Matches were organized into four family clusters, one per couple. Within each family the DNA confirmed the documented descent. Across families the matches ran 8 to 34 cM, levels that suggest kinship but can arise by chance. The analysis concluded that segment triangulation and Y-DNA would be needed to go further, and that the Chicago line, with no paternal cousins to compare, was the main obstacle.
Rebuilding the father, reading chromosome 2
The researcher’s father, Thomas Kenny Hamall, was reconstructed from his children’s DNA. With his rebuilt paternal kit as a comparator, a cluster of matches on chromosome 2 was identified, triangulated across platforms, and confirmed as coming through the Hamall side. Tree work on the cluster named one common ancestor, Peter McGeough (1788–1869), and tied one descendant of Susan Hamill McCanna directly to the Hamall paternal line.
The Four Family Clusters
Before asking whether the families were related to each other, each family’s own testers had to be shown to descend from the couple on record. They do. Matches within each family fall where the documented relationships predict.
| Family | Testers | Children’s lines | Matches within the family | Matches to the Chicago line |
|---|---|---|---|---|
| Henry Hamill & Mary McMahon Chicago | 9 | 2 independent: Mary Ann Hamill Byron; Owen Hamall | 15–20 cM between the two lines | — |
| Owen Hamill & Ann King Wisconsin, Nebraska | 8+ | 3–4 | 52–74 cM between children’s lines | 21–34 cM (23andMe); 10–17 cM (Ancestry) |
| Charles McCanna & Susan Hamill Joliet | 9 | 3: Mary Anna McCanna Kelly; Ellen “Nellie” McCanna Sheridan; Frank McCanna | 27–67 cM between children’s lines | 8–15 cM, a few to 23 cM |
| James Hamill & Ann Gartlan Montana, Missouri, Ireland | 14 | 5: James, Patrick, Henry, Anna, Bridget | 23–228 cM between children’s lines | Indirect only, through shared matches |
Four testers descend from Owen’s sister Mary Ann Hamill Byron; three of them are a father and two daughters, so they count as two independent data points. Five testers descend from Owen himself, but all five are children of Thomas Kenny Hamall and share one father’s DNA between them. The direct line narrowed to one surviving son in each of three generations, so there are no paternal first, second or third cousins to compare. Conventional triangulation could not work on this line. Step Six describes how that was overcome.
Matches Across the Families
The pattern is consistent. The amounts are small. Both facts matter.
Chicago and Wisconsin
On 23andMe, M.H.M. of the Chicago line shares 21 to 34 cM with descendants of three different children of Owen Hamill and Ann King. On Ancestry the same pairing runs 10 to 17 cM, again across several branches. Matching more than one branch of a family, on two platforms, is what close kinship between Henry and Owen would produce. Amounts in this range can also arise by chance, so on their own these matches are suggestive and no more.
| 23andMe, cM | M.H.M. Chicago | R.V.E. Engel | L.L. Bucklin | L.F. H.P. Hammel | C.F. Gartlan |
|---|---|---|---|---|---|
| M.H.M. (Henry & Mary) | — | 34 | 33 | 21 | 23 |
| R.V.E. (Anna Hammel Engel) | 34 | — | 74 | 26 | 16 |
| L.L. (Mary E. Hammel Bucklin) | 33 | 74 | — | 52 | 16 |
| L.F. (Henry Patrick Hammel) | 21 | 26 | 52 | — | 0 |
| C.F. (James & Ann Gartlan) | 23 | 16 | 16 | 0 | — |
Testers are shown by initials; the ancestor each descends from is in parentheses.
Chicago and Joliet
Two Chicago siblings match McCanna descendants in all three of Susan and Charles’s documented children’s lines, mostly at 8 to 17 cM and up to 23 cM. Within the McCanna family the same testers share 27 to 369 cM between cousin lines, as expected. A separate tester, D.G., a descendant of Catherine Hammill Linstead, shares 28 to 32 cM with descendants of Nellie McCanna Sheridan, a stronger link still under investigation.
Chicago and Dian
No direct match of weight connects the Chicago testers to the James Hamill and Ann Gartlan family by conventional matching; the connection appears through shared matches. C.F., a Gartlan-line descendant, shares 16 to 23 cM with Chicago and Wisconsin testers.
| Shared amount | How it was read |
|---|---|
| 100 cM and over | A relationship within about four or five generations, almost always inherited from a common ancestor. |
| 30–99 cM | Probably related, the common ancestor perhaps five to seven generations back. |
| 15–29 cM | Possible distant kinship; can occur between unrelated people, especially on a single segment. |
| 8–14 cM | Frequently shared by chance. Not evidence of relationship without triangulation, shared matches or documents. |
In an intermarried parish population, descendants can share DNA through several ancestral lines at once, which inflates the totals. Cross-family amounts in this case are therefore read as evidence that the families come from one community, not as measures of how closely any two founders were related.
Rebuilding Thomas Kenny Hamall’s Paternal DNA
The man whose DNA would answer the question died without testing. Four of his children, read against their mother, could stand in for him.
The method
Each child inherits half of their DNA from each parent. Comparing each child’s kit with their mother’s and setting aside what came from her leaves that child’s paternal share. Merging four children’s paternal shares gives a composite of their father’s genome. The reconstruction was made with the Borland Genetics workflow from three 23andMe kits and one FamilyTreeDNA kit, with the mother’s kit as the phasing reference, and loaded to GEDmatch as a research kit.
The test
The fifth child, E., tested at Ancestry and was not used to build the kit. Her comparison against it is therefore independent: whatever she shares with it came from the same father, not from the algorithm. Adding her would have filled more of the rebuilt genome. She was kept out so the reconstruction could be tested.
Four children typically capture 75 to 90 percent of a parent’s genome. A match that shares DNA with Thomas Kenny Hamall only in the stretches that went to E. will not show against the kit. GEDmatch also allows research kits one-to-one comparisons only, so candidates have to be found through the children’s own match lists first.
| Child | Platform | Shared cM | Coverage | Estimated relationship |
|---|---|---|---|---|
| A. | FamilyTreeDNA (input) | 3,585.0 | 99.92% | Parent–child |
| C. | 23andMe (input) | 3,454.1 | 96.27% | Parent–child |
| K. | 23andMe (input) | 3,146.6 | 87.70% | Parent–child |
| M.H.M. | 23andMe (input) | 3,082.1 | 85.90% | Parent–child |
| E. | Ancestry (independent) | 2,500.8 | 69.71% | Parent–child |
GEDmatch one-to-one comparisons against the rebuilt kit. E.’s lower figures reflect the smaller overlap between the Ancestry and 23andMe chips (148,461 SNPs compared, against about 435,000 for her siblings), not a weaker relationship.
Any match can now be compared directly with the father. DNA a match shares with the rebuilt kit came down the Hamall side, not through the researcher’s mother’s family. That single step turned cluster patterns into paternal-line evidence.
The Chromosome 2 Cluster
Found four ways, triangulated on three platforms, and traced through the father before any claim was made.
Found independently
The cluster surfaced through 23andMe’s relatives-in-common tool anchored on the largest cluster match, through GEDmatch’s segment mapper on the top fifteen matches, through GEDmatch’s match-both-kits search, and through MyHeritage’s chromosome browser. Each pointed to the same stretch of chromosome 2.
Triangulated
MyHeritage’s triangulation tool confirmed two overlapping groups, nine matches in all, sharing one segment with the researcher and with each other. The smallest region all three platforms detect, the conservative core, runs from 44,599,253 to 50,854,610 on chromosome 2, about 6.25 Mb.
Traced through the father
Triangulation shows a shared segment; it does not show which parent it came from. Ten matches compared one-to-one with the rebuilt kit share DNA with Thomas Kenny Hamall directly: M.S., P.H., B.R., D.O., T.S., S.M., M.R., J.L., A.D. and J.S. P.H. shares three separate segments with him, on chromosomes 2, 4 and 16, which points to a closer relationship than a single segment would.
Three groups within the cluster
Segment positions on 23andMe separate the cluster into three groups, and 23andMe’s own clustering of shared matches, working independently, sorts the matches the same way: an older Hamill group shared by Donaghmoyne, Wisconsin and Michigan Hamill branches; a Louth-side group carrying Trainor, Gartlan and related names; and a Donaghmoyne core.
| MyHeritage triangulation | Segment | Matches |
|---|---|---|
| Group 1 | 44,343,676–51,515,672 9.4 cM | Six |
| Group 2 (inside Group 1) | 44,599,253–50,854,610 8.9 cM | Seven |
The surname signature
Independently of the DNA, the documented ancestors of the cluster matches carry a short list of recurring surnames, led by Hamill, McGeough, Gartlan and Finegan, concentrated in southeast Monaghan and neighboring Louth. That geographic concentration is what an endogamous parish population looks like, and not what broad Irish ancestry would produce.
Several Hamill-surnamed testers share nothing with the rebuilt kit at 7 cM, and their trees run to Antrim, Down or Tyrone Hamill families. The name is necessary for the Hamill branch of the cluster and is not sufficient.
A named common ancestor: Peter McGeough (1788–1869)
Three matches, J.C., G.W. and B.U., share a segment with identical end points, 36,862,239 to 59,041,432, each 21.09 cM. Identical boundaries in three people point to recent shared descent. Their trees converge on Peter McGeough of Donaghmoyne: G.W. and B.U. through one granddaughter line, as first cousins once removed; J.C. through a separate branch. Two further cluster matches descend from Peter’s son Michael McGeough (1806–1881). He is the first specific common ancestor identified in the cluster. His wife’s identity differs between trees and remains open.
A founding couple tied to the Hamall line: M.S.
M.S. (1925–2019) descends from Charles McCanna and Susan Hamill (1835–1917) through their daughter Ellen “Nellie” McCanna Sheridan. Compared with the rebuilt kit, she shares a 21.8 cM segment with Thomas Kenny Hamall on chromosome 4. This is the first segment-level evidence linking one of the four founding couples directly to the Hamall paternal line. Her tree names Susan’s father as James Hamill; the source behind that attribution has not yet been seen.
The Y-DNA Line
Autosomal DNA follows every ancestor. Y-DNA follows one: father to son, the line that carries the surname.
A male-line descendant of Owen Hamall, K.H., took the Big Y-700 test in February 2024; his son A.H. confirmed the result, matching at every marker level with three differences in 669, as expected for one father-to-son step. The terminal haplogroup is E-FTD95657, a branch of E-M35, a lineage of Mediterranean and North African deep origin. No one outside the two of them yet tests on this branch.
The Hamill Surname Project has about thirty tested Hamill, Hammill, Hamel and Hamall men. Every other line is an R haplogroup (chiefly R-M222, R-M269, R-S673 or R-S588) or an I haplogroup. The Hamall line is the project’s only E line. Its father-line ancestor shares no common paternal ancestor with any other tested Hamill man within the genealogical era.
If Henry Hamill, Owen Hamill of Wisconsin, James Hamill of Dian and Susan Hamill’s father were brothers in the male line, a direct male-line descendant of any of the others will also test E-FTD95657. An R or I result would rule out father-line kinship for that family, while leaving the autosomal network untouched; the connection would then run through mothers and grandmothers. One Y-67 or Big Y-700 test from a male-line Hammel of Wisconsin, a male-line Hamill of the Dian family, or a male-line relative of Susan would answer the question for that line.
It cannot place the family in a townland, and it cannot say how an E line came to carry the Hamill name: adoption of the name, an unrecorded event in the male line, or a separately named family that took it are all consistent with the result.
How the Evidence Was Weighed
Each finding is placed in one of five tiers according to what supports it. The tiers separate what the DNA shows about the community from what it can and cannot say about particular relationships.
Segment triangulation, checked on more than one platform and traced through the rebuilt kit or a father–son test.
Records that state the relationship, with DNA at the expected levels.
Consistent DNA and documentary evidence that does not yet decide the specific relationship.
Threshold matches or a documentary lead not yet tested.
Visible only through shared-match lists, without segment data.
Several cluster members’ online trees carry the same ancestor back to the 1650s, including Hamill-surnamed testers who share no DNA with the rebuilt kit. That pattern suggests a shared published genealogy copied from tree to tree, not independent documentation. Attributions before 1800 in testers’ trees are treated as unverified unless confirmed from primary sources.
Conflicts and Unsettled Readings
McKenna or McCanna
The record saysThe 1857 marriage and son Pat’s 1858 baptism both read McKenna in the Donaghmoyne index. The family appears in Joliet as McCanna.
WeighedTwo Irish entries a year apart agree, so McCanna is most likely an American spelling. Both indexes could share a transcriber’s habit; the original register page will settle it.
Two Owens
The record saysOwen Hamill married Ann King in Donaghmoyne in 1846 and died in Wisconsin in 1858. Owen Hamall, son of Henry and Mary, was born in the 1840s and died in Chicago in 1898.
WeighedTwo men of one name, parish and generation, with separate records on two continents. Published trees have merged them; they are kept apart here.
Ann King or Nancy King
The record saysThe 1846 marriage and the 1846 baptism of daughter Mary name Ann (Anne) King. One tester’s tree has Nancy King.
WeighedThe register is preferred. Nancy is a common familiar form of Ann.
An Anne King at a Dian baptism, 1878
The record says“Anne King” stands sponsor to Martha, daughter of James Hamill and Ann Gartlan, in 1878.
WeighedAnn (King) Hamill was in Nebraska by 1880 and widowed in Wisconsin in 1858; a return to Donaghmoyne is unlikely. Most probably another woman of the name, perhaps a relative. Worth following.
James Hamill of Anaconda’s birth
The record saysJames, son of James Hamill and Ann Gartlan, baptized in Donaghmoyne 2 September 1873. His Montana death certificate of 1951 gives his birth as 12 September 1873.
WeighedThe baptism, made at the time, is preferred. The certificate also records him as widowed in July 1951, which conflicts with trees that have his wife Catherine (Kate Gartlan) dying in the 1960s.
Peter McGeough’s wife
The record saysOne tree names her McEneaney; another Bridget, maiden name blank; a third Bridget Woods.
WeighedUnresolved. The segment evidence for Peter as common ancestor does not depend on which wife it came through. Donaghmoyne marriages of about 1808–1815 are the place to look.
Open Questions and the Path Forward
The DNA has done what DNA at this distance can do. The remaining questions are about records.
Documentary
The parents of the four Hamill spouses. Catholic Qualification Rolls 1778–1790, Tithe Applotment Books 1823–1838, surviving estate records, the original Donaghmoyne register pages for the four marriages, and the earlier Inishkeen registers, where some of the first entries for these families have already been found.
“James Hamill” as Susan’s father. Find the source behind the attribution in M.S.’s tree, then test it against what is known of the other three founders’ parentage. The single most valuable documentary lead.
Andrew Hamill (1793–1866). T.S.’s ancestor, of the right generation; the 1841 census of England and Donaghmoyne records for his family.
The McMahons. Louth Catholic records of the 1790s to 1810s for the families in A.M.C.’s tree, and for the two Mary McMahons of the network.
Mary Hamill (b. about 1800), Trainor’s wife. Her birth parish, for the Michigan branch of the older Hamill group.
Genetic
Y-DNA recruitment. A male-line descendant of Owen Hamill of Wisconsin (through sons James or Henry Hammel), of James Hamill of Dian (through James of Anaconda, Patrick or Henry of Missouri, or Peter), or of Susan’s Hamill family.
More one-to-one comparisons against the rebuilt kit as cluster matches upload to GEDmatch, including the MyHeritage-only matches.
D.G.’s link to the McCanna family, 28–32 cM, once raw data is on a platform with a chromosome browser.
P.H.’s segments on chromosomes 4 and 16, to see who else shares them.
When a line has no cousins left to test, the parent can sometimes be rebuilt from the children, and one child held back to check the work. Keep network-level findings and relationship-level findings apart, and say which one each claim is. A surname is a reason to look, never evidence of kinship by itself.
The segment data, the Y-DNA results and the record images each have a page of their own. Every living tester on these pages is identified by initials only.
Return to Case Study Chromosome 2 Triangulation Y-DNA Analysis Documentary Sources The Owen Hamall Mystery