The book’s account, as printed: What DNA Can and Cannot Say (full text). This page keeps the research log behind it, with every dated finding.
Genealogy and genetics answer different questions. The records reconstruct who Charles was; DNA tests whether the biological line agrees. This page reports both tests in full, including the matches they did not produce, so a reader can weigh the genetic evidence against the documentary conclusions on the claims dashboard.
All three match lists were read again on 18 October 2026. Nothing new bears on Charles.
- AncestryDNA (about 22,100 matches): ThruLines still proposes no parent for Charles; the five matches it routes through him all descend from his son. The newest matches share 18 cM or less.
- MyHeritage (7,199 matches): no Theory of Family Relativity; the newest matches share about 23 cM or less.
- FamilyTreeDNA: the same four Big Y matches; Family Finder (14,376 matches) has no Konze and no Trendelburg, Deisel or Hofgeismar.
Searched on all three sites: Konze, Kontze, Konzen, Kunze, Gans, Gansen and Gannsen, and births at Trendelburg and Deisel in the matches' trees. The Gans matches are small (about 8–21 cM) and lead to Dutch, Scandinavian, British or Rhineland families, none to Kreis Hofgeismar. The book's DNA chapter reports the results as of this date and points here for later ones.
Y-DNA — the direct paternal line
Terminal subclade: I-Y38863 · Haplogroup: I2, the I-M223 branch.
The Y chromosome passes essentially unchanged from father to son, so it tracks one line — Charles's patriline — back through the centuries, independent of every other ancestor. A Big Y-700 reads that line at high resolution and places it on the human paternal tree by its terminal SNP.
The same Big-Y data was independently interpreted by YFull (the descendant’s sample; ID withheld), which places the line on the identical terminal subclade I-Y38863 (YTree v14.03, formed ~750 ybp, TMRCA ~700 ybp ≈ 1300 CE). On YFull this subclade has exactly three members: the sample of Charles's tested descendant, another tester’s sample — flagged Poland [PL-14] with YFull's “yid” Ashkenazi origin tag — and one further sample (tester sample IDs are not published here). So the two men sharing Charles's exact YFull branch are an Ashkenazi Pole and one further sample — the same Poland/east-European signal FamilyTreeDNA shows, reached on a separate tree. The full Y-STR panel (~780 markers, reviewed 6 June 2026) is an unremarkable I2-M223 modal haplotype (DYS393=13, DYS390=22, DYS19=16, DYS391=10, DYS385=16-17) with no rare or null markers — consistent with the placement and adding no conflicting signal.
What it shows
I-M223 is an old, broadly European paternal haplogroup. But the relevant signal is Charles's specific terminal subclade, I-Y38863, which on both the FamilyTreeDNA and YFull trees sits inside an Ashkenazi-Jewish, east-central-European cluster (Poland, Ukraine, Belarus, Romania, Russia, with some German presence) — not the northwest-European distribution of the parent haplogroup as a whole. So while the deep haplogroup is unremarkably European, the subclade that actually defines Charles's male line points toward a possible Ashkenazi origin — on the present two-to-three-tester evidence (developed in the next sections).
How recent is the shared ancestor?
This is the part that makes the Y-DNA genealogically useful rather than merely deep-historical, and two labs, analyzing the same single Big-Y file, agree on a medieval window (an independent re-interpretation of one kit, not a second sample). FamilyTreeDNA's Discover estimates place the I-Y38863 branch's formation near 1400 CE and its common ancestor near 1450 CE; YFull's independent tree (v14.03) dates the subclade's formation to ~750 years before present and its most recent common ancestor to ~700 ybp — roughly 1300 CE. Either way the shared ancestor is medieval — a few centuries before Charles's 1835 birth — so I-Y38863 is a comparatively young subclade: a Konze man who tested into it would share a paternal ancestor with Charles within a genealogically meaningful window, a far more probative test than a deep-haplogroup match.
On both trees the branch's members cluster in east-central Europe — Poland, Ukraine, Belarus, Romania, Russia, and Germany — the Ashkenazi heartland, not a specifically Hessian or German-Christian distribution. Several of the YFull samples on and just above this branch carry YFull's “yid” origin tag (its marker for Ashkenazi-Jewish submitters), which is what makes the next section's reading independent of FamilyTreeDNA.
What it does not show
- No Konze surname match. Across both the two close Big Y matches and the wider set of Y-STR (111-marker) matches, none carries a Konze / Kontzen / Contze surname. In keeping with the project's living-relatives policy, living testers are not named on this page. What the match list does reveal is discussed in the next section.
- Subclade geography includes but does not pinpoint Hesse. The branch's German / Polish / Russian spread is consistent with a German origin, but a three-tester branch with a ~1450 CE common ancestor cannot resolve to a region within Germany, let alone to a parish.
- A recent-ish common ancestor cuts both ways — it neither excludes Konze nor confirms Gans. Because the branch's ancestor sits near 1450 CE, a Konze match would carry real genealogical weight. But the absence of one among only three testers — whose surnames are not Konze and whose lines fixed surnames at different times across this range — may simply mean that no Konze male-line descendant has Big-Y tested yet. Equally, the Trendelburg Konze are a Christian family documented from at least 1773, so a Konze patriline would need a conversion or non-paternal event at some generation to carry an Ashkenazi-clustered I-Y38863. Since October 2026 the same is true of a Gans patriline: no Jewish Gans family existed at Trendelburg or Deisel (household lists of 1823 and 1858), so a local Gans line would also need such an event further back. The project therefore treats the Y-result as neutral between Gans and Konze.
The match cluster — an unexpected Ashkenazi paternal line
The most striking result is not in the haplotree but in who else sits on it. Reading the Y-STR match list (eight matches, tested to 111 markers or Big Y) by their position on the tree, a clear pattern emerges: Charles's closest paternal-line matches are predominantly Ashkenazi-Jewish lineages from the Ashkenazi heartland of east-central Europe.
| Position relative to Charles's line | Paternal origin(s) | Character of the match(es) |
|---|---|---|
| Same terminal subclade (I-Y38863) — two Big Y matches | Poland, Russia | Both Ashkenazi — the two closest matches by clade; their earliest-known paternal ancestors (b. ~1840 and 1866) are documented Ashkenazi families. |
| One branch up (I-FGC24153) — one Big Y match | Lithuania | Ashkenazi — earliest-known ancestor (b. 1856) carries a Hebrew given name; the family's surname was later Anglicized. (FamilyTreeDNA’s refined tree places him on I-FGC24153, closer to Charles’s line than its parent clade I-BY2640.) |
| Parent clade (I-BY2640) — one Big Y match, about 29 SNPs from Charles’s line | Belarus | An east-European Jewish family (earliest-known paternal ancestor). |
| Nearby branches (I-S23612, I-M223) — five Y-STR matches | Austria, Germany, Sweden, + unspecified | Mostly recognisably Ashkenazi surnames (one earliest-known ancestor b. ~1640); a single Swedish line is the likely exception. |
FamilyTreeDNA's Match Time Tree shows the same structure: both men who share Charles's exact terminal subclade (I-Y38863), and the closest match one branch up (I-FGC24153), are Ashkenazi Jewish — three independent lines, with documented earliest-known ancestors born in Poland, Russia and Lithuania in the mid-1800s, all converging on Charles's line around 1400–1450 CE. The fourth Big Y match, further up on I-BY2640, is also an east-European Jewish family, so all four Big Y matches on the line are east-European Ashkenazi. This is the most striking pattern in the genetic evidence, though it rests on only four testers.
Across the 96 Y-STR matches and the Family Finder list, no Konze, Gans or Kunze paternal line and no Hessian place was found (match list re-read 10 Oct 2026). The Ashkenazi-cluster reading below does not favor Gans, because no Jewish Gans family lived at Trendelburg or Deisel (1823 and 1858 lists, research note, 10 Oct 2026); the book makes the case that he was possibly Karl Ludwig Hofedietz of Deisel, whom a Konze family possibly raised (the book’s case, 28 Oct 2026; not proven). MyHeritage offers no Theory of Family Relativity and no record match for Charles. The decisive test is still a Big Y on a documented Konze man and on a Gans man from the Trendelburg area.
What it most likely means
Charles's strict paternal line — his father's father's father, and so on — descends from a lineage that is, on the present evidence, Ashkenazi-Jewish in origin. Three independent Ashkenazi lines coalescing with Charles's around 1400–1450 CE are most simply read as a common paternal ancestor of that period who was himself part of this Jewish lineage — though with only two-to-three testers this is a working inference, not a demonstrated fact. That reading of the deep male line does not, at least, rest on a single company's analysis: YFull's independent tree reaches the same result, placing the line in the same I-Y38863 subclade, with the same medieval coalescence, among samples it flags as Ashkenazi — though this is an independent re-analysis of the same Big-Y file, not a second tested sample. What remains genuinely open is when and how that line entered the Christian, Hessian, Konze-surnamed milieu in which Charles was born and baptized — by an early assimilation centuries before his birth, or by a later undocumented-paternity event closer to it.
What it does not mean
- It does not make Charles or his documented family Jewish. The Konze family appears in Hessian Evangelical parish registers; Charles married at a German-Evangelical church and his children were baptized there. The Y-DNA sits beneath that documented Christian identity — it speaks to ancestral origin, not to the religion the family practiced in the 1800s.
- It does not date the Jewish ancestor. The signal could reflect an ancestor many generations before Charles. Neither the AncestryDNA nor the MyHeritage ethnicity estimate flags a European-Jewish region, yet the FamilyTreeDNA list carries several small autosomal matches to Ashkenazi-surname individuals (see the match list below) — together pointing to a Jewish contribution that is real but limited and/or deep, and diluted across the generations down to the tested descendant.
- The matches are not close cousins. Genetic distances are moderate (GD 7–10 at 111 markers; 8–10 differing Big Y STRs), consistent with the ~1450 CE common-ancestor estimate rather than a recent shared grandfather.
Why it matters to this investigation
The finding intersects two unresolved threads, and is logged as a new line of inquiry on Open Questions:
- The never-found Konze baptism. One explanation for years of negative parish searching is that Charles's biological paternal line is not, in fact, the documented Konze family — i.e., an undocumented-paternity event somewhere up the line. The Ashkenazi Y-signature is consistent with (though does not prove) such a scenario.
- The illegitimacy hypothesis. The project has long entertained the possibility that Charles was born illegitimate (the basis of the multi-parish January-1835 illegitimacy scans; since 8 Oct 2026 the birth year is treated as uncertain, c. 1833–1841, so those scans cover one year of the range). An unknown biological father would carry whatever paternal line he carried — and could be the source of this Y-result.
Both remain hypotheses. The Y-DNA neither confirms nor refutes them; it adds a concrete, independent constraint that future record work can be tested against.
This is not a new direction for the project — the Y-DNA independently supports an old one. From August 2025, the investigation pursued a Hessian-Jewish paternal-origin inquiry with the Trendelburg researcher, working from documented local Jewish families: the 1740 Trendelburg “Protected Jews” (Seeligmann Ganß — a protected Jew at Trendelburg since 1720, read in the image of the 1731 list, and very probably the man written “Seelmann Gandt” on the 1740 list; no Jew of the canton bore the name Gans in 1812 — and neighbors), the Meimbressen Gans / Segelbaum family, the Christian Gans of Lippoldsberg, and the Stern / Kleeburg line of Hofgeismar. That inquiry was set aside in May 2026 because its basis was a weak surname-resemblance argument (Gans → Gannsen) and because the specific candidates were eliminated or never tied to Charles by a record. The Y-DNA result arrives at the same destination by a completely different road — biology rather than name-similarity — and so independently supports the direction while leaving those specific eliminations standing. It is consistent with the Working conclusion that the name may come from either convergent pool, Gans or Konze (since 5 Jul 2026), and it neither confirms nor excludes the Gans family; it does make the documented Jewish presence in Charles's own locality worth a fresh, evidence-led look. The reactivated thread is tracked at Open Questions · deep paternal origin.
No Jewish Gans family lived at Trendelburg or Deisel. Read in the image (research note, 8 Oct 2026): the October 1823 list of Jewish households in Kreis Hofgeismar (HStAM 18 Nr. 2631) has six households at Trendelburg (Gutmann, Grüntler?, Rosenblatt, Levi Dannenberg, Goldschmidt, Plaut?) and three at Deisel (Nathan Dannenberg, Joseph Ohmsberg, the widow of Levy Katzenstein) — no Gans. The grave list of the Trendelburg Jewish cemetery, which served Deisel and Trendelburg (HHStAW 365/797, 1746–1927), has no Gans either. With the Jewish birth registers of the surrounding communities negative for 1833–41, the idea that Charles was born into a Jewish Gans family at Trendelburg or Deisel has no support. This does not touch the Y-DNA result itself (an Ashkenazi-linked paternal line can lie many generations back), and it does not weaken the Gans surname hypothesis, which rests mainly on his own G-initial signatures; it closes only one route to it.
The 1858 household list confirms it. Read in the image (research note, 10 Oct 2026): the list of every self-supporting Jewish household in Kreis Hofgeismar, drawn up at Hofgeismar on 9 October 1858 (HStAM 18 Nr. 2631, images 156–170; 169 households, 808 persons), has six households at Deisel (Brandenstein, Rosenmeier, Rosenthal, Ohmsberg, Meierhof, Wertheim) and four at Trendelburg (Rosenblatt, Dannenberg, Bloch, Giesberg) — no Gans. The only Gans in the Kreis was Nathan Wolf Gans of Grebenstein. The Staatsarchiv Marburg’s finding aid lists no Trendelburg or Deisel Synagogenbuch after 1812, which explains why no 1824–45 Jewish register has been found for either place. The Jewish route to the Gans surname is not supported and is exhausted online; only two offline files remain (Meimbressen before 1846; HStAM 180 Hofgeismar Nr. 666). The Y-DNA pattern itself stands: an Ashkenazi-linked paternal line can lie many generations back. Charles’s own life was Protestant throughout (St. Marcus marriage 1866, the children’s baptisms, his burial).
Re-checked 7 June 2026 (public FTDNA Discover data): the subclade roster is unchanged — still three tested men (earliest-known paternal origins Germany, Poland, Russian Federation; TMRCA ~1450 CE, 95% CI 1102–1673). Their public group-project memberships add quiet corroboration: of the three, two belong to the Avotaynu Jewish Census project, with single memberships across the German-Jewish (GerSIG), Sephardi-Ashkenazi-Connections, JewishGen-Scandinavia, and Jewish-DNA projects, one in a Jewish surname project, and one each in the Germany and Poland geographic projects — i.e., the subclade's other members publicly research their lines as Ashkenazi lines.
The new match of July 2026. A FamilyTreeDNA alert of 2 July 2026 flagged “two new Y-DNA matches”; filtering the match list by date isolates one new tester, added 1 July 2026 and counted at two marker levels (genetic distance 1 at 25 markers, 4 at 37 markers). It is a distant paternal-line match: it drops below the match threshold at 67 and 111 markers (despite the tester having tested to 111), and is not a Big Y match. Its haplogroup is reported only as I-P222, FamilyTreeDNA's current backbone label for the whole I-M223 branch (the company now assigns unrefined, STR-only predictions to I-P222 rather than I-M223), so it sits far upstream of Charles's terminal I-Y38863 and cannot be resolved into or against his subclade. Its structured fields are empty, but the tester’s free-text profile (read 18 July 2026) says he is searching for his own male-line surname: his third-great-grandfather was born out of wedlock in Hannover in 1795 and took his mother's surname, the true paternal surname possibly having been Neumann. Three observations, all held at low weight. First, the pattern — illegitimate birth, mother's surname adopted, paternal surname uncertain — mirrors one already documented in the Gans line (the 1844 Hellmuth/Gans case), but it was the ordinary practice of the period across German states, not a fingerprint of any one family. Second, Hannover borders this project's Weser-corridor parishes, yet “Hannover 1795” names neither a parish nor a year inside Charles's birth window. Third, as a guard against a false lead: this Neumann is a surname in Hannover and is not the Naumann of the Hümme records, which is a place name — different spelling, different grammatical role, no demonstrated connection. The genetic weight is low: genetic distance 1 at 25 markers but 4 at 37, with the tester dropping out of the list at 67 and 111, is the signature of a distant relationship. The match neither strengthens nor weakens any conclusion.
Match individuals are not named here, per the project's living-relatives policy; the pattern is described from surnames, recorded earliest-known ancestors, and paternal countries of origin in the FamilyTreeDNA match list (retrieved 30 May 2026; Discover re-checked 7 June 2026; match list re-checked 2 July 2026). Kit identifiers can be supplied to bona-fide researchers on request.
Autosomal DNA — the whole-ancestry estimate
Autosomal DNA recombines each generation and reflects all of a person's ancestry, not one line. The test here is from a descendant of Charles, so Charles is only one of many ancestors contributing to the estimate — and a smaller share with each generation of distance. The kit has been run through two independent vendor reference panels — AncestryDNA and MyHeritage — which segment Europe differently but agree on the essentials.
The Germanic component (AncestryDNA)
Among the descendant's fifteen reported regions, the Germanic cluster dominates the Western-European portion:
| Region | Share |
|---|---|
| Southern Germanic Europe | 33% |
| Northwestern Germany | 16% |
| The Netherlands | 5% |
| Germanic / Western-European subtotal | ~54% |
The remainder is largely British-Isles and Celtic ancestry (Northern Wales & NW England, Central Scotland & Northern Ireland, the Hebrides, Connacht, Donegal, Isle of Man, Leinster, and Southeastern England) entering from other lines in the tree, plus several smaller regions.
The same kit through a second panel (MyHeritage v2.5)
MyHeritage's independent reference panel (Ethnicity Estimate v2.5, retrieved 7 June 2026) returns ten regions for the same descendant — all of them European:
| Region | Share |
|---|---|
| Germanic | 36.4% |
| English | 18.5% |
| Scottish and Welsh | 17.0% |
| Dutch | 12.0% |
| Irish | 3.8% |
| French | 3.7% |
| Danish | 3.4% |
| East European | 2.9% |
| North Italian | 1.2% |
| Breton | 1.1% |
| Germanic / Dutch subtotal | ~48.4% |
The two vendors draw their region boundaries differently — MyHeritage's single “Germanic” region spans roughly what Ancestry splits into “Southern Germanic Europe” and “Northwestern Germany” — but they converge on the same picture: a dominant Germanic / Dutch block of roughly half the genome (~48% MyHeritage, ~54% Ancestry), a large British-Isles / Celtic remainder (English, Scottish and Welsh, Irish, and Breton total ≈40% at MyHeritage), and small trace regions. Notably, neither vendor reports a European-Jewish region — the same below-threshold silence discussed under the match list below. Agreement between two independently built reference panels strengthens confidence in the broad admixture picture, but it does not relax either of the limits that follow.
What it shows — and its two limits
A Germanic block of roughly half the genome — reported by both vendors — is consistent with Charles's documented German origin and with the broader family reconstruction. But two limits keep it from confirming anything specific:
- It is a descendant's admixture. By the time the line reaches the tester, Charles contributes only a fraction of the genome; the 54% Germanic signal aggregates Charles's line with every other German ancestor and cannot isolate his.
- The regions are broad. Ancestry's "Southern Germanic Europe" and "Northwestern Germany" — and MyHeritage's single "Germanic" region even more so — each span large territories. Hesse falls within them but is not resolved by them — the estimates cannot distinguish a Hessian ancestor from a Bavarian, Rhinelander, or Saxon one.
The FTDNA match list — what ~14,300 autosomal matches add
Beyond the ethnicity estimate, FamilyTreeDNA's Family Finder returns about 14,300 autosomal matches for the tested descendant. Their structure is itself informative:
- One immediate-family match — a close relative who has also tested, whose tree carries the Gannsen / Konze surnames. It confirms the kit sits where expected in the family but adds no independent evidence.
- No close cousins on the German side. Beyond that relative, the nearest match shares only ~119 cM (a 2nd-to-4th-cousin range), and the mid-range matches are predominantly British-Isles / American surnames — the autosomal signal is dominated by the tester's other lines, not Charles's.
- No match to the documented Hessian families. Searching every match's surname and ancestral-surname list returns no Konze / Kontzen, Kreichelt, Grandjot, or Herwig connection at a meaningful segment size. The Hessian reconstruction remains unconfirmed autosomally — the same gap as on the parish side.
- A faint Ashkenazi echo of the Y-line. The list does contain several small matches (≈9–12 cM, single segments) to individuals whose ancestral-surname lists are distinctly Ashkenazi — including, notably, the same paternal surname that heads Charles's closest Y-subclade match (a different individual). On their own these are weak: single small segments are exactly what Ashkenazi endogamy produces in abundance and can be coincidental. But arriving alongside the Y-DNA result, they are consistent with genuine Ashkenazi ancestry in the paternal line. They are not, however, independent confirmation of the Y reading: the same endogamy that makes the Y-subclade Ashkenazi-clustered is also what scatters these small segments, so the two are largely one signal seen twice rather than two. They do not date it.
- The echo localizes to a specific segment (7 June 2026). A full segment export (15,610 segments) shows the Ashkenazi-surname matches are not scattered: 35 matches carry ≥5 cM segments overlapping one block on chromosome 17 (≈54–64 Mb), many with near-identical breakpoints — suggestive of a shared founder haplotype rather than random noise (and distinct from the known chr-17 pile-up region at 43–45 Mb). This is the same block on which the tester, his father, and the MyHeritage German paternal cluster triangulate. If these segments sit on the paternal copy of chromosome 17, the German cluster and the Y-line's Ashkenazi signature would converge on a single inherited segment — a testable claim: FTDNA's paternal/maternal bucketing (pending an account-linking step) will assign them definitively. Until then this is an unverified but striking alignment.
- X-DNA: structurally uninformative. The X chromosome never passes father-to-son, and Charles sits on the tester's strict paternal line — so X-matching can never bear on the Charles question from a male-line descendant's kit. This avenue is closed by inheritance logic, not by absence of data.
- mtDNA: not pursued (and not informative for Charles's line). No mitochondrial test has been run. mtDNA traces the strict maternal line, which from the tested descendant leads away from Charles entirely, so it could not speak to his paternal-origin or surname question. It would, however, be the correct tool for a different open thread — testing the maternal side of the unwed-mother / illegitimacy hypothesis raised below — and is noted here for completeness of the test plan rather than as a gap in the Charles inquiry.
The apparent tension with the ethnicity estimates — neither of which flagged a European-Jewish region — is not a real contradiction: small Ashkenazi segments can fall below an estimate's reporting threshold, especially once diluted across the generations down to a descendant. Read together, the two suggest an Ashkenazi paternal contribution that is real but limited and/or deep.
The MyHeritage match list — a German-resident paternal cluster (June 2026)
The same descendant's kit returns 6,896 matches at MyHeritage (6,991 on a re-check of 18 July 2026; the ninety-five added in the interval run 9–21 cM, carry no target surname, and include a single Germany-resident tester at 9.2 cM — all below the 30 cM threshold this project monitors), whose tester base is far more continental-European than FTDNA's or Ancestry's. Crucially, both of the tester's parents and a sibling have also tested there, so every match can be assigned to the paternal or maternal side by shared-match triangulation — a discriminating power the FTDNA list lacks. Reviewed 7 June 2026:
- The documented-surname gap persists. Searching all matches' names and ancestral-surname lists returns no Konze / Kontzen, Gannsen, Kreichelt, Grandjot, or Herwig — the same negative as FTDNA, in a second database.
- 327 matches reside in Germany. The largest (≈50–80 cM) resolve to the maternal side and carry Swabian / Württemberg surnames — a reminder that not every German signal belongs to Charles.
- A paternal German cluster — the most promising autosomal lead to date. Two German testers and close relatives of one of them match the tester at ≈52–60 cM, match the tester's father and not the mother, and share ≈52 cM with each other — one ancestral couple plausibly links the cluster to the father's German ancestry. The vendor's estimate places both German testers at 3rd-cousin depth to the father: a most-recent-common-ancestor couple born roughly 1800–1840 — Charles's own generation or one above it. The cluster's (shallow) trees carry ancestral surnames pointing to two regions — Lower Saxony / Hannover on one side and Westphalia on the other (the specific surnames belong to living matches' family lines and are withheld here per the privacy policy; available to bona-fide researchers on request).
- The cluster is not yet assigned to a line. The father descends equally from Charles and from Auguste Kreichelt, whose family was Hannoverian (Clausthal, Oberharz; emigrated 1853). The cluster could therefore descend from a sibling of Charles who stayed in Germany (→ the long-sought German anchor that could decide between two surname leads (Gans and Konze); the book’s case (28 Oct 2026) is Karl Ludwig Hofedietz, raised by a Konze family or from Auguste's collateral kin (→ the first autosomal corroboration of the Kreichelt reconstruction) — the Hannover-region surnames fit the Kreichelt geography at least as well as the Hessian one. Either resolution advances the project. Correspondence with members of the cluster opened June 2026; one member had independently contacted the project years earlier, naming one of these ancestral lines. The discriminating question — Oberharz or Hofgeismar district — is now with them.
- The cluster triangulates. A one-to-many chromosome-browser run (tester × the German tester's family × the tester's father) confirms one triangulated segment on chromosome 17 (≈54.2–64.3 Mb, 12.4 cM, 4,608 SNPs) shared by all four kits — a true identical-by-descent segment from a single common ancestor, inherited through the father. One cluster member's kit has shared-segment viewing disabled, so whether their segment sits on the same block awaits their reply.
- AutoClustering isolates the same group — and nothing else. MyHeritage's AutoClusters analysis (run 7 June 2026; 100 matches ≥~30 cM in 17 clusters) places the three cluster members in one cluster of exactly three, the only Germany-resident paternal cluster in the entire match list. Shared-match clustering, the members' ≈52 cM overlap with each other, and the chr 17 triangulation now agree independently: one ancestral couple likely stands behind the group. The same run also resolved the remaining large German matches into two maternal clusters (the Swabian block and a second family group), clearing them from the Konzen-side candidate list.
- The faint Ashkenazi echo replicates. Small matches (≈9–20 cM) to individuals with distinctly Ashkenazi ancestral surnames recur inside the paternal cluster's shared-match lists, and only four matches reside in Israel, all ≤20 cM — the same weak-but-consistent pattern as the FTDNA list, now visible in a second database, and still undated.
The kit was attached to the project's family tree on MyHeritage on 7 June 2026, activating automated tree-matching (Smart Matches™ live immediately — the import returned 3,187 individuals and 2,635 Smart Matches, which are tree-to-tree matches and not DNA matches; Theories of Family Relativity™ pending the next recalculation). Resolved 18 July 2026: six weeks and one recalculation later, filtering the full match list for Theories of Family Relativity™ returns none at all — zero theories across all 6,991 matches, not merely none for the German cluster. This is the expected result rather than a disappointing one, and it is worth stating plainly: a Theory is built by tracing a documented path through other users' trees to a shared ancestor, and the shared ancestor this project needs sits one rung above Charles — in exactly the undocumented gap the investigation exists to close. Tree-matching cannot bridge a link that no tree records. Granular workings — side assignments, segment figures, the surname-sweep table, and the AutoClusters parse — are in the research notes (research note, 7 Jun 2026); living matches are not named here per the privacy policy.
Re-check across Ancestry & MyHeritage, 28 June 2026
A cross-platform pass to test whether the trees or the autosomal matches can name Charles's parents — they still cannot, but the German paternal cluster was sharpened slightly.
- The cluster is still unassigned, and the deciding answer is still outstanding. Correspondence with the two Germany-resident testers, opened 7 June, remains unanswered; Theories of Family Relativity™ still returns none for them — expected, because the most-recent-common-ancestor couple sits one rung above Charles, in exactly the unrecorded zone the whole project is trying to fill.
- A new triangulation observation weakly disfavors the Kreichelt reading. Reviewed live, the cluster resolves to a most-recent-common-ancestor couple at the tester's father's great-great-grandparent level (born ~1780–1820) — i.e. the parents of Charles, of Auguste Kreichelt, or of one of the two other German lines at that depth. Decisively, across the cluster members' (shallow) trees and the 57 DNA matches they share with the tester, no Kreichelt / Kreigelt / Christoph surname and no Clausthal-Zellerfeld place appears — even though the tester's Kreichelt line is documented in the same tree all the way to Clausthal — and one tester's deepest recorded ancestral place lies in eastern Hesse (the Werra valley), not the Harz. This does not settle the line, but it sits at least as comfortably with a Hessian (Konze) origin as with Auguste's Oberharz one; the common couple still cannot be named.
- AncestryDNA confirms the trees hold no answer. The same day, ThruLines® for Charles proposed no parent above him — the only five DNA matches running through Charles all descend via his son (Charles Jr.), i.e. known cousins of the documented line. A search of all 21,853 AncestryDNA matches for the surname Konze (and similar) returned only relatives already linked to the project's own family tree — no match carries an independently documented Konze line — and a public-member-tree search found only the project's own tree (and copies of it), none attaching parents to Charles.
Net: as of 28 June 2026 the documentary trees and the autosomal match lists still do not name Charles's parents or siblings; the MyHeritage German paternal cluster remains the single best living lead, now leaning marginally toward a Hessian rather than an Oberharz origin, and turns on the members' reply or on Y-testing a documented Trendelburg–Deisel Konze male. Workings: research note, 28 Jun 2026.
A Kontze surname hit, tested and ruled out (19 July 2026)
Weekly monitoring of the MyHeritage kit produced its first-ever hit on the ancestral-surname search for Kontze — which, on investigation, turned out not to bear on Charles. It is recorded here because a negative that has been checked is worth more than one that has not.
- The surname search surfaced two matches where it had always returned none. Every prior weekly scan (14–28 June, 5–12 July) returned zero for Kontze; on 19 July it returned two testers — a mother and her child — both managed by a single user and both sitting in the same ~1,780-person German tree, at 14.2 cM and 8.5 cM (single- and two-segment, both flagged low confidence).
- Examined the same day, the pair does not belong to Charles's line. The match page shows no shared ancestral surnames and no shared ancestral places between the trees. Of the 37 DNA matches the closer tester shares with the descendant, none is a member of the established German paternal cluster — the shared-match group is instead overwhelmingly Dutch, Belgian, and French. And the tree's own principal surname (withheld here per the privacy policy) is an East-Westphalian name centered on the Gütersloh district, roughly 150 km northwest of the Hofgeismar–Trendelburg area and in a different record system.
- Assessment. A Kontze almost certainly sits somewhere among the ~1,780 people in that tree — the search behaves fairly literally, returning nothing for Konze or Konzen and different people for Kunze and Kuntze — but it is not the reason for the shared segment, which is small enough (largest block 8 cM) to be a distant or coincidental match on an unrelated line. The documented-surname gap therefore persists: no Konze / Kontzen, Gannsen, Kreichelt, Grandjot, or Herwig connection survives scrutiny at a meaningful segment size, in either database.
- The rest of the picture is unchanged. The paternal German cluster stands as before (its three members at ≈52–60 cM); no new match above the 30 cM monitoring threshold appeared; and filtering the full list for Theories of Family Relativity™ still returns none.
Net: the 19 July surname hit is a checked negative, not a lead. It changes nothing about the cluster or the conclusions, and the highest-yield next step remains Y-testing a documented Trendelburg–Deisel Konze male. Workings and the parallel documentary note are in the monitoring log (19 July 2026).
What the DNA changes — and what it doesn't
Nothing in the genetic results promotes a claim to Confirmed, and nothing refutes one. The Y-DNA and autosomal evidence both corroborate a German / Germanic origin in general while remaining silent on the two questions the project most wants answered: was the parish Trendelburg, and was the surname Gans or Konze? Those conclusions continue to rest on the documentary record (the surname trail and the “Freundburg” / “Deisse” origin decodings) and are held Working accordingly. See the claims dashboard for how this feeds the load-bearing claims, the methodology · genetic evidence section for the source-rating treatment, and open questions · DNA for the live thread.
The highest-yield next step
Because the Y-DNA line is the one piece of evidence that could directly test the surname conclusion, the most valuable single action is targeted recruitment:
- Y-test a documented male-line Konze / Kontzen or Gans descendant (ideally Big Y-700) — a Konze from the Trendelburg–Deisel area, or a Gans from the Oberweser or Kreis-Hofgeismar Gans families. If his line lands on or near
I-Y38863, that paternal line would be strongly confirmed. A non-match would not exclude Charles from the Konze name: the illegitimate, step-household and maternal-surname hypotheses all predict a non-match even if he bore the name. A match would confirm; a non-match tests only a legitimate Konze patriline. - Widen the autosomal net by inviting known descendants of the candidate Hessian families to test or to share kits, and by uploading existing kits to additional matching databases where permitted.
- Fish for autosomal matches against the documented Kreichelt, Grandjot, and Herwig descendants to confirm the U.S.-side reconstruction on its own genetic footing.
- Resolve the MyHeritage paternal cluster (above): correspondence is open with its German members; their answer to the Oberharz-vs-Hofgeismar question assigns the cluster to the Kreichelt or the Konze line. Attaching the kit to a family tree on MyHeritage (to unlock automated tree-matching against the cluster) is the highest-leverage free action.
- Activate FTDNA Family Matching: with the father's kit linked as a parent, FTDNA buckets all ~14,300 matches paternal/maternal — immediately testing whether the chromosome-17 Ashkenazi founder block sits on the paternal copy (the convergence claim above) and giving the FTDNA list the same side-resolution the MyHeritage analysis already enjoys.
If you descend from the Trendelburg–Deisel Konze / Kontzen family, or from any of these lines, the project would welcome contact — see Contribute.
Genetic results summarized on this page: FamilyTreeDNA Big Y-700 (haplogroup I2-M223; terminal subclade I-Y38863), AncestryDNA and MyHeritage (Ethnicity Estimate v2.5) autosomal estimates, the FTDNA (retrieved 30 May 2026; match list re-checked 2 July and 18 July 2026) and MyHeritage (retrieved 7 June 2026; re-checked 18 & 19 July 2026) match lists, and the FTDNA all-segments export (7 June 2026). Page first published . Living DNA matches are not named on the public site per the privacy policy; kit identifiers can be supplied to bona-fide researchers on request.