5. Maternal-Line Direct Ancestor Chart

Traced from genealogy.db (GEDCOM-derived, internal DB extraction — classified as a derivative source throughout; no original vital records or sources table citations are attached to these individuals unless noted). Generation numbers count Cody as Gen 1.

Every link above (famc/fams) resolves to an existing DB individual record — no ancestor in this chain is absent or unlinked. The gap originally flagged (Mildred's parents = Cody's maternal-great-grandparents) is not a missing-person gap; both Patrick J Logan and Annie Logan exist in the DB with birth data. The real issues are (1) a surname/parentage anomaly at Mildred herself, addressed below, and (2) incomplete death/marriage data noted in Section 7.

Source note: Barnaby's internal DB traversal (task-2, 2026-09-07) is the source for the full chain above; Oscar (task-3) and Angela (task-4, hive/agents/angela-mszdgen3/task4-evidence-analysis.md) independently cross-checked Gen 3–5.

6. Documented Conflicts — Maternal Line

6.1 RESOLVED — Mildred's surname: "Kern" vs. "Logan"

• Conflict: Mildred's DB individual record carries the surname "Kern," but she is linked in family F237 as a child of Patrick J Logan & Annie Logan, alongside six full siblings who all carry the surname "Logan" (Margaret, Irene, Gertrude Veronica, Dorothy Ellen, James Patrick, Annamay). Her own notes field reads "aka Mickey Kern."
• Resolution: "Kern" is a married surname, not her birth surname. Mildred's own family records show she married twice — family F199 (m. Gerald A. Green, 3 Jun 1948, Salem, Rockingham Co., NH) is the marriage represented in this chart, and a separate DB family record F168 shows an earlier marriage to a Russell Kern. Her birth surname is Logan.
• Corroboration: An independent, external WikiTree profile — Logan-7083 — matches Mildred exactly on birth date (31 May 1930), death date (30 Jun 1994), and birthplace, listing "Kern" only as a LastNameOther field, and naming her father via a WikiTree ID that resolves to Logan-8841, Patrick James Logan, an exact match (DOB 26 May 1891, Ireland) to I420027268496. WikiTree is itself an unsourced, user-compiled derivative tree, but it is a genuinely independent compilation (different platform, different informant history), so its agreement with the DB's F237 structure carries real, if modest, corroborative weight.
• Documented but resolved discrepancy — 1940 census "Granddaughter" note: Mildred's 1940 U.S. Census entry records her Relation to Head of House as "Granddaughter," which directly conflicts with daughter status under F237. However, all seven of Patrick & Annie's other recorded children were born 1921–1930 — none was old enough by 1930 to have parented a child, so there is no candidate biological parent within the household under a literal "granddaughter" reading. The most economical explanation is an enumerator or transcription error, a well-documented census error class. An informal-fostering scenario (Mildred raised by grandparents while biologically another relative's child) cannot be fully excluded, but no record anywhere supports an alternate parentage line, and Oscar's independent WikiTree search likewise found no corroborating profiles for that alternative.
• Proof Level: PROBABLE — not Proven. No birth certificate or vital record has been examined; the conclusion rests on internally consistent DB family structure plus one independent but itself-unsourced external tree. See Angela's full evidence analysis (hive/agents/angela-mszdgen3/task4-evidence-analysis.md) for the complete comparison matrix and reliability rankings.

6.2 RESOLVED (minor) — Mildred's death place

• DB records (1992 residence event and 1994 death event) independently agree: Newfield Hamlet, Tompkins Co., NY.
• WikiTree's Logan-7083 profile gives a conflicting single field: "Newport, Herkimer Co., NY" — a different county roughly 150 miles away, not a plausible transcription variant.
• Resolution: The DB's internally double-corroborated pair of records is retained as of record. WikiTree's entry is flagged as more likely erroneous (a possible confusion with "Newport" as a place name recurring across several NY counties), but this remains unconfirmed absent a death certificate or obituary.

6.3 OPEN / UNRESOLVED — Katie McSorley's age at marriage

• Conflict: Katie McSorley (I420027268567) has a recorded birth of "abt 1878," New Hampshire, but family F194 records her marriage to Thomas James McSorley in 1885 — implying she married at approximately age 7, a physical/legal impossibility.
• Analysis: Neither DB field carries a source citation. The one firm anchor available is daughter Annie's independently-attested birth (21 Nov 1893), which only narrows the true marriage date to "sometime before late 1893" — it does not indicate which of the two conflicting fields (Katie's birth year, or the 1885 marriage date) is in error. A WikiTree profile, McSorley-246, matches Thomas's birth year (1853) but conflicts on birthplace (Kilskeery, Co. Tyrone, Ireland vs. the DB's Massachusetts) and offers no data on Katie or the marriage — it neither confirms nor resolves the age conflict, and it isn't even certain it is the same Thomas.
• This conflict is explicitly NOT resolved. No record found so far identifies which field is wrong, or whether the spousal linkage itself is sound. "Likely a transcription error in one of the two dates" cannot be upgraded to Probable without identifying which one.
• Proof Level: UNPROVEN.
• Status: Flagged for a round-2 external search (in progress via Oscar, task-5) targeting NH/MA vital records for a Katie McSorley birth record and the McSorley–Logan marriage record, including alternate spellings ("Catherine McSorley," "Sorley"/"McSorly") and the possibility that "1885" is a transposition of a later year.

7. Research Gaps Remaining

• Patrick J Logan (I420027268496) — death place is a malformed string ("MA,,,"), with no town identified and no death date recorded at all.
• F237 (Patrick J Logan x Annie Logan) — no marriage date or place recorded for the couple's own marriage.
• F214 (Theodore Daratt x Alice Hoover, paternal-maternal great-grandparents) — no marriage date or place recorded.
• Missing death dates/places, direct line, Gen 4–5: Annie Logan (no death info at all); Ellen Logan (Gen 5, none); Michael Logan (Gen 5, death year only, no place); Thomas James McSorley (Gen 5, none); Katie McSorley (Gen 5, none — also the open age-conflict above); Theodore Daratt (death year only, no place); Alice Hoover (death year only, no place).
• Katie McSorley's age-at-marriage conflict (Section 6.3) — genuinely unresolved pending NH/MA vital-records research.
• None of these represent an absent or unlinked direct ancestor — every famc/fams link in the traced chain resolves to an existing DB record. The gaps are data-completeness issues (missing dates/places/citations), not missing people.

7.1 Further Research Leads

• Peter McSorley Sr. & Mary McSorley (Ireland) — extends the McSorley line beyond Section 5. The maternal-line chart in Section 5 stops at Gen 5 with Thomas James McSorley (I420027268605, Annie Logan's father). The DB's famc/fams chain extends this two further generations: Thomas James McSorley → father Thomas McSorley (I422151680890, b. 1828/1829 Ireland, d. 14 Mar 1903, Cambridge, MA, famc F253) → father Peter McSorley (I422154178720, b. 1805, Omagh, Co. Tyrone, Ireland, d. 16 Mar 1865, Cambridge, MA). Death date/place for this Peter McSorley — 16 Mar 1865, Cambridge, Middlesex Co., MA — is rated PROBABLE per GPS-standard source audit: a FamilySearch index citation (ARK 1:1:NWVD-WRP, collection "Massachusetts, State Vital Records, 1638–1927") matches the pre-existing DB entry (age 59, widowed, laborer, birth ~1806 Ireland, within normal estimation variance of the DB's 1805/Omagh). This citation is an index match only — the original record image has not yet been examined. New to the index record: it names his own parents as "Peter" and "Mary" McSorley, both born Ireland (surname beyond McSorley not given). This parentage is UNPROVEN — a lead only, not an established fact: it rests on a single informant-supplied source with no surname beyond McSorley and no distinguishing dates or places, so it cannot yet distinguish this "Peter Sr." and "Mary" from any other Irish-born McSorley couple. Neither has been added as an individual record. A future researcher should treat "Peter McSorley Sr." and "Mary McSorley" (parents of the Peter McSorley who died 1865 in Cambridge, MA) as the next brick wall up this line — first obtaining the original 1865 death record image to check for additional identifying detail, then searching MA/Ireland (Co. Tyrone area) records for a McSorley couple consistent with a son born ~1805 in Omagh. (Source verification: worker-angela, GPS audit, 2026-09-07.)

• Samuel Winslow Robinson (I420187259064) — INVESTIGATED AND FOUND NOT CONNECTED to Cody Lee Hall's pedigree; do not pursue further as ancestry. Samuel (no famc_id; b. 1 Jan 1821, Massena, St. Lawrence Co., NY; d. 18 Apr 1890, Massena, NY; m. 29 May 1844, family F165) was initially flagged as a promising brick wall, and a WikiTree profile (Robinson-451, WikiTree person Id 325005) was found matching his birth/death dates and places exactly, listing parents Oliver Robinson (b. 1785 Pomfret, CT; d. 1838 Massena, NY) and Eunice (Chamberlain) Robinson (b. 1792 VT; d. 1864 Massena, NY). However, a subsequent structural audit of Cody's full computed direct-ancestor set (394 individuals, traced recursively via famc_id/families.husband_id/wife_id from Cody upward) plus one generation of collateral kin found zero connection: of the 19 Robinson individuals in the database, only one colonial-era pair (Abigail Robinson, b. 1690, family F40, and one collateral) traces into Cody's pedigree. Samuel Winslow Robinson's entire household (himself, wife, and all 4 children) is a fully isolated family island with no traceable link to Cody Lee Hall's ancestry. The WikiTree lead above is retained here only as a record of research already done — it is not part of Cody's confirmed or probable ancestry, and should not be re-investigated as if it were. (Pedigree-connection audit: worker-barnaby, 2026-09-07; original WikiTree lead: worker-oscar, 2026-09-07.)

• Eliza P. Blish (I420027595069) — INVESTIGATED AND FOUND NOT CONNECTED to Cody Lee Hall's pedigree; do not pursue further as ancestry. Eliza (no famc_id; b. 24 Feb 1839, Winooski, Chittenden Co., VT; d. 12 Feb 1902, Claremont, Sullivan Co., NH; m. Albert G. Blish, 24 Sep 1859, Colchester, VT, family F262) was flagged as a genuine brick wall with specific, researchable place names, but a WikiTree searchPerson query and a general web search found no profile or record matching her known data (near-misses Blish-224, Blish-19, Blish-20 all have the wrong birthplace, death place, or generation). Separately, a structural audit of Cody's full computed direct-ancestor set (394 individuals) plus one generation of collateral kin found that none of the 26 Blish individuals in the database — including Eliza — trace into Cody's pedigree by any famc/fams chain. The entire Blish surname (family F262) is a fully isolated family island, unconnected to Hall/Green/Logan/Daratt/McSorley ancestry. Retained here only as a record of research already done; not part of Cody's confirmed or probable ancestry. (Pedigree-connection audit: worker-barnaby, 2026-09-07; negative search result: worker-oscar, 2026-09-07.)

• James Hanna (1730–1800, I420106471379) & John Hanna (1756–1845, I420106471441) — Hanna line matches, plus a documented (not adopted) alternative-parentage claim. James Hanna (b. 1730, d. 1800, Renick, Greenbrier Co., VA) matches WikiTree profile Hanna-1432 on exact birth and death dates and place. That profile names James's wife — previously a blank slot in family F453 — as Mary Jane McIlwrath (WikiTree Mcllwrath-1, b. 1717 Ireland, d. 1798 Virginia, m. 1742, PA); proof level POSSIBLE, corroborated by the exact date match on James himself and by his son's WikiTree bio (below) naming a daughter "Nancy" matching our Nancy Agnes Hanna (b. 1793), but not yet independently verified against a primary VA/PA record. James's son John Hanna (b. 1756, d. 11 Apr 1845) matches WikiTree profile Hanna-1434 on a double exact-date match (birth year 1756 AND death date 11 Apr 1845); proof level PROBABLE, per a rich corroborating bio (Revolutionary War POW at Charleston, Greenbrier Men at Pt. Pleasant 1774, moved VA → Jackson Co., Ohio 1817–1820, buried Madison Cemetery, Rempel, Jackson Co., OH; source FamilySearch Family Tree PID L627-W9R via WikiTree) that also names a daughter "Nancy" who stayed in Virginia in 1817, again corroborating Nancy Agnes Hanna. Parentage conflict, resolved by decision (not by evidence): our database records James Hanna's father as James Hanna (b. 1718); WikiTree Hanna-1432 instead names his father as William Hanna (b. 1703, Scotland) and Grissel Williamson — a completely different lineage. Per an explicit human decision, our database's existing lineage (James Hanna b. 1718) is being kept as-is unless/until a primary source is found; WikiTree's claim is documented here as an unadopted alternative, not an open toss-up. Structural famc links were not changed. (Research: worker-oscar/god, 2026-09-07.)

• Charles Russell Leonard II (I420088075101) — paternal-line parentage lead, proof level revised to POSSIBLE after GPS audit. Two community trees independently name parents for Charles: (1) WikiTree profile Leonard-979 — exact death-date match on Charles Russell Leonard II himself — gives Russell Leonard (b. 1808/9, Chesterfield, Hampshire Co., MA) and Sophronia Burrill (b. 1815 NY, d. 1883, Moravia, Cayuga Co., NY) as parents, and its bio text cites a DNA corroboration figure (35–38 cM Ancestry matches, updated Aug 2025) for the father attribution; (2) FamilySearch Family Tree profile LRT6-XP1 (checked live, logged in as Cody L Hall) independently lists the same two parent names/years, plus a matching spouse (Mary Polly Brown) for Charles, confirming it is the same tree subject rather than a shared GEDCOM mirror between the two sites. This research report initially rated the finding PROBABLE on the strength of that two-source agreement plus the cited DNA figure (research: worker-oscar/god, 2026-09-08); a subsequent GPS-standard audit revised the proof level DOWN to POSSIBLE (audit: worker-angela, 2026-09-08), for two reasons that should be read as caveats, not dismissals: (a) two community trees agreeing establishes same-subject-identity (matching spouse and death date) but not independent derivation of the parentage claim itself — community trees are commonly cross-populated by the same underlying researchers, and neither profile's own source citations have yet been checked for a shared origin; (b) the cited 35–38 cM DNA match is a secondhand figure quoted in bio text, not a verified match detail pulled from an actual DNA account, and is not being counted as evidence until it is. Russell Leonard and Sophronia Burrill have not been added as individuals in the GEDCOM/DB pending further verification. Concrete next steps to upgrade proof level: (1) check both WikiTree Leonard-979's and FamilySearch LRT6-XP1's own source citations to determine whether they derive from a shared origin or are truly independent; (2) search Hampshire Co., MA vital/church records for Russell Leonard's birth (~1808/9, Chesterfield) and Cayuga Co., NY vital/probate records for Sophronia Burrill (d. 1883, Moravia) to establish a primary-source basis; (3) if DNA is to be relied on for the father attribution, pull the actual match detail (shared cM, tree overlap, common ancestor path) from the DNA account rather than the secondhand bio figure. Until then this remains a possible, not probable or confirmed, lead — accuracy over a stronger-sounding headline. (Research: worker-oscar/god; GPS audit: worker-angela; both 2026-09-08.)

• Harriett Veitch (I422458924676) — genuine open brick wall, IS pedigree-relevant. Harriett (b. 1839, birthplace unknown; d. 12 Oct 1923, buried Ufford, Muskoka District, Ontario; family F52) appears in the database only under her married surname — she is the wife of Adam Veitch (I420034571492, b. 9 Nov 1839, probably Waddington, NY; d. 21 Apr 1917, Ufford, Muskoka, Ontario). Unlike Robinson and Blish above, the Veitch line is confirmed connected: a structural audit of Cody's full ancestor set found 5 direct blood ancestors and 14 collateral relatives among the 19 Veitch individuals in the database (chain Thomas → Thomas Adam → John → Ann Veitch, families F435→F223→F186→F48). Adam Veitch himself is collateral kin — a sibling of direct ancestor Thomas Adam Veitch — making Harriett, his wife, one step off the direct pedigree by marriage. Her true maiden name and parents remain unfound: WikiTree searchPerson (surname Veitch, given name Harriett/Harriet variants) returned no matching profile, and the FindAGrave memorial for her husband (226090305) that might link to her own memorial returned HTTP 403 and could not be retrieved. Known family movements: Waddington, NY area → Stormont/Dundas/Glengarry Co., Ontario (1860s–70s, per children Daniel b. 1864, Elizabeth b. 1870, Wellington A. b. 1876) → Muskoka/Ufford, Ontario (by ~1900s). A separate, older Adam Veitch + Ann Nesbett of Waddington, NY (per an 1855 memorial record) may be our Adam's parents via FAMC F48 in the GEDCOM — not yet checked, a possible side lead for the Veitch line itself, not for Harriett. Recommended next steps: (a) retry the FindAGrave memorial via an alternate/authenticated fetch method; (b) search Stormont/Dundas/Glengarry Ontario (or St. Lawrence Co., NY) marriage records for Adam Veitch × Harriett, likely pre-1864; (c) manually browse bracebridgemuskokahistory.com's "Veitch Index." (Pedigree-connection audit: worker-barnaby, 2026-09-07; brick-wall research: worker-oscar, 2026-09-07.)

Current Brick-Wall Research Queue (Gen 5–9)

Task-16: gen 5-9 direct-ancestor brick walls (regenerated 2026-09-08)

Computed via BFS from Cody Lee Hall (I420027268461) through famc_id -> families -> husband/wife. 73 direct ancestors in generations 5-9 have no famc_id (blank/unlinked parents) -- this is the real researchable frontier, separate from (and broader than) the "8 untouched surname lines" side-quest, which only covered collateral/large-count surnames, not necessarily direct ancestors.

Already resolved/attempted, skip: - I422154178720 Peter McSorley (gen6) -- 1865 MA death record found, parents lead (Peter Sr./Mary) documented, unproven. - I422675186332 John Kennelly (gen6) -- attempted, inconclusive (too common a name in Cork).

Best next candidates (real dates+places, most researchable), in priority order: 1. I420088075101 Charles Russell Leonard II (gen6) -- b.25 Nov 1830 Union Springs, Cayuga, NY / d.23 Nov 1908 Groton, Tompkins, NY. Full US dates+places, strong candidate. 2. I422620250515 Marie Louise Desrochers (gen6) -- b.3 Mar 1828 St. Frederic, Drummondville, Quebec / d.25 Feb 1882 Barton, Orleans, VT. Quebec parish records likely findable. 3. I420027268595 Philomene Perusse (gen5) -- b.1853 Canada / d.1876. Quebec French-Canadian, worth a targeted search. 4. I420173649406 Julius Schneider (gen5) -- b.1864, no place recorded yet. 5. I420173649427 Elizabeth W Pries (gen5) -- b.1865, d.1943, no place recorded yet. 6. I420027596320 Peter Barraby (gen6) -- b.1820 Canada, French / I420027596321 Maria Billings (gen6) -- b.1821 Canada -- a couple, worth researching together. 7. I420035597116 Euphemia Brunton (gen7) -- b.1788 Crailing, Roxburghshire, Scotland / d.1860 Waddington, St Lawrence, NY -- this is the "elder Adam Veitch's wife" side-lead Oscar flagged during the Harriett Veitch search (FAMC F48) -- may resolve two brick walls at once.

Remaining 60+ entries (mostly gen7-9, sparser data, many blank dates/places -- lower priority, colonial-era-adjacent) are in the full BFS output; regenerate via the same script if this file goes stale (root=I420027268461, filter famc_id blank + generation 5-9).

DB pre-check on top 7 candidates (Barnaby, 2026-09-08) -- before Oscar spends browser budget

Checked all 7 for notes_json flags (none exist yet, all []) and for surname/family cross-references elsewhere in genealogy.db that narrow the search. Findings:

1. Charles Russell Leonard II (I420088075101) -- no cross-refs found, no variant spellings elsewhere in db. Family F3 (wife I420088075094, married 1853 "of Cayuga Lake, Union Springs, Cayuga, NY", 8 children) is otherwise fully self-contained. Straight NY vital/church record search, nothing to narrow further from the db side.

2. Marie Louise Desrochers (I422620250515) + 3. Philomene Perusse (I420027268595) -- these two brick walls are IN-LAWS of each other, already linked in the db. Desrochers' family F294 (husband Moise Valée/Vallée, I420110641009, b.2 Jun 1823 Yamaska, Pierre-De Saurel, Quebec -- famc_id F142, already traced one gen further back) has son Onesime Vallée (I420027268488, b.1 Apr 1853 Barton, Orleans, VT), who married Philomene Perusse on 25 Dec 1872 in East Albany, Orleans Co., VT (family F183). So both brick-wall women's in-law/husband side is already well documented with real Quebec parish origin (Yamaska) + VT settlement (Barton/East Albany, Orleans Co.) -- a strong, narrow research target: Yamaska-area Quebec parish records for Valée/Desrochers, and Orleans Co. VT vital records for the Perusse/Vallée marriage tying back to her own parents.

3. Julius Schneider + 5. Elizabeth W Pries (I420173649406 / I420173649427) -- siblings via shared famc_id F171 with Freda Schneider (I420087915149, b.1888, fams_ids F92) -- i.e. F171 is the Schneider/Pries parents' family (blank husband/wife in families table -- not yet even entered as individuals, just referenced by id). No further narrowing available in-db; F171's parent records are essentially empty placeholders. Lowest-information candidates of the 7.

4. Peter Barraby (I420027596320) + Maria Billings (I420027596321), family F172 -- strong existing forward chain already in the db, worth leading with. Their children: Benjamin Barraby (I422146674744, b.1830 Canada French), William Baraby (I420027596322, b.15 Aug 1840 St Albans, Franklin Co., Vermont -- note the one-r spelling variant "Baraby" already present in-db, confirming spelling drift -- d.1920 Westminster, Windham Co., VT, m.1865 Sophia Parrel/F157), and Peter Barraby Jr. (I420027268598, b.Jan 1845 Canada, d.1910 Claremont, Sullivan Co., NH, m.1872 Cordelia "Delia" Cushion, family F184, 7 children in Claremont/Acworth/Goshen, NH). No notes_json flags on any of them, but the family is richly documented forward into NH/VT vital records through the 1870s-1920s -- the strongest lead of the 7 for indirect proof: NH/VT descendant vital records (deaths of Peter Jr./William, both with informants) may name the immigrant parents' origin parish in Quebec.

5. Euphemia Brunton (I420035597116) -- family F279 (husband Samuel Porteous, I420035597115, b.1784 Maxton, Roxburghshire -- already has famc_id F263, i.e. already linked one generation further back in the db, m.17 Nov 1804 Eckford, Roxburgh, Scotland). Their daughter Ann Veitch (née Porteous), I420034571488, b.1810 Jedburgh -- confirmed married into F48, matching the existing Veitch/Harriett research thread. Since Samuel Porteous's own parents are already in the tree (F263), the real open brick wall here is narrower than it looked: only Euphemia Brunton's own parentage (not Porteous's) is unresolved -- Roxburghshire OPR (Old Parish Register) baptism search for an Euphemia Brunton b.~1788 in Crailing parish is the precise target.

Bottom line for Oscar: items 6 (Barraby/Billings) and 2+3 (Desrochers/Perusse/Vallée) have the richest existing in-db forward documentation to anchor a search from (specific NH/VT towns, marriage dates/places already known for descendants) -- start there. Item 7 (Brunton) is narrowed to a single-generation, single-parish OPR search. Items 1, 4, 5 have no additional narrowing available from the db and remain as originally scoped in the priority list above.

Side-quest status (8 untouched surname lines, separate audit)

• Walker/Makepeace/Hanna/Wheeler: CONFIRMED pedigree-relevant, but ZERO brick walls in gen5-9 window (Barnaby, 2026-09-08) -- fully linked already, nothing to research here.
• Bergeron/Blish/Robinson: CONFIRMED NOT connected to Cody's pedigree (Barnaby) -- do not research further as ancestry (Blish/Robinson leads already investigated and documented as such in the research report).
• Veitch: CONFIRMED pedigree-relevant. Harriett Veitch (I422458924676, married name only) already researched -- inconclusive/negative, well documented. Her husband Adam Veitch's own parents (via FAMC F48, item 7 above) is a fresh, not-yet-tried lead.

8. Family Oral History

Methodological note (Evidence Explained standard): The material in this section is oral tradition / family testimony, not documentary evidence, and is presented separately from and at a lower evidentiary tier than Sections 5–7 above. Informant: Denice (Green) Hall (I420027268464), Cody Lee Hall's mother, relayed via a family member (the human requester) on 2026-09-07. As to Denice's own witnessed childhood experience, this is primary information (an eyewitness account of events she lived through) but it remains undocumented — no vital record, court record, or contemporaneous written account corroborates it. It should not be read with the same certainty as the census, family-structure, and cross-platform evidence cited above; it is included because family testimony is itself a legitimate genealogical source class, properly labeled as such. This section does not alter, and should not be read as bearing on, the parentage or proof-level conclusions in Section 6 (Mildred Logan/Kern's own parentage; the McSorley age conflict) — it concerns a separate, later-life chapter of Mildred's story, decades after those events.

Testimony, as relayed:

• Mildred Claire (Logan) Green/Kern (I420027268557) married Gerald A. Green (I420027268583) in family F199 (3 Jun 1948, Salem, Rockingham Co., NH). She was later also known as "Mildred Kern," per DB family record F168 (husband: Russell Kern, I422146679883 — read directly from genealogy.db; no birth/death data recorded for him), from a subsequent relationship with Russell "Russ" Kern. Per the human's account, this was described as "her time she spent as Russ Kern's partner," language that may indicate a domestic partnership rather than a second formal marriage. This is unconfirmed either way — the DB's F168 record does not itself specify marriage vs. partnership, and no external record has been located to settle it. It should not be asserted as a marriage without further documentation.
• Per family oral history, Mildred left Gerald Green and their children when daughter Denice (Green) Hall was somewhere between ages 12 and 16 — approximately 1969–1973, calculated from Denice's recorded birth date of 4 Aug 1957 (this date range is a derived estimate from the oral account, not itself independently documented).
• Denice subsequently helped raise her younger siblings.
• The human's account further relays that both Mildred and Gerald drank heavily and fought, and that Denice recalls Mildred leaving her and her siblings with their father before departing.

Cross-reference: Barnaby's parallel DB notes on this material (task-11) were pending confirmation of individual IDs at the time this section was written; Russell Kern's ID (I422146679883) was confirmed directly against genealogy.db (family F168) in the interim. If Barnaby's task-11 output adds further IDs or DB notes, this section should be revisited and cross-referenced accordingly.

8.1 DNA Evidence: Half-Sibling Status of Diane and Dawn Green

Methodological note: This subsection combines two evidence classes that must not be conflated. (1) The DNA-match claim itself is reported family testimony, not an examined DNA report — no raw match data, centimorgan (cM) figures, or third-party testing-service export is present anywhere in genealogy.db or the source GEDCOM. It is recorded here as the human's account, pending verification. (2) A structural fact independently present in the source document — family F426 — is original/derivative database evidence, directly citable and verifiable in genealogy.db, and exists wholly independent of the DNA claim.

Reported DNA testimony: Per the human's account, DNA match testing shows that Diane M. Green (I420027268477, b. 19 Jun 1954) and Dawn Green (I420027268588) each share roughly half the DNA (cM) with their siblings that those siblings share with one another — a pattern consistent with half- rather than full-sibling relatedness. No cM figures or testing-service report were provided to this research effort; if the human supplies the underlying DNA export, this subsection should be updated with the actual figures and the claim's proof level revisited.

Independent structural corroboration (genealogy.db): Separate from and prior to the DNA claim, the DB already records two distinct family structures for Mildred's children: - Family F199 (Mildred x Gerald A. Green, m. 3 Jun 1948) lists seven children, all with Gerald A. Green as father: Dorothy L Green (I420027268538), Donald E Green (I420027268497), Dennis Green (I420027268566), Denice (Green) Hall (I420027268464), Debra Lynn Green (I420027268540), David Green (I420027268476), and Doug Green (I420027268520). - Family F426 lists Diane M Green (I420027268477) and Dawn Green (I420027268588) as children of Mildred (I420027268557) alone — husband_id is blank in this record. No father is named anywhere in the DB for either Diane or Dawn.

This structural separation was present in the source data independent of the DNA claim, and corroborates it: it is consistent with Diane and Dawn having a different, unidentified father than the seven F199 children, which would make them half-siblings to that group rather than full siblings.

What this does NOT establish: The DB does not identify, and this report does not speculate about, who Diane and Dawn's biological father might be. No record in genealogy.db names one. This subsection states only that a father is not identified in any record available to this research effort — nothing further should be inferred.

For context/consistency: Jacquelyne Weaver (I420027268516) is confirmed elsewhere in the DB as Dawn Green's daughter, via family F205 (father: I420027268546).

Proof Level: PROBABLE / REPORTED — not Proven. The F426 structural separation is real, citable, independently-sourced corroboration; the specific DNA cM data is reported testimony only and has not been independently examined. Should the human provide the actual DNA test results, this claim can be upgraded pending that review.

Research Task Log