Gurney Genealogy Library Updated 4 August 2026 by Allen Gurney

These research notes are provided as-is and contain supplementary working research.

Identity — the shape, the matrix, and the two families

Every candidate for the Braintree emigrant is tested against one specification of what the man and his household must look like. This unit holds that specification, the matrix that scores every non-zero candidate against it, and the structural finding the matrix produced — which is the most useful result the July 2026 work delivered and was not visible until the attributes were laid out side by side.

The shape as specified

Fifteen attributes, each a scale rather than a binary, none mandatory.

Attribute
0 surname Gurney or a variant
1 a household with Mary, Richard and/or John Jr., in that order
2 a Mary born c.1628
3 a Richard born c.1630
4 a John born c.1633
5 a father named John
6 a mother named Mary
7 the father born c.1603
8 a John × Mary marriage c.1622–28
9 a John × Mary marriage c.1628–40 (second-marriage hypothesis)
10 the father’s father named Richard
11 a region known for Great Migration participation
12 a region connected to Weymouth, Massachusetts, or the colonial John’s cohorts
13 the father a tailor
14 an event c.1636–39 whose context could support emigration

Attribute 14 is read throughout as a documented push event.

The matrix

✓✓ strong · ✓ soft or partial · ~ unknown or untested · ✗ absent · ✗✗ contradicted

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
B — son of Francis (11%) ✓✓ ✗ ✗ ✗ ✗ ✓ ✗ ✓ ~ ~ ✗✗ ✓✓ ✓ ✓✓ ✓✓
Ackworth (8%) ✓ ✓ ✗ ✗ ✓ ✓✓ ✓✓ ~ ✗ ✓✓ ~ ✓ ✗ ~ ✓
Weston Turville 1604 John (5%) ✓✓ ~ ~ ~ ~ ✓✓ ~ ✓✓ ~ ~ ✗ ✓ ✓✓ ✓✓ ✓
Cheddington 1608 (6%) ✓✓ ✗ ✗ ✗ ✗ ~ ~ ✓ ✗ ✗ ✓✓ ✓ ✓✓ ✓ ✓
C(ii) Berkhampstead (6%) ✓✓ ✓✓ ✗ ✓✓ ✓ ✓✓ ~ ✓✓ ✗ ✗ ✗✗ ✓ ✗ ~ ✓✓
D London draper (4%) ✓✓ ✗ ✗ ✗ ✗ ~ ✗ ✓✓ ✗ ✗ ✗✗ ✓✓ ✓ ✓✓ ✓
Epping / North Weald line (3%) ✓ ~ ~ ~ ~ ✓✓ ~ ~ ~ ~ ✓✓ ✓✓ ✓ ~ ~
Hitcham (3%) ✓✓ ✓ ✓✓ ✗ ✗ ✓✓ ✗ ✓ ✗ ✗ ~ ✓ ✓ ✗ ✗
Newgate apprentice (2%) ✓✓ ✗ ✗ ✗ ✗ ✗ ✗ ✗✗ ✗ ✗ ~ ✓ ✓✓ ~ ✗
Weston Turville 1627 father (2%) ✓✓ ✗ ✗ ✗ ✗ ✓✓ ~ ✓✓ ✗ ✗ ~ ✓ ✓✓ ✓✓ ✗
Bucks / Herts / Beds group (3%) ✓✓ ✗ ✗ ✗ ✗ ✓ ✗ ~ ✗ ✗ ~ ✓ ✓✓ ✓ ~
A(1603) Stewkley (2%) ✓✓ ✗ ✗ ✗ ✗ ~ ~ ✓✓ ✗ ✗ ✗ ✓ ✓ ~ ✗
Gurley, Westminster (2%) ✓ ✗ ✓✓ ✗ ✗ ✓✓ ✓✓ ✓ ✓ ✗ ~ ✓ ~ ~ ~
C(i) Berkhampstead (eliminated, scored for calibration) ✓✓ ✓ ✗ ✗ ✓ ✓✓ ~ ✗✗ ✗ ✗ ~ ✓ ✗ ~ ✗
Norwich St Peter Mancroft (2%) ✓✓ ~ ~ ~ ~ ✓✓ ✓ ~ ~ ~ ~ ✓✓ ✓ ~ ✓
Stewkley 1611 (2%) ✓✓ ✗ ✗ ✗ ✗ ~ ~ ✗ ✗ ✗ ✗ ✓ ✓ ~ ✗
Eythorne, Kent — eliminated, scored for calibration ✓✓ ✓ ✓ ✗ ✓✓ ✓✓ ✓✓ ~ ✓✓ ✗ ~ ✓ ✗ ~ ✗

The three record-state rows — no record survives (21%), documented but never assembled (11%), documented but in a source not yet searched (5%) — are not people and cannot be scored. Together they hold 37%. C(i) is eliminated on his own age and carries no share; he is scored here for calibration only, as Eythorne is. Evidence

The structural finding: two disjoint families

The candidates fall into two families, and no candidate is in both.

Family one — documented households with the wrong shape. C(ii), Hitcham, Ackworth and the eliminated Eythorne each have a household: a father, sometimes a mother, children with dates. What they lack is the right children. C(ii) has a John and a Richard in the right order but no Mary in the parish at any date. Hitcham has a Mary satisfying the hard bound and nothing else at all. Ackworth has the father and mother right and the children wrong. Eythorne has almost everything and is eliminated by two burials.

Family two — documented men with no household at all. Candidate B, Cheddington 1608, A(1603), Candidate D and the two Weston Turville men are, respectively, the best trade-and-corridor fit, the only candidate reached by prediction from the naming custom, the exact age match, the best trade match, and the best combination of age, trade and colonial-side channel. Every one of them scores zero or unknown on attributes 1 through 4. They are men without wives or children on record.

The emigrant needs a man of the right age and his household, and every candidate supplies one half. That is the actual structure of this problem, and it is the strongest argument in the file for the record-state rows: the man’s own baptism and his household’s baptisms are probably not in the same searchable place — one side is lost, unindexed, or in a class not yet opened. That is precisely what the not-yet-searched and no-record rows describe. Evidence

Per-candidate notes where the score needs explaining

Candidate B is contradicted by attribute 10. Its whole thesis is that John’s father was Francis; the shape predicts Richard. B is strong on the contextual attributes — corridor, trade lineage, and the 1634 total land sale as a push event squarely inside the attribute-14 window — and empty on every attribute that describes an actual household, because no document names a John of this family, a wife, or a child. B is a context without a household.[1]

Ackworth is the only candidate scoring attributes 5 and 6 together, and attribute 9 is its strongest card. The shape’s explicit allowance for a second-marriage hypothesis is exactly what the standing objection to Ackworth does not accommodate: the household is ruled out because Mary Shed’s 1647 Braintree marriage puts her parents’ marriage by about 1630, which a June 1636 marriage cannot produce — decisive only if 1636 was John’s first marriage, and the register entry does not say. It gives two names and the word married, which is the usual form for this parish and this hand, so its silence on marital status is not itself informative. Against it: neither manuscript witness to the 1637 baptism names a parent, and the entry now has three competing readings — John Thomas Gurnoe on the original face at 28 January, John Gurnoe junior on the facing leaf at 19 January, and the providers’ John Thomas Gurnoe at 19 January. A “junior” reading is neither a compound forename nor the “John son of Thomas” artefact, and it would attach the baptism to the 1636 marriage rather than detach it.[2][3]

C(ii) is the best household in the file and the worst on attribute 2. John 1624 and Richard 1626 give two of three colonial sons in the right order; the father’s own age, c.1598–1602, is the best documented fit to “aged 50 or thereabouts”; and its attribute-14 case is the strongest of any candidate — two children buried within three weeks in February and March 1638/9, a documented family catastrophe inside the exact window. Attribute 14 is stronger than it looks on the index alone, because the register image assigns the parish’s other burials of those eighteen months to other households: Joane of November 1637 is a widow and Richard of June 1638 is the sonne of Thomas. What C(ii) suffered was a two-child catastrophe, not a household collapse that would explain the register silence without an emigration; and the same reading leaves his Richard of 1626 with no recorded death anywhere, so attribute 3 is a live child rather than a buried one.[4] But no Mary Gurney is baptised at Great Berkhamsted at any date under any spelling or any father, 1544–1888, and no John Gurney marries anywhere in Hertfordshire between 1600 and 1654. Attribute 10 is contradicted: C(ii)'s father was C(i), a John.[5]

Hitcham stands at 3% after attribute 6 was damaged. Its case is attribute 2 — a Mary baptised 22 January 1630/1, satisfying the hard bound — plus attribute 5. But Hitcham is a true singleton: one Gurney baptism in the entire Buckinghamshire index, with no mother recorded. Five miles away, John Gurney married Susan at Farnham Royal on 30 March 1629, and that couple has no children recorded anywhere in the seventeenth century, a national parent search returning nothing before 1770; Farnham Royal itself records no John Gurney children. So there is a named, documented, childless-on-record John Gurney couple within five miles and twenty-two months of the Hitcham baptism, whose wife is Susan. That does not eliminate Hitcham, but the mother of its one child is unknown and now has a plausible alternative who is not a Mary.[6][7]

Weston Turville is the strongest contextual profile in the file, and it is two men rather than one. A man buried there on 4 February 1625 cannot father a child baptised there on 2 September 1627, so the parish held at least two John Gurneys and the old single row was scoring a composite.

The row that matters is the newer one. John Gurney, baptised 6 May 1604 to Thomas Gurney, scores attribute 7 at full strength — forty-nine at the 1653 deposition, the closest documented fit in the corpus — and attribute 13 through his father, a documented parish tailor, which is how a country tailor’s craft actually passed. Attribute 12 is now evidenced rather than asserted: a primary deposition of 1688 names Wendover, Lee and Ashton Clinton as the home parishes of four documented Weymouth settlers, two of whom hold or bound John Gurney’s own Weymouth land, and Weston Turville adjoins Ashton Clinton. He is the only candidate in the field carrying attributes 7, 12 and 13 together. His household attributes are ~ rather than ✗: they are untested, because nothing yet says whether he is the 1627 father, the 1625 burial, or a man who left. Attribute 10 is a miss — his father is Thomas, the same miss as C(ii) and D.[8]

The 1627 father keeps the old scoring as a residual, and may simply be the same man.

Candidate D holds the best direct trade match — his father Robert was described as a tailor at Old Change — and an excellent attribute 7 from his 1623/4 Drapers’ admission by redemption. Empty household, attribute 10 contradicted.

The Newgate apprentice fails attribute 7 decisively — birth c.29 September 1615, twelve years off — and cannot satisfy 1 through 6 at all, being an indentured servant. Its only strength is attribute 12, which it holds absolutely: he was physically in Massachusetts in 1636.

The two rows added at the close of the work both illustrate the split rather than breaking it. Gurley of Westminster is family one taken to its limit: father John, mother Mary, a daughter Mary baptised 1631 satisfying the primary bound exactly, and the household’s records stopping in 1633 — then no Richard, no John, three surplus daughters, and a surname requiring an n→l confusion that sits at the bottom of the plausibility scale and was reached only by the provider’s own variant expansion.[9] Epping / North Weald is the first candidate that is mostly ~ rather than mostly ✗: its household attributes are untested, not absent, because the coverage negative that would have closed them is void. It scores ✓✓ on attribute 10 for a reason no other candidate can match — the parish’s Gurn- line alternates John and Richard across three generations, which is the naming behaviour attribute 10 was derived from.[10] Evidence

Norwich St Peter Mancroft is the same profile as Epping with a better mother field and a worse naming argument. A child named John Gurny was buried at St Peter Mancroft on 10 February 1639, and the burial transcript indexes his parents: father John Gurny, mother Marg. That scores attribute 0 strongly — Gurny is a full-weight form — attribute 5 strongly, attribute 6 softly on the weighting adopted below, attribute 11 strongly and attribute 12 softly, since Norwich is the city of the 1611 Gurney–Rybett marriage and the whole Norfolk corridor; and attribute 14 softly, because the household is losing a child inside the departure window. Attributes 1 to 4 and 7 to 10 are ~ rather than ✗: no baptism, marriage or later record has been attached to this household, so its shape is untested, not absent. Against it, the buried child is himself named John while the colonial John Jr. survived to adulthood in New England, so this is not the colonial son, and nothing yet connects the household to any candidate. And the household is now challenged rather than merely untested: the Norfolk Record Office transcription published by Ancestry — an independent witness to the same register — does not contain the burial, with coverage controlled twice in the same collection. The scores above should not be raised before the original register is read.[11]

Eight corrections to the shape

In descending order of how much each would change a score.

1. Margaret belongs on the mother axis as a substantial soft hit. Surname and forename variants are a weighted scale, not a binary. In seventeenth-century hands g and y carry near-identical looping descenders, and Margaret was routinely contracted by clerks to Marg. or Mgt. to save paper and ink — so the one letter distinguishing the written token from Mary is exactly the letter most likely to be ambiguous. A Marg. written with a looping g reads as Mary to a later clerk, a nineteenth-century transcriber or a modern indexer, and a Mary is equally extendable to Margaret. The corruption is introduced by the transcription chain and runs in both directions. The corpus supplies a worked example: Tittleshall with Godwick, Norfolk, 1620 returns the same child twice, father Henry in both, with the mother indexed Anne in one transcript and Mary in the other.[12] The adopted weighting:

Mother token Weight on attribute 6 Basis
Mary, Marie, Marye, Maria, Marya, Mria full orthographic variants of one name
Marg, Marg., Mgt, Margaret, Margarett, Margret substantial (soft hit) g/y descender ambiguity plus routine abbreviation; corruption runs both ways
Margery, Margerie partial shares the abbreviated stem Marg. but is a distinct name
Molly, Polly excluded in this period eighteenth-century hypocorisms, anachronistic before c.1700

The same logic governs the father axis, and had already been in force in the sweeps: the John family is scored across John, Jhon, Jon, Jno, Joh, Johis, Johannes, Johan, Johane, Johanne, and — importantly — Jone and Joan, which in these indexes are sometimes a Latin or contracted Johannes misread as the female Joan. The county-set data contains live instances of both directions of failure: at Norwich St Giles in 1618 the father of Joseph Garnam is indexed Jone, and at King’s Lynn in 1615 the father of Gregorye Gurnall is indexed Katheryne.[12]

Admitting Margaret changed no candidate score on the baptism arm, and that is the whole extent of the negative. Every instrument in the stack that produced it indexes baptisms; the burial arm was never run, and when it was, it returned a father-John, mother-Marg. Gurney household at Norwich St Peter Mancroft on the first pass. The standing result is therefore “no such household in the baptism registers searched,” not “no such household anywhere,” and it must be stated that way rather than repeated.[13] The obvious instrument for closing the gap does not exist: FamilySearch’s record-search parent fields rank but never filter — a deliberately nonsensical mother name returns essentially the whole surname set with the same top hit as no filter at all — and a wildcard in the mother field fails closed to zero. No parent-axis negative may be built on that instrument, so the burial and marriage arms have to run on FindMyPast and FreeREG with positive controls.[14] Within the baptism arm the sweep is real. Across FreeREG’s 585 captured rows, five households carry a Margaret-family mother — Norwich 1638, Bardwell in Suffolk 1643 and 1645, and Mamble and Bayton in Worcestershire 1637 — and none has a father in the John family. Across 705 rows from eight FindMyPast county baptism sets, seven rows carry a Margaret-family mother, to fathers Robert, Edward, Michael and William, and none has a father in the John family at any date. The national cross-collection parent search bounded 1610–1650 returns eight rows for father John with mother Margaret, not one of them a genuine Gurney — the surname column reading ?, Fosse, G...Ds and G?, fuzzy-match noise from Cornwall, Cheshire and Scotland. The negative is interpretable because the identical query shape with mother Mary returns the known Eythorne household: the instrument was working when it returned nothing for Margaret.[15]

2. Attribute 2 has a primary bound and attributes 3 and 4 do not. Mary Gurney married Daniel Shed at Braintree in 1647, so at a minimum marriage age of sixteen she was born by 1631 — a primary constraint. The “c.1628” is compiled tradition from Sprague (2001) and the History of Weymouth (1923), and Richard c.1630 and John c.1633 rest entirely on those same compiled estimates with no primary record behind them. Score attribute 2 against “born by 1631” and treat c.1628 as soft; mark 3 and 4 explicitly as compiled-estimate attributes worth less than 2. This matters directly — Hitcham’s Mary of January 1630/1 satisfies the primary bound exactly while looking three years off the traditional date.[16]

3. Attributes 8 and 9 are mutually exclusive, so a candidate scores one or the other. Scoring both double-counts the marriage.

4. Attribute 10 is a derived prediction, not an independent observation. It comes from reading the naming custom forwards — the eldest son takes the paternal grandfather’s name — and it is conditional on Richard being the eldest son. It directly contradicts Candidate B, whose father is Francis, and C(ii) and D, whose fathers are John and Robert. A hit on 10 is a hit on a hypothesis, and a miss is weak evidence, because naming customs were followed loosely.

5. Add a departure attribute, with its coverage control attached. The household’s records stop between 1635 and 1641 with no burial or probate for the head, in a parish whose burials are demonstrably covered. This is separable from attribute 14 — a push event explains why he might go; this records that the household stops. Without the coverage control it is worthless: two false departure gaps have already been produced by omitting it. On a properly controlled test only Ackworth currently passes, and it now passes at a standard no other candidate approaches: 170 Smith burials in the parish against zero Gurney burials at index level, an image-level sweep of the register across 1617–1628 and 1632–1644 finding no Gurney-family form beyond the two known entries, and a national absence of the surname Gurnoe anywhere in England between 1637 and 1762.[17][3]

6. Add a penalty for children with no colonial counterpart. A household with the right three children and no others fits better than one with the right two and four extras. C(ii) carries Michaell and Elizabeth; the Gurley household carries Elizabeth, Ellynor and Alexander. As specified, the shape can only reward matches and never penalise surplus.

7. Attribute 0 is a weighted scale, not a single tick. Gurney / Gurny / Gourney / Gurnie at full weight; Gurne / Gurn as near-certain contractions; Gurnard / Gurnat at moderate-to-high, because the Heralds’ Visitation of London of 1634 records letters patent of 26 July 1633 to “Richard Gurny, alias Gurnard, Sheriff of London” and to “Burton Gurny, alias Gurnard, descended from the Gurneys of Kendall,” and the Norfolk line’s own canting crest is a gurnard fish borne by Thomas Gournay I before 1465; Gorne / Garne / Gernne as ordinary vowel drift; Gurnett / Gurnel / Gurner lower — and lower is never exclusion. Garne in Gloucestershire and Gurley in Westminster are probably distinct surnames with their own lineages, but that is a weight, not a class boundary.[18]

8. Peter, Isaac and Sarah are excluded from scoring, not treated as missing children. No Peter Gurney-variant baptism exists anywhere in Great Britain 1627–47 and no Isaac 1632–52; both men first appear in colonial records, so their absence in England is expected and must not count against a candidate. The same applies to Sarah, who has no primary record at all and is absent from Torrey’s compendium of New England marriages before 1700.[19]

A ninth rule arrived from the elimination side. Three undated FamilySearch role-row households, one of them earlier described as the single most important loose end in the work, were dated on a second instrument and resolved to 1687, 1704 and 1710 — every one late, none within a century of the window. An undated role row is not evidence of an in-window household until it is dated elsewhere.[20] Evidence

Counting is the wrong metric

Because attribute 0 is a scale, a count of Gurney-variant records is not a meaningful quantity: the total depends entirely on where the scale is cut. The object of the search is an individual record matching a corner of the colonial shape, not a total of Gurney baptisms. Any statement phrased as a variant count should be restated as a match or partial-match statement, with the weight named.[18]

Crosslinks

  1. Identity — Candidate B: Norfolk parentage. Cross-reference, not a source; the underlying records are cited there. ↩︎

  2. Identity — Ackworth, Yorkshire. Cross-reference, not a source. The “John son of Thomas” reading is the same indexing artefact already caught at Toddington, Bedfordshire, where a compound forename in the index resolves to a relationship phrase in the register. ↩︎

  3. Ackworth St Cuthbert parish register, read at image level 2 August 2026 in Ancestry collection 2256, West Yorkshire, England, Church of England Baptisms, Marriages and Burials, 1512–1812 (West Yorkshire Archive Service partnership), on two parallel films of the same run, 32355_249046 and 32355_249045. The 6 June 1636 marriage is at image 01583 and reads “John Gurnoe & Mary Burton married — June 6”, with no marital status at 98% confidence; the January 1637 baptism is at image 01584 with the alternate-hand reading on the facing leaf. The whole-page sweep covers 1617–1628 and 1632–1644 with every visual near-match adjudicated; 1629–1631 is not held. Diplomatic readings, confidences and the excluded near-matches at sources/corpus_supplement/ackworth-st-cuthbert-register-1617-1644-gurnoe-burton-readings.md. Source ID: ancestry-west-yorkshire-1512-1812. ↩︎ ↩︎

  4. Parish register, Great Berkhamsted, Hertfordshire, burials, page 165 of 175 in the Findmypast Hertfordshire Burials image series (record GBPRS/HERT/007567815/00277, transcript GBPRS/D/72723126); master retrieved 1 August 2026, SHA-256 21beb9013d738bef7980f8b1a1da3af0e109213ccbcbfadf53bc7db80cfd46e5, held local-only under sources/media/findmypast-hertfordshire-burials/_local/ because the register is © Hertfordshire Archives & Local Studies. The entries read “Richard the sonne of Thomas Gurney was buried the xxviij of June” and “Joane gurney a wido was buried the vij of November.” Diplomatic transcription at sources/corpus_supplement/findmypast-hertfordshire-berkhamsted-burial-page-165-1638.md. Source ID: findmypast-hertfordshire-burials. ↩︎

  5. Identity — Berkhamsted: Candidates C(i) and C(ii). Cross-reference, not a source. ↩︎

  6. FindMyPast Buckinghamshire Baptism Index, authenticated session, 29 July 2026: datasetname=buckinghamshire+baptism+index&sid=103&lastname=gurn*&keywords=hitcham returns one row for the parish in the whole index — “Mary Gurny, 1631, father John, mother blank, Hitcham.” The same set scoped to Farnham Royal returns four rows, none to a father John: Ann Gurney 1623 and Rebecca Gurney 1630 (father Silvester), Thomas Gurney 1630 (father Thomas), and Eliza Gurney 1855. Source ID: findmypast-bucks-baptism-index. ↩︎

  7. FindMyPast cross-collection parent search, 29 July 2026: sid=102&fatherfirstname=john&fatherfirstname_variants=true&motherfirstname=susan&motherfirstname_variants=true&fatherlastname=gurney&fatherlastname_variants=true, no year bound → 49 rows, every one 1770 or later. So no John Gurney and Susan household has any recorded child in the seventeenth century. The marriage is John Gurney and Susan at Farnham Royal, 30 March 1629, from the Buckinghamshire Marriage Index; the couple is independently documented in Buckinghamshire Archives D-BAS/40/303 and /305 (1633), where John Gurney of Farnham Royal, yeoman, and Susan his wife both sign by mark, Susan being named as daughter of Oliver Cogeram. Source IDs: findmypast-uk-parish-baptisms; findmypast-bucks-marriage-index; bucks-archives-catalogue. ↩︎

  8. Identity — Candidate A and the Buckinghamshire cluster, the Weston Turville section — cross-reference, not a source. The parish events are FamilySearch, England, Buckinghamshire, Church Records, 1217–1994 (collection 4380170): John Gurney baptised 6 May 1604, father Thomas Gurney; John Gurny buried 4 February 1625; Anna Gurney baptised 6 June 1624, father Thomas Gurney; Elyzabethe Gurney baptised 2 September 1627, father John Gurney. Thomas Gurney of Weston Turville, tailor, 1633, is in the published Buckinghamshire probate index, and his son Francis’s 1622 Clothworkers’ apprenticeship is in the livery-company sweep. The 1688 deposition of Jonas Humphrey and Robert Randall naming Wendover, Lee and Ashton Clinton is Massachusetts Archives vol. 129, p. 16, printed in the History of Weymouth, Massachusetts (1923); Randall holds John Gurney’s East Field two acres and William Brandon of Ashton Clinton bounds his Mill Field in the 1643 Weymouth property record. Source IDs: fs-england-buckinghamshire-church-records-1217-1994; england-wales-published-probate-indexes; rollco-other-companies-gurney-1573-1653; history-of-weymouth; weymouth-land-grants-book-ms. ↩︎

  9. FindMyPast, authenticated session, 29 July 2026, cross-collection parent search with fatherfirstname=john&fatherlastname=gurley unexpanded, bounded 1613–1653: six results, of which five are the Westminster household — Eliz 1627, Ellynor 1628, Alexander 1629, Mary 1631, Sara 1633, all to father John and mother Mary, in England Births and Baptisms 1538–1975 — and the sixth is an unrelated John Gurley at Berwick-upon-Tweed, Northumberland, 1625. No Gurley baptism to this father is recorded after 1633. The household is confirmed on the Westminster county set. Source IDs: findmypast-uk-parish-baptisms; findmypast-county-baptism-sets-2026-07-sweep. ↩︎

  10. FindMyPast Essex Baptisms, authenticated session, 29 July 2026, lastname=gurn*&keywords=epping, no year bound: 28 results, every event double-indexed as both Epping, All Saints and Epping Upland, All Saints, giving 14 distinct events. The seventeenth-century sequence is Richard Gurnat 1616 (father John); John Gurne 1622 (father Robert, mother Sara); Richard Gurnet 1649 (father Richard); John Gurnet 1654 (father John); Elisabeth Gurnet 1659 (father John, mother Anne) — a line alternating John and Richard across three generations, in the county with the strongest Great Migration weight. No Gurn- baptism earlier than 1616 appears in the set for either parish, and whether that is absence or absent coverage cannot be settled on this instrument, because a year bound and a place keyword cannot be combined in a single-dataset query. Source ID: findmypast-county-baptism-sets-2026-07-sweep. ↩︎

  11. Ancestry, Norfolk, England, Church of England Baptisms, Marriages, and Burials, 1535–1812 (collection 61045, in partnership with the Norfolk Record Office), searched 2 August 2026. A Gurn* wildcard scoped to St Peter Mancroft returns the parish’s whole Gurney-variant run — a baptism of 1560, then nothing until 1713 — with no seventeenth-century entry of any kind. Controls in the same collection: Smith at St Peter Mancroft returns a burial of 28 February 1641, and a Gurn* burial search across all Norfolk for 1633–1645 returns Anna Gurney at Hempnall, 6 April 1639. The place parameter here ranks rather than filters, so these are ranked reads of a covered parish rather than enumerations. Source ID: ancestry-norfolk-1535-1812. ↩︎

  12. FindMyPast, authenticated session, county baptism and marriage index sets, sweep of 29 July 2026; 705 captured rows across Norfolk, Essex, Northamptonshire, Oxfordshire, Cambridgeshire, Middlesex, Westminster and Berkshire, each against the stems gurn*, gourn*, gorn*, gern*, girn*, garn*. Tittleshall with Godwick duplicate: Norfolk Baptisms Index, 1620, Katerine/Katherine Gurnay, father Henry in both rows, mother indexed Anne in one and Mary in the other. Norwich St Giles 1618, Joseph Garnam, father indexed Jone; King’s Lynn St Margaret with St Nicholas 1615, Gregorye Gurnall, father indexed Katheryne. Working dataset and per-query coverage ledger at sources/intake/archive/g13-july-2026-sweeps/fmp-county-sweep-2026-07-29/results.tsv and PROGRESS.md. Source IDs: findmypast-county-baptism-sets-2026-07-sweep; findmypast-norfolk-baptisms-index. ↩︎ ↩︎ ↩︎

  13. Findmypast, England Deaths & Burials 1538–1991, transcript R_272863566, read in an authenticated session 31 July 2026: John Gurny, burial 10 February 1639, St Peter Mancroft, Norwich, Norfolk; father John Gurny; mother “Marg.” Index transcript only; no register image has been opened, and no baptism, marriage or later record is yet attached to the household. The same underlying event, through the FamilySearch index, is what the March 2026 working case file read as an adult John Gurney’s death and used to eliminate a Norwich St Peter Mancroft candidate; because the transcript indexes parents, the deceased is far more likely a child, and that elimination is void. Fuller treatment at research/topics/john-gurney-candidate-others.md — cross-reference, not a source. Source IDs: fs-england-deaths-burials-1538-1991; fs-england-births-christenings. ↩︎

  14. FamilySearch record search, controlled 2 August 2026, all queries f.recordCountry=England: q.surname=Gurn* → 90,074; with q.fatherGivenName=John → 73,356; with q.motherGivenName=Margaret → 71,965; with the deliberate nonsense q.motherGivenName=Xyzzyqq → 71,523, and the same top hit as the unfiltered query. The parent-name parameters therefore rank rather than filter, extending the already-recorded behaviour of the date, place and spouse parameters. Separately q.motherGivenName=Marg* returns a flat zero against 71,965 for Margaret, so a wildcard in that field fails closed — a silent zero-maker of the class catalogued on FindMyPast (keywords on a wrong place token; sortby accepted and ignored). Source ID: fs-england-births-christenings. ↩︎

  15. FreeREG (Free UK Genealogy), sweep of 28 July 2026, 585 captured rows across the Gurney surname family, re-cut for Margaret-family mothers on 29 July 2026: Norwich 12 August 1638 (Ellin, father Will, mother Margret); Bardwell, Suffolk, 18 June 1643 and 18 May 1645 (Robert, father Edward, mother Margarett); Mamble and Bayton, Worcestershire, 17 September 1637 (John Gurner, father Thomas, mother Margery). Working dataset at sources/intake/archive/g13-july-2026-sweeps/freereg-g13-sweep-2026-07-28/results.tsv with per-query coverage in the accompanying PROGRESS.md. Source ID: freebmd-freereg. FindMyPast cross-collection parent search (sid=102, Great Britain, parish baptisms), 29 July 2026, father forename John and surname Gurney both variant-expanded, bounded yearofbirth=1630&yearofbirth_offset=20: with motherfirstname=margaret variant-expanded, 8 results with no Gurney-family surname among them; with motherfirstname=mary variant-expanded, 16 results including Eythorne, Kent (John 1638, Thomas 1635, and the duplicate transcriptions Gorney Edward 1641, Gurnay John 1638, Gurner Thomas 1635) and the Westminster Gurley household. The Mary query is the positive control for the Margaret query: same shape, same bounds, known household returned. Source ID: findmypast-uk-parish-baptisms. The county-set half of the negative is the 705-row sweep at [12]. ↩︎

  16. Mary Gurney’s marriage to Daniel Shed at Braintree in 1647 is in the Braintree town records (Samuel A. Bates, ed., Records of the Town of Braintree, 1640 to 1793, 1886) and in Torrey, New England Marriages Prior to 1700. The compiled birth estimates for Mary c.1628, Richard c.1630 and John Jr. c.1633 are Waldo Chamberlain Sprague, Genealogies of the Families of Braintree, Mass., 1640–1850 (Boston: NEHGS, 2001), p. 695, and the History of Weymouth, Massachusetts (1923), vol. 3, p. 251. Source IDs: braintree-records-1640-1793-1886; torrey-new-england-marriages-prior-1700; sprague-braintree; history-of-weymouth. ↩︎

  17. The two false departure gaps are the Berkhamsted “no Gurney burial at any date” claim, disproved by sixteen Gurney-variant burials 1579–1643 in the same parish, and the Stewkley “register silence after 1614” claim, disproved by baptisms of 1615, 1616, 1618 and 1626. Both are worked at 33-berkhamsted-candidates.md and 31-candidate-a-aylesbury.md. The Ackworth control is 170 Smith burials against zero Gurney burials in the parish, from FindMyPast’s Yorkshire burial coverage; the elimination-instrument argument is 71-refactor-burials-as-the-elimination-instrument.md. Cross-references, not sources. ↩︎

  18. Heralds’ Visitation of London, 1634 (Harleian MS 1476), printed in Daniel Gurney, The Record of the House of Gournay, Part III (1848), p. 533, recording letters patent of 26 July 1633 to “Richard Gurny, alias Gurnard, Sheriff of London,” and to “Burton Gurny, alias Gurnard, descended from the Gurneys of Kendall.” The Norfolk line’s canting gurnard-fish crest, borne by Thomas Gournay I before 1465, is on the G21 fact sheet. The retraction of the earlier claim that Garne, Gurnett, Gurnat, Gurnard, Gurnel and Gurner are “other surnames” — a conjecture built on five index rows spanning three generations with no register image read — is 69-refactor-essex-tested-and-the-surname-confusion.md. Source IDs: dg-rec-pt3; heralds-visit-london-1633. ↩︎ ↩︎

  19. FindMyPast, all parish-baptism collections, wildcarded across the Gurney surname family, Great Britain: no Peter baptism 1627–1647 and no Isaac 1632–1652. Peter first appears in George Madison Bodge, Soldiers in King Philip’s War (1891), p. 114 (“Peeter Gurnay,” Weymouth, Dedham muster 10 December 1675); Isaac in the Plymouth Colony court order of 1 March 1663/4. Sarah has no primary record and does not appear in Clarence Almon Torrey, New England Marriages Prior to 1700. Source IDs: findmypast-uk-parish-baptisms; bodge-soldiers-king-philips-war-1891; familysearch-fulltext-search; torrey-new-england-marriages-prior-1700. ↩︎

  20. FindMyPast, authenticated session, 29 July 2026, single-dataset queries (sid=103). Sussex Baptisms, lastname=gurn*&firstname=grace: 3 results, all Grace, Horsham 1704, father John, mother Mary (indexed twice as Gurne, once as Gurney). Herefordshire Baptisms, lastname=gurn*&firstname=mar*: 57 results, among them Mary Gurney, Stoke Lacy 1687, father John, mother Margaret. Warwickshire Baptisms, lastname=gurn*&firstname=ann*: 16 results, among them Ann and Anne Gurney, Coleshill 1710, father John, mother not indexed. Each dates a household previously carried only as an undated FamilySearch role row. FamilySearch’s role-aggregated rows carry no event date and its date filter does not bind on them, which is what produces the illusion of an in-window household. Source ID: findmypast-county-baptism-sets-2026-07-sweep. ↩︎