Gurney Genealogy Library Updated 4 August 2026 by Allen Gurney

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

Identity — assessment

John Gurney’s English origin is open. No record made in his lifetime names his father, and the work of July and August 2026 did not find one. What it did do was close the searchable index layer on three axes and change the shape of the answer: the leading named candidate stands at about eleven percent, and three rows that describe the state of the record rather than a man hold thirty-seven percent between them — more than every named candidate combined.

That is not a retreat into “nothing is known.” A great deal has been eliminated, and the eliminations are findings. What changed is that the elimination stopped producing a survivor.

The structural finding

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

Documented households with the wrong shape — the younger Berkhamsted John, Hitcham, Ackworth, and the eliminated Eythorne — each have a father, sometimes a mother, and children with dates. What they lack is the right children. Documented men with no household at all — Candidate B, the Cheddington 1608 John, the 1603 Stewkley baptism, Candidate D and the two Weston Turville men — are 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 Weymouth channel respectively, and every one of them scores zero or unknown on all four child attributes.

The emigrant needs a man of the right age and his household, and every candidate supplies one half. 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 the strongest argument in the file for the record-state rows, and it is why they lead.[1] Evidence

The probability model

Nineteen rows, summing to 100. Nothing is estimated more finely than a whole point, and a row at two percent means “possible but poorly supported” rather than a measured quantity.

Candidate / category Basis
No record survives or is indexed anywhere 21% The searches run against this are broad — livery-company rolls, Norwich freemen and apprentice registers, Prerogative Court and archdeaconry probate across five counties, four parish-register providers, and the full child, parent, mother and marriage axes. Seventeenth-century registers are lost, defective and unindexed at a rate that makes a whole household’s disappearance ordinary rather than remarkable. Raised by one point from Ackworth, whose third support failed with nothing rising to take its place — a candidate weakening with none rising is exactly what this row describes.[2]
Documented but never assembled into a person 11% A John Gurney whose records do appear in the surveyed sources but who has never been reconstructed as a man: scattered single events never linked, or records wrongly merged into another household. The mechanism is the best-evidenced in the file — Candidate A resolved into two men, Stewkley into two households, Berkhamsted into a father and a son and then into a parish of four — but every one of those defects was found and resolved, which drains this row rather than filling it. Reduced by four points, which fund the Weston Turville rows below: a baptism, a tailor father, a Clothworkers brother and a parish adjacency all sat in the corpus and had never been joined, which is precisely this row’s own mechanism firing again. The Norwich St Peter Mancroft household is the same move made earlier.[3]
B — son of Francis Gurney and Margaret Rybett 11% The leading named candidate and the only one with a circumstantial case: the verified 1611 Norwich marriage, a documented Norfolk family in the right window, the 1634 liquidation as motive, the corridor, the Puritan network, and a trade that ran in the family. Empty on every attribute describing an actual household, and no document names a John of this family at all — which is the comparison that matters against the 1604 Weston Turville row below.[4]
Ackworth, Yorkshire 8% The only unaccounted-for John × Mary marriage in England before 1660, and the only candidate scoring a father John and a mother Mary together. The block is the date: a June 1636 marriage postdates the colonial daughter Mary’s birth, so it can only be the emigrant’s as a second marriage, and the register entry gives two names and the word married — the usual form, which settles nothing either way. The departure signature is the strongest in the file, image-verified across twenty-four years plus a national surname absence 1637–1762; and the compiled-tradition support fails outright. One support fails, one strengthens, one is untouched — hence one point, not more. Reclassified from an origin candidate to a possible departure point: the groom was an incomer and Ackworth is the bride’s parish.[5]
John baptised Weston Turville 6 May 1604, son of Thomas the tailor 5% The best documented age fit in the corpus, and it had never been scored — filed under “generational predecessors” on the mistake that a baptism is a fathering event. Forty-nine at the 1653 deposition of “aged 50 or thereabouts”; son of a documented parish tailor, which is how a country tailor’s trade actually passed; in a parish adjoining Ashton Clinton and two to three miles from Wendover and Lee, the three parishes now documented to have sent men to Weymouth. He is the only man in the field carrying age, trade and a colonial-side channel together. Held down hard by an unresolved parish burial of 4 February 1625 that could be him, and by having no household. Explicitly a holding value pending one document: the 1633 will of his father resolves the row to roughly zero or roughly ten.[6]
John baptised Cheddington 1608, son of Richard 6% The only candidate reached by prediction rather than by noticing a name match: the naming custom read forwards requires a paternal grandfather named Richard, and this is the household it points to. Father’s probate 1626. No household of his own on record.[7]
Documented, but in a source or class not yet searched 5% Split out of the assembled row, and evidenced rather than speculative. The listable residue: the mother axis has never been run on burials or marriages, and the one time a burial arm was opened it returned a father-John, mother-Margaret-family household at the first attempt; the national parent pool misses whole county sets; there is no Suffolk baptism set; three populous counties have no marriage set; thirty-one county-sweep queries were capped, of which seventeen are now enumerated in full; the south-Midlands licence series has never been searched; East Dereham’s 1611–15 returns are lost. Berkshire marriages have come off this list: the set has now been run.[8][9]
C(ii) — the younger Great Berkhamsted John 6% The best documented age fit in the field and the strongest push event — two children buried within three weeks in early 1639. The register image removes the objection that used to sit against him: the Richard buried in the parish on 28 June 1638 is the sonne of Thomas, so his own Richard of 1626 leaves the record alive. Still against him: no Mary Gurney in the parish at any date, and no John Gurney marriage anywhere in Hertfordshire 1600–1654.[10][11]
Norwich St Peter Mancroft — John Gurny and “Marg.” 2%, flagged A real household with a secure surname, a father John and a Margaret-family mother together — more than the Westminster Gurley household has, and in the city of the 1611 Gurney–Rybett marriage. It surfaced on the burial arm, which no mother-axis sweep had touched. Held rather than moved, and now challenged: the Norfolk Record Office transcription published by Ancestry — an independent witness to the same register rather than another copy of the International Genealogical Index — does not contain the burial, with coverage controlled twice. The row was entered on a single derivative transcript and a closer-to-source one disagrees. The original register is the deciding evidence and this household must not be weighted upward before it is read.[12][13]
D — son of Robert Gurney, draper of Old Change 4% The best direct trade match — a father described as a tailor at Old Change — and an excellent age from his 1623/4 Drapers’ admission by redemption. Held rather than eliminated because the 1662 hearth-tax identification, though strong, rests on continuity of an address across twenty-four years.[14]
The Epping and North Weald Gurnat–Gurnet line 3% The only candidate whose household attributes are untested rather than absent, and the only one scoring strongly on the naming prediction: the parish’s Gurn- line alternates John and Richard across three generations, in the Great Migration’s own county. The coverage negative that would have closed it is void.[15]
The 1636 Newgate apprentice 2% A structurally sound elimination on family status and chronology — born c.29 September 1615, twelve years off — held above zero only because its test depends partly on the colonial children’s birth years, which are compiled estimates.[16]
Weston Turville 1627 father 2% The residual of the row that was split. He fathered Elyzabethe on 2 September 1627, so he was born by about 1602 and was about fifty-one in 1653 — but he may simply be the 1604 man above, and no record says otherwise. Carried separately only because nothing proves the identification. Empty on the entire child core. The burial negative is genuine: the parish is covered, with Gurney burials at 1621, 1625 and twice in 1633, and the 1627 father is not among them.[17]
Hitcham head 3% Fathering 1631, so born by about 1606. The child baptised there is a Mary, and across an England-wide sweep on two independent axes this is still the only baptism in the country carrying a colonial child forename to a father John that belongs to no otherwise-accounted-for household. Against it: a true singleton with no mother recorded, no seventeenth-century burial coverage in the parish at all, and a documented John Gurney × Susan marriage five miles and twenty-two months away. The missing mother is probably an artefact of the document class: what survives is a bishop’s transcript, and the mother’s name at a baptism is exactly what a clerk copying a year’s entries for the diocese drops first. The parish register is a separate, richer document at the same archive and nobody has asked for it.[18][19]
Buckinghamshire, Hertfordshire and Bedfordshire heads cleared as a group 3% East Claydon, Chesham, Wing, Cublington, Hulcott, Linslade, Monks Risborough, Amersham, Ivinghoe, Haddenham and the rest. The Buckinghamshire arm is genuinely swept: the county’s John Gurney marriages 1572–1704 have been read from both sides and not one bride is a Mary. The Bedfordshire arm is not. Eaton Bray — a John Gurney baptised there 26 September 1606 to a father John, five miles from Stewkley and seven from Toddington — is the group’s strongest member on age and has never been worked.[20][21]
A(1603) — the Stewkley baptism, if distinct 2% Still among the closest age matches in the corpus to “aged 50 or thereabouts,” and still not fathering children anywhere visible. Reduced because the claim that he has no burial is false: the Buckinghamshire Burial Index records John Gurney buried at Stewkley 21 September 1615, inside a covered cluster. One image resolves him to elimination or survival — but it must be the parish register, not the bishop’s transcript the index is drawn from, because qualifiers such as “elder,” “younger” or “son of” are exactly what a transcript drops.[22][19]
John and Mary Gurley of Westminster 2% A documented household that fits the parents and the eldest daughter and fails the sons: father John, mother Mary, Mary baptised 1631 satisfying the primary bound exactly, records stopping 1633 — then no Richard, no John, three surplus children, and a surname at the outer edge of the variant scale.[23]
John baptised Stewkley 1611, son of Robert 2% A second, contemporaneous Stewkley household. About forty-two in 1653, a poor fit against “about 50,” and no record of his own after baptism.[24]
Other named and out-of-corridor candidates 2% Toddington, Winkfield, Eythorne, Earsham, the London and Middlesex households, and the out-of-corridor sweeps, each carrying an individuating record that places it elsewhere. Held rather than retired: naming two further households at the close of the work does not exhaust the unenumerated remainder.[25]

Candidates A(i), A(ii) and C(i) are eliminated outright and carry no probability. C(i), the elder Great Berkhamsted John, goes on his own age: fathering from 1597 puts his birth at c.1570–75, a ground immune to the same-name masking problem that weakens register-continuity eliminations.[10] A(i), the 1641 certificate man, is taxed in Northamptonshire and no longer in the Aylesbury half-hundred in 1641 and is a tenant at Walgrave in 1650; A(ii), the St Mary Aylesbury father, is fathering children in England 1638–1653.[26] Evidence

Reading the table

No option exceeds twenty-one percent, and the leading rows are not people. Three of them describe the state of the record rather than a man, and together they hold thirty-seven percent.

Each of the four leading rows implies different work, which is the practical value of separating them. The unassembled row is tested by re-sorting records already in hand — repeatedly productive through 2026, and it fired again in August when the Weston Turville tailor’s household was assembled out of four records the corpus already held. The not-yet-searched row is tested by opening named classes that have never been opened: the south-Midlands licence series, the Suffolk gap, the Buckinghamshire wills. The no-record row is not tested by records at all. Candidate B is tested by finding a marriage or a child in the Rivett-orbit parishes, or by a kinsman’s will naming a Gurney beyond the seas.

Y-DNA is not the lever on the no-record row that this file has been calling it. The FamilyTreeDNA Gurney project holds very few datapoints for this line; samples that small limit what any comparison can support; and across thirteen generations even a small per-generation probability that a son was not fathered by his recorded father becomes a materially plausible break in the genetic line. The test is therefore weak in both directions — a match would fix surname-level paternity at best and would not confirm any particular household, and a mismatch would not disprove one, because a single break anywhere in thirteen generations produces the same result. The real levers on that row are the record classes never opened: the south-Midlands archdeaconry marriage-licence series, the Suffolk baptism gap, and the Buckinghamshire wills.[27]

Why the unassembled row is smaller than the merge defects would suggest. That row describes a man hiding in plain sight — records already searched, never joined up into a person. Merge defects of exactly that kind keep turning up: Candidate A, Stewkley, Berkhamsted, Denton, Toddington, Linslade, Kelham, Bishop’s Stortford. But each one was found and resolved, and a resolved defect drains this row rather than filling it.

The residue is also now countable, which it was not while the father axis was incomplete. The completed sweep yields thirty-one localities for 1615–1649. Classified, the genuinely unattached single events in the colonial window that carry a colonial child forename number one — Hitcham — and Hitcham already has a named row of its own. Of the six “scattered single events” the row was once built on, four are assembled: Ivinghoe 1640 is the marriage of John Gurney and Rebecka Cooker on 22 October 1640, Great Kimble 1619 is the Owlswick father-and-son marriage settlement, and Weston Turville and Haddenham carry named rows. The two that remain, Aylesbury 1644 and 1669, fall outside the window entirely.[3]

Why Ackworth and Hitcham carry as much as they do. Not because new supporting evidence appeared for either, but because the funnel closed around them: when every other household in the searched corpus resolves, the two that do not carry more of the weight. That is a legitimate inference and it is also a warning — both remain single-record candidates with real blocks, and neither should be read as rising toward identification.

Why Candidate B does not gain from the emptying. B would benefit from an emptied corpus, because its auxiliary predicts a man who would leave almost no English paper. An auxiliary that explains away absence must not draw confidence from further absence. That rule is what holds B at eleven.

Falsifiability, and the two named records that have now been read. A probability table is exactly the kind of artefact that looks like rigour while resting on judgement. Two guards apply: each row names the evidence it turns on, and the numbers are stated so that they can be moved by a record rather than by a mood. Three specific records were named as able to move them materially, and two have now been read.

  • The Great Berkhamsted burial register for 28 June 1638 reads “Richard the sonne of Thomas Gurney”, which is the branch under which C(ii) was to recover what he had lost. He does, from 4% to 6%.[11]
  • The Ackworth parish register for 1636–37 was to settle three things at once. It settled one and closed the other two rather than answering them: the marriage states no marital status, so the date block cannot be broken from that document; the baptism names no parent on either of two manuscript witnesses, which disagree with each other; and the register cannot say whether “John Thomas” is a forename or a relationship phrase. A named record can close a route as well as open one, and that is what happened here. Ackworth moves one point, and it moves for a reason the file did not predict — the failure of its compiled-tradition support.[5]

The one that remains is the Epping and North Weald registers at the Essex Record Office. Ancestry now carries two Essex Record Office-derived transcriptions of them, but both are index-only, so the image route is unchanged.[28] Three new records join the list, and two are in one building: the 1633 will of Thomas Gurney, tailor of Weston Turville, and the Weston Turville burial register for 4 February 1625, which decide the 1604 John between roughly zero and roughly ten percent; and the St Peter Mancroft register for February 1638/9, which decides whether the Norwich row rests on a real entry or an indexing error. Evidence

What the eliminations do not clear

Clearing named households is not the same as clearing the record, and the counting was optimistic in a specific way: the older enumeration counted households as reconstructed, which repeatedly proves to be too few men. Candidate A was one household and is two; Candidate C was one and is two; Great Berkhamsted alone holds four contemporaneous Gurney households rather than the one it was carried as.

The residual is not an abstraction. The London hearth-tax returns of 1662–1666 list eleven Gurney or Gurny households across the City, Westminster and inner Middlesex — Walter Gurney at Greene Dragon Court in St Margaret’s Westminster, William Gurny with seven hearths on the south side of Fleet Street, William Gurney at Cock and Key Alley in St Dunstan in the West, Ann Gurney with five hearths at Two Crane Court in the same parish, and seven more — a denser London Gurney population than any of the candidate searches had surfaced.[29] None of them is a candidate: they are all alive and in London after the colonial John was dead. What they show is how many Gurney households the record can hold without any of them being visible to a search built around a name, a wife and a child set.

The emigrant’s own family leaves no English trace

Clearing other men is not the same as producing John’s own family in the record, and four independent sweeps have now failed to.

No John-and-Mary Gurney household in England baptises a Mary, a Richard and a John between 1627 and 1634. No Gurney household anywhere in Britain 1620–1640 has a father in the John family and a mother in the Mary family except Eythorne, Kent, which is eliminated on two burials — and one further household appears only when the surname is expanded to the outer edge of the variant scale, Gurley of Westminster, which fails on the sons. Every John Gurney-cluster groom in England 1600–1660 is now named with his bride, and not one marries a Mary — a negative sharper than “none found,” because nine Gurney men married Marys in the same holdings, so it is specifically the Johns who have none.[30]

The fourth negative is narrower than it was stated, and the correction matters. The mother axis, swept for the first time in July 2026, returned no Gurney household anywhere pairing a father in the John family with a mother in the Margaret family. Every instrument that produced that result — FreeREG, eight county sets, and the national cross-collection parent search — indexes baptisms. The burial arm was never run. It has now been run once, and it returned such a household at the first attempt.[31]

Only four of John’s children were born in England, at most. Peter and Isaac were almost certainly born in New England: no Peter Gurney-variant baptism exists anywhere in Great Britain 1627–1647, and no Isaac 1632–1652. That leaves Mary, Richard, John and possibly Sarah — and Sarah has no primary record at all and does not appear in the compiler Torrey’s survey of New England marriages before 1700.[32]

Sarah’s doubtful existence does not bear on the naming argument, despite appearances. It is tempting to reason that the prediction of a paternal grandfather named Richard depends on Sarah not existing, because Richard has to be the eldest child for the custom to point at him. That confuses two separate sequences: the naming custom counts son-slots and daughter-slots separately, so a daughter — real or not — cannot displace Richard from the eldest-son position. Evidence

A John-and-“Marg.” household at Norwich, on the record arm nobody had searched

A Gurney household with a father in the John family and a mother in the Margaret family exists, and it is in Norwich. The record is a burial: John Gurny, buried 10 February 1639 at St Peter Mancroft, Norwich, the transcript indexing father John Gurny and mother “Marg.”[12]

Three things follow.

  1. The standing negative must be restated rather than repeated. “No father-John, mother-Margaret Gurney household anywhere” was true of the baptism registers searched and was never tested on any other record class. This household surfaced the first time a burial arm was opened.
  2. It voids an old elimination. The March 2026 working case file cleared “Norwich, Saint Peter Mancroft” on the ground “died in England, buried 10 February 1639” — reading the entry as the burial of an adult John Gurney. A transcript that indexes parents for the deceased points to a child, so nothing about the record eliminates anyone, and the household behind it has never been assessed at all. That row does not survive into the current published case file, so until now the fact was carried nowhere in the live account.
  3. Marg. is a substantial soft hit rather than a miss, under the weighting the project already adopted: g and y carry near-identical looping descenders in these hands, and Margaret was routinely contracted to four letters and a mark of suspension, so the one letter separating the written token from Mary is the letter most likely to be ambiguous.

What it is not. The buried child is himself named John, and the colonial John Jr. survived to adulthood in New England, so this is not the colonial son. No baptism, marriage or later record has been attached to the household, and the mother’s name is known only from an index abbreviation. The row entered the table at 2%.

And a second transcription of the same register does not contain the burial. The Norfolk Record Office transcription published by Ancestry is an independent witness to the same parish register rather than another copy of the International Genealogical Index, and its St Peter Mancroft Gurney run is one baptism of 1560, then a gap of a century and a half. Coverage was controlled twice — the collection returns a Smith burial in that parish in 1641, and a Gurney-family burial elsewhere in Norfolk in the same years — so the absence is interpretable. It cuts against the row. The entry behind it may be a parish misattribution or a date error in the derivative index; equally, one derivative index disagreeing with another is not proof either way, and this interface ranks rather than filters on place, so these are ranked reads rather than enumerations. What it does establish is that the row was entered on a single IGI-derived transcript and the closer-to-source witness does not corroborate it. The original St Peter Mancroft register is now the deciding evidence, and the household must not be weighted upward before it is read.[13]

The next moves are that register and a Norwich sweep for the couple’s other children and their marriage — and the sweep needs a new design, because the obvious route through FamilySearch’s parent fields does not exist.[33]

The patronage question

The strongest network evidence in this case does not attach to John. In October 1627 Brampton Gurdon wrote to John Winthrop about the widowed “mrs. Gurny,” whose son’s wardship was moving through Winthrop’s own office at the Court of Wards; the Gurneys of West Barsham, in the decade John emigrated, stood inside the personal circle of the men who founded Massachusetts Bay.[34] That is a fact about the senior branch. Candidate B’s John is two steps from it — a nephew of the ward line through his father Francis — and no record connects him to the network at all.[35]

Whether John reached Massachusetts through that network, or simply out of the same county at the same time as thousands of others, is unresolved, and it matters: a documented patronage route would turn a plausible corridor into a mechanism. The test that would settle the chain is a patron-side search of the Norwich diocesan institution books for a 1627–28 presentation to the West Barsham vicarage made by right of the Gurney wardship. It cannot identify the emigrant; it can turn the Gurdon–Winthrop connection from strong construction into citable fact.[36] Evidence

Anderson’s “Boston”

Anderson’s directory gives John’s New England settlements as Boston and Braintree, and his origin as “Unknown.”[37] No Boston record names him. The Boston attribution is most plausibly the same conflation that produced the family tradition of an 1615 birth in Southwark: the 1636 runaway apprentice named John Gurney, a distinct and younger man, was bound to John Newgate the feltmaker and lived in Newgate’s Boston household — a servant’s position that would generate neither church admission nor property record.[16] A short unrecorded stay at Boston by the elder John, before he surfaces at Weymouth in June 1641, remains possible: the town was under a day from Weymouth and traded with it constantly. But a transient lodging is invisible in principle, so the possibility can neither be tested nor used. Evidence

Where the number can go

Three levers carry most of the remaining value, and they attack different rows.

One archive request, four targets. The highest-value action available on this problem is offline, cheap and unblocked: a single request to the Centre for Buckinghamshire Studies covering the 1633 will of Thomas Gurney the tailor (registered copy We 29 62, original file Wf 29 234), the Weston Turville burial register for 4 February 1625, the Hitcham parish register for 22 January 1631, and the Stewkley parish register for 21 September 1615. In each case ask for the parish register or the original will, not the bishop’s transcript — the Hitcham and Stewkley targets have been treated throughout as “the register entry” and are not; they are the thinner diocesan fair copies.[19]

Images, not indexes. What is left elsewhere is register images with specific predictions attached: Epping and North Weald Bassett at the Essex Record Office, where a line alternating John and Richard across three generations sits in the Great Migration’s own county and the index cannot settle whether an earlier John exists; and the St Peter Mancroft register, which now has two derivative indexes disagreeing over it. The Ackworth register has come off this list — it has been read, and it answered what it could.

Sources never opened. The mother axis on burials and on marriages, which produced a household the first time it was tried and has otherwise never been run — and which must be run on FindMyPast and FreeREG, because FamilySearch cannot carry it;[33] the south-Midlands archdeaconry marriage-licence series across three archdeaconries in two dioceses; the Suffolk baptism gap; the Norwich licence bonds of 1606–11; the Buckinghamshire wills; and the Eaton Bray household in the Bedfordshire sets.

And one question that would move Ackworth further than a point in either direction: whether Gurnoe is a real north-western surname or a Yorkshire clerk’s rendering of an incomer’s unfamiliar Gurney. The whole pre-1700 English Gurnoe corpus is four records in three households, on a Cheshire–Lancashire–London axis with nothing in Norfolk, Suffolk or Essex.[5]

A marriage record for John and Mary in the Rivett-orbit parishes around East Dereham, or a kinsman’s mention of a Gurney “beyond the seas” in one of the unread Norfolk family wills of the 1650s, would lift Candidate B materially. Failing that, continued negative searching leaves the identification where it is: open, with a leading named candidate at eleven percent and no candidate near identification.

Crosslinks

  1. 38-the-shape-and-the-two-families.md. Cross-reference, not a source; the underlying records are cited there. ↩︎

  2. The negative-search stack behind this row: the Merchant Taylors’ Company apprentice and freedom data (Source ID: ukda-9263-mt-apprentices-scott-2024); Millican’s Register of the Freemen of Norwich, 1548–1713 and the Rising–Millican index of Norwich apprentice indentures (Source IDs: millican-register-freemen-norwich-1934; rising-millican-norwich-apprentices-index-nrs29-1959); the Prerogative Court of Canterbury Gurney corpus and the UK Data Service PCC will index 1384–1858 (Source IDs: tna-pcc-gurney-elimination-corpus; ukda-4816-pcc-wills-index); the published England-and-Wales probate indexes (Source ID: england-wales-published-probate-indexes); FindMyPast’s parish-baptism, county and burial sets, FreeREG, Ancestry’s IGI-derived sets and FamilySearch (Source IDs: findmypast-uk-parish-baptisms; findmypast-county-baptism-sets-2026-07-sweep; freebmd-freereg; ancestry-england-select-births-christenings-9841; ancestry-england-select-marriages-9852; fs-england-births-christenings); and, for the Norfolk arm specifically, Ancestry’s Norfolk parish-register collection 1535–1812, searched on the same family signature with no matching cluster returned (Source ID: ancestry-norfolk-1535-1812). ↩︎

  3. The merge defects and their resolutions are worked at 31-candidate-a-aylesbury.md (Candidate A, Stewkley), 33-berkhamsted-candidates.md (Berkhamsted), and 36-other-eliminations.md (Denton, Toddington, Linslade, Kelham, Bishop’s Stortford). The father-axis enumeration yielding thirty-one localities for 1615–1649 is the 598-row parent sweep of 28 July 2026, working dataset at sources/intake/archive/g13-july-2026-sweeps/fmp-parent-sweep-2026-07-28/results.tsv with its coverage ledger. The count of more than forty distinct John Gurney heads of household in England between 1600 and 1670, against which this row is read, comes from the May 2026 supplementary same-name sweep across twelve Gurney surname variants and eighteen English parish, burial, probate and licence collections. Source IDs: findmypast-uk-parish-baptisms; findmypast-john-gurney-2026may-supplementary-same-name-sweep. ↩︎ ↩︎

  4. Identity — Candidate B: Norfolk parentage. Cross-reference, not a source. ↩︎

  5. Identity — Ackworth, Yorkshire. Cross-reference, not a source; the register images, the surname distribution and the tradition test are cited there. The register was read at image level on 2 August 2026 in Ancestry collection 2256, West Yorkshire, England, Church of England Baptisms, Marriages and Burials, 1512–1812, films 32355_249046 and 32355_249045; diplomatic readings at sources/corpus_supplement/ackworth-st-cuthbert-register-1617-1644-gurnoe-burton-readings.md. Source ID: ancestry-west-yorkshire-1512-1812. ↩︎ ↩︎ ↩︎

  6. Identity — Candidate A and the Buckinghamshire cluster, the Weston Turville section. Cross-reference, not a source. The records 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; Elyzabethe Gurney baptised 2 September 1627, father John Gurney (Source ID: fs-england-buckinghamshire-church-records-1217-1994) — with Thomas Gurney of Weston Turville, tailor, 1633, in the published Buckinghamshire probate index (Source ID: england-wales-published-probate-indexes) and the 1622 Clothworkers’ apprenticeship of his son Francis in the livery-company sweep (Source ID: rollco-other-companies-gurney-1573-1653). How the 5% was reached, stated so it can be argued with: a rough reading gives the 1604 man perhaps a 55–65% chance of being alive and available — he is either the 1625 burial, the 1627 father, or a man who left the parish — and conditional on being alive his contextual profile would justify roughly 10–12%. Five percent is close to that product once the 1627 father is carried as a separate 2% residual, and the pair is deliberately a holding value pending one document. ↩︎

  7. John Gurney baptised Cheddington 5 August 1608, son of Richard Gurney and Jane Founteine, the father’s probate 1626; reached by prediction from the naming custom rather than by name match. Worked at 54-refactor-naming-convention-and-the-cheddington-household.md and carried in 31-candidate-a-aylesbury.md. Cross-references, not sources. ↩︎

  8. The residue is itemised and each item evidenced at 73-refactor-the-funnel-closed-and-the-unassembled-row.md §“The reallocation” and 75-refactor-margaret-county-sets-and-loose-ends.md §6. The pool defect is proven rather than assumed: the national parent search returns neither the Lamport, Northamptonshire household nor the Earsham household, both of which FreeREG and the Norfolk county set hold. Source IDs: findmypast-uk-parish-baptisms; findmypast-county-baptism-sets-2026-07-sweep; freebmd-freereg. ↩︎

  9. The county-sweep coverage ledger at sources/intake/archive/g13-july-2026-sweeps/fmp-county-sweep-2026-07-29/PROGRESS.md, reconciled row by row on 31 July 2026. Its own summary said nine high-noise queries were capped; the rows themselves carry thirty-one with a status of exactly CAPPED — eighteen on the stems gurn*, gourn*, gorn*, gern* and girn*, and thirteen on the high-noise garn* stem — across Norfolk, Essex, Northamptonshire, Oxfordshire, Cambridgeshire, Middlesex, Westminster and Berkshire, reporting 13,764 raw results of which 509 first-page rows had been retained. Seventeen of the eighteen lower-noise rows have since been enumerated in full through the provider’s own chronological pagination — 2,724 displayed rows, resolving 2,442 of the archived continuation rows — and none contains a baptism or marriage with an indexed father in the John family and an indexed mother in the Mary or Margaret family. One lower-noise row (Norfolk Baptisms gurn*) and all thirteen garn* rows remain. Twelve of the transcripts read in that continuation are retained at sources/corpus_supplement/findmypast-county-gurney-variant-baptism-transcripts-1606-1643.md. The Berkshire item is separate: FindMyPast’s Berkshire Marriages Index, previously never searched, was run unbounded on lastname=gurn* on 31 July 2026 and returned 88 rows, of which the only seventeenth-century entries are Johannis Gurney × Marie Tow at St Michael, Bray, 18 June 1616 (GBPRS/M/251030761/1, and a duplicate rendering as Gurnet Johnnis) and John Gurney at Winkfield in 1643 (GBPRS/M/251341139/1) — no Gurney-variant marriage indexed in the set between 1617 and 1642. This is a set-specific result, not a county-wide negative. Source IDs: findmypast-county-baptism-sets-2026-07-sweep; findmypast-berkshire-marriages-index. ↩︎

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

  11. Parish register, Great Berkhamsted, Hertfordshire, burials, page 165 of 175 in the Findmypast Hertfordshire Burials image series (record GBPRS/HERT/007567815/00277, against transcript GBPRS/D/72723126). The unmodified master, 3,716 × 3,780 pixels, SHA-256 21beb9013d738bef7980f8b1a1da3af0e109213ccbcbfadf53bc7db80cfd46e5, was retrieved 1 August 2026 and is held local-only at sources/media/findmypast-hertfordshire-burials/_local/, the register carrying the notice “© Hertfordshire Archives & Local Studies. Not to be reproduced without permission.” Under the heading 1638 the entry reads “Richard the sonne of Thomas Gurney was buried the xxviij of June”; on the left leaf, “Joane gurney a wido was buried the vij of November.” The opening spans September 1637 to mid-July 1638, so the William Gurney burial of 1 October 1638 falls on the next page and remains unread. Diplomatic transcription and crop manifest at sources/corpus_supplement/findmypast-hertfordshire-berkhamsted-burial-page-165-1638.md. Source ID: findmypast-hertfordshire-burials. ↩︎ ↩︎

  12. 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 opened. The same underlying event, through the FamilySearch index of it, is what the March 2026 working case file read as an adult John Gurney’s death, giving the relative field as “John Gurny, Marg.” and glossing it “most plausibly the deceased’s parents” while still eliminating the row on “died in England.” Fuller treatment, with the household’s remaining unknowns, 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. ↩︎ ↩︎

  13. 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 — Susan Gurney baptised 2 March 1560 to a father William, then a gap of a century and a half to Mary Gurner in 1713, a Richard Gurnell marriage of 1733, an Eliz. Gurney burial of 1719, and the eighteenth- and nineteenth-century Samuel Gurney banking household — with no seventeenth-century entry of any kind. Coverage controlled twice 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 in this interface ranks rather than filters, so these are ranked reads of a covered parish rather than enumerations, and one derivative index disagreeing with another is not proof either way. Fuller treatment at research/topics/john-gurney-candidate-others.md — cross-reference, not a source. Source ID: ancestry-norfolk-1535-1812. ↩︎ ↩︎

  14. Identity — Candidate D: London Draper, Old Change. Cross-reference, not a source. ↩︎

  15. FindMyPast Essex Baptisms, authenticated session, 29 July 2026, lastname=gurn*&keywords=epping, no year bound: 28 results double-indexed as Epping, All Saints and Epping Upland, All Saints, giving 14 distinct events, of which 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) and Elisabeth Gurnet 1659 (father John, mother Anne). No Gurn- baptism earlier than 1616 appears for either parish, and whether that is absence or absent coverage cannot be settled on this instrument. Source ID: findmypast-county-baptism-sets-2026-07-sweep. ↩︎

  16. Origin — trade and training, G13-RI-000075, and Origin — traditions, G13-RI-000100, which reads the 1615-birth tradition backward from the 1636 court order’s “aged 24” clause. Cited here cross-unit. ↩︎ ↩︎

  17. FindMyPast Buckinghamshire Burial Index, authenticated session, 29 July 2026: Weston Turville is covered, with Gurney burials at 1621, 1625 and twice in 1633, and the 1627 father is not among them — so this parish’s negative, unlike Hitcham’s, is genuine. The Gurney tailor Thomas who died in the parish in 1633 is in the published Buckinghamshire probate index. The household reconstruction that splits this row is at 31-candidate-a-aylesbury.md — cross-reference, not a source. Source IDs: findmypast-bucks-burial-index; england-wales-published-probate-indexes; fs-england-buckinghamshire-church-records-1217-1994. ↩︎

  18. FindMyPast Buckinghamshire Baptism Index, 29 July 2026: lastname=gurn*&keywords=hitcham returns one row for the parish in the whole index — Mary Gurny, 1631, father John, mother blank. The parish has no seventeenth-century burial coverage in the burial index at all: the Smith control returns a single burial dated 1827, so “no burial, no probate” is hollow there. The alternative parentage is John Gurney × Susan, Farnham Royal, 30 March 1629, a couple with no child recorded anywhere before 1770. Source IDs: findmypast-bucks-baptism-index; findmypast-bucks-burial-index; findmypast-bucks-marriage-index; findmypast-uk-parish-baptisms. ↩︎

  19. Both Buckinghamshire image targets are bishop’s transcripts — the abbreviated fair copies a parish sent the diocese each year — and both sit in the same series at the Centre for Buckinghamshire Studies: D/A/T/106 for Mary Gurny’s baptism at Hitcham, 22 January 1631, and D/A/T/160 for John Gurney’s burial at Stewkley, 21 September 1615. A transcript is systematically thinner than the register it copies: the mother’s name at a baptism and qualifiers such as “elder,” “younger” or “son of” at a burial are the first material to go. So the blank mother at Hitcham and the missing age or relationship at Stewkley are as likely to be artefacts of the copying as facts about the original. The parish registers are separate, richer documents surviving at the same archive. The digital routes are both closed: FamilySearch’s indexed Buckinghamshire collection is FindMyPast third-party data with images withheld — confirmed at the Hitcham record itself, which displays its D/A/T/106 reference and the Centre for Buckinghamshire Studies as affiliate but serves no image — and the film route resolves Stewkley’s transcripts to catalogue 760791, film 1999454, image group 4010724 with stated coverage including 1607–1618, for which Explore Images returns zero results in an authenticated session. Checked 2 August 2026. Source IDs: findmypast-bucks-baptism-index; findmypast-bucks-burial-index; fs-england-buckinghamshire-church-records-1217-1994. ↩︎ ↩︎ ↩︎

  20. FindMyPast Buckinghamshire Marriage Index, 29 July 2026, tested from both sides. Every John Gurney marriage in the county 1572–1704: Alice 1572, Elizabeth 1573, Alice 1628, Susan 1629, Elizabeth 1630, Avis 1638, Rebecca 1640, Mary 1663, Susannah 1691, Anne 1699/1700, Elizabeth 1704 — five in the emigration window, not one bride a Mary. From the bride side, the earliest Mary to marry a John Gurney in the county is Mary Kidgell at Edlesborough in 1663, two years after the colonial Mary died at Braintree. Three independent instruments agree that the colonial marriage is not in the Bucks/Herts belt. It tightens rather than eliminates, since grooms married in the bride’s parish. Source ID: findmypast-bucks-marriage-index. ↩︎

  21. FindMyPast Bedfordshire Baptisms, transcript R_951758452, read 31 July 2026: John Gurney, baptised 26 September 1606, Eaton Bray, Bedfordshire, father John Gurney, mother not indexed. Worked with its parish context at 31-candidate-a-aylesbury.md — cross-reference, not a source. Source ID: findmypast-bedfordshire-baptisms. ↩︎

  22. FindMyPast Buckinghamshire Burial Index, 29 July 2026: John Gurney buried at Stewkley 21 September 1615, bishop’s transcript D/A/T/160, with no age and no relationship recorded, inside a cluster with Elizabeth 1614, Martha 1615 and Robert 1618; the parish is demonstrably covered. He is either the 1603 boy at twelve or his father, and the register image resolves it. Source ID: findmypast-bucks-burial-index. ↩︎

  23. FindMyPast Westminster Baptisms, 29 July 2026: datasetname=westminster+baptisms&sid=103&lastname=gurley → 12 rows read with the parent columns, of which five are one household at St Clement Danes, all to father John and mother Mary — Eliz 1627, Ellynor 1628, Alexander 1629, Mary 1631, Sara 1633 — with a separate Alce Gurley 1634 at St Margaret Westminster to a father William. A national run for 1620–1640 returns 51 rows showing the same cluster plus unrelated Gurleys in Northumberland and Yorkshire. Gurley requires an n→l confusion, which is not one of the well-attested substitutions, and unlike Gurnard it has no documented alias attaching it to a Gurney; five consecutive entries across seven years all read Gurley, which suggests a clerk who knew the name. Source IDs: findmypast-county-baptism-sets-2026-07-sweep; findmypast-uk-parish-baptisms. ↩︎

  24. Worked at 31-candidate-a-aylesbury.md. Cross-reference, not a source. ↩︎

  25. Identity — other eliminations. Cross-reference, not a source. ↩︎

  26. The National Archives, Kew, E 115/180/113, certificate of residence recording John Gurney’s 1641 move from the Aylesbury half-hundred into Northamptonshire (Source ID: tna-e115-180-113-certificate-of-residence-1641); FamilySearch, “England, Births and Christenings, 1538–1975,” the Saint Mary Aylesbury family group of five children baptised 1638–1653 (Source ID: fs-england-births-christenings). Worked at 31-candidate-a-aylesbury.md. ↩︎

  27. Research state — open questions, G13-RI-000056 — the Braintree-line kit 576097 (R1b-DF19 › R-Z27053 › R-FTD83678) against the untested English-line kit 365744 in the FamilyTreeDNA Gurney project (Source ID: ftdna-gurney-ydna), cited here cross-unit. The downgrade rests on three compounding limits rather than on any new datum: the project’s small number of datapoints for this line, the inference limits of small samples, and thirteen generations of compounding non-paternity risk between the Braintree emigrant and the living testers. Reviewed and accepted 2 August 2026. ↩︎

  28. Ancestry carries two Essex Record Office-derived collections, Essex, England, Church of England Baptisms, Marriages and Burials, 1538–1812 and Essex, England, Select Church of England Parish Registers, 1518–1960, searched 2 August 2026. Both hold Richard Gurnat, 1616, father John Gurnat, and both are index-only — rows render without the View affordance Ancestry’s image-bearing collections carry. The Select collection additionally shows a John Gurnat death or burial row the research layer does not hold, which would date the Epping household; it was not opened. Validation note: sources/validations/ancestry-essex-collections.md. Source IDs: ancestry-essex-1538-1812; ancestry-essex-select-parish-registers-1518-1960. ↩︎

  29. Mark Merry, London Hearth Tax project database (Centre for Metropolitan History, 2010), published with London Hearth Tax: City of London and Middlesex, 1666 and London Hearth Tax: Westminster 1664 at British History Online — eleven strict Gurney/Gurny household entries, 1662–1666, across the City, Westminster and inner Middlesex. Named here: Walter Gurney, St Margaret’s Westminster, Greene Dragon Court, 1 hearth, 1664; William Gurny, St Bride Fleet Street, south side of Fleet Street, 7 hearths, 1666, returned “empty”; William Gurney, St Dunstan in the West, Cock and Key Alley, 3 hearths, 1666; Ann Gurney, St Dunstan in the West, Two Crane Court, 5 hearths, 1666. Underlying manuscripts in The National Archives, E 179 series. Extract: sources/corpus_supplement/london-hearth-tax-1662-1666-gurney-cluster.md. Source ID: bho-london-hearth-tax-merry-2010. The 1662 entry for a John Gurney of St Augustine Watling Street is Candidate D’s: see Identity — Candidate D, G13-RI-000122, cited here cross-unit. ↩︎

  30. The child-set negative is the England-wide two-parent sweep worked at 60-refactor-matching-by-family-not-by-man.md. The mother-axis and marriage-arm results are 73-refactor-the-funnel-closed-and-the-unassembled-row.md §§“The marriage arm” and “The mother axis”: FreeREG’s 178 marriage rows deduplicate to 97 Gurney-cluster events across England 1615–1645, every John-cluster groom named with his bride and not one a Mary, against nine Gurney men who did marry Marys in the same holdings; and the parent instrument run England-wide for 1620–1640 with both parents named returns Eythorne alone. The corrected form of the both-parents claim, admitting the Gurley household at the outer edge of the variant scale, is 75-refactor-margaret-county-sets-and-loose-ends.md §4. Each of these negatives is bounded by its instrument’s coverage, which is stated in the units cited. Source IDs: freebmd-freereg; findmypast-uk-parish-baptisms; ancestry-england-select-marriages-9852; findmypast-bucks-marriage-index; findmypast-hertfordshire-banns-marriages. ↩︎

  31. The result being qualified is the July 2026 mother-axis sweep: FreeREG’s 585 captured rows re-cut for Margaret-family mothers, 705 rows from eight FindMyPast county baptism sets, and the national cross-collection parent-baptism search bounded 1610–1650, each returning no Gurney household with a father in the John family and a mother in the Margaret family. Every one of those instruments indexes baptisms; none indexes burials or marriages. Worked at 75-refactor-margaret-county-sets-and-loose-ends.md §§1–2 and 38-the-shape-and-the-two-families.md, correction 1 — cross-references, not sources. Source IDs: freebmd-freereg; findmypast-county-baptism-sets-2026-07-sweep; findmypast-uk-parish-baptisms. ↩︎

  32. 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. ↩︎

  33. FamilySearch record search, controlled 2 August 2026, all queries f.recordCountry=England: q.surname=Gurn* returns 90,074; adding q.fatherGivenName=John returns 73,356; adding q.motherGivenName=Margaret returns 71,965; and adding the deliberate nonsense q.motherGivenName=Xyzzyqq returns 71,523 with the same top hit as the unfiltered query. The parent-name fields rank; they never filter, so no count from a FamilySearch parent-name query means anything and no parent-axis negative may be built on the instrument. This extends the already-recorded rank-not-filter behaviour of its date, place and spouse parameters. Separately, a wildcard in the mother field fails closed: q.motherGivenName=Marg* returns a flat zero while Margaret returns 71,965 — a silent zero-maker of the class already catalogued on FindMyPast. Source ID: fs-england-births-christenings. ↩︎ ↩︎

  34. Brampton Gurdon to John Winthrop, 20 October 1627, Winthrop Papers, vol. I (Boston: Massachusetts Historical Society, 1929), p. 363 — Gurdon writes on behalf of the widowed “mrs. Gurny who is gardyon to her sonn,” presenting through Winthrop and Emmanuel Downing at the Court of Wards. Source ID: mhs-winthrop-papers-gurdon-to-winthrop-1627. ↩︎

  35. Origin — wardship and patronage network, G13-RI-000083 (the senior West Barsham branch inside the founders’ Court of Wards circle) and G13-RI-000084 (why that is a weak lever for John’s own origin), cited here cross-unit. Francis Gurney’s own place in that family is Identity — Candidate B, G13-RI-000146. The 1627 identification is corroborated from a second direction by Edward Gournay’s own will, which asks to be buried near Martha Gournay at West Barsham — his mother, the widow of the 1615 Norris memorandum and the “Mrs Gurny” of the Gurdon letter. Source ID: tna-pcc-west-barsham-gournay-wills. ↩︎

  36. The Clergy of the Church of England Database (theclergydatabase.org.uk) indexes diocesan institution and subscription books, including Norwich. The unidentified clergyman “Warford,” presented in 1627 and absent from both Venn’s Alumni Cantabrigienses and Foster’s Alumni Oxonienses, is the open question the same search would answer: Origin — wardship and patronage network, G13-RI-000085, cited here cross-unit. The Gurney family’s advowson of the West Barsham vicarage is recorded in Blomefield’s History of Norfolk (Source ID: blomefield-norfolk), which is what makes a wardship-era presentation traceable to them. ↩︎

  37. Robert Charles Anderson, The Great Migration Directory: Immigrants to New England, 1620–1640: A Concise Compendium, 2nd ed. (Boston: New England Historic Genealogical Society, 2025), p. 158: settlement “Boston, Braintree”; origin “Unknown.” Validation note: sources/validations/anderson-gmd-2015.md. Source ID: anderson-gmd-2015. ↩︎