The D'Arcy Archive
Hannah's baptism · found 13 September 2026

One vowel

The document this archive twice declared missing was indexed the whole time under a spelling its own alias list did not carry.

HANNAH SANIGOR, baptised 18 AUGUST 1804, BERKELEY, GLOUCESTERSHIRE — father JNO SANIGOR.

England Births & Baptisms 1538-1975 (FamilySearch index) — FindMyPast serves the same index, so the two sites are one reading documented

This archive has spent weeks looking for Hannah Saniger's baptism, and on two separate days concluded it was missing from every index money could buy. It is not missing. It is spelled SANIGOR, and that is one of the very few spellings this site's own alias list did not carry.

The search index here knows eleven forms of this surname — Sainger, Sanigar, Sanigaer, Sinager, Sinneger, Sinnegar, Sinegar, Synegar, Synager, Singer. It did not know Sanigor. Neither did the variant matching on the subscription site, which returned her marriage and nothing else however the query was phrased. One vowel.

And with it, the whole family

John Saniger's children at Berkeley

John Saniger's children, all baptised at Berkeley, all naming him as father:

Ann Sanigar1798
William Saniger6 September 1801
Jno Saniger18 September 1803
HANNAH SANIGOR18 AUGUST 1804 the direct line
Thos Saniger3 July 1808 almost certainly the Thomas Sinegar who lived a mile from her in Bristol

It also closes the Thomas Sinegar question. He and Hannah are both baptised at Berkeley to John Saniger — she in 1804, he on 3 July 1808 — and he spent his life in the Bristol parish next to hers, naming a daughter Lydia, a daughter Hannah, and two sons John. Thomas →

The honest caveat

One thing does not fit, and it is the year.

The family tree gives Hannah's birth as 23 July 1805. This baptism is 18 August 1804 — eleven months earlier. There is no second Hannah baptised to John Saniger at Berkeley, and no burial for a Hannah between the two dates, so the ordinary reading is that the tree's day and month were carried from somewhere reliable and the year slipped by one. The alternative — two daughters of the same name, the first dying unrecorded — is possible and has no evidence at all.

her baptism, very probably; the year unresolved

And a name that is not yet evidence

Her mother is named, and the name is not evidenced.

FamilySearch's shared tree gives John Saniger a wife LYDIA JONES, 1770–1851, married at Berkeley in 1797. That would name the seventy-year-old Lyddya Sinegar living in Hannah's house in 1841 and settle the oldest small mystery on this site.

It is not sourced. John Saniger's profile there carries fourteen sources and every one of them is a baptism — his own and his children's. No marriage record is attached, and no record anywhere in this archive gives Lydia a surname. JONES is somebody's assertion until a Berkeley marriage register says otherwise. a name to test, not a fact

AND THIS ARCHIVE'S OWN DATA HOLDS FOUR MORE FIELDS THAT THIS PAGE NEVER MENTIONED: Lydia Jones born 13 MARCH 1770 AT WOTTON-UNDER-EDGE, Gloucestershire; parents WILLIAM JONES and SARAH; died 7 JULY 1851 AT BERKELEY. Every one of those is testable and none has been tested. A death at Berkeley in July 1851 is inside civil registration and should sit in the September quarter of the GRO index for the Thornbury district. A baptism at Wotton-under-Edge in March 1770 to William and Sarah Jones is an ordinary register lookup. A week was spent deciding the surname was unevidenced without trying either.How that was missed →

Lyddya Sinegar, aged seventy →
Gloucestershire probate, 1801–1858

And there is no will to ask.

A will names children, and that is the shortest route from John Saniger of Berkeley to his daughter Hannah. The Gloucestershire probate index for the right period — Gloucestershire Wills & Administrations 1801-1858, the Consistory Court of Gloucester, 14,195 entries — was read this fortnight under every opening this surname takes. It holds exactly TWO of them.

1813ABRAHAM SYNEGAR, CARDMAKER, of DURSLEY — administration, 1813/8
1822JOHN SANIGAR, CARPENTER, of OLDBURY-ON-SEVERN, Thornbury — will, 1822/194
Neither is Berkeley. The John Sanigar of 1822 is the Thornbury outlier this archive already knew about, ten miles down the Vale and in the wrong parish; a cardmaker at Dursley makes the combs that card wool, which is the Stroudwater trade the reconstitution found. THERE IS NO SANIGER OF BERKELEY IN THE GLOUCESTERSHIRE PROBATE INDEX AT ALL, which means John Saniger left no proved will and no administration in the diocesan court — the ordinary outcome for a man with nothing much to leave, and a closed door rather than an unopened one.searched, empty, and the coverage established

Searched with wildcards rather than spellings — SAN*, SIN*, SYN* and SWA* return every form the name takes and 27 surrounding families besides, so no spelling can hide inside this result the way SANIGOR hid a baptism. Control: SMITH in the same index returns 277.

Settled · 20 September 2026

The 1804 entry, and what was in the column nobody reads

The complete Saniger pass left one entry it could not place: Jno Saniger, Berkeley, 1804, with the father's column empty — the only Berkeley baptism of the 158 without a father named. Three readings were offered: a second son born the same year as Hannah, a mis-indexed copy of the 1803 entry, or a different family. It is none of them.

ChildBaptisedFatherMother
Jno Saniger 18 September 1803 father JNO SANIGER mother not given Hannah's brother
Jno Saniger 11 February 1804 no father mother MARY SANIGER not this family
The two records sit at consecutive index numbers, five months apart, which is what made a duplicate look likely. The transcripts settle it. The 1803 entry names a father and no mother; the 1804 entry names a mother and no father — and the mother is MARY, not Lydia. A father's column left empty beside a named mother is how a parish register records a base-born child, and it is not a clerical slip. documented
And it was found the same way as everything else this week. The result row showed «Father's first name(s): —» and stopped. The mother is on the transcript, one click further in. That is the fourth question this month answered by the field a results page does not show, after a Sawyer, a Cossens and a Gairden.
The atlas · 22 September 2026

And on the map she was twenty kilometres from Berkeley

The spelling that hid her baptism hid her again, in a place nobody was looking. This archive's own atlas had her birthplace recorded as «Berkley, Gloucester, England» — the family tree's idiosyncratic spelling, the one the 1851 census also uses, and the one [/thomas-sinegar](/thomas-sinegar/) has named since 13 September. It never reached the map as Berkeley.

So she sat alone on 51.865 -2.246, a coordinate a gazetteer returns when it cannot resolve «X, Gloucester, England» — the county written as its county town. Two other villages were on the same point, Brimpsfield and Ashton Underhill, which have nothing to do with her or with each other. That is how the pin was caught: not by anyone checking Hannah, but by three unrelated places being in one spot.

She is on Berkeley now, with the rest of them. John Saniger, Lydia, Catherine Cotton, Margaret Richards, William and Thomas Saniger — eleven people on the parish the whole English line comes out of. It changes no finding on this page. It changes where the archive was drawing the woman the findings are about.

documented The map →

Her mother, tested · 16 September 2026

Her mother, tested

The block above was written on 13 September and ended by saying that four fields about Lydia were «testable and none has been tested». All four have now been tested, in one afternoon, on the subscription this archive already had. Two are confirmed from record, one is wrong, and one has nothing behind it.

The claimVerdictGradeWhat was found
Died 7 July 1851 at Berkeley CONFIRMED documented LYDIA SANIGAR, death registered in the September quarter of 1851, Thornbury district — GRO volume 11, page 329. Berkeley is in Thornbury registration district, and a death on 7 July falls in that quarter. She died under her married surname, which is itself evidence that a Saniger marriage happened.
Born about 1770 at Wotton-under-Edge CONFIRMED documented The 1851 census gives her birthplace, in the enumerator's own spelling, as «Wooton under Edge», and her age as 81. Her baptism is not in England Births & Baptisms — no Lydia Jones at all is baptised in Gloucestershire between 1765 and 1775, against 939 Jones baptisms in the county in those eleven years — but the place and the year do not need it. They are on a record she was alive to answer.
Married John Saniger at Berkeley in 1797 WRONG AS STATED disputed There is a Lydia marrying a John at Berkeley in that window, and she is the only one in twenty-one years: LYDIA JONES, 8 January 1798. Her husband's surname, in the transcript rather than in the result row, is SAWYER. Berkeley has 272 marriages indexed for 1793-1803, so the parish is covered, and no Saniger marries there in that decade under any of eight spellings.
Her parents were William Jones and Sarah NOTHING BEHIND IT not claimed It rests entirely on the maiden name, and the maiden name rests entirely on the marriage that turned out to be a Sawyer's. JONES is now disputed rather than untested. Wotton-under-Edge survives; Jones does not.
Where JONES probably came from. A Lydia marrying a John at Berkeley in 1798, one year out from the date the tree gives and in the right parish, is a very close match on everything a result row displays — first name, spouse's first name, place, year. The only field that rules it out is the spouse's surname, and that field is not in the result row. It is one click further in, on the transcript. The likeliest account of «Lydia Jones» is that somebody stopped one click early, twenty years ago, and every tree since has copied it. inferred
What holds. The 1841 census has a seventy-year-old Lyddya Sinegar in Hannah's house in Bristol. The 1851 census has Lydia Sanigar, 81, born Wotton-under-Edge, back at Ham in Berkeley. The GRO registers her death there months later. Lydia is real, she is a Saniger by marriage, her age is right in three records and her birthplace is on one of them. It is her maiden name and her wedding that were invented. documented
Three tests · 20 September 2026

Where Lydia did not marry

JONES was disputed on 16 September, when the marriage the whole internet cites turned out to belong to a Sawyer. That left a question rather than an answer: where did John Saniger marry, and to whom? Three days of the obvious tests have not found it, and the shape of the failure is now worth as much as a find would have been.

The neighbours at Ham, 1851 not found

Her household's transcript identifiers run in sequence, so the street can be walked either side of her door. Samuel and Mary Ann Nelmes and their baby on one side; Stephen Smith, farmer, 57, born Berkeley, his wife Sarah and three servants on the other. No relative of Lydia's is named in either, and there is no Jones in the village.

Her own parish not found

She was born at Wotton-under-Edge, and a bride usually married at home. One Lydia marries a «Jno.» at Wotton-under-Edge in the whole window — Lydia Howell, 2 April 1794. The result row shows the groom's forename and stops there. The transcript gives his surname: COSSENS.

Every spelling, the whole country not found

No SANIGER, SANIGAR, SAINGER, SANIGOR, SANEGAR, SINEGAR, SINNEGAR or SANNIGER marries anywhere in England between 1790 and 1802 in England Marriages 1538-1973. The last before the gap is John Saniger to Elizabeth at Thornbury in 1790; the next after it is William Saniger to Mary at Thornbury in 1803.

And the controls hold, which is what makes this worth publishing. A null about a parish is worthless unless the parish is in the book. Berkeley has 272 marriages indexed for 1793-1803. Wotton-under-Edge has ten full pages of them for 1792-1802, and not one carries any spelling of this surname. Both of the two parishes where this marriage should be are well covered, and it is in neither. documented
What that leaves. Four readings, and this archive cannot yet choose between them. They married somewhere else — a licence could take them to any parish in the diocese. They married in a register this transcription does not include, which is an ordinary gap and not a mystery. The entry exists but is indexed under a spelling not yet tried — nine have been, and the surname has run to at least fifteen. Or they did not marry at all, which the archive raises because a woman keeping the surname and dying under it does not by itself prove a wedding. not claimed
And one of the four readings is now closed. The third reading above — that the entry exists under a spelling not yet tried — was the live one, because this archive had just found two spellings it did not know and had been caught out by both. It can be retired. On 21 September the question was asked without any spelling at all.

The surname box takes wildcards. `SNG*` across Gloucestershire for 1793-1803 returns 122 rows, and all 122 were read. Six of them are really this surname — Mary Sannigar at Stroud in 1794, James Senigar at Bristol in 1795, Daniel Sinigar at Gloucester in 1800, Sarah Sinigor at Stinchcombe in 1801, William Saniger at Thornbury in 1803. Not one is a John. Not one marries a Lydia.

Then the same question with no county at all: a John of any `SNG*` surname marrying a Lydia anywhere in England between 1780 and 1820 returns sixteen rows, and the only one whose surname is even arguable is John Sanger and Lydia at Salisbury in 1788 — a Wiltshire form this archive excludes by name, sixty miles off, ten years before their first known child.

The wildcard proves itself on the way past: those 122 rows contained SINIGOR and SINNIGER, two forms missing from this archive's spelling table, which had been rebuilt and called complete that morning. So a search that cannot miss a spelling has now been run, and it does not find them. Three readings remain, not four: they married in a parish this transcription does not include, they married somewhere no one has thought to look, or they did not marry. Further searching of this index cannot decide between them. The Berkeley register itself can, which is what the letter to Gloucestershire Archives is for. documented

What would settle it. The Berkeley marriage register itself, 1795 to 1799, read by eye rather than through an index — the same answer the RCB Library gave about Wicklow, and for the same reason. Failing that, the Gloucester diocesan marriage licence bonds and allegations, which survive separately from the registers and would name both parties, their parishes and often her father.
Two censuses · 16 September 2026

Clapton, in the village of Ham

Following Lydia into the census does something the baptism registers could not: it puts this family at an address. Both censuses place them in Ham, one of the three hamlets of Berkeley parish that the 1608 muster names — and the manor whose court roll of 1587 sits in John Smyth's own collection, MS 3549, as «Ham and Hinton». The family this archive has been tracing through spellings turns out to have been living, in living memory, in the place the archive has been reading fourteenth-century deeds about.

1841 — Clapton, Ham, Berkeley

NameAgeRelationship
John Sanigar35 b. Gloucestershire
Anne Sanigar40 b. Gloucestershire
William Sanigar6 twinb. Gloucestershire
John Sanigar6 twinb. Gloucestershire
Mary Sanigar3 b. Gloucestershire

HO107/347, book 6, folio 3, page 1, schedule 761. Ages of adults are rounded down to the nearest five in 1841, so «35» covers a man of 37.

1851 — Village of Ham, Ham

NameAgeRelationship
Ann Sanigar50 Head, widow«Formly tailors wife» · b. Gloucestershire
William Sanigar16 SonGeneral labourer · b. Berkeley
Mary Ann Sanigar13 DaughterScholar · b. Berkeley
Lydia Sanigar81 MOTHER IN LAW, widow«Formly labourers wife» · b. «Wooton under Edge»

HO107/1957, folio 561, page 13, schedule 46, Berkeley sub-district.

The twins, and why the index looked wrong. The complete baptism sweep found two Berkeley baptisms in 1835 to a father John — a John and a William — and read as index entries alone that looks like the duplication this set is full of. It is not. The 1841 census has William 6 and John 6 in the same household. They are twins, and the two baptisms are two children. documented
Who the head of that house was. Lydia is mother-in-law to the widowed Ann of 1851, so Ann's late husband was Lydia's son. The children William and Mary Ann were both baptised at Berkeley naming a father JOHN — William and John in 1835, Mary Ann in 1838. A John Sanigar's death is registered in the Thornbury district in the first quarter of 1850, which is why Ann is a widow in the March of 1851. The John Saniger baptised at Berkeley on 18 September 1803, Hannah's brother, was 46 in 1850. Everything fits and nothing yet proves it: the identification rests on a first name, a parish and an age, and the death certificate has not been ordered. inferred
And what it does to the medieval work. This archive has spent a fortnight on Swonhungre, the fourteenth-century form of the name, and on the hamlet of Saniger in Berkeley that the surname came from. It had no evidence that the modern family was still anywhere near it. Now it has two censuses putting them at Clapton in Ham, which is the next hamlet, in the same parish, in the same manor. That is proximity, not descent — no document joins the 1841 household to the 1390 wardship deed, and this archive has already invented one family out of thin air this month. But it is the first time the two halves of the work have been in the same field.
Tested · 16 September 2026

Her mother, tested

The block above was written on 13 September and ended by saying that four fields about Lydia were «testable and none has been tested». All four have now been tested, in one afternoon, on the subscription this archive already had. Two are confirmed from record, one is wrong, and one has nothing behind it.

Died 7 July 1851 at Berkeley
CONFIRMED documented

LYDIA SANIGAR, death registered in the September quarter of 1851, Thornbury district — GRO volume 11, page 329. Berkeley is in Thornbury registration district, and a death on 7 July falls in that quarter. She died under her married surname, which is itself evidence that a Saniger marriage happened.

Born about 1770 at Wotton-under-Edge
CONFIRMED documented

The 1851 census gives her birthplace, in the enumerator's own spelling, as «Wooton under Edge», and her age as 81. Her baptism is not in England Births & Baptisms — no Lydia Jones at all is baptised in Gloucestershire between 1765 and 1775, against 939 Jones baptisms in the county in those eleven years — but the place and the year do not need it. They are on a record she was alive to answer.

Married John Saniger at Berkeley in 1797
WRONG AS STATED disputed

There is a Lydia marrying a John at Berkeley in that window, and she is the only one in twenty-one years: LYDIA JONES, 8 January 1798. Her husband's surname, in the transcript rather than in the result row, is SAWYER. Berkeley has 272 marriages indexed for 1793-1803, so the parish is covered, and no Saniger marries there in that decade under any of eight spellings.

Her parents were William Jones and Sarah
NOTHING BEHIND IT not claimed

It rests entirely on the maiden name, and the maiden name rests entirely on the marriage that turned out to be a Sawyer's. JONES is now disputed rather than untested. Wotton-under-Edge survives; Jones does not.

Where JONES probably came from. A Lydia marrying a John at Berkeley in 1798, one year out from the date the tree gives and in the right parish, is a very close match on everything a result row displays — first name, spouse's first name, place, year. The only field that rules it out is the spouse's surname, and that field is not in the result row. It is one click further in, on the transcript. The likeliest account of «Lydia Jones» is that somebody stopped one click early, twenty years ago, and every tree since has copied it. inferred
What holds. The 1841 census has a seventy-year-old Lyddya Sinegar in Hannah's house in Bristol. The 1851 census has Lydia Sanigar, 81, born Wotton-under-Edge, back at Ham in Berkeley. The GRO registers her death there months later. Lydia is real, she is a Saniger by marriage, her age is right in three records and her birthplace is on one of them. It is her maiden name and her wedding that were invented. documented
FindMyPast · found 16 September 2026

All one hundred and fifty-eight, read

FindMyPast's England Births & Baptisms 1538-1975 carries a column this archive has never had from an index: the father's first name. For SANIGER it holds 158 baptisms, and 156 of them name a father. All eight pages have now been read in baptism-year order, 1571 to 1909. 115 of the 158 are Berkeley — the rest are Bristol (9), Thornbury (7), Stone and Tidenham (4 each), Dursley and St Botolph Aldgate (3 each), King's Stanley and Stonehouse (2 each), and one apiece at Frampton on Severn, Slimbridge, North Nibley, Wheatenhurst and Newent. Two strays sit far outside all of it, both in 1688: Nether Whitacre in Warwickshire and Stoke Prior in Worcestershire.

ChildBaptisedFatherParish
Ann Saniger1798 JohnBerkeley already on this page — see the correction below
William Saniger1801 JohnBerkeley
Jno Saniger1803 JnoBerkeley
Jno Saniger1804 none indexedBerkeley unexplained
Thos Saniger1808 JohnBerkeley
Hannah Saniger1746 ThomasBerkeley two generations early
Hannah Saniger1772 WilliamBerkeley entered twice; still early
John Saniger1835 JohnBerkeley a second John's children
William Saniger1835 JohnBerkeley
What the pass gives. The four SANIGER children of John Saniger of Berkeley — Ann 1798, William 1801, John 1803, Thomas 1808 — now stand with their father named in the index, not asserted by a tree. With Hannah Sanigor, 18 August 1804, that is the complete set of five, and every one of them names JNO or JOHN. This is the deepest line in the archive and until this week the only thing holding the sibling group together was other people's trees. documented
A correction, one day old. Yesterday this section said “Ann has never had a baptism” and presented her 1798 entry as a new find. She had been on this page since 13 September, four blocks above, in the sibling table, as Ann Sanigar 1798. I read one page of an index, matched a name against what the tree asserts, and did not read the page I was writing on. It is the same entry under a different vowel. Recorded as own error 33.
The null under SANIGER is now a real null. There is no Hannah in the set between 1777 and 1817 under this spelling. The only two are Berkeley 1746, father Thomas, and Berkeley 1772, father William — the second entered twice. That is no longer a guess about an index: every one of the 158 has been read. She is there under SANIGOR, which is why she was found on 13 September and not before, and this pass is the coverage that makes the SANIGER absence worth stating.
And what the set is not. Thirty-three of the 158 are duplicates — 125 distinct name-year-place combinations, so one entry in five is the same child indexed twice, sometimes four times: Mary, daughter of Samuell, appears four times in 1688, and Sarah, daughter of William, four times at Bristol in 1839. The Berkeley run also has holes, and they are the holes of a register rather than of a family: 1632 to 1676, forty-four years with nothing, then 1749–1769, 1777–1786, 1808–1817 and 1826–1835. Counting this surname's presence by decade from this set alone would be counting the survival of paper.
It also closes something this archive got wrong. Own error 32 was a Forest-of-Dean branch of this family that I invented and then built on. Here is a third independent test of it: 158 baptisms across 268 years contain no Awre, no Blakeney, and nothing anywhere west of the Severn. Bigland said so, Rudder said so, the 1608 muster said so, and now the parish registers say so.

Two notes on method

  1. An alias list is a research instrument, and mine had a hole in it. Every "not found" this archive published for Hannah was really "not found under ten of the eleven spellings I happened to think of". The list now carries sixteen, including Sanigor, Sanigear, Sinigar and Swanhanger. Where the other holes were →
  2. And on Ancestry. Asked to look on Ancestry, which holds the Gloucestershire registers FindMyPast does not. It is not signed in on this machine and I will not enter anyone's password, so it was not used. FamilySearch was signed in, is free, and had the record.

Six generations at Berkeley → · The surname in 1608 → · Why I thought she was not from there → · The errors →