556 records · 81 families · searched 13 September 2026

The Sanigers, reconstituted

Every Saniger record in Gloucestershire, read individually and assembled into households — to answer one question: is Hannah in any of them? She is not.

Read with a correction, 13 September 2026. This reconstitution excluded Hannah Saniger from all 81 Gloucestershire households because the 1841 census ticked her as born outside the county. The 1851 census names her birthplace: Berkeley, Gloucestershire. The counts and the families below are sound; the conclusion drawn about Hannah was not. Where she actually came from →

This archive's oldest claim is a Gloucestershire family. Its only link to that family is one line in a Queensland death register naming Miriam Wakefield's mother as Hannah Saniger. Searching for Hannah by name had failed. Repeatedly →

So instead of searching for her, the county was reconstituted: all 556 Saniger records opened one by one, the father and mother read off each baptism, and the children assembled back into the households they came from. 81 families. Then ask which of them could be hers.

None of them. Not one family in Gloucestershire has a mother called Lydia — the name the tree gives Hannah's mother. Not one is a John and Lydia. There is exactly one Hannah baptised in the whole dataset, at Stinchcombe in 1730, and only three families are having children at all between 1795 and 1815, when Hannah should have been born. searched and empty, comprehensively

Method

What was actually done

Records opened	556	205 baptisms, 200 marriages, 151 burials
Baptisms naming a father	196	of 205 — the rest give the child only
Households assembled	81	38 with two or more children; 43 with one

Grouped by parish, father's forename and mother's forename, and split wherever more than twenty-five years separates one baptism from the next — otherwise a grandfather and a grandson of the same name merge into one impossible household spanning a century. Duplicate transcriptions of the same event, which are everywhere in this county because several parishes are transcribed twice, are counted once.

The answer to the question

Three tests, three negatives

Asked of all 81 households	Found
A mother called Lydia the tree names Hannah's mother Lydia Jones	None. Not one household in the county empty
A father John with a wife Lydia the tree names her father John Saniger	None. empty
A Hannah baptised anywhere	One, at Stinchcombe on 2 July 1730 — daughter of Benjamin and Sarah, seventy-five years too early not hers
Any household bearing children 1795–1815	Three. Daniel and Elizabeth at Cam, George and Harriet at Dursley, William and Mary at Falfield — no John, no Lydia none of them hers

And this is where it stops being a puzzle and starts being an answer. The 1841 census entered Hannah Wakefield as born Out of County — not born in Gloucestershire. She is not in these eighty-one households because she is not from here. A week of failed searches was not bad luck; it was looking in the wrong county. The census that says so →

Eight counties, and a simpler explanation

Where else she is not

Gloucestershire was the obvious county and it is now exhausted. So were the others within reach of Bristol: Somerset, Wiltshire, Monmouthshire, Herefordshire, Worcestershire, Shropshire and Oxfordshire. No Hannah Saniger born about 1805 in any of them — and the searches were control-tested, with more than five hundred Smiths returned for the same twenty years in each county group, and 358 Saniger-variant records in Wiltshire alone. eight counties, searched and empty

Which points at the least exotic answer available. Bristol straddled a county boundary. The parishes north of the Avon were in Gloucestershire; those south of it — Bedminster, Temple, St Mary Redcliffe — were in Somerset. An enumerator in St Paul's, returning his book under Gloucestershire, would enter a woman born half a mile away across the river as “Out of County” and be perfectly correct.

Hannah may simply have been born in Bristol, on the wrong side of the river — and that is precisely the ground where FreeREG holds four registers and 1,551 records. a hypothesis, and the cheapest one What would test it →

Who they actually were

A cloth-trade family of the Stroud valleys

Since the records were all opened anyway, they describe a family this archive had never looked at properly. The occupations are unmistakable:

Trades recorded

Blacksmith	7
Weaver	6
Labourer	5
Shoemaker	5
Plaisterer	5
Coachman	4
Yeoman	3
Farmer	3
Currier	3
Clerk in Holy Orders	2
Tucker	2

Where they lived

Cam	32
Dursley	14
Lower Cam	8
Painswick	6
Berkeley	5
Woodmancote	5
Hallmer	4
Oldbury	4
Piddington	3
Rectory Cottage	3

Weavers, tuckers, curriers, blacksmiths and shoemakers, living at Cam, Lower Cam and Dursley. A tucker finishes woollen cloth; a currier dresses leather. These are the trades of the Stroudwater cloth district in its working century, and they are what this family did. Two Clerks in Holy Orders and three yeomen are the whole of the gentility.

The households

Every family of three children or more

Parish	Father and mother	Children	Span	Baptised
Dursley	Abraham & Jane	10	1698–1721	John 1698, William 1701, Susannah 1704, Jane 1707, Thomas 1711, Thomas 1712, Jane 1715, Tho. 1720, Thomas 1721
Cam	Daniel & Elizabeth	8	1729–1763	Daniel 1729, Mary 1733, Mary 1735, Mary 1755, Henry 1757, Sarah 1760, Sarah 1762, Sarah 1763
Cam	Daniell & Betty	8	1741–1752	William 1741, William 1742, Solomon 1745, Solomon 1746, Solomon 1747, Betty 1749, Ann 1751, Ann 1752
Cam	Thomas & Sarah	8	1770–1791	Betty 1770, Thomas 1773, Daniel 1775, Mary 1778, Deborah 1780, Thomas 1782, Deborah 1788, George 1791
Dursley	William & Elizabeth	6	1731–1761	Margaret 1731, Ann 1732, William 1733, Samuell 1752, Mary 1757, Elizabeth 1761
Cam	John & Prudence	5	1677–1682	Richard 1677, John 1677, Daniel 1679, John 1680, Mary 1682
Dursley	William & Mary	5	1704–1710	Mary 1704, William 1705, William 1707, John 1708, Martha 1710
Cam	Daniel & Elizabeth	5	1813–1826	John 1813, Daniel 1815, Elizabeth 1820, Eliza 1823, Caroline 1826
Painswick	Thomas & Eleanor	5	1844–1849	Joseph Thomas 1844, Helen 1846, John 1846, Jane 1848, Emily 1849
Dursley	William & Ann	5	1846–1848	George 1846, Charles 1846, Henry 1846, Prudence 1846, William Sydney 1848
Dursley	James & Emma	5	1887–1895	Ethel Louisa 1887, Francis Henry 1888, Charles James 1890, John Sidney 1894, Arthur George 1895
Thornbury	Thomas & Jane	4	1750–1760	William 1750, John 1755, Isaac 1760, Jacob 1760
Berkeley	James & —	3	1671–1675	James 1671, Elizabeth 1673, Edward 1675
Stinchcombe	Benjamin & Sarah	3	1723–1730	Sarah 1723, Mary 1726, Hannah 1730
Thornbury	William & Elizabeth	3	1779–1784	Edward 1779, Jane 1784, Elizabeth 1784

The largest household in the county is Abraham and Jane of Dursley — ten children between 1698 and 1721. Their Thomas, baptised 17 March 1711/12, is the one this archive briefly took for its own ancestor before reading the father's name. How that failed →

Where the families are

Parish	Households
Dursley	33
Cam	20
Thornbury	8
Berkeley	5
Stinchcombe	4
Painswick	2
Falfield	2
Marshfield	2
King's Stanley	1
Tetbury	1
North Nibley	1
Bromsberrow	1
Boxwell	1

Dursley and Cam hold fifty-three of the eighty-one. Berkeley — the parish this archive has named in every account of itself since it was built — holds five, and cannot hold more, because its transcript stops in 1677.

27, not eleven

How this name is actually spelled

The Berkeley page lists eleven spellings. Asked of one record set on one basis — England Births & Baptisms 1538-1975, searched nationally, every form put the same way — this name is spelled 27 ways, of which 20 return anything at all. Together they hold 373 index entries between 1565 and 1909, and the commonest is the one the family uses. Entries are not children: the five largest forms have been counted one by one and about one entry in six is a repeat. The measurement →

This table was recounted on 21 September 2026 and its numbers changed. It previously held nineteen forms counted as entries in the parish registers read for this archive — a narrower source, drawn differently. Those counts were not wrong. They were not comparable, which is why SANIGOR and SYNIGAR could not simply be added to them when they were found. The old table is kept in the data rather than deleted. The sweep this is drawn from →

Spelling	Occurrences
SANIGER	158	the form this archive uses
SANIGAR	56
SAINGER	30
SINEGAR	28
SINNEGAR	26	the commonest in the registers
SYNAGER	10
SANNIGER	10
SANEGER	8
SANEGAR	7
SINIGER	7
SINEGER	6
SINNEGER	5
SYNIGAR	5
SANIGOR	4
SANNEGAR	4
SANIGEAR	3
SINIGAR	2
SINNIGER	2
SINAGER	1
SANNIGAR	1
SINIGOR	0	nought here — but a Stinchcombe marriage of 1801
SANIGRE	0	nought here — and the only form that carries Thomas's 1828 marriage
SANYGER	0	the 1608 muster's form — in no register anywhere
SYNIGER	0	the 1608 muster's form — in no register anywhere
SYNEGAR	0	tried, and returns nothing
SANIGAER	0	tried, and returns nothing
SENIGAR	0	tried, and returns nothing

All 27 are now folded together by this site's own search, so any one of them finds the rest. Try it →

FreeBMD — the GRO index

The same three deaths, in the other index

The table above is the LOCAL register — the registrar's own book, register and entry number. FreeBMD indexes the GRO's quarterly volumes instead, which is a different index of the same events. Searched on 15 September, it carries all three of the Thornbury deaths this archive cares about, and gives the reference a GRO certificate is ordered against.

Name	Quarter	District	GRO index	Local register
JOHN SANIGAR	March quarter 1850	Thornbury	vol 11, page 401	local register 3, entry 146
LYDIA SANIGAR	September quarter 1851	Thornbury	vol 11, page 329	local register 3, entry 265
MARY ANN SANIGAR	December quarter 1851	Thornbury	vol 11, page 369	local register 3, entry 282

Two indexes, built from different books by different people, agree on three deaths in the same district within twenty-one months. The pairing of local entry to GRO volume is an IDENTIFICATION — same name, same district, same year — not a proven match, because neither index names a parent, an age or an address. What it does settle is that Lydia can now be ordered from the GRO as well as from Gloucester.documented

And the control that made the null meaningful. FreeBMD's first three answers here were NOUGHT, and all three were false. A search across 1837–1875 returns nothing at all — the form silently refuses a range that wide and redisplays itself, which looks exactly like an empty result. The control was CHARLOTTE DARCY, whose death this archive already holds from the GRO index: searched for 1869 alone she comes back «Darcy Charlotte 73 Brighton 2b 152», the reference this site already carried. Narrowing the Sanigar search to three years then produced all three deaths at once. lore

Gloucestershire Registration Service Index

Twenty-four deaths, with the register and the entry number

The Gloucestershire Registration Service and the Gloucestershire Family History Society between them have indexed the registrars' own books and put them online free. It is not the GRO index: it gives the LOCAL register and entry number, which is what a certificate is ordered against. documented

Coverage first. Coverage checked before the search, not after. For deaths the districts that matter are complete: Thornbury/Berkeley 1837–1948, Dursley 1837–1937 (20 registers), Wotton-under-Edge 1837–1937 (19 registers), Uley 1837–1906.

Year	Surname as indexed	Forename	Register	Entry
1838	SANIGAR	Sarah	1	149
1845	SANIGER	Susan	2	418
1849	SANIGER	Catharine	3	430
1850	SANIGAR	John	3	146
1850	SANIGER	Joseph	4	33
1851	SANIGAR	Lydia	3	265
1851	SANIGAR	Mary Ann	3	282
1856	SANIGER	Elizabeth	5	285
1857	SANIGER	Eliza	5	394
1857	SANIGER	Samuel	4	257
1859	SANIGER	Mary Ann	4	499
1859	SANIGER	Samuel	4	449
1860	SANIGER	Mary	5	96
1861	SANIGER	Sarah	6	435
1861	SANIGER	Susanna	6	432
1862	SANIGER	Charlotte	6	464
1862	SANIGER	Edward Thomas	6	462
1862	SANIGER	Joseph	6	465
1867	SANIGAR	Keturah	6	131
1872	SANIGER	Charlotte	7	125
1877	SANIGER	Ann	8	174
1880	SANIGER	Frederick James	8	462
1882	SANIGER	Margaret	23	304
1882	SANIGER	William Henry	23	353

And the one that was being looked for

LYDIA SANIGAR, died 1851, register 3, entry 265. This archive has wanted her since it began: she is the Lyddya Sinegar aged seventy in Hannah's house in 1841, and she is the wife whose maiden surname the marriage of about 1797 would give. documented

What an index is and is not. The index gives a year, a register and an entry — not a date, an age, an abode or a maiden name. It locates the certificate; it does not replace it. What it does do is make the certificate orderable against the local register rather than the GRO index. On the queue →

Ten marriages, with the other name

The marriage index gives both partners, which the GRO quarterly index does not do for this period without a second lookup.

Year	Saniger	Married	Reference
1838	Hester SANIGAR	William WHITE	C7/15
1841	Harriet SANIGAR	George FORD	C7/123
1849	Thomas SANIGER	Charlotte FORD	C7/181
1855	Catherine SANIGER	John STEPHENS	190
1858	William SANIGER	Caroline STOCKWELL / WORKMAN	C7/1168
1862	Thomas SANIGER	Sarah Bendall WATTS	C7/1204
1879	Emily Sophia SANIGER	Robert DORE	C3/2320
1880	Rachel Harris SANIGER	James SINNETT	30x No 1430
1891	Albert Thomas SANIGAR	Ellen Kate YOUNG	34 No 1168
1891	Arthur Edward SANIGER	Louisa POWELL	34 No 1154

One oddity, left as found. One entry appears twice under one reference, C7/1168 in 1858, with the bride as both Caroline STOCKWELL and Caroline WORKMAN. One event, two surnames for the same woman — a remarriage, a step-parent's name, or an index error. It is left exactly as the index gives it.

Twenty-seven births

The birth search accepts a MOTHER'S MAIDEN NAME as a search term, which is the field that turns an index into a family. It is not displayed in results, but it can be searched against — so a maiden name guessed from a marriage can be tested against a birth.

Year	Surname	Forename	Reg/Entry
1838	SANIGER	Mary Ann	190
1845	SANIGER	Susan Ann	3/253
1846	SANIGAR	Ann	3/488
1848	SANIGER	Catharine	4/435
1849	SANIGER	Susan	5/110
1850	SANIGER	Joseph	5/169
1851	SANIGAR	Mary Ann	4/495
1851	SANIGER	Emily Sophia	5/340
1852	SANIGER	Fanny	6/62
1854	SANIGER	Female	6/385
1856	SANIGER	Edward Thomas	7/143
1858	SANIGER	Keturah Fanny	7/408
1860	SANIGER	Susanna	8/348
1862	SANIGER	Joseph	9/8
1863	SANIGER	Albert Thomas	9/241
1864	SANIGER	Hester Jane	9/8
1865	SANIGER	Arthur Edward	9/166
1866	SANIGAR	Eva Emma	9/343
1868	SANIGAR	Thomas Browning	10/95
1869	SANIGER	Florence Elizabeth	10/361
1871	SANIGAR	Charlotte	11/45
1872	SANIGER	Francis William	11/256
1873	SANIGAR	Amy Charlotte	12/92
1875	SANIGAR	Edith	12/349
1876	SANIGAR	William Henry	13/122
1878	SANIGAR	Margaret	13/412
1879	SANIGER	Frederick James	14/216

What the three indexes do together

Read across, births and deaths in the same index start closing on each other. Nine names appear in both, at intervals that read as short lives:

Catharine	born 1848	died 1849	about one year
Susanna	born 1860	died 1861	about one year
Edward Thomas	born 1856	died 1862	about six years
Keturah Fanny	born 1858	died 1867 as Keturah	about nine years
Frederick James	born 1879	died 1880	about one year
Margaret	born 1878	died 1882	about four years
William Henry	born 1876	died 1882	about six years

And two who lived. EMILY SOPHIA, born 1851, marries Robert Dore in 1879. ALBERT THOMAS, born 1863, marries Ellen Kate Young in 1891. ARTHUR EDWARD, born 1865, marries Louisa Powell in 1891, a fortnight or so from his brother by the look of the references — 34 No 1154 and 34 No 1168.

And none of it is proved. Every one of these joins is an inference. The index gives a surname, a forename, a year and a register reference, and nothing that proves the Catharine born in 1848 is the Catharine who died in 1849. In a district where the same dozen forenames repeat for a century, that is a real risk. They are set out because they say where to spend certificate money, not because they are established. matched on name and date, not on a document

A spelling hid her, for the fourth time

Searched as SANIGER, the 1849–1853 window returns two people and neither is Lydia. Searched as SANIGAR it returns three, and she is the second. The surname box does prefix matching when Exact Match is left unticked, so the only safe search for this family is the stem SANIG — which returns all five.

Hannah's baptism was hidden under SANIGOR. Her brother's marriage was hidden under SANIGRE. The Sussex sisters were hidden under ARCY. Lydia's death was hidden under SANIGAR. The pattern is now four for four, and it is the single most reliable fact about researching this family. method

And where they went

Twenty-nine Sanigers in Bristol, 1841

The Vale families did not stay in the Vale. On census night in 1841 there are 29 of them in Bristol alone, under five spellings — and one of them is Lyddya Sinegar, aged 70, living in the Wakefield house.

Name as enumerated	Born about
Rachael SANGER	1766
Lyddya SINEGAR	1771
Rachael SANGAR	1781
Hannah SENGEIR	1801
William SENGEIR	1801
Martha SINEGAR	1810
Mary SUNIGAR	1811
Thomas SINEGAR	1811
Charles SUNIGAR	1816
John SINGER	1816
John SANGER	1816
Caroline SANGER	1821
Ann SINGER	1821
William SANGER	1821
Susan SANGER	1821
James SANGER	1826
John SENGEIR	1826
George SENGEIR	1827
James SINEGAR	1830
Hannah SENGEIR	1831
Lydia SINEGAR	1832
Thomas SINEGAR	1834
Ann SENGEIR	1836
Emma SUNIGAR	1836
Martha SINEGAR	1839
Sarah SENGEIR	1839
John SANGER	1840
Elizabeth SUNIGAR	1840
Stephen SINGER	1840

This is the migration that makes the family findable and unfindable at once. A cloth-trade family leaves a collapsing rural industry for a city, and in doing so passes out of the parish registers that recorded it for two hundred years and into a city whose registers are almost entirely untranscribed — four registers and 1,551 records for the whole of Bristol. They are visible in 1841 because a census caught them, and invisible on either side of it. What each database holds →

A mariner at Chepstow, and how a Vale surname crosses water. EDWARD SANNIGER, mariner of Chepstow, Monmouthshire, died 1806, and two archives hold him. IR 26/352/152 at Kew is the legacy-duty abstract of his will — «Abstract of Will of Edward Sanniger, Mariner of Chepstow, Monmouthshire. Proved in the Llandaff.», 28 November 1806 — and LL/1806/24 at the National Library of Wales is the will itself, Diocese of Llandaf, parish and township of Chepstow.

Chepstow is across the Severn from Berkeley, and a mariner is how a Vale surname gets to the other side of an estuary. SANNIGER is one of the forms in the table above. Nothing yet joins him to this family; what he shows is the mechanism by which the name could travel, which is worth more than another unconnected entry in a count.

Found on 20 September and published here on the 22nd, after an audit asked whether every finding in [/searched](/searched/) had reached a page. This one had not: two references, a date and a trade, sitting in the register where only somebody already looking for him would ever read them. documented

What the table counts

One basis, stated

One basis, stated: England Births & Baptisms 1538-1975 on FindMyPast, searched nationally, every form asked the same way. 27 spellings tried, 20 productive, 373 index entries between 1565 and 1909 — and that is a floor, not a total. SINIGOR and SINNIGER were added on the evening of 21 September, found by a single wildcard query over eleven years of one county, hours after the table had been rebuilt and called complete.

The column counts entries, not children. On 21 September the five largest forms were read row by row and counted as events rather than as rows: about one entry in six is a repeat, and the repetition sits almost wholly in SANIGER and SANIGAR, whose parishes were transcribed more than once. The measured figures are on [/spellings](/spellings/), where they are derived from the count rather than written out again here.

This replaces a table of nineteen forms whose counts were entries in the parish registers read for this archive — a narrower and differently-drawn source. Those counts were not wrong; they were not comparable, which is why SANIGOR and SYNIGAR could not simply be added to them when they were found. The old table is kept in the data as spellingsOld rather than deleted.

SANYGER and SYNIGER still count nought, which is the finding that survived the change of basis: they are the 1608 muster's forms, in Smyth's Men and Armour and in no register anywhere. SANIGRE counts nought here too and is in the table anyway — it is the only form under which Thomas Sinegar's 1828 marriage is indexed, and a spelling that is invisible in one set and decisive in another is exactly what this table exists to show. SINGER and SANGER are excluded — 127 and 40 in Gloucestershire alone, and they are ordinary English surnames rather than forms of this one; counting them would drown the table in strangers.

one stated basis, and a floor not a total

The chain that failed → · The household, 1841 → · The certificate that would settle it →