Corrections

was Dec 15 1969

now 15 December 1968

how Johannesburg Chevrah Kadisha burial register: died 15.12.1968, buried 16.12.1968, Westpark block L grave 211, Yahrzeit 24 Kislev 5729 — which is itself 15 December 1968.

was 1945

now not documented; deported to Auschwitz 26 February 1943

how Bundesarchiv Gedenkbuch en1100957, which records the deportation and deliberately gives no date of death.

was Nov 9 1991

now 8 November 1991

how Johannesburg Chevrah Kadisha burial register: died 08.11.1991, buried 10.11.1991, Westpark block U grave 2469.

was Georg Leibholz

now Oskar Leibholz

how Census of 17 May 1939 via Mapping the Lives: Gertrud and Oskar share an address, Waghäuseler Str. 9/10 Wilmersdorf. Georg was at Michaelkirchstr. 26, Mitte, with Lydia née Perl.

was son of Sally Leibholz and Friederike Schmul

now NOT A BROTHER — the dates forbid it. Probably a cousin or nephew of the same family

how Jacob was born at Schubin on 8 August 1884 and Georg at Hammerstein on 7 June 1884 — SIXTY-TWO DAYS APART. Two sons of one mother cannot be. Both dates are the Bundesarchiv's, independently of the family, and that arithmetic is what this correction rests on. *** THE WORDING BEFORE 20 SEPTEMBER 2026 ALSO SAID 'two hundred kilometres apart' AND CALLED GEORG 'a different Leibholz family'. BOTH ARE WITHDRAWN. *** SALLY LEIBHOLZ, JACOB'S OWN FATHER, WAS BORN AT HAMMERSTEIN ON 7 JULY 1847 and married at Szubin on 24 December 1874 (corrected 27 September 2026 from 24 November, which was the family record's date and is not the register's). The family had people in both towns, so the distance between them is not evidence of anything and the surname at Hammerstein is not a stranger's. What the dates rule out is BROTHERHOOD; what they leave standing is a cousin or a nephew in the town Sally came from. The Leibholz page has said exactly that in its 'A likelier reading' panel since the start — 'right surname, right family, wrong generation' — and this correction contradicted it. It no longer does.

was This entry formerly read: "marriage to Unknown Da Rocha, before 1941 · WAS: married · NOW: NO SUCH MARRIAGE — she was a spinster in February 1948", sourced from the Germiston Duplicate Original Marriage Register no. 34 of 1948, column 5 Personal Status: SPINSTER. On that basis the Ivone page published "She had never been married", "it did not happen", and "there is no marriage record to find, because there was no marriage".

now *** WITHDRAWN 27 SEPTEMBER 2026. THE MARRIAGE HAPPENED AND THE FAMILY TREE HAD IT RIGHT. *** Ivone married FREDERICO PINTO DA ROCHA on 18 NOVEMBER 1939 at the 1.ª Conservatória do Registo Civil de Lourenço Marques, and that marriage was dissolved by divorce, by a sentence of 6 DECEMBER 1945. Both facts are averbamentos on her own birth registration. She was therefore not a spinster at Germiston in February 1948 but a divorcée of two years, and the Portuguese state still recorded her as DIVORCIADO on her death certificate in 2010.

how Assento de nascimento n.º 447-D, fls. 224, Conservatória dos Registos Centrais, Lisboa, 7 May 1987, transcribed from the 1.ª Conservatória do Registo Civil de Lourenço Marques; averbamentos 1 and 2. Photographed from the family's own papers and supplied on 27 September 2026. WHY THE GERMISTON REGISTER LOST: it records what was declared in a South African office about a status acquired in another colony, while the assento is the continuously maintained register of her own civil status, naming a husband, an office, a wedding date and a court sentence. A divorcée declaring herself single abroad is ordinary; a registrar inventing all four is not. THE LESSON IS THE COST: this archive struck a true fact out of the family record on one word in one column, and then built three panels on the strike. Working in data/ivone-assento-nascimento-1922.tsv.

was dress designer

now dress designer (1933); shop keeper (1948)

how Johannesburg marriage register no. 858 of 1948, column 6: Shop Keeper. The 1933 emigration record gives dress designer. Both are true of one man.

was no date; Sasolburg

now 2 April 1963, Sasolburg

how THE TREE'S OWN DUPLICATE, not an outside record. The harvest holds this man twice: 9500005, born December 1923 at Johannesburg, with a death place and no date; and 9500160, listed as HIS SON, of exactly the same name, born circa 1923 at Pretoria, died 2 April 1963 at Sasolburg. A father cannot be born in the same year as his son, and the two die in the same small Free State town. They are one man, and the date of death is on the duplicate. *** INDEPENDENT CONFIRMATION OBTAINED, 20 SEPTEMBER 2026, AND IT DID NOT NEED THE ESTATE FILES. *** "South Africa, Civil Death Registration, 1953-1967" — Entry for JOHANNES ANTONIE OELOFSE DE WITT, 2 APRIL 1963, SASOLBURG, ORANGE FREE STATE, aged 40, born 1923 at PRETORIA (FamilySearch ark:/61903/1:1:DD9Y-PLT2). A national death register says what the tree's uncited duplicate said. IT ALSO SETTLES WHICH HALF OF THE DUPLICATE WAS RIGHT ABOUT THE BIRTHPLACE — Pretoria, not Johannesburg. AND IT RAISES A DOUBT ABOUT THE BIRTH DATE: the death register makes him 40 on 2 April 1963, and a second marriage found the same day — to ELSIE MARIE GEYER on 25 FEBRUARY 1950 (ark 1:1:6ZHR-NVYN) — makes him 27 then. The two ranges overlap at April 1922 to February 1923, and the tree's "December 1923" falls outside both. Declared ages are not proof; two of them thirteen years apart, agreeing against the same date, are worth recording. See data/dewitt-confirmed.tsv.

was son of Johannes Antonie Oelofse de Witt, born circa 1923 at Pretoria

now NOT A SECOND PERSON — the same man as 9500005, duplicated

how Same name, same birth year as his supposed father, same place of death, and a father-son relationship that is arithmetically impossible. The duplication also invented a wife for him — 'Unknown de Witt', 9500161, who exists only to be his mother — and gave Delina de Witt a half-brother who is really her father. Found by tools/audit-tree.py, 20 September 2026, which flags a parent under fourteen at a child's birth. *** DELIBERATELY NOT COLLAPSED INTO 9500005. *** The DUPLICATES map in tools/build-register.py collapses records only where the pair shares AN IDENTICAL BIRTH AND DEATH DATE, and says in its own comment that those are not guesses. This pair does not meet that bar: the birth places differ, Johannesburg against Pretoria, one birth is 'circa' and only one record carries a death. It is a strong inference and not an identity, so both records stay in the register with this correction attached to them, which is what corrections are for. *** 20 SEPTEMBER 2026: THE BAR HAS BEEN PARTLY MET AND THE RECORDS STILL STAY APART. *** The civil death register gives the birthplace as PRETORIA, which is the duplicate's birthplace and not 9500005's Johannesburg, so the two records now differ on a point where an outside source backs one of them. That strengthens the inference that they are one man and does not turn it into an identity, because nothing yet ties the Johannesburg birth to anybody at all. Both records stay in the register.

was a node in the family tree on the front page, flagged “Removed — she was nobody”

now OFF THE TREE. Kept in the evidence layer and nowhere else

how DAVID'S DIRECTION, 20 September 2026: “if this person is a nobody she really doesnt need to be in the tree. keep it in evidence but not on the main screen.” HE IS RIGHT AND THE PANEL WAS SELF-INDULGENT. Edwina Ann Wear was born EDWINA ANN GREEN and married BRUCE WALLACE WEAR — Leonard's younger brother — in 1980. She is a Wear by marriage and NO BLOOD RELATION TO DERRICK AT ALL. This archive had spent a day on her, decided she was a coincidence, and then left her standing in the tree across all three columns under a heading announcing that she had been removed. A withdrawal does not need a monument, and a family tree is for the family. WHAT SURVIVES AND WHERE: her burial at West Park on 8 July 2014 in the cemetery-register table on the Leonard page; the cancelled estate request 33157/2014 on the status page; the negative in data/negatives.json that uses her to bound the Johannesburg cemetery search; the reasoning on the DNA page; and Bruce's own 1980 marriage on his timeline, because that is a fact about Bruce. ONE THING CHANGED WITH HER. The PHYLLIS AGNES ROBERTSON node existed to explain Edwina's mother. It now points where it always should have — at GILLIAN AGNES LOUISA NEL, born Wear, born eleven months into that marriage and carrying Agnes in her own name.

was flew from Berlin to Melbourne — this archive's own reading, published 20 September 2026

now sailed from Bremerhaven to Melbourne on the ship FLAVIA, arriving 17 September 1965

how NAA: A2559, 1965/86/85-88, item ID 5435314 — the Australian half of the same nominal roll, whose title states the ship, the port and the date. The error came from reading the mode of travel off the Arolsen FOLDER heading, “TRANSPORT BY AIRPLANE”, instead of off the sheet. A folder heading describes the folder.

was 54.35 N, 21.15 E — a pin placed before the village had been identified, and fifty kilometres from the village

now 53.935 N, 20.846 E — Zerbuń, Gmina Jeziorany, powiat olsztyński

how German Wikipedia's coordinates for Zerbuń, the village this archive identified as Sauerbaum on 20 September 2026. Checked because a DIFFERENT claim was being verified — the distance from Sauerbaum to Bischofstein — and the arithmetic came out at 33 km against a published 15. The distance claim was near enough (17.8 km from the real village). THE PIN WAS NOT. The map had been showing Auguste Luwinski's birthplace fifty kilometres from where she was born, since before the village was identified at all.

was LEIBHOLZ. Read from the OCR text layer of the 1980-81 Berlin telephone book, which prints "Leibholz Erna 3 35 84 05 20 Stteitsu.39", and published as the last appearance of the Leibholz name in the post-war Berlin run.

now LEIPHOLZ — the eighth spelling. The 1975-76 and 1979-80 telephone books print the same woman, the same address and THE SAME TELEPHONE NUMBER 3 35 84 05, with the P clear in both.

how ZLB Berlin, Berliner Telefonbuch 15849358_1975-76 p.1444, 15849358_1979-80 p.1704 and 15849358_1980-81 p.1770, read 21 September 2026. *** WHAT ACTUALLY PROVED IT WAS NOT A BETTER SCAN BUT THE ALPHABET EITHER SIDE OF HER. *** In every volume where the name is genuinely LEIBHOLZ it is printed between Leibfried and Leibich. In the 1980-81 volume that line sits between LEIPERT WERNER and LEIPNER BODO. Leipholz belongs in that gap and Leibholz cannot be in it, so the surrounding names settle the letter that the OCR could not. The same technique — read the alphabet, not the word — is what found Erna Frida Else Leipholz in the Schöneberg death index the same afternoon.

It mattered beyond one letter. The claim it supported was that a LEIBHOLZ survived in Berlin until 1981, which was this archive's latest trace of the name in the city. That trace belongs to the other spelling, and the correction makes the two Ernas visible: ERNA FRIDA ELSE LEIPHOLZ died on 13 January 1975, and ERNA LEIPHOLZ of Streitstraße 39 is in the telephone book for six years afterwards. TWO WOMEN OF THE SAME NAME, which the misreading had hidden inside one.

was "After May 13 1942". The family record gives a FLOOR and not a date, and the register turned that floor into a year: year() takes the first four digits it finds, so the page carried 1942 as though it were the year she died, with nothing to say it was a bound.

now MAY 1942, at KULMHOF (CHEŁMNO) EXTERMINATION CAMP. Day not recorded.

how Bundesarchiv Gedenkbuch entry en1169257, read 22 September 2026: Elli Struzik née Leibholz, born 07.11.1915 in Berlin, of Berlin (Mitte), deported from Berlin to the Litzmannstadt (Łódź) ghetto on 1 November 1941 and from there to Kulmhof on 13 May 1942; place of death Kulmhof, date of death 00.05.1942. The birth date matches the family record to the day, which is what identifies her.

IT EXPLAINS THE FAMILY RECORD RATHER THAN CONTRADICTING IT. "After May 13 1942" is not a guess — it is the DEPORTATION DATE, the last day anybody knew she was alive, written down as the only honest thing that could be said. The Gedenkbuch closes the gap by a fortnight and names the place. THE YEAR ON THE PAGE DOES NOT CHANGE and that is the point: this archive was right by accident, because a floor and the truth happened to fall in the same year. Joachim Luwinski's "Before 1955" has no such luck behind it and is still only a bound, which the bloodline and the timeline now say in words.

was "ERNA — 1980-81 telephone book — the last Leibholz in the run", and, in the summary line of the same file, "then Erna alone in 1980". THE SPELLING HALF OF THIS WAS ALREADY CORRECTED ON 21 SEPTEMBER AS 9500904 — she is a LEIPHOLZ — but two things were left standing: data/berlin-postwar-luwinski.tsv and the Leibholz page were never amended to match 9500904, so a reader of either still met the old surname; and NOBODY HAD ASKED WHO THE LAST LEIBHOLZ ACTUALLY WAS.

now HILDE LEIBHOLZ, at Wiesbadener Straße 57, is the last of the name in the post-war Berlin run. She is in NINE CONSECUTIVE telephone volumes, 1972-73 to 1980-81, in the flat Felix held before her — including the very 1980-81 volume from which Erna was called the last. Erna is in that volume too, three places further on in the alphabet, under the other spelling.

how ZLB Berlin telephone books, PI_ANCHOR 15849358, read at the alphabetical slot on 22-23 September 2026 rather than by name search, which is what turned Hilde up: five of the eleven readable volumes mangle one or other of these surnames and a name search misses them. Supersedes nothing in 9500904, which stands; this records the claim that was built on top of it. Working in data/leipholz-telephone-slot-audit.tsv.

was KRONENSTRASSE 59-60, published on the Leibholz page, in data/berlin-adressbuch-names.tsv, in data/jacob-leibholz-deportation.tsv and in two research-log entries.

now KRAUSENSTRASSE 59-60. The house number is unchanged.

how Berliner Adressbuch, ZLB Berlin. FOUR READINGS IN THREE VOLUMES, none of which says Kronen: 1925 page 1871 prints it cleanly, "JaqueB Leibholz & Co, Konfektion, W8, Krausenstr. 69. 00"; 1934 page 1462 "W8 Kraufen-/strasse 69. 60"; 1938 page 1633 in the householder run "W8 Krausen-/strasse 59. 60"; and 1938 page 3385, a different section of the same volume, "Leibholz J & Co W8 Kraujeuilr 69. 60". THE DECIDING EVIDENCE IS THE OCR ARTEFACT RATHER THAN THE SPELLING: the f of "Kraufen" and the j of "Kraujeu" are both this scanner's rendering of the long s, and Kronenstrasse has no s in that position. Found on 22 September while sweeping the householder runs for row 260, and the wrong street survived because both streets are real and both sit in the Konfektion quarter around Hausvogteiplatz.

was "So Jacob's 1913 registry office is one of Standesamt Charlottenburg I to IV", published on Kurt's page and in the research log.

now ONE OF CHARLOTTENBURG I, II OR III. Charlottenburg IV did not exist in 1913 — its earliest volume of any kind is 1922 — and the unnumbered office closes in 1897.

how Landesarchiv Berlin, LABSA, 23 September 2026. THE ERROR CAME FROM GUESSING THE OFFICE NAMES RATHER THAN ENUMERATING THEM, which is the same root as this archive's 143-against-207 miscount of the Charlottenburg holdings the day before: the search matches an office name exactly, and the list of names is only available from the form's own autocomplete endpoint. The three surviving candidates hold 114 name-index volumes between them. Working in data/labsa-charlottenburg-inventory.tsv and data/labsa-berlin-catalogue.tsv.

was "Henrique was born on 11 March 1941 to an unmarried mother... His father gave him a surname and, on the evidence of that surname, acknowledged him. Ivone did not marry him." The page also concluded that his birth registration was "the only document likely to name his father — there is no marriage record to find, because there was no marriage."

now HE WAS BORN INSIDE A MARRIAGE — sixteen months after his parents' wedding of 18 November 1939 and four years before their divorce. He carries Pinto da Rocha because his parents were married to each other when he was born, which is a simpler explanation than acknowledgement by surname. His father has a forename: FREDERICO PINTO DA ROCHA. THERE IS ALSO A MARRIAGE RECORD TO FIND, at the 1.ª Conservatória do Registo Civil de Lourenço Marques, 18 November 1939.

how Averbamentos 1 and 2 to assento de nascimento n.º 447-D. NOT A PROOF OF PATERNITY: a husband at the date of birth is a legal presumption, not a certificate, and nothing here excludes another father. What it removes is the archive's reason for treating his surname as evidence of acknowledgement, and it replaces one unlocated document with two.

was 1924, from the family tree, and repeated on her page as the heading "Lourenço Marques, 1924".

now 4 March 1922

how Assento de nascimento n.º 447-D, Conservatória dos Registos Centrais, Lisboa — "dia quatro, mês Março, ano mil novecentos e vinte e dois" — and independently her 2010 death certificate, which carries 04/03/1922 in its own boxes. Her own declared ages contradict each other and both: 22 at Germiston in February 1948 and 24 in June 1948, when she was 25 and 26.

was no death recorded. The page said only that she died and was buried at Porto, and called the missing date "the largest single hole in this archive".

now 14 March 2010, at home at Rua Nova das Marinhas 82, Gulpilhares, Vila Nova de Gaia

how Certificado de óbito n.º 12893536, Ministério da Saúde: died at 04:30 in a private dwelling, of cardio-respiratory failure with dementia contributing, aged 88. Supplied by the family 27 September 2026. The day is confirmed by the family and is consistent with the form: the physician signed on 16 March, two days after. This archive first read the boxed digits as "the 16th or the 17th" — the 17th was impossible on the certificate's own evidence, since a doctor cannot certify a death that has not yet happened.

was “Forty-four of the fifty-two are Berlin — Schöneberg 6, Wilmersdorf 6, Charlottenburg 5, Friedrichshain 5, Tiergarten 4, Neukölln 4, Mitte 4, Prenzlauer Berg 3, Kreuzberg 1, three plain Berlin, two with two boroughs each.” Work list row 260, written 22 September 2026.

now FORTY-SEVEN of the fifty-two have a Berlin residence. Thirty-eight carry one borough — Schöneberg 6, Wilmersdorf 6, Charlottenburg 5, Friedrichshain 5, Mitte 4, Tiergarten 4, Neukölln 4, Prenzlauer Berg 3, Kreuzberg 1. THREE carry two boroughs each, not two: Hans, Laura née Rothenberg and Sophie née Rothenberg. FIVE are plain “Berlin” with no borough, not three: Else née Soldin, Helene Kamnitzer, Willi, and the two under the LAIBHOLZ spelling, Georg and Lidia. And ONE is Berlin second after another town, Heinz Paul Hugo at “Paderborn (Umschulungslager Grüner Weg); Berlin (Neukölln)”. Five of the fifty-two are not Berlin at all.

how Recounted mechanically from data/leibholz-gedenkbuch.tsv on 11 October 2026, all fifty-two rows. THE ROW'S OWN FIGURES NEVER ADDED UP TO ITS OWN TOTAL: 6+6+5+5+4+4+4+3+1 = 38, plus three plain and two dual = 43, written as forty-four. The arithmetic was wrong before the data was. The two LAIBHOLZ rows are the likeliest cause of the undercount — they are the Gedenkbuch's own alternative spelling of Georg and Lydia of Mitte and are easy to read past. Working in data/adressbuch-boroughs-1938-1943.tsv.

was “Two men stripped of their trade in a single year, 1940 to 1941 … WILHELM ISRAEL LEIBHOLZ of NO18 Weinstraße 6 goes from VERTRETER, a commercial traveller, to BAUARBEITER.” data/berlin-adressbuch-names.tsv, 20 September 2026.

now THE FALL STARTS A VOLUME EARLIER AND HAS THREE STEPS, NOT ONE. 1938: Wilhelm, KAUFMANN, NO18 Weinstraße 6. 1939: Wilhelm, VERTRETER, same house. 1940: Vertreter. 1941: Wilhelm Israel, BAUARBEITER. A merchant became a commercial traveller in the 1939 volume, before the compulsory name appears beside him at all, and a building labourer in the 1941 one.

how Berliner Adressbuch 1939, ZLB 34115495_1939 page order 1719, the Leibholz run read at the page image on 11 October 2026 — “—Wilhelm Vertret NO18 Weinstr 6”. WHY IT WAS MISSED: the 20 September reading recorded the 1939 forenames but not the 1939 trades, so the only trade change visible in the file was the one between the two volumes that did carry them. THE CORRECTION DOES NOT WEAKEN THE ORIGINAL POINT AND SHARPENS IT: the archive read the 1940-41 step as marked, because “Israel” is printed beside it. The 1938-39 step carries no marking of any kind and happened anyway.

was “Married Frances 29 February 1956, sued her for RESTITUTION OF CONJUGAL RIGHTS in 1957” — on Kurt's page in two places and in his timeline as “1957 · Files for divorce again”.

now IT IS 1956, THE YEAR OF THE MARRIAGE. Witwatersrand Local Division, “ILLIQUID CASE. RESTITUTION OF CONJUGAL RIGHTS. KURT LUWINSKI VERSUS FRANCES (FORMERLY SEFTON, FORMERLY BERNER, BORN HOSTOWSKY)”, reference ZA TAB WLD_0_1878/1956_1. He married her on 29 February 1956 and sued her before the year was out. THE MARRIAGE LASTED MONTHS, NOT A YEAR AND A HALF.

how National Archives of South Africa, the NARSA Search Portal, 11 October 2026. *** AND THE 1957 HAD NOTHING BEHIND IT. *** The claim appears nowhere in this archive's data layer — no file, no citation, no NAAIRS reference — only in the prose of one page and the timeline built from it. THE LIKELIEST ORIGIN IS A CONFLATION with WLD 2542/1957, STANDARD BANK OF SOUTH AFRICA LTD v K LUWINSKI, which this archive has held since 14 September and which is a different action against a different plaintiff. A year read off the nearest case in the list. The reading that the restitution action is 'the legal signature of a wife who has left' is unchanged and is strengthened: she left inside the first year.