The Luwinski Archive

Ivone de Figueiredo

Lourenço Marques, 4 March 1924 — Porto

She was born in a Portuguese colonial town on the Indian Ocean, had a son at seventeen, married four men in Johannesburg between 1948 and 1953 — two of them Jewish, one of them an Afrikaner, the first a Portuguese she left behind — and was buried at the other end of the empire, in Porto.

Her name is Mota Figueiredo Documented

The family tree calls her Ivone / Mota “Yvonne” Figueiredo, and everyone — this archive included, until the South African national archives index was read properly — took Mota for a stray nickname sitting next to the anglicised Yvonne.

It is not a nickname. It is half her surname. The Transvaal Supreme Court, 1948:

Illiquid case divorce. Kurt Luwinsky versus Ivone (born Mota Figueiredo).
— National Archives of South Africa, Transvaal depot, Witwatersrand Local Division, WLD 1950/1948

She was born Ivone Mota Figueiredo, in all likelihood Ivone da Mota Figueiredo.

Which gives us her parents' surnames

Portuguese naming order puts the mother's surname before the father's. Her son's name already demonstrated the rule working in this family — Henrique Figueiredo Pinto da Rocha, his mother's name then his father's. Applied to her own:

Surname
Her fathera FIGUEIREDO
Her mothera MOTA — most likely da Mota

It is the first thing this archive has ever known about Ivone's parents, and it came out of an index entry eleven words long. Inferred from a documented name

The names

She appears in the record under five, which is itself the story. The constant is Figueiredo: the name she was born with, and the one the court reached for in 1951 when it had to say who she actually was.

As writtenWhere
Ivone Mota FigueiredoHer birth name in full — the Supreme Court of the Transvaal, 1948
Ivone de FigueiredoAs the 1948 Johannesburg marriage register has it
Ivone / Mota “Yvonne” FigueiredoAs the family tree carries her — “Mota” unexplained, “Yvonne” the anglicisation Johannesburg gave her
Ivone de WittThe name she married Kurt Luwinski under, in June 1948
Ivone LuwinskiThe name she married Morris Salkow under, in February 1953
Ivone, formerly de Witt, born FigueiredoThe Supreme Court of the Transvaal, 1951

Her son's name gives up her father's Inferred

Her first son was Henrique Figueiredo Pinto da Rocha. Portuguese naming puts the mother's surname before the father's. So Figueiredo is hers, and Pinto da Rocha is the father's — the man the tree can only call “Unknown da Rocha”.

Which makes Pinto da Rocha, not da Rocha, the name to search on.

Lourenço Marques, 1924

Lourenço Marques — Maputo since 1976 — was the capital of Portuguese East Africa, a deep-water port built on the traffic of the Witwatersrand goldfields: the Rand's nearest sea, joined to Johannesburg by rail since 1895. A Portuguese family in Lourenço Marques in 1924 was living four hundred miles from Johannesburg on a line that ran straight there, which is the plainest available answer to how she came to be married in Germiston twenty-four years later.

Three court files, all now identified

Kurt sued her three times. The national archives index has all three, in the Transvaal depot, and they can be ordered.

CaseYearAs the index has it
WLD 1950/19481948Divorce. Kurt Luwinsky v. Ivone (born Mota Figueiredo)
WLD 497/19511951Restitution of conjugal rights. Kurt Luwinski v. Ivone (formerly de Witt, born Figueiredo)
WLD 2368/19511951Divorce. Kurt Luwinski v. Ivone (formerly de Witt) (born Figueiredo)

Watch what changes. In 1948 she is not “formerly de Witt” — only born Mota Figueiredo. By 1951 she is. Either the first clerk left it out, or the de Witt marriage was not being recognised in the earlier action. A divorce summons of this period normally recites both parties' full names, ages and domiciles and the date and place of the marriage, so these three files are worth ordering on their own account.

The 1951 divorce was granted Documented

The index lists the actions; it does not say how they ended. That is now settled from outside the court. When Kurt married again on 29 February 1956, the registrar entered his personal status as Divorcee and noted in the Remarks that an order had been filed — the decree he had to produce before a marriage officer would proceed. A registrar in Johannesburg saw the document that ended this marriage and wrote that he had seen it.

Which also dates the end of Ivone's second marriage to somewhere between 1951 and February 1956, and she had already remarried inside that window — Morris Isaac Salkow, 13 February 1953. See Kurt for the register entry in full.

Four marriages

DateHusbandHimWhere
before 1941A da RochaName not recordedSon Henrique, born 11 March 1941
2 Feb 1948Johannes Antonie Oelofse de Wittb. Dec 1923 Johannesburg; d. SasolburgGermiston
19 Jun 1948Kurt Israel Luwinskib. 1912 Berlin; d. 1969 JohannesburgJohannesburg, cert. 858
13 Feb 1953Morris Isaac Salkowb. 12 Oct 1921 Pretoria; d. 9 Nov 1991 Johannesburg, buried West Park

The Germiston register, read at last Documented

It was attached to the family tree as a photograph and never opened. Register no. 34 of 1948, Germiston, 2 February 1948:

Johannes Antonie Oelofse de WittYvonne Figueiredo
Age2522
BornTransvaalLourenço Marques
Personal statusBachelorSPINSTER
OccupationMiner, undergroundDressmaker — firm
Residence77 Browning Street, Fairview42 Joel Road, Berea

Married by special licence no. 156884, dated the same day, without antenuptial contract. Witnesses B. Green and E. E. Naude; marriage officer A. J. Rossouw, Magistrate's Office, Germiston. Both parties signed: J.A.O. de Witt and Y. Figueiredo — so the Y-spelling is not only what officials wrote about her, it is what she wrote herself.

Read again at full resolution. The first reading of this frame was made from the 1,195-pixel copy the website serves; the original is 4,931 × 3,929. Five cells came out differently and have been corrected here: the register number (34, not 54), the licence number, both witnesses, and the husband's address. Self-corrected Nothing in the argument below changes — spinster, Lourenço Marques and the age of 22 all read the same at either size.

Which means there was never a da Rocha marriage Corrects the tree

She was a spinster in February 1948. She had never been married. The family tree records “Marriage to: Unknown Da Rocha, before 1941”it did not happen.

Henrique was born on 11 March 1941 to an unmarried mother of sixteen. His father gave him a surname and, on the evidence of that surname, acknowledged him. Ivone did not marry him. She came to Germiston in 1948 still single, and her son stayed in Mozambique for twenty-one years.

It also makes his birth registration at Lourenço Marques the only document likely to name his father — there is no marriage record to find, because there was no marriage.

And the gap is four months and seventeen days Unexplained

2 February 1948: spinster, marries de Witt.
19 June 1948: divorced, marries Kurt Luwinski.

Both are on official registers, both signed by her. A divorce granted inside four and a half months was not ordinary in South Africa in 1948, and this archive does not understand how it was done — annulment, a void marriage, or a court moving faster than courts moved. The Germiston divorce file is the next document. And the pattern then repeats at once: Kurt filed against her within weeks of their own wedding.

They were in the same trade Inferred

She was a dress maker. Kurt, on his 1933 emigration record, was a dress designer — his mother's trade — and by 1948 a shop keeper. Two people in the Johannesburg rag trade, four months apart on the same page of her life. It is the first plausible account of how they met, and it is pleasingly ordinary.

What the two registers showed about the dates Resolved

The archive previously called these dates a contradiction and guessed the February one was wrong. It was not. Both marriages are real, both registers have been read, and the four-and-a-half-month gap is exactly what it looks like. What was wrong was the assumption that no divorce could fit inside it.

And the register itself now answers it Documented

The 1948 entry has been read — it was attached to the family tree as a photograph, unopened. Column 5, Personal Status, gives Kurt as Bachelor and Ivone as DIVORCED.

She was not still married to de Witt in June 1948. That marriage had ended. So the February 1948 date is wrong, as suspected — the de Witt marriage was earlier, and she came to the second one a divorcée, under her first married name.

The Luwinski date is solid: it is on the certificate, number 858, with her age given as 24. The February date carries the same citation as the June one, which means it was never independently sourced. The likeliest reading is that the de Witt marriage was earlier and the February 1948 date belongs to something else — its dissolution, most obviously. But she was already Mrs de Witt when she married Kurt, so the de Witt marriage is real and came first. The Germiston register settles it.

The son she left, and the son she kept

Henrique was born on 11 March 1941, when she was seventeen. He did not reach Johannesburg until November 1962 — twenty-one years later, and a grown man. In between, his mother had married three more times in another country and had a second son there. Henrique married Dawn Elizabeth Hendey on 18 May 1968, and died on 27 October 2022, outliving his wife by thirteen years.

Her cousins are alive, and there are dozens of them

Two of the Portuguese DNA matches were opened in full. Both resolve to common ancestors at great-grandparent depth — which is exactly where Ivone's own parents and grandparents sit. Match holders are living and are not named here; their families' surnames are, because those are the search terms.

MatchShared DNABest readingShared matches with Derrick
A match in Portugal, b. 1974108.7 cM, 5 segments2nd cousin's son, 52.1% — common ancestors are great-grandparents36
A match in Portugal, b. 1949133.2 cM, 4 segments3rd cousin 41%; 2nd cousin 17%37

What the closer one implies Inferred

If the second-cousin's-son reading holds, then Derrick and that match's mother — born 1952, died 2018 — are second cousins, sharing one set of great-grandparents. Derrick's Portuguese great-grandparents are Ivone's grandparents. So Ivone's grandparents should sit among the forebears of that family, whose recorded surnames are Jorge, Silvestre and Conde.

The second match's line adds Gaspar and Matos, through a father born 1902 and dead in 1962.

Five surnames to take to the Portuguese registers, where before there were none.

And the clusters do not touch. Every one of the seventy-odd shared matches across both lists is Portuguese — Ferrão, Dias, Fernandes, Fogaça da Mata, Pereira, Da Costa Oliveira, Amorim Meirelles, Queirós, Dos Santos Pais, Machado, Vargas, Carneiro, Matos, Souza Pereira, Pires. Not one British name appears among them. Derrick's matches split cleanly into a maternal Portuguese side and a paternal British one, which is the paternity finding arriving from the other direction.

A small human detail, and a warning for searchers

In June 1948 she gave her age as 24, which is right for a birth on 4 March 1924. In February 1953 she gave it as 25. She was 28. She took three years off. Any search keyed to a 1927 or 1928 birth year is chasing her, not a different woman.

Her names accumulate rather than replace: by 1953 the register carries her as Ivone de Figueredo de Witt Luwinski, all three at once — which at least tells us the clerk had the earlier marriages in front of him.

And one name, from a passport register Candidate only

Her father must be a Figueiredo. Searched across every Portuguese archive indexed on Digitarq, exactly one Figueiredo took a passport for Lourenço Marques before she was born: Augusto Pereira de Figueiredo, 21, of Eixo in Aveiro, on 22 October 1904. It may well be nothing — Figueiredo is among the commonest names in Portugal and the index is a fraction of the whole. The case for and against him →

Porto

She died and was buried at Porto. Not Lourenço Marques, not Johannesburg — the metropole, which she may never have lived in before. No date is recorded, which is the largest single hole in this archive: a Portuguese death certificate names the deceased's parents, and naming Ivone's parents is the whole object of the exercise.

Her line is the one part of this family the DNA confirms outright Documented

Three of Derrick's ten closest matches are in Portugal, sharing 133, 111 and 109 centimorgans, all three flagged by the testing service as having Portugal in common with him. Against 60.6% Portuguese in the ethnicity estimate. Whatever else is in doubt — and a great deal is — Ivone is not.

Why she is not simply looked up Tested

A search of MyHeritage's birth, marriage and death collections for Ivone Figueiredo returns nine results: eight Brazilian, one Floridian. Portuguese and Portuguese-colonial civil registration is not in these indexes at all. Her records exist — in Porto, in Maputo, in Lisbon — and none of them can be reached from a desk. This is not a search done badly; it is a search that cannot succeed, and knowing which is worth something.

A coincidence in the same index, logged and not believed

Two Johannesburg files: WLD 1467/1971, a divorce — Margaret Figueiredo (born Harrison) v. Estácio Figueiredo — and WLD 506/1970, an ex parte application by the same woman.

Derrick's single closest DNA match, at 333.9 cM, is a Harrison in the United Kingdom, on the paternal side.

Be careful with this. A woman marrying into a Figueiredo family is no blood relation of Ivone's, so this is not a link between the two surnames in any genetic sense. What it is: evidence that an English Harrison family was in Johannesburg in exactly these years, in the same Portuguese-Johannesburg world Ivone lived in — and Derrick's paternal cluster is English with a South African connection. Harrison is also a very common name. Speculation, recorded so it is not lost

Still to find