Open Questions

1 Who were Ljubomir's parents? answered

What is known

This was the archive's central doubt: the join at generation three rested on a man with no document. The Senj marriage register of 8 December 1920 names the groom's parents outright — Juraj Blažević and Tereza r. Žubrinić — and confirms Ljubomir's birth as 1 March 1892.

What has been tried

It was attached in the family tree the whole time and absent from the GEDCOM export, which is why the archive spent its first weeks treating a documented join as an open doubt.

What would settle it

Nothing further. What replaces it is question 1b.

The 1920 register →

1b Who was Juraj Blažević's father? — a house number, and a second witness substantive

What is known

Four more registers were read in September 2026 and the man is no longer a bare name.

He has an occupation and a house. His son Marijan's baptism, 12 April 1895, enters him as “Jurko Blažević i Tereza rođ. Žubrinić, selj.” — peasant — living at Smokvica kbr. 114.

And a second register names the couple. Anka Blažević's marriage to Pavao Kosina on 20 November 1919 gives her parents as Juraj Blažević i Tereza r. Žubrinić — independently of the 1920 entry for their son Ljubomir.

And the house is now written out in full. Antonija “Tonka” Blažević's marriage to Ginro Babić on 1 December 1912 enters her parents as “††Toma i Tereza Blažević, poljodjelci iz sela Smokvica kbr. 114” — not a number in a baptismal margin but the household stated in plain Croatian, both of them marked dead. That is a fourth independent document for house 114 and the first to name Toma and Tereza together with the number. It hardens the household the whole argument stands on; it still does not name Juraj's father.

And the godparents close the circle. The tree gives Juraj two younger siblings, Marija (b. 1880) and Ivan (b. 1882) — children of Toma at house 114. The godparents at Marijan's 1895 baptism are entered as “Ive Blažević, sin Tomin” and “Marija Blažević, ćer Tomina”: the son and the daughter of a Toma, of the right names and the right ages, standing for the child of the man who lives in Toma's house. With the house number demoted to a kin-group, the godparents carry the weight — and they carry it well. Godparents are chosen, and a man's own brother and sister are the classic choice.

What has been tried

The house turned out to be a zadruga, not a household — this archive's own correction. Toma and Terezija Pećanić are recorded at Smokvica 114 in 1882 and 1884, and Toma was born at Br. 114 in 1838. But a control document — the 1890 baptism of Frane Blažević — puts Pavao Blažević, son of Luka, at Smokvica 114 as well. Luka and Toma's father Anton are both sons of Ilija, so Pavao and Toma are first cousins, and the number covers an extended-family compound. It places Juraj in Ilija's kin-group; it cannot name his father. The zadruga →

September 2026 — and the patronymic turns out to be deliberate. Reading the 1888 baptism of Pavao's daughter Terezija, also at Smokvica 114, gives a godfather called Ivo Blažević sin Matin — son of Mate. So two different men named Ivo Blažević stood godfather at this one house within seven years, told apart only by their fathers' names. That is why the priest bothered with a patronymic at all: there were several Ivo Blaževići about, and “sin Tomin” in 1895 is a distinction he chose to draw, not a formula. It also shows the compound from the outside — one house number holding Toma's children, Pavao's children, Jurko's children, an Ivo son of Toma and an Ivo son of Mate.

September 2026 — and the godparent column says it plainly. Four baptisms at Smokvica 114 have now been read, and the godparents are drawn from two distinct Blažević sibling-sets:

1888  Ivo Blažević sin Matin · Anica Tomljanović
1894  Šime Pećanić · Mara Blažević Matina
1895  Ive Blažević sin Tomin · Marija Blažević ćer Tomina
1896  Nikola Blažević · Ana Krpan kći Ilijina

Mate's children and Toma's children, both standing at the same house, plus a Pećanić — Toma's wife's family. That is what a zadruga looks like from the godparent column: one number, several fathers, and the children of one brother standing for the children of another. It still does not name Jurko's father; it does make the compound something you can see rather than infer.

What would settle it

It is still not the sentence that says so. What would give it: Jurko's own baptism in the Sv. Jakov Krmpote register, or his marriage to Tereza Žubrinić before 1892. Two of his children, Marijan (1895) and Ruža (1897), have no marriage recorded anywhere — and every marriage entry re-names the parents, so those are two more chances at the same answer.

1c Who was Milka's first husband? answered

What is known

Zvonimir M. Vukelić, gostioničarski pomoćnik — innkeeper's assistant — born 13 January 1883, married at Senj on 4 November 1907 (marriages p.294 no.30 — read at full resolution 13 September 2026, image 299 of the film). He died in 1915, aged thirty-two, leaving Milka a widow at twenty-eight. Their children Miroslava “Mica” Vukelić (1907–1993) and Josip Vukelić (1913–2004) are Hedviga's half-sister and half-brother — and were in the tree all along.

What would settle it

Nothing further on identity. What remains is 1d.

Milka's life →

1d How did Zvonimir Vukelić die in 1915? — and a correction worth doing

What is known

His baptism has been read: born 13 January 1883 at Senj, baptised 25 February, son of Petar pl. Vukelić — plemeniti, noble — and Lojza. Milka married into a titled family and was widowed out of it at twenty-eight.

And a third occupation, from a third document. On 16 August 1914 he stands witness at the wedding of Milka's sister Franjka, entered as “Zvonko M. Vukelić, pivar” — brewer. The trajectory now reads gostioničarski pomoćnik (1907) → pivar (1914) → gostioničar, bankar (at death). It is also the first document found that places him inside this family other than his own marriage entry.

What has been tried

This archive proposed the Austro-Hungarian casualty lists, reasoning that he was thirty-two in the second year of the war. That reasoning was probably wrong. The tree places his death at Senj, in his own town, and gives his occupation as gostioničar, bankar — innkeeper and banker, a considerable rise from the gostioničarski pomoćnik who married Milka in 1907. A man who dies at home, in a settled trade, is not a Verlustliste entry.

What would settle it

The Senj death register for 1915, and nothing else. The casualty lists are behind a subscription for uncommon surnames and are, on this evidence, the wrong book.

1e Which came first — the wedding or the child? worth doing

What is known

Milka and Zvonimir married on 4 November 1907 by episcopal dispensation, with the banns read once on 3 November rather than three times. Their daughter Miroslava was born in 1907. A dispensation, a one-day wedding and a child in the same year make a familiar shape — but the register does not date the birth, and this archive does not claim to know.

What has been tried

Weakened 13 September 2026. The Senj marriage of Martin Kosina and Angjelika Boras, 12 January 1888, carries the same shape: dispensed by the Senj ordinariate on 7 January, banns read once only on the 8th, married on the 12th. He was 47 and she 30, and there is no suggestion of anything in it. A single dispensation in this diocese looks like ordinary administrative practice, not evidence — so this question rests on the birth date alone, and that is exactly what is missing.

What would settle it

Miroslava Vukelić's baptismal entry, Senj 1907, which gives the day.

2 How did Oto Blažević die? critical

The spine depends on it.

What is known

Two accounts in the same note, from two named relatives: a snakebite, and killed in the war in Slavonia with no known grave. A photograph of the family grave now corroborates the second witness in its checkable parts — Oto's name is on the stone, and Leo and Katica were added later exactly as she described. It still does not establish the cause.

What would settle it

A search of the Yugoslav war victims register (Popis žrtava rata) and Croatian State Archives military death records for Oto Blažević, b. 1923, Senj. An entry — or a confirmed absence — settles it.

The full page →

3 When did Santou Codron leave Rhodes? substantive

What is known

Born on Rhodes in 1913; alive in South Africa to marry Hedviga about 1973. The island's Jewish community was deported on 23 July 1944 and roughly 150 of 1,615 survived. Whether he left before the war — as ~2,000 Rhodes Jews did, for the Congo and Rhodesia — or survived the deportation is not known here, and the difference is enormous.

What would settle it

South African immigration and naturalisation papers giving arrival year and last residence; the Rhodes Jewish Museum community registers and Yad Vashem deportation lists, searched for Codron.

The full page →

4 Is the Port Pirie branch the same Žubrinić family? critical

The spine depends on it.

What is known

Now traceable three generations back from the emigrant: Ivan Žubrinić (1860–1930) → his father Joannes, b. 19 June 1837, Forest Ranger → Elias “Ilija” (1806–1845) → Casparus (1767–1829). And there it stops: in the tree, Casparus's parent is a node literally named “Brothers of Casparus (It Seems) Žubrinić” — a researcher's guess written into the structure. Nothing joins the two lines — and as of 13 September 2026 it is possible to say exactly what the apparent join is made of. Walk Casparus upward and the chain reads: Casparus 1767 → “Brothers of Casparus (It Seems)” → “C Žubrinić Xubrinich Zubrinic” → “Žubrinić Zubrinix Xubrinic Zubrinic Otocac”. The second is a researcher's note. The third is an alphabet — one of eighteen nodes whose entire given name is a single letter, holding everyone whose name starts with it. Ninety-two people are attached that way.

Walk Toma upward and he stops at a bare node called Žubrinić after one step. So the two lines meet in a filing system and nowhere else.

But they are born a year apart in one small village. Casparus 14 January 1767; Toma 1768, at Šumećica 2 — and the 1834 Frontier baptism this archive has read is at Šumećica 24. A surname held by ~312 people on earth, two men born a year apart in the same hamlet. The Frontier →

What has been tried

The 1760s baptisms do not exist in any collection reachable from here — checked 13 September 2026. FamilySearch's Otočac holdings begin in 1834. This question's stated next step has been unavailable the whole time, and better image access does not change that.

The same parish holds a better document, and this archive had never looked at it. Church Census (Stanje duša) 1710–1846 — a status animarum, filmed at the Croatian State Archives as Otočac knj. 1, 384 images. It is not a register of events but a standing list of households, and its columns are the ones this question actually needs:

Pagus         the village
Domus         THE HOUSE NUMBER
Nomina Familiæ  every member, relationship spelled out
Dies/Mensis/Annus  a birth date for each

A baptism register would give two fathers' names. This book would give the house — and for a question about whether two men born a year apart in one hamlet belong to one zadruga, the house is the better answer. It covers 1767 and 1768.

What it was judged to cost. It is not in FamilySearch's full-text index — tested, the query returns nothing — and there is no index at the front of the book. The Pagus is written sideways in the gutter and vanishes into the binding shadow on many frames, so on a first sitting it could not even be navigated by village reliably. That meant about 110 openings read one at a time, looking for Žubrinić in the name column and Šumećica in the margin. A sitting, not a lookup — but a sitting that could actually be taken.

14 September 2026 — the census was read, and it does not hold this family. The Pagus margin turned out to be perfectly legible: the village is written bottom-to-top in the gutter, so the earlier sitting's rotation rendered it upside down. Rotated the other way, twenty-two openings gave the village order of the whole book — Ottosaz · Luka · Dubrava · Novoselia · Lieška · Kostelievo Selo · Obilje · Beljue · Shumechicza · Luka · Extinctæ — and put Šumećica at images 102–106. Five openings, not a hundred and ten.

All five were read. The households are Dašović (Domus 9), Štabinčić (10), Orehović (4), a Paulus household, a household of Andreas and Stephanus, Janežić, Ollar, Kovačić, Orehovački. There is no Žubrinić at Šumećica in this book, and no Domus 2 and no Domus 24. The Extinctæ section was read too; the ruled census stops about image 112 and what follows has no Pagus column at all.

The likeliest reading is that the Šumećica block was written once, early, and never re-made — its households carry births into the 1840s because members were added to existing families, not because new families were entered. And there is outside support for that: the 1913 Hungarian gazetteer gives Šumećica a population of 200, and Degen's Flora Velebitica calls it a Häusergruppe, a group of houses south-west of Otočac. Two hundred people is roughly twenty-five households against the census's seven or eight. The book is not the whole village. notes/otocac-status-animarum.md

And a warning about the other book on the same film. The Otočac marriage register 1859–1872 was opened the same day and read: images 129–147, Latin, ruled, pag. 10 = 1862, pag. 18 = 1865. It has no parents column. It gives a name, a trade, a village, a house number, an age and whether the person was single — and never a father. So it cannot push any line upward, and no plan should assume it will. One entry was read in full: 25 September 1859, Nicolaus Žubrinić, vicedecu… — a Military Frontier rank — of Prozor, aged 24, marrying Anna Dujmović of Prozor, aged 18, with Lucas Žubrinić among the witnesses. Aged 24 in 1859 means born about 1834–35, and ahnentafel 10 is a Nicolaus Žubrinić born 1834. It is deliberately not claimed as his. The tree gives ahn 10's wife as Ana Marić, not Dujmović; the surname is thick on the ground in this parish; and this archive has already once announced a direct ancestor as a duplicate on exactly this kind of name-and-year coincidence.

15 September 2026 — the earlier death book is read end to end, and it holds a Thomas who is not Toma. The Otočac Liber Mortuorum 1780–1836, images 156 to 332, every opening. It buries four grown men of the name, and only one has an age that puts his birth in the 1760s: 27 February 1808, Thomas Xubrinich, annorum 40 — born between February 1767 and February 1768.

Toma's own death was read the same week, in the next book — 4 January 1838, An. 70, ex Pago Shumcchicza N° 2 — which puts his birth in 1767 or 1768 as well. So the parish buries two men called Thomas Žubrinić born in the same two years, and the 1808 entry gives no village and no house. Read on name and age alone it would have been claimed as Toma and been wrong. It is recorded instead as a candidate for the other Thomas — the husband of the census's Martha vidua defti Thomæ at Obilje, whose son Paulus was born February 1805 — and as nothing more.

The same book puts Žubrinići at Šumećica and Obilje from 1789, forty-nine years before the 1838 entry: Mathæus Xubrinich ex Sumechicza, Ann. 26, corroborated four months later by Rosalia filia dfti Mathei Xubrinich ex Sumechicza, his orphan daughter; a second Šumećica house under Josephus; and Lucia Xubrinich ex Obilje, Annor. 66. Five villages in 1789 — Šumećica, Obilje, Dubrava, Potok, Selca. No house numbers: the ruled Domus column does not begin until 1828.

And a caution about Casparus. 20 November 1806 buries Casparus Xubrinich annos circiter 40, morte improvisa — born about 1766, where the Casparus of Dubrava 60 who died in 1829 aged 62 is born about 1767. Both fit the census's b. 14 January 1767 on age alone. That identification stands on the house number and a four-point census match, not on the age — but it should be read with this beside it. notes/otocac-status-animarum.md

What would settle it

So this question no longer belongs to the status animarum, and the hundred-and-ten-opening plan should not be attempted. The instrument is the baptism register. The 1846–1858 book writes the household into its Conditio column as «Rustici Confiniarii e Šumećica N° 2», and the record index gives the image ARK of every indexed entry, so each one opens at the page directly. About 138 Žubrinić baptisms at Otočac, 1834–1862 — twenty-three to fathers called Michael, two to a Nicolaus.

One caution, established the same day. The 1834–1846 book does not record house numbers: its Conditio cell holds a single status word, checked at December 1839 and March 1845. Ahnentafel 10's own baptism of January 1834 — «Confiniarius, ex pago Shumechicza N° 24» — is an early-book habit that the parish abandoned. The house numbers are readable for 1846–1858 and for the first years of the 1834 book, and not in between. notes/otocac-house-numbers.md

And separately, Tereza Žubrinić's baptism, 29 November 1867 at Otočac — which Otočac baptisms, stopping at 1858, fall nine years short of.

4b When did Elias Žubrinić and Ana Untersinger actually die? worth doing

What is known

The explanation is in how the dates are written. Six people in this family carry the bare string 1845 — no day, no month, no place, no source: Elias, Ana Untersinger, Phillipus Žubrinić, Martinus Žubrinić, a Maria born 1838, a Maria born 1809, and Nicolaus Kraljić. The only two 1845 deaths in the whole archive that carry a real date — Jelena Katnić on 21 November, Mihovil Kalanj on 13 March — belong to other families entirely.

An epidemic entered in a register produces days and months. A year applied in bulk produces exactly this. So 1845 is far better read as somebody's estimate for a group who dropped out of the record together — which also dissolves three of the contradictions, because children born in 1847 are not impossible if the parents' death year is a guess.

What has been tried

This question used to read “what killed them in 1845?”, on the reasoning that two adults dying in one year suggests an epidemic or a famine in 1840s Lika. Tested on 13 September 2026, that did not survive.

Across the whole 15,643-person export, 1845 carries 27 deaths against a mean of 25.8 for 1820–1880 — no spike whatever. Restricted to deaths at a Croatian place it carries none at all, and the worst Croatian years in this archive are 1875, 1880 and 1867.

What would settle it

Any decade of the Otočac death register between 1830 and 1860. Not to find an epidemic — to find out when these six actually died. On Untersinger, a German surname in Lika which the Frontier's administrative class would explain, see the connected families.

4d Which Maria Orešković, and was there a second wife? — NO SECOND WIFE, THREE MEN answered

What is known

Answered 13 September 2026, from the tree's own structure. With the surname fold fixed — it had been excluding 63 Žubrinići — the question comes apart cleanly.

Ahnentafel 20's children are a coherent set: Nicolaus 1834 · Daniel 1837 · Francisca 1840 · Maria 1841. Four children in seven years at Otočac. That is one marriage, contracted about 1833, behaving exactly as a marriage should.

And the tree holds a second Michael — Michael Miho Michaël Žubrinić, married to Maria Marković, whose children are Theodorus 1841 · Josephus, 16 March 1845 · Marta 1849 · Magdalena 1852.

That is the entry this archive read. The Otočac baptism of 16 March 1845 — Michaël et Maria Žubrinić nata Marković, the one that made this question suspect a second wife — is his, and the tree has held him all along.

What has been tried

What went wrong is a filing error. The 1845 scan is tagged to ahnentafel 20. This archive read it, found a wife née Marković where it expected née Orešković, and invented a second marriage to explain the difference. There was no second marriage. There are at least three men of the name — the third being the ††Mate Žubrinić i Marija rodj. Orešković named at the 1910 marriage as parents of Nikola, b. 1880, who is also wrongly hung on ahnentafel 20, thirty-nine years after that coherent set ends.

What would settle it

Nothing further on the second wife — there was none. What remains is smaller and still worth doing: which of the three couples Maria Catharina Orešković, born 8 September 1835, actually belongs to. She cannot be the mother of children born 1834 to 1841, so she is the third couple's, and her whole Orešković–Sekula ancestry moves with her.

The Otočac marriage register around 1855–70 would name that couple outright. And a small tidy: Josephus 1845 is in the tree twice, both entries dated 16 March 1845 — one baptism recorded as two people.

4c Can the Žubrinići be joined into one family? — NO, NOT FROM HERE worth doing

What is known

Both numbers this question was built on were wrong, and the answer changes with them. The tree holds 422 people of the name, not 232 — the surname fold split on a bracket but not a slash, so Žubrinić / Xubrinich never matched. And they resolve into 90 loose ends, which is high for a reason the old figure never named: 92 people hang off an A–Z index, eighteen nodes whose whole given name is a single letter.

What has been tried

Of the 90 loose ends, 79 are bare names — no date, no place, nothing to search on. Four have both. So joining them is not the largest piece of work outstanding here; it is work that cannot be done from this archive at all, and saying otherwise was the mistake. The fragments, counted →

What would settle it

A systematic read of the Otočac registers (and Prozor, Šumećica, Kompolje), rather than record-by-record hint-following. This is a job for someone who can work in Latin parish hands and knows that X spells Ž before about 1850.

5 Are Antić and Papić one family or two? — ONE, and a by-name explains it answered

What is known

The tree's own structure answers most of this. Of 15,643 people there are 7 Antić and 117 Papić, and they never interleave: no family anywhere contains both an Antić and a Papić child. The two names touch at exactly one point — Georgius Juraj “Jure” Antić Papić, ahnentafel 12, born at Selce on 11 March 1838, who is entered carrying both surnames and the word Papa. Every one of the nine cross-name parent-child links in the tree is a child of his. Everything above him is Antić; everything below is Papić.

So this was never a question about two families. It is a question about one join — and the register says what happened at it.

Ahnentafel 24's baptism, Vrbnik, 14 March 1806: patre Joanne, Matre Marthæ Antich Papa, nomine Josephum. The family is Antić. Papa is its house by-name — underlined by the scribe exactly as he underlines Loncsarich Franich and Antich Dudin on the same page. A nadimak, the device that tells one household from another where a village shares a surname, doing the same work as the house numbers at Smokvica.

1806  Vrbnik  Antich Papa — surname Antić, by-name Papa
1835  Selce   Josephum ANTICH PAPA — ahn 24's own marriage
1835  Selce   Euphemia ANTICH ALITER PAPA — his sister, same day
1838  Selce   ahn 12 born, carrying every form at once
1885  Senj    his son's marriage enters the father as PAPA alone
after  every descendant written Papić

The gap in the middle closed on 14 September 2026, from a scan that had been sitting in this archive unread. Selce, 22 November 1835: Copulavi Josephum Antich Papa, filium Denati Joannis et Margarittæ, cum virgine Maria Jellicich, filia Denatorum Georgii et Margarittæ — underlined by the scribe, exactly as at Vrbnik. That is ahnentafel 24 marrying ahnentafel 25 and naming ahnentafel 48, 49, 50 and 51, and it carries both names on the man himself.

Found, 14 September 2026 — and it answers the surname question and opens a name question. Ahn 12's baptism is image 186 of the Selce births film 3:1:3QSQ-G99C-5963-L, right-hand page, first entry:

Die 11ª Martii 1838 · Domus 51 — Ego Vincentius Rombal Parochus Loci Baptizavi GREGORIUM … conjugum Josephi Antich Papa, filii Denati Joannis et …csarich, et Mariæ, filiæ Denatorum Georgii et Margarittæ.

The identification is certain — the tree's exact date, the same house 51 as the parents' 1835 marriage, the same priest, and both sets of grandparents matching. The surname question is answered: the priest wrote Antich Papa for the father, so the by-name is carried at the font too.

A by-name displaced a surname across three generations, which is an ordinary Croatian process, and there is now a document at three of the four steps.

What has been tried

And the formula is the parish's, not the family's. The entry immediately above ahn 24's marriage enters his sister as “Euphemia Antich, aliter Papa”. Nine such pairs have now been read out of this one parish, in three different hands across thirty-three years:

Antich     aliter / dictus  Papa
Antich     aliter          Dudich  (twice, 1835 and 1842)
Jellichich  dictis          Jurešina
Lonzarich   dicti           Hel
Franich     aliter          Lonzarich
Puxich      aliter          Lonzarich
Badrovich   aliter          Bagatello
Penes       aliter          Paschul

And the scribes underline them. Jellichich dictis Jurešina is underlined in 1809 by Michaël Rombol, exactly as Antich Papa is underlined in 1835 and 1838 by Vincentius Rombal. Selce recorded a house by-name beside the surname as a matter of course, for at least six different surnames, and marked it as what it was. Papa is one instance of a parish-wide system — which makes this archive's reading the ordinary explanation rather than the clever one.

But the given name is not what the tree says. The tree has ahn 12 Georgius Juraj “Jure”. The register, underlined by the scribe and read at the highest magnification the scan allows, says Gregorium — Gregorius, Croatian Grgur. The same line writes Georgii twice, for the bride's father, and it does not look like this word. Stated as a reading, not as a correction: against it, a man known all his life as Jure is an odd Grgur, whose short form is Grga. Either the tree misread it when it was built, or the priest slipped, or I have. notes/selce-1838-baptism.md

The 1885 marriage was read at full resolution on 13 September 2026, and it did not go as expected. The word is Papa — a clean round a, no abbreviation stroke, room to spare in the column. This archive had assumed the missing -ić was an artefact of a 960-pixel scan. It was not: the priest wrote the by-name, in 1885, for a man whose children were all written Papić. That is the displacement caught in the act. See 5c.

What would settle it

Ahn 12's marriage to Matia Pilipić, and his death entry of 29 January 1912, each name him again in another hand, and Selce marriages run to 1878 — which is what would settle Gregorium against Georgius.

And a guess, offered as one. Ahn 49, Margarita “Martha” Lončarić — the Martha of the 1806 entry — has a mother recorded as Anna Papić. A woman actually surnamed Papić marrying into this Antić household one generation before the by-name first appears is exactly how a nadimak arises: the household becomes known by the wife's family. It would explain where Papa came from. It is also a single undocumented name in a tree, and nothing here tests it — the Selce marriage of Pauli Lončarić and Anna Papić would.

5b Two sisters born eighteen days apart substantive

What is known

Franjka Papić and Šulka Papić are both daughters of Josip Papić and Antonija r. Prpić — Milka's parents, so both are Hedviga's aunts. Franjka's marriage register (Senj, 16 August 1914) gives her birth as 20 January 1895. Šulka's (Senj, 21 December 1919) gives hers as 7 February 1895. That is eighteen days, and they cannot both be right for full sisters.

Neither date is this archive's inference: each is written in that woman's own entry. A marriage register records what the bride said, not what the baptism book says, so either could be the loose one. They might be twins with garbled dates; one might belong to a different Papić couple, and there were at least three having children in Senj in these years.

What has been tried

It is not a tree error. All four Papić sisters' marriage registers have now been read, and the tree reproduces every one of them exactly: Ana 25 May 1888, Božica 9 December 1891, Franjka 20 January 1895, Šulka 7 February 1895. So the registers disagree with each other and the tree faithfully carries both — which moves the question off the transcription and onto the books.

What would settle it

The Senj baptism register for January–February 1895 settles it. Both entries would fall within three weeks of each other in the same book, and it would take one sitting.

5c Four of the most important documents were thumbnails — three are now read at full size answered

What is known

The median scan in this archive is 2,934 pixels wide and several are 4,500. Four were 960, and they were not minor ones: the 1920 marriage that answered question 1, the 1885 marriage that names four direct ancestors, the 1907 marriage that is Milka's hidden first, and a 1859 birth already superseded by a 3,569-pixel copy.

On 13 September 2026 the three that mattered were found on FamilySearch and read from the deep-zoom tiles — about six times the width of the tree's copies, and nine times on the crops that decided anything.

p.192 no.5  3 May 1885    image 196  3:1:3QSQ-G99X-17PK
p.294 no.30  4 Nov 1907   image 299  3:1:3QS7-L99X-175J
p.366 no.41  8 Dec 1920   image 371  3:1:3QSQ-G99X-17M7

Two other questions fell out of the 1885 row. The word this archive could not read — the groom's father's surname — is Papa, plainly, with a clean round a, no abbreviation stroke and room to spare in the column. It was never a resolution problem: the priest wrote the by-name. Question 5. And Lucia, ahnentafel 15, who carried no surname at all, is entered Luca rodjena Pèr-pić — hyphenated over two lines, the grave accent standing for the syllabic r. She is a Prpić who married a Prpić at Krivi Put. One of the seven nameless women on the direct line has a name.

What has been tried

The method, since it was wrong twice before it was right. It first said the viewer does not repaint its canvas under automation. It repaints perfectly well. It then said the tiles were unreachable — 401, no CORS headers, tainted canvas — and that was my error: I tested the wrong host. The viewer pulls tiles from sg30p0.familysearch.org, which is cross-origin and does all of the above. The same tiles are served same-origin from www.familysearch.org — no preflight, no taint, cookies sent automatically. Fetch …/deepzoomcloud/dz/v1/<ark>/image.xml for the size, then the 256-pixel tiles at level 13, and stitch.

A third correction, to the navigation. This page said image 374 is page 350. It is not — the film runs about page = image − 5, fixed by four measured points, and image 372 is END OF ITEM. Nor does the ?i= parameter drive the viewer. What does: every filmstrip thumbnail carries alt="Image N" and its ARK in the same tag, so the whole number→ARK map scrapes out of the DOM and the tiles can be fetched with no navigation at all. The method is in tools/familysearch.md; the transcriptions are in notes/reading-2026-09-13-familysearch.md.

What would settle it

What is still not read. One witness at the 1920 wedding is Franjo Bora—, brijač, a barber whose surname runs into the column rule and is illegible at every crop tried. Boras is one of this archive's own surnames, which is precisely why it is left unread rather than guessed.

1f Twenty-two years, an emigration and a whole country critical

The largest single hole in the life of the person this archive exists for.

What is known

Hedviga is at Senj in April 1953, when her son Robert is born. She is next seen arriving in South Africa in April 1975, and the entry says “from Zimbabwe”. Between those two dates this archive holds nothing whatever — not a departure, not an arrival, not an address. An entire country is known only from two words on a later record.

So the crossing happened after April 1953 and before 1973, when she is said to have married Santou Codron. Twenty years wide, and everything about the family's life in Rhodesia sits inside it — three children schooled somewhere, a husband working somewhere, and a marriage ending, since Ivan died back at Senj in 1995.

What has been tried

The family remembers it as 1948, a post-war move for better and safer conditions. The record will not allow 1948: their second and third children were born at Senj in December 1951 and April 1953. But 1948 is a real year here — Ladislava “Ada” was born at Crikvenica on 5 March 1948 — and a remembered date attaching itself to the wrong event is the commonest fault in this whole archive.

A migration leaves more paper than a birth does, and none of it has been looked for, because until September 2026 this archive did not know there was a Rhodesian decade to look in.

What would settle it

· Southern Rhodesia immigration and permit records, 1953–1965
· South African permanent-residence and naturalisation files, 1975
· Yugoslav passport applications — leaving was not a private act
· School enrolments in Salisbury or Bulawayo for three children
· Whatever ended the marriage, since he returned and she did not

The National Archives of Zimbabwe in Harare holds the immigration series.

6 Why is Antonija Prpić buried at Mirogoj? worth doing

What is known

She was born at Krivi Put 136 in 1859 and died at Senj in 1935 — but is buried at Mirogoj in Zagreb, 200 km away. Nothing in the record explains it. A child living in Zagreb is the obvious guess and it is only a guess.

What would settle it

The Mirogoj burial register, which gives the grave, the date and who paid for it.

7 Two deaths on 30 October, six years apart worth doing

What is known

Ivan Anton Defranceski died at Senj on 30 October 1995; Hedviga died at Johannesburg on 30 October 2001. Both dates come from the tree, not from certificates seen here. A shared date is exactly what memory tidies into place.

What would settle it

Both death certificates — Croatian civil registration for Senj 1995, South African for Johannesburg 2001.

8 What was the Senj tobacco factory? worth doing

What is known

Two women of this family worked there — Milka as a supervisor, Anka Blažević (1893–1953) on the floor. The tree gives the building's address and says it is now a textile factory. Its founding date, size and closure are unconfirmed here; the machinery is said to have gone to Rovinj after the Second World War.

Its scale can now be read off one page. Senj marriages p.359 covers 20 November – 22 December 1919, entries 44 to 51. Six of the eight brides — Anka Blažević, Marija Ilić, Ivka Babić, Marija Babić, Božica Zrinski and Šulka Papić — are entered radnica u tvornici duhana, and entry 45's witness Ivan Antić is a mašinista in it.

And the register shows the whole hierarchy, because it records each person's grade:

radnica / radnik   worker
vratar   doorman
mašinista   machinist
mestar   master
nadglednica / nadglednik   supervisor
nadčinovnik   senior clerk

Milka is a nadglednica; so is Antonija Blažević. Two women of this family reached supervisor, which makes the factory a career and not merely a wage.

It was also a route out. Antonija “Tonka” Blažević married Ginro Babić in 1912 — nadglednik u tvornici duhana na Rieci — and went to Rijeka with him; the witness at that wedding, Andrija Babić, held the same grade at the same Rijeka factory. Floor in Senj, supervisor in Rijeka.

What has been tried

Three pages across three registers, counted the same way:

p.303 · 1909 · 3 of 5 brides
p.350 · 1918 · 3 of 6 brides
p.359 · 1919 · 6 of 8 brides

This is not one unusual page. Across at least 1909–1919 the factory is the standing occupation of the young women marrying in Senj, and men appear in it too — Mate Tadić, radnik u tvornici duhana, and Ivan Antić, mašinista.

Ten pages have now been counted the same way, 1909 to 1919: 3 of 5 brides, 3 of 5, 4 of 5, 4 of 5, 4 of 5, 3 of 4, 3 of 6, 7 of 8, 6 of 8. It is not one unusual page or one unusual year.

What would settle it

The Senj town museum and Croatian industrial-history literature on the state tobacco monopoly after the patent of 20 November 1850.

9 Three place errors that will mislead someone data quality

What is known

(a) Juraj Blažević's birthplace is geocoded to Smokvica, Općina Smokvica, Dubrovačko-neretvanska županija — that is Smokvica on Korčula, 400 km south. The village meant is Smokvica Krmpotska near Novi Vinodolski. (b) Ivan Žubrinić of Port Pirie is given “Senj, Split-Dalmatia”; Senj is in Lika-Senj. (c) A FamilySearch citation for Nicolaus Žubrinić's Otočac christening is catalogued under “Osekovo, Belovár-Körös Megye, Hungary”.

What has been tried

Half-done, 13 September 2026 — the harm is stopped even though the tree is not fixed. The gazetteer already folded these, so the village counts have been right for a while. What was never fixed is the place where a reader actually meets the string: the person page, which printed it raw. It now carries a line saying the county named is a real place 400 km (a) and 250 km (b) from the event, and what the place actually is. Two people carry such a string; both are annotated.

What would settle it

Still outstanding: correcting (a) and (b) in the MyHeritage tree at source, which is the only way the other five sibling archives stop repeating them. (c) is FamilySearch's own catalogue heading, not the tree's — an Otočac christening register filed under an unrelated Hungarian district — and is noted wherever the citation is reused.

10 The photographs are not in the export answered

What is known

The GEDCOM contains no media records at all. Seventy items have now been swept out of the MyHeritage media library across the whole line — 34 of them register scans, certificates and gravestones, many at 4,500 px. Three are read out in full: the 1920 marriage register, the 1951 birth extract, the family grave.

What would settle it

Nothing further for the direct line. The same sweep has not been run on the wider register, and the site library holds about 4,500 items in total.

All of them →

11 Who is in the photographs? substantive

What is known

Most of the recovered photographs carry no caption but a camera filename — 1595624858.9990501, 1134835622, IMG_5752. Faces this archive can name are guesses from resemblance and from the ages of children in shot.

What has been tried

And five of them were not photographs. Five items were being published as portraits of ancestors born between 1806 and 1838; they are stock question-mark placeholder graphics, one with a visible Adobe Stock watermark. They are now excluded from the build. The five →

What would settle it

Somebody in the family looking at them. This is the one open question here that no archive can close and a single phone call could — so the pictures have been gathered onto a single contact sheet meant to be put in front of a relative, with what is being asked of each one spelled out.

The pictures →

12 Why does Ivan Defranceski have Arolsen card files? answered

What is known

They are his, beyond reasonable doubt: all three give the birth as 25 June 1925 at Crikvenica, matching the 1951 civil birth extract exactly. They place him at Licht und Kraft, Reutlingen, as an Hilfsarbeiter — unskilled labourer — from 21 October 1941, aged sixteen, lodging at Brunnenstraße 7, Pfullingen, with a second Reutlingen employer from January 1942.

What would settle it

Nothing further on identity. What replaces it is 12b and 12c.

The full reading →

12b Was he a forced labourer or a free one? critical

What is known

The 1950 municipal questionnaires carry three printed lines designed to answer exactly this — Wurde der Arbeiter restlos entlohnt? (was he paid in full), Dienstverpflichtet (compulsorily assigned), Freier Arbeiter … seit (free worker, since). All three are blank on both cards. The collection is titled a card file of persecutees, but that is the name of a zone-wide registration file, not a finding about anyone in it. This archive claims neither.

What would settle it

A direct enquiry to the Arolsen Archives for the full file on Defranceski, b. 25.6.1925 — they answer families without charge, and the online collections may hold employment or transport records these three cards were extracted from.

12c How long was he in Germany? substantive

What is known

The second employment runs from 26 January 1942 to 28 November 194? — the last digit of the year runs into the column rule and cannot be read. 1942 gives ten months in Germany; 1944 gives just over three years. His daughter is born at Crikvenica in 1948, so he was home by then, and nothing at all is documented between the card and that birth.

What would settle it

A higher-resolution scan of the insurance card, or the Arolsen file itself. The second employer's name is also unread — an old German hand, possibly beginning Otto.

9b “Scanned” did not mean unread — FIXED answered

What is known

The coverage page grades every person on the direct line, and 12 were marked scanned — defined there as an image is attached but has not been read out.

So scanned actually meant “has an image and no swept register row”, which is a different claim. The real backlog was eight, not twelve — and it is now none. Eight readings on 13 and 14 September cleared it, among them the 1835 Selce marriage that closed question 5's middle step and the 1806 baptism that opened 5d.

What has been tried

On 14 September 2026 all twelve were worked through, and four had already been read. Ahnentafel 42 and 43's 1835 baptism is quoted on /frontier; ahnentafel 106's 1754 and ahnentafel 214's 1779 entries were both read and written up on 10 September. The tool did not know, because it decides the state from the data alone and had no way to learn that a page or a note has read a document.

What would settle it

Fixed the same day. A new file, sources/readings.psv, lists every document that has actually been read out, one row per person per document, each with its citation — and tools/coverage.py now grades from it. A baptism counts as a reading for the child and for both parents, because it names all three.

The effect on the direct line: read went from 19 to 31, and scanned from 12 to none. Nothing was re-graded that had not genuinely been read; the tool simply stopped being ignorant of eight weeks of work. The file has to be maintained by hand, which is the honest cost — a reading that nobody records still will not count.

5d Vrbnik or Selce? — SELCE, and the label was the error answered

What is known

Reading the 1806 baptism at full resolution on 14 September 2026 turned up a conflict the archive had not noticed in its own pages.

What the entry says — and it confirms question 5 exactly: Die 14. Martii Aº 1806. baptizavi ego Michael Rombol C:L: infantem lege natum patre Joanne, Matre Marthæ Antich Papa, nomine Josephum, susceptoribus conjugibus Judice Mathia Jellichich et pia consorte Margarita. Two details are new: the godfather is the village judge, and a Jelličić — ahnentafel 24's godfather is of the family he would marry into twenty-nine years later — and the entry below confirms Lonzarich Franich and Antich Dudin, both underlined.

This archive holds four registers in the hands of one family of priests — Michaël Rombol in 1806 and 1809, Vincentius Rombal in 1835 and 1838 — and the same by-names run through all of them, on Krk in 1806 and on the mainland in 1835.

What has been tried

The problem was the parish. The header reads Anno Dñi 1806 in the original hand. Beneath it, in modern block capitals, someone has added * 40 NOVORODJENIH and VRBNIK — finding-aid annotations, not the register's own statement. But a cataloguer's note carries weight, and it was corroborated: the officiant of the next entry is a Canon of Dobrinj, which is on the island of Krk, as Vrbnik is.

Either the family moved from Krk to the coast between 1809 and 1835, which would make the tree's Selce birthplace for ahnentafel 24 wrong and would explain the by-names travelling because the families carried them. Or all four are one parish and the pencil is a mislabel, in which case the tree is right and a visiting canon is unremarkable. For a few hours this archive asserted both at once: /searched called the scan Vrbnik baptisms while the tree gave ahnentafel 24 a birthplace of Selce.

What would settle it

Closed the same day, and by neither of the routes first proposed. Instead of hunting a parish heading, the Selce deaths of 1806 were opened — image 108 of the Selce film — and they are written per me Mich: Rombol Capell: Loci. The 1806 baptism signs itself Michael Rombol C:L: — the same abbreviation for Capellanus Loci. Same man, same office, same year, in a film labelled Selce.

Two more things from that one page: the Antići are demonstrably at Selce in 1806 — Die 20 Junii 1806. Helena vidua Antonii Antich, annorum circiter 50 — and the tree independently gives ahnentafel 24's birthplace as Selce.

So the pencil was wrong and the tree was right. The family did not cross from Krk: they are at Selce in 1806, 1809, 1835 and 1838, under two generations of Rombol priests. /searched's label has been corrected.

The one loose thread, kept rather than tidied: the Canon of Dobrinj officiating at the entry below. Dobrinj is on Krk and Selce is Senj–Modruš, so he is a visitor from another diocese — worth noticing, not enough to outweigh a priest signing his own register in his own hand.

13b Two Selce registers disagree with each other about when she died data quality

What is known

This question said something else for a few hours on 14 September 2026, and it was wrong. It said the register contradicted the tree about ahnentafel 51. The tree is right.

The Selce Liber Defunctorum, entry 147, house 204: “Die 8ª Maji 1842 — Margareta Jellicich, Uxor Georgii…” That is the tree's date to the day.

The real conflict is between two registers in the same parish, in the same priest's hand. Twice — at her daughter's marriage in 1835 and again at her grandson's baptism in 1838 — Maria Jelličić is entered as filia Denatorum Georgii et Margarittæ. Read at 3.6×, the word is unmistakably the genitive plural: the -orum ending is plain, and the line above it writes Denati for the groom's father with a clear -ti. The priest wrote twice that both her parents were dead, and then buried her mother seven years later.

What has been tried

Two explanations, and this archive cannot choose between them. Either the formula was careless — denatorum reached for as a set phrase when only the father was dead — or there are two Margarita Jelličićs, wives of two men called Georgius, and the one buried in 1842 is not ahnentafel 51. The 1842 entry is at house 204; nothing yet fixes which house the Jelličići of 1835 lived in.

Georgius was read the same day, and it does not settle it. He is in the same book: Die 1ª Aprilis 1832 · house 39 · Georgius Jellicich, Antonii — the tree's date exactly, plus a house and a patronymic the tree did not have. But 39 is not 204. A widow moving to a child's house in ten years is entirely ordinary, so this proves nothing either way.

And a third house, read the same day, makes the house useless here. Ahnentafel 25's own baptism of 22 March 1809 puts the family at house 110. So the line runs 110 (1809) · 39 (1832) · 204 (1842) — three houses in thirty-three years. Either they moved twice, Selce renumbered, or these are not one household. The house number is not a reliable identifier in this parish, and the argument made from it is weakened in both directions.

One real discriminator did turn up. The 1809 baptism enters the family as Jellichich dictis Jurešina — with a by-name, underlined. The Margareta buried in 1842 carries none. That is suggestive and not conclusive: a death entry is terser than a baptism, and by-names are dropped as often as used. But it is the first thing found that could separate two Jelličić households rather than merely failing to join them.

What would settle it

What is left: the Selce marriage of Georgius and Margarita, before 1809, which would give her maiden name — the tree says Brozović — and whether it carries Jurešina. Selce marriages run from 1790.

13 The gravestone contradicts the register data quality

What is known

The Senj stone gives Ljubomir's birth as 1891. The 1920 marriage register, written in his lifetime from his baptismal certificate, gives 1 March 1892. This archive follows the register.

What has been tried

14 September 2026 — the Rijeka baptismal entry cannot be reached. Rijeka was carried in this archive's parish table as unknown, not checked. It is nothing of the kind: Župa Rijeka is filmed 1590–1920, with baptisms running almost continuously from 1594 and alphabetical Kazalo indexes to births, marriages and deaths. But the birth volumes go

… Rođeni 1880-1886 · 1886-1890 · 1890-1891 · 1900-1901

There is no 1892–1899 birth register on the film, and the birth index runs 1800–1889 and then 1895–1902. Ljubomir's baptism falls in the gap twice over — by image and by index. It is not on FamilySearch in any form. sources/familysearch/holdings.psv

What would settle it

The Rijeka baptismal entry of 1892 would settle it beyond argument — and would also independently name his parents a second time. That is not a dead end so much as a redirection: the entry exists, in the Rijeka state archive or the parish, and a letter asking for it has been drafted and is waiting to be sent. This question now turns on a letter rather than a sitting.