What is being done
The work list
What is being done here, as opposed to what is known.
The register the others do not keep
90 open, 185 in all
Every other register on this site answers what we know. What Has Been Read says what has been searched, negatives included; Open Questions says what the records have not answered; Corrections says where a document and the family tree disagree. None of them says what is actually being worked on, who is waiting for whom, or what was deliberately abandoned and why — so that lived in a chat window, which is the one place that does not survive the session. This is that page. It is kept in the same shape as the other seven archives in this estate, and a build check refuses if it slips.
1. Running — a machine or a person is at it now this archive
Reading the Scilla registers for the Donato and Arena lines
The live sweep. Scilla's civil registration begins in 1809 — not 1816, as this archive believed for eight days and paid for — and the volumes read easily at native width, so this is the one quarter of the family where the registers still have more to give. Recent: the 1844 marriage Tavola, entry 17, which named Domenico Donato's parents and Nunziata Sofi's and turned the Michele Donato of 1845 from a lead into a documented brother. Findings go to what has been read and corrections; this row only says the work is open.
2. Done — finished and published this archive
Give documented-joins.tsv and corrections.tsv the same checker treatment as documented-additions.tsv
On 15 September a new `tools/check_additions.py` was written, and the first thing it did was catch six people this archive told readers were not in the family tree when they are — Giovanni Arena b.1854, Concetta Russo, Domenico Donato, Nunziata Sofi, Antonino Prostamo and Maria Fragalà. One row said «She was not in the family tree at all» about a woman we had already graded a parent edge for, which is only possible between two people who are both in the export. `documented-joins.tsv` and `corrections.tsv` are still unvalidated by anything. The lesson is not that those six were wrong; it is that nothing was looking. Done — and it was done the same day it was raised. `tools/check_corrections.py` validates both files: `corrections.tsv` for slug, source, confidence and false-absence claims, and `documented-joins.tsv` for id format, self-merges, the `basis` vocabulary and missing evidence. It found thirteen rows using confidence labels the page never explained.
3. Done — finished and published this archive
Domenico Donato's own birth — Scilla, 1816 to 1818
DONE, and not by the eighty-opening sweep. The processetti had it: Scilla, Nati 1816, atto 207 — 18 DECEMBER 1816, declared by his father Candeloro the same day. It matches to the day the thirty-eight he is given at his son Michele's birth on 3 October 1855. This row was queued as expensive and was closed cheaply by doing row 18 first, which is the whole argument for mapping a series before sweeping it.
4. Next — ready to start this archive
The other children of Domenico Donato and Nunziata Sofi — Scilla, 1845 to 1860
Two are known, Candeloro in 1847 and Michele in 1855, and eight years sit between them. Naming custom predicts a Maria for the paternal grandmother Maria Paladino and a Pasquale for the maternal grandfather Pasquale Sofi. Same indexing problem as row 3: most of those years are unindexed. AND NOW THERE IS A SECOND REASON TO DO IT: a child of CANDELORO DONATO in the Scilla births of 1810-1830 would settle whether Domenico's mother was Marta Paladino, as his birth act says, or Maria, as the 1844 Tavola says. THE PREDICTION LANDED, 16 September. This row reasoned from naming custom that the couple should have a PASQUALE, for the maternal grandfather Pasquale Sofi. PASQUALE DONATO died at Scilla on 20 April 1937, «Padre Domenico, Madre Annunziata Sofi», with a wife, Generosa Costa. The Maria predicted for Maria Paladino is still not found, and the row stays open for her and for the births themselves.
5. Next — ready to start this archive
Briatico marriages 1843–1849 and 1853–1858; Zambrone births 1811–1839
The Calabrian quarter. Briatico is where the 1849 keystone act came from, and the surrounding years have never been swept. Zambrone's birth run should have its coverage mapped before anything is promised, the way Scilla's has now been.
6. Next — ready to start this archive
Piedimonte Etneo — Mariano Mazza's death, and Rosario Mazza's marriage about 1870
The Sicilian quarter, and the oldest open item here. Rosario's marriage would name Mariano Mazza with an age and a wife and would independently confirm the join that attached the hundred and ten unconnected Piedimonte Mazza to this family. Seven years of the death run are already closed. THE MARRIAGE IS READ, AND SO IS THE VOLUME BESIDE IT. Matrimoni 1871 atto 14, 30 April 1871 — and now the allegati, fascicolo 14, whose printed publication certificate of 16 April 1871 calls all four parents VIVENTE with a trade and a residence apiece. Mariano Mazza was alive, a bracciante, at Piedimonte, aged about sixty-two, which is the first bound this archive has ever had on his death. What the marriage act does NOT give is his age, so the death run is still the route to it. Seven years of the death run are closed; the rest are not.
7. Done — finished and published David
Trove and the 1942 internments — swept by name, and the newspapers do not name them
Worked so far through Trove's web interface, which is flaky enough that several queries had to be retried and scroll-waited. An API key would turn this into clean JSON. Asked on 15 September; not blocking anything else, and it will be done by hand if no key appears. Trove's newspapers also largely stop at 1955, so the yield may be modest either way. SHARPENED 16 SEPTEMBER: this is no longer a speculative search. The NAA has now given two named internees in this family — Rosario Mazza (MP1103/1 Q8110) and Mario Crupi, held at Loveday from 18 August 1942 to 30 December 1943. Trove would carry what the newspapers said about it — the round-ups of north Queensland Italians in 1942 were reported, and internees' names sometimes printed. Two names and two dates is a far better starting point than «the 1942 internments».
SWEPT 27 SEPTEMBER, AND THE ANSWER IS NO. Both men searched by full name. «Rosario Mazza» returns fourteen Queensland articles and not one is about internment — a wine fine, a premises warning, an eviction case, a police court. They may be his and none is attached. «Mario Crupi» returns six, all Innisfail, Cairns, Tully and Townsville, which is the right district and the wrong subject.
Trove holds the camp, not the roll. Loveday returns 1,085 articles — «ITALIAN INTERNEES AT LOVEDAY: First Train Load Arrives», 12 June 1941 — and covers its existence closely and its inmates anonymously. Trove's newspapers run to 1954, so the 1942–43 window is fully covered: this is an absence in the source, not a gap in it. The NAA file stays the only place either man is named.
8. Blocked — waiting on a person or an institution David
Geneanet Premium — the 897 Piedimonte Mazza behind the paywall
A decision, not a task: whether this archive pays for a Geneanet subscription to see 897 indexed Piedimonte Etneo Mazza. Everything reachable without it has been taken.
9. Blocked — waiting on a person or an institution David
Three registers that exist on paper and not on Antenati
Each would have to be ordered from, or visited at, an Italian state archive. Pizzo's 1837 cholera register — only its index survives online, acts 209 to 325, and the act wanted is 327, two past the end. Zambrone's deaths 1813–1821 — simply missing from the digitised series. Piedimonte Etneo Morti 1880 act 122, Mazza Gaetano — indexed but never filmed. Listed together because they share one answer: an order to the Archivio di Stato, or nothing.
10. Struck — founded on a mistake, kept rather than deleted this archive
Mount Gravatt burial records — blocked on finding an endpoint
RAISED ON A STALE PREMISE, AND STRUCK THE SAME DAY. This came off a five-day-old queue and I carried it over without checking. Mount Gravatt was already done — `data/australia-graves.tsv` has held Domenico Francesco Prostamo and Maria Assunta Prostamo in adjacent plots, Mon 3B 607 and 608, since an earlier session. The real work was documenting the endpoint, which is now in what has been read, and it produced a correction and two leads. See rows 27 and 28.
11. Next — ready to start the kit & estate
Tell the other archives that a kit pin will not move on a short SHA
Found while adopting this very page. All eight archives are pinned at kit `b16bea7`, which predates `WorkList.astro`, so every one of them has to bump the pin before it can import the component. Two traps on the way: `npm install` will not move a dependency whose old SHA is still in `package-lock.json`, and writing a short SHA into `package.json` fails silently — the install reports success and the old files stay. It took the full forty-character SHA plus an explicit `npm install @daviddef/archive-kit`. Worth saying once to seven sessions rather than being discovered seven times.
12. Next — ready to start the kit & estate
Offer check_additions.py to the estate, if other archives keep a documented-additions list
It refuses the build on: two rows sharing a name; a parent not older than the child; a parent dead before the child was born; a generation not one above the child's; a loop; a generation that disagrees with the export's own; and a row claiming someone is absent from the tree when the export answers to the name. The last rule is the one that caught six published falsehoods here. If the other archives keep people the tree lacks, they have the same exposure.
13. Done — finished and published a machine
Warn the estate that csv.writer with QUOTE_NONE will truncate a file halfway
Recorded because it nearly destroyed data and the next session will reach for the same tool. Writing `corrections.tsv` back through Python's `csv.DictWriter` with `quoting=csv.QUOTE_NONE` raised on a field containing a double quote — after writing nine of ninety-two lines. Restored from git. A TSV is tab-separated and nothing else: write it with a plain `"\t".join(...)` and assert no field contains a tab or a newline.
14. Done — finished and published this archive
Messina before 1820 — abandoned, and why
A deliberate stop, not a failure. Michele Donato was born at Messina about 1814 and his parents Candeloro Donato and Maria Paladino married earlier still. Probing the whole province year by year — 1815, 1819, 1820, 1821, 1825, 1840 — returns nothing before 1820 and 366 volumes in 1820. Sicily had no Napoleonic period, so there is no earlier series to fall back on. Anything before 1820 is a parish record held by the Archdiocese of Messina and is not on Antenati. So Candeloro Donato is where the Scilla line stops, exactly as Pasquale Prostamo is where the Calabrian one does. A second wall may sit behind it: the comune of Messina appeared in no year probed although its series claims 1069 volumes, which is plausibly the earthquake of 28 December 1908 — recorded as a thing to check, not as a finding.
15. Done — finished and published this archive
The rest of the Scilla 1844 marriage Tavola — stopped on legibility
Entry 17 was read and gave the Donato marriage. The remaining entries were pulled, contrast-stretched and tripled in size, and are at the limit of reading — the Archivio's own cards on the volume say *inchiostro sbiadito* and *grinze*, and they are fair. The S section at image 55 holds entry 39 and it is not a Sofi. Stopped deliberately: the remaining entries are other families, and guessing at faded handwriting is how this archive produced «Caterina Muso» for Caterina Majo.
16. Struck — founded on a mistake, kept rather than deleted this archive
Re-read Scilla Nati 1854 to settle two disagreeing readings
Raised in error, by this session, and struck the same day. The claim was that the 1854 birth had been read two ways and never re-opened. It had: `data/scilla-reads.tsv` settled it long ago. The 1854 Tavola reads «Arena Giovanni di Antonio e Arlotta ROSA, 20 April»; the act itself, act 76, says SERAFINA and says the thirtieth, and the ages close it to the month. The index is simply wrong — and worth remembering, because trusting it once produced a false relationship, making the boy Rocco's cousin rather than his brother. What actually remained was stale prose in corrections, now fixed.
17. Next — ready to start this archive
Map which Scilla birth years carry an annual index, 1816 to 1865
One end-of-volume check per year, done once, so that every later sweep is a lookup instead of a gamble. Known so far: 1853, 1856, 1859 and 1860 have a Tavola; 1817 and 1849 have none. Today's work would have been cheaper if this map had existed, which is the argument for building it before the next sweep rather than after. THE RANGE IS WRONG: it is 1809 to 1865, not 1816 to 1865. Scilla has a napoleonic fondo of seventeen registers with Nati for 1810, 1811, 1812 and 1815. Arks in `scilla-arks.tsv`.
18. Done — finished and published this archive
Map Scilla's marriage processetti bundles — the highest-yield document type here
DONE, AND IT PAID. Scilla's processetti bundles, never mapped before: 1817-1819 (sig 1760), 1820-1823 (1761), 1824-1827 (1762), 1832 (1764), 1840-1841 (1766) and 1842-1845 (sig 1767, ark an_ua2002480, 1935 images) — which covers 1844. Fascicolo 20 is Domenico Donato and Nunziata Sofi, and fogli 1 and 2 are their own birth acts. See what has been read and corrections. The series appears to stop after 1845; 1850 and 1860 have none.
19. Done — finished and published this archive
Retry FamilySearch — it reaches Briatico to 1910, where Antenati stops
UNBLOCKED, AND THE HOPE IN IT DID NOT SURVIVE CONTACT. The site-wide Error 15 has cleared; FamilySearch works, needs no login, and shows both parents on the results page. Add `q.surname.exact=on` or the surname match is uselessly fuzzy. But Briatico is not indexed there. Twenty-six exact Prostamo births 1896-1912 in the Vibo Valentia *Tribunale* collection are all Mileto, Comparni, San Giovanni or Pizzo — and Francessco Prostamo, born 16 November 1900 at Briatico, is not among them. See what has been read.
20. Running — a machine or a person is at it now this archive
The 58 shared names left unmerged since the export was parsed
People who share a name and were deliberately left apart rather than merged on a guess. Each needs a documented decision in `data/documented-joins.tsv` — merged on evidence, or held apart with the reason — not a judgement call. This is the oldest untouched data-quality item here. STARTED: `data/audit.json` turned out to be computed BEFORE the documented joins ran, so it named both sides of merges already made — «Giuseppa Melluso» was published as an open question when her whole cluster was two ids a join had united. `tools/build.py` now recomputes it after every merge: 58 → 57, and three other clusters shrank.
21. Next — ready to start this archive
The nine unnamed crossings, starting with Domenica Prostamo in March 1951
Nine of eighteen crossings are documented. Domenica Prostamo, March 1951, is in the index under neither Prostamo nor Mazza. Also open: Michael Antonio Polistena, 28 May 1950, and Francesca Prostamo, who sails on the Roma with Francessco on 10 October 1957 and is in no tree at all. RE-OPENED 16 SEPTEMBER. Domenica Prostamo's March 1951 crossing was the one this row thought was nearly settled. It is not: the Ugolino Vivaldi identification has been read off the manifest and disproved (row 110). She is back to a month and a year with no ship. What the disproof also teaches is how to search now. The NAA index gives a name, a date and a ship; the manifests give a destination and a port of debarkation, and those are what actually distinguish one Italian family from another in 1951. A woman joining her husband in Brisbane should appear on a manifest marked for Sydney with a Queensland address — not Fremantle, whatever the item title says. That is a much sharper filter than a surname, and it can be applied to every 1951 arrival out of Genoa rather than only to the ones whose name nearly matches. 20 SEPTEMBER: ONE OF THE THREE IS LOCATED, AND DOMENICA IS NOT. Francesca Prostamo has her own record — NAA: BP25/1, *«Trungadi [nee Prostamo], Francesca — arrived Brisbane on the T/v Roma 21 October 1957»*. She is the daughter the manifest already showed, travelling with her father; what is new is that the card carries a married surname, which is a name to search on rather than a second woman. And it corrects her father's arrival and hers. This archive had Francesco Prostamo at Fremantle on 10 October, graded *documented*. The Brisbane aliens series says Brisbane, 21 October — the ship's landfall against the man's debarkation, eleven days apart. Both true; the second is the one that explains a Queensland family. Domenica Prostamo is still nowhere. POSTAMO, the spelling Queensland's registrar uses on her daughter's death registration, returns zero from NameSearch — control-tested against PROSTAMO, which returns thirteen in the same minute. She exists in the National Archives only inside her husband's file title, *J25 1971/1074, «Mazza, Sebastiano; wife Domenica [nee Prostamo]»*. The destination-and-port method this row describes is still the way in.
22. Next — ready to start this archive
How the Arneytown Polistena join the Scilla Polistena — asserted, not documented
The New Jersey Polistena are connected to the Brisbane ones by assertion and nothing else. This is the exact shape of the failure this archive has made four times — a join that looks obvious and has no act behind it. Either document it or mark it as unproven on the page.
23. Next — ready to start this archive
Joseph Polistena's WWII draft card, and CEMLA for the Buenos Aires Prostamo
Two cheap outside sources. The US draft registration gives Joseph's unit and his New York address. CEMLA, Argentina's immigration database, is the only route to the crossing behind twenty-three Argentine events that have no ship attached to them.
24. Blocked — waiting on a person or an institution David
Prepare a captioning sheet for the 133 photographs
Saved before the MyHeritage links expired on 16 September and almost entirely uncaptioned. No record will ever supply these; only Michael and his relatives can. The archive's job is to put the most identifiable ones in front of them in a form that is easy to answer.
25. Next — ready to start this archive
Settle whether Domenico Donato's mother was Marta or Maria Paladino
His birth act of 18 December 1816 reads «e da Mar-ta Paladino, di anni trenta»; the 1844 marriage Tavola calls her Maria. The act is primary and contemporaneous, and this index series has already been caught calling Serafina Arlotta «Rosa» in 1854 — an error that briefly turned a brother into a cousin. But the same Tavola entry gets the *other* mother exactly right. Published as Marta, with the alternative on the page. Another child of Candeloro in the Scilla births of 1810-1830 would end it.
26. Done — finished and published this archive
Map the processetti of Briatico — done; Piedimonte Etneo still to do
BRIATICO MAPPED. Bundles exist for 1824, 1826, 1828, 1830, 1840, 1845 and 1850 — not 1820, not 1822 — and each year returns two or three arks which are DIFFERENT COMUNI in the circondario: 1826 gives Briatico and Sciconi, 1828 adds a Conidoni bundle covering 1828-1851. Read the manifest label before assuming the town. AND THEY ARE ARRANGED UNLIKE SCILLA'S, which decides what a search costs: Scilla's are per-marriage fascicoli each opening with a cover sheet naming both spouses, so the right file is found by reading covers. Briatico's 1826 bundle has no cover sheets at all — a continuous run of loose notificazioni with the names buried in the body. So the route in is the marriage register first, for the act number. Piedimonte Etneo not yet mapped.
27. Done — finished and published this archive
Test the Pinnaroo Donato against the Queensland index — done: named like us, not of us
Tested, and the answer is no — or at least, not provably. The Queensland index documents the whole family: Annunziato Donato and Concetta Briganti had five children registered in Brisbane — Rocco 1912, Concetto 1913, Peppino 1914, Candeloro 1915, Nunzio 1917 — and four of their deaths repeat the same parents. And Annunziato's own death of 28 February 1954 names his parents: ROCCO DONATO and CONCETTA ROMANO — not Michele Donato and Maria Sofi. So the Pinnaroo Candeloro is a *Rocco's* grandson, not this line's. The names stay suggestive and that is all: Candeloro, Annunziato, Rocco, Nunzio and Concetto are exactly the Scilla Donato pool, and Concetta Romano carries a surname this archive has already met at Scilla. A cousin line is plausible. A documented link is absent, and this is the join this archive has got wrong four times. Closed as a lead that did not hold rather than left hanging.
28. Blocked — waiting on a person or an institution this archive
Mario and Maria Mazza of Mount Gravatt — outside the index's reach
Cannot be tested here. Queensland's public death index stops at 1996, and Mario Mazza died in 2015, Maria in 2014. The Mario Mazza who does appear died in 1986 and is a different man, son of Rosario Mazza and Rosa Grasso. So the shared-plot lead at Mount Gravatt is neither confirmed nor refuted — it is simply out of the index's coverage. The routes that remain are a death certificate ordered from the registrar, or Michael's own knowledge. Recorded as blocked rather than quietly dropped.
29. Done — finished and published this archive
Giuseppe Ventrice × Caterina Cichello's marriage — FOUND, Zambrone, 2 March 1824
Found. Zambrone, Matrimoni 1824, atto 2 — 2 March 1824, before Giovanni Sarone, sindaco. Giuseppe Ventrice, 34, born at S. Costantino, contadino, «figlio del fu Domenico e della fu Anna Maropoli»; Caterina Cichello, 26, born at Zambrone, «figlia del fu Francesco e della fu Vittoria Mollo». Hunted across Briatico 1822–27 and Zambrone 1822–26 before this, and it sat three images from where the sweep had reached. See what has been read and corrections.
30. Next — ready to start this archive
Sweep the whole Briatico marriage series 1816–1858 from its Tavole
The cheapest thing on this board, by a distance. Briatico's marriage volumes are six to thirteen images and every one carries an annual Tavola naming BOTH SETS OF PARENTS — «Nomi e cognomi dei Sposi | PATRIA | Nomi e cognomi de' genitori | Giorno della celebrazione». So a whole year, with four parents per couple, costs one image. The arks are sequential — ark(Y) = an_ua87654 + (Y − 1840), now confirmed backwards to 1823 — so the entire series of forty-odd years could be read for about forty images. Marriages run three to five a year; the whole of Briatico's nineteenth century is perhaps two hundred couples. That is small enough to read all of it and stop guessing which year to open.
31. Done — finished and published this archive
Check whether Zambrone's deaths really are missing 1813–1821 — YES, and now proved
Answered by opening the volume instead of inferring it, and the archive had it right. Zambrone Morti 1811–1839 runs 1811, then 1812 ending with its own death Tavola at images 37–38, then a printed title page — and image 40 is 1822. The years 1813 to 1821 are physically absent: nine years, gone. What it costs: Domenico Ventrice died 13 January 1815, squarely inside the gap, so the Fama/Maropoli question cannot be settled from Zambrone's own register. What it gained anyway: the 1822 death Tavola names a «fu Franc.co e Dom.ca MAROPOLI» — so Maropoli is a real Zambrone surname of exactly this period, which is what the 1824 reading needed. And the book is cheap: every year ends with a Tavola naming the dead *and their parents*. One image a year.
32. Done — finished and published this archive
Read the Zambrone marriage acts of 1818–1825 — done, and it found the act
Done, and it worked. The 1823 acts (images 15–17) and the 1824 acts (20–26) were read. Act 2 of 1824 is the Ventrice marriage. And it exposed a reading error of mine: the Tavole in this volume leave their year blank — the printed header says «nell'anno ___» and nobody filled it in — so I had deduced the wrong years from position, calling the Tavola at image 27 «1824» when its three couples are the acts of 1823. Rule: in this series trust the acts for the year and the Tavola only for the names.
33. Done — finished and published this archive
Zambrone 1811–1817 — no longer needed for this marriage
Raised as the fallback if 1818–1825 came back empty. It did not: the act is 2 March 1824. Closed unread.
34. Done — finished and published this archive
Caterina Cichello's own parents — FOUND: Francesco Cichello and Vittoria Mollo
The prize, and it came in the same act. «Figlia del fu FRANCESCO e della fu VITTORIA MOLLO» — both already dead when she married at twenty-six in 1824. A couple that appears nowhere else in this archive and nowhere in Michael's tree. Note the surname: a Mollo is also Pasquale Piatto's mother in the Zambrone marriage of 1823, so they were a local family and Vittoria was probably born in the town.
35. Next — ready to start this archive
Sweep the whole Zambrone marriage book from its Tavole, 1811–1858
Phase 2. The same trick as Briatico and the same volume already half-read: 253 images, forty-seven years, an annual Tavola naming both sets of parents. Reading every Tavola is perhaps forty images and would give every marriage in Zambrone for half a century, with four parents each — Ventrice, Cichello, Melluso, Anile, Barbuto and Prostamo all in one pass. Do this once and stop opening the book a year at a time.
36. Next — ready to start this archive
Map which Briatico and Zambrone BIRTH and DEATH volumes carry a Tavola
Phase 2. The marriage Tavole are mapped; births and deaths are not. Zambrone's births are one volume for 1811–1839 and its deaths another for 1811–1839, so the same one-image-a-year economics may apply — or may not. Establish it before promising any birth or death sweep in either town.
37. Done — finished and published this archive
Build one coverage table: comune × record type × years × is it indexed
Phase 2, and the thing that would have saved the most time today. Four towns, four record types, a hundred and fifty years, and the answers are scattered across `briatico-arks.tsv`, `scilla-arks.tsv`, `zambrone-arks.tsv`, `searched.tsv` and my own memory. One table, on the site, saying for every comune and record type which years exist, which ark holds them, how many images, and whether there is an index. Every sweep after it becomes a lookup. BUILT, 21 SEPTEMBER, AND IT IS BIGGER THAN THE ROW EXPECTED. `tools/build_holdings.py` reads the five ark banks and publishes one table on What Has Been Read: comune × fondo × record type × year span × volumes × how many can be opened today × whether an index was seen. 874 volumes across four comuni, 259 of them openable now. The *openable* column is counted from the banks rather than asserted, so the page cannot overstate what this archive can reach — the same discipline as the «listed here» column beside it. And a blank in the index column means «not recorded», never «no index». Only Scilla's restauration bank carries a `tavola` column at all, which is why the figure is three out of 874 and why it is not a coverage claim. What the table makes obvious at a glance, and what no amount of re-deriving from five files ever did: Piedimonte's *Morti* run to fifty volumes and not one container is bought; its twenty *Matrimoni processetti* are untouched; and Scilla's *Morti* 1866–1943 are eighty-one volumes with seventeen bought.
38. Done — finished and published this archive
SCILLA — Filippo Arena × Domenica Costa: all four years read, and the line stops here
Closed, and it is a wall rather than a gap. The marriage had exactly one place it could be — the four years between Scilla's first civil register in 1816 and their son's birth on 25 November 1819. All four are now read: 1816 and 1817 act by act, forty marriages each and the sindaco's own count on the cover; 1818 through its Tavola at thirty-three; 1819 through its A section. Filippo Arena is in none of them. So the marriage is pre-1816, therefore pre-civil, therefore a parish record of the Diocese of Reggio Calabria and not on Antenati. The Arena line stops at Filippo — the fourth wall of its kind here, beside Candeloro Donato, Pasquale Prostamo, and the nine missing years at Zambrone. And the negative is meaningful rather than merely empty: the Arena were in the town throughout. A *Giuseppe Arena, marinaro*, and *Antonia Polimeni* give a daughter in marriage in 1816; a *Giuseppa Arena* is a bride's mother in 1817. The surname is present in both years. This couple is not.
39. Done — finished and published this archive
Giovanni Arlotta × Maria Costa — found, 7 September 1828
Phase 3. Serafina Arlotta was born 13 June 1830 to this couple, and Giovanni was born about 1779, so the marriage is likely 1800–1829 — half of which is inside Scilla's civil registration. He declared his own grandson's birth in 1854 at seventy-five and gave the boy his name, so he is well attested and badly dated. THREE OF THEIR CHILDREN FOUND, 16 September, though not the marriage. The Scilla death index gives DOMENICO ARLOTTA (d. 14 March 1915, husband of Maria Pontillo), ANNA ARLOTTA (d. 16 March 1915, wife of Antonino Costa) and GIUSEPPE ANTONIO ARLOTTA (d. 2 June 1918), every one «Padre Giovanni Arlotta, Madre Maria Costa». Serafina Arlotta was an only child in this archive until today. Domenico and Anna died two days apart, which is worth reading the acts for. And Anna is the important one — see row 118. THE PROCESSETTI ARE THE ROUTE, 20 SEPTEMBER. Giovanni Arlotta and Maria Costa married before June 1830. Scilla's processetti run 1810–1865 unbroken and their daughter's 1851 bundle (`an_ua2002487`) may carry their death acts; their own bundle would be 1824–1827 (`an_ua2002461`) or 1828–1831 (`an_ua2002463`), both now in `scilla-arks.tsv`. And Scilla's registers begin in 1809, not 1816 — this archive had been planning round a wall seven years too late. AND NOW THERE IS A REASON BEYOND COMPLETENESS. Giovanni Arlotta's age is in dispute by eleven years — forty at Serafina's birth in June 1830, seventy-five at his grandson's in April 1854, both his own declarations in person. His marriage act would give the age a third time and name his parents, which is the only thing that will settle whether the Domenico Arlotta generation belongs to this family at all. Married before June 1830; bundles 1824–1827 `an_ua2002461` and 1828–1831 `an_ua2002463`. It would also test the candidate for Maria Costa's father — a Domenico Costa, *commerciante*, forty-six, who stood as a witness *produced by Giovanni Arlotta* at Serafina's registration. STARTED 20 SEPTEMBER; RATE-LIMITED PART WAY. Containers bought and saved: 1828–1831 `LD6PQKn` (1296 images) and 1824–1827 `5xYoOBO` (1416 images). A ten-point thumbnail scan of `LD6PQKn` puts 1830 at roughly images 700–1000 — wrappers headed *«Anno 1830»* at 740 and 880 — so 1829 is roughly 450–700. The volume has no year index; its last images are ordinary documents, so the fascicolo wrappers are the only finding aid, and they are found by shape (see row 147). The cheaper route is blocked, not closed. An annual marriage Tavola gives the act number, and the processetti fascicolo number follows it — two reads instead of a wrapper crawl. The ark lookups for Scilla Matrimoni 1828 `an_ua2002545`, 1829 `an_ua2002548`, 1830 `an_ua2002551` all returned 403 Forbidden and were recorded as a block, not as three absent registers. Only the containers are owed; one paced page load each once the limiter lifts. Note the alphabetisation risk: Scilla's marriage Tavola for 1818 sorts by the groom's *given* name and 1819 by surname. Read the section headers before concluding Arlotta is absent. FOUND AND READ, 20 SEPTEMBER. Scilla, Matrimoni 1828, atto 28. Civil promise 30 August 1828; banns posted Sunday 3 August; church wedding 7 September 1828. GIOVANNI ARLOTTA, twenty-eight, *marinaro*, son of the late Domenico Arlotta and of Anna Barbaro. MARIA COSTA, twenty-six, daughter of Rocco di Costa, *marinaro*, and of Pasqualina Idone — and both her parents attended and consented in person. It confirms the Domenico Arlotta generation and dates it: *«fu Domenico»*, so dead before 30 August 1828, twenty-six years tighter than the «before April 1854» this archive carried. Three new people: Anna Barbaro, Rocco Costa, Pasqualina Idone. And it breaks two ages. Giovanni is 28 here, 40 on his daughter's 1830 birth act, and 75 on his grandson's of 1854 — three of his own declarations across twenty-one years. The marriage act wins. *Trenta* written as *quaranta* explains 1830; 1854 is not explicable that way and stays open — see row 85. Maria Costa is 26 here, so born c. 1802, where this archive entered *c. 1807* a few hours earlier off the 1830 act. Corrected, with the old figure left on the record. A candidate withdrawn by the same act: the Domenico Costa who witnessed Serafina's registration in 1830 was recorded as a candidate for Maria's father. Her father was Rocco. That is what the grade was for. Method note, and it is worth as much as the act: the 1828 annual marriage Tavola names both spouses' parents — four parents per marriage, one index page, no act needed unless ages matter.
40. Done — finished and published this archive
SCILLA — who is the Giuseppe Arena of generation seven? Answered: he was never an ancestor
Answered by row 42 rather than by a register. He was on the spine at ancGen 7 for one reason only: he is Orazio Arena's father, and Orazio was believed to be Rocco's. With that edge dropped he is not an ancestor at all — `onSpine` false, no generation. He remains a real person in Michael's tree; he is simply not in this family's direct line. Asking «is he real or an artefact?» turned out to be the wrong question: he is real, and his *ancestry* was the artefact.
41. Done — finished and published this archive
SCILLA — the Pontillo above Giovanna: three more ancestors resting on a contradicted surname
Sharpened, and it is the same shape as the Orazio Arena problem just resolved. The tree gives Giovanni Polistena's mother as Giovanna Pontillo. Two independent records call her Giovanna ROMANO — Scilla Nati 1883 atto 206, and the New Orleans manifest of 1908 taken from his own statement at twenty-four, half a world away. This is the oldest resolved conflict in the archive. The edge to her is kept, because she is almost certainly the same woman under a wrong surname rather than a different person. What is not safe is what sits above her: if she is a Romano, then Giuseppi Pontillo and Nunziatta Pontillo — whom the tree gives as her parents, and who sit on the spine at generation six — are not this family's ancestors at all. Three people, resting on a surname the registers contradict. What would settle it: her own marriage or death act at Scilla, naming her parents. Until then the edge is graded `tree` with the problem written on it, rather than quietly drawn. RESOLVED 16 SEPTEMBER, and the answer explains where the wrong surname came from. A fourth record calls her ROMANO — her husband's own death act of 1924, «Coniuge Giovanna Romano» — and a fifth, their daughter Santa's of 1920, «Madre Giovanna Romano». And the same act names the husband's mother as ANNA PONTILLO. So the export took Camillo Antonio's parents and hung them on his wife: the tree's «Giuseppi Pontillo» is his father GIUSEPPE POLISTENA, and the Pontillo is his mother ANNA. That is why no Pontillo family could ever be found behind her — there is not one. Two asserted ancestors dissolve and two documented ones replace them. AND HER OWN ACT CAME IN LATER THE SAME DAY, which ends it. «GIOVANNA ROMANO, died Scilla 15 July 1930, Padre ANNUNZIATO, Madre ANGELA ZIRILLI, Coniuge Camillo Antonio Polistena.» Her own surname, her own parents and her own husband on one document. Five records now call her Romano and this is the first to name her parents — so the three ancestors this row said were resting on a contradicted surname are resolved: two are withdrawn and two documented ones replace them.
42. Done — finished and published this archive
SCILLA — Orazio Arena and Giovanna Zirilli: the edge is dropped
Done, and the site has stopped arguing with itself. It was proving on /corrections/ that Rocco Arena is not Orazio's son while still drawing that parentage on every chart. The edge is now dropped in `parent-links.tsv`, with the whole case written on the row. The case, five independent reads: Rocco's own birth act (Nati 1861/228) and his own marriage act (Matrimoni 1888/63) both name Francesco Antonio Arena and Serafina Arlotta; act 45 of 1860 gives that couple a son Giuseppe, and the tree gives Orazio exactly two children — Rocco and Giuseppe — so the tree took Francesco Antonio's household whole; Orazio and Giovanna's real children are documented across 1853–1860 and include neither; and the obvious namesake, the Rocco of 1858, was tested and excluded. Three people left the blood line with it — Orazio, Giovanna and Orazio's father Giuseppe Arena. 62 ancestors became 59. Orazio and Giovanna stay on the site as the real couple this archive found; they were simply not his parents. Rocco now has no parents on the chart, because his real ones are not in the export. That is row 51's problem, exactly.
43. Next — ready to start this archive
BRIATICO — the Potenzoni generation eight: Anile, Barbuto, Lesto, Grasso
Phase 3. Four parentless spine people in one place — Fortunato Anile × Antonia Grasso, Giuseppe Barbuto × Saveria Lesto, all of Potenzoni and Mandaradoni, hamlets of Briatico with their own registers. Antonia Grasso is dated 1809, so her birth is inside civil registration. This is the densest untouched cluster on the spine.
44. Next — ready to start this archive
BRIATICO — the Melluso: Fillipo b.1855 × Rosa Vitrale, and Caterina b.1860
Phase 3. Both dated and both parentless, and both late enough that their births and marriages are squarely inside the registers. Caterina Melluso married Alfonso Prostamo and died in 1957. The Melluso are one of the four surnames that returned nothing in Brisbane's council cemeteries, so their story is still entirely Calabrian.
45. Next — ready to start this archive
BRIATICO — settle Caterina Di Meio against Caterina Majo
Phase 3, and a live contradiction. The export gives Pasquale Prostamo's wife as Caterina Di Meio, born 1762. The Pizzo register of 1829 calls her Caterina Majo — and this archive has already corrected its own reading of that name once, from «Muso». Two spellings, one woman, and she is generation eight. Her marriage or death act would end it.
46. Next — ready to start this archive
BRIATICO — what would it actually take to prove the 1697 Prostamo?
Phase 3, and a decision rather than a search. Franco Antonino Prostamo and Agnere Polito are dated 1697 and 1699 and sit at generation ten with nothing behind them. Civil registration in Calabria begins in 1809, so they are a century pre-civil and only parish and notarial records could reach them. Write down what the route would be and what it would cost, so the tree's oldest claim stops being quietly carried.
47. Next — ready to start this archive
PIEDIMONTE ETNEO — Mariano Mazza, and the Nicotra couple behind Maria
Phase 3. Rosario Mazza b.1850 is parentless on the spine and his father is MARIANO MAZZA, known only from his son's 1873 birth act. And John Nicotra × Venera Cassaneti (g5) are parentless too, known from a Queensland death registration rather than any Italian act. The Sicilian quarter is the thinnest of the four. GENERATION SIX IS NO LONGER JUST A PAIR OF NAMES. The 1871 publication certificate gives Mariano Mazza and Maria Catalano their trades (bracciante, industriosa), their residence (Piedimonte) and a dated proof of life (16 April 1871), and does the same for Gaetano Bonanno and Anna Cavallaro. And the Nicotra half has a route it did not have: the Piedimonte MARRIAGE index names both spouses, one page-pair per year — 1871 is read and holds no Nicotra. Sweeping 1866–1885 is about twenty crops. Row 161. THE NICOTRA HALF IS OPEN. Matrimoni 1872 atto 20 gives Giovanni Nicotra's parents (Rosario Nicotra × Venera Murabito, of Mascali, both living) and Venera Cassaniti's (Gaetano Cassaniti × Maria Currenti, of Piedimonte, both dead by 1872). Generation six now exists on both Sicilian lines — Mazza/Catalano and Bonanno/Cavallaro from the 1871 certificate, Nicotra/Murabito and Cassaniti/Currenti from the 1872 act. What is left on this side is MASCALI, not Piedimonte. Row 173.
48. Done — finished and published this archive
Sweep Queensland BDM for every family surname — done, and mostly empty for a reason
Done: Donato, Mazza, Crupi, Prostamo, Polistena, Melluso, Nicotra, Anile, Barbuto, Arena. The one real find was the Crupi — three brothers of Angela nobody here knew about. The emptiness elsewhere is informative rather than disappointing. Polistena and Anile return *nothing at all*; Prostamo returns exactly one record for a family with a dozen people in Brisbane. That is the shape of the index, not of the family: births stop in 1926, marriages in 1951, deaths in 1996, and these people arrived in 1949 and mostly died after 1996. Queensland BDM is the wrong instrument for a post-war migration and this archive now knows the size of that hole. Arena returns seventy-five and not one is ours — Byrne, Kerby, Pollard, Timothy Leo: an Irish-Australian family entirely unconnected to Scilla. A surname sweep without checking parents would have produced a dozen false joins. And a correction to my own expectation: the Melluso and Prostamo registrations I took for finds were already in `qld-bdm.tsv`, recorded by an earlier session with better analysis than mine. What was genuinely missing is that Gennaro Melluso, Maria Limardo and Carmela Maria Borello had never become people. They are now — which is row 51's lesson in miniature: a fact in one file does not reach the reader until it is surfaced in another.
49. Done — finished and published this archive
AUSTRALIA — the electoral rolls, which nobody here has touched
Phase 4. Australian electoral rolls are public, near-annual, and give address and occupation for every adult. For a family that arrived between 1925 and 1957 they would place every household in Brisbane year by year, and they are the only source that tracks people who neither died nor married in the window. Entirely unused here. See row 62: the rolls are not free. Ancestry or the NLA, and that is David's call. UNBLOCKED 16 SEPTEMBER: David has taken a full month's Findmypast subscription, and Findmypast carries Australian electoral rolls. See row 99. CLOSED 16 SEPTEMBER by row 99. Phase 4 is done as far as Findmypast can take it: one Queensland roll year, 1959, plus 1949. Everything after 1959 is row 104.
50. Done — finished and published this archive
AUSTRALIA — the WWII internment and naturalisation files at the NAA
Phase 4. Italians in Queensland were interned from 1940, and Trove has already shown Brisbane Prostamo naturalisations. The NAA holds both series and its index is searchable. Michele Mazza died in 1926 and the rest arrived later, so the internment years cover the generation that settled. OPENED 16 SEPTEMBER, and the internment half is no longer hypothetical. NAA: MP1103/1, Q8110 — «Prisoner of War/Internee: Mazza, Rosario; Date of birth - 20 February 1899; Nationality - Italian», 1942–1944. Michael Mazza's great-uncle was interned as an enemy alien. Behind it sit BP242/1 Q21227, his Queensland investigation case file (1940–1948), and J1737 MAZZA R, a Civil Aliens Corps record of 1944 — which is what happened to interned Italians after release: they were put to labouring work under guard. All three are Open. All three are paper. This is a whole chapter of one man's life that the tree records as nothing at all, and it connects directly to row 7, the Trove search for the 1942 Brisbane internments. Still to do: the same sweep for Crupi, Arena and Polistena, and the A367/D1901/MP14 items that may be the same man under other agencies. CLOSED 16 SEPTEMBER. Two men in this family were interned, and one of the cards is digitised. MARIO CRUPI — Angela Crupi's brother — NAA: D4028, CRUPI MARIO. Born 28 November 1905 at Piedimonte, province of Catania; cane cutter at Ingham; entered Australia 13 June 1928 at Brisbane; five feet eight, brown eyes, black hair, *scar on left thigh, wart on right side of face*. The reverse reads: «18.8.42 LOVEDAY INTERNMENT CAMP. 30/12/43 Released and entrained for Brisbane.» Sixteen months. And the birthplace is a structural finding: Piedimonte Etneo is the Mazza's own town, so the Crupi and the Mazza came from the same village — which is how the marriage of Rosario Mazza and Angela Crupi was made, and which this archive did not know. Also found: NAA J25 1971/3900, «Arena, Rocco; wife Angelina [or Angela] [nee Polistena]», so *both* pairs of Michael's grandparents have a Queensland immigration file. Both need the access request in row 115. What is left is row 117 — the naturalisation half for the Arena and Polistena, and the other Crupi cards.
51. Done — finished and published this archive
Let the charts draw people the export does not contain — half done
The reader-facing half is done. `tools/build_documented_line.py` walks the additions' own father/mother names up from every row that carries an `inTree` slug, and `/lines` now publishes the result: thirteen ancestors no chart on this site can draw, each with the relation, the dates, the trade and the citation, under the export person where the drawing stops. The deepest is three generations above Giovanni Arena — Francesco Antonio Arena, then Filippo Arena and Domenica Costa, and on the mother's side Serafina Arlotta, Giovanni Arlotta and Domenico Arlotta. `/bloodline/` shows none of them. What is NOT done, and cannot be done here: the charts themselves still cannot draw these people, because `Bloodline.astro` and `lib/bloodline.js` are the shared kit's and this archive leaves the kit alone. The gap is now visible and cited rather than silent, which is most of the value — but drawing it needs the estate. See row 63. DONE, 20 SEPTEMBER, AND THE BLOCKER WAS A MISREADING OF THE KIT. This row said the charts could not draw these people because `Bloodline.astro` and `lib/bloodline.js` belong to the shared kit. True of *those* two. `/tree/` does not use them. It uses `AncChart` with the kit's `lineage()`, and `lineage()` takes `parentsOf(row)` from the caller — with a comment in `kit/lib/ahnen.js` that could not be plainer: *«returning a row that is not in the file is allowed and is the point: a parent known only as a name on a register is still a real thing the chart should draw, and refusing it would quietly shorten the line.»* The kit had been waiting to be asked. `tools/build_documented_parents.py` emits the map and `/tree/`'s own `parentsOf` falls back to it only where the export gives a person no parents at all — never over the top of the tree, because the chart is a drawing of the tree and quietly redrawing it would be lying about the source. The drawn people are graded documented, and unlinked, because they have no person page to link to. Nothing in the kit changed and nothing was forked.
52. Running — a machine or a person is at it now this archive
STRUCTURE — raise the graded fraction of parent links
Phase 5, and it is measurable. Of 748 parent edges, 727 are ungraded — drawn dashed because the tree merely asserts them. Only 11 are `read` and 10 `index`. Every act this archive has read should have upgraded an edge and most have not. Work back through `register-reads.tsv` and `scilla-reads.tsv` and grade what is already proved. STARTED. Six spine edges graded from acts already read: Nunziata Donato to Michele Donato and Maria Sofi (Nati 1892/162), Giovanni Polistena to Antonino Polistena (Nati 1883/206), and Antonino Prostamo to Pasquale Prostamo (the *avo paterno* certificate in the 1849 Briatico bundle). `read` goes from 11 to 15. Forty-two spine edges remain ungraded, and most of them cannot be graded because no act has been read for them — which makes this row a measure of research done, not of tidying. 27 NOW, AND THE CEILING IS VISIBLE, 21 SEPTEMBER. Row 53's checker was built and immediately answered the question this row could not: of 741 ungraded edges, only four belong to a person whose act has been read, and 714 are ungraded because no act exists for the child — which is honest, not untidy. Two of the four are now graded (Antonio Arena to Rocco Arena and to Generosa Costa, from the New York birth index), one must NOT be graded because row 45 is arguing about it, and one is a read act that says nothing about the parent. So this row is not a tidying job with 700 items in it. It is a research measure, and it will move when acts are read, not when links are edited.
53. Done — finished and published this archive
STRUCTURE — a checker for the gap between an act read and an edge graded
Phase 5. The natural follow-on from row 52, and the way to stop it recurring: if a read act names a parent and both people are in the export, the edge should carry a grade. A checker can list the ones that do not. Same shape as `check_additions.py`, which found six published falsehoods the hour it existed. BUILT, 21 SEPTEMBER. `tools/check_grades.py`, in the build as `check:grades`, advisory by default and `--strict` for a day when the list should stay at zero. It lists every edge the chart draws dashed for a person whose act this archive has read — and it grades nothing, because only someone who has read the act knows whether it names the parent. Two false-positive classes had to be cut before it was worth anything. It matched the word *register* inside *«Brisbane Catholic Cemeteries register»* and reported twenty burials as gradeable; and it reported Ellis Island manifests. A burial plot names who lies beside whom; it does not name a father. It also now skips `disputed-edges.tsv`, because an edge this archive is deliberately arguing against is not a gap. Four real candidates out of 741 ungraded edges, and one live error. The export's mariano-mazza — a twentieth-century grandson of Rosario — was still carrying two corrections keyed to his slug that describe his great-grandfather: *«born about 1809»*, *«BRACCIALE in 1845»*. Yesterday's fix removed the record and left these. Detached, with the reason written on the row.
54. Done — finished and published this archive
Review the outstanding searches — done, and every one now has a row against it
Done, and it was measurable because another session had already built the instrument. `check:covers` came into the build from the kit and reported «work list accounts for 0 of 24 outstanding items» — a gate that counts what a list does NOT say, which is the one thing `checkworklist` could never see. One of the twenty-four was simply finished: «Queensland BDM — beyond the six registrations» was closed by today's sweep and is now marked done with what it found. The other twenty-three are now covered by fourteen work-list rows carrying explicit `covers` keys, and `unreconciled` is down from 24 to 0. The build now reads *«accounts for 23 of 23»*. That number cannot silently rot: a covers entry pointing at a key that no longer exists fails the build in the other direction.
55. Next — ready to start the kit & estate
ESTATE — reconcile the people this archive shares with Falco and D'Arcy
Phase 5, and it is not a Mazza problem alone. The spine runs straight into two sibling archives: Josephine Falco and Ian D'Arcy sit here at generation two with their daughters Stefanie and Cristina. The same people are recorded in three archives, each from its own export, with no mechanism to notice when they disagree. `checkshared.py` compares components; nothing compares people. Propose one.
56. Blocked — waiting on a person or an institution David
MICHAEL — confirm the recent deaths this archive found in council records
Phase 6. Brisbane City Council's register gives Domenico Francesco Prostamo, died 10 January 2025, and Maria Assunta Prostamo, died 30 October 1991, in adjacent plots at Mount Gravatt. The 2025 death is eight months old and the family will know it first-hand. Worth confirming with Michael rather than publishing from a cemetery index alone — and worth asking gently.
57. Blocked — waiting on a person or an institution David
MICHAEL — the four questions only a living relative can answer
Phase 6. Beyond the photographs in row 24, there are four things no record will settle: whether the Brisbane Donato of Pinnaroo carrying the name Candeloro are kin; whether Mario and Maria Mazza of Mount Gravatt are; what the family calls the Buenos Aires Prostamo; and which of the 58 shared names are one person. A short list of specific questions, with the evidence beside each, is worth more than another month of registers. A fifth question now: who was Venero Crupi, born 1935 and buried at Nudgee in December 2023 — the only Crupi in either Brisbane register, and Angela Crupi had three brothers here nobody in this archive knew about until today.
58. Blocked — waiting on a person or an institution this archive
Francesco Cichello and Vittoria Mollo — who were they?
Blocked behind the gap, and now demonstrably so. Both were dead before March 1824 and neither appears in the death Tavole of 1822, 1823 or 1824 — so they died in the nine years the register does not hold. They come back into reach only through row 59. Note also that neither surname is rare at Zambrone: a Maria Mollo died in 1822, a Marianna Mollo and an Anna Cichello in 1823. No join on one of these names alone.
59. Blocked — waiting on a person or an institution David
The second copy of Zambrone's 1813–1821 deaths — and it is probably at the comune, not on film
Reframed, and the reframing changes who can do it. Under the Napoleonic and Bourbon codes every act was written twice: one register kept by the comune, one sent to the tribunale, and the tribunale's copies are what ended up in the Archivio di Stato. Antenati digitises the Archivio di Stato. This volume's own manifest says so — «Archivio di Stato di Catanzaro › Stato civile napoleonico › Zambrone › 1811-1839». So the nine missing years are missing from the court's copy, and the copy that may still survive is the Comune di Zambrone's own, in its Ufficio di Stato Civile. That is corroborated by the fact that the act existed in 1849, when the chancery certified a copy of Domenico Ventrice's death for a marriage at Briatico. Something was still being copied from, thirty-four years after he died. So this is a letter, not a search — an enquiry to the Comune di Zambrone, and it is the only route to four people: Domenico Ventrice, Anna Maropoli, Francesco Cichello and Vittoria Mollo. *Noted while checking: FamilySearch's catalogue was erroring out («Something Went Wrong») on 15 September, so the film holdings for Zambrone have NOT been ruled out and should be re-checked when it is back.*
60. Done — finished and published this archive
Sweep the Zambrone death Tavole for the family surnames — 1822–24 done, and negative
Done for the years that matter, and negative. The Tavole of 1822 (image 51), 1823 (65) and 1824 (78) name everyone who died at Zambrone in those years with their parents, and neither Francesco Cichello nor Vittoria Mollo is among them — nor Anna Maropoli. All three were already dead by March 1824, so all three almost certainly died inside the missing 1813–1821, behind the same gap as Domenico Ventrice. Four wanted people, one gap. What it did settle: Maropoli appears twice more in these three years, including an infant Giuseppe Maropoli dead 23 March 1824. The surname is unmistakably local. 1825–1839 remain unread but no longer urgent.
61. Next — ready to start this archive
Re-check FamilySearch's catalogue for Zambrone when it is working
Small and specific. The catalogue UI returned «Something Went Wrong» for every query on 15 September, and its JSON endpoint accepts `q.place` but ignores it, returning two million hits. So the film holdings for Zambrone have not actually been ruled out — only failed to be checked. Retry, and look specifically for a Zambrone or Monteleone *stato civile* film covering 1813–1821, before concluding that row 59's letter is the only route.
62. Done — finished and published this archive
Electoral rolls are now the right instrument, not BDM
Promoted out of row 49 because today's sweep proved the case. Queensland BDM cannot see this family: they arrived in 1949, married after the marriage index closes in 1951, and mostly died after the death index closes in 1996. Polistena returns zero; Prostamo returns one. Electoral rolls have none of those limits. They are near-annual, they run through the whole post-war period, they give address and occupation, and they capture people who neither died nor married inside an index window — which is precisely this family's problem. For a migration that lands in 1949, the rolls are the primary source and the registrar's index is the secondary one. Start with Brisbane 1949–1980 for Prostamo, Polistena, Mazza and Arena. TRIED AND BLOCKED ON A SOURCE. FamilySearch has no Australian electoral rolls for these surnames — its Australian holdings here are Victorian Petty Sessions and Find a Grave. The rolls are on Ancestry, which is paid, and at the National Library of Australia. So this needs either a subscription decision from David or an NLA route, and it is blocked until one of those exists rather than being quietly open. UNBLOCKED 16 SEPTEMBER: David has taken a full month's Findmypast subscription, and Findmypast carries Australian electoral rolls. See row 99. CLOSED 16 SEPTEMBER, and the conclusion needs qualifying. The rolls were the right instrument and they did deliver — Sebastiano Mazza at Wacol, Rosario Mazza at Leichhardt Street, both with trades. What was wrong was the assumption that Findmypast holds them near-annually. For Queensland it holds 1903, 1906, 1913, 1922, 1934, 1941, 1949 and 1959 and nothing after. The near-annual run is real but it is at the NLA and on Ancestry. See row 104.
63. Struck — founded on a mistake, kept rather than deleted the kit & estate
Ask the estate whether the kit's chart can take a second source
Raised by row 51, and it is not a Mazza-only problem. `Bloodline.astro` and `lib/bloodline.js` draw one array of people — the archive's export. Any archive that also keeps documented people the export lacks can therefore prove an ancestor and not draw them. Here that is fifty-four people and thirteen direct ancestors, including three generations above Giovanni Arena. The ask is small and probably general: let `graph()` accept a second, clearly-styled set of nodes — drawn differently, the way a `tree` edge is already drawn dashed — so an archive can show the difference between what the family remembers and what the records prove. Worth asking before anyone builds a private fork of the chart, which is exactly how this estate got two Pedigrees 7.6KB apart. WITHDRAWN, 20 SEPTEMBER. The ask was already answered before it was written. `lineage()` has always taken `parentsOf()` from the caller and its own comment invites exactly this. The Mazza archive now draws its documented ancestors on `/tree/` without a line of kit change (row 51). What may still be worth raising is narrower: `Bloodline.astro` and `lib/bloodline.js` take one array and have no such hook, so `/bloodline/` still cannot show them. That is a real question for the estate; *«can the chart take a second source»* was not, and this row asked the wrong one.
64. Next — ready to start this archive
The processetti: the 1885/1888 bundles do not exist, and the 1851 one does
Phase 7. The method that has produced this archive's two biggest finds, pointed at a generation it has never touched. Processetti carry the birth acts of both spouses and the death acts of any dead parent. Briatico 1849 broke the Calabrian line open; Scilla 1844 fascicolo 20 gave Domenico Donato's birth and Caterina Cichello's parents. Four family marriages sit in one narrow window and none of their bundles has been opened: Giuseppe Arena × Maria Martello (Matrimoni 1885 atto 73), Giovanni Arena × Concetta Russo (1885 atto 86), Rocco Arena × Generosa Costa (1888 atto 63), and Michele Donato × Maria Sofi. Their papers would give birth extracts for eight people and death acts for whichever parents had already died — including, possibly, Filippo Arena and Domenica Costa, whose marriage the registers cannot reach but whose deaths they can. CHECKED 20 SEPTEMBER, AND THE ROW IS HALF WRONG AND HALF BETTER. Listing Scilla's whole holding — 466 registers — shows twenty-two volumes of marriage processetti running 1810 to 1865 and then stopping. The Italian-era fondo, all 267 registers of it, holds no processetti at all. So the bundles behind Giuseppe Arena × Maria Martello (1885), Giovanni Arena × Concetta Russo (1885) and Rocco Arena × Generosa Costa (1888) are not on Antenati and this half of the row is closed, not pending. But the marriage that matters most is inside the window. Francesco Antonio Arena × Serafina Arlotta, 7 February 1851, atto 13 — the bundle is the volume 1849–1852, `an_ua2002487`, now in `scilla-arks.tsv`. That fascicolo should carry both spouses' birth extracts and the death acts of whichever of their four parents had died by February 1851 — Filippo Arena, Domenica Costa, Giovanni Arlotta, Maria Costa — and the Scilla bundle of 1842–1845 carried a dead grandparent as well. One volume, three rows: 68 Filippo Arena and Domenica Costa's deaths, 39 Giovanni Arlotta × Maria Costa's marriage, 85 Domenico Arlotta. This is now the cheapest unopened thing in the archive and it is row 147.
65. Running — a machine or a person is at it now this archive
SAN COSTANTINO CALABRO — mapped, and it may walk round the Zambrone gap
Mapped for the first time, and it is bigger than expected: 205 volumes, 1810 to 1912, including Nati 1812–24, Morti 1813–27, Matrimoni 1813–32 and forty-eight processetti volumes. The strategic point: Zambrone's deaths are missing 1813–1821, and four wanted people died inside that gap. San Costantino's Morti 1813–27 covers exactly those years. If Anna Maropoli — Giuseppe Ventrice's mother, already «la fu» in March 1824 — died in her husband's own town rather than at Zambrone, her act survives. The gap may be walkable *round* rather than through. Not yet swept: the 1813–27 death volume is 325 images and is not one comune — image 12 is San Costantino, images 24 and 40 are Francica. The comune line must be read on every act. Giuseppe's own birth, about 1790, stays out of reach.
66. Running — a machine or a person is at it now this archive
PIZZO 1823–1860 — Maria Fragalà's death, and the Prostamo who stayed
Phase 7, and it is reachable where the rest of the Prostamo line is not. Pizzo's registers begin in 1823, and Maria Fragalà was alive at Pizzo in June 1849 when her son married at Briatico. Her death is therefore *inside* the surviving run — one of the few Prostamo events on this line that is. The same volumes would hold any other Prostamo who did not move to Briatico. The 1837 cholera register is not digitised, but 1823–1836 and 1838–1860 are a different question and have never been opened. Series listed: 751 volumes, earliest 1823 — the wall confirmed from the series itself. Reachable: Nati and Morti from 1823, Matrimoni 1823–29, and thirty-four processetti volumes. Maria Fragalà's death is the target and it falls inside the surviving run. Antonino Prostamo's marriage of about 1810–13 does not. THE NAME INDEX DOES NOT REACH THEM, 21 SEPTEMBER, AND THAT IS ITSELF ODD. Pizzo is one of only three comuni this family touches that Antenati has indexed by name — births, 1868–1903. *Prostamo* at Pizzo returns nothing at all, control-tested against *Arena at Scilla* (69) in the same breath, so it is a real nil and not the limiter. Either the Prostamo had left Pizzo by 1868, or the indexing does not reach them — and the first is checkable against what this archive already holds, since Antonino Prostamo's line is at Briatico by the 1849 marriage bundle. Recorded so the index is not tried again for this surname.
67. Next — ready to start this archive
THE HAMLETS — Potenzoni, Sciconi, Conidoni and Daffinà as series in their own right
Phase 7, and it follows from a trap already recorded. Briatico and Zambrone searches keep returning arks that turn out to be Daffinà, Sciconi or Conidoni — separate comuni in the same circondario with their own registers. That has been a nuisance three times. It is also an opportunity: the generation-eight Briatico cluster — Fortunato Anile × Antonia Grasso, Giuseppe Barbuto × Saveria Lesto — is recorded at Potenzoni and Mandaradoni, not Briatico proper. Map these four series properly and the densest untouched part of the spine becomes searchable.
68. Next — ready to start this archive
Filippo Arena and Domenica Costa — alive in February 1851, and the window is 1851–1888
Phase 7, and the consolation prize for row 38. Their *marriage* is pre-civil and gone. Their deaths are not. Both were born in the 1790s, both were alive when their son married in 1851, and Scilla's death registers run from 1816 — so they died inside a surviving run, most likely in the 1850s–1870s. A death act gives an age, a trade, an address and often the parents. It is the only remaining route to a generation above Filippo. UNBLOCKED 16 SEPTEMBER. The pre-1866 Scilla death series is now listed — 51 volumes, 1817–1865 — in `scilla-arks.tsv` (row 130). This row can start. A CHEAPER ROUTE APPEARED, 20 SEPTEMBER. Sweeping the annual death indexes for them means 1862–1888 — twenty-seven years, and no container is saved for any of them, so twenty-seven rate-limited page loads before a single image. Instead: their son's marriage processetti, Scilla 1849–1852, `an_ua2002487`. If either was dead by 7 February 1851 the bundle holds the death act itself. If both were alive it says so, and the sweep starts in 1851 instead of 1862. Either way one volume replaces twenty-seven. Row 147. NARROWED FROM BOTH ENDS, 20 SEPTEMBER. The 1851 processetti give them ages: Filippo Arena twenty-five in November 1819 (b. c. 1794) and Domenica Costa twenty-seven (b. c. 1792). The same bundle contains no death act for either, where a neighbouring fascicolo binds them in when a parent is dead — so both were very likely alive on 7 February 1851. Filippo is called *«di Filippo»*, not *«fu»*, in act 228 of December 1861, so he was alive then too; both are *«fu»* in Francesco Antonio's death act of 1901. The window is 1862–1888, since the A blocks of 1889–1908 hold neither. Twenty-seven years, no containers saved. A cheaper attack first: their sons' and daughters' marriage processetti stop in 1865, so a family marriage in 1862–65 would carry their death acts if they had died — and the volume 1862–1865, `an_ua2002507` is already in `scilla-arks.tsv`. CONFIRMED ALIVE, 20 SEPTEMBER, AND NOT BY AN ABSENCE. The 1851 annual marriage Tavola names both spouses' parents and marks the dead with «fu» — the 1828 Tavola writes Giovanni Arlotta's father as *«fu Dom.co»* beside his living mother, and *«fu»* appears on the same 1851 page in other entries. Entry 1 is «Arena Francesco e Arlotta Serafina», and its parents are «Filippo, e Costa Domenica» and «Giovanni e Costa Maria» — neither with «fu». So the window closes from the left. Both were alive in February 1851; Filippo is still *«di Filippo»*, not *«fu»*, in act 228 of December 1861; both are *«fu»* in Francesco Antonio's death act of 1901; and the A blocks of 1889–1908 hold neither. They died between December 1861 and 1888 — twenty-seven years, unindexed. A cheaper instrument exists, but it does NOT reach them the way I first wrote. Every annual marriage Tavola in this series dates the life status of four parents in one opening — but a marriage act names the spouses' own parents, not their grandparents. The Arena weddings of 1885 and 1888 are Francesco Antonio's *sons*, so those Tavole date Francesco Antonio and Serafina Arlotta — already known alive in 1888 — and say nothing whatever about Filippo and Domenica. *Corrected the same day it was written.* What would actually work: find a SIBLING of Francesco Antonio — another child of Filippo Arena and Domenica Costa, born at Scilla in the 1810s–1830s — and read the Tavola of the year that sibling married. This archive has never looked for one. Domenica Costa was twenty-seven when Francesco was born in 1819, so there are very likely elder children, and the annual birth Tavole for 1813–1818 are now reachable: Scilla's registers begin in 1809, not 1816 (row 17). Otherwise it is the death indexes, 1862–1888 — twenty-seven years, an A block apiece, and no containers saved for any of them. THE PREMISE OF THIS ROW IS HALF GONE, 20 SEPTEMBER. It was written as the *only* route to a generation above Filippo. It is no longer: their marriage act was found — Scilla, Matrimoni 1815 atto 29 — and it names all four of their parents (row 151). The row still stands, for what it was always also for: a death act gives an age, an address and a trade, and it would settle the five-year disagreement about Domenica Costa's birth — twenty-eight at her wedding in 1815 (b. c. 1787), twenty-seven at her son's birth in 1819 (b. c. 1792). It is no longer urgent, and it drops behind the cheaper work the 1815 act opened up.
69. Done — finished and published this archive
SCILLA DEATHS — the Arena/Arlotta search past 1896: answered
Phase 7, and it is finished. The A block of Scilla's annual death index was read for 1889–1896 without them, then the Antenati name index showed neither of them as the deceased between 1910 and 1943, which narrowed forty-seven unread years to thirteen. The twelve readable years of that window have now been swept — row 128 — and both deaths are found: Francesco Antonio Arena, 29 August 1901 (*Morti 1901, atto 96*) and Serafina Arlotta, 24 April 1908 (*Morti 1908, atto 55*), both at Via Grotte number two. A negative had to be narrowed on the way. A searched row claimed 1889–1902 clear of both; he died in 1901, inside that span. The row now claims 1889–1896, which is what it can stand behind. A negative is only worth the pages actually turned.
70. Done — finished and published this archive
SCILLA — when and where did Michele Donato die?
Phase 7. He is alive and fifty-two on the New Orleans manifest of March 1908, sailing for Quincy, Florida, and the 1910 US census shows he did not stay — his wife and household remained at Via Grotte throughout. So he came back, and he has a death somewhere. Scilla's death registers, or a second crossing. He is the last man on the Donato line whose end is unknown. CLOSED 16 SEPTEMBER. Michele Donato died at Scilla on 10 May 1936 — «Padre Domenico, Madre Nunziata Sofi, Coniuge Maria Sofi», all three matching this archive exactly. He was fifty-two on the New Orleans manifest of March 1908 and lived another twenty-eight years after coming home. The Donato line now has no man whose end is unknown.
71. Next — ready to start this archive
SCILLA — Nunziata Donato × Giovanni Polistena's marriage, about 1916
Phase 7, and it is the act that joins two quarters. They sailed on the same ship in February 1908 — his line immediately after her father's on the Liguria manifest — and married about eight years later. That act would name four parents and tie the Donato quarter to the Polistena quarter with a document instead of an inference. It would also, if his mother is named, settle whether she was a Romano or a Pontillo — which is row 41 and three ancestors. AND HE MAY HAVE FOLLOWED THE FAMILY TO QUEENSLAND. NAA: BP9/3, ITALIAN POLISTENA G — *«Polistena, Giovanni born 1883 — arrived in Brisbane on the ORMONDE, 21 February 1928»*. This archive's Giovanni Polistena was born at Scilla in 1883, act 206; the year matches exactly and Brisbane is the right port. Graded candidate, because the entry carries no parents and Polistena is a comune as well as a surname. If it is him, he crossed to New Orleans in 1908 and to Brisbane in 1928 — and the marriage this row wants sits between them.
72. Next — ready to start this archive
SCILLA — Maria Sofi's parents, and the second Sofi marriage
Phase 7. Michele Donato married a Sofi, as his father had. Maria Sofi sits on the spine at generation five with no parents at all. Given that Nunziata Sofi's parents were found in the 1844 processetti, hers should be findable the same way — their marriage act, or its bundle. Two Donato men marrying two Sofi women in successive generations is either a close family tie or a small town, and the acts would say which. TRIED AND FAILED BY INDEX, 16 September. The Scilla name index holds four women called Maria Sofi and not one is named as the wife of a Michele Donato; the one who died on 3 August 1913 carries no relations at all. Her parents are not reachable this way and the row stays a volume read — which row 121 predicted, since the index is deaths only and her parents would be named in a *marriage* act.
73. Done — finished and published this archive
STRUCTURE — extend the documented line DOWNWARD as well as up
Phase 8. `build_documented_line.py` walks *up* from every addition anchored to the export, and found thirteen ancestors no chart can draw. It does not yet walk down. Maria Donato, Fortunato Arena, Francesco Giovanni Rocco Arena and Candeloro Donato (b. 1847) are documented children of export people, and they are just as invisible. Same tool, opposite direction, and it would show the households whole rather than only the lines. Done. The tool walks both ways now: 15 documented ancestors above export people and 4 documented children below them, all published on /lines.
74. Done — finished and published this archive
STRUCTURE — a page for the four walls, now that all four are established
Done. `data/walls.tsv` and a section on /lines: four quarters, four walls, four different causes, each with the evidence that established it and the route that remains. Filippo Arena — four complete years of Scilla marriages read, eighty acts, to prove an absence. Candeloro Donato — a province probed year by year to find registration starting in 1820. Pasquale Prostamo — Pizzo begins 1823. Domenico Ventrice — a volume opened to find nine years physically missing between two images. The page says the thing that matters: a wall established by reading is a finding; a wall assumed from silence is a shrug, and the difference is most of what this archive is for.
75. Done — finished and published this archive
STRUCTURE — check_links for the graves and the ships
Phase 8. `check_additions` and `check_corrections` each caught real published errors the hour they existed. `data/australia-graves.tsv` and `data/naa-ships.tsv` are still unvalidated — no slug check, no date sanity, nothing to notice a burial dated before a birth or a grave hung on a person who died elsewhere. Today's Find a Grave double-index, the same woman in Victoria and Ohio, is exactly the shape a checker would catch. Done. `tools/check_places.py` is in the build. It refuses a burial before a birth, a crossing that lands before the traveller was born or after they died, a grave row with no ground, and one person buried in two different cemeteries — the exact Find a Grave failure found today. Tested by planting both faults. And it caught me first: reading `who` as a slug in `naa-ships.tsv` produced twelve confident false failures, because column 1 is the slug and column 2 is the display name. Fixed in the checker, not the data.
76. Done — finished and published this archive
Retire notes/QUEUE.md into the work list — done
Done. QUEUE.md now opens «This file is history, not a to-do list» and says why: four items were picked up from it in one day and three were already finished. Everything still live is migrated — the Piedimonte Etneo threads as rows 89–96, the rest already covered by existing rows. It is kept rather than deleted, because deleting the record of what was once believed is precisely what this archive tells other people not to do. The difference that matters is the gate: `check:worklist` refuses a row with no owner, and `check:covers` refuses an outstanding search with no row against it. The queue had neither and drifted in five days.
77. Done — finished and published this archive
STRUCTURE — give the Crupi somewhere to live on the site
Phase 8. Angela Crupi's three brothers — Pietro, Antonino and Mario — went in as documented additions today, and with their parents that is six Crupi. They have no page, no section, and no place in the four-quarters story, because they arrived through a marriage rather than a descent. A short section, probably on /brisbane/, saying who they were and how they reached the family. DONE, 21 SEPTEMBER. A section on /brisbane/, which is where they actually are, built from `additions.json` so it cannot drift from the data. It carries what the family tree does not: that *«Angela Cupo»* is Angela Crupi, born 20 December 1900; that she had three brothers in Queensland and nobody knew; that Pietro and Antonino came into Brisbane together on the Otranto on 19 August 1930 as consecutive entries in the aliens register, and Angela followed on the Viminale in 1937 with two daughters; and that Antonino was interned at Loveday, a boot repairer at Ingham, born at Piedimonte Etneo — the Mazza's own town. And it closes on the thing worth carrying away: Pietro's missing aliens card looked like a gap and is not one — *«Alien Registration Certificate NOT ISSUED»*, because he was naturalised in 1932. An explained absence and an unexplained one look identical until somebody checks.
78. Done — finished and published this archive
Find the Crupi brothers' graves — not in either Brisbane register
Searched both registers, wildcards included, and they are in neither. Brisbane City Council returns zero Crupi across its nine cemeteries. The Catholic Archdiocese returns exactly one, and it is none of the three: Venero Crupi, born 14 October 1935, died 4 December 2023, Nudgee plot LA-3299. So Pietro, Antonino and Mario are buried somewhere neither register covers — and their death registrations carry no country code, so they died in the Brisbane district, which makes the absence stranger rather than less. Venero is left as a lead and nothing more. Venero is a Catania name and Angela Crupi was born at Catania in 1900, so he fits the family's origin exactly — and fitting an origin is not a descent. He died in December 2023, well within living memory. That makes him a question for Michael rather than for an index, and it belongs with row 57.
79. Running — a machine or a person is at it now this archive
AUSTRALIA — Trove swept to the end of what it will serve
Unblocked 27 September. Trove's single-page app never painted a result list in this session's browser; the API returns JSON and needs no JavaScript. Seven certificate-of-naturalization entries found — Prostamo Angela and Antonino Giacomo (16 November 1961) and Assunta (3 November 1966); Polistena Maria Luisa (22 January 1959), Maria and Annunziato (March 1969); Melluso Francesco (6 July 1961). They are in `data/trove-gazette.tsv` as candidates, not joins.
And the notices are thinner than this row hoped. It expected «a date, a street address and a former nationality». These give a name in an alphabetical list and a date — nothing to tell two men of one name apart. The Graceville address this archive already holds for Maria Assunta came from somewhere else.
Still to sweep: Mazza, Arena, Donato, Nicotra and Anile — the last needs a phrase query, because a bare `Anile` returns 5.8 million results by matching inside longer words.
SWEPT OUT 27 SEPTEMBER. Every Australian person this archive holds searched by full name. Michele Mazza, Ignazio Mazza, Venera Mazza, Alfonso Prostamo, Caterina Melluso and Giovanni Polistena all return nothing. Sixteen Queensland hits in total, and the two that matter are withheld: a naturalisation ceremony report for SEBASTIANO MAZZA, Bundaberg, 6 July 1962 — «THREE NEW CITIZENS ... three former Italian citizens received certificates of...» — and a Family Notice naming him, 28 September 1963. Both carry `status: currently unavailable`; Trove indexes the post-1955 Queensland papers and will not serve their text. A copyright wall, not an absence. See row 184.
AND THEN RE-OPENED THE SAME DAY. «Swept out» meant swept out of the NEWSPAPER category. Trove has seven others, and the archival one holds four Nicotra wills at Queensland State Archives. See row 185. A source is not exhausted because the part of it you know about is.
80. Running — a machine or a person is at it now this archive
Angela Polistena × Rocco Arena — documented as a fact, still undated
Phase 9. They are Michael's grandparents on the Arena side and their marriage is not documented anywhere in this archive. Queensland's marriage index closes in 1951, so if they married before that it is reachable there and if after, it is not — which is itself worth establishing. Nudgee's register, the Catholic parish, or Michael. 20 SEPTEMBER: IT IS DOCUMENTED, JUST NOT DATED. NAA: J25 1971/3900, Brisbane, 1953–1971: *«Arena, Rocco; wife Angelina [or Angela] [nee Polistena]»*. J25 is the Queensland immigration and naturalisation correspondence series and its titles name a couple together with the wife's maiden name — which is precisely what this row said was missing. What it changes: the marriage is no longer an undocumented family claim. What it does not do: give a date. The file is «not yet examined», which means nobody has ever asked for it; an access examination is free and takes about ninety days, and it would very likely settle the date and place. That is a request for David to lodge, alongside the other Brisbane items. The Queensland marriage index still closes in 1951 and still has not produced them, which now reads as evidence they married after 1951 — or in the church only.
81. Next — ready to start the kit & estate
ESTATE — offer check_corrections.py and the phase structure
Phase 10. `check_corrections.py` reads the confidence vocabulary out of the page itself, so the legend and the data cannot drift apart — it found thirteen rows using labels the page never explained. Any archive with a corrections page has that exposure. And the work list's phase numbering turned 57 rows from a pile into a roadmap without changing the schema; worth passing on as a convention rather than a code change.
82. Next — ready to start the kit & estate
ESTATE — warn about the stale Vite cache that fails a build twice
Phase 10, and small. Two consecutive builds here died on «Cannot find module dist/pages/people.astro.mjs». I first diagnosed it as a race between concurrent builds. It was not — it was a stale `.astro` / `node_modules/.vite` cache, and `rm -rf dist .astro node_modules/.vite` fixed it immediately. Every archive in the estate runs the same Astro version and will meet it.
83. Done — finished and published this archive
Write down what today taught about instruments, not just about people
Phase 10. Four findings today were about how to read, not about the family, and each one will otherwise be rediscovered: a Tavola may alphabetise by given name and change key between adjacent years; some Tavole leave the year blank and must be dated from the acts; Queensland BDM cannot see a 1949 migration because its windows close in 1926, 1951 and 1996; and Find a Grave will index one person at two cemeteries when the names resemble each other. Three are in memory; none is on the site. /method/ is where a reader would look. DONE, 21 SEPTEMBER, AND LONGER THAN FOUR. Nine instruments are on /method/, built from `data/instruments.tsv`: the annual Tavola, the act-head montage, the marriage processetti, the marginal annotation, the register index at s_size=500, the container fetch from inside the origin, the Riveli di anime e di terre, the Queensland BDM download endpoint, and NAA item titles. Every row carries a CAUTION and the builder refuses one without it. An instrument described without the way it fails is how a method becomes a superstition — the best tool in this archive is the Tavola, and it changes its alphabetical key between adjacent years. And the page stopped lying about its own audit. It said *«nine were audited and five had already failed»* in fixed prose, beside a link to a page that counts for itself. It reads twelve and seven now, because it counts too.
84. Next — ready to start this archive
BRIATICO — Caterina Cichello's death, which would confirm her parents independently
Phase 7. She was alive at Briatico in June 1849, a widow of eighteen years, when her daughter married — and her parents Francesco Cichello and Vittoria Mollo are known from one document only, the Zambrone marriage act of 1824. Her death act would name them again, from a different clerk in a different town, which is the standard this archive holds a new couple to. Born about 1798, so she died after 1849 and Briatico's death registers run on.
85. Next — ready to start this archive
Domenico Arlotta — confirmed, dated, and one age still wrong
Phase 7. He is dated «before 1779» and «dead before April 1854», and he exists in this archive on the strength of a single phrase — his son Giovanni declaring a grandson's birth in 1854 as «figlio del fu Domenico». He is the deepest Scilla ancestor and has never been looked for. Scilla's registers begin in 1816 and he died before 1854, so his death act may be inside the run even though his birth cannot be. That would give an age, a trade, and possibly his own parents. UNBLOCKED 16 SEPTEMBER. The pre-1866 Scilla death series is now listed — 51 volumes, 1817–1865 — in `scilla-arks.tsv` (row 130). This row can start. SCILLA BEGINS IN 1809, NOT 1816 — checked 20 September. The napoleonic fondo holds seventeen registers: Matrimoni and Morti for 1809, Matrimoni/Morti/Nati for 1810, 1811, 1812 and 1815, and Matrimoni 1813–1814. Domenico Arlotta was born before 1779 and dead before April 1854, so his death is reachable in the civil registers and his son's 1851 marriage processetti (`an_ua2002487`) may hold the act outright. All seventeen arks are in `scilla-arks.tsv`. AND HE MAY NOT BE THIS FAMILY'S AT ALL. The Domenico Arlotta generation rests entirely on the 1854 act calling the declarant *«Giovanni Arlotta, figlio del fu Domenico, di anni settantacinque»*. Serafina's own birth act of 1830 makes her father forty, not fifty-one. Either an age is badly wrong or the 1854 declarant is a different Giovanni Arlotta — and if he is, Domenico Arlotta leaves the tree. Settling it is row 39, not this row. CONFIRMED AND TIGHTENED, 20 SEPTEMBER. His son's marriage act of 30 August 1828 calls him *«fu Domenico Arlotta»* — so he was dead before August 1828, not merely before April 1854, and the generation does not leave the tree, which was the real risk. He also has a wife now: ANNA BARBARO, alive at that wedding. What stays open is his son's age, not his existence. Giovanni Arlotta is 28 in 1828 (b. c. 1800) and 75 in 1854 (b. c. 1779). Both acts call him *marinaro* and both call him a son of a late Domenico. Either the 1854 clerk wrote an age badly wrong, or there were two Giovanni Arlotta at Scilla, both sons of a late Domenico — in which case the 1854 declarant who named his grandson after himself may not be Serafina's father at all. His own birth is out of reach either way: c. 1800 or c. 1779, both fall before Scilla's civil registration begins in 1809, so it is a parish record or nothing.
86. Next — ready to start this archive
Michele Donato came back from America — find the return crossing
Phase 9. He sailed from Palermo on the Liguria in February 1908 for Quincy, Florida. The 1910 US census for the whole of Gadsden County holds fifteen Italians and neither him nor Giovanni Polistena — so Quincy was where they were going, not where they stayed, and his wife and household stayed at Via Grotte throughout. He came back and there should be a manifest. Departure from a US port 1908–1910, or an Italian arrival. It would also be the last dated fact anyone has about him.
87. Done — finished and published this archive
Generosa Costa's documented parents now reach the reader — and she was the only case
Done, and the spot-check was worth running. Her 1888 marriage act names «FIGLIA DI ANTONIO... E DI ANNA ARLOTTA», and both parents were already documented people in this archive with nothing connecting them to her — so no chart or page could show that Rocco Arena's wife had known parents. An anchor row now joins them, and the documented line goes from 13 ancestors to 15. I then checked whether the same was true anywhere else, by parsing every addition's note for «X's father/mother» and testing whether X was an unanchored export person. Generosa Costa was the only real case. That is worth knowing: the gap was one row, not a systemic backlog.
88. Done — finished and published this archive
Audit the six confidence levels — done, and one label was carrying three jobs
Audited all ninety-four rows against the definitions. `certain` is used consistently — all ten rest on *two documents set against each other*, which is exactly what the legend says: an act against its chancery copy, a declared age against a birth act. `probable` was carrying three different jobs: an ambiguous reading, a conflict between documents, and an inference from circumstance. The first two belong together; the third does not. Maria Rosa Prostamo and Mattea Prostamo — both reasoned from BIRTHPLACE, with nothing ambiguous to read — moved to `inferred`, and both the `inferred` and `probable` legends on /corrections/ now say plainly what separates them. `inferred` goes 1 → 3, `probable` 6 → 4.
89. Next — ready to start this archive
PIEDIMONTE — Maria Nicotra's mother: «Monti Venera» or «Venera Cassaneti»?
Migrated from QUEUE.md, and it is a live conflict. The 1898 marriage proclamation at Piedimonte Etneo calls her «Monti Venera»; the Queensland death registration of 1963 calls her «Venera Cassaneti». Same woman, two surnames, and she sits on the spine at generation five. This is the same shape as Anna Fama against Anna Maropoli and Caterina Di Meio against Caterina Majo — three name-conflicts now, each between an Italian act and a later copy, and they may be worth a page of their own. A CANDIDATE, FROM THE SPELLING THIS ROW'S SISTER ESTABLISHED. Piedimonte Nati 1848 atto 45: «CASSANITI VENERA — Cassaniti Gaetano / Venuti Maria», born the 14th. Right name, right town, and the arithmetic fits twice over — twenty-seven when Maria was born about 1875, seventy-seven when Venera Cassaneti died in 1925. It does not settle the «Monti Venera» conflict and it is not yet a finding: row 161 read fourteen Piedimonte birth indexes and found no Maria Nicotra, so either this household left the town or this is a different Venera. The test is the marriage index, not another birth sweep. HALF SETTLED, AND THE CONFLICT IS NOW THREE-WAY. She is VENERA CASSANITI, born about 1848 at Piedimonte, daughter of the late Gaetano Cassaniti and the late Maria Currenti — her own marriage act of 7 October 1872 says so, and Nati 1848 atto 45 fits it to the year. So «Cassaneti» is a Brisbane registry counter's spelling and nothing more. «MONTI VENERA» is not explained. The 1898 Piedimonte marriage proclamation still calls her that, and no act touched so far accounts for it. That is what this row now holds — one name in three forms, two of them reconciled and the third outstanding. AND HER DAUGHTER'S BIRTH ACT NAMES HER AGAIN, in the same spelling: Mascali Nati 1875 atto 144, *«da VENERA CASSANITI sua moglie»*. Two acts, three years apart, two comuni, one spelling. «Monti Venera» remains the outstanding third form.
90. Next — ready to start this archive
PIEDIMONTE — Mazza Veneranda × Mariano Cavallaro, 1 October 1920
Migrated from QUEUE.md. Another child of Salvatore Mazza and Maria Nicotra, absent from the tree — and probably the link to the Brisbane Cavallaro, of whom the export holds thirteen. The act is dated exactly, so it is a single lookup rather than a sweep.
91. Next — ready to start this archive
PIEDIMONTE — the other Mazza sons: Giovanni b.1875, Gaetano b.1880, Santo b.c.1882
Migrated from QUEUE.md. Rosario Mazza's other sons, none of them in the family tree. Giovanni is the one to read first: he is the declarant at Salvatore's death in 1929, aged fifty-four, which both dates him and shows he was still at Piedimonte when the rest of the family had gone to Queensland. Act 64 of 1875.
92. Next — ready to start this archive
PIEDIMONTE — who is Pasquale Mazza, and is the generation compressed?
Migrated from QUEUE.md, and it is a structural question rather than a lookup. Pasquale Mazza, b. c.1829, bracciante of Via Terremorte, married Concetta Greco. Is he a third brother of the line, or their father's brother? The queue also carried a «compressed-generation hypothesis» — that a Salvatore Mazza of Via Terremorte born about 1845 collapses two generations into one. Both turn on the same acts, and the Mariano Mazza frontier (row 47) cannot be settled without answering them.
93. Next — ready to start this archive
PIEDIMONTE — Innocenzio Mazza × Domenica Cavallaro, about 1835–40
Migrated from QUEUE.md. A marriage from before Antenati's Piedimonte run begins, so it may be unreachable the way three of the four walls are — but that has never been established, and establishing it is cheap. If Sicilian civil registration begins in 1820 (row 14) then a marriage of 1835–40 is *inside* the run, not outside it, and this item may have been written off too early. OPENED, AND THEY ARE REAL. The annual Tavola of Piedimonte Nati 1847 gives at act 31 «MAZZA SEBASTIANO — Mazza Innocenzio / Cavallaro Domenica», born the 25th. The couple was at Piedimonte and having children in 1847, so the marriage this row thought might be unreachable is bracketed from above by a real document. Next: open act 31 for their ages, trades and street, which is what dates the marriage. Found by reading a whole Tavola for an unrelated question — the argument for reading the whole table rather than scanning for one name.
94. Next — ready to start this archive
Check whether RAITI joins the hundred and ten to the family
Migrated from QUEUE.md, and it is the kind of small thread that opened the Donato line. Raiti appears both in the tree's unjoined hundred and ten AND as the surname of Antonino Mazza's wife. If those are one family, it is a second bridge between the Piedimonte cousins and the direct line — and the first bridge, the Rosario Mazza join, moved 102 people onto the spine.
95. Next — ready to start this archive
Michele Mazza, 1925–26 — where is he buried, and what does his passenger record say?
Migrated from QUEUE.md. His Queensland death registration of 1926 carries the «C» code — a country registration, outside Brisbane — and he is not at Nudgee, so where he lies is unknown. Separately his 1925 Brisbane passenger record at the NAA would name a next of kin and a destination address, which for the first of this family to arrive is worth more than it sounds.
96. Done — finished and published this archive
Camillo Antonio Polistena's own birth act, about 1835
Migrated from QUEUE.md. Scilla's births begin in 1816, so a birth about 1835 is inside the run and reachable. He is the oldest Polistena this archive has, and the Polistena quarter is the thinnest of the four — the surname returns nothing at all in the Queensland index and has no documented ancestor above him. CLOSED 16 SEPTEMBER, and not by his birth act — by his death act. Camillo Antonio Polistena died at Scilla on 19 October 1924, «Padre GIUSEPPE, Madre ANNA PONTILLO, Coniuge GIOVANNA ROMANO». The thinnest quarter gains two documented ancestors where it had none above him. His birth act, about 1835, is still worth having for the age and the address, but the parents no longer depend on it.
97. Done — finished and published this archive
Guard the searched.tsv key column — appending a row silently corrupts it
Raised because I broke it twice in one session. `data/searched.tsv` gained a `key` first column from another session, and `build_searched.py` deliberately never recomputes it — the docstring says why: hand-added keys were once wiped and the gate then reported «nothing outstanding» for an archive with twenty-four outstanding searches. But nothing stops a session appending a row in the old five-column shape. I did, twice, and every field shifted one place right: `part` landed in `key`, a date landed in `status`, and the rows became invisible to the gate that exists to count them. A checker should refuse a row whose key is empty or does not match `src/<slug>-<6 hex>`, and refuse a status outside the vocabulary — which has drifted to seven values: done, outstanding, not-yet, part, superseded, correction, wall. Some of those I invented myself without declaring them. DONE 16 SEPTEMBER. `tools/check_searched.py`, wired into the build as `check:searched`. The shape check runs first and reports by physical line, because once a row has shifted the named columns cannot be trusted to mean what they say — and on the actual bug it prints *«5 columns, expected 6 — this is the shifted-row bug; first cell is «part»»*, which is word for word what happened. It also refuses a key that is not `src/<slug>-<6 hex>`, a duplicate key (two searches counting as one), a status outside the seven words in use, and a `done` or `part` row with an empty result — a finished search with nothing written in it being counted as a negative result. One thing learned writing it: I invented the status vocabulary instead of reading it, and the first run threw 43 false failures. The seven real words are `done`, `outstanding`, `not-yet`, `part`, `superseded`, `wall`, `correction` — and only `outstanding` is what `check:covers` counts, which is worth knowing before anyone adds an eighth. There is no legend on /searched/ to read them out of, unlike the confidence words on /corrections/, so they are pinned in the checker with a comment saying so.
98. Done — finished and published this archive
FINDMYPAST — unlock Our Lady of Loretto: eighty-four Arena baptisms, indexed and locked
THE PAYWALL IS OPEN AND THE FIRST PASS HAS PAID. The subscription includes the Catholic Heritage Archive: 265 Arena baptisms across New York's Catholic parishes, 18 at Our Lady of Loretto, transcripts live. Rocco Arena and Generosa Costa's whole New York household, found in one search: Antonio b. 3 Oct 1896, Serafina b. 15 Aug 1898, Clementina b. 21 Dec 1900, and a second Antonio b. 13 Oct 1902. An address at last — 141 Mulberry in 1896–98, 198 Mulberry by 1901 — corroborating the 1899 Ellis Island manifest that sends Giovanni Arena to «brother Rocco, N. York, Mulberry St». And a fifth Rocco Arena baptism excluded on the mother's name: Giuseppe, 1911, the Bronx, mother *Teresa Siniscaldi*. A surname sweep would have taken all five. Still to do: the remaining 247 Arena baptisms for the wider colony, and the same collection's marriages and burials, which have not been touched at all. DONE 16 SEPTEMBER, and it closed smaller and harder than it opened. The 247 remaining Arena baptisms were never the way to this family — the parent fields are, and they are now exhausted. Father Rocco + mother Generosa returns exactly the four known children and nothing else, which audits the American harvest as complete. Father Giovanni + mother Concetta returns zero, so Rocco's brother — who sailed in 1899 to join him — baptised no child in New York, and came as a sojourner rather than a settler. The marriages are excluded by arithmetic, not sampling: both Arena men were already married at Scilla before they sailed, in 1885 and 1888, and no child of Rocco's was of marriageable age in America before the household left. And there are no burials to sweep — the Catholic Heritage Archive for New York is baptisms and marriages only, so the infant death of the first Antonio cannot be tested here at all. See row 113.
99. Done — finished and published this archive
FINDMYPAST — the Australian electoral rolls, which unblock rows 49 and 62
Done — all seven surnames this row named, and the answer is half of what it promised. Prostamo, Polistena, Mazza, Arena, Donato, Melluso and Crupi, each swept in «Australia Electoral Rolls» and sorted to bring the Queensland rows onto the first page. It found Michael's grandfather. SEBASTIANO MAZZA, builder's labourer, Progress Road, Wacol, sub-district Inala, on the 1959 roll — the post-war migrant belt, ten years after he landed, and the first record of him in Australia between his arrival and his grave. It also found ROSARIO and ANGELA MAZZA at 225 Leichhardt Street on consecutive roll numbers, which is one household, and ROSARIO SALVATORE MAZZA at Paddington in 1949, the same man a decade earlier. But the instrument is smaller than this row assumed. Findmypast holds only eight Queensland commonwealth years and the last is 1959 — for a family that arrived in 1949 that is *one usable roll*, not the near-annual run promised here. See rows 104, 105 and 106 for what that leaves. Recorded in what has been searched, which carries the Findmypast log. The Crupi were the second prize. Angela Crupi's three brothers — whom the Queensland death index named and neither Brisbane cemetery register could find — are all at Tully on the 1959 roll with trades and streets. Polistena, Melluso and Rocco Arena return nothing, and that absence has a shape: of the adults who landed in 1949, only Sebastiano Mazza was enrolled by 1959. See row 108 for the two surname clusters this turned up.
100. Done — finished and published this archive
FINDMYPAST — the nine unnamed crossings, and the Australian passenger lists
Row 21's other half. The trial already CORROBORATED five Prostamo in the Australian passenger index — Caterina 1969, Domenico F 1956, Francesca 1957 — and separately failed to find Domenica Prostamo's 1951 voyage in «Australia, Inward, Outward & Coastal Passenger Lists 1826-1972». With full access, sweep that collection properly for all ten surnames rather than by targeted lookup, and test the *Ugolino Vivaldi* candidate of 1 March 1951 that `corrections.tsv` carries at `candidate` confidence. DONE 16 SEPTEMBER, and the answer is that this was the wrong shop. For twentieth-century Australian arrivals the collection is the NAA passenger index — every post-1945 row carries a NAA barcode and links back to recordsearch.naa.gov.au — and this archive has already worked that index directly, for nothing. The Ugolino Vivaldi candidate could not be tested: Findmypast serves the identical record, barcode 30133197, with no age and no sex, so the candidate grade stands. One new crossing did fall out: ROCCO PROSTAMO, 10 August 1949, the *Napoli*, Genoa to Melbourne — a sixth Prostamo and not one in the tree. And the subscription does hold something genuinely new: the Victorian inward, outward and coastal lists, 1852–1923, which are not in the NAA index at all. See rows 110 and 111.
101. Done — finished and published this archive
FINDMYPAST — Rocco Arena's naturalisation, and the American records
Covers the outstanding «NARA New York» search. The trial established Rocco Arena's four crossings (1892, 1894, 1899, 1913) and that by 1913 he was travelling as a US citizen — which means a naturalisation exists and has never been found. It also found no petition under his name, so the search needs the subscription's wider American sets: NARA indexes, WWI and WWII registrations, and the 1900–1930 censuses for Mulberry Street. DONE 16 SEPTEMBER. The naturalisation is still not found, but the census settled the family's movement and bounded the search. Rocco Arena appears in the United States census exactly once, in 1900 — one hit in the whole country, right birth year — and never again. The 1910 census indexes 543 Arena households and not one holds a Rocco, nor a Generosa, nor a Serafina, nor a Clementina. The 1920 and 1930 Rocco Arenas are all born after 1882. Three independent absences now say the household went back to Scilla: no census after 1900, no Generosa on the 1913 manifest, no New York draft card for the son born on Mulberry Street. See row 112 for where the naturalisation actually is.
102. Done — finished and published this archive
FINDMYPAST — re-run the four PAYWALL and BLOCKED rows in findmypast.tsv
Twenty-six rows of trial work exist and four of them are not answers, they are locks. Two PAYWALL rows (the New York Roman Catholic transcripts, twice), and one BLOCKED row recording that the session signed out before the work was finished. Re-run each one and convert it to a real result — found or negative — rather than leaving a paywall standing in the record as though it were evidence. A lock is not a negative and this archive should not let one look like it. DONE 16 SEPTEMBER. No PAYWALL or BLOCKED row remains in the log. The two parish-transcript locks opened and became the American harvest — four Arena baptisms, both Mulberry Street addresses, two namesakes excluded on a mother's name. The signed-out row became a method note. And turning the third lock produced a retraction. The locked row had claimed the transcript carried «the mother's name and the godparents». The mother's name is there and has done more work than anything else this week. The godparents are not in the record at all — no such field, and no image behind it, so they are unreachable at any tier and exist only in Our Lady of Loretto's own register. *That claim was a guess about the contents of a locked record, written down as though it were a finding*, which is exactly the failure this row existed to fix.
103. Blocked — waiting on a person or an institution David
FINDMYPAST — decide what to do before the month expires
A calendar item, not a research one. The previous trial expired the day it was most needed and the record still carries a BLOCKED row from the moment it lapsed. A month is enough for rows 98 to 102 if they are done deliberately and in that order — the locked parish transcripts first, because they are the only thing here that cannot be got anywhere else. Worth agreeing a date to review what is left, rather than discovering the end of it the same way as last time. UPDATED 16 SEPTEMBER: rows 98 to 102 are all done, on the first day. The subscription has been worked out rather than used up, and the honest summary is that Findmypast had less for this family than the work list expected, and gave it up quickly. What it delivered: Sebastiano Mazza on the 1959 roll at Wacol; the Crupi brothers at Tully; Rocco Arena's absence from the American census proved across the whole country; the parish transcripts opened and audited. What it turned out not to hold: Queensland rolls after 1959, any Australian arrival the free NAA index does not already have, any New York burial, and any godparent. So the question is no longer «what else can we get before it expires» — it is whether the month is worth renewing at all. On this evidence, no: the open work now points at Ancestry or the NLA for the post-1959 rolls (row 104), the NAA for naturalisations and one manifest image (rows 105, 110, 112), and free New York city records (row 113). The one thing still unmined here is the Victorian passenger lists (row 111). Worth a decision, not a diary date.
104. Blocked — waiting on a person or an institution David
The Queensland rolls after 1959 — which means the NLA or Ancestry, not Findmypast
Row 99 hit a wall that a subscription cannot climb. Findmypast's last Queensland electoral roll is 1959. Sebastiano Mazza is on it at Wacol; Domenica Prostamo is not on it at all, under either name — most likely because she had not yet been naturalised, since her husband would have been naturalised first. Every year that would settle that is outside this collection. The near-annual Queensland run to 1980 is held by the National Library of Australia and by Ancestry, and it would follow this household through Wacol year by year, catch Domenica's first enrolment, and place Frank Mazza as he came of age. This is David's call, not the archive's — it is a second paid service, and the Findmypast month is already running.
105. Done — finished and published this archive
Sebastiano Mazza's naturalisation, which the 1959 roll proves exists
An enrolment is a proof of citizenship, and that is the quiet finding in row 99. Only British subjects could be enrolled to vote, so Sebastiano Mazza's presence on the 1959 Queensland roll means he was naturalised between his arrival in 1949 and 1959, and a naturalisation file exists for him. The NAA holds them — series A712, A714 and the certificate registers — and this archive already knows how to work the NAA index from the passenger work. The file would give his birth date, his birthplace, his occupation, his address, and often the ship and date he arrived, which would corroborate the crossing independently. And it should be run for Domenica Prostamo at the same time, because her naturalisation date is exactly what would explain her absence from the 1959 roll. DONE 16 SEPTEMBER — the file exists and it is cited. NAA: J25, 1971/1074, «Mazza, Sebastiano; wife Domenica [nee Prostamo]», Brisbane, 1950–1971. It names both of them and her maiden name, and its date range covers their whole Australian lives to 1971. It is paper, it is not digitised, and it is «not yet examined» — so it needs an access examination request before anyone can read it. That is row 115, and it is not this archive's to make. It may also settle row 21: a Queensland immigration file opening in 1950 should record how and when Domenica came.
106. Running — a machine or a person is at it now this archive
Andrea Mazza — three dates now, and still no parents
Tully is not a coincidence and it is not yet a relative. Rosario Mazza's daughter Carmela Mazza was born at Tully in 1926, so this family had a household in that exact sub-district; and the 1959 roll puts an ANDREA MAZZA, timber cutter, at Cardwell in the Tully sub-district. No Andrea appears anywhere in the tree as it stands. A roll carries no parents and no birth year, so it cannot decide him. What could: the Queensland BDM birth index for Tully in the 1920s and 1930s, which *is* inside its window and would name his parents; or Michael, if the family remembers a branch that stayed north. Left as a lead rather than attached — the archive has been offered two Arena namesakes in New York this week and excluded both on a parent's name. THE TULLY PREMISE IS BACK, BY A DIFFERENT ROUTE — 20 SEPTEMBER. Row 116 knocked out the reason this row was interesting: Carmela Mazza was not born at Tully in 1926, so no Mazza household stood there then. But a family household did. Angela Crupi's brother Antonino Crupi was a cane farmer at Dean Street, Tully, naturalised there in 1946, and Pietro and Mario Crupi were at Tully too. The family's Tully connection is real and it runs through the Crupi, not the Mazza. That changes what Andrea Mazza would mean. A Mazza at Cardwell in the Tully sub-district in 1959 is a man living where his in-laws' family already was — which makes him more worth checking and no more proved. The test is unchanged: the Queensland BDM birth index for the Tully district, which names parents. DATED, 20 SEPTEMBER. NAA: BP4/3, ITALIAN MAZZA A — «Mazza, Andrea — naturalisation certificate granted 27 NOVEMBER 1934», held at Brisbane. Naturalisation of that period wanted about five years' residence, so he was in Australia by about 1929 — thirty years before the roll that found him, and long before the post-war migration this row has been reading him against. With his Queensland death on 31 July 1975 that makes three dates. He is still not a documented relative, and the certificate names no parents. Two things would settle him: his death certificate, which names parents where the index gives only the surname *Mazza*; and the naturalisation file, which carries a birthplace — and if that birthplace is Piedimonte Etneo this stops being a lead and becomes a question.
107. Done — finished and published this archive
Never record a negative from an untested search parameter — the rule this sweep nearly broke
Raised and closed the same day, because it nearly put a falsehood in the record. On Findmypast's results URL a wrong parameter fails in two different ways and neither announces itself. `state=Queensland` and `page=2` render a filter chip and report «0 results found» — word for word what a genuine negative looks like. `street=progress road` is discarded in silence and the unfiltered count comes back unchanged, so the search you think you ran is not the one that ran. The first was caught only because the *unfiltered* search plainly listed Queensland rows. The rule now in the log: control-test a parameter against a search known to return rows before believing any zero it produces. This is the same family of trap as `fathersfirstname` for `fatherfirstname` earlier in this work, and as the multi-comune trap in the Antenati registers — *the instrument answering a different question than the one asked.* Written up in what has been searched.
108. Next — ready to start this archive
Donato and Nicotra in Queensland — two clusters off this family's own spine, unopened
The electoral sweep walked into two surnames nobody here has looked for in Australia. DONATO: 65 rows, and the Queensland block holds a CANDELORE DONATO enrolled in 1941 and 1949, a CANDO DONATO in 1939 and 1959, and a DOMENICO DONATO in 1959. CANDELORO DONATO IS ONE OF THE FOUR WALLS — born about 1783 at Scilla, the point where the Donato quarter stops. He is plainly not the same man, and *that is the interest*: southern Italian families name a grandson for his grandfather, so a Candeloro Donato in Queensland is what a descendant of the wall looks like from this end. Domenico Donato is the name on the 1816 birth act that came out of the Scilla processetti this week. NICOTRA: 39 rows, almost all Queensland, a family enrolled continuously from 1934 to 1959 — Filippo, Gregorio, Giuseppe, Carmela, Leonarda, Leonard. Nicotra is Sebastiano Mazza's mother's name, Maria Concetta Nicotra. Neither is claimed and neither can be, from rolls alone — no parents, no birth years. The route that could decide them is the Queensland BDM birth and marriage indexes, which *are* inside their windows for the 1930s and 1940s and do carry parents.
109. Done — finished and published this archive
Which Antonio Crupi — answered by a naturalisation, not by the roll
A small ambiguity, written down so it is not quietly resolved the wrong way later. The 1959 roll holds two Antonio Crupi: a cane farmer at Dean Street, Tully, and a farmer at Silkwood East in the Innisfail sub-district. Angela Crupi's brother ANTONINO CRUPI died 15 April 1982 (reg 1982/52783), son of Francesco Crupi and Maria Vecchio. A roll cannot tell which man he is — it carries no parents and no birth year. The Tully one is the better bet only because his brothers Pietro and Mario are both at Tully, and *better bet* is not evidence. What would settle it: the 1982 death registration's district, or the marriage of an Antonino Crupi registered 24 August 1940 (1940/C/3314), which would name a wife to match against Nellie Crupi, the woman sharing Dean Street with the Tully man. ANSWERED 20 SEPTEMBER, AND BY THE TEST THIS ROW ITSELF NAMED. NAA: A714 57/18792 — *«CRUPI, Antonio; naturalised 7 November 1946; place of birth Piedimonte, Italy; year of birth 1907; place of residence Tully, QLD; [W] CRUPI, Nellie»*. Piedimonte and 1907 tie it to Angela Crupi's brother's aliens card; Tully and Nellie tie it to the cane farmer of Dean Street on the 1959 roll. The Silkwood East farmer is a different man. Found while reading Antonino's aliens card for row 117 — the question was answered by a document nobody went looking for. And it adds a move: a *boot repairer at Ingham* in October 1939, a *cane farmer at Tully* by 1946, where Pietro and Mario were too. One loose end tidied, not proved: the card registered in October 1939 already says *married*, which sat badly against a Queensland marriage of 24 August 1940. The same card carries 1943 annotations, so *married* was very probably added later and the 1940 marriage is the real date — very likely to Nellie. Stated as a reading.
110. Done — finished and published this archive
The original Ugolino Vivaldi manifest at the NAA — free, and the only thing that can settle Domenica's crossing
Row 100 proved the subscription cannot do this and pointed at what can. Domenica Prostamo's 1951 crossing is graded candidate on a good argument — right month, right year, right port, right married surname, and «Domenico» one letter from Domenica. What it lacks is an age or a sex, and the index line has neither, on Findmypast or at the NAA. The manifest image does. RecordSearch serves barcode 30133197 as a scanned page, free, and the handwritten entry would give the age and sex outright — and show everyone else on that page, which is how a two-year-old Salvatore Mazza would appear, since infants are not separately indexed. This is the single cheapest upgrade available to this archive: one image, no subscription, and it converts a candidate into a documented crossing or kills it. DONE 16 SEPTEMBER — and it killed the candidate rather than confirming it. This row said the image would «convert a candidate into a documented crossing or kill it». It killed it. The master list of trip 2/51 carries one Mazza, no. 476, bound for Narracan, Victoria. There is no Prostamo aboard under any spelling, and no second Mazza, though the list names children individually — so no two-year-old Salvatore. And the port column settles it: unlabelled, but every `-F-` passenger appears in the Department of Immigration's separate *Schedule of Alien Passengers for debarkation at Fremantle*, one for one. MAZZA Domenico is `-M-`, and he is absent from that schedule, which runs MARTINELLI straight to MERENDINO. Domenica Prostamo was going to her husband in Brisbane; this man landed at Melbourne for a Gippsland shire. Retracted in the corrections, regraded *flagged*, and the two voyage rows are off the crossings. Her March 1951 crossing is unknown again — see row 21.
111. Done — finished and published this archive
The Victorian passenger lists 1852–1923 — the part of the subscription that is genuinely not the NAA
Found by accident while proving row 100 a dead end. «Australia, Inward, Outward & Coastal Passenger Lists 1826-1972» is two different things stitched together. The post-1945 half is the NAA index re-served. The pre-1940 half is not: ids of the form ANZ/VCPL and ANZ/MIGR/VICTORIAPASSLIST are the Victorian inward, outward and coastal lists, which the NAA index does not hold. The Mazza search returns rows there from 1891, 1896, 1897, 1898, 1900, 1901, 1910, 1911, 1914 and 1919. Why it might matter: this archive's Australian story starts in 1925 with Michele Mazza on the *Palermo*, and everything before that is assumed to be Italy. These lists cover the years when it might not have been. Sicilian and Calabrian men worked passages to Victoria and went home again, and a Mazza or a Prostamo on an outward list would be exactly that. Speculative, cheap to test, and nobody has looked. DONE 16 SEPTEMBER, and the guess was wrong for reasons worth writing down. The Victorian material is mostly coastal — Strahan to Melbourne, Fremantle to Melbourne — which is people already in Australia moving around it, not people arriving from Italy. And the coastal rows are mostly nameless, carrying a surname, a sex and an age and no given name, so a real ancestor there could not be recognised anyway. The named overseas Victorian rows are British subjects. The control makes it a negative rather than a shrug: searched by name, Michele Mazza's 1925 crossing from Messina to Brisbane comes straight back. The collection can see this family when it is there. Before 1925 it is not. The Australian story still starts where this archive has always said it starts.
112. Next — ready to start this archive
Rocco Arena's naturalisation is a pre-1906 New York court filing — which no subscription holds
Row 101 could not find the papers and instead worked out when they must have been filed. He is in America from 1892, he is in the 1900 census, he is gone from the whole country by 1910, and he sails in 1913 as a US citizen. So the declaration and petition were filed in New York between 1892 and about 1906. That is exactly why Findmypast cannot see him: its United States Naturalization Petitions collection begins after the 1906 standardisation, and before 1906 these cases were heard in state, county and city courts as well as federal ones. The finding aid is a NARA index to petitions filed in New York City courts, 1792–1906 — on microfilm, held by the National Archives at New York City, and on no subscription site. *(The series number should be confirmed before anyone writes to them; it is believed to be M1674.)* What has changed is the size of the problem: the search is now bounded to a fourteen-year window in one city, instead of a lifetime in a country.
113. Next — ready to start this archive
The first Antonio Arena's death, 1896×1902 — free records, and neither has been tried
An inference this archive has published, and the record that would settle it is free. Rocco Arena's first son Antonio was born 3 October 1896; a second Antonio was born 13 October 1902. This archive reads that as the first boy dying in between — the same reading it makes of Orazio Arena's two sons named Domenico at Scilla. It is sound and it is still an inference. Findmypast cannot test it: there is no New York Roman Catholic burials collection — baptisms and marriages only, and «Congregational Records» is empty even of Russo. Two places can. The New York City death certificates at the Municipal Archives are indexed from 1862 and searchable for nothing; and the interment registers of Calvary Cemetery, where the Italian Catholic dead of lower Manhattan were mostly buried. A death between 1896 and 1902 for an infant of 141 or 198 Mulberry Street is a narrow, well-specified search. It would also date the family's last years in America, which matters now that the census says they were gone by 1910.
114. Next — ready to start this archive
Salvatore Mazza's arrival, which fell with his mother's
Collateral damage from row 110, and it should not be left implicit. This archive placed Salvatore Mazza (b. 1948) on the *Ugolino Vivaldi* at `inferred`, on the reasoning that infants are not separately indexed and he belonged on his mother's voyage. The reasoning was fine and the voyage was wrong. Reading the manifest also removed its own premise: this master list names children individually — the Mellino family of eight, the Puusaari of six, a Campagna aged 1 — so an infant *would* have been listed. He has no documented crossing now. The tree records a Fremantle arrival out of Genoa with no date; that, and the family account of a son not yet three, are all that remain. He travels with whatever voyage row 21 eventually finds for his mother.
115. Blocked — waiting on a person or an institution David
Ask the NAA to examine J25 1971/1074 — the family's own naturalisation file
The single best unread document about Michael's grandparents, and getting at it is a form, not research. NAA: J25, 1971/1074, «Mazza, Sebastiano; wife Domenica [nee Prostamo]», Brisbane, contents 1950–1971. Its access status is «not yet examined», which means nobody has ever asked for it. Anyone can ask, it costs nothing, and the NAA then examines it for release — the usual wait is about ninety days. Why it is worth the form: a Queensland immigration and naturalisation file covering 1950 to 1971 would carry Sebastiano's naturalisation, his address history, his occupation, and — since she is named on it — how and when Domenica actually arrived, which row 21 badly needs now that the Ugolino Vivaldi is disproved. This archive cannot lodge the request; it needs a person with an account and a postal address. Michael is in Brisbane, where the file is.
116. Next — ready to start this archive
Carmela Mazza cannot have been born at Tully in 1926
A contradiction found by accident, and it broke one of my own arguments the same day. The tree has Carmela Mazza born 1926 at Tully, Queensland. But NAA: A12508, 31/1454 has her arriving at Brisbane on 7 December 1937 aboard the *Viminale*, aged eight, as an Italian alien, with her mother Angela Crupi and her sister Maria. A child born at Tully in 1926 would be eleven and an Australian-born British subject, needing no alien declaration at all. The record is not loose about ages — Maria's «age 15» matches the tree's 1922 exactly. So she was born in Italy about 1929. What it cost: I had argued that the Tully timber cutter of 1959 (row 106) was interesting because a Mazza household already stood at Tully in 1926. On this evidence it did not. The Tully connection survives, but through the Crupi brothers, not through a birth that did not happen. What would settle it: the Queensland BDM birth index, which covers 1926 and would simply show whether a Carmela Mazza was registered at Tully at all.
117. Done — finished and published this archive
The aliens cards — Antonino's read, Pietro's explained away
Row 50 found the seam and did not empty it. Mario Crupi's card (D4028) turned out to be digitised, free, and the richest document this archive holds on anyone in Australia — birth date, birthplace, occupation, arrival, a physical description, and an internment camp. There should be more. Wanted: cards for PIETRO and ANTONINO CRUPI, the other two brothers, both of whom are on the Queensland electoral rolls; for ROSARIO MAZZA, to set against his internment dossier; and for ROCCO ARENA and the POLISTENA, who arrived in 1949 and may be too late for the wartime series. Note the series split: D4028 is Queensland aliens registration; B78 and BP25/1 and SP908/1 are the Victorian, Queensland and New South Wales equivalents, and a family scattered across states will be scattered across series. Search by surname in all record categories, then check each item for «View digital copy» — that is what separates an hour's work from a request form. DONE, 20 SEPTEMBER. CRUPI swept across all record categories — 233 items, read end to end at 200 to a page. ANTONINO CRUPI'S CARD IS DIGITISED AND IS READ. NAA: D4028, item 9209610. Born 5 October 1907 at Piedimonte Etneo — his brother Mario's town and the Mazza's. A boot repairer at Ingham. Entered Australia 2 September 1930, port Brisbane. Five feet four, black eyes, black hair, sturdy; the fourth finger of his right hand deformed. Registered at Ingham police station, 3 October 1939. AND HE WAS INTERNED — THE THIRD MAN IN THIS FAMILY. The front is stamped LOVEDAY 35924; the reverse records him moved *«ex Gay[thorne] to 14A»* on 16 March 1943 and released and returned to Queensland on 22 October 1943. Rosario Mazza on his own dossier, Mario Crupi at Loveday from August 1942, Antonino released October 1943. Angela Crupi's husband and two of her brothers, all three behind wire in the same eighteen months. The censor kept a sentence of his. A mail extract of 2 April 1943, from a letter to his wife in Queensland, is copied onto the card: *«… Our freedom will come soon and we shall always cry Vive Italia.»* It is the only sentence of his own this archive has, and it survives because an official wrote it down as evidence against him. PIETRO HAS NO CARD, AND THAT IS THE ANSWER, NOT A GAP. NAA: BP4/3 records *«Alien Registration Certificate not issued»*, the folio noting he was naturalised 23 February 1932 — by 1939 a British subject, outside the series. The brothers came together: Pietro and Antonino both arrived in Brisbane on the OTRANTO, 19 August 1930; Mario two years earlier on the MARIA CRISTINA, 20 May 1928. Naturalisations: Antonino 7 November 1946 at Tully, wife Nellie; Mario 26 April 1940 at Ingham, *single*; Pietro 23 February 1932. And a trap one barcode away: item 9209609 is another *CRUPI Antonio* — born Serrata, Calabria, arrived Melbourne 1926, nobody here. The run 9209609 / 9209610 / 9209611 is stranger, brother, brother. Adjacency in an alphabetical card series is not kinship. What is left: the ARENA and POLISTENA cards, who arrived in 1949 and are probably too late for the wartime series — and one open file, row 148. AND THE ARENA CARDS DO NOT EXIST, 20 SEPTEMBER. This row wanted aliens cards for Rocco Arena and the Polistena and guessed they would be «too late for the wartime series». They are. Rocco Arena's only record in the entire NAA name index is J25 1971/3900, the file that names him with his wife — no card, no naturalisation, no arrival item of his own. He landed in December 1953, after the wartime series closed. A caution came with it. Narrowing ARENA (1,650 items) to Rocco Arena gives nineteen, and one of them is an internment — a Sydney man born 31 August 1909, with an Aliens Tribunal objection and a dossier. He is not this family: this archive's Rocco Arena arrived thirteen years later, and the Rocco born at Scilla in 1861 died there in 1932. Guard written.
118. Done — finished and published this archive
Rocco Arena married his first cousin — and one act away from proving it
A question this archive asked itself and left open has almost certainly been answered. The row for ANNA ARLOTTA, Generosa Costa's mother, noted that Rocco Arena's mother was Serafina ARLOTTA and his wife's mother was Anna ARLOTTA, and said in as many words that whether they were sisters *had not been tested*. The Scilla death index tests it. Anna Arlotta, died 16 March 1915, is «Padre GIOVANNI ARLOTTA, Madre MARIA COSTA, Coniuge ANTONINO COSTA» — and Giovanni Arlotta and Maria Costa are Serafina Arlotta's own parents. Her husband's name matches the Antonio Costa the 1888 marriage act gives as Generosa's father. So the two women were sisters and Rocco Arena and Generosa Costa were first cousins. **Graded *probable*, not *certain*, and deliberately. Arlotta and Costa are both common at Scilla and nothing read so far rules out a second Anna Arlotta married to a second Antonino Costa. One act settles it: the 1915 death act itself, which gives her age and parents in full, or Generosa Costa's own birth act. Both are inside the surviving run. CERTAIN as of 16 September. The Costa sweep supplied the two acts that were missing: Antonino Costa's own death, naming Anna Arlotta as his wife, and Generosa Costa's, naming Antonino Costa and Anna Arlotta as her parents. Rocco Arena and Generosa Costa were first cousins** — his mother Serafina and her mother Anna were both daughters of Giovanni Arlotta and Maria Costa.
119. Done — finished and published this archive
Turn /open-questions/ into rows
202 lines of prose. Shape: the question, what is known, what has been tried, and the record that would answer it. `kit/components/Questions.astro`.
Raised by the UI alignment pass, 2026-09-16. It is recorded here because it is research this archive owns, not something to guess at from outside. [State corrected from «open» to «next» on 16 September — «open» is not in the estate vocabulary and the build gate refuses it. The row is otherwise untouched.] DONE, 21 SEPTEMBER. Fifteen questions in `data/questions.tsv`, rendered by the kit's `Questions` component — five open, ten answered — and the builder refuses a row with no `settles` field, because a question without the record that would answer it is a regret and not a queue item. The extraction caught what the prose was hiding. Three questions were headed ANSWERED and still sat undifferentiated beside the open ones. Two answered on 20 September had never been added at all — Filippo Arena's marriage, and how far above him the Arena line goes. And the *«is Salvatore Mazza in the Sicilian birth registers»* negative, nine indexes across three comuni, has since been overtaken: he is Nati 1873 atto 3, so all nine were read in the wrong place. The answered ones stay on the page, which is the decision worth recording: a question that vanishes when it is settled leaves a reader unable to tell a thing never asked from a thing asked and closed, and it hides the shape of the mistakes.
120. Done — finished and published this archive
Extract /sources/ into data
70 lines inline — the smallest, and so the cheapest of the five to move.
Raised by the UI alignment pass, 2026-09-16. It is recorded here because it is research this archive owns, not something to guess at from outside. [State corrected from «open» to «next» on 16 September — «open» is not in the estate vocabulary and the build gate refuses it. The row is otherwise untouched.] DONE, 21 SEPTEMBER, AND IT WAS NOT A REFACTOR. The list is `data/sources.tsv` now, and every row carries `seen` — *act, index, tree, lore, wall* — which is the only thing about a source that really matters. Twenty-five sources: six read in the original, fifteen index-only, two walls. And extracting it exposed a page that had been underselling this archive by a year's work. It opened *«almost every source below is an INDEX cited by the family tree, not a document this archive has seen»* and listed no Italian civil register at all; it closed under the heading *«What is not here yet»* with *«No Italian civil register has been read for this family. That is the largest single gap.»* Both were true on the day the archive was built. The old headline is kept on the page, struck — an archive that quietly deletes its own wrong headline is harder to trust than one that shows the correction.
121. Done — finished and published this archive
Does Antenati's name index cover BIRTHS and MARRIAGES too, or only deaths?
The single highest-value question in the work list, because the answer re-prices a dozen rows. Today's sweep queried `/search-nominative/?cognome=X&localita=Scilla` and every result came back an atto di morte, 1910–1943. That may be all Scilla has indexed, or it may be all that these four surnames happened to return. Not the same thing, and the difference is enormous: if marriages are indexed, row 71 — Nunziata Donato × Giovanni Polistena, about 1916 — is answered by a query. If births are, so is Camillo Antonio Polistena's birth act and half of row 4. Test it properly: query a surname with no locality and read the *tipologia* facet, which will say what act types exist and how many of each. ANSWERED 16 SEPTEMBER, and the answer is no. Every act in Scilla's name index, across every year from 1880 to 1943, is an atto di morte. There are no births and no marriages in it at all. Coverage is 1911–1943 solidly — between 279 and 704 indexed persons a year — with only a scatter before that: four in 1880, one in 1885, a handful each in 1902, 1905, 1906, 1908, 1909, 1910. What it costs: row 71, row 137 and half of row 4 stay volume reads. The hope that one query would answer them is gone. What it gains: the window it does cover is exactly when this family's Italian relatives were dying, which is why it gave up four deaths in an afternoon. And `anno=` is the year parameter, control-tested — `anno=1880` returns 4, matching the facet, while `estremoRemoto=` is silently ignored and returns all 13,777.
122. Done — finished and published this archive
The Scilla surnames are all swept — and the summaries cost a second sweep
DONE, AND IT WAS ALREADY DONE. This row was written when four surnames had been swept and listed eight more as wanted: Costa, Sofi, Zirilli, Martello, Russo, Paladino, Romano, Pontillo. Every one of them was run by rows 128, 130, 131 and 132 within the following two days, and this row was never closed. It has now been checked by re-running seven of them. What the re-run bought, and what it cost. It bought a NEGATIVE that answers row 131 — Mariantonia Musolino is not in the index at all, Scilla having only seven Musolino records — and four households written out into `corrections.tsv`: Giovanna Romano's three sisters, the four children of Lorenzo Costa and Generosa Polistena (including Generosa Costa's own father Antonio, d. 1 June 1915), Domenico Arlotta's wife Maria Pontillo, and the guard now standing on the two Generosa Costas (row 139). It cost seven rate-limited queries that had already been paid for once, because the first pass wrote «69 indexed persons, 26 of them the deceased» and then named eight. The rule, and it is the second time today it has been bought: when a query returns a BOUNDED set, write the set down. A paragraph is cheaper than a second search, and the second search is rate-limited.
123. Done — finished and published this archive
Generosa Costa's death — absent from the Arena index, so look under Costa
Her husband's death act of 1932 names her as his wife, so she was alive or recently dead in February 1932. She does not appear as *the deceased* anywhere in the Arena results, which is expected — she would be indexed under COSTA. Worth doing on its own because she is the last of the New York household whose end is unknown: Rocco died in 1932, their daughter Serafina in 1912, and the two Antonios and Clementina have no recorded end at all. DONE 16 SEPTEMBER. Generosa Costa died at Scilla on 6 June 1916 — «Padre ANTONINO COSTA, Madre ANNA ARLOTTA, Coniuge ROCCO ARENA». She died sixteen years before her husband, who followed on 8 February 1932. Her parents are exactly those the 1888 marriage act gave, now from a second and independent document. And her father's own act came with it: Antonio Costa, died 1 June 1915, «Padre LORENZO COSTA, Madre GENEROSA POLISTENA» — two new ancestors on a line that stopped dead at him, and the reason for his daughter's name.
124. Done — finished and published this archive
Read the four acts the index only summarises
An index line is not an act — this archive says so about NAA item titles and the same rule applies here. Four acts are now identified and none has been opened: Rocco Arena, 8 February 1932; Giovanni Arena, 31 March 1936; Michele Donato, 10 May 1936; Camillo Antonio Polistena, 19 October 1924. Each would give an age — which dates a birth — plus an address, a trade, and the two declarants. Camillo Antonio's is the most valuable: his age would date the birth this archive has only as «about 1835». ONE OF THE FOUR READ, 16 September — Camillo Antonio Polistena, and it was worth it. His age is eighty, so he was born about 1844, not 1835: the archive's figure was inferred from his son's birth and is nine years out. He was a pescatore, he died at fifty minutes past midnight at Rione Monacina, and the act calls his parents *«dal fu Giuseppe»* and *«dalla fu Pontillo Anna»* — both already dead — and him *«marito di Romano Giovanna»*. A witness is «ARENA ROCCO FU FILIPPO, di anni trentasei» — not this family's Rocco Arena, who was sixty-three in 1924, but the one who died in 1938. Exactly the namesake row 132 warns about. The method is now cheap: the volume's own *indice annuale* is the last image, it gives the act number, and the acts run two per image — three downloads instead of seventy-five. Still to read: Rocco Arena 1932 (`an_ud16599915`), Giovanni Arena 1936 (`an_ud16600296`), Michele Donato 1936 (`an_ud16600302`). TWO OF FOUR READ. Rocco Arena, Scilla Morti 1932 parte I atto 9: he died at Via Prof. Zagari at 18:15 on 8 February 1932, a pescatore, «vedovo di Costa Generosa» — a widower of fifteen years and eight months. Both parents and the wife match the archive exactly. One tension recorded rather than smoothed: the act says *settantuno*, seventy-one, which would put his birth before February 1861 — but his birth act is Nati 1861 atto 228, late in the year, and the 1900 US census says September 1861, making him seventy. The birth act wins; a death-act age is what two elderly neighbours told the registrar. Still to read: Giovanni Arena 1936 (`an_ud16600296`), Michele Donato 1936 (`an_ud16600302`). ALL FOUR READ, 16 September. Giovanni Arena died at Via Grotte 72 on 31 March 1936, a marinaio, aged eighty-one — exact against his birth of 30 April 1854 — «marito di Russo Concetta», so his wife outlived him. Michele Donato died at Via Grotte 66 on 10 May 1936, also a marinaio, aged eighty. That is a correction: the tree has Michele born in 1851, which would make him eighty-five here and fifty-seven on the 1908 New Orleans manifest that calls him fifty-two. Both records point at 1855–56 and the tree is five years out. And the four acts together produced something none of them says alone: Giovanni Arena at Via Grotte 72, Michele Donato at Via Grotte 66, Fortunato Arena born at 70, Giuseppe Arena at 75, Maria Donato at 21, Giovanni Polistena at 47. Three of the four Scilla quarters lived on one street — which is how they kept marrying each other, and it was sitting in the archive unnoticed.
125. Done — finished and published this archive
Serafina Arena's death act, 1912 — does it record that she was born in America?
A small question with a large answer if it goes the right way. She was baptised at Our Lady of Loretto, Manhattan, on 15 August 1898 and died at Scilla on 15 August 1912. If the Scilla act names her birthplace as New York, that is the return migration written down in an Italian register — the only document here that would state it outright rather than by inference from three absences. If it simply says Scilla, that is worth knowing too, because the index's «Nascita: Scilla» field may be a default rather than a claim. ANSWERED 17 SEPTEMBER, and the answer is that it is a different child. The act says «di mesi tre, residente in Scilla, nata in questo Comune» — three months old and born at Scilla. So this archive's row about her was wrong and is retracted. The Serafina baptised at Our Lady of Loretto on 15 August 1898 did not die on her fourteenth birthday; the matching dates were a coincidence and I built on it. What the act found instead is a child nobody here knew of: Rocco Arena and Generosa Costa had a second Serafina, born at Scilla about May 1912 in the Quartiere Chianalea, dead at three months — the name given again fourteen years after the first, exactly as they did with their two sons named Antonio. And it answers the row's real question better than a birthplace line would have: the household was at Scilla and *still having children* in 1912. That is the return from America seen from the inside. The New York Serafina has no known fate again.
126. Done — finished and published this archive
Domenico and Anna Arlotta died two days apart — read both acts
14 and 16 March 1915, brother and sister, both children of Giovanni Arlotta and Maria Costa. Two deaths in one family in one week is not ordinary and the acts will say why: an epidemic, a fire, or a boat. Scilla is a fishing town and the sea takes siblings together. Whatever the cause, both acts carry ages, which would date Serafina Arlotta's siblings and bracket the marriage row 39 wants. DONE 17 SEPTEMBER, and the row's premise was wrong. Both acts are read. Neither gives a cause — Italian civil death acts do not record one. And the arithmetic removes the mystery: Scilla registered about 125 deaths in 1915, one every three days, so *two deaths in a week is the ordinary rate*. A third, Arena Filippo on 17 March, makes three in four days and means nothing either. I called it «not ordinary» and it was. What the acts do give is better than a cause. Both name «da fu Giovanni e da fu Costa Maria», so the sibling link — and the whole cousin chain resting on it — now stands on the acts rather than an index line. Their ages date them: Anna about 1835, Domenico about 1842, against Serafina's documented 13 June 1830. Domenico was a marinaro, husband of Maria Pontillo; Anna a casalinga, wife of Antonino Costa. And the address is the real find: both died in CHIANALEA, the fishing quarter — where Rocco Arena and Generosa Costa's infant daughter had died three years before. An eighty-year-old and her brother of seventy-three, same quarter, two days apart, is most simply one household and one illness. The register will not say so and neither will this archive.
127. Done — finished and published this archive
Settle whether Rocco Arena married his first cousin — one act
**Row 118 grades it *probable*; this is the act that would make it *certain*. Either Anna Arlotta's death act of 16 March 1915, which gives her age and parents in full, or Generosa Costa's own birth act, which would name her mother with a patronymic. Both are inside the surviving run.** What is being ruled out is a second Anna Arlotta married to a second Antonino Costa — unlikely, and *unlikely* is not *ruled out*. SETTLED 16 SEPTEMBER, and not by reading an act — by the Costa side of the index falling into place. Four acts now agree and none contradicts. Anna Arlotta's death of 16 March 1915 names her parents as Giovanni Arlotta and Maria Costa, who are Serafina Arlotta's parents, and her husband as Antonino Costa. His own death of 1 June 1915 names her back. Generosa Costa's death of 6 June 1916 names that same couple as her parents. And the 1888 marriage act gave them in the first place. **Row 118 is upgraded from *probable* to *certain*.** What would still break it — two different couples named Giovanni Arlotta and Maria Costa at Scilla in the same decades — would have to be pointing four mutually consistent acts at the wrong pair.
128. Done — finished and published this archive
Francesco Antonio Arena and Serafina Arlotta — both found, in the same house
DONE, 20 SEPTEMBER 2026, AND THEY DIED AT THE SAME ADDRESS. FRANCESCO ANTONIO ARENA died 29 August 1901 at seven ten in the morning — *Morti 1901, parte I, atto 96* — a *pescatore* of eighty-two (eighty-one and nine months in fact), son of the late Filippo and the late Domenica Costa, husband of Serafina Arlotta. SERAFINA ARLOTTA died 24 April 1908 at four in the morning — *Morti 1908, parte I, atto 55* — a *filatrice* of seventy-seven (exact), daughter of the late Giovanni Arlotta, *pescatore*, and the late Maria Costa, *tessitrice*, and widow of Arena Antonio. Both at Via Grotte number two, Scilla, six years and eight months apart. She kept the house and died in it, with Rocco in America and Giovanni home from it. Twelve annual indexes read, 1897 to 1908. 1909 is not on Antenati at all (row 129) and did not need to be. THE CATCH THAT NEARLY HID HER. The 1908 index enters her as «Arena Serafina, fu Giovanni» — her *husband's* surname over her *father's* forename — while the act says Arlotta. Eleven A blocks were swept for an Arlotta and she was found only because Arena is also an A. When sweeping an index for a married woman, read her husband's initial too. Six Arena leads came out of the sweep and none is read — row 144.
129. Next — ready to start this archive
The two Scilla death volumes Antenati does not hold — 1888 and 1909 (no longer blocking)
Found by listing the series rather than by failing to find someone in it, which is the better way round. The Morti run is complete from 1866 to 1943 except for 1888 and 1909 — and 1909 sits inside the thirteen-year window where Francesco Antonio Arena and Serafina Arlotta must have died. This is the same shape as the Zambrone gap: a hole in the digitisation, not necessarily in the record. The second copy is the route — these registers were kept in duplicate, one at the comune and one at the Tribunale, and it is the Tribunale copy Antenati filmed. Belongs with row 9. AND THE INDEX IS NOT A BACK DOOR INTO THEM, checked 16 September. The name index lists two acts under 1909, which looked at first like content from a volume browse cannot reach. It is not: their document arks — `an_ud16595370` and `an_ud16595649` — fall inside the ark ranges of the 1911 and 1912 volumes. They are 1909 deaths written into later registers, which is ordinary. The 1888 and 1909 volumes are genuinely absent and the second copy remains the only route. THE URGENCY IS GONE, 20 SEPTEMBER. This row mattered because 1909 sat inside the window in which Francesco Antonio Arena and Serafina Arlotta had to have died. They did not die in it. He died 29 August 1901 and she 24 April 1908, both found in volumes Antenati does hold (row 128). The two holes are still real and still worth recording as holes — 1888 is the year of Rocco Arena's own marriage, so a death in his household that autumn would be unreachable — but nothing this archive is currently looking for is behind either of them. Downgraded from a blocker to a note about the digitisation.
130. Done — finished and published this archive
List the Scilla death series BEFORE 1866 — rows 68 and 85 cannot start without it
`scilla-arks.tsv` holds 55 Nati and 54 Matrimoni volumes for 1816–1865 and NOT ONE Morti volume. The restaurazione death series has never been listed, and two rows depend on it entirely: row 68, Filippo Arena and Domenica Costa, who died most likely in the 1850s–70s, and row 85, Domenico Arlotta, dead before April 1854. One page load does it — the same `s_size=100` trick that returned all 81 post-1866 volumes this afternoon. Do it before either row is attempted, for the reason row 17 gives about mapping before sweeping. DONE 16 SEPTEMBER: 51 volumes, 1817–1865, in one page load, now in `scilla-arks.tsv`. The deaths begin in 1817, a year after the births — so anyone reasoning «Scilla's civil registration begins in 1816» about a *death* is reasoning one year wrong. 1846 and 1854 are doubled. The reusable part is the series id: `serie=2002440` carries all three act types for the restaurazione period and is filtered by `tipologia=`, exactly as `16594786` does for the italiano period.
131. Next — ready to start this archive
Orazio Arena's second wife is not in the index — the marriage act is the only route
A whole second marriage, found as a by-product. The Scilla death index gives Orazio Arena four children by Giovanna Zirilli — Teresa (d. 1913), Giuseppe (d. 1922), Maria Annunziata (d. 1933) and Alessandro Antonio (d. 1936) — and a fifth, DOMENICO ARENA (d. 30 December 1923), whose mother is MARIANTONIA MUSOLINO. Musolino is a name this archive has never held. Either Orazio married twice or there are two Orazio Arenas at Scilla — and this archive already knows this man had *two sons both named Domenico*, which is exactly the situation where a second wife gets overlooked. The marriage act would settle it. CHECKED 20 SEPTEMBER, AND SHE IS NOT THERE. The name index holds seven Musolino records at Scilla in total, of which two are the deceased — Domenico Musolino (d. 17 April 1911, son of Bruno Musolino and Grazia Longordo) and Francesca Musolino (d. 7 November 1928). Neither is Mariantonia. She exists in the index only as a name inside her son Domenico Arena's act of 1923. So she died outside 1911–1943, or away from Scilla, and the index route is closed. What is left is Orazio Arena's second marriage act — Scilla's civil marriages 1866–1913 are all on Antenati, his first wife Giovanna Zirilli bore children to 1860, and Domenico Arena (d. 1923) is the only known child of the second marriage, so the wedding sits somewhere after 1866. The annual marriage indexes are one A-block read apiece.
132. Done — finished and published this archive
The two Antonio Arenas — the guard is written
A collision to guard before it does damage, in the spirit of row 97. This family's FRANCESCO ANTONIO ARENA married SERAFINA ARLOTTA, and the index renders him simply «Antonio» on his sons' death acts. But the same index holds an ANTONIO ARENA married to a SERAFINA DE LUCA — parents of Grazia (d. 1913), Anna (d. 1917) and Annunziata (d. 1919). Same given name, same wife's given name, same town, same decades. Anyone sweeping «Antonio Arena + Serafina» will merge two families. The mother's surname is the discriminator and it must be checked every time — which is exactly how two Arena namesakes were excluded in New York this week. WRITTEN, 20 SEPTEMBER. `corrections.tsv` now carries the separation, graded certain. It names the discriminator — Arlotta or De Luca, read every time, never inferred — and it records that this archive has already sprung this trap once: the Orazio Arena who died at Scilla in August 1895 was taken for the tree's Orazio, the supposed father of Rocco. He is the son of **Antonio Arena and Serafina *De Luca*, a different household entirely. The second family is now fully drawn from the index: three more children, Grazia (d. 14 February 1913), Anna (d. 8 March 1917, wife of Rocco Paladino) and Annunziata (d. 12 April 1919). And the collision is worse than the row knew, because the whole Arena surname at Scilla resolves into at least seven households — see the map filed under the name index. Antonio × Serafina is one collision in a field of them. AND NOW IT IS A DOCUMENT, NOT A GUARD. The 1828 annual marriage Tavola names all four parents: «Arena Antonio e De Luca Serafina — Orazio e Morabito Letteria / Filippo e Alfonzetti Anna — 10 Ag.o 1828 — atto 26». So the other Antonio Arena married his Serafina in August 1828 and was the son of Orazio Arena and Letteria Morabito, while this archive's Francesco Antonio married his in February 1851 as the son of Filippo Arena and Domenica Costa. Different decade, different parents, different wife's family. And it explains the name that caused the trouble: their son Orazio, who died at Scilla in 1895 and was once taken for the tree's Orazio, is named for his own grandfather** — the custom working exactly as predicted.
133. Done — finished and published this archive
PIEDIMONTE ETNEO — is it name-indexed? The Crupi are from there too
The NAA card gave Mario Crupi's birth as 28 November 1905 at Piedimonte, province of Catania — the Mazza's own town. So the Crupi and the Mazza came from one village, and Angela Crupi's birth of 20 December 1900 with her parents Francesco Crupi and Maria Vecchio should be reachable in the registers the Mazza line already uses. First test whether Piedimonte Etneo is in the name index at all — if it is, two birth acts and a set of parents come back for two queries. If it is not, the acts are still there to be read by volume, with exact dates in hand, which is the cheapest register work there is. ANSWERED 16 SEPTEMBER: Piedimonte Etneo is NOT name-indexed. «Nessun risultato trovato» on a locality-only query. Angela Crupi's birth of 20 December 1900 and Mario Crupi's of 28 November 1905 are not reachable this way — but the NAA card gave both dates to the day, so the volume read is as cheap as a register read ever gets: a known comune, a known year, a known date.
134. Done — finished and published this archive
Are Briatico, Zambrone and Pizzo name-indexed?
The Calabrian half of this family has been worked entirely by volume-reading, and it may not have needed to be. Scilla turned out to be name-indexed for deaths 1910–1943; nobody has checked whether Briatico, Zambrone, Pizzo or San Costantino are. If any is, several open rows collapse into queries — row 84 (Caterina Cichello's death), row 44 (the Melluso), row 66 (Maria Fragalà's death at Pizzo). One paced fetch per comune answers it. Do it before any further Calabrian volume sweep. ANSWERED 16 SEPTEMBER, and mostly no. Briatico, Zambrone, San Costantino Calabro and Potenzoni are NOT indexed — all four return «Nessun risultato trovato». Pizzo IS, but for births, about 5,900 of them across 1868–1903 — and it holds no Prostamo, which is expected since the family was at Briatico by then. So row 84, row 44 and row 66 all stay volume reads. The important by-product: Pizzo being indexed for *births* while Scilla is indexed for *deaths* proves the act type varies by comune, so «deaths only» was never an Antenati fact.
135. Done — finished and published this archive
Build the name-index coverage map: comune × act type × years
The same argument as row 17, one level up, and it will pay better. Row 17 wants to know which Scilla volumes carry an annual index; this wants to know which comuni and which years Antenati has indexed by name at all — because where it has, the volume index does not matter. Today's work was done blind: the coverage was discovered by querying four surnames and noticing every hit was a death between 1910 and 1943. One table, built once, and every later question becomes «is it indexed» before «which volume». FIRST PASS DONE 16 SEPTEMBER — the eight comuni this family touches. Indexed: Scilla (deaths, 1911–1943), Pizzo (births, 1868–1903), Catania (deaths, 1911+). Not indexed: Piedimonte Etneo, Briatico, Zambrone, San Costantino Calabro, Potenzoni. Two rules fall out of it. The act type varies by comune — assuming Scilla's «deaths only» held everywhere would have been wrong at Pizzo. And `localita=` does not mean «registered here»: it matches any place attached to the person, and `localita=Catania` returns people who died at Vicopisano in Tuscany. It read cleanly at Scilla only because those people were born and died in one town. Still to do: turn this into the actual table, and extend it past this family's comuni if the estate wants it. FOLDED INTO THE HOLDINGS TABLE, 21 SEPTEMBER. The name-index map is published beside the volume holdings on What Has Been Read, because the two questions are one question asked at different heights and a reader should not have to hold both pages open. Indexed: Scilla deaths 1911–1943, Pizzo BIRTHS 1868–1903, Catania deaths 1911 onward. Not indexed: Piedimonte Etneo, Briatico, Zambrone, San Costantino Calabro, Potenzoni. The act type differs between comuni and that is the trap — assuming Scilla's «deaths only» held everywhere would have been wrong at Pizzo, where it is births.
136. Done — finished and published this archive
The Loveday records — Mario Crupi's internment, from the camp's own side
His aliens card says «18.8.42 LOVEDAY INTERNMENT CAMP / 30/12/43 Released and entrained for Brisbane» and that is all it says. The NAA holds the camp's own series — nominal rolls, internee service and casualty forms, and the objection and appeal files that record why a man was interned. The why is the part worth having: north Queensland Italians were interned on the word of neighbours, and the files carry the allegation and sometimes the appeal. Rosario Mazza's dossier (MP1103/1 Q8110) should be worked the same way, and row 7's Trove search now has two names and two dates to search on. DONE 16 SEPTEMBER, and the «why» turned out to be a different question than expected. His service and casualty form (MP1103/1, Q30124) is digitised and gives the whole arc on one sheet: taken at his own address, Bemerside, Ingham, on 20 April 1942; Gaythorne, then Cowra on 5 May, then Loveday on 17 August 1942; a tribunal in January 1943; released 29 December 1943 — twenty months. And the record that makes it bite is the other one. NAA A714, 9/4300: he was naturalised on 26 April 1940, two months before Italy entered the war. It did not protect him — the casualty form still records his nationality as «Italian (Nat.)». It also proved something this archive had refused to claim. The 1959 roll put a Mario and a Maria Crupi at Tully with both address columns blank, and I declined to pair them on adjacency. The next-of-kin line pairs them properly: *«Maria CRUPI … Relationship: Wife»*. And it dates the marriage — he is *Single* in April 1940 and married by April 1942. Five files remain unopened: BP242/1 Q25743 (his Queensland investigation case file, which is where an allegation would be), A435 1946/4/2800, D1901 C594, MP1103/2 Q30124 and J25 1955/16646.
137. Done — finished and published this archive
Camillo Antonio Polistena — born 21 September 1844, found
Row 96 closed on his death act, which gave his parents. His birth act would give the arithmetic a check and the parents a second witness. Scilla's births begin in 1816, so a birth about 1835 is inside the run. What makes it worth doing now rather than before: we know the parents to look for — Giuseppe Polistena and Anna Pontillo — so the read is a confirmation with a known answer, which is the kind that catches a wrong identification. His age at death in 1924 will narrow the year first, which is row 124. THREE YEARS READ ACROSS, 20 SEPTEMBER, AND HE IS IN NONE OF THEM. His son's 1883 birth act makes him forty-eight, so about 1835; the complete P section of the annual Tavola was read for 1834, 1835 and 1836 — 629 births — and there is no Camillo Antonio Polistena. The shape of the absence is the finding. Those 629 births hold exactly one Polistena. The P section at Scilla in the 1830s is Paladino, Pirrotta, Patafio, Papalia, Pizzarello. The Polistena are barely there, which may be why this quarter stops where it does — and *Polistena* is itself a comune in the province of Reggio Calabria. The one entry is a lead with a flaw. Entry 191 of 1836: *«Polistina Annunziata, di Giuseppe e Martello Anna, 2 Dicembre»*. Right given names for his parents, wrong surname for the mother — the death act says Anna Pontillo. And four entries earlier stands *«Pontillo Giovanni, di Giuseppe e Paladino Giuseppa»*: a clerk copying that column had both surnames under his hand. Act 191 of 1836 settles it and is three images away. Then widen. A declared age is a range, and this archive has already been burned reading three points inside a window as a search of the window. 1830–1833 and 1837–1840 are next; the containers are not bought. THE LEAD WAS OPENED THE SAME HOUR AND IT PAID. The act is 204, not 191: that Tavola's *«Num. d'ordine»* is an index row and the index carries no act number at all — act 191 is a Squillaci birth at Solano. The acts run in date order, so the birth date is the only key, and two probes found it. Scilla, atto 204 of 1836. On 2 December 1836 the midwife Teresa Arena, forty-four, declared a girl born that morning at eight, in her parents' own house, to ANNA PONTILLO, thirty-five, lawful wife of **GIUSEPPE POLISTINA, thirty-eight, *marinaro* — and named Annunziata Polistina. So the index was the liar and the death act was right. The mother is Pontillo, not Martello, and Annunziata is Camillo Antonio's sister** — the first sibling he has ever had in this archive. The guess about *why* the index erred held too: four entries earlier in the same P section stands *«Pontillo Giovanni, di Giuseppe e Paladino Giuseppa»*. And the Polistena quarter gets its first two dates. Giuseppe Polistena born about 1798, a marinaro — a third Scilla sailing family beside the Arena and the Arlotta — and Anna Pontillo about 1801. Camillo Antonio's own birth is still not found. With his parents dated, a son born about 1835 fits them exactly; the years still to read are 1830–1833 and 1837–1840. SIX CONSECUTIVE YEARS NOW, 1833–1838, AND HE IS IN NONE OF THEM. The complete P section of every annual Tavola. But his parents are in two. Rocco Polistina, 13 December 1833 — *«Giuseppe e Pontillo Anna»*, the index spelling her correctly this time — and Annunziata Polistina, 2 December 1836. The three years between hold no Polistena birth at all, and they are not thin years: 201, 206 and 222 births. A child born to this couple about 1835 is simply not there. So «forty-eight» in September 1883 is out by at least two years, and he was born before 1833 or after 1838. The search moves to 1830–1832 and 1839–1840, and it should not stop at the first year that fits. One thing to carry: the 1837 Tavola has *«Pirrotta Salvatore — Rocco, e Polistina Annunziata»* — an adult Annunziata Polistina bearing children two months after the infant one was born. The baby was very likely named for her, and she is a relative this archive has not identified. 1839 ADDED: SEVEN CONSECUTIVE YEARS NOW, AND A SECOND POLISTENA HOUSEHOLD. Entry 153 of 1839 is *«Polistina Maria — Francesco e Longardo Giuseppa — 30 Maggio»* — not Giuseppe and Anna Pontillo, so not a sibling. Seven years, about 1,450 births, four Polistena in total: Rocco (1833) and Annunziata (1836) to this archive's couple, Maria (1839) to the other, and nothing else. With a surname that scarce the temptation to join any two bearers is strong and must be resisted — the discriminator is the mother. Guard written. What is left: 1830, 1831, 1832 and 1840 onward. Containers bought for 1832, 1840 and 1839; 1830 and 1831 are not. The naming custom argues he is older, not younger: Rocco is the eldest son found and under Calabrian custom the first son takes the father's father's name, which would make Rocco Polistena the grandfather — leaving *Camillo Antonio* as a second son named for the mother's father, i.e. a Camillo Pontillo. That reading puts him after December 1833, which the register denies. So either the custom is not being followed here, or he is the elder and 1830–1832 is where he is. 1832 ADDED: EIGHT CONSECUTIVE YEARS, 1832–1839. Entry 152 of 1832 is *«Polistina Domenica — Francesco e Longordo Giuseppa — 17 Dicembre»* — the other household again. Eight years, about 1,630 births, five Polistena in total: Domenica (1832) and Maria (1839) to Francesco and Giuseppa Longordo; Rocco (1833) and Annunziata (1836) to Giuseppe and Anna Pontillo. Two households running in parallel, and between them they account for every Polistena birth in eight years. Camillo Antonio is not one of them. So he was born in 1830 or 1831, or in 1840 or later, and «forty-eight» in September 1883 is out by at least three years. Containers bought: 1832, 1839, 1840. Not bought: 1830, 1831. Take 1840 first — it is in hand and costs nothing but images. FINISHED THE WINDOW: ELEVEN CONSECUTIVE YEARS, 1830–1840, AND HE IS IN NONE OF THEM. 1840 gives a third sibling — *«Polistina Fortunata, di Giuseppe e Pontillo Anna, 28 Gennaro»*. So this archive's couple have Rocco (Dec 1833), Annunziata (Dec 1836) and Fortunata (Jan 1840) — three years apart, twice — with 1830, 1831, 1832, 1834, 1835, 1837, 1838 and 1839 all read and empty of them. So Rocco is the eldest and Camillo Antonio cannot be earlier. He must be 1841 or later, and «forty-eight» in September 1883 is wrong by at least six years. The naming custom agrees: a first son takes the father's father's name — making Giuseppe's father a Rocco Polistena — and a later son the mother's father's, which would make Camillo Antonio the namesake of a Camillo Pontillo. Both point forward. The eleven years also map the surname. Eight Polistena births in about 2,100 children, across three households: Giuseppe × Anna Pontillo, Francesco × Giuseppa Longardo, and **Luigi × Marianna Baviera — whose son born 21 June 1831 is called *Francesco Antonio Polistina*, one word from our man, right town, right decade, different parents. Guard written. Next: 1841–1848. Anna Pontillo was about thirty-five in 1836 and could bear to about 1846. No container for any of those years is bought. THE NEW WINDOW IS STARTED: 1841 AND 1842 ARE CLEAR. 1841 holds no Polistena at all**; 1842 holds one, *«Polistina Salvadore, di Domenico e Costa Caterina»* — a fourth household. 1843 is a doubled year and the index is in the other ark. `an_ua2002874` holds the acts and closes at act 200 with no Tavola; `an_ua2002877` is where the *indice* will be — the same split this archive met in 1852 and 1875. Its container lookup returned 403, caught by the control test and recorded as a block. AND THE METHOD ITSELF HAS A HOLE IN IT. Entry 160 of 1841 reads *«Palmisano Gio. Battista — Giuseppe e Anna Pontillo — 26 Agosto»*. A Giuseppe married to an Anna Pontillo, right town, right decade, wrong surname. The way these Tavole are being read is by scanning the *parents* column for «Giuseppe e Pontillo Anna» — which walks straight into this entry, because what separates them is the child's surname, the column a parent-scan skips. Guard written. Still to read: 1843 (needs the second ark), 1844–1848. AND ROCCO POLISTINA'S ACT IS READ — atto 174 of 1833. It confirms the parents from a second independent document (**Giuseppe thirty-four, *marinaro*; Anna thirty-three, against thirty-eight and thirty-five in 1836 — both landing on Giuseppe b. 1798–99 and Anna b. 1800–01), and it shows the same midwife, Teresa Arena, delivered both children. Its margin carries a marriage added thirty-six years later:** *«L'anno 1869, il giorno 12 Luglio … ha contratto matrimonio in questo Comune con Colorina Bri[…]»*. Rocco lived to thirty-five and married at Scilla on 12 July 1869. That opens row 154 — these acts annotate themselves, and nobody here has been reading the margins. 1843, 1844 AND 1845 ADDED — SIXTEEN CONSECUTIVE YEARS NOW, 1830 TO 1845. 1843: no Polistina child (one as a *mother*, the third sighting of the adult Annunziata). 1844: two Polistina, both from the other households — Domenico × Costa Carolina, Francesco × Longordo Giuseppa. 1845: partial — seven of eighteen lines in the P block are too faint to read, and that is recorded as partial rather than as a negative. The family's shape is now fixed. Giuseppe Polistena and Anna Pontillo have children in December 1833, December 1836 and January 1840 — and in the twelve other years read, nothing. Anna was about thirty-five in 1836, so her childbearing plausibly ends around 1840–42, which is exactly what the register shows. Which leaves a contradiction worth stating rather than smoothing. His 1924 death act names these two as his parents; his son's 1883 act makes him forty-eight. There is no year in their childbearing window where he fits, and the window has been read across. Four readings, none preferred: he was born elsewhere and the family came in; the death act's parents are wrong; the Tavola omitted him (these indexes *do* drop entries); or he is a child already found under a name nobody has connected. The test that separates them is row 155. THE SEARCH WAS AIMED AT THE WRONG DECADE, AND HIS MARRIAGE ACT SAYS SO (row 155). He was thirty-six on 23 January 1881, so born about 1844 — not 1835. Eleven years of Tavole were read around a figure that was nine years out. He was born at Scilla, on his own sworn word, so the act exists. And 1844 and 1845 are both already in hand and both incompletely read: 1844's P block was covered for entries 127–138 without establishing where it ends, and 1845 is recorded as partial with seven of eighteen lines too faint. Go back to those two years first. FOUND. 1844 TAVOLA, ENTRY 142: «Polistina Camillo — Giuseppe e Pontillo Anna — 21 Settembre». Born at Scilla on 21 September 1844. The arithmetic is exact, which is what makes it safe. Thirty-six years and four months at his marriage on 23 January 1881 — the act says *trentasei*. Thirty-eight years and eleven months at his son's birth on 16 September 1883 — that act says forty-eight, ten years wrong, and it is the figure this row searched around. And the reason it took sixteen years of sweeping is mine. The 1844 P block was read earlier the same day; the crop covered entries 127–138 and stopped, and this row was written saying the end of the block had not been established. He is four entries past where I stopped. Instead of one more crop on the year already open, I swept 1845, 1830–1832 and five marriage years. The rule it buys: *a partial reading recorded as a negative is the same error as a 403 recorded as a nil.* Where a section is read, find its end — the next section letter — before writing «not there». 1845 is still marked partial and should be finished on the same principle. A fifth Polistena household came with it: entry 143 is *«Polistina Tommaso — D. Antonio e Paladino Annunziata»*, the *D.* being Don. Next, and cheap: the act itself — row 156.
138. Next — ready to start this archive
Donato–Polistena — four ties, and the first one now has a date
The index turned up a second one and the archive only knew the first. Row 71 is about Nunziata Donato × Giovanni Polistena, about 1916. But the Scilla deaths also give MARIA DONATO (d. 3 April 1911), «Padre Candeloro Donato, Madre Domenica Romano, Coniuge GIUSEPPE POLISTINA» — a Donato woman married to a Polistena man a generation earlier. Her father is a Candeloro Donato, which is the name of both this archive's wall and Domenico Donato's son of 1847. Two families that intermarried twice are usually close to begin with, and the earlier act would say how. MORE THAN TWO, CHECKED 20 SEPTEMBER. Writing the Polistena and Donato index sets out in full turns two ties into four: (1) Maria Donato, d. 3 April 1911, «Padre Candeloro Donato, Madre Domenica Romano, Coniuge Giuseppe Polistina» — the one this row was written for. (2) Raffaele Polistena × Teresa Donato, whose three children all die at Scilla — Santa (12 July 1926, wife of Rocco Catalano), Maria (19 July 1927, wife of Camillo Matia) and Luigi (8 December 1933, husband of Annunziata Matrà). (3) Giovanni Polistena × Annunziata Donato, whose daughter Teresa dies 5 January 1933. (4) Nunziata Donato × Giovanni Polistena, about 1916 — row 71. And a fifth pattern worth naming: Rocco Polistena, d. 7 August 1923, is the husband of a Nunziata Polistena — a Polistena married to a Polistena, the same surname endogamy this archive has already met among the Bueti and the Vizzari. Two families that intermarry four times in two generations are one neighbourhood, and the Romano are in it too: Maria Donato's mother is a Domenica Romano, and Giovanna Romano's sister Concetta married a Rocco Donato. SIX, NOT FOUR — 20 SEPTEMBER. The National Archives adds two more, in Queensland: NAA J25 1967/9873, *«Polistena, Giuseppe; wife Maria Maddalena [nee Donato]»*, and NAA J25 1970/3023, *«Donato, Concetta [nee Polistena]»*. Two families that intermarry six times over two generations and two hemispheres are not two families who happened to meet. They were one neighbourhood at Scilla — Via Grotte — and they were still one neighbourhood in Brisbane. ONE OF THEM IS DATED, 20 SEPTEMBER. Scilla, Matrimoni 1882, atto 48: LUIGI CARMELO POLISTINA × ANNUNZIATA MATRÀ, 23 November 1882 — and he is *«figlio di Raffaele e di Teresa Donato»*, matching the name index's *«Luigi Polistena, d. 8 December 1933 — Padre Raffaele, Madre Teresa Donato, Coniuge Annunziata Matrà»* across fifty-one years. Note which generation it dates. Teresa Donato is the mother here, so the Donato–Polistena join in this line happened a generation earlier than this row had it — Raffaele Polistena × Teresa Donato, before about 1852, is now a named target. And both of them were alive in November 1882, which is the first date either has ever had.
139. Done — finished and published this archive
The two Generosa Costas — the guard is written
Rocco Arena's wife was Generosa Costa. So was Pasquale Donato's. The death index gives PASQUALE DONATO (d. 20 April 1937), «Padre Domenico, Madre Annunziata Sofi, Coniuge GENEROSA COSTA» — and Pasquale is Michele Donato's brother, so this is inside the family, not a stranger who happens to share a name. That is exactly the circumstance in which two women become one in a tree. They are almost certainly different: Rocco's Generosa was born about 1868 and married in 1888. Write the distinction down now, while it is obvious, rather than discovering the merge later. WRITTEN, 20 SEPTEMBER. `corrections.tsv` now carries the separation explicitly, against `generosa-costa`, and it is graded certain rather than inferred, because the two women are separated by evidence and not by assumption: Rocco's Generosa was already dead on 6 June 1916, twenty-one years before the other's husband died in 1937, and their parents differ — Antonino Costa and Anna Arlotta for the first, unknown for the second. A third Generosa sits a generation above them: GENEROSA POLISTENA, Antonino Costa's mother, for whom Rocco's wife was almost certainly named. Three women of one name in three generations of one town, and the guard names all three.
140. Done — finished and published this archive
Tell the estate that Antenati has a name index
This is not a Mazza finding, it is an estate finding, and the other archives are working Italian registers too. `/search-nominative/?cognome=X&localita=Y&s_size=100` returns acts with father, mother and spouse already extracted, and for Scilla it covers deaths 1910–1943. Falco and Defranceschi have been reading volumes the same way this archive has. Worth a note carrying the query form, the `localita=` caution — `fq=` and `facets=` are silently ignored and return the unfiltered count — and the parsing rule that only rows with a `p.morte` value are the deceased; the rest are relatives seen from inside someone else's act, and reading them as deaths invents people. SENT, 21 SEPTEMBER, to the landing site for circulation, with the query form and all three cautions — and with the one this row did not have: the act type differs by comune. Scilla is indexed for deaths 1911–1943 and Pizzo for BIRTHS 1868–1903, so an archive carrying «deaths only» across from one to the other would produce a confident negative on a series that is indexed the other way round. Five of the eight comuni this family touches are not indexed at all, so the annual Tavola remains the instrument for those — and that is written up on /method/ beside it. AND THE FORM WAS TESTED BEFORE IT WAS ALLOWED TO STAND. The Booyzen session's rule, arriving the same hour: *«a method note is code, and an untested method note is untested code»* — they had recorded archive.org identifiers with a suffix that does not exist and every id in their own note returned nothing. This archive had just sent a query string to three other archives. It was re-run: it works, and the `p.morte` caution is now confirmed by example rather than asserted — the index returns *«Rocco Arena — Figlio Serafina Arena — Atto di morte 15 Agosto 1912»* beside *«Serafina Arena — Morte: Scilla 15/08/1912 — Padre Rocco Arena»*. The dead child is Serafina. Reading the first row as a death would have killed Rocco Arena twenty years early, with a plausible date.
141. Blocked — waiting on a person or an institution David
The two Queensland investigation case files — why these men were interned
The casualty form says what happened to Mario Crupi. These files would say why. NAA: BP242/1, Q25743 — «Crupi, Mario — Queensland investigation case file», 1942–1948, Brisbane, Box 152. And NAA: BP242/1, Q21227 — the same series for Rosario Mazza, 1940–1948. Both are Access status OPEN, so neither needs an access examination — unlike row 115. They are simply paper, in Brisbane, undigitised, and someone has to ask for them. Why they matter more than the camp records. North Queensland Italians were interned largely on the word of neighbours and informants, and it is the investigation file that carries the allegation, the evidence offered for it, and sometimes the appeal. Mario Crupi had been naturalised since April 1940 and was taken anyway; this is the file that would say on what grounds. Michael is in Brisbane, where both files are. A reading-room visit would open both in an afternoon, or the NAA will quote for copies.
142. Done — finished and published this archive
The spine still calls Antonino Prostamo untested, and additions.json says he is proved
Two files in this repo disagree about the same man. `spine.json` gen 6, `antonino-prostamo-1785`, is drawn on /direct-line/ as a joint that will not bear weight — one of the four above 1785 said to rest on the family tree alone. `additions.json` says the opposite: «The fourth generation, carried as UNTESTED until 14 September 2026 and now documented four times over. A MARINARO like his father and his son.» It gives a source (Briatico, Matrimoni 1849, atto 1 — Antenati an_ua87663), a trade, and a death: 31 December 1837 at Pizzo, in the cholera, at two in the morning on the last night of the year. It also carries `inTree: antonino-prostamo-1785`, so the two files certainly mean one man.
Which is right is this archive's to say, not mine. Raised because it travels: Our-Family reads this spine to decide what a child is told, and it currently tells them that generation is only what the family says. If the spine is the stale one, the children's page is repeating a caution the archive has already retired — and the 1697 line is one generation shorter in doubt than it looks.
ANSWERED 17 SEPTEMBER, and the spine was not the stale one — the person page was. `spine.json` is right to show what it shows. It is built from Michael's export, and this archive's standing rule is that the charts draw the export as given because the corrections are proposals, not edits. So the two files are *meant* to differ. What was wrong is that the difference never reached the man. `build_person_extras.py` gathered a person's records from four places — Nudgee, the NAA index, the tree's own citations and `read-people.tsv` — and from nowhere else. It never read `documented-additions.tsv` or `corrections.tsv`, which is where most of this archive's evidence lives. So Antonino Prostamo's page carried the sentence *«No record has been read for this person. What is known of them is what the tree asserts, and it is unverified»* — about a man documented four times over. Twelve people were in that position, including Giovanna Pontillo, whose surname six records settled yesterday, and Salvatore Mazza, Sebastiano's own father. Fixed: both files are now read, keyed the way each keys a person — additions by `inTree`, corrections by `slug`. Citations go to `records`, prose to `corrections`, under the heading the page already had for it: *«What a register says about this person, which the tree does not»*, followed by its own statement that the chart above still draws the tree. Nothing is applied. The archive just stops being silent about it. Guarded: `check_links.py` now refuses the build if a person named in either file has a page carrying neither a record nor a correction. Verified against a reintroduced bug — it exits 1 and names them. And the caller was right about the reach: Our-Family reads this spine, so the children's page was repeating a caution this archive had already retired. Raised by the estate, not by us, which is the argument for the landing page existing.
143. Next — ready to start this archive
The merchant seamen series at Kew — the one thing the UK archives might hold
Raised by David, and the only part of his suggestion that survives testing. A sweep of The National Archives' Discovery API found nothing at all for this family — Prostamo, Polistena, Cichello, Zirilli, Fragalà, Briatico and Piedimonte Etneo all return zero, and the handful of Crupi, Arlotta and Melluso hits are provably other people. That is structural: this family went Italy→Australia and Italy→New York and never through Britain, and Kew's passenger series (BT 26, BT 27) cover UK ports only. But the seamen series are a different matter. BT 372 registry cards and BT 99 crew agreements record an Italian place of birth — the sweep turned up entries reading «place of birth, Napoli» and an Arlotta born at Udine. And this family is full of mariners: Giovanni Arena and Michele Donato are both «marinaio» on their 1936 acts, Domenico Arlotta on his of 1915, Antonino Prostamo «marinaro», Rocco Arena «pescatore». If any of them ever signed on a British-registered ship, they are in these series. Two cautions before anyone spends a day on it. Most nineteenth-century crew agreements are not at Kew at all — the bulk went to the Maritime History Archive in Newfoundland, and Kew keeps a sample. And a Scilla fisherman working the Strait is a far likelier thing than one signing British articles. Long shot, cheap to test, and the API makes it a query rather than a trip.
144. Next — ready to start this archive
Three Arena leads read and closed; three left
Fallout from row 128. Sweeping twelve annual death indexes for two people turned up six other Arena acts, each with a patronymic. Three are read, 20 September, and all three close. **~~ARENA GIOVANNA, *di Rocco*, 1908 parte II atto 2~~ — NOT ROCCO'S DAUGHTER. Nine, born at Scilla, dead at Marseille 1 February 1905; mother Francesca Martello. What it opened instead is row 145. ~~ARENA NICOLINA, *di Giuseppe*, 1906 atto 41~~ — NOT GIUSEPPE'S DAUGHTER. She was forty, born about 1866, when Rocco's brother Giuseppe was six. Via Grotte 17; mother the late Marianna Romano; husband Giuseppe Cesoriero. ~~ARENA COSMO, *di Filippo*, 1904 atto 122~~ — NOT ROCCO'S NEPHEW. Three months old, Vico San Michele 13, son of Filippo and Clementina Furfari. The act gave the father no age, so the plan was to sweep ten years of marriage indexes — seven of them with no container saved, a rate-limited page load each. Instead the Antenati name index, which carries the spouse, answered in one GET: Filippo Arena died at Scilla 17 March 1915, «Padre Giovanni Arena, Madre Natalina Longordo, Coniuge Clementina Furfari». A different Filippo. The order of operations is the lesson:** when the question is *whose son was this man*, ask the name index before planning a volume sweep. The three unread: Arena Raffaele *fu Giuseppe* (1901, atto 63); Arena Rocco *di Alessandro* (1906, atto 141); Arena Santo *fu Rocco* (1908, atto 108). All three now look like other households — see the seven-household map filed under the name index — so they are low priority, kept because a named act should not be silently dropped. What the three closures cost and bought: they cost nine image fetches and they bought the question on row 146.
145. Next — ready to start this archive
PARTE II — the deaths abroad, in eighty-one volumes nobody here has opened
Found by accident chasing a lead from row 144, and it is the biggest thing the Scilla death series has left to give. Every Scilla death volume has a parte II at the back, just before the annual index, and parte II is where deaths abroad are transcribed — on the order of the *Procuratore del Re* at Reggio Calabria, from certificates issued by foreign authorities, at a family's request. The 1908 parte II runs to five acts and three of them are foreign: GULLI ROCCO, 27, a *bracciante*, husband of Lofaro Maria, dead in NEW YORK on 17 August 1907 of a fractured skull — transcribed from an English-language certificate *with a translation*, at his widow's own request; ARENA GIOVANNA, nine, dead at MARSEILLE; and CACCIOLA CARMELA, seven, dead at TUNIS at ten at night. Three continents in four acts, in a town of a few thousand. Why it matters here. Rocco Arena crossed to New York four times between 1892 and 1913 and Giovanni Arena followed him in 1899. This archive has always treated the American end as the American record's problem. If any Arena, Arlotta, Donato or Polistena died outside Italy and the family asked for it, the act is in a parte II — and it names the foreign city, the date, often the cause, and the parents. A New York death certificate, in other words, sitting in a Calabrian register. The sweep is affordable. Parte II is a few acts per volume, at a fixed position — the last images before the *indice annuale*, which `tools/antenati_index.py sheet` already finds. The annual index lists parte II entries in its own column (*Parte* = 2ª), so the index alone says whether a year has any worth opening. Eighty-one volumes, 1866–1943, and none read. AND ROW 146 GIVES IT A TARGET. Three of Rocco Arena's four brothers vanish from Scilla's registers after 1898. Parte II is the first place to look for them.
146. Next — ready to start this archive
Two of Rocco Arena's brothers vanish after 1898 — the third was a sister
A silence, not a gap — and it took two complete searches to see it. Francesco Antonio Arena and Serafina Arlotta had five known sons: Filippo (b. 14 Feb 1852), Domenico (b. 24 Mar 1853), Giovanni (b. 30 Apr 1854), Giuseppe (b. 5 Mar 1860) and Rocco (b. 20 Dec 1861). Only Rocco and Giovanni die at Scilla. Rocco on 8 February 1932, Giovanni on 31 March 1936 — both in the Antenati name index with their parents named. Filippo, Domenico and Giuseppe are in none of it. Not among the twenty-six indexed Arena deaths of 1910–1943. Not in the A block of any annual index from 1897 to 1908 (row 128). Not in 1889–1896 either. And Giuseppe was certainly still there in June 1898, burying his own ten-month-old son Francesco Giovanni Rocco at Via Acquagrande 30. Then nothing. The reading that fits is that they left. Their brothers Rocco and Giovanni both crossed to New York — Rocco four times from 1892, Giovanni in 1899 to join him — and Giovanni's bride's father, Paolo Martello, was already in America in 1885. Scilla was going before this family went. Where to look, cheapest first. 1. Row 145 — parte II of Scilla's own death volumes, where deaths abroad are transcribed at a family's request. If one of them died in New York and the family asked, the act is there and it names the parents. 2. The Ellis Island and NARA manifests for a Filippo, Domenico or Giuseppe Arena of Scilla, 1892–1914 — the same route that documented Rocco's four crossings and Giovanni's one. 3. Their marriages at Scilla: Giuseppe married Maria Martello in 1885 (atto 73) and Giovanni married Concetta Russo the same autumn (atto 86). Filippo and Domenico have no marriage in this archive at all, and a marriage act would give a wife whose own death may be indexed. The negative that would be worth as much as a finding: if none of the three is in parte II and none is on a manifest, then they died somewhere that never told Scilla, and the archive should say so plainly rather than leave three men trailing off. ONE MORE STRAW, 20 SEPTEMBER. Filippo Arena's birth act, Nati 1852 atto 48, has a blank margin. Scilla's birth acts carry marriage annotations for marriages celebrated in the same comune after 1866 (row 154), and any marriage of his falls in that window. A blank margin does not prove he never married — it is consistent with a marriage somewhere else, which is precisely what «he left Scilla» predicts. Set beside his absence from the death index 1889–1943 and from the name index of 1910–1943, it is a third silence pointing the same way. The same test is worth running on Domenico Arena (b. 1853), whose act has not been opened in the register at all. ONE OF THE THREE WAS NEVER A BROTHER, 20 SEPTEMBER. *«Domenico Arena, b. 24 March 1853»* is DOMENICA ARENA, a daughter: act 53 of 1853 says *«una femina»*, *«di dare alla fanciulla il nome di Domenica»*, and the baptism column names Domenica Arena. A married daughter leaves the Arena surname and stops appearing under it, which is not a disappearance at all. So the row is two men, not three: Filippo (b. 1852) and Giuseppe (b. 1860). And Filippo's silence got louder. His birth act's margin is blank in a volume where other acts carry marriages written back in 1876, 1881 and 1883 — so he did not marry at Scilla. Third silence, same direction. AND THE MARGIN STRAW IS WITHDRAWN, SAME DAY. I wrote an hour ago that Filippo Arena's blank margin was evidence he did not marry at Scilla. Giuseppe Arena's birth act is blank too, and Giuseppe married at Scilla in 1885. The annotation was simply not always made. Filippo's blank margin means nothing, and this row rests where it did: on his absence from the death index 1889–1943 and from the name index of 1910–1943. RE-TESTED AGAINST A SET NEARLY THREE TIMES LARGER, 21 SEPTEMBER, AND IT HOLDS. The name index was re-queried in full — not to redo the search but to test the query form this archive had just broadcast to three other archives, on the principle that an untested method note is untested code. It works, and it returns 152 rows for Arena at Scilla, about sixty-nine people, where the earlier sweep saw twenty-six. Only two carry «Madre Serafina Arlotta» — Rocco, 8 February 1932, and Giovanni, 31 March 1936. Neither Filippo nor Giuseppe is anywhere in the sixty-nine. A negative that survives the set trebling is worth more than the first one was, and it points the same way. And the pull paid sideways, which is what a full index read does: GIUSEPPE ARENA, died 27 March 1925, «Padre Domenico, Madre Teresa Polistena» — the first evidence that Filippo's brother Domenico's 1838 marriage produced anyone, and an independent confirmation of Teresa Polistina eighty-seven years later in a different series. His daughter Teresa Arena died there in 1943, named for her grandmother.
147. Done — finished and published this archive
The 1851 processetti bundle — opened, and it dated two generations
The cheapest unopened thing in this archive. Francesco Antonio Arena married Serafina Arlotta at Scilla on 7 February 1851, atto 13. Antenati holds the *processetti* volume for that window — 1849–1852, `an_ua2002487` — found by listing Scilla's 466 registers in one page load (row 64). What a processetti fascicolo carries: the birth extracts of both spouses, and the death acts of any parent already dead. The Scilla bundle of 1842–1845 carried a dead grandparent as well, and the Briatico bundle of 1849 broke the whole Calabrian line open — a cholera death, three generations of descent, a birth act, an *atto di notorietà* and five new names, in twenty minutes after two days of sweeping the wrong comuni. The four parents it could name, with their death acts attached: Filippo Arena and Domenica Costa (row 68 — otherwise twenty-seven years of unindexed sweep), Giovanni Arlotta and Maria Costa (row 39). And possibly Domenico Arlotta (row 85), the deepest name on the Scilla side, dead before April 1854. What it cannot do: these bundles stop in 1865. The 1885 and 1888 Arena marriages have no processetti on Antenati at all. Method: the ark is in `scilla-arks.tsv`; the container costs one rate-limited page load, then manifests and images are free. Bundles are filed in act order, so fascicolo 13 of 1851 is found by reading the act numbers on the wrappers rather than by counting images — and the volume spans four years, so find the 1851 block first. OPENED 20 SEPTEMBER, AND IT PAID. Container `58rXBoX`, 1845 images; fascicolo 13 of 1851 begins at image 1190, wrapper reading *«Volumetto de documenti pel matrimonio di Francesco Arena e Serafina Arlotta — fogli quattro»*. Foglio 1 is Francesco Antonio Arena's birth act itself, act 186 of 1819 — the act, where this archive has always cited the Tavola. Filippo Arena, twenty-five, marinaro (b. c. 1794) and Domenica Costa, twenty-seven (b. c. 1792): two generation-seven people who had no dates at all. Born at twelve o'clock in the family's own house; baptised next day. And Filippo signed it, in a register where almost every act ends *«perché illetterati»*. Foglio 2 is Serafina Arlotta's, and it is act 89, not act 2 — this archive has been citing a Tavola row number in three places. Maria Costa, twenty-three (b. c. 1807), her first date of any kind; Giovanni Arlotta, forty — which conflicts by eleven years with his own 1854 declaration of seventy-five, and is flagged rather than averaged. And the absence is evidence: four sheets, no death act of any parent, where fascicolo 3 of the same year binds in *«Atto di morte del padre e dell'avo paterno dello Sposo»*. The ordinary reading is that all four parents were alive on 7 February 1851 — graded *inferred*, because the bride was twenty and no consent sheet is listed either. Navigation, which is the reusable part: fascicolo wrappers are findable by shape in a montage of thumbnails at width 430, no text read. In 1851 they fall at images 1100 (n. 3), 1190 (n. 13), 1280 (n. 24), 1430 (n. 37). Do not navigate by the *Num. d'ordine* on the atti di notificazione bound inside — a separate series at a different rate. What is left in this bundle: sheets 3 and 4, the banns and the certificate of canonical celebration. Low value beside the first two, but unread. AND THE INFERENCE IT RESTED ON IS NOW CONFIRMED. The bundle's four-sheet index held no death act for any parent, which was an argument from absence and graded *inferred*. The 1851 marriage Tavola settles it positively: entry 1's parents are written without *«fu»*, in a series that marks the dead with it. All four were living in February 1851, and the correction is now graded *documented*.
148. Next — ready to start this archive
Antonino Crupi's Queensland investigation file — open access, unread
NAA: BP242/1, Q16634 — «Crupi, Antonio and Maria — Queensland investigation case file», 1939–1957, held at Brisbane, access status OPEN. Found while reading Antonino Crupi's aliens card (row 117). Why it matters. The card says only that he went into Loveday compound 14A in March 1943 and came out in October. The investigation file is the document that put him there — BP242/1 is the Queensland series of the Commonwealth Investigation Branch, and its files carry the informant's allegation, the police report, the internment recommendation and the objection hearing. Mario Crupi has one too — BP242/1 Q25743, also open, also unread. «Antonio and Maria» is the file title; Antonino's wife was Nellie by 1946, so the Maria may be his mother Maria Vecchio, an earlier wife, or a different couple entirely. The title is a finding aid, not a record — it has to be opened before anything is claimed from it. Neither file appears to be digitised, so this is a reading-room request or a digitisation order, not an afternoon's work. Worth putting on David's list beside the other Brisbane items.
149. Next — ready to start this archive
Two Prostamo in Western Australia in the First World War
Thirty years earlier than anything this archive holds in Australia, and in the wrong state. NAA series PP14/3 — applications for registration as an alien, Perth — holds two items: `ITALIAN/PROSTAMO G`, GABRIELE PROSTAMO (item 5137304) and `ITALIAN/PROSTAMO F`, FRANCESCO PROSTAMO (item 5137303), both dated circa 1916–1920, both access status Open. Why this is worth opening. This archive's Australian story starts in 1937 with Angela Crupi and runs through the post-war assisted passages. Prostamo is a Briatico and Pizzo surname and it is not common. Two men of it registering as enemy aliens in Perth during the First World War means an Italian migration this family knows nothing about, a generation before the one it has documented — and the registration form of that period carries birthplace, date of birth, arrival, occupation and address. What it is not, yet. A surname and a state. Neither man is in the tree, and Western Australia is not where this family settled. The test is the birthplace on the form: if either says Briatico, Potenzoni or Pizzo, it is the same stock; if it says somewhere else, Prostamo is a wider surname than this archive assumed, which is worth knowing too. Check the item detail for «View digital copy» first — PP14/3 may not be scanned, in which case this is a Perth reading-room request rather than an afternoon.
150. Next — ready to start David
Four Brisbane files that name this family and nobody has ever asked for
These are requests, not searches. Four National Archives files name people on this archive's spine, are held at Brisbane, and carry the access status «not yet examined» — which means nobody has ever asked to see them. An access examination is free and takes roughly ninety days, and anyone can lodge one. 1. NAA: J25 1971/3900 — *«Arena, Rocco; wife Angelina [or Angela] [nee Polistena]»*, 1953–1971. Michael's grandparents on the Arena side, whose marriage is documented nowhere else (row 80). 2. NAA: BP242/1 Q16634 — *«Crupi, Antonio and Maria — Queensland investigation case file»*, 1939–1957, access Open. The file behind Antonino Crupi's internment (row 148); his brother Mario's is BP242/1 Q25743, also open. 3. NAA: J25 1962/5930 — *«Prostamo, Domenico Francesco; wife Assunte Maria [nee Melluso]»*, 1952–1966. The marriage of Maria Assunta Melluso, buried at Mount Gravatt in 1991 under a surname nobody could attach to a man. 4. NAA: J25 1972/23086 — *«Melluso, Gregorio; wife Carmela; and children Giuseppe, Maria [surname Prostamo]»*, 1959–1974. Gregorio Melluso's household, which Queensland's own index cannot show — it holds exactly one Melluso registration in its entire run. Why they are worth the wait. J25 is Queensland immigration and naturalisation correspondence: the files carry arrival details, addresses, sponsors, and the family relationships spelled out. BP242/1 is the Investigation Branch: informant, police report, internment recommendation, objection hearing. Every one of these would answer a question this archive currently records as open. Owned by David because lodging the examination request is his to do, and the ninety days start the day it is lodged.
151. Done — finished and published this archive
Filippo Arena × Domenica Costa's marriage — the only route to generation nine
This archive reaches generation eight on the Scilla side and stops. Filippo Arena was twenty-five and Domenica Costa twenty-seven when their son Francesco was born on 25 November 1819 — ages read today off the act itself, in the 1851 processetti. Their own parents are unknown, and there are only two documents that would name them: their marriage act, or Filippo's death act (row 68, a twenty-seven-year unindexed sweep). The marriage is the cheaper of the two. The window is about 1812–1819, and four years are now read: 1819 — no Arena. Tavola alphabetised by surname; its A section holds one entry, Abbate Grazia. 1818 — no Filippo Arena. Tavola alphabetised by the groom's given name, so Filippo sits under F, and the F section is Migliarino, Polimeni, Pizzarello, La Valle. *Searching the A section here would have produced a confident, complete and entirely false negative* — the caution this archive already wrote, paying for itself. 1817 and 1816 have no Tavola at all. 1817 runs to act 41 then blank forms; 1816 to act 42, and the archivist's card reads VOLUME TARLATO — worm-eaten. Each costs a full read of about forty openings, one act to an opening. What is left, cheapest first. (1) The napoleonic marriage volumes — 1809, 1810, 1811, 1812, 1813–1814, 1815 — whose arks are in `scilla-arks.tsv` and whose containers have not been bought. Scilla's registration begins in 1809, so those years exist and nobody here has opened one. (2) The unindexed 1816 and 1817, act by act. (3) The processetti for the same window — `1810–1814` and `1815–1816` are held, and a bundle carries the death acts of dead parents, which would reach generation nine even if the Tavola does not. Containers in hand: 1816 `wQYJYDb`, 1817 `5K9P9J9`, 1818 `LPaGar8`, 1819 `0AJEJ8b`. THREE MORE YEARS FALL, 20 SEPTEMBER. 1812 (Tavola, seventeen marriages, all read), 1811 (Tavola; F section is di Franco, Arbitrio, Bova) and 1810 (indice; F section is Filippo Laganà, Vita, Martello, Zirilli, Fava) hold no Filippo Arena. 1809 has no Tavola. All three napoleonic indexes use the groom's given name as well. So the marriage is not in 1810, 1811, 1812, 1818 or 1819. What is left is 1809, 1813–1814, 1815, 1816 and 1817 — every one of them without an index. All nine containers are now saved. The cheapest remaining route is therefore the processetti, not the acts: bundles are found by their wrappers, which name the couple in three words, and Scilla holds `1810–1814` (`an_ua2093352`), `1815–1816` (`an_ua2093354`) and `1817–1819` (`an_ua2002455`) — the whole remaining window, in three volumes. A bundle also carries the death acts of dead parents, so it reaches generation nine even where a Tavola would only name them. And the sweep paid for itself sideways: the 1811 Tavola gave Candida Costa, daughter of Rocco Costa and Pasqualina Idone — Maria Costa's sister, found in an index opened for somebody else. A Tavola that names four parents will keep doing that, which is an argument for reading the whole section rather than scanning for one surname. THE PROCESSETTI ARE BOUGHT AND SIZED, 20 SEPTEMBER, AND THE ROUTE IS CONFIRMED BUT NOT CHEAP. Containers: 1810–1814 `LaNPkQJ` (947 images), 1815–1816 `LqA4a7D` (364), 1817–1819 `02Ro183` (880) — 2,191 images for the five unindexed years. Wrappers exist and are the finding aid. Image 22 of the 1815–1816 volume reads *«Anno 1815, N.o 4 — Documenti pel Matrimonio di …»* over a mostly blank left page — the same form as the 1849–1852 bundles. But a twenty-point sample at nine-image spacing surfaced only one clearly legible wrapper, so they are neither evenly spaced nor reliably distinguishable at thumbnail size in this volume, which is catalogued *VOLUME TARLATO* and heavily stained. So this is a deliberate pass, not an afternoon: roughly 250 thumbnails at nine-image spacing across the three volumes, read in montages of twenty, with each candidate wrapper cropped to confirm the names. Start with 1815–1816 — it is the smallest and covers two of the five years. Everything needed is now in hand and nothing here needs buying again. CLOSED, 20 SEPTEMBER, AND NOT BY THE PLANNED ROUTE. The processetti were never opened. Filippo Arena married Domenica Costa at Scilla on 12 NOVEMBER 1815 — Matrimoni 1815, atto 29, in the napoleonic volume `57O1erJ`, one of the five unindexed years. What cracked it was not a new source but a new way of reading an unindexed volume: in these registers an opening puts the *end* of one act on the left page and the *head* of the next on the right, and that head carries the act number, the date, the groom's name, his age, his trade and his father. Cropping that band and tiling it ten to a sheet reads a whole volume in four montages. Three volumes — 1815, 1816, 1817, one hundred and eleven acts — fell in an afternoon, where this row had budgeted 2,191 processetti thumbnails. All three were read to the closing statement, so each volume's end was found before anything was written about it; 1816 and 1817 are complete negatives. The act names four people this archive had never seen: ORAZIO ARENA, marinaro, and SETTERIA MORABITO — Filippo's parents, both alive, both standing beside him and consenting, because at twenty-two he was still a minor under the code; and GIOVANNI COSTA, dead at Scilla on 29 March 1790, and MARTA ALFONZETTI, Domenica's parents. Filippo signed the act, in the same hand that signs his son's birth in 1819. Two things this row got wrong, recorded rather than quietly fixed. Its header promised generation nine; Filippo is generation seven here, so his parents are generation eight — the same depth already reached on the Arlotta–Costa side. And the wall it was built to climb was this archive's own mistake: the marriage was called pre-civil and lost only because the search had started in 1816. Scilla begins in 1809. The sea trade now runs four generations — Orazio, Filippo, Francesco Antonio, Rocco — and above Orazio there is no civil register at all. What is left is parish, and that is row 158.
152. Next — ready to start this archive
The Civil Aliens Corps — Rosario Mazza is in it and the Crupi are not
Internment was not the end of it, and this archive stops at the camp gate. NAA: J1737, MAZZA R — «Mazza Rosario, CAS 15000 — Civil Aliens Corps (CAC)», Brisbane, 1944, access OPEN. The CAC was the labour corps that interned Italians were released *into* rather than freed from — road, timber and farm work under military administration — and CAS 15000 is Rosario Mazza's service number in it. The three men line up. Mario Crupi was at Loveday from 18 August 1942 and released 30 December 1943. Antonino Crupi was moved to compound 14A in March 1943 and released 22 October 1943. Rosario Mazza's Corps file is 1944. They were moving through the same machinery within months of each other — Angela Crupi's husband and two of her brothers. TESTED THE SAME DAY AND THE GUESS WAS WRONG. I wrote here that the Crupi brothers would have J1737 files too. They do not. The complete CRUPI name index — 233 items — holds no J1737, no Civil Aliens Corps and no CAS number for any Crupi. Rosario Mazza has one; neither of his wife's brothers does. Why is unresolved: the Corps was not universal, and men released in October and December 1943 may simply have gone back to the cane. What the test found instead is that both Crupi internment dossiers exist and are open — MP1103/1 Q7838 for Antonino and MP1103/2 Q30124 for Mario — and that those serials are the ones written on their aliens cards, which ties card to camp for both men. What to look for now: NAA: BP628/5, the *«Name Index and registration cards for internees dossiers Queensland — MACAUDO to MYSLIM»* (1939–1962, Brisbane, open), which is the finding aid to the Queensland dossiers and whose range covers Mazza. This is the part of the story the family is least likely to have been told, and it is on open access.
153. Done — finished and published David
A Trove API key — asked for on 15 September, granted on 27th, and it unblocked the source in an afternoon
Granted 27 September 2026, valid to December 2026. Stored in `.env`, which is gitignored — this repository is public and the key is never committed, printed, or written into a data file. `tools/trove.py` reads it from there and refuses to search without it, rather than querying anonymously and reporting an empty result as a finding.
It worked the same afternoon: seven naturalisation entries across Prostamo, Polistena and Melluso, and the eight-day block on row 79 is closed. The failure had always been the tool, never the source.
154. Running — a machine or a person is at it now this archive
The margins — worth reading when full, worth nothing when empty
This archive has read a dozen Scilla birth acts and never once looked at the left margin. Atto 174 of 1833 has written down its edge, in a different hand: *«L'anno 1869, il giorno 12 Luglio … ha contratto matrimonio in questo Comune con Colorina Bri[…]»*, signed by the Uffiziale and dated from Reggio on 26 August 1869 — thirty-six years after the birth. The printed form asks for it. The left column is headed *«MARGINE per scriversi alcune notizie riguardanti gli atti di nascita»*; the right is the baptism, which this archive does read. One column is the sacrament; the other is everything that happened to the person afterwards. What a margin can hand over free: a marriage date and a spouse, with no marriage register opened at all — and for years where the marriage Tavola is missing or the volume is unindexed, that is the whole answer. The acts to go back to, all already fetched and all free to re-crop: Francesco Antonio Arena, Nati 1819 atto 186 (read in the 1851 processetti — *and the processetti copy may not reproduce the margin, so the register itself must be opened*); Serafina Arlotta, Nati 1830 atto 89; Rocco Arena, Nati 1861 atto 228; Giovanni Arena, Nati 1854 atto 76; and the Arena births of 1852, 1853 and 1860. Cheap, mechanical, and it uses containers already bought. THREE TESTED THE SAME HOUR, AND THE RULE IS NARROWER THAN IT LOOKED. Rocco Polistina (1833): margin carries his 1869 marriage. Serafina Arlotta (1830): blank — she married in 1851. Filippo Arena (1852): blank — any marriage of his falls after 1866. The reading that fits all three: the annotation is an Italian-era habit, absent from an 1851 marriage and present on an 1869 one, and it is written only when the later act is registered in the same comune — the note says *«ha contratto matrimonio in questo Comune»* in so many words. So a blank margin is weak evidence. For a pre-1866 marriage it means nothing at all. For a post-1866 one it is consistent with a marriage elsewhere — which for Filippo Arena is the reading the rest of the evidence already supports (row 146). Still worth doing for Rocco Arena (b. 1861, m. 1888 at Scilla), Giovanni Arena (b. 1854, m. 1885 at Scilla) and Giuseppe Arena (b. 1860, m. 1885 at Scilla) — all three married at Scilla after 1866, which is exactly the case where the annotation appears. And for Domenico Arena (b. 1853), where a blank would be one more straw and an annotation would be a marriage nobody has. THE CONTROL IS RUN AND THE GRADING GOES UP. The 1852 volume does annotate margins — act 55 carries *«Scarfone Rocco contrasse matrimonio in Scilla con Vita Maria»*, 1881 — and 1853 does it twice on one opening (1876 and 1883). So Filippo Arena's blank margin is evidence: he did not marry at Scilla. *(And a correction of my own inside the hour: I first took writing at the edge of the 1852 act-48 opening for act 49's margin. It is act 49's baptism column. The control had to be run on a different opening before the blank meant anything.)* And chasing a margin turned up something far bigger — see row 137 and the retraction filed against Domenico Arena. AND THEN THE CONTROL FAILED, WHICH IS WHY YOU RUN CONTROLS. Giuseppe Arena married Maria Martello at Scilla on 12 September 1885 — act 73, read here in the original — and his own birth act, Nati 1860 atto 45, has an empty annotation box. A man who demonstrably married in the comune after 1866 was never annotated. So a blank box is not evidence of anything, and the reading filed against Filippo Arena is withdrawn in full. What survives: a *present* annotation is free evidence — Rocco Polistina's 1869 marriage came that way, and marriages of 1876, 1881 and 1883 sit on other acts. A bonus when it is there, silence when it is not. And the box moves between volumes. 1852–53 use a left column headed *«MARGINE per scriversi alcune notizie…»*; 1860 uses a right box headed *«MENZIONI prescritte dagli articoli 64, 65 e 106»*. Same purpose, opposite side of the page — so anyone reporting a blank must say which side they looked at.
155. Done — finished and published this archive
Camillo Antonio's marriage — FOUND, 23 January 1881, and he is nine years younger
Sixteen years of Scilla's birth register have been read and he is in none of them (row 137). His parents' children come in 1833, 1836 and 1840; there is no year he fits. The birth register cannot settle it, so stop asking it. His marriage act can. He married Giovanna Romano — her death act of 15 July 1930 names him as her husband, and their son Giovanni was born at Via Grotte 47 on 16 September 1883, when Camillo Antonio was entered as forty-eight. So the wedding falls roughly 1870–1882. What a Scilla marriage act of that period gives: both spouses' ages, both birthplaces, and all four parents. That is his age a third time, from his own mouth, and — decisively — whether he was born at Scilla at all. If the act says he was born somewhere else, sixteen years of negative reading stop being a puzzle and become the answer. The ground is already bought. Scilla's Italian-era marriages run 1866–1913 without a gap and every ark is listed in `scilla-arks-italiano.tsv`; containers are saved for 1874–1877 and 1885–1890. The annual index of each year is one A-block read, and P for Polistena is a short section at Scilla — eight births in eleven years, so the marriages will be sparse too. Start at 1875 and work outward, since a man of about forty in 1883 marrying a woman young enough to bear a son that year most likely married in the middle of the window. STARTED 20 SEPTEMBER. 1876 IS CLEAR; 1875 HAS NO INDEX. 1876 carries an *«Indice annuale per gli Atti di Matrimonio»* in a good format — every spouse alphabetical, the two halves of each couple braced together against one register number, so one reading gives both names and the act. Its P range is Pirrotta, Patafio, Paladino ×3 and Piemi. No Polistena. 1875 has no Tavola at all — the volume ends in blank forms and a printed *Parte II* title page, and there is only one ark per Italian-era marriage year at Scilla, so no second volume carries the index. 1875 must be read act by act, about forty openings. So the indexes are not uniform across this series either — 1876 and 1877 have them, 1875 does not. Check the tail before promising a year. Left: 1870–1874, 1875 (by acts), 1877–1882. Containers saved for 1874–1877. 1874 AND 1877 ADDED, BOTH CLEAR. 1874's P section is Parisi, Pizzarello ×2, Piria, Pirrotta, Patafio; 1877's is Patafio, Paladino, Panuccio, Pirrotta, Papalia, Parisi. No Polistena in either. The index format is the same good one in all three years — every spouse alphabetical, the two halves of each couple braced against one register number — so a year costs one crop once the container is in hand. That makes the rest of this window cheap. Noted and not joined: act 44 of 1877 marries an Arlotta Maria Concetta to Pietro Panuccio. Serafina Arlotta's known siblings are Anna, Domenico and Giuseppe Antonio; a Maria Concetta is not among them and nothing here makes her one. Left: 1870–1873 and 1878–1882 (containers not bought), and 1875 act by act. 1879 AND 1880 ADDED, BOTH CLEAR. 1879's P is Procopio, Papalia ×3, Paladini, Paladino, Prigandè; 1880's is Paspiglia, Piria, Pirrotta, Paladino, Patafio, Papalia ×2. No Polistena in either. Five years read and clear: 1874, 1876, 1877, 1879, 1880. 1875 has no index; 1878's container lookup hit a 403 and was not retried. Left: 1870–1873, 1878, 1881, 1882. Three things noted and not joined. 1879 act 22 marries an Arlotta Giuseppe — Serafina's brother Giuseppe Antonio Arlotta is recorded here as the husband of Angela Nunziata Pizzarello and the bride in this index does not obviously match, so the act is wanted before anything is claimed. 1879 act 8: Russo Giovanni × Arena Fortunata — the Russo–Arena pairing again, six years before Giovanni Arena married Concetta Russo. 1880 act 40: Riganatì Antonio × Arena Teresa — and the Teresa Arena this archive holds, Orazio's daughter born 1853, died in 1913 as the wife of *Pietro Bova*. A different woman or a first marriage; nothing decides it here. 1878 AND 1882 ADDED. 1878 is clear — Paladino, Porpiglia, Palermo, Pizzarello. 1882 holds two Polistina and neither is him: *«Polistina Giuseppe × Zagari Giovanna»* and *«Polistina Luigi Carmelo × Matrà Nunziata»*, act 48. That is worth more than a negative. Seven years of this index are now read and 1882 is the first to hold the surname at all — so the Polistena were marrying at Scilla by 1882, and the earlier silence is not an artefact of the index. And «Luigi Carmelo» fits a household already known: the 1831 birth Tavola gives *«Polistina Francesco Antonio, di Luigi e Giovanna Baviera»*. Both 1882 acts are worth opening regardless. A Scilla marriage act of that date gives both spouses' ages, both birthplaces and all four parents — so either one would say whether the Polistena of Scilla were born there or came in, which is the question row 137 could not answer from sixteen years of births. Left: 1870–1873 and 1881 (not bought), 1875 (no index, act by act). AND A NAVIGATION NOTE, BOUGHT WITH FOUR WASTED PROBES. The 1882 volume is split into two series and both are numbered from one — image 33's margin reads *«Numero 4 — Colosi Giuseppe Luigi Vincenzo / Zagari Giuseppa Maria Raffaela»*, which is Parte II act 4. Parte II begins around image 31. The index gives a «Numero di Registro» without saying which part, so an index number cannot be resolved to an image without checking both series; and several images in the thirties are blank versos showing only bleed-through. The cheap way in: every act carries a margin label with its number and both spouses' names, so a montage of the narrow left-margin strip across a run of images identifies every act without reading a single act body. That is the trick that worked on the 1874 volume and it should have been used here first. AND THE TRICK WORKED. A montage of the narrow left-margin strip across images 24–32 read the whole volume's act numbers and both spouses' names in one look: acts 45, 47, 49, 51 on successive images, and Parte II beginning at image 30. Act 48 is the second act on image 25, and it is read — see the 1882 row in scilla-reads. FOUND. SCILLA, MATRIMONI 1881, ATTO 4 — 23 JANUARY 1881, six seventeen in the evening. *«Polistina Camillo Antonio, di anni trentasei, marinaro, nato in questo Comune, figlio del fu Giuseppe e della fu Anna Pontillo»*; *«Romano Giovanna, di anni ventisei, filatrice, nata in questo Comune, figlia di Annunziato e di Angela Zirilli»*. THE AGE IS THE FINDING. Thirty-six in January 1881 is a birth about 1844 — and his son's 1883 act makes him forty-eight, a birth about 1835. Nine years apart, and the 1835 figure is what row 137 swept eleven years of birth Tavole around. The answer was never in them. He was born at Scilla — *«nato in questo Comune»* — so the birth act exists and the sweep missed it. And 1844 is inside the swept range: that year's P block was read for entries 127–138 without establishing where the block ends, and 1845 is on record as partial, with seven of eighteen lines too faint. Those two years are the shortest road left and are already paid for. Three more things it gives. Giuseppe Polistena and Anna Pontillo were both dead by January 1881 — the first bound either has had, against their last sighting alive in January 1840. The parentage is confirmed from a second document, forty-three years before the 1924 death act. And Giovanna Romano is dated: twenty-six, born at Scilla about 1854, with both her parents living — the fourth record to call her a Romano and the first in which she is the subject.
156. Done — finished and published this archive
Camillo Antonio's birth act — read: atto 137 of 1844
His birth is dated — 21 September 1844 — but only from the index (row 137). The act has not been opened, and this archive's own standing rule is *read the act, not the index*: that rule has been vindicated four times here, most recently when a Tavola line read as *Domenico* turned out to be Domenica, a daughter. It is cheap. The 1844 Tavola carries no act number, but the register runs in date order, so a birth of 21 September sits two or three probes into container `LNolOVK`, which is already bought. What the act would add. Both parents' ages — dating Giuseppe Polistena and Anna Pontillo a third time, against *thirty-four and thirty-three* in 1833 and *thirty-eight and thirty-five* in 1836. The house, the hour, the declarant (a midwife delivered both his siblings), and the witnesses. And its margin, which on a man who married at Scilla in 1881 may carry that marriage — though row 154 records that a blank there proves nothing. It would also confirm the given name. The index says Camillo; every later record says Camillo Antonio. The act will say which he was christened. HUNTED THE SAME HOUR AND NOT FOUND — five acts eliminated. The 1844 volume maps cleanly: image N carries folio N−2, with the even act on the left page and the odd act on the right. Image 73 = acts 140/141, image 74 = acts 142/143. Checked and excluded, all in the September run: 138 (declarant Domenico Spoletti), 139 (child Giuseppe), 140 (declarant Cosmo Bellantoni, child *Marianna Violante*), 141 (declarant Domenico Calarco, baptised at the parish of Solano), 142 (declarant Giovanni Macrì, child *Rosantonio*, born the 28th). None is Polistina. So either the act sits outside 138–142 — the September run plausibly spans acts 135 to 145 — or the Tavola's «21 Settembre» is the birth date against a later declaration, or that faint line reads a different day. The finding does not depend on it: the Tavola gives the parents and the date, and the marriage act's *trentasei* is exact for September 1844. But the act has not been read, and this archive's own rule is that the index is not the source. Next pass, and do it this way: montage the act-number-and-date band — region `pct:50,10,48,20` at width 900 — across images 68–76, which covers acts 132 to 149 in one look, instead of opening them one at a time as I did. FOUND — ATTO 137, not 142. Two montages did what five one-at-a-time probes could not: the even acts across images 68–77 mapped September to acts 129–143, and the odd acts across 68–73 put a Polistina at act 137. *«È comparso Giuseppe Polistina, di anni cinquanta, marinaro … quale ci à presentato un maschio … nato da Anna Pontillo, filatrice, sua moglie legittima … nel giorno ventuno corrente mese … di dare al fanciullo il nome di Camillo Antonio.»* The baptism column names Camillo Antonio Polistina. So the Tavola was right about the date and the parents and wrong about the number — 142 was its row, 137 is the act. Third time today. And the act gives the full given name, where the index gives only *Camillo*. And the margin carries his marriage, annotated *«Scilla 25 Gennaio 1881»* — two days after the wedding of 23 January 1881. The birth act and the marriage act point at each other across thirty-seven years, and the identification is closed from both ends. One figure recorded and not averaged: Giuseppe Polistina is fifty here, against thirty-eight in 1836 and thirty-four in 1833 — births of 1794, 1798 and 1799. The two earlier agree; this one does not.
157. Done — finished and published this archive
Orazio Arena's other children — the Scilla birth registers, 1809 onward
A new surname to sweep, and the cheapest lead the 1815 act opened. Orazio Arena and Setteria Morabito were both alive and living at Scilla in November 1815, and their son Filippo was their minor child — the word the act uses is *«figlio minore»*, meaning under twenty-five, not youngest. Scilla's birth registers run from 1809 and its marriage registers from the same year, so any child of theirs born or married from 1809 on is in a volume this archive can already reach, and most of those volumes have an annual Tavola. A sibling of Filippo is worth more than a name: the Tavola of the year a sibling married dates the life status of Orazio and Setteria, which is the instrument row 68 wanted and could not use on Filippo's own children. Start with the marriage Tavole 1816–1840 under A, which are one page each, before touching the birth registers; the marriage index is alphabetised by the groom's given name in some years and by surname in others, so read the whole section, not the letter you expect. Containers for 1816–1819 are already in `scilla-arks.tsv`. THE VOLUMES ARE OPEN, 20 SEPTEMBER. Twenty-one marriage containers, 1820 through 1843, were fetched in one sitting from inside the antenati origin -- six at a time, five seconds apart -- with no 403, and their image counts came free off the IIIF manifests. `scilla-arks.tsv` now holds every Scilla marriage volume from 1809 to 1845 with a container against it. Nothing in this row needs buying again; what is left is reading. AND THE NAPOLEONIC SWEEP FOUND AN ARENA HOUSEHOLD NOBODY HAD MET. Scilla Matrimoni 1813, atto 21: *«DIEGO ARENA, di anni trentasette, di professione SARTORE … figlio maggiore di DOMENICO ARENA, di anni SETTANTA, di professione BARBIERE, domiciliante in questo Comune, e di ANNA MELIDONI, di anni cinquantaquattro»*, both present and consenting; the bride Maria Giuseppa Pontillo, sixteen, daughter of Giovanni Pontillo, commerciante, and Diana Idone. Banns at Campo as well as Scilla. Domenico Arena was born about 1743 — earlier than any Arena in this archive, Orazio included — and he is a barber, where the spine is four generations of sailors. He is NOT claimed as Orazio's father, and the temptation to claim him is the point. Scilla has more than one Arena family; a shared surname in a small town is how this archive nearly hung a birth act on the wrong Mariano Mazza. What would settle it is Orazio Arena's own marriage, which is pre-1809 and parish — row 158. In the meantime the household is recorded as three documented people with no edge to the spine. Also banked from the same two volumes: the first year of the register marries a Fortunata Donato (1809 act 2) and a Giacoma Arena (1809 act 10) — both brides, both surnames present in the town from the beginning. AND THE SAME THREE INDEXES GIVE THE SHAPE OF THE SURNAME. Read whole, the napoleonic years hold six Arena and not one of them is Filippo's household: Diego Arena the tailor as a groom (1813), with his father Domenico Arena the barber, seventy, born about 1743; Giacoma Arena (1809) and Giovanna Arena (1812, daughter of Carlo Arena and Eugenia Ateja) as brides; Natalina Arena as a mother (1811); and Antonia Arena, *«fu»* and so dead before February 1811, wife of Domenico Falcomatà, whose two daughters marry in 1811 and 1812 — the same parents in both entries, so two sisters and one household. That is THREE Arena fathers at Scilla in one decade — Domenico the barber, Carlo, and Orazio who is Filippo's — with nothing between them. The finding is not a line; it is the shape of a surname in a town, and a standing warning against the join this archive would most like to make. DONE, AND ON THIS ROW'S OWN STATED METHOD. The row said: *find a sibling of Filippo, and the Tavola of the year that sibling married will date Orazio and Litteria.* It did exactly that, in the second year opened. Scilla, Matrimoni 1821, atto 29 — GIUSEPPE ARENA, twenty-seven, MARINARO, *«nato in questo Comune … figlio di ORAZIO ARENA, di professione marinaro, domiciliato ivi, e di LITTERIA MORABITO, domiciliata ivi»*, married Maria Teresa Polistina on 13 November 1821; the promessa was taken six days earlier before Tommaso Cutelli — *the same sindaco who married Filippo in 1815*. Three things at once. A brother for Filippo, the first sibling this archive has ever found for him. Orazio and Litteria alive on 7 November 1821 — named with no *«fu»* on either, six years past the only other date for them. And LITTERIA confirmed a third time, after the 1815 act and the baptism extract, which retires the last doubt about this morning's correction. And the Tavola was enough on its own — 1821 alphabetises by surname, so Arena sits at the top of the book and the whole test was one crop. The A section also holds ARENA ROCCO × COSTA DOMENICA (act 20) with different parents: a fourth Arena family at Scilla, and a second Domenica Costa. What is left is not this row: Orazio's own marriage is pre-1809 and parish (row 158), and the remaining Tavole 1822–1843 would find more siblings but cannot go higher.
158. Blocked — waiting on a person or an institution this archive
Above Orazio Arena there is no civil register — the parish records of Scilla
The Arena spine now ends at a man born in the 1760s or early 1770s, and the state was not yet writing. Civil registration in the Kingdom of Naples begins in 1809; everything above Orazio Arena and Setteria Morabito, and everything about Giovanni Costa, who died at Scilla on 29 March 1790, is parish. There is one concrete handle. The 1815 marriage act does not merely say Giovanni Costa was dead — it cites his death record and says it *«si conserva»*, is preserved. So a Scilla parish death register covering March 1790 existed in 1815. Scilla's parishes fall under the Archdiocese of Reggio Calabria–Bova; its historical archive is not on Antenati and not on FamilySearch's Italian civil collections. Blocked, and honestly so: this archive has no access, no catalogue and no evidence of digitisation, and the 1908 Messina earthquake, which levelled much of the strait, is a real risk to the holdings themselves. The next step is not a search, it is a question: write to the Archivio Storico Diocesano at Reggio and ask what survives for Scilla before 1809. That is row 150's kind of task — a human one — and it is offered to David, not claimed here. AND IT IS NO LONGER A BLIND QUESTION, 20 SEPTEMBER. The 1815 marriage bundle (row 159) quotes the Scilla parish books by name and names the men who held them: the *«liber Baptizatorum»* and the *«liber Defunctorum»* of the Chiesa Matrice, kept by the archpriest D. Giuseppe Ingigneri and searched for these extracts by the economo curato Leonardo Cantore Bandiera, who signed all three in May and October 1815. The baptism book was open at 1787 and 1793, the death book at 1790. So the letter to Reggio can name books, custodians and years instead of asking a general question -- which is the difference between an enquiry that gets answered and one that does not.
159. Done — finished and published this archive
The 1815 marriage file — three leaves that reach 1787, 1790 and 1793
Opened the same day the marriage act was found, and worth more than the act. Row 151 had budgeted a blind sweep of 2,191 processetti thumbnails across three volumes to *find* the marriage. Once the act was found in the register instead, the bundle stopped being a search and became an address: year 1815, number 29. The wrappers are the index — each carries the year, the folio range, the bundle number and the couple in three words — so a scan at five-image spacing, then two passes at step two, landed on it in about forty thumbnails. Image 149 of `LqA4a7D`: *«1815 | 114 a 116 | N.o 29 | Scilla | Documenti per il Matrimonio di FILIPPO ARENA E DOMENICA COSTA | fogli tre».* Three Latin extracts, and every one of them paid. Folio 114 — Filippo's baptism: *«Die vigesima quinta mensis Martii ... millesimo septingentesimo nonagesimo tertio, Scyllae ... infantem HODIE NATUM ex Horatio Arena et LITTERIA MORABITO coniugibus, cui nomen imposuit PHILIPPUS NUNTIATUS».* Born and baptised 25 March 1793, godmother Rosaria Bombarda — and his mother's name is Litteria, not the «Setteria» this archive read off the civil act six hours earlier. Folio 115 — Domenica's baptism, 16 June 1787, by the archpriest Giuseppe Ingigneri himself, *«cui imposui nomen DOMINICA JOSEPHA»*, godmother Giuseppa Cotronei. That settles the five-year dispute: twenty-eight at her wedding was right, and the twenty-seven her husband gave at their son's birth in 1819 was five years wrong. Folio 116 — her father's death, 29 March 1790, aged about forty: *«Joannes de Costa ... FILIUS QUONDAM JOSEPHI ET JACOBAE TURCO, legitimus vir Marthae Alfonzetti»*, ill in bed, confessed to Rev. D. Paolo Turco, given the Viaticum, anointed, and buried in the church of S. Maria di Porto Salvo. One word does the work. *Quondam* — the late — makes GIUSEPPE COSTA and GIACOMA TURCO generation nine, the first time the Scilla side has reached that far, and dates them both dead before March 1790. And the lesson is the reverse of the one row 151 taught. There the register beat the bundle. Here the bundle beat the register: it corrected a given name, corrected two birth years, and reached two generations the civil registers cannot touch, because it copies parish books that begin nineteen years before the state did. Where a marriage bundle survives, read it even when the act is already in hand.
160. Done — finished and published this archive
Piedimonte Etneo relisted — the archive held 103 of 457 registers
Asked why the Sicilian generations were still stuck, the first honest answer was: nobody had finished looking at what exists. `pe-arks.tsv` held 103 Piedimonte Etneo registers. The comune has 457. One page load of the register index at `s_size=500`, parsed in page context and reconciled group by group against the live listing — every fondo, type, ark range, count and year list — before a row was written. What had never been listed at all: every *stato civile italiano* year after 1863; all the allegati and pubblicazioni; the *Diversi* and *Cittadinanze*; and twenty volumes of `Matrimoni, processetti` 1841–1860 — the exact instrument that broke the Scilla wall this afternoon (row 159), sitting unopened for the other half of the family. And nine volumes nobody knew were there: *«Riveli di anime e di terre»* of the Deputazione del Regno, 1714 and 1747–1752 — Sicilian household tax returns that list a head of household with his wife, his children and their ages. Row 162. One hard limit, recorded so it is not rediscovered: the holding stops at 1893. Venera Cassaneti's death in 1925 and Salvatore Mazza's marriage to Maria Nicotra about 1900 are not on Antenati and never were — those are the comune at Piedimonte, or the Tribunale copy at Catania.
161. Done — finished and published this archive
John Nicotra and Venera Cassaneti — the side nobody has ever searched
Two of the three cards in generation three are blank, and it is not because the records are gone. Searching this archive for *Cassaneti* returns nothing at all — not a read, not a negative, not a work-list row. Maria Nicotra (1875–1963) married Salvatore Mazza and is buried at Nudgee; her parents come to us only through a Queensland death registration, which is also where the «Maria Concetta» comes from, and where *Nicotra* is garbled to *Lecota* on her son Michele's. Nobody has opened a Sicilian act for any of them. 1875 is read and eliminated, 20 September. The annual index holds exactly one *Nicotra Maria*, act 34, born 6 March 1875 to Alfio Nicotra, sixty-eight, and Angela Previtera — and her margin reads *«unita in matrimonio con Nicito Giuseppe fu Vincenzo, atto del 29.7.91»*. She married at Piedimonte in 1891 and stayed. The method is now fixed and it is cheap. Piedimonte's Italian-era birth volumes each carry an Indice Annuale alphabetical by surname in the last three images. For each year: read the index for *Nicotra Maria*, then crop that act's margin, which names the husband. Ours married a MAZZA; every other Maria Nicotra at Piedimonte will name somebody else. One index page and one margin crop per year. The window is 1870–1885 — the tree's years are wrong often enough in this family that a single year is not a search (the range rule). Containers for 1866–1893 are one batched fetch away; only 1875 is bought. And if she is in none of them she was not born at Piedimonte, which would itself be the finding, because it would mean the Nicotra came in from another comune. 1879 GOES TOO, AND IT COST ONE INDEX READ. The N section of the 1879 Indice Annuale holds Neri, Nicodemo and Nucifora and no Nicotra whatever. The same page also closed a second question for free: its M section has no Salvatore Mazza, which excludes the tree's «Salvatore Mazza 1879» at this comune and leaves Nati 1873 atto 3 standing with a negative behind it. Fourteen containers are now banked — 1870–1874, 1876–1883 — so the remaining years are reading, not fetching. One route tested and closed: Salvatore's own birth act carries a marginal annotation column, and in this series those name the spouse and the comune of the marriage. His is empty. The act four lines below it carries a full one, so the practice existed and was simply not applied to him — and a blank box proves nothing. His birth act will not tell us where he married. And the post-1893 half needs a different archive entirely. Antenati's Piedimonte holding stops at 1893, so the Mazza–Nicotra marriage of about 1900 and Venera Cassaneti's death in 1925 are not there. The route is the Tribunale copy — «Italy, Catania, Catania, Civil Registration (Tribunale), 1866–1929», already cited elsewhere in this archive — which runs thirty-six years further than the Archivio di Stato's filming. SEVEN YEARS CLOSED, 20 SEPTEMBER: 1870, 1871, 1873, 1874, 1875, 1876, 1879. Every N section read in the original and quoted whole on what has been read. The Nicotra ARE at Piedimonte — a Filippa in 1870, a Salvatore and a Giovanni in 1871, an Ignazio in 1873 — and in seven years they produce no Maria except the 1875 one whose margin marries her to a Nicito. That is now worth saying out loud as a hypothesis, not a conclusion: her birthplace in the family tree may be her *husband's* town copied across. Salvatore Mazza was born at Piedimonte; their son Rocco was born at CATANIA in 1899. If the household had already moved down to the city, Maria may be a Catania girl and no amount of Piedimonte sweeping will find her. Do not act on that yet — seven years is seven years, not the whole window, and 1872, 1877, 1878 and 1880–1884 are still unread with containers in hand. But read them to close the hypothesis, not in hope. A cheaper instrument for the rest: the 1870 and 1872 indexes are written *«Surname Given DI FATHER»*, so where a year does that, a *«Nicotra Maria di Giovanni»* would identify itself in the index without opening an act. 1872 MAKES EIGHT, AND ITS M SECTION PAID BETTER THAN ITS N. No Maria among *Nicotra Angelo 147, Nicotra Giovanna 158, Nucifora Francesca 162* — but the same index gave «Mazza Mariano — 32», and act 32 read in the original is Salvatore Mazza's elder brother, born 12 February 1872 to *«Rosario Mazza DI MARIANO, d'anni ventisette»* and *«Leonarda Bonanno di Gaetano»* (row 163). And the Catania hypothesis just got weaker, not stronger. That Mazza birth is in CONTRADA SAN BASILIO; the 1875 Maria Nicotra who is *not* ours was born in VIA SAN BASILE to Alfio Nicotra, bracciante. The Mazza and the Nicotra were neighbours in the same lane. That is not evidence of anything on its own — half the town were braccianti and lanes are long — but it is a reason to finish reading Piedimonte before reaching for a city of a hundred thousand people. 1877, 1878 and 1880–1884 remain, containers in hand. THE WHOLE DECADE IS NOW READ, AND IT IS EMPTY. 1877 holds a single N entry, *Nicodemo Maria 85*, and no Nicotra at all; 1878 holds *Nicotra Alfio* twice and *Nicodemo Giuseppe*. With the eight already done that is 1870 through 1879 complete — ten consecutive annual indexes, every N section read in the original and every entry named on what has been read. The Nicotra are there the whole time and never produce her. A decade is not a sample, and the year the tree gives her is the year she is most clearly absent from. Still to read before anything is concluded: 1866–1869 and 1880–1884. And 1877 cost nothing and paid twice, because its M section gave a second Rosario Mazza whose wife is also a Leonarda — Leonarda Monacorso, in via Grotta Nicodemo — which is exactly the act that would have been hung on the tree's Orazio Mazza. Recorded as a correction. TWELVE CONSECUTIVE YEARS, 1870–1881. 1880's N section is one line, *Nucifora Francesco 92*; 1881's runs *Nicotra Antonina 24* across a page break into *Nucifora Concetta, Nucifora Agata, Nicotra Mariano, Neri Concetta*. Both halves of that split section were read — a section that ends at the foot of a page is exactly where a partial reading becomes a false negative, which this archive has paid for once already. A lead left standing: the 1880 M section carries «Mazza Gaetano — 66», and *Gaetano* is what the custom gives a second son, after the mother's father — and Leonarda Bonanno's father is Gaetano Bonanno. It may be another child of Rosario and Leonarda or another of the town's Mazza households. Not opened, nothing claimed. — OPENED, AND IT IS THEIRS. *«Mazza Rosario, di anni trentadue, bracciante … da BONANNO LEONARDA, industriosa, sua moglie … via S. Basilio … il nome di GAETANO»*, born 5 May 1880. The street and the mother's surname are what settle it against the other Rosario-and-Leonarda in the town. Row 165. AND THE OTHER SIDE OF THE QUESTION ANSWERED BETTER THAN THE FIRST. Twelve years of Italian indexes have not produced her; her own Australian file names her in one line. NAA: BP25/1, ITALIAN MAZZA M, Brisbane, Item ID 5041955, access Open: *«Mazza [nee Nicotra], Maria — Nationality: Italian — arrived Sydney 20 April 1962 on PHDCG».* Three things at once. It confirms Nicotra as her maiden name from an Australian government record rather than from the family tree. It gives an arrival this archive did not have — widowed in 1929, she came out at about eighty-seven to her sons and died at Brisbane the next year. And an alien registration file of this kind routinely carries a date and place of birth, which is the one fact the whole Piedimonte sweep is trying to reach. It is not digitised, so it is a copy request at Brisbane and not a search — row 164, offered to David alongside the four files already waiting in row 150. FOURTEEN CONSECUTIVE YEARS, 1870–1883. 1882 gives *Nicotra Nunzia 65, Nucifora Rosaria 103, Nicodemo Francesco 114*; 1883 gives *Nicotra Rosa 27* and four Nucifora. The Piedimonte hypothesis has now been tested across the whole plausible window — she died in 1963 at eighty-eight, which puts her birth about 1875, and 1870–1883 brackets that by five years either side. This row now waits on row 164, her own alien registration file at Brisbane, rather than on more Italian indexes: 1866–1869 and 1884–1893 remain, but reading them would be sweeping outside the window rather than closing it. And the same two pages answered a question nobody had asked of them: the 1883 M section holds *Mazza Rosaria 117* and *Mazza Venera 118* — consecutive numbers — and no Orazio, so the tree's «Orazio Mazza, born 1883» has no act there. Flagged, not disproved. AND THE SURNAME THIS ARCHIVE HAS BEEN SEARCHING IS PROBABLY WRONG. The Queensland index was swept by surname: «CASSANETI» returns nothing at all — Maria's own 1963 death registration is the only instance of that spelling in the entire Queensland index, taken down from an informant at a Brisbane registry — while «CASSANITI» returns five, and Cassaniti is the ordinary Catanese form. It is also what the Piedimonte birth index writes: a *Cassaniti Sebastiano* and a *Cassaniti Leonardo Maria* in 1875 alone. A Sicilian register searched for «Cassaneti» will return nothing and the negative will mean nothing. The same sweep also closes a question nobody had asked: eighteen Nicotra registrations in Queensland and not one other names John Nicotra or Venera Cassaneti as a parent, so Maria seems to have had no sibling who died or married there. The Queensland Nicotra are Maccarone, Patanè, Scalia and Carbone households, and none of them is hers. AND THE NAME INDEX CANNOT REACH HER EITHER, 21 SEPTEMBER. *Nicotra* at Catania returns eight deceased and not one is hers — Concetta at Catenanuova, three Giuseppe and a Nunziata at Vicopisano, which is in Tuscany. That is this archive's own `localita=` caution demonstrated in its own data: the parameter matches any place attached to a person, not the place of death, so a query scoped to Catania returned people who died four hundred miles away. *Cassaniti* at Catania returns nothing, control-tested in the same breath against *Arena at Scilla*, which returns sixty-nine. This row still waits on row 164, the Brisbane file. AND THE MARRIAGE INDEX IS A DIFFERENT INSTRUMENT FROM THE BIRTH INDEX. Fourteen years of BIRTH indexes produced nothing. Piedimonte's annual «Indice Annuale per gli atti DI MATRIMONIO» names both spouses — groom above, bride below — on one page-pair for the whole year, so a year is eliminated without opening an act. 1871 is read in full: twenty-six acts, all twenty-six couples written out on what has been read, and NO Nicotra groom and no Cassaniti bride. (There is a «Cassanoti Giovanni» at act 12; he marries an Amante Rosaria.) Sweeping 1866–1885 is about twenty crops, and only the 1871 container is bought. AND THE BIRTH TAVOLA HAS HANDED OVER A CANDIDATE FOR HER MOTHER: Nati 1848 atto 45, «Cassaniti Venera — Cassaniti Gaetano / Venuti Maria». Row 89. This row no longer waits only on row 164. FOUND, AND THE FOURTEEN EMPTY YEARS HAVE AN EXPLANATION. Piedimonte Etneo, Matrimoni 1872 atto 20, 7 October 1872: *«sono comparsi GIOVANNI NICOTRA, di anni trentuno, bracciante, domiciliato e residente a MASCALI, figlio di ROSARIO … e di VENERA MURABITO, contadina, domiciliati e residenti pure in Mascali. E VENERA CASSANITI, di anni ventiquattro, qui domiciliata, figlia del fu GAETANO e di fu MARIA CURRENTI.»* Four new people at generation six and the two blank cards on `/tree/` are filled. HER FATHER WAS A MASCALI MAN. This row read fourteen consecutive Piedimonte birth indexes and the Nicotra never produced her — because a wife went to her husband's comune, and the children of this marriage were registered at MASCALI, which this archive has never opened and never listed. The Catania hypothesis is retired: it was invented to explain a silence that has a simpler cause. This row no longer waits on row 164 at all. Row 173 opens Mascali. AND SHE IS FOUND. Mascali, Nati 1875, atto 144 — born 8 December 1875, father Giovanni Nicotra, mother Venera Cassaniti. This row's fourteen consecutive Piedimonte indexes were read in the wrong town, and the 1872 marriage act said so in six words: *«domiciliato e residente a MASCALI»*. Closed.
162. Done — finished and published this archive
The Riveli di anime e di terre — Piedimonte households in 1714 and 1747–1752
This archive has been saying «Sicily begins in 1820, and before that it is parish records or nothing». That is wrong, and the register index says so. Antenati holds nine volumes for Piedimonte from the Deputazione del Regno: *Riveli di anime e di terre*, 1714 (`qJkjONB`, 283 images) and 1747–1752 (`noPNme4`, 73 images, plus four more volumes). **A *rivelo* is a household tax return sworn by the head of the house, and it lists the house, the head, his wife, his children and their ages — a census of souls and land, a century before any civil register in Sicily. For a family whose Sicilian ceiling is Mariano Mazza, born about 1809, this is the only source that can reach behind him without a parish archive. What it can and cannot do.** It cannot prove descent: a Mazza household at Piedimonte in 1747 is not *evidence* that it is this Mazza household. What it can do is give the shape — how many Mazza households there were, what given names ran in them, and whether the surname is native to the town or arrived later — which is exactly what the repeated Mariano–Rosario–Orazio naming would be tested against. Read it for the shape first and make no joins. Start with the smaller 1747–1748 volume. THE 1747–1748 VOLUME IS READ, AT LEAST ITS INDEX, AND THE ANSWER IS A NEGATIVE WORTH HAVING. Images 1–14 are an alphabetical index with cut letter tabs down the fore-edge — and it is alphabetised by given name, so there is no shortcut: all fourteen openings, about eighty households, were read. There is no Mazza. There are two MAZZEO — *Onofrio Mazzeo, folio 5* and *Silvestro Mazzeo, folio 29* — and the ending was cropped at native resolution on both because it is exactly the kind of near-match that has fooled this archive before. Mazzeo is not Mazza until something says it is. What makes the negative meaningful is how recognisable the rest of the town is: Pagano, Currenti, Grasso, Musumeci, Vasta, Puglisi, Torrisi, Moschella, di Mauro, Ricceri, Zappalà, Patanè — the same surnames that fill the birth indexes of the 1870s and 1880s, read this week. The register is legible and the town is continuous, and the Mazza are not in it. Set beside the nineteenth-century evidence — no Mazza birth at all in 1844 or 1846, two in 1845, four marriages in sixteen years — that is a second piece pointing the same way: a small family, possibly a recent one. Recorded as probable, not documented. What is left in this row: the 1714 volume (`qJkjONB`, 283 images), which is four times the size and may or may not carry an index of its own; and the two Mazzeo folios themselves, 5 and 29, which would say whether that household looks anything like a Mazza one. AND THE TWO MAZZEO FOLIOS ARE READ. The folio numbers sit in the top outer corner with an older number struck through, so the map is five probes wide: folio = image − 16. Folio 5: *«Rivelo che Io ONOFRIO MAZZEO … in virtù di Bando promulgato sotto li 30 Aprile 1747. ANIME: Notato Onofrio Mazzeo d'anni TRENTA; et CARMINA moglie; et GIUSEPPE figlio.»* Folio 29: *«Rivelo che Io SILVESTRO MAZZEO … ANIME: M.ro Silvestro Mazzeo d'anni TRENTACINQUE; et MARIA moglie»*, and then his house — *«un loggio terraneo di pietra a cruda, scagliato dentro e di fuori di calce, per proprio abitare, consistente in una stanza in piano»*: one ground-floor room of rough stone, limewashed inside and out. The surname is settled — MAZZEO four times over, in two hands, never Mazza. And the limit of the source is now known, which is worth as much as the content. A rivelo runs *«Rivelo che Io …»* naming the declarant and the bando, then ANIME, then BENI STABILI in close detail, then the dues. At Piedimonte only the HEAD's age is given — Carmina and Giuseppe and Maria are named and not aged. So a rivelo here will not hand over a wife's or a child's birth year, which is the thing one hopes for from a census. It gives the household and the house. Read it for shape, exactly as this row said. THE 1714 VOLUME IS READ TOO, 21 SEPTEMBER, AND ITS OWN ARITHMETIC IS WHAT SAVED THE READING. Images 1–17 are a summary index — *Nomi | Somme di Lordo | di Netto | Cagione di Riveli | folio* — alphabetised by given name again, 142 households. Twelve of the seventeen were read first, and two of them looked like blank versos at thumbnail size. The running totals said otherwise: p6 closes at 1468 and p8 opens with a *riporto* of 2556; p8 closes at 3774 and p10 opens at 4023. Eleven households nobody had read. Both pages were fetched at full width and hold entries 73–80 and 106–108. A double-entry book checks itself, and this one caught a negative that would otherwise have been eleven households short. No Mazza and no Mazzeo in 1714. So the sequence is: 1714 neither; 1747–1748 two MAZZEO and no Mazza; the civil registers from 1820 Mazza and no Mazzeo — while Pagano, Currenti, Grasso, Musumeci, Vasta, Puglisi, Torrisi, Moschella and dell'Aquila run through all three. That is a shape, not a descent. A rivelo is a tax return; a household can be missing from one; and Mazzeo is not Mazza until a record says so. Recorded as *probable*.
163. Done — finished and published this archive
Mariano Mazza, born 12 February 1872 — the son the naming custom promised
Found in an index opened for somebody else, which is the third time that has happened this week. The 1872 birth index was being read for a Maria Nicotra. Its M section held «Mazza Mariano — 32». Act 32, read in the original: *«L'anno milleottocentosettantadue, nel giorno tredici Febbraro ... è comparso ROSARIO MAZZA DI MARIANO, d'anni ventisette, bracciante ... un bambino di sesso maschile ... nato il giorno dodici corrente Febbraro alle ore dieci pomeridiane, dalla di lui moglie LEONARDA BONANNO DI GAETANO ... contrada SAN BASILIO ... al quale figlio dichiara di dare il nome di MARIANO.»* Witnesses Leonardo Grasso di Sebastiano, fifty, and Carmelo Russo fu Venero, forty. Three things fall out of one act. (1) Rosario and Leonarda have a fifth child, and he is the eldest — the tree gives them four. (2) The first son is named for his father's father, which is precisely what the Sicilian naming custom predicts, and it is now the third independent pointer at Mariano Mazza (b. c. 1809) after the 1845 birth act and the *«di Mariano»* patronymic. (3) Rosario is twenty-seven in February 1872, so born about 1845 — a third statement against the tree's 1850, this one his own at a registration desk. One thing not resolved. Salvatore was born 5 January 1873, eleven months after this boy. Either Mariano died in infancy or the two births are ten and a half months apart. The Piedimonte death registers of 1872–1875 would say, and they are listed and unbought. And the custom is being used the way this archive says it should be — to rank a candidate, never to make an identification. It did not find Mariano; the index did. It only explains him.
164. Next — ready to start David
NAA BP25/1, ITALIAN MAZZA M — re-scoped: it is a Giovanni question now, not a Maria one
RE-SCOPED 23 September 2026. This row was written to answer *where was Maria Nicotra born*, and said «her own file probably says». That question is answered — Mascali, Nati 1875 atto 144, read in the original — so the file is no longer the cheapest way to it.
It is still worth asking for, for a different reason. NAA: BP25/1, ITALIAN MAZZA M, Item ID 5041955, Brisbane, access Open, not digitised: *«Mazza [nee Nicotra], Maria — Nationality: Italian — arrived Sydney 20 April 1962 on PHDCG»*. An Italian alien registration file states a birthplace and usually the applicant's parents. Maria's would name Giovanni Nicotra and Venera Cassaniti, and may give Giovanni's own birthplace as the family understood it — which is exactly the fact that forty-five read index years have failed to settle. A statement by his daughter is not a register, and would be recorded as what it is.
165. Done — finished and published this archive
Gaetano Mazza, born 5 May 1880 — and the custom calls a second child in a row
A lead written down as a lead, and then tested. The 1880 birth index carried «Mazza Gaetano — 66». That was recorded here as *«it may be another child of Rosario and Leonarda or another of the town's Mazza households; not opened, nothing claimed»* — because Piedimonte holds three Mazza households and one of them is a second Rosario with a second Leonarda (row 163 and the corrections). Act 66, read in the original: *«L'anno milleottocento ottanta, addì nove di Maggio … è comparso MAZZA ROSARIO, di anni trentadue, bracciante … nella casa posta in VIA S. BASILIO … da BONANNO LEONARDA, industriosa, sua moglie … il nome di GAETANO»*, born at half past four on the afternoon of 5 May 1880. Witnesses Dimauro Salvatore, twenty-two, and Zappalà Sebastiano, twenty-four. Registered before Ignazio Dr. Fragalà, assessore delegato, not the sindaco. What settled it was the street and the mother's surname, not the name Mazza and not the name Rosario. The other household is in via Grotta Nicodemo and its Leonarda is a Monacorso. And the naming custom has now called two children in a row. Mariano in 1872 for the father's father; Gaetano in 1880 for the mother's father, Gaetano Bonanno, who is named as such in the 1872 and 1873 acts. First son for the paternal grandfather, then the maternal grandfather — the order the custom prescribes. It still did not find either of them; the index did. It explains them. One disagreement recorded rather than smoothed: Rosario is *«trentadue»* here, which would put his birth in 1848, where he is twenty-seven in 1872 and twenty-eight in 1873 and his own birth act says 2 June 1845. A declared age is a statement, not a measurement.
166. Running — a machine or a person is at it now this archive
Caterina Melluso, born Briatico 1860 — a wall with not one record against her
She is on the chart and she is blank. Caterina Melluso, born at Briatico in 1860, died 1957, married Alfonso Prostamo and had six children including Francessco Prostamo, who came to Brisbane in 1957. The family tree gives her no parents, and this archive has never opened a single record for her — her person page carries none. Two documents would do it: her birth act of 1860, which names her father with his age and her mother; or her marriage to Alfonso, about 1880, which names both couples. And the first thing checked was not the records but the listing, because that is what Piedimonte taught today. `briatico-arks.tsv` holds eighty registers. Briatico has 555 — the facets read *Catanzaro (440)* and *Vibo Valentia (115)*. So the same under-listing was sitting on the deepest line in the archive. What is established and what is not. A listing at `s_size=500` returned 500 of the 555 and showed the years running 1810 to 1930 with nothing between 1859 and 1910. That is not yet a finding. Fifty-five registers were not in the sample, and the Vibo Valentia holdings are exactly where a later series would sit. A 500-of-555 listing reported as a gap would be the same error as a half-read index section reported as a negative — paid for twice already this week. `s_size` above 500 times out and the rate limiter closed the session before year probes could settle it. Next, and cheap: one complete listing, or three year probes — 1860, 1880, 1900. If the gap is real, Caterina is a Vibo Valentia archive enquiry or a comune letter, not a search. If it is not, her birth act is one container away.
167. Done — finished and published this archive
Candeloro Donato × Marta Paladino — a marriage the 1816 error had closed
Found by fixing a sentence, not by a search. This archive wrote, of Candeloro Donato: *«the first age this archive has ever had for him, and the last it will get — Scilla's civil registration begins in 1816, so his own birth and marriage are pre-civil … HE IS WHERE THE SCILLA LINE STOPS».* Scilla's registers begin in 1809. His own birth, about 1783, is still out of reach. His marriage is not necessarily. Their son Michele was born at Messina about 1814 and Domenico at Scilla on 18 December 1816, so the marriage is before 1814 — and 1809 to 1813 at Scilla are civil and reachable. What is already read, from the Arena sweep, at no extra cost: 1810, 1811 and 1812 through their annual indexes and 1815 groom by groom — no Donato. 1816 and 1817 were read groom by groom too, and hold none. What is left is 1809 and 1813–1814, the two napoleonic volumes with no index at all — and both fall to the act-head montage that read 1815, 1816 and 1817 in an afternoon. Containers are banked: 1809 `LD6vJel`, 1813–1814 `0MnkW4l`. And if it is in neither, the marriage is pre-1809 and the conclusion this row corrects turns out to have been right for the wrong reason — which is still worth writing down, because a right answer resting on a false premise is not evidence. SIX YEARS CLEAR, 20 SEPTEMBER, AND THE TWO UNINDEXED VOLUMES ARE READ. 1809 (twenty-five acts, the first year Scilla ever registered a marriage) and 1813–1814 (forty-four acts) were read groom by groom off the act head, both to their closing statements. No Donato groom in either. With 1815, 1816 and 1817 that is six years clear. Three are not clear and this row will not pretend they are: 1810, 1811 and 1812 were read through their Tavole for the letter F, on the Arena hunt — and a *Candeloro* sits under C or D. Those three openings have to be looked at again, which is three crops, not three volumes. If they are clear too, the marriage is pre-1809 and parish, and the conclusion this row corrects turns out to have been right for the wrong reason — which is still worth writing down. CLOSED THE SAME DAY IT WAS OPENED, AND THE ANSWER IS THE ONE THIS ARCHIVE ALREADY HELD. The three indexed years were re-read whole rather than for one letter — and they are richer than expected: 1810's indice names both spouses, and the 1811 and 1812 Tavole name both spouses AND all four parents. 1810's C section is one line (Corro Weingarten); 1811's is one line (Candido Giordano); 1812 has no C groom at all. No Donato appears anywhere in any of the three — not as a spouse, not as a parent. So nine consecutive years, 1809 to 1817, hold no Donato marriage at Scilla. Michele was born at Messina about 1814, so the marriage is before 1814; nine civil years exclude it; therefore it is pre-1809 and parish. The conclusion is what this archive said this morning. The reasoning is entirely different, and that was the point — it used to rest on a start year that was false. A right answer resting on a false premise is not evidence. It now rests on nine complete readings.
168. Done — finished and published this archive
More Arena siblings, and the year Orazio and Litteria die
Opened by row 157 closing. Finding one sibling took one crop, because the Scilla marriage Tavole are alphabetised by surname in the Italian-era and restauration years and name both spouses and all four parents — so Arena sits at the top of the book and a whole year is a single A-section crop. Twenty-one containers are banked, 1820 to 1843, and 1820 and 1821 are read. Nineteen years remain and each is one crop. What they can give. More children of Orazio Arena and Litteria Morabito — there is no reason to think Filippo and Giuseppe were the only two. And the year the parents die: the Tavola marks the dead with *«fu»*, so the first Arena entry that writes *«fu Orazio»* brackets his death between that year and November 1821. What they cannot give. Anything above Orazio. His own marriage is pre-1809 and parish — row 158. One caution carried forward: Scilla holds at least four Arena households in this period — Orazio's, Domenico the barber's, Carlo's, and the Rocco of 1821 act 20 whose parents are different again. The parents' column is the only thing that tells them apart, and it is in the Tavola, which is why the Tavola is the instrument and not the act. TWELVE YEARS READ, 1820–1831, AND A SECOND BROTHER. The method held: one crop a year, except that the Tavola starts at image n−3 in some years and n−2 in others, so both get fetched. 1828 opens on him. *«ARENA ANTONIO e DE LUCA SERAFINA — Scilla — ORAZIO, e MORABITO LETTERIA / fu Filippo, ed Alfonzetto Anna — 10 Ag.o 1828 — 26».* Neither parent carries «fu», so Orazio and Litteria were both alive on 10 August 1828 — thirteen years after Filippo's wedding, seven after Giuseppe's. So Orazio and Litteria had at least three sons: Filippo (1793), Giuseppe (c. 1794) and Antonio. And the surname's shape is now plain. Eight Arena grooms in twelve years, in seven households: Rocco (1821, mother Vittoria Fiorenza); Antonio (1822) and Luciano (1827), both sons of *Giuseppe ed Antonia Polimeni* and therefore brothers; Carlo (1826, *fu Antonino e Vittoria Molinari*); Antonino (1827, *Giuseppe e Carmela d'Amico*); Bonaventura (1827, *Rocco e Domenica Samari*); and Orazio's two. Without the parents' column these eight men are one indistinguishable surname — which is why the Tavola is the instrument and the act is not. Four years have no A section at all (1823, 1829, 1830, 1831) and three have a single A entry that is not an Arena. Twelve years remain, 1832–1843, containers banked. AND THE SWEEP FOUND WHAT IT WAS BUILT TO FIND, 1832–1838. The 1838 Tavola opens on *«ARENA DOMENICO e POLISTINA TERESA — FU ORAZIO, e Lett.a Morabito — 4 7bre 1838 — 20».* The «fu» is the finding. Orazio Arena is named living in the Tavole of 1821 and 1828 and dead in 1838, so he died between 10 August 1828 and 4 September 1838 — ten years, every one of them covered by a Scilla death register this archive has already listed. Letteria Morabito outlived him; she carries no *«fu»*. And Domenico is a fourth son, after Filippo, Giuseppe and Antonio. Two of the four married Polistina sisters — Giuseppe took Maria Teresa in 1821 and Domenico took Teresa in 1838, both *«figlie di fu Raffaele Polistina e Letteria Morabito»*. Two sisters and two brothers, or one widow marrying her brother-in-law; nothing read settles it and the death registers would. Two more Arena grooms, neither ours: Francesco in 1833 (*«fu Santo, e Fiorenza Vittoria»*) — which also supplies the cut-off father's name for the Rocco Arena of 1821, same mother, so they are brothers and their father is Santo Arena; and Santo in 1836 (*«fu Domenico, e Autieri Santa»*). Five years remain, 1839–1843. THE RUN IS COMPLETE, 1820–1843, TWENTY-FOUR CONSECUTIVE YEARS. 1839 has no A section; 1840 and 1842 have one A entry each and neither is an Arena; 1841 holds *Arena Rocco × Martello Domenica* («fu Santo») — either the Rocco of 1821 remarrying as a widower or a namesake in the same household — and 1843 holds *Arena Salvadore, calzolaio, × Polistina Maria, filatrice* («Gaetano e Lomanti Camilla»). No child of Orazio Arena marries after 1838, which is what four sons already married by then would predict. The whole run: twelve Arena grooms in about eight households, three of them Orazio's — Giuseppe 1821, Antonio 1828, Domenico 1838, with Filippo's 1815 marriage just outside the range. One form change worth knowing for the next sweep: from 1842 the Tavola adds a Professione column for both spouses. And one practical warning: the Tavola's first page sits anywhere between n−4 and n−2 with no rule, so several years cost a second fetch just to find it. What this row cannot do is go higher. Orazio's own marriage is pre-1809 and parish. What follows it is row 169 — his death, now bracketed to ten years, and a death act names parents.
169. Done — finished and published this archive
Orazio Arena's death — ten years, and every volume is listed
The narrowest window this archive has ever had on a death, and it was bought with one word. The 1828 marriage Tavola names Orazio Arena living; the 1838 one writes *«fu Orazio»*. He died between 10 August 1828 and 4 September 1838. Every volume in that window is already listed. Scilla's *restaurazione* death series runs 51 volumes, 1817–1865, in `scilla-arks.tsv` — so the ten years cost ten containers and ten A-section crops, and the annual death Tavole are alphabetised the same way as the marriage ones. What it would give. A death act names the age, the trade, the address and often the parents — which is the only remaining route to Orazio's own father, a generation this archive cannot otherwise reach because his marriage is pre-1809 and parish (row 158). And LETTERIA MORABITO is in the same window and beyond it. She outlived him — no *«fu»* in 1838 — so her death is 1838 or later, and the same series runs to 1865. Read his first: a widow's death act names her husband, so finding him narrows her, and finding her does not narrow him. CLOSED THE SAME AFTERNOON IT WAS OPENED, IN THE SECOND YEAR READ. The death Tavole turn out to be the same instrument as the marriage ones — surname order, a parents column, and for a married person the spouse in it — so a year costs one A-section crop. Scilla, Morti 1830: *«ARENA ORAZIO — Scilla — MARINARO — Mar.to MORABITO LETTERIA — 12 Luglio».* And the act gives a generation. Atto 64, reached by walking the dates rather than the index, because the Tavola's number is its own row and not the act: *«ORAZIO ARENA, MARITO LITTERIA MORABITO, nato in questo Comune, DI ANNI SETTANTANOVE, di professione MARINARO, FIGLIO DEL FU FILIPPO ARENA E DELLA FU SANTA UNGARO»*, dead at six in the morning of 12 July 1830 in his own house, declared by Diego Arena, twenty-two, and Prospero Arena, twenty-five, both marinari and both *«nipote del defonto»*. Three things. The death is exact to the hour. He was born about 1751, nineteen years earlier than this archive had inferred by working back from his sons and assuming a young father. And his parents are named — FILIPPO ARENA and SANTA UNGARO, generation nine. The naming custom, confirmed after the fact rather than used to identify. Orazio's first son, born 25 March 1793, is a Filippo — named for this Filippo. This archive has spent four months using the custom to rank candidates and refusing to make a join with it; this is what it looks like when a record catches up.
170. Running — a machine or a person is at it now this archive
Litteria Morabito's death, and two young Arena who called him uncle
Three questions left standing by row 169, all of them cheap. LITTERIA MORABITO outlived him by at least eight years. Orazio died 12 July 1830 and she is named without *«fu»* in the 1838 marriage Tavola, so her death is 1838 or later and Scilla's death series runs to 1865. A widow's death act names her husband, so the A section of each year from 1838 will find her the same way his did — one crop a year, and *«vedova di Arena Orazio»* in the parents column is the whole test. DIEGO ARENA, twenty-two, and PROSPERO ARENA, twenty-five, declared his death and the act calls both *«nipote del defonto»* — which in Italian is nephew or grandson and the act does not choose. Born about 1808 and about 1805, they are too old to be sons of Filippo or Giuseppe, who married in 1815 and 1821 — which points at nephew, and a nephew means Orazio had a brother. Their birth acts are pre-1809 for Prospero and 1808–1809 for Diego, so one of them may be reachable and the other is parish. And FILIPPO ARENA THE ELDER and SANTA UNGARO are the end. Orazio was born about 1751, so they belong to the 1720s, and nothing civil exists. That is row 158 — the parish books of the Chiesa Matrice, whose custodians the 1815 marriage bundle already names. SIXTEEN CONTAINERS BANKED, 20 SEPTEMBER — Scilla deaths 1828–1836 and 1838–1844 — so what is left is reading. And the first year read has already exposed a flaw in the plan. The 1838 A section holds no Litteria: *Arlotta Mariano, Arlotta Caterina, Arlotta Antonino, Arena Rocco (11 Agosto), Arlotta Mariano, Arlotta Giuseppe*. But reading the A section is only HALF a search for a married woman, and this archive knows it: Serafina Arlotta is entered as *«ARENA Serafina, fu Giovanni»* in the 1908 index — her husband's surname over her father's forename. So Litteria Morabito may sit under A for Arena or under M for Morabito, and the M section has not been read for a single year. Two crops a year, not one. Recorded as *part* rather than done so that nobody later mistakes six A sections for a search. And one name to come back for: the Arena Rocco who died 11 August 1838 has not been identified against the seven Arena households of the town. FIVE YEARS READ AT BOTH LETTERS, 21 SEPTEMBER: 1838, 1839, 1840, 1841, 1842. Each year's A and M sections both, which is what the flaw above demanded — and every M section is quoted whole on the Scilla reads. Eight Morabito deaths across the five years and not one is a Litteria or a Letteria. Still PART, not done. The window is not closed: Scilla's deaths run to 1865 in the restauration fondo and to 1943 in the Italian one. Containers are banked to 1844. And the by-product is a warning rather than a lead: five Arena deaths turned up in those A sections and none belongs to Orazio's household — Rocco 1838, Giuseppa 1839, Giuseppe and Santo 1841, Giuseppe 1842, whose parents read *«Ant.o e Ros…»*, *«Fran.co e Falcomatà»* and *«Santo e Martello Angela»*. Seven Arena families in one town, and the parents' column is the only thing that tells them apart.
171. Next — ready to start this archive
A status this archive cannot write: read, and the source is empty
Raised by the Falco archive, 21 September, and it is a real hole. They found Arienzo births 1836: a Tavola with a printed heading, ruled columns, and every cell empty. Their point is the one that matters: *«that is nastier than absence, because it files itself in your notes as ‘not read’ and somebody re-opens it.»* `searched.tsv` cannot say it. The vocabulary is `done`, `outstanding`, `not-yet`, `part`, `superseded`, `wall`, `correction`. A blank-but-present source is not `done` (nothing was found), not `wall` (nothing blocked it), and emphatically not `not-yet`. It would go in as `done` with a note, and the note is the only thing stopping the next person opening it again. This archive does not have a clean instance yet, which is why this is a row and not a patch. What it has is the adjacent class: present but unreadable — Scilla Matrimoni 1816 catalogued *VOLUME TARLATO*, the 1842 death Tavola carrying an *INCHIOSTRO SBIADITO* card, the 1845 birth index with seven of eighteen lines too faint. Those are recorded in prose and would be better recorded in the status. Do not invent vocabulary for a case that has not happened. When the first genuinely blank source turns up here, add ONE word and update `check_searched.py` with it — and tell the estate, because Falco has the same gap and reached it first. AND THREE CONCRETE CASES ARRIVED THE SAME DAY. Mascali births 1876 and 1877 were read at the N section, every entry, and their indexes carry no paternità column — so the Nicotra entries there are neither found nor absent. 1874, whose index *does* carry the column, is a clean negative from the identical amount of work. The status this archive cannot write is the difference between those two readings, and writing them both as «part» loses it. AND HERE IS THE CASE MADE AGAINST THIS ARCHIVE'S OWN WORK, THREE HOURS APART. «Cassaniti» in the Antenati name index returns nothing for the whole of Italy, and that was filed as a control-tested real negative. The same afternoon the archive read Cassaniti Gaetano in a Piedimonte Tavola and Venera Cassaniti in two acts. The instrument was read correctly, the source was empty, and the emptiness was a fact about the instrument — Piedimonte and Mascali are not name-indexed. `done` is wrong, `wall` is wrong, and `part` loses the distinction entirely.
172. Done — finished and published this archive
PIEDIMONTE — sweep the marriage indexes 1866–1885 for a NICOTRA groom
The cheapest unopened route into the thinnest quarter of the family. Piedimonte's Italian-era marriage volumes each carry an annual index that names both spouses for every act, on a single page-pair. One crop-pair per year eliminates the year. 1871 is done — twenty-six couples, no Nicotra. The rest, 1866–1870 and 1872–1885, need one batched container lookup (the arks are in `pe-arks.tsv`; only 1871 carries a container) and then about twenty crops. What it would settle: the marriage of Giovanni Nicotra and Venera Cassaniti, which names both sets of parents and would carry generation five of the Sicilian quarter — the side row 161 calls the one nobody has ever searched. And it tests row 89's candidate — the Venera Cassaniti born at Piedimonte in 1848 — without another birth sweep. Mind the DUPLICATA: these volumes are photographed twice and the numbers at the top right are folios, not act numbers. 1870 AND 1872 DONE, AND 1872 WAS THE ANSWER. 1870's N section is one line — «Nucifora Michele, Lo Coco Nicoletta, 15». 1872's N section is «NICOTRA GIOVANNI, E CASSANITI VENERA — 20», and act 20 is legible whole on a single page. The instrument worked exactly as predicted: two crops a year, no act opened to eliminate a year. 1873, 1874 and 1875 have containers banked and are no longer needed for the Nicotra question — but they remain worth reading for the other Piedimonte households this archive is tracking, and the Mazza sons' marriages fall in the 1890s and beyond, past Antenati's run.
173. Done — finished and published this archive
MASCALI — opened, and Maria Nicotra's birth act found in it
Closed 21 September 2026. Matrimoni 1872 atto 20 put Giovanni Nicotra «domiciliato e residente a MASCALI», so his children were registered in his comune and not his wife's. Mascali, Nati 1875, atto 144: Maria Nicotra, born 8 December 1875 at half past nine in the evening, contrada Sopra Pietrafitta, declared two days later by her father in person. The act is on the Mascali page.
498 volumes listed and banked. What is still open here has its own rows: the birth Tavole need re-reading whole (row 180) and no death has been sought (row 183).
174. Next — ready to start this archive
Re-test this archive's own zeros with the most suspect parameter removed
A control licenses the SHAPE of a query and says nothing about the VALUE. This archive's standing rule — *a zero is not recordable until a control in the same form and session returns a known result* — was satisfied in the Lerena archive on 21 September and the zero was still false. A parent-pair query with the mother's surname returned nothing; dropping the surname returned seven, two of which that archive had already published, one of them displaying the very surname it had asked for. The name sits unparsed in the given-name field, so supplying a surname excluded every record the parser failed on. Larriera has one spelling and controlled fine; Villademoros has four. What to do here. Every zero this archive has recorded from a MULTI-FIELD query should be re-run with the most suspect parameter removed — the rare surname, the `localita`, the year range — to see whether the count moves. The candidates are the name-index nils: *Prostamo* at Pizzo, *Cassaniti* at Catania, *Nicotra* at Catania, *Mariantonia Musolino* at Scilla. This archive already knows the other face of the same fault — `localita=` at Antenati matches loosely and returned deaths at Vicopisano, Tuscany, for a Catania query. Loose in one direction, silently exclusive in the other. A field you did not put there yourself is a field you do not control. DONE THE SAME HOUR, AND IT FOUND SOMETHING WORSE THAN A FALSE ZERO. All four nils were re-run with `localita` removed and the raw page inspected. The nils are genuine — «cognome=Cassaniti» with no locality at all returns *«Nessun risultato trovato»*, and the Musolino control still returns 7. But two things were wrong anyway. (1) The detector. This archive reads a count by matching «N risultati», and the zero page does not contain that string — its count is a sentence in a 266 KB page. So «no match» was indistinguishable from a throttled page or a broken selector, and only a second signal told them apart. (2) The meaning. «Cassaniti returns nothing» was filed as a negative *about this family*. Within three hours this archive read Cassaniti Gaetano in the Piedimonte 1848 birth Tavola and Venera Cassaniti in two acts. They were in the registers the whole time; Piedimonte and Mascali are simply not name-indexed. The nil was true and carried no information. Row 171.
175. Next — ready to start this archive
The nils that cannot be checked by a reader — nine, after the checker was checked
`tools/check_nils.py` is written, wired into the build and reports 27. Of 220 searched rows, 173 assert no nil and 20 assert a nil *and say what the instrument could have reached*. Twenty-seven assert a nil and say nothing about reach. They may every one be sound — the checker cannot judge that, and deliberately exits 0. What each needs is one sentence: which comune, which years, which record types the instrument actually covers. The worst of them are the ones that sound most conclusive — «Trove — Prostamo before 1954», «Find a Grave — Prostamo and Melluso across Australia», «TNA Discovery — every surname and town of this family», «Antenati — Archivio di Stato di Messina, whole province». Each is a sentence about a family that may turn out to be a sentence about a search engine, exactly as «Cassaniti returns nothing» was. Coverage fails in three places and all three render the same empty page: at the collection, at the query shape, and at a single field. Only the third is ever evidence about a family, and only once the first two are excluded. Work through the 27 in the order the checker prints them. CORRECTION, THE SAME DAY: THE NUMBER WAS 27 AND IT IS NINE. 27 came straight out of a regex written an hour earlier and never tested against the rows it judges. Three errors, all in the tool. Its nil pattern matched rows full of findings — «not one» caught *«NOT ONE AT NUDGEE»* inside a Find a Grave row reporting eleven burials with plot numbers. Its reach pattern missed plain English: *«BT 26 and BT 27 cover UK PORTS ONLY»* and *«CIVIL REGISTRATION IN THE PROVINCE OF MESSINA BEGINS IN 1820»* are reachability sentences, and were flagged for not containing the word «coverage». And the note claiming it would «under-report rather than over-report» was backwards — it over-reported three to one. Fixed by looking for a BOUNDARY of place, date or record type instead of for a vocabulary. Three rows have since been given the sentence they lacked, two of them badly needed: the Piedimonte birth-index sweeps for Maria Nicotra now say outright that the instrument could only ever reach children registered *at Piedimonte*, and she was registered at Mascali. Nine remain, and the honest lesson is the one this archive keeps paying for — a tool that judges evidence is evidence, and an untested one is an untested claim.
176. Done — finished and published this archive
The 115 documented people are invisible on the Atlas — fixed
The map draws places from the EXPORT's people, and the archive's best-documented people are not in the export. That is why Zambrone — a comune with a read register run and a recorded gap in its deaths, 1813–1821 — has no marker at all, and why twelve of thirteen coverage rows match instead of thirteen. `documented-additions.tsv` now holds 115 people, every one of them added from a register this archive opened: the whole Ventrice line at Zambrone, Mariano Mazza and Maria Catalano at Piedimonte, Rosario Nicotra and Venera Murabito at Mascali, Gaetano Cassaniti and Maria Currenti. None of them is on the map. Why it is not a one-liner. An addition carries a place as prose inside `source` and `note`, not as a structured field — there is no `birthPlace` to join on. Giving them coordinates means either adding a place column to the additions file or parsing the comune out of the source string, and the second is the kind of guess this archive does not make. The honest first step is the column. And it would answer a question the Atlas cannot currently answer at all: where the documented spine actually stands, as against where the family tree happens to name a town. DONE, 22 September. `documented-additions.tsv` now carries a `place` column, and the Atlas has a fourth layer, The documented — 89 people across 6 comuni, 79 of them in no family tree. The column is NOT derived from `source`, and that distinction is the whole job: the source names the comune of the record, and the two come apart exactly where it matters — Rosario Nicotra is named in a Piedimonte act and was a Mascali man, and the act says so in the same line. Parsing the source would have put him on the map in the wrong town. So the column is seeded from the source's leading comune only where that prefix is a known comune, overridden by hand where a document says otherwise, and left blank for 25 rows where nothing states it — blank means *not recorded*, never *nowhere*. ZAMBRONE HAS A MARKER FOR THE FIRST TIME: eight documented people, no exported person, and a comune whose register run this archive has read. Four gazetteer coordinates added (Mascali, Giarre, Riposto, Zambrone), each checked against the bare-name fallback — unlike the Soriano Calabro entry, none of these corrects a collision, and the note says so. And `check_additions.py` now refuses an unknown place, because a free-text spelling would stop matching the gazetteer and the marker would vanish with nothing failing — tested by breaking it.
177. Running — a machine or a person is at it now this archive
GIARRE — «nato a Giarre», eleven years that do not hold him, and a first year that changes the question
A sixth comune, arriving with a document rather than a hope. Piedimonte's *Matrimoni allegati* 1872 fascicolo 20 prints «nato a ___» separately from «residente a ___» and fills them GIARRE and MASCALI.
Eleven Giarre birth years, 1838–1848, are read end to end — main runs and the appendices that sit past Z — and he is in none of them. For a day this archive let that stand for the town. Then 1820, the first year Sicily registered anything, was opened: it holds two untitled Nicotra households, one father a Rosario a generation too early to be Giovanni's. The decade was wrong, not the town.
And Giarre Matrimoni 1844 atto 54 marries the *Dottor Don* Alfio whose parents Riposto had already named in 1838 — one man, and «of S. Giovanni» in one register, «of Giarre» in the other. San Giovanni Montebello is a frazione of Giarre, which is the third reading of Giovanni's problem, documented in his own surname. Nine marriage years and the early births remain: row 181.
178. Next — ready to start this archive
The 25 documented people with no place recorded
89 of the 114 documented additions now carry a `place` and are drawn on the Atlas. Twenty-five do not, because their source is an index or a registration rather than a comune act — the Antenati name index, Queensland death registrations, New York parish baptisms, FindMyPast. Each needs a person who has read the record to say where the person belongs, and a blank is the honest value until then. It is not a parsing job. The rows whose source begins with a comune were filled deterministically; these are the ones where the comune is mid-sentence or absent, and guessing from prose is what row 176 was careful not to do. Work through them against the act, not the string.
179. Running — a machine or a person is at it now this archive
RIPOSTO and MASCALI marriage Tavole, swept for Rosario Nicotra × Venera Murabito
Twelve Riposto years (1835–1846) and eight Mascali years read whole. Rosario is in none of them. What the sweep did find is Riposto Matrimoni 1844 atto 39: *Nicotra Salvadore*, patria Giarre, son of Alfio Nicotra and Rosa Pappalardo, marrying Maria Barbagallo of Riposto — the father of the two Riposto children the birth sweep had already turned up in 1845 and 1847. He is not Giovanni's father and the pages say so twice. He settles the argument behind the search: see Riposto and Giarre.
180. Next — ready to start this archive
RE-READ the Mascali birth Tavole whole — the junction method is not safe at this comune
Mascali Matrimoni 1847 writes its sections in the order C, N, M. 1841 and 1835 put V before S. So at Mascali «the alphabet runs M straight into P» does not prove there is no N section — the N may sit somewhere else entirely, and the only safe reading is the whole Tavola end to end.
The eight Mascali marriage years on the Nicotra line were all read that way and are sound. The nine Mascali birth years of 1839–1848 were read at the M–to–next‑letter junction, which was the method before this was known. They are recorded here as provisional until each one has been read whole. Nothing found is withdrawn — Alfio Nicotra and Carmela Sudieri are still there in 1840 and 1847 — but the *negatives* are weaker than they were written.
181. Next — ready to start this archive
GIARRE, the early years — Salvadore Nicotra's birth act and the marriage Tavole
Two sweeps, one queue. Giarre Nati 1821–1826: Salvadore Nicotra married at Riposto in 1844, so his own act sits about 1815–1824, and 1820 — the first year Sicily registered anything — already holds two untitled Nicotra households, one of them a Rosario a generation too early to be Giovanni's father. Salvadore's act would name Alfio Nicotra's age and trade and say whether Alfio, Simone and Rosario are brothers. Giarre Matrimoni 1835–1846: where Rosario × Venera Murabito would be if he married in the town his son was born in.
Containers for Giarre Matrimoni 1835–1844 are now banked, and 1844 is read — it gave the Don Alfio who is «of S. Giovanni» at Riposto and «of Giarre» at Giarre, and 1840's whole Tavola holds no N at all. Nine marriage years and the 1821–1826 births remain. Blocked by the rate limiter and nothing else. Twenty-one ark→container lookups were attempted on 22 September and exactly one returned. The limiter is cumulative, not per-interval — a three-second pause buys about twenty-eight lookups, not immunity — and a fifteen-minute wait after a lock was not enough. The images themselves are never throttled. This is a queue of about twenty-five containers a session, not a wall.
182. Next — ready to start this archive
The MASCALI RIVELI of 1616 — read the name index through, all of it
Forty-six volumes at Palermo, not Catania, and not civil registration: *riveli di beni e di anime*, household returns to the Sicilian parliament. 1616 is two centuries before anything else this archive has read and the oldest source of any kind in it.
The first volume is open and it is name-indexed, with the letter tabs cut into the fore-edge of the book. The catch decides the method: the index is ordered by FORENAME, not by surname — the N tab leads to Natale and Nicolò, not to Nicotra. So the whole index has to be read through, which is perhaps twenty images rather than five hundred and forty-six. The surnames on one opening are already the ones the 1840s registers are full of: Barbagallo, Musumeci, Finocchiaro, Grasso, Gagliano, Pinnisi, Strano, Arcidiacono.
No line of descent will come out of this — nothing joins 1616 to 1875. What it can do is say whether the surname was in this town before the registers begin, which is a different and honest question.
183. Next — ready to start this archive
The DEATHS of Rosario Nicotra and Venera Murabito — 296 volumes, not one opened
Both were alive and unmarked by «fu» in October 1872, in an act that writes «fu» for a witness two lines above. Neither has been sought in any register of any town.
Mascali holds 78 Morti volumes, Giarre 147, Riposto 71. They were «of Mascali» in 1872, so Mascali first. An Italian-era death act names the deceased's parents, which would reach generation eight — one further than their own marriage would. That makes this better value than the marriage sweep it has been queued behind, and it is the only line of attack that does not depend on finding Giovanni first.
184. Next — ready to start this archive
Three Queensland articles Trove has indexed and will not show — and one of them contradicts row 105
Trove matched these inside the text and serves `status: currently unavailable`; both article URLs 404. The words exist and are reachable another way — the National Library, or the paper's own microfilm.
- News-Mail (Bundaberg), 6 July 1962, «THREE NEW CITIZENS» — snippet: *«Supported by friends and relatives, three former Italian citizens received certificates of...»*. SEBASTIANO MAZZA is indexed in it.
- News-Mail (Bundaberg), 28 September 1963, «Family Notices» — names him again.
- A Queensland «Family Notices» of 17 February 1958 naming a CONCETTA MAZZA.
AND THE FIRST ONE FIGHTS SOMETHING THIS ARCHIVE ALREADY ARGUES. Row 105 reasons that Sebastiano Mazza was naturalised before 1959, because only British subjects could enrol and he is on the 1959 roll at Wacol. A naturalisation ceremony at Bundaberg in July 1962 does not fit, and Bundaberg is three hundred and fifty kilometres away. Either it is a different Sebastiano Mazza, or one of the two readings is wrong — and the archive holds a naturalisation file for him, NAA J25 1971/1074, which would settle it.
185. Next — ready to start this archive
FOUR NICOTRA WILLS at Queensland State Archives, open access, and a North Queensland cluster nobody has looked at
Found through Trove's archival category, which this archive had never searched — rows 7, 79 and 153 are all about newspapers.
Series S6220, Supreme Court (Northern) Ecclesiastical Files (Wills), Queensland State Archives, all OPEN ACCESS:
- NICOTRA, SALVATORE PHILIP — 1955 — item ITM938108
- NICOTRA, LEONARDA — 1962
- NICOTRA, VINCENZO — 1965
- NICOTRA, FILIPPO — 1966
A will file names the deceased, the death date, the address, the occupation, the executors and the beneficiaries — usually the whole household. That is better than anything Trove's newspapers hold for this family.
All four are Townsville district, and north Queensland has now come up three times. The only two Nicotra family notices in the entire newspaper sweep are the *Evening Advocate* (Innisfail), 1 May and 17 September 1953. The Lands Department holds three Nicotra land files for Innisfail and Cardwell Shire. And this archive already places the Crupi brothers at Tully, forty kilometres from Innisfail.
Whether any of it is Maria Nicotra's family is unknown. She died at Brisbane in 1963, four hundred kilometres south, and nothing here is attached to anybody. What makes it worth the trip is that a will would settle identity outright rather than adding another unjoined name.
Also recorded and not pursued: «Naturalisation — Central District — MAZZA, James, 1898» at Rockhampton, half a century early; and court files under Cassaniti and Murabito, both Nicotra-line surnames, dated 2014–2018 — modern, and touching the living.
Nothing in this archive is in that state.
The other work lists
Each archive keeps its own, in the same shape. Nobody has a list of everybody — that is the point: the work sits with whoever can do it.
**How to read the order.** The numbers are the order work was raised. From row 32 they run as a roadmap: **32–34** finish the Ventrice hunt; **35–37** the cheap sweeps the annual-Tavola trick unlocked; **38–47** the spine frontier town by town, which is where the parentless ancestors actually are; **48–50** Australia; **51–55** the structural work that decides whether the site can show what the archive knows; **56–57** what only a living relative can answer; **58–63** what the September walls left behind. **64–88 are the second tranche, raised after those walls were established:** **64–72** the Italian records the walls did *not* close — processetti, death registers, and two towns never queried; **73–77** structure; **78–80** Australia; **81–83** the estate and the method; **84–88** the loose ends today's work exposed. Items founded on a mistake are struck through rather than deleted, the same way a retracted claim is.
Where the other registers are
This page is deliberately narrow. It does not hold findings, and it is not a list of questions about the family — those have their own pages and are better kept there.
- What has been read — every search made, negatives included, so nobody repeats one.
- Open questions — what the records have not answered.
- Corrections — every place a document and the family tree disagree, and the people the tree does not hold.
- What changed — how the archive has moved since it was built.