The register

Domenica, Dominica, Lucia, Michiela, Pascha, Pasqua, Veneranda, Maria, Michaela, Cattarina, Anna — about forty-nine of the children's records sit on film 005497893. That film's catalogue creator reads «Rimokatolička crkva, Župa Vižinada». It is not Vižinada's. Tested image by image earlier the same day, two of the hundred and twelve town-less records on that reel are Vižinada's; the rest are on the second item bound to it, the books this archive identified in September as Vodnjan's. So the largest block of «lost daughters» are Vodnjan's own daughters, and the Vodnjan marriage register had already said so. Read in September, it gave «four women of the name married out before the books open — Domenica, Antonia Maria, Lucia and Veniera», recoverable only because their daughters' weddings named their mothers. The index has been holding their children the whole time, in rows with no place on them. Inferred — proven not to be Vižinada's; Vodnjan is the only other parish on the reel

Francisca Defranceschi — nine children, 1831–1846, by Antonius Cerneca — has been the head of this queue since it was built. All nine of her rows are on film 005497886: Svetvinčenat, the deaths-and-marriages reel. She is not alone there. Veneranda Defranceschi's nine rows are on 005497885, Svetvinčenat's baptism reel, and Sinforosa Defranceschi is on 005497886 too — and Sinforosa is already in the nineteen Svetvinčenat marriages read out of that book this morning: «Andrea Vernier × Sinforosa Defranceschi, 1801». Two routes, run hours apart, landing on the same woman.

Marietta Defranceschi — three children 1843–1871 — is at Dubrovnik, two rows on the city parish's reel and one on another. Antonia Defranceschi, 1796, is at Rijeka. Dominica and Domenica Defranceschi have five rows between them on Labin's reel. Catharina Defranceschi has one at Barban. None of those four towns would ever have been guessed from the rows, because the rows say nothing at all.

Placing the rows by parish does something the queue could not: it splits the names apart. «Maria Defranceschi» is not a woman. Her forty-five rows land in at least thirteen parishes — Fažana nine, Pula six, Vodnjan four, Ližnjan three, Medulin three, Vižinada two, Labin two, and one each at Svetvinčenat, Pazin, Rovinj, Dubrovnik and Krbune. Some of those are certainly the same person moving; most cannot be. «Lucia Defranceschi» is three women — Pomer, Vodnjan and Ližnjan. «Veneranda» is two, Svetvinčenat and Vodnjan. «Dominica» is two, Vodnjan and Labin. And the reverse also holds: «Mariana Defranceschi» and «Mariana De Franceschi» are one woman, both at Krbune, counted twice because of a space. So the figure of twenty-six was a count of spellings in an index, not of people. This archive has said several times this week that a first count has been too big. This one was too big and too small at once.

Eight reels remain unnamed, and they hold the rest: 005494789 (nine rows of a Maria Defranceschi, the single largest unresolved block), 005494797, 005494798, 005494788, 005497818, 005497822 (five rows of Mariana — expected to be Krbune, where a Mariana De Franceschi × Vincentius Marziol was found the same day), 005497825, 005497835. 005497826 came back carrying three parishes at once — Pazin, Pomer and Plomin — and needs the ark test rather than the catalogue. Eighteen rows matched no record in any harvest and need re-fetching one at a time. The work stopped where it did because FamilySearch returned 403 for the second time in an afternoon. It is a throttle, not a wall.

The clearest household to come out of the queue. Lucia Defranceschi × Matthaeus Lazar — written Lazaro, Lazar and Lazar again — at Pomer, the village on the bay south-east of Pula. Antonius Matheus 1870 · Matheus 1873 · Jacobus 1875 · Zorica «or Albina» 1880 · Josephus 1882 · Catharina 1884. And in the same year as the last of them, the eldest daughter's wedding: Catharina Lazar × Joannes Zuccon, 1884, an entry that names Lucia as the bride's mother and so closes the circle the queue was built on — a woman recoverable only because somebody else's marriage had to name her. Pomer was one of the twenty-one towns found this morning to be in this archive's register but not in its atlas, on the strength of three rows all spelt Defrančichi and none of them ours. It has a De Franceschi family in it after all.