The Madrid death record is from 1891.
The witness line says born in Xàbia, Marina Alta.
I underlined Xàbia, checked adjacent years for the same name, and booked the trip.
Train to Valencia before first coffee, bus to Xàbia, uphill to the municipal archive.
I kept the request short. Padrón 1880 to 1900, municipal minute extracts if they exist.
Two padrón volumes arrived.
One household fits the surname if I allow a single vowel shift in the father’s given name, address inside the Old Town catchment.
A second entry two years later shows a son listed as jornalero. I copied folios, address, and occupations. Minute extracts gave me nothing on the first pass. I marked bad harvest years to recheck, since movement often follows those.
San Bartolomé parish.
Index by year is usable. Script climbs the line in Valencian. Marriage window 1885 to 1890. One from 1887 puts a godparent with the same surname cluster as the Madrid witness. Spelling varies in the expected ways. I copied date, page, and witnesses, then set a baptism window based on padrón ages and handed in a written request.
While they pulled, I walked to the address. Narrow lane, stone walls, heavy door, air that stays cool even when the street does not. Three corners away the new builds start, white and glass and clean edges. Estate windows full of stock. If you track the market you will find plenty of villas for sale in Javea including Ibizan style properties with flat roofs and smooth walls. I prefer stone and a door that shuts like it means it.
Back at the parish desk there are two baptisms inside my range, parents’ names stable, surname written as in the marriage. No direct appearance of the Madrid witness yet, but the cluster repeats at the same address.
I added notary protocols to the plan in case a move left a paper trail. Also port and cooperage material. One line calls a godparent tonelero, which fits seasonal barrel work and explains why a household shifts one lane and comes back.
Afternoon in the municipal room again. I asked for the padrón years before and after the two hits. The family slides one lane one year, returns the next. Ages move in step. A daughter disappears, likely a marriage or a move. I listed every spelling variant in the section and marked which ones are phonetic slips and which look like distinct branches.
I built a quick evidence table in the notebook, four columns, nothing fancy. Source and date. Location and folio. Name as written. Role and link strength zero to three.
The strongest link today is the 1887 marriage with the godparent surname and the occupational hint, plus a padrón entry that fixes the family to a street within the same parish catchment.
The weak point is assuming the Madrid witness equals the padrón son whose recorded age runs two years high for the Madrid event window. That could be rounding. It could be another man.
I wrote three tests for tomorrow. Resident lists in municipal minutes tied to sanitation or conscription. Parish deaths 1888 to 1892 to confirm the household shape. Ask at the diocesan archive in Valencia about duplicate parish copies if Xàbia has gaps.
I called my cousin in Ohio in the evening. Read the names slowly. He said his mother used one variant when she was angry at his grandfather. I logged it as oral, unverified, worth checking after I have a second hard point on paper. Then I rewrote the best entries by hand to fix the letter forms.
Early start again. Parish marriages for collateral lines. Municipal fines by street.
Back to the padrón with fresh eyes. If it resolves, I will request permission to transcribe the marriage and baptisms in full. If it does not, I still leave with two solid references, a tighter household map, and a next step that is clear.