A record of the places themselves — the cities, fortresses, and churches that stand behind the prints and the family lines, as they are now. Two entries are below, in the order they were visited; the rest are still being sorted.
Torre de Belém
c. 1719
26 July 2026
We climbed the tower on a clear morning at the end of July and stood on the terrace for a while, which is the second photograph. The plate beside it comes from the same book as the view of Nagyvárad in the collection — the 1719 Frankfurt edition of Mallet's geography, Figure CXXVII where Waradin is Figure L.
The German caption above the platemark reads der thurm Belem oder Bethlehem — the tower Belem, or Bethlehem — glossing the Portuguese name for a German reader. Belém is Bethlehem: the district takes its name from the hermitage of Santa Maria de Belém that stood on the shore before the Jerónimos Monastery replaced it. Inside the engraving the banderole still reads TOVR DE BELEM in French, because only the caption was re-lettered for the German market and the copper plate itself never was.
Mallet came to geography by way of soldiering — sergeant-major and inspector of fortifications, part of his career spent in the service of the King of Portugal before he returned to France to teach mathematics at the court of Louis XIV. His Description de l'Univers is a soldier-engineer's world: harbours, siege works and city profiles drawn as if by someone assessing them. That service does not prove he stood on the Tagus and drew this tower from life; the book leans heavily on earlier printed sources and its views are frequently conventionalised. But he knew the country at first hand, which is more than can be said for most compilers working up a view of Lisbon in a Paris workshop.
What changed, and what did not
Three centuries separate the two views, and almost everything that differs between them is explicable.
- The waterThe tower was built on a small basalt island close to the north bank, not out in midstream, and it did not move when the river did. The shoreline was progressively filled and built out until the island was absorbed into it. Mallet's plate already shows the tower reading as part of the shore; today it stands in an ornamental basin reached by a timber causeway.
- The skylineThe crown of merlons, the bartizan cupolas, and much of the decorative stonework in the photograph are the work of the military engineer António de Azevedo e Cunha, who in 1845–46 pulled down the barracks grafted onto the building and restored it to a nineteenth-century idea of the Manueline. The engraving predates that campaign by roughly 130 years.
- The bastionThe low projecting work in front of the tower is genuine and original — the hexagonal gun bastion that gave the building its purpose. It survives, and it is recognisably the same element in both images.
- The shippingMallet fills his river with sail because that was the point of the place: Belém guarded the approach to the harbour, and every vessel in the plate is passing under its guns. The photograph's empty water is the more eloquent comment on what changed.
What the engraving is doing is not portraiture but position — a fortified tower commanding a river approach, with the traffic it exists to control passing beneath it. Its documentary value is its date: before the earthquake of 1 November 1755, and before the restoration that gave the tower much of the silhouette it has today.
The Valley of Andorra
“THE VALLEY OF ANDORRA, IN THE PYRENEES, A SEQUESTERED MOUNTAIN REPUBLIC.”
The letterpress caption below the image. Note what it is doing: it is telling a British reader what Andorra is rather more than where the view was taken.
Andorra sits in a cluster of high valleys on the southern flank of the eastern Pyrenees, drained by the Valira and hemmed in by France to the north and Spain to the south. In 1890 it had no railway and, effectively, no carriage road: you reached it by bridle path over passes that closed with snow. That isolation is precisely what the caption is selling, and it is why a London weekly thought a picture of an empty valley was worth a page.
It was never a republic
Andorra is a co-principality and has been one since 8 September 1278, when a paréage agreed at Lleida divided its lordship between Roger-Bernard III, Count of Foix, and Pere d'Urtx, Bishop of Urgell. The Foix half passed to the French crown when Henry of Navarre, Count of Foix, became Henry IV of France, and it descends today to the French head of state; the other half is still the Bishop of Urgell's. Nineteenth-century writers called the place a republic almost by reflex, because its day-to-day government lay with an elected General Council and because the fact worth remarking on was that a miniature self-governing polity had survived at all, unabsorbed by either neighbour.
The editorial logic of the original page bears that out: records of the same leaf describe a second illustration printed beside this one, showing the town and castle of Foix in France as the seat that anciently held the lordship of Andorra. The two views went out as a single explanatory unit — the valley itself, and the French half of the medieval title that still governed it.
Two hands, neither of them named
There are two marks inside the image. Lower left, cut into the foreground stonework, is a cursive signature that reads most plausibly as Hennings, though Hensings and Jennings are defensible readings of the same strokes. By the conventions of the illustrated press that position carries the draughtsman, the artist who put the design on the block — and no draughtsman of that name is presently identifiable among the hands working for the London weeklies in 1890. Either the reading is wrong, or the drawing was worked up in the office from a photograph or a traveller's sketch by someone whose name never reached the reference literature.
Lower right, in block capitals, is GREENAWAY: the engraver's position, the craftsman who cut the block. A Greenaway wood-engraving hand is well attested in the London illustrated press — John Greenaway (1816–1890), who engraved for the papers and the book trade, and who was Kate Greenaway's father and first teacher. He is the obvious candidate, and he died in the year this plate was published. But a shared surname is not an attribution, and workshop practice of the period means the mark may name a firm, or a son, rather than the man. Recorded as a candidate, not a finding.
What the sheet can and cannot prove
- PublicationThe Illustrated London News, 1890 — strongly supported. Three independent trade records describe the identical page and agree on both paper and year. That is strong, but it is still trade attribution rather than documentary proof
- Issue date13 September 1890, reported — open. It comes from the least reliable of those records, which names a paper that had ceased publication in 1863
- Why the object cannot settle itThe leaf was trimmed out of its page, so it carries no head-of-page dateline, and the verso holds unrelated letterpress rather than the Andorra article. The text a reader would want went to the guillotine with the rest of the page
- The way inThat unrelated verso text is unique to one page of one issue. A distinctive phrase transcribed from it and searched against the digitised run would identify the leaf outright, and with it the date, the article, and very likely the artist credit
3 August 2026
The valley in the plate is a place you can still stand in. We went in August; these are three of the things that are still there.
The round tower at Santa Coloma is the one building in the country nobody confuses with another: a twelfth-century Lombard bell tower, circular where every other bell tower in Andorra is square, built against a nave older than itself. The bridge is the other kind of survival — one of the dry-laid stone pack-bridges that carried the only routes in and out before there were roads. It is the world the 1890 caption meant by sequestered, and enough of it is still standing to make the word legible.
The engraving is a view from a height of a valley with almost nothing in it. What has changed since is not the shape of the ground but the reason anyone can see it: in 1890 the picture was worth printing because so few readers would ever get there.
Why this page exists
It is the other half of the Collection. I buy prints of places I have seen with my own eyes; the rule is deliberate, and it is what keeps the collection from becoming a set of pictures of anywhere. This page shows the same places as they stand now, three or four centuries after the engraver saw them.
More entries are coming: Nagyvárad and the fortress at its centre, and the other sites in Hungary and well beyond it that have earned their way into the collection. The photographs exist; they are not yet in any order worth showing.