Eight cities · 41 historical sheets
Old maps, laid over the cities they became.
JuxMaps stretches scanned historical maps onto their true geographic position, so you can fade or swipe between a city’s past and its present. Built for researchers, journalists and anyone curious about how a place got its shape.
London
Reference points in London
How the fit is checkedAn old map is only useful once it sits in the right place. These are long-lived features — rivers, gates, cathedrals, high ground — that appear on both the historical sheet and the modern map. Tap one to fly there and judge the alignment yourself. If a landmark lines up but the streets around it do not, the difference is history, not error.
Old maps of London
Select a sheet to lay it over the live map above. Select it again to take it off. Each card shows the date on the badge; where the source record carries no date, the badge says so.
How the tool works, in plain terms
A scanned old map is just a picture. It has no idea where on Earth it belongs. Turning it into something you can lay over a modern map takes five steps — four of them happen before JuxMaps ever loads.
-
1
Start with a scan
A library or archive photographs an old map at high resolution. Some sheets here are enormous: Rocque’s 1746 London is roughly 30,000 pixels across, assembled from 24 separate plates.
-
2
Find the same place twice
Someone opens the scan beside a modern map and clicks pairs of matching points: this church tower here, the same church tower there. Each pair is a ground control point.
Good control points are things that do not move — a cathedral, a city gate, a bridge crossing, a hilltop. Coastlines and field boundaries make poor ones.
-
3
Stretch the picture to fit
Software uses those pairs to warp the scan — stretching, rotating and bending it until the control points land on their real coordinates. Cartographers call this georeferencing, or rubber-sheeting, because the image is pulled like a sheet of rubber.
Old maps were drawn before modern surveying, so the stretch is never uniform. The fit is excellent near control points and looser between them.
-
4
Cut it into tiles
The warped map is sliced into thousands of small square images, one set per zoom level, in the same grid the modern map uses. That is what lets you zoom and pan smoothly instead of downloading a gigapixel image.
-
5
Juxtapose and judge
JuxMaps loads the historical tiles into a layer sitting directly above the modern base map, in the same coordinate system. Fade dissolves between the two; Swipe draws a curtain you drag across the screen.
The reference points above are the check on all of it: fly to one and see whether the old map agrees with the modern one.
-
6
Or add one yourself
Steps 1–4 already happened for thousands of maps on Map Warper and Wikimaps Warper, open platforms where anyone can georeference a scan. Found one that isn’t in our gallery? Paste its link under any city’s gallery and JuxMaps will fetch it, confirm it actually serves map tiles, and place it — nothing is uploaded to us.
This deliberately stops short of accepting a raw image upload. Turning a scan into something that can sit on a map is steps 2–3 above, and doing that well takes real georeferencing tools, not a quick fit in a browser. Send well-chosen maps to Map Warper instead — once they’re georeferenced there, they can come straight back here.
What this prototype does not claim
- Alignment is approximate. Every overlay was warped by a volunteer contributor, not by this project. Fit quality varies from sheet to sheet, and we have not independently measured the error on any of them.
- Dates come from the source record. Where an uploader did not enter a date, the gallery badge reads “undated” and the Sources list says so. We have not verified dates against library catalogues.
- Control points are not published. Map Warper holds the actual point pairs for each sheet, but its export is not open to automated access, so we cannot show you counts or error figures here. Each Sources entry links to the record so you can inspect it directly.
- Reference-point coordinates are measured; the choice of them is editorial. Positions were resolved through OpenStreetMap’s Nominatim service; which landmarks to include, and the founding dates shown, are ours and approximate.
- Do not measure off these overlays. Use them to see change and to generate questions — then go to the archival source before publishing a claim.
- Maps you add yourself are session-only, and unreviewed by us. They disappear when you close the tab, unless the page URL still names them, in which case they’re re-fetched and re-checked live — never stored. We don’t vet them editorially the way we did the curated gallery; the same tile-serving check runs, but title, date and description come straight from the archive record.
Where these maps come from
Every historical overlay here is served from an open georeferencing platform, where contributors upload scans of old maps and align them by hand: Map Warper for most cities, and Wikimaps Warper for the British Admiralty charts of Cape Town. Each entry below names its archive and links back to the record, where you can see the scan, the contributor and the control points used.
Map Warper
Open georeferencing platform by Tim Waters. Supplies the historical tile layers for seven of the eight cities. mapwarper.net
Wikimaps Warper & Wikimedia Commons
The Wikimedia-hosted instance of the same software, serving the British Admiralty charts of Table Bay. warper.wmflabs.org
OpenStreetMap
Base map data, and the Nominatim service used to resolve reference-point coordinates. Data under ODbL. openstreetmap.org
Esri & OpenTopoMap
The street and light/dark canvas base map styles, and satellite imagery, from Esri with data by Esri, HERE, Garmin, Maxar, Earthstar Geographics and OpenStreetMap contributors; topographic tiles from OpenTopoMap (CC-BY-SA). esri.com · opentopomap.org
Leaflet
The open-source mapping library that draws and synchronises both layers. leafletjs.com
Named collections
Where a record identifies its holding institution we credit it in the list above — for example the LLILAS Benson collection at UT Austin, and the Survey of Palestine.