Statistics
The Itinerant Crown
A digital humanities project by Creagh Factor. The source texts were digitized from public-domain scans of the printed calendars; this project builds an independent analysis layer on top of them.
The court in motion, year by year
Labelled dots: where instruments were dated that year (size = share), usually but not always where the king was. Arrows: how the single busiest dating place shifted from one month to the next (a coarse proxy for the court's movement, not a recorded itinerary).
What's in this map?
The dot marks the dating place, where the great seal was when an instrument was sealed. Usually the king's household, but the chancery could sit apart from the king, more often late in the period. Continental seats (Bordeaux, Antwerp, Ghent…) appear too; pan out to see them. Only dated, geocodable entries are shown.Where the court travelled, by place
Pick a seat to see where the great seal came from and went next: its network of moves, ranked.
Arrived from
Departed to
What counts as a "move"?
A move links two consecutive dated entries whose places differ and fall within 60 days of each other (the window is enforced in code, so long jumps across gaps between rolls are excluded). This is a far richer signal than the monthly map below, catching secondary seats like Norwich. Distances are straight-line, so real road journeys were longer; same-day entries at different places add some noise.Plan a journey: from one seat to another
First we show journeys the seal actually made between the two places, in date order (the dark line) — runs where one was a later stop than the other on a continuous itinerary. If it never travelled between them in sequence we say so, and only then show a hypothetical route stitched from the busiest separate corridors (soft gold) — clearly not one trip. Type to search the place boxes.
Other possible corridors
The whole network of seats
Every place the seal moved between, drawn at its real location. Click a place to light up its connections; overlay the real medieval geography below for context. For the same links as an abstract, explorable graph (sized, popularity-adjustable), see the location network.
The base map is CARTO's label-free style (no modern roads or place names behind the data). Terrain relief is Esri World Shaded Relief, so hills and river valleys read without clutter. © OpenStreetMap contributors.
Reading this network & sources
Circle size is how often a place was a stage on the seal's travels; line weight is how many journeys ran along that corridor, both computed from the dated, geocoded entries (see the place explorer's note).
The overlay layers are real published datasets, not derived from the rolls, shown for geographic context. Each is external research data, reprojected to the map but otherwise unmodified:
- Gough Map roads (c.1355–1360, up to a century later than the corpus, so it shows the road network near the period) — Oksanen, E. & Brookes, S. (2024) The Routes and Roads of the Gough Map: GIS Database. Archaeology Data Service. doi:10.5284/1124312.
- Navigable rivers (river courses with direct and indirect evidence of medieval traffic) — Oksanen, E. (2019) Inland Navigation in England and Wales before 1348: GIS Database. Archaeology Data Service. doi:10.5284/1057497.
- Bridges to c.1250 — Brookes, S., Rye, E. & Oksanen, E. (2019) Bridges of Medieval England to c.1250. Archaeology Data Service. doi:10.5284/1053676. © UCL Institute of Archaeology, CC BY 4.0.
- Monastic houses (dissolution-era, so values postdate the corpus; shown because abbeys were common lodging and dating places) — after the MonasteryQuest map, stainedglassattitudes.com.
See the reign corridors instead
The centralization curve
Share of dated instruments issued at Westminster, plotted for all rolls together and for the Patent and Close rolls separately. The shaded bands mark major wars and crises for orientation; their spans are standard textbook date ranges (some grouped or rounded), not computed from the corpus.
Is the jump real, or just the Patent rolls ending?
Good instinct — partly the latter. Close-roll business sat at Westminster far more often than Patent-roll business, and the Patent Rolls here load only through the early 1340s. So when Patent drops out of the mix, the combined “all rolls” line rises toward the Close-roll share even if neither roll’s own Westminster share changed. The dashed (Patent) and solid (Close) lines let you separate the two: a rise that shows up within the Close line is real centralization; a rise only in “all rolls” around the 1340s is mostly this coverage/mix shift.Numerator: entries dated at Westminster. Denominator: all dated, geocodable entries that year (≥20 required, so early years are sparse). “Westminster” is the palace/abbey complex; nearby London seats (the Tower, St Paul’s) are counted separately, so the true capital share is if anything understated.
Approximate straight-line km between monthly seats
What's in this number?
Straight-line distance between the busiest dating place of each consecutive month, summed per year. This is a lower bound on real travel (roads wound; the court doubled back). A year reads as "still" if government sat in one place all year, but also if its months lack dated, geocodable entries, so dips overstate stillness where coverage is thin.War on paper: protections and pardons
Why the collapse after the 1340s?
Letters of protection are a Patent Roll instrument, and in this corpus the Calendar of Patent Rolls has been ingested only through the early 1340s; the later Patent volumes aren't loaded yet. The Close Rolls continue to 1377, so the corpus doesn't end, but protections and pardons fall to near zero because their source roll stops. This drop is a coverage gap, not a real decline in wartime paperwork. Once the later Patent volumes are added, expect Crécy (1346) and the 1359–60 campaign to show clear peaks.Chancery output by roll
What's in this number?
Dated entries enrolled per year, split by roll. This tracks what has been ingested and survives in the calendars, not original chancery volume. Henry III's calendars are heavily abbreviated (so his bars are short); the Patent Rolls series stops in the 1340s while the Close Rolls run on, so the later period is Close-only.Where government sat, reign by reign
Percentages are of that reign's dated, located entries, not of all its business, since undated and unplaced entries are excluded. The pie aggregates the whole corpus.
Record months away from Westminster
Busiest single months at any seat other than Westminster, usually a parliament, a campaign base, or the court abroad. Counts are of dated, located instruments only.