Atlas layer
Treaty frontiers, 1829-1947
What each settlement moved, drawn as dated phases - and every line editorial, because not one instrument in the ledger supplies geometry this project could use.
- Canonical ID
dacia-treaty-frontiers- Period
- 1829–1947
- Role
- historical
- Category
- conflict campaigns frontiers
- Status
- published
Part of Corpus Chartarum Daciae
What you are looking at
Eight lines across seven phases, showing where the frontiers of the Romanian lands stood after each settlement between the Treaty of Adrianople in 1829 and the Paris Peace Treaty of 1947. Move the year slider and the map shows the arrangement in force at that moment.
Every line on this layer was drawn by this project. None of them is a digitised boundary.
How to read this layer
Three kinds of line, told apart without relying on colour:
| Line | Appearance | What it is |
|---|---|---|
| Treaty line | solid, heavy | A frontier an instrument actually fixed. |
| Reconstruction | dashed, medium | A later scholarly redrawing of where it ran. |
| Proposal | dotted, thin | A line somebody put forward that was not adopted. |
Weight and dash carry the distinction so it survives for readers who do not see the hues, and so a printed screenshot keeps its meaning.
Where two sources give different lines for one moment, both are kept. They are not averaged. An averaged frontier would be a line nobody proposed, presented with more confidence than either source earns.
The gap this layer exists to make visible
A treaty can settle a frontier completely and still give you nothing to draw.
The instruments in the ledger say things like “the Pruth shall form the boundary” or “the frontier shall follow the line laid down by the Commission” - legally sufficient, cartographically empty. The delimitation commissions that turned those words into ground truth produced protocols and maps this project has not obtained, and in several cases which may not survive.
So the evidence for where a frontier was and the geometry showing where it ran come from different places, and this layer keeps them apart. The ledger records what each instrument establishes. The line records what this project drew in order to show it. Nothing here should be read as a delimitation.
Historical scope
1829 to 1947, as seven dated phases. Each feature carries its own validity range, so the year slider reveals and hides individual lines rather than the whole layer: at 1860 you see the post-Adrianople arrangement, at 1920 the Trianon settlement, and the intervening proposals appear only in the years they were live.
Sources and evidence
The full instrument-by-instrument ledger is published as the treaty and frontier source ledger. For each instrument it records the date, the source and locator, what the evidence establishes, what it does not establish, the geometry provenance and the confidence and review state.
The short version: the instruments are cited from Hertslet’s The Map of Europe by Treaty and the standard treaty series. Confidence attaches to the fact of the settlement, which is generally high - these are published international instruments. It does not attach to the linework, which is editorial in every case.
Reconstruction and uncertainty
Three things are uncertain here and the layer is explicit about each.
The line’s position. Drawn from the instrument’s description and from published small-scale maps. In river sections it is defensible to within the river; across plains and mountains it is indicative.
Which line was in force. Some settlements were signed, modified and ratified across several years. The phase dates follow the instrument that fixed the arrangement, not the ratification, and the ledger records where those differ.
Whether a proposal mattered. A proposal is kept because a rejected frontier is evidence about what was thought possible at the time. Keeping it is not a claim that it was ever close to adoption.
Editorial decisions
Editorial lines are never presented as source geometry. Every feature
carries geometry_provenance: editorial_reconstruction, and the dash pattern
shows it on the map without the reader having to open a panel.
Competing lines are retained rather than resolved. Disclosure over tidiness: a map that shows one line where the sources give two has made a scholarly decision silently.
Modern boundaries are available as context and labelled as anachronism. Comparing a 1913 frontier with today’s border is a legitimate thing to want; the modern boundaries layer exists for it and says what it is.
Data fields
line_type, interpretation_status, confidence, source_id and
review_status are all filterable on the map. Their meanings are in
the Dacia data fields; the shared geometry
rule behind them is in
the shared Dacia GIS method.
Data and downloads
Served as GeoJSON under CC BY 4.0 from the content-addressed release, so a given URL always returns the same bytes. Attribution: Terra Chartarum; instruments after Hertslet and the KAN-351 source ledger; all linework editorial. That last clause must travel with any reuse.
Citation and reuse
- Version
-
dacia-treaty-frontiers-kan352 - Licence
- CC BY 4.0
- Attribution
- Terra Chartarum; instruments after Hertslet and the KAN-351 source ledger; all linework editorial
- Last reviewed
- 2026-08-23
Cite this dataset
Attribution must travel with any reuse. Where a layer carries a third-party licence, that licence governs the data regardless of anything on this page.
Advanced and technical
Schemas, source data, build history and raw assets. Nothing here is needed to understand what this page describes; it is here so the work can be checked and reused.