The evidence files
Ten files, indexed by subject. Each one takes a single observation or a single body of source material and sets out what is seen, what the flat-earth reading makes of it, and what the mainstream reading makes of the same thing.
This is the index to the rest of the archive. Nine of the ten files below work the same way: they state what is observed, what the flat-earth reading of that observation is, and how mainstream geodesy or astronomy accounts for it. Where a number is quoted, the file says where the number came from rather than asking you to take it on trust.
The tenth is different in kind. Objections, answered states the strongest case against the level-earth reading before answering it. If you only read one file on this site, read that one: a reading that cannot survive its best objection is not worth the rest of the archive.
Read in this order. First What the horizon does — it is the observation most people can check the same afternoon. Then Water finds its level, because every other argument leans on it. Then Try it yourself so that you have one test of your own rather than ours. Then The model, to see how the pieces are supposed to fit together. Finish with Objections, answered and the glossary — the files use terms such as dip, refraction and azimuthal equidistant precisely, and the glossary is shorter than the arguments. If you would rather measure than read, the downloads page carries a printable field log sheet, a worked example log and two plain-text briefs.
For the whole archive in one place — the reading order, a line on each page, and the files the site is built from — see the site map.
The files
What the horizon does
Why a ship disappears hull-first, what refraction does to a sightline, and what a level line over 4.5 km of still water actually shows.
ObservationWater finds its level
Canals, aqueducts and long pipelines are surveyed from a level reference, and still water is the reference everyone agrees on.
ObservationSun and moon overhead
Apparent size through the day, the seasons, the analemma, and the reading in which both bodies circle above a level plane.
ObservationTwo skies, one ceiling
Northern and southern star rotations, the missing measurable parallax, and the pole problem both readings have to face.
Source workThe firmament in the old texts
What the Hebrew, Greek and early English cosmologies describe above the earth, with the words and the translations shown side by side.
Source workMaps drawn level
Azimuthal equidistant projection, the Gleason map, and what the route-map and flight-path arguments can and cannot show.
The modelThe model in full
Plane, dome, circling sun and moon, a southern ring of ice — the whole picture assembled in one place, with its open questions named.
Do it yourselfTests you can repeat
Six tests with ordinary tools: spirit level, laser, long lens, aircraft, two observers and one stopwatch.
Steel-manObjections, answered
The strongest arguments against the level-earth reading, stated fairly and at full strength, and how far each one actually gets.
ReferenceGlossary
Firmament, azimuthal equidistant, dip of horizon, parallax, refraction, spheroid — the working vocabulary, defined plainly.
How each file is put together
The evidence files keep to the same order so that you can compare them directly, and so that you can tell a measured claim from an argued one at a glance. Where a file rests on a historical or textual claim rather than a physical observation, it says so in the first paragraph and names the edition or translation it is working from.
- Observation — what a person sees or measures, with the conditions and the equipment recorded.
- The reading — the flat-earth account, stated as a claim that could turn out to be wrong.
- The mainstream reading — the geodetic or astronomical account, given as its own sources state it, not as a caricature.
- What would settle it — the sightline, measurement or dataset that would decide between the two readings.
Two conventions are worth knowing before you read anything else. First, a quotation is a quotation and a paraphrase is labelled as one; if the wording of a source could not be checked, the file paraphrases and says so. Second, a number without a source is not used. Where the two readings both survive a test, the file says that too, rather than picking a winner for rhetorical effect.
If you think a file has got a fact wrong, or you know a source we have missed in either direction, the contact page says what to send and how corrections are handled.
This page presents the flat-earth interpretation as its subject matter. It is a belief-based reading of the observations, not a scientific finding — see About for how we frame claims and for the mainstream sources.