The Verification Venue · pointed at the most repeated wrong sentence in navigation
One Minute of Latitude Is Not One Length
Ask what a nautical mile is and you will be told: one minute of latitude. The sentence is false everywhere, and this page shows why without arguing. The Earth is not a sphere, so a minute of meridian arc has a different length at every latitude, shortest at the equator and longest at the poles. The modern mile declines to compete: it is a definition, exactly 1852 metres, planted inside the range the planet allows. Drag the slider and watch the true minute travel under the fixed line; then let a knotted rope and a sand glass count the same mile out of pure arithmetic.
Below is the true length of one minute of meridian arc at every latitude, computed from a named ellipsoid by the standard series. The curve moves only when you change the Earth; the straight line never moves at all, because nothing measured it. Everything printed is recomputed in front of you, and the check panel at the bottom re-derives all of it a second way.
Pick an Earth: every figure below recomputes
These are the defining parameters of five named reference ellipsoids. None of them is the actual sea-level Earth, which undulates away from every ellipsoid by tens of metres. The shape of the result survives that; the millimetre digits belong to the model. ↓
One minute of arc here
· m
versus the fixed mile: ·
verdict: ·
The fixed line crosses the curve
·
Poleward of these latitudes the true minute runs long; equatorward, it runs short. The line itself is 1852 m because a conference said so in 1929, not because anyone measured it there.
Drag toward the equator and the minute shrinks below the line; drag toward either pole and it swells above it. The line cannot follow you: it is a definition, not a measurement.
What changes along the curve is the meridian radius of curvature: the radius of the circle that best hugs the meridian at your latitude. On a sphere it is the same everywhere, and a minute of latitude is one fixed slice of one circumference. On an ellipsoid flattened at the poles, the meridian bends more gently as you approach the poles, so its local radius grows, and with it the length of a minute. The page evaluates it two ways: the closed form below, and the four-term Fourier series in the eccentricity that draws the curve (Snyder 1987, independently rederived for the verifier).
Read the crossing latitude off the panel above; it sits in the middle forties, printed to three decimals, computed by bisection rather than quoted. The crossing is defined on the centred-arc minute, the finite one-minute span the curve draws. The local-limit minute, the radius of curvature times the angle printed in the formula line above, is a different quantity: it meets the fixed line a few hundred-thousandths of a degree away, and the page prints both without presenting either as a check on the other. Poleward of the crossing, your minute out-measures the mile. Equatorward, it falls short. A navigator who steers by minutes and calls them miles drifts, in opposite directions, on either side of that parallel. None of this is a scandal of measurement. It is the gap between a sentence about the world, "a minute of latitude", and a sentence about a decision, "1852 metres, exactly, by agreement". The curve is the Earth talking. The line is a committee.
Layer two: the instrument that carried the unit, a chip log
Sailors measured a ship's speed by throwing a wooden chip attached to a line knotted at fixed intervals, and timing the run with a sand glass. The knots that paid out were the speed, and the word knot stuck to the unit. Notice what the device asserts: a spacing and a duration together encode one distance per hour, spacing × 3600 ÷ seconds. Set the two dimensions below and the implied unit falls out of school arithmetic. Change either one and you have declared a different mile. The instrument and the unit are the same statement made twice.
Your log declares
· m
·
Knot spacing that matches your latitude
· ft
·
Drag it and watch the implied unit slide along the band of minutes the Earth permits, drawn above.
A shorter glass needs tighter knots to count the same mile. Only the product matters; the sliders are two ways of holding it.
Three reported pairings
Pairings varied by service and by century; these three are commonly cited representatives of the scattered record, not a canon (what is and is not pinned down is stated in the research README). The arithmetic is indifferent to the choice: whatever pair you set, the product IS the unit your log counts.
Press it, then sweep your latitude from equator to pole and watch the required spacing. Here is the quiet result hiding in the arithmetic: the mile cannot be a minute of latitude, because the minute will not hold still, yet a single rope with a single spacing serves the whole ocean, because the minute's entire journey fits inside a fraction of a percent. The definition fails as geography and succeeds as engineering. The readout above prints the full required range, computed, so you do not have to take the sentence on faith.
The check · every number recomputed in front of you
Each row below is computed in this page from the defining parameters of a named ellipsoid, by the same series the curve uses. Its second-to-last column recomputes the meridian quadrant by Simpson integration of the meridian radius, sharing no code with the series beyond sine and square root: if the series is wrong, the columns disagree and the row turns red. The sphere row keeps WGS 84's radius and deletes its flattening, isolating what oblateness alone does; with no flattening the quadrant must equal πa/2 exactly, and the fixed mile must miss the minute everywhere. A row passes at two millimetres or better.
| earth | a (m) | 1/f | min @ 0° | min @ 90° | crossings | quadrant, series | quadrant, Simpson | |diff| mm | verdict |
|---|
For the reader's current selection (·), the same arithmetic with the numbers substituted:
What is exact: the definitions (1852 m, 0.3048 m, the ellipsoid parameters), the series' neglected term (bounded well under a millimetre at these eccentricities), the bisection (converged past 1e-12 degrees), and the Simpson quadrature (far finer than the printed digits). What is assumed: that the Earth is an ellipsoid of revolution, which the geoid politely declines to be. Run the independent re-derivation yourself: node research/one-minute-of-latitude/verify-one-minute-of-latitude.mjs
What's idealised here, and what's exactly true
Exactly true. The international nautical mile is exactly 1852 metres, agreed at the 1929 International Hydrographic Conference in Monaco; it is a definition and the page treats it as one. The international foot is exactly 0.3048 metres. The ellipsoid parameters are the defining constants of their named systems. The series used for the curve is the exact antiderivative of the exact meridian-radius integrand, truncated at the sixth power of the eccentricity; the neglected next term is bounded below a tenth of a millimetre for every eccentricity used here, and the Simpson column in the check panel verifies the bound empirically, row by row.
Idealised. The Earth is modelled as an ellipsoid of revolution. The geoid, the actual sea-level surface, departs from any such ellipsoid by up to roughly a hundred metres in elevation, so a minute measured along the real ocean's level surface differs slightly from the one printed here. Rotation, gravity anomalies and terrain are absent. Minutes of longitude are a different quantity entirely and appear nowhere on this page. The chip-log arithmetic assumes the line pays out straight and the glass runs honestly.
Representative, not universal. After 1852 was agreed internationally, the United Kingdom went on counting a mile of exactly 6080 feet and the United States a mile of exactly 6080.20 feet for decades; both come to about 1853.2 metres at today's international foot, and the feet in use then differed from today's in the seventh digit, so treat those conversions as good to a few parts per million. The knot-and-glass pairings offered above are representatives from a scattered record. What survives every scatter: the ordering (the minute is shortest at the equator, longest at the poles), the crossings in the middle forties, and a total pole-to-pole spread of about one part in a hundred.
Deliberately absent. Histories of how the metre and the mile were fixed tell of survey expeditions whose popular retellings have accumulated folklore of their own. This page recounts none of that saga and computes none of its numbers. The geodesy above stands without it.