Is the Earth flat? The shape of the Earth, in plain sight.

• 1How he did it (~240 BCE). At noon on the solstice, the Sun stood straight over Syene, so a deep well there cast no shadow. At the same moment in Alexandria, about 800 km north, a vertical stick cast a 7.2° shadow. That angle is 1/50 of a full circle, so the whole way around is about 50 times the 800 km gap, or roughly 250,000 stadia. Depending on the length of a stadion, that comes out near 39,000 to 46,000 km. The real number is 40,075 km. He got it right the first time, with sticks and shadows.
• 2The test that settles it. Flat-earthers say a small, nearby Sun could make those shadows too. So add more measuring stations at other latitudes. The near-Sun idea then needs a different Sun height for every pair of cities, and those cannot all be true at once. A far-off Sun over a round Earth fits every station with one answer. The Sun also keeps the same angular size, about 0.5°, all day and everywhere. A Sun moving closer and farther overhead could not do that.
• 3The horizon dips. The higher you stand, the more the horizon sits below level. The angle is about √(2h/R), which works out to roughly 1.6° from a 1,000 m peak. You can measure it with a surveyor's scope, and it grows with height. On a flat, endless plane the horizon would stay right at eye level. It doesn't.
• 4The horizon backs away as you climb. Its distance grows with the square root of your height, about 3.57·√h km. That is a real, height-based edge, which is what a ball gives you, not the endless view of a flat plane. Surveyors, gunners, and microwave engineers all plan around the drop, roughly 8 inches times the miles squared.
• 5“It looks flat” is just what a big ball looks like up close. The curve drops by about d²/2R over a line of sight d. On a 6,371 km radius that is only about 2 m over 5 km. A ball that huge looks flat to the naked eye, so you find the curve with tools, not by staring. Those tools measure things like horizon dip, hidden height, and long sightlines. We also know the size well now. Satellite measurements (the World Geodetic System of 1984, the WGS84 standard) put the equatorial radius at 6,378,137 m, a circumference near 40,075 km. That matches Eratosthenes to within a few percent.

“Flat” is a measured quantity, and Earth fails the test. In a lab, a surface earns the word flat only after a laser interferometer counts the fringes. Lay a reference optical flat on a surface under monochromatic light and each dark fringe marks half a wavelength of gap, about 316 nanometers (nm) for a red He-Ne laser. A good reference flat is certified to a fraction of one fringe, roughly λ/20, near 32 nm. The table takes Earth’s own curvature drop (d²/2R) and writes it in those same fringes.

Deviation from flat grows with the square of distance. At tabletop scale Earth is flat to interferometric standards, the curve sits below a thousandth of a fringe. By 1 m a reference flat could resolve it, and over a few km it is millions of fringes. The same laser interferometry that certifies a surface flat, the Fizeau interferometer, is the technology a ring-laser gyro uses to clock Earth’s spin [475].

Sources: Eratosthenes' measurement [43] · horizon geometry & refraction [8] · circumference [4] · modern circumference (WGS84) [443]. → Curvature data rows

• 1The original mistake. In 1838 Rowbotham sighted along the Old Bedford River just above the surface, saw a distant marker, and called the water flat. There were two problems. Light skimming close to the water passes through the strongest temperature layers, which bend it down to follow the curve. And one line of sight on its own cannot tell "flat" apart from "curved plus bent light."
• 2Wallace's fix (1870). Alfred Russel Wallace set three markers at the same height over six miles. On a flat surface they would all line up. Over a curve the middle one rides higher than a line joining the two ends. It did, by about the amount expected, and Wallace won the £500 "Bedford wager." John Hampden never accepted the result and harassed Wallace for years, which shows how hard belief can hold on against evidence.
• 3Why you need three points. The raised middle marker is the mark of a bulge, and it still shows up even when the bending is steady across the whole line. That is the same idea behind the curve calculator in Entry 3. Modern surveys, using good targets and a measured bending value, get the same result.
• 4It is really a lesson about light. Skim the water and bending can fake a flat result. Raise the sightline, use several targets, and the curve comes back. Rowbotham did not disprove the curve. He ran into refraction without knowing it. (See the Optics section.)
• 5A neutral judge called it for the curve, and 1901 confirmed it. The 1870 bet had referees. The neutral one, John Henry Walsh, editor of The Field, ruled that Wallace’s three-marker sighting showed the curve, and gave him the bet. Hampden fought it for years in pamphlets and courts. In Hampden v Walsh (1877) the court ruled the bet was a wager, void and unenforceable under the Gaming Act 1845, so Wallace, even though he had won the experiment, was ordered to hand the £500 back [474]. Hampden was separately jailed for libel and for threatening to kill Wallace. But the measurement held. In 1901 Henry Yule Oldham, a geography lecturer at Cambridge, ran the test again with three equal-height poles and found the bulge once more. The fight was always about the bet, never the curve.
• 6A modern rerun: Thompson v. Garcia (2019). The same fight reached a Georgia court. Flat-earther Zen Garcia offered a cash reward (a $5,000 contest; the suit later sought $15,000) to anyone who could show Earth’s curve by real-world experiment. William Thompson tried, was refused, and sued for breach of the contest agreement. He lost, because Garcia’s rules required a result matching the “8 inches per mile squared” figure, the same naive formula this entry corrects, the one that ignores eye height and refraction. No sound observation can match that number, so the contest could not be won by design [472]. The court ruled only on the contest terms. It did not find Earth flat, or that curvature cannot be shown, though the case is often miscited that way [473]. Two centuries after the Bedford wager, a court again settled the contract, never the shape.

Sources: Bedford Level — Rowbotham & Wallace [44] · refraction [8]; the 1870 wager verdict & 1901 replication [250]; Hampden v Walsh (1877) [474]; Thompson v. Garcia (2019) [472][473]. → Curvature data rows · → curvature & horizon tools

• 1Geometry: the hidden base. Past the horizon (about 4.7 km for a 1.7 m eye, farther if you go higher), the bottom of a target is blocked by the bulge of water or land. A zoom makes the visible part sharper, but it cannot raise a base that sits behind the curve. That is the "sinking ship," or missing-foundation, look.
• 2Refraction: looming. Temperature layers, especially over water, bend light downward and can lift a hidden object partly back into view. This is called a superior mirage, and it explains many "impossible" shots. It is an effect of the air over a curved Earth, not a sign of a flat one. (See the Optics section.)
• 3Resolution: what a zoom really does. A longer lens shows finer detail. It does not change the line of sight. "Zoom in and it comes back" only brings back detail that was still above the horizon. It never brings back a base that is below it.

The record — 443 km, and why it was perfectly possible

The longest confirmed line-of-sight photo (a Guinness World Record) is Marc Bret's 2016 picture of Pic Gaspard in the French Alps (3,883 m), taken from Pic de Finestrelles in the Spanish Pyrenees (2,826 m). The two peaks are 443 km apart. People often call it proof of no curve. It is the opposite. The shot works only because of the curve, and it shows the curve.

• ①Height buys you horizon. Horizon distance grows with the square root of height: d ≈ 3.86·√h km in normal air. A 2,826 m peak sees about 205 km to its horizon. A 3,883 m peak sees about 240 km. Add those and the horizons meet at about 445 km, almost exactly the 443 km gap. The two summits are just barely able to see each other over the bulge.
• ②The photo shows the curve. You do not see the whole Alps, only a thin row of summits poking over the horizon. Do the math and about the lower 2,800 m of the Écrins massif is hidden behind the Earth's curve. Only the top few tens of meters clear it. That hidden base is the curve, caught right in the frame.
• ③Right at the edge. The summits sit within tens of meters of the cutoff, so the shot needed the unusually clear, strongly bending air reported that dawn (a polarizing filter cut the haze). That is why it is a record and not an everyday snapshot. On a flat Earth there would be no horizon to clear and no hidden base, and we would photograph far-off low things all the time.

Interactive — how much is hidden by the curve?

Enter an observer height, a target height, and a distance. The tool gives the horizon distance, how much of the target the bulge hides, and how much clears it. Slide the refraction value k (0 = vacuum, 0.13 = normal air, about 0.2 = strong bending) to see how much the air changes a long shot.

Infrared photography

Longer wavelengths scatter less. Rayleigh scattering drops as 1/λ⁴, so near-infrared (about 850 nanometers (nm)) scatters roughly 6 to 16 times less than blue light. That is why landscape and long-distance photographers use red or near-IR filters. They cut through haze and bring back sharp far-off detail that visible light loses to scatter. But infrared helps with haze, not geometry. The curve sets the horizon, not the color of the light. Infrared makes an object that is already above the horizon look clearer. It cannot show a base hidden behind the bulge. Long IR shots show the same sinking-ship, hidden-foundation pattern as normal light, just with less haze. If anything, infrared bends a little less than visible light, so it lifts hidden objects slightly less, not more.

Still to come for this entry: observer-height tables and worked examples (the Toronto skyline, lighthouse ranges) added to the calculator above.

Sources: horizon geometry & refraction [8] · longest-line-of-sight record [41] · atmospheric scattering & infrared [42]. → Photography rows (Optics)

• 1The dip has a formula. The horizon sits below true horizontal by θ = arccos(R/(R+h)), which is close to √(2h/R). That is about 0.045° at standing eye height (2 m), about 1° from a 1,000 m hill, and about 3° (2.98° with refraction) from a jet at 33,000 ft. A flat Earth says 0° at every height.
• 2You can measure it yourself. Sailors have used a “dip of the horizon” correction in star navigation for centuries. Even at the shore it is about 2 to 3 arcminutes, which the naked eye can pick out. From a plane window, apps like Theodolite or Dioptra show the horizon sitting clearly below eye level. You don't need special gear.
• 3“Rises to eye level” is a trick of having nothing to compare it to. With no instrument, the brain treats the horizon as level, because the dip is small and there is no reference line beside it. Add a true horizontal line and the flat answer (always 0° dip) fails. The globe's answer matches the numbers.
• 4Visual, optical, radio, radar: all the same curve. There isn't one horizon but a family of them. Each is the distance at which the surface curves out from under the line of sight. Light bends a little in air, so the optical horizon reaches about 7 to 8% past the bare geometric one. Radio and radar waves bend more, which engineers handle by pretending Earth's radius is 4/3 its real size. That pushes the radio horizon to about 4.12√h km, against the visual 3.57√h. Each one grows only with the square root of your height and backs away as you climb. That is the mark of a ball. A flat Earth would have no horizon at all.
• 5Below every horizon is a hidden zone. Because the surface curves away, everything past the horizon sits in a zone the curve hides. A coastal radar cannot see a low ship or a sea-skimming missile beyond its radar horizon. That is why low-flying aircraft slip under coverage, and why over-the-horizon radar has to bounce signals off the ionosphere or hug the surface. Once in a while a temperature inversion ducts the waves, bending them further around the curve, so a far-off station shows up well past its normal range. That only makes sense if there is a curve to bend around. The hidden zone and the ducting both need a round Earth.

Sources: horizon-dip geometry & the navigation correction [96] · the visual, optical, radio & radar horizons [206]. Builds on the curve of Entry 1 and the long-range sightings of Entry 3. → Curvature data rows · → curvature & horizon tools.

• 1“8 in × mi²” is the wrong quantity. It is a close approximation of the surface’s drop below a level line from your eye (the exact drop is R(sec − 1)). It stays accurate to a fraction of a percent out to a few hundred km. But it measures a tangent drop, not what is hidden, and a tall, far object can still poke above the bulge. This calculator skips the shortcut and uses exact sphere geometry: a target an angle β past your horizon is hidden by R′(sec β − 1).
• 2Eye height moves the horizon a lot. With your eyes at 2 m, the horizon is about 5 km away. From a 100 m hill it is about 36 km. Every meter of height lets you see farther and hides less of what lies beyond. That is why “but I can still see it” rarely matches the simple drop figure.
• 3Refraction bends the ruler. Air thins with height, so light curves gently downward. That flattens the Earth a little for a line of sight, and the standard fix uses an effective radius about 7/6 of the real one. The toggle shows how much that alone changes the result. This is not special pleading. It is the same geometry surveyors and sailors have used for centuries (Entry 4, 154).
• 4Put numbers on it: the simple figure nearly doubles the truth. Take the calculator’s default: a 50 m target 16 km away, eyes at 2 m. The “8 inches per mile²” drop over about 9.9 miles is about 66 feet (about 20 m), which flat-Earthers present as “hidden.” But 2 m of eye height puts your horizon about 5 km out, so only the 11 km beyond it does any hiding. The honest geometric figure is about 9 m, and with standard refraction about 8 m. So about 40 m of that 50 m target still stands above the bulge, which is just what you see. The simple rule overstates the hiding more than twofold because it forgets the horizon.
• 5Refraction changes, which is why the “anomalies” cut both ways. The 7/6 figure is an average. Over cold water under warm air the bending gets stronger (looming, a superior mirage) and can lift a hidden object fully into view. That is the source of most “I saw impossibly far” clips, and of the cherry-picked frames where the Pontchartrain towers briefly look straight (154, 150). A real curve plus real, changing air explains both the usual hiding and the rare over-the-horizon catch. A flat plane explains neither, since on a plane nothing is ever hidden in the first place.
• 6A laser “seen at the same height” across a lake shows the curve, not a flat plane. The test runs like this. A laser at one end of a lake 16 miles wide is caught at the far end near the same height, and the caller calls it no curvature. Work the geometry. At 6 feet of height on each end, the far shore hides 120 feet below your line of sight. A 6-foot laser sits 114 feet down inside that bulge. Drop to 3 feet and it grows worse, not better: 136 feet hidden, the beam buried 133 feet. A lower eye has a nearer horizon, so more of the water lies beyond it and the bulge piles higher. On a globe that laser should be out of sight. So if it arrives, light bent to reach it. Over 16 miles the air had to curve the beam near 87 percent of the Earth’s own curvature at 6 feet, and 93 percent at 3 feet. That is an atmospheric duct. It forms over cold, calm water with warmer air above, the same effect that lifts distant ships over the horizon. Now the tell. A flat plane would show that laser every day, in any weather, because nothing stands in its path. A globe shows it only when the duct forms. Run the test at midday in a light wind, and the beam is gone. Visibility that changes with the weather is the curvature, not its absence. The one way to see across by plain geometry is to climb, not drop: near 45 feet at each end lets the two horizons meet with no refraction. Fix the geometry instead, as the evenly spaced Pontchartrain towers above do, and the drop is right there. 2 6
• 7Take the air out and the doubt goes with it: LIGO is built around the curve. The lake test above cannot rule out refraction, because it runs through open air. The clean version removes the air, and it already exists. LIGO is a gravitational-wave observatory whose two arms are each a 4-kilometer laser beam inside a straight steel tube held at near-perfect vacuum. With no air, there is no refraction left to argue about. The beam is dead straight, but the ground is not. Over 4 kilometers the Earth curves away from that straight line by about 1.25 meters. This was not a number the builders assumed. It was a construction problem they had to solve. Pour the slab level, the way you would for a highway, and the beam would leave the corner station and strike the far mirror about a meter too high. So they surveyed the whole path against the curve of the Earth and set the far ends higher off the ground than the corner. Caltech states it in plain words on its own page: over each arm “the Earth curves away by nearly a meter.” Two of these observatories run today, in Washington and Louisiana. A straight laser, in a vacuum, that had to be aimed for a round Earth, or it would miss. [677] 197

Sources: horizon & hidden-height geometry [96]. Try the matching cases in Entry 2 (Bedford) and 154 (the Causeway & salt flats). → Curvature data rows · → curvature & horizon tools.

• 1The threshold (Lynch, 2008). Daytime tests put the lowest altitude for seeing the horizon curve at or just below 35,000 ft (about 10.7 km), and only with a wide field of view in clean air. A jet cruises near 11 km, right at that edge. So “it looks flat from my window” is just what a globe would give, not a sign of a plane.
• 2Red Bull Stratos. On 14 Oct 2012 Felix Baumgartner rode a balloon to 38,969.4 m (127,852 ft, confirmed by the FAI) and stepped off. At that height the horizon dips about 6° below eye level and lies about 700 km away. The curve is plain in the footage.
• 3The fisheye dodge fails. Flat-Earth videos blame “GoPro barrel distortion.” But the exact camera and lens were identified and the footage was de-fished, and the curve stayed. Here is a clean test: a real horizon curve stays curved when you rotate the camera, while lens distortion turns with the frame. Put the horizon through the center of the lens and the distortion drops out.
• 4Amateurs get the same result. Hobby high-altitude balloons and rockets often reach 25 to 40 km with corrected (rectilinear) lenses and record the same curve. No space agency needed. The dip you measure climbs with altitude just the way a ball would (Entry 4, Entry 5).

Sources: Red Bull Stratos / FAI exit altitude [135] · curvature-visibility threshold, Lynch 2008 [136] · refraction & dip [8]. → Curvature data rows · → curvature & horizon tools

• 1Zoom magnifies, it doesn’t move the horizon. Magnification enlarges the image. It adds no new line of sight. Light from a hull hidden behind the bulge never reaches the lens, so zooming in only enlarges the water and sky in front of it. When more of a distant object “returns,” the cause is refraction or looming bending the rays, the same air-bending measured in Entry 4, not a missing curve (Entry 3, Entry 6).
• 2What measures the curve. A theodolite reads horizontal and vertical angles to a fraction of an arc-second on a leveled, calibrated circle. A total station adds laser ranging. Aimed at the sea horizon from a known height, it reads the dip directly, and that dip grows with altitude just as a ball requires (Entry 4). The Bedford Level, redone with a leveled instrument instead of the naked eye, shows the midpoint bulging up (Entry 2).
• 3The sextant proves the globe through latitude. A marine sextant reads the angle between a star or the Sun and the horizon to about a tenth of an arc-minute. The measured height of Polaris is your latitude, and the noon Sun gives it too. Those angles change as you travel north or south at a steady 1° per 111 km, and they only fit one consistent global grid on a ball (Entry 5). A flat plane cannot reproduce them.
• 4Past 125×, the pixels are invented. The P1000’s real optics stop at 125× (24–3000 mm). Its “250× Dynamic Fine Zoom” (6000 mm) and full digital zoom (up to 500× / 12000 mm) are digital. The processor guesses pixels between the real ones and adds no true optical detail, which is why reviewers warn against it. A tiny 1/2.3″ sensor and clear barrel distortion settle it: this is a consumer camera, not a measuring tool.
• 5Surveyors don’t ignore the curve, they subtract it. Open any leveling manual and you find a standard “curvature and refraction” correction for long sights. The curve drops the far target by about 0.0785 D² meters, refraction lifts it back by about 0.0112 D², for a net correction near 0.0673 D² meters, with the sight distance D in kilometers. That is about 7 cm at 1 km, and it grows with the square of distance. So careful leveling either subtracts it from every reading or cancels it by balancing fore- and back-sights. Engineers building railways, canals and pipelines treat Earth’s curve as a printed line-item, not a debating point. It is in every surveying textbook and every total station.

Sources: Nikon P1000 optics — 125× optical, digital “Dynamic Fine Zoom” [145] · digital zoom adds no optical detail [146] · theodolite angular precision [147] · sextant accuracy & latitude [148] · horizon dip & refraction [8] · surveying curvature & refraction correction [262]. → Optics & navigation data rows

• 1Where 69 comes from. The nautical mile was set as one arc-minute of latitude, so 60 nautical miles = 1° ≈ 69 statute miles (111.3 km). That is the same as 40,075 km ÷ 360. It is an arc length over a center angle, nothing more.
• 2Three different things. The center angle (1° per 69 mi), the “drop” below a tangent line (the famous 8-inches-per-mile² figure, Entry 5), and the horizon dip you see (Entry 4) are three separate things. The whole error is treating the center angle as a visible tilt or a per-mile drop.
• 3Why pilots don’t “pitch down.” “Level” means at a right angle to local gravity, which turns with the surface. An autopilot holding altitude follows the curve on its own. The turn is only 1° per 69 miles, about 0.002° per second even at jet speed, far too slow to feel (57).
• 4The real giveaway is longitude. Degrees of latitude stay about 69 mi apart, which fits a ball. But degrees of longitude shrink with the cosine of latitude, down to about 34.6 mi at 60° and zero at the poles, where every meridian meets. An endless flat plane, or a flat-Earth map disc, cannot pull the meridians together to a point.
• 5A degree of latitude isn’t even constant, and that pins down the shape. Earth is an oblate spheroid, slightly flattened at the poles, so a degree of latitude is not 69 miles everywhere. It runs from about 110.6 km (68.7 mi) near the equator to about 111.7 km (69.4 mi) near the poles, where the surface curves more gently. That 1 km difference was the whole point of the 18th-century French Geodesic Missions, which measured a degree in Lapland (Maupertuis, 1736) and at the equator in Peru (Bouguer and La Condamine, from 1735). The polar degree came out longer, which proved Earth is flattened at the poles, as Newton predicted. Not egg-shaped, and certainly not flat. The very number flat-Earth maps borrow is a measured feature of a spinning, slightly squashed ball.

Sources: length of a degree of latitude & longitude [137] · the nautical mile (one arc-minute) [138] · circumference [4] · the degree-of-latitude expeditions [261]. → Curvature data rows

• 1What he did (~240 BC). Eratosthenes knew that at noon on the summer solstice the Sun stood straight over Syene (modern Aswan). A vertical stick there cast no shadow, and sunlight reached the bottom of a well. At the same moment in Alexandria, due north, a vertical gnomon cast a shadow of about 7.2°, which is one-fiftieth of a full circle. Two sticks, one moment, one angle.
• 2The arithmetic. If the Sun is far enough away that its rays arrive parallel, that 7.2° is the angle at Earth’s center between the two cities. So the arc between them is 1/50 of the whole way around. The cities were about 5,000 stadia apart, giving 50 × 5,000 = 250,000 stadia. Depending on which stadion he meant (about 157–185 m), that is roughly 39,000–46,000 km. The true polar circumference is 40,008 km, so he landed within a few percent to about fifteen, from sticks and shadows.
• 3The flat-Earth reading. Flat-Earthers accept the shadows but reject the far Sun. A small Sun a few thousand kilometers up, throwing spreading rays over a plane, would make the same 7.2° difference, and you can work a Sun height back out of the two numbers. With only two data points, that idea cannot be tested. So you take more.
• 4Three sticks break the flat model. Measure the noon shadow at three or more latitudes along a line of longitude. On a globe with a far Sun, the angle grows in step with distance, a straight line. Under a near Sun at a fixed height it grows as arctan(distance ÷ height), a clearly curved relation that steepens as you near the point under the Sun. The measured relation is dead straight. The same data that “fit” a close Sun for one pair refuse to fit it for three.
• 5One Earth, one answer, versus many Suns. Every pair of cities, anywhere, on any clear day, gives the same circumference, about 40,000 km. The near-Sun model has to invent a different Sun height, a different Sun, for each pair and each season, and it never settles on one sky. A close Sun can also stand straight over only one point. Yet on the solstice the noon Sun is overhead all along the Tropic of Cancer at once, a whole circle of latitude. A far Sun lighting a ball explains that. A local Sun over a disc cannot (92).
• 6You can still do it. Schools across dozens of latitudes repeat Eratosthenes each equinox, phone in their shadow angles, and recover about 40,000 km together. You need only a stick, a sunny noon, and a friend a few hundred kilometers north or south (175). Cross south of the equator and the shadows lean the other way, toward the north, just as a ball requires (170, 171).
• 7The crowd has already redone it. The flat-Earth debunker SciManDan had viewers around the world measure a vertical stick’s shadow at their own local noon on the same solstice, then pooled the results. Plotted against latitude, the shadow ratios fell along the curve a globe predicts, not the straight line a flat Earth would give, and three readings along one meridian reproduced Eratosthenes’ circumference to within about four percent. No ancient authority, no space photos, and no single traveler are needed, just hundreds of ordinary people with sticks agreeing on a round Earth. You can add your own reading, or run the open school-pairing data at eratosthenes.eu. [547]

The full walkthrough. What this measurement proves, what it cannot prove on its own, the flat-Earth escape routes and how each one closes, and the modern crowds who repeat it across many latitudes and always land on a globe. Read the Eratosthenes page →

Sources: Eratosthenes’ measurement [299] · the method, assumptions & modern reproductions [300]. See also Entry 8, 170, 175. → Curvature data rows.

• 1Two versions, two dead ends. The simple concave Earth makes predictions that are plainly wrong. The fancy one makes no testable prediction at all. Neither can win.
• 2The horizon decides it. On the inside of a sphere the horizon would climb toward eye level and distant ships would tower upward. In reality the horizon sits below eye level and ships sink hull-first. That is the convex signature, measured at every shoreline (see Entry 4).
• 3The rectilineator was surveyor’s error. Teed’s 1897 Florida-coast survey claimed to track the ground curving up over four miles. It was really built-up mechanical mis-alignment in a hand-leapfrogged device, not a measurement of concavity.
• 4If it can’t be told apart, it’s empty. The bent-light version is a mathematical inversion of the globe through a sphere. Coxeter noted it copies every observation exactly, so it predicts nothing new. It is the globe in disguise, not a real alternative.
• 5It survives only by bending light. Straight-line light is confirmed on its own, by laser levels, geodetic surveying, and line-of-sight microwave links. Take away the made-up curved light and the concave model falls straight back into the convex globe.
• 6The history. Cyrus Teed renamed himself “Koresh” after an 1869 electric-shock vision and founded the Koreshan Unity. Their commune at Estero, Florida is now a state historic site. The idea drew a few hundred followers and died with its prophet in 1908.

Sources: Teed’s Cellular Cosmogony, the Koreshan Unity and the rectilineator survey [418]; the inversion-geometry argument that a bent-light concave model is observationally identical to the globe [419].

• 1“Level” is not “flat.” Still water settles at a right angle to local gravity, an equipotential surface. On a ball that surface is curved (the geoid). Water “finding its level” is just what a curved ocean does.
• 2The meniscus is tiny. Surface tension shapes water only within a capillary length, about 2.7 mm. A meniscus is liquid climbing a container wall. Scale the container up to a pond and gravity pulls the surface down to the local equipotential. It never bows up across open water.
• 3Sea level really does curve. Satellite altimetry shows the geoid varying from about −106 to +85 m off a smooth ellipsoid, following gravity (where mass sits), while staying curved overall. The drop is about 8 in × (miles)².
• 4Water is never truly “flat.” Its shape is set by whichever force wins. At small scales surface tension takes over and pulls water into spheres: raindrops, beads of dew, the floating blobs astronauts play with in orbit. At planet scale gravity wins and the surface settles onto a curved equipotential. Twice a day the Moon lifts the whole ocean into two tidal bulges (Entry 42). A calm lake only looks flat because the curve is gentle across a few miles, not because water can’t hold a curve.
• 5Why the deep ocean is cold. The seafloor sits on top of ~2,900 km of mantle and crust, and rock is a poor conductor (~2–3 W/m·K). Heat from inside the Earth (~0.087 W/m²) is roughly 3,900× smaller than the average from the Sun (~340 W/m²). The ocean is warmed from above, and cold, dense water sinks at the poles and fills the deep at 0–4 °C. This says nothing about Earth's shape, but it is a common "gotcha," so it is answered here.
• 6Water doesn’t need walls, it needs a force. “Water has to have a container” misses the point: water takes its shape from the forces on it, not from a box. A falling raindrop is a self-contained sphere held by surface tension. Water released on the Space Station balls up into a floating globe. Comets and icy moons hold their water in the vacuum of space by their own gravity. The ocean is the same story at planet scale: a thin shell pulled onto a sphere by gravity, with no wall at any “edge,” because “down” points to the center everywhere, so there is no edge to pour over.
• 7Water without a container, everywhere you look. A dewdrop on a leaf, a falling raindrop, and a floating water blob released by an astronaut on the ISS all hold a rounded shape with nothing touching them but surface tension. A comet hauls a mountain of water ice through the vacuum with no vessel at all. The ocean is the same idea scaled up: water held to a spinning sphere by gravity, settling onto the curved geoid. Not one of them needs a wall. Water takes its shape from whichever force wins, surface tension or gravity, never from a container.

Sources: geoid / sea level [28] · surface tension & capillary length [29] · ocean circulation & deep temperature [30] · geothermal vs solar flux [31] [417]. → Water data rows

• 1Two meanings of one word. “Flat” can mean planar (no curve) or smooth and level (no bumps, no slope). Flat-Earth arguments quietly swap the second meaning for the first. Water being “level” is a statement about gravity, not geometry.
• 2What a level measures. A bubble level and a plumb bob both point along local gravity, which is “down.” Over a long distance, “down” here and “down” 100 km away are not parallel. They fan toward Earth’s center by about 0.9° per 100 km. The tool re-levels at each spot. It never proves a plane.
• 3Still water is the geoid. Calm water settles onto a surface of constant gravitational potential, the geoid. That surface is gently lumpy but a sphere overall. “Sea level” is curved by its very definition.
• 4It is testable. The Bedford Level done right (Entry 2), bridge towers that lean apart at the top (Entry 16), and the abyssal plains (Entry 14) all show “level” following a curve, not a plane.
• 5A mountain bends ‘level’ sideways. If ‘level’ were a fixed flat plane, no nearby object could tilt it. Yet in 1774 the Astronomer Royal Nevil Maskelyne, on the Scottish mountain Schiehallion, measured a plumb line being pulled toward the mountain’s mass, about 11.6 arc-seconds. He found it by comparing the apparent vertical against the stars on the north and south sides. ‘Down’ points to the local center of gravity, not to a single flat plane, and it leans toward big masses. The same experiment gave the first good figure for the density of the whole Earth.

Sources: the geoid & “level” as an equipotential [139] · refraction & sightlines [8] · WGS 84 figure [4]; the Schiehallion plumb-line deflection [248]. → Curvature data rows

• 1“Level” follows the geoid, not a plane. A free water surface sits at a right angle to local gravity. That equipotential surface, the geoid, wraps around the Earth, so “dead level” water is gently curved, not flat (Entry 12). Nothing drains off because there is no edge to run toward. Down points to the center at every point.
• 2Suez: no locks, and none needed. The Suez Canal cuts about 193 km of flat desert between the Mediterranean and the Red Sea, two seas at essentially the same level, so it is a single sea-level channel with no locks. A waterway that long lying “flat” is just what a curved geoid at constant elevation looks like. The water is level, which means curved.
• 3Panama: locks because the land rises. The Panama Canal is shorter (about 82 km) yet needs three flights of locks, because its route crosses high ground. Ships are lifted 26 m (85 ft) to Gatún Lake, carried across, then lowered back to the ocean. The locks track the land’s height above sea level, a continental divide, not any curvature of the water.
• 4The calm-canal test finds the curve. Flat-earth lore loves long, still waterways, so use one. Sight a leveled instrument down a six-mile drainage canal and the far marks ride below the line while the midpoint bulges up about 1.5 m, the mark of a 6,371 km ball (Entry 2). The very stillness that is supposed to prove flatness is what lets the curve show (Entry 5).
• 5A canal climbs to a summit, and has to be fed from the top. Where the land rises, the canal has to go over it. The Canal du Midi (opened 1681) climbs 189 m from Toulouse to the Seuil de Naurouze, a high point where water drains both ways, west to the Atlantic and east to the Mediterranean, then locks back down the far side, 91 locks in all. Its short summit stretch loses water downhill in both directions, so Pierre-Paul Riquet had to pipe in mountain streams through tens of kilometers of feeder channels and a dedicated reservoir just to keep the top filled. If ‘water finds one flat level’ described the world, no canal would ever have to climb a hill or hold a reservoir at its peak. They do, because ‘level’ follows the real ground.

Sources: Suez — sea-level, lock-free, ~193 km [158] · Panama — three lock flights lifting ships 26 m to Gatún Lake [159]; the Canal du Midi summit climb [249] · “level” = the geoid, and the Bedford Level result (Entry 2, Entry 12). → curvature & water data rows

• 1The strong version is true. At 3,000 to 6,000 m depth, abyssal plains are the smoothest large surfaces known. Height changes of 1 to 2 m over hundreds of km, slopes under 1:1000. The Sohm Plain alone spans about 900,000 km².
• 2What “flattest” measures. That figure is a local slope, how much the floor rises or falls per km. It is a relief number, and says nothing about whether the whole sheet curves. A pool table is “flat,” and so is any short stretch of a 1°-per-69-mile arc (Entry 8).
• 3They lie on the geoid. Turbidity currents blanket the rugged volcanic crust with fine sediment until the top matches the local equipotential, the curved “level” surface (Entry 12). Their smoothness is evidence that the sea settles onto a ball.
• 4The numbers do not add up to a disc. Spread a slope under 1:1000 across an ocean basin thousands of km wide, and the surface has dropped and curved by kilometers compared with a flat tangent plane. That is a ball’s bulge, hidden because every local patch is smooth.
• 5Even ‘sea level’ is a curved, bumpy surface. The plains lie beneath the ocean’s mean surface, and that surface is neither flat nor a perfect sphere. It is the geoid, the equipotential that still water settles to. Because Earth’s mass is spread unevenly, it rises and falls by up to about ±100 m off a smooth reference ellipsoid. There is a broad 106-meter ‘low’ in the Indian Ocean south of Sri Lanka, and highs elsewhere. Satellite missions like GRACE, GOCE, and radar altimeters map this gently rolling shape from orbit, and geodesists happily call the planet a slightly bumpy potato. The abyssal plain drapes over that curved equipotential, so ‘flat’ describes only the local sediment, laid down on a globe whose very sea level is measured, curved, and bumpy.

Sources: abyssal-plain gradient & extent [140] · the geoid / “level” [139] · curvature geometry [8] · the geoid undulation & Indian Ocean low [263]. Kin to the “flat” places on land (154). → Water & oceans data rows

• 1The shadow that maps the core. S-waves can't travel through liquid, so they disappear everywhere past about 103° from the epicenter. That reveals a liquid outer core starting at the Gutenberg boundary, about 2,890 km down. P-waves bend at that boundary, leaving a ring with no direct arrivals between about 103° and 142°. The size and shape of those shadows fix the core radius (about 3,480 km) and the planet's radius. The input is angular distance from the quake; the result is which waves arrive when.
• 2It only works on a sphere. The shadow zones are even rings at fixed angular distances no matter where on Earth the quake is, which is what a sphere with nested shells predicts. A flat disc of any thickness gives no such universal angle, and no reason for S-waves to cut off at one consistent arc. Inge Lehmann even found the solid inner core (1936) from faint P-waves sneaking into the shadow.
• 3Anyone can check the data. Global seismic networks publish arrival times openly, and the PREM reference model (1981) was built from thousands of quakes and predicts travel times to within seconds. It can be repeated, it can be falsified, and it can't be staged. The "experiment" runs for free every time the Earth shakes (compare Eratosthenes' surface measurement, Entry 1).
• 4The numbers behind the model. PREM was worked out from roughly 1,000 free-oscillation (normal-mode) periods and about 1.75 million P- and S-wave travel-time readings, a huge, openly published dataset that fixes the layer depths and the 6,371 km radius. Lehmann’s 1936 inner core came from a single faint signal, PKIKP, that should not have reached the shadow zone unless a separate solid core bent it inward. The interior is rebuilt from data, not assumed.
• 5The picture of the core is the shadow. The demand is for a photograph, but no camera reaches 2,890 km down, and none is needed. Every large earthquake X-rays the whole planet, and the shadow it casts, mapped above, is the picture of a liquid outer core wrapped around a solid inner one, drawn by thousands of seismometers for free every time the ground shakes. You can watch the waves sweep through the interior, quake by quake, in the IRIS Seismic Waves Viewer. [512]

Interactive — trace the rays, find the shadow

Every ray is traced by Snell’s law through a layered Earth, not drawn by hand. The shadow zone is nowhere in the code. It appears because the rays bend. Switch the outer core to solid and watch the S-wave shadow vanish.

Sources: seismic shadow zone & Earth's layered interior (Gutenberg, Lehmann, PREM) [72] · Lehmann 1936 & PREM 1981 [445]. Complements the surface measurement of Entry 1 and the gravity of Entry 17. → interior data rows · → seismic tools

• 1Each tower is plumb, so they can’t be parallel. A tower built straight up points at Earth’s center. Two such towers 1,298 m apart lean slightly outward. Over the 211 m height of the Verrazzano towers that opens a 41.3 mm gap at the top. It was a stated design requirement in 1964, recorded by the bridge authority and its engineers.
• 2You can weigh the curve with a tape measure. From the tower height (211 m), the spacing (1,298 m) and the 41 mm spread, similar triangles give Earth’s radius at roughly 6,400 km, within a few percent of the true 6,371 km. A flat Earth predicts exactly zero spread.
• 3It is not only bridges. The same bookkeeping shows up wherever structures get long: the 3.2 km Stanford linear accelerator is built straight through a curving Earth, London’s Crossrail tunnels were aligned for curvature, and LIGO’s 4 km arms (Entry 41) follow the chord, not the surface.
• 4Every precise survey already subtracts the curve. Long before any bridge, leveling and geodetic surveying carry a standard curvature-and-refraction correction. The Earth’s level surface falls away from a horizontal line of sight by about 0.0785 D² meters (D in km), trimmed by roughly a seventh for refraction to a combined ≈0.0675 D² m, about 7 cm at one kilometer, growing with the square of distance. Surveyors either balance fore- and back-sights to cancel it or compute it for long shots. To the people who measure the ground, the curve is not a fringe claim. It is a line item in the textbook.

Sources: curvature in large structures (Verrazzano-Narrows Bridge) [94] · the surveyor’s curvature-and-refraction correction [217]. Kin to the horizon geometry of Entry 1 and LIGO’s 4 km arms (Entry 41). → Curvature data rows.

• 1Cavendish (1798). Lead spheres attract each other inside a sealed torsion balance, giving G = 6.674×10⁻¹¹. Two inert weights pulling together in still air has no density or buoyancy explanation. It is gravity, and it let us "weigh the Earth."
• 2Schiehallion (1774). A plumb line clearly bends toward a mountain. Gravity points slightly sideways toward a large mass. Pure "down = density" predicts zero sideways pull.
• 3g is not uniform. It runs 9.780 m/s² at the equator to 9.832 at the poles and weakens with altitude, and tides require a 1/r² pull from the Moon and Sun. A single upward acceleration would read the same everywhere and produce no tides.
• 4Newton → Einstein. Newton's inverse-square law (1687) predicts orbits. General Relativity (1915) sharpens it where the two part ways: Mercury's 43″/century, starlight bent 1.75″ (1919 eclipse), redshift (Pound–Rebka 1959), gravitational waves (LIGO 2015).
• 5GPS proves it daily. Every receiver applies a ≈ +38 µs/day relativistic clock correction (SR −7, GR +45), or position drifts by about 10 km/day. Gravitational time dilation is a working engineering fact in your pocket.
• 6The deepest law: things fall the same no matter what they’re made of. Gravity speeds up a feather and a cannonball at the same rate. Gravitational and inertial mass are equal. The MICROSCOPE satellite confirmed this to about one part in 10¹⁵ in 2022 (lunar laser ranging tests the Earth and Moon to about 1 part in 10¹³). That exact match is the keystone Einstein built general relativity on. “It’s just density” explains none of it.
• 7Gravity can be mapped, and it isn’t uniform. Surface gravity runs about 0.5% stronger at the poles than the equator, partly from the spin, partly from the equatorial bulge (Entry 33). Gravimeters and the GRACE satellites (Gravity Recovery And Climate Experiment) chart these tiny changes into a global gravity map, the geoid. The way GRACE does it is worth picturing, because it is very hard to fake. Two satellites chase each other around the Earth about 220 km apart, and they measure the gap between themselves to about one micron, the width of a blood cell. Fly the leading one over a heavy region, a mountain range or a buried aquifer, and the extra pull tugs it forward, widening the gap; then the trailing one arrives and the gap closes again. That breathing distance, tracked since 2002, weighs the water moving around the planet: draining aquifers, melting ice sheets, rising seas. [657] A flat plane with “density” has nothing to map, and no center for the pull to point at. A spinning, slightly flattened mass has this signature and no other. It is the same satellite-geodesy toolkit that measures frame-dragging in Entry 120.
• 8Submarines navigate by the gravity map. The pull of gravity varies slightly from place to place, by amounts measured in milligals, where one milligal is roughly a millionth of the surface value. A device called a gravity gradiometer senses these variations, and a submarine matches them against a pre-surveyed map of the seafloor to fix its position, using no satellites and emitting no signal of its own. The method reaches about 100 meters of accuracy and cannot be jammed or spoofed. A flat Earth under one uniform upward acceleration would offer no such map, because the pull would read the same at every point. [548]

Sources: Cavendish [16] · Schiehallion/Maskelyne [17] · tests of GR [18] · GPS relativity [19]. → Gravity data rows

• 1Newton describes it, Einstein explains it. Newton’s inverse-square law predicts the motion of moons, planets and spacecraft to extraordinary accuracy. General relativity goes deeper: mass-energy curves spacetime, and free objects follow the straightest path through that curvature. It is confirmed by Mercury’s perihelion, the bending of starlight, GPS timing, and the direct detection of gravitational waves (Entry 41). “We can’t explain it” is just false.
• 2It isn’t “just density and buoyancy.” Buoyancy does not compete with gravity. It needs it: a fluid pulled down by gravity pushes less-dense things up (Entry 26). Take away the fluid and density stops mattering. In a vacuum a feather and a steel ball hit the floor together (Entry 35). Density explains the order in which things settle within a fluid. Gravity is what makes them settle at all.
• 3We measure the pull directly, down to milligrams. Gravity is universal: every mass attracts every other. Cavendish measured the attraction between lead spheres in 1798. In 2021 physicists in Vienna measured it between two gold balls of about 90 milligrams, the size of sesame seeds, and the result matched Newton and general relativity. The same universal pull acts on single atoms: cold-atom gravimeters time atoms in free fall to measure g with great precision.
• 4Weakest force, longest reach, and one honest gap. Gravity is about 10³⁶ times weaker than the electric force, so inside an atom it is completely negligible. But electric charges cancel over large bodies while mass only adds, so gravity wins at planet scale. What we still lack is a quantum theory of gravity. No graviton has been found, and general relativity and quantum mechanics aren’t yet unified. That is the deepest open problem in physics, not a reason to doubt that things fall.
• 5“Which gravity are you using?” They all agree on the fall. The gambit treats Newton, Einstein and quantum gravity as rival answers to “does the apple fall?” They aren’t. They are nested approximations, each with a known range. Newton’s law is the weak-field, low-speed limit of general relativity (for everyday masses the two agree to about one part in a billion), and general relativity is in turn the classical limit any quantum theory must reduce to. Demanding “which gravity?” to dispute that things fall is like demanding “which arithmetic?” to dispute that 2 + 2 = 4. Newtonian or relativistic, the measured g is 9.8 m/s² and the orbit comes out the same. The frameworks part ways only at the extremes that daily life never touches: near light speed, near black holes, at the Planck scale. Several working models is the mark of a mature science, not a contradiction.
• 6It isn’t electrostatics either. A related claim swaps density for static electricity: “gravity is really the electric force.” They are easy to tell apart. Gravity is only ever attractive, because mass comes in one sign and always pulls. Charge comes in two signs, so like charges repel. No object ever falls up toward a heavy mass the way two like charges fly apart. The forces also act on different things: gravity on mass-energy, electricity on charge. Between an electron and a proton the electric attraction is about 2.3 × 10³⁹ times stronger than their gravity. Yet ordinary matter is almost perfectly neutral, so all that huge electric force cancels out, leaving only the feeble, never-canceling pull of mass to steer planets and stars. And gravity can’t be screened: a Faraday cage zeroes an electric field completely, but nothing known blocks, reflects, or cancels gravity.
• 7Drop a charged ball and a neutral one, they fall together. The deciding test is free fall. In a vacuum every object speeds up at the same rate no matter its mass, material, or net charge. A feather and a hammer land together (Entry 35), and the equivalence principle is confirmed to about one part in 10¹⁵ (the MICROSCOPE satellite, Entry 25). An electric force does the opposite. It acts only on charged bodies and sorts them by charge-to-mass ratio, flinging a charged pith ball while ignoring a neutral one. And Cavendish (1798), repeated with 90-milligram gold spheres in 2021, measured attraction between electrically neutral masses, where electrostatics predicts almost nothing. Both forces are real and both obey an inverse-square law (Newton’s and Coulomb’s). They are clearly not the same force.
• 8Faraday tried to unify them, and failed honestly. The idea that gravity is really electromagnetism is older than the flat-Earth movement, and it was tested by the man who built the field concept. In 1849 Michael Faraday set out to find an experimental relation between gravity and electricity, dropping and lifting heavy weights inside coils to see whether falling matter would induce a current. Nothing appeared. He published the negative result rather than bury it, and wrote that it gave “no proof that such a relation exists.” A decade later he tried once more and the Royal Society declined to print the result. The link has been looked for by the best experimentalist of the age and has not been found. 191 [573]

Sources: gravity measured between ~90-mg gold spheres, matching Newton & GR [163] · gravity is universal, the weakest force, not yet unified with quantum theory [163, 164] · gravity vs. the electric force [185] · buoyancy needs gravity (Entry 26), vacuum free-fall (Entry 35), gravitational waves (Entry 41). → gravity data rows

• 1Scalar = size only. Mass (5 kg), density (1000 kg/m³), temperature (20 °C), speed (30 km/h), energy, time. No direction, and they add like ordinary numbers.
• 2Vector = size + direction. Velocity (30 km/h north), acceleration, displacement and force all carry a direction. Two equal vectors pointing opposite ways cancel to zero. Two scalars never cancel that way.
• 3A force is a vector. A push or pull with a size and a direction, measured in newtons. Weight is the force of gravity on a mass: W = m·g, pointed toward Earth’s center. Mass is the scalar. Weight is the vector it produces.
• 4Why density alone can’t make things fall. Density is a scalar, with no “down” built in. Sinking and floating are a contest of forces in a fluid (weight pulling down versus the buoyant force pushing up), and both of those forces exist only because gravity provides the downward direction (Entry 26).
• 5Mass versus weight. Your mass (a scalar) is the same on the Moon. Your weight (a vector force) is about one-sixth, because g is smaller there. Treating the two as one quantity is the heart of the “it’s just density” mistake.
• 6This is why “down” is real and local. Gravity is the vector that aims every “down” at the center of the Earth (Entry 29). A scalar like density could never pick out a direction, let alone the same center from every side of a globe (Entry 18).

Sources: Scalars versus vectors, and force as a vector [371]. Builds on buoyancy and Archimedes (Entry 26), why “down” is local (Entry 29) and how gravity works (Entry 18).

• 1The four forces, strongest to weakest. The strong force binds quarks into protons and neutrons and holds nuclei together (range ~10^-15 m; carriers: gluons). Electromagnetism binds atoms and molecules and carries light and electricity (infinite range; carrier: the photon). The weak force drives radioactive decay and helps power stellar fusion (range ~10^-18 m; carriers: the W and Z bosons). Gravity binds planets, stars and galaxies (infinite range; carrier: the proposed, still-undetected graviton).
• 2Gravity is staggeringly weak. Between two protons, gravity is about 10³⁶ times weaker than their electric repulsion. That is why a small fridge magnet (electromagnetism) beats the pull of the whole Earth (gravity) to lift a paperclip.
• 3So why does the weakest force run the universe? The strong and weak forces reach no farther than an atomic nucleus. Electromagnetism is long-range but mostly cancels, because matter carries equal positive and negative charge. Gravity is the only force that is long-range and always attractive and never cancels. So across planetary and cosmic distances it wins, shaping orbits, stars and galaxies.
• 4What each force acts on. The strong and weak forces act on subatomic particles (quarks and the like). Electromagnetism acts on electric charge. Gravity acts on all mass-energy, and nothing is exempt, which is why everything falls the same way (Entry 18).
• 5Newton’s gravity. A force between any two masses, F = G·m₁m₂/r², along the line joining them. Simple, and accurate enough to fly spacecraft to the planets.
• 6Einstein’s gravity. Mass-energy curves spacetime, and objects fall because they follow the straightest path, a geodesic, through that curved geometry: “matter tells spacetime how to curve; spacetime tells matter how to move.” It reduces to Newton’s law for weak fields and slow speeds, and adds what Newton cannot: Mercury’s perihelion shift, starlight bending at the Sun, gravitational waves (Entry 41) and the time dilation GPS must correct (119, Entry 25).
• 7Same thing, deeper map. Moving from Newton’s force to Einstein’s geometry refines how we describe gravity, not whether it exists. Both say the apple falls. Gravity is also the one force not yet merged with quantum theory, the open problem of quantum gravity (178). But its existence and direction are not in any doubt.

Sources: The four forces — strengths, ranges and carriers [372][373]; gravity from Newton’s force to Einstein’s geometry [374]. See how gravity works (Entry 18), spacetime curvature (Entry 25), GPS relativity (119) and the unsolved quantum-gravity frontier (178).

• 1Gravity acts by mass, not by permission. Weight = mass × g. The same g (~9.8 m/s²) pulls a ~0.5-gram butterfly and the ~1.4×10²¹ kg of the oceans alike. Gravity does not “decide” to hold one and release the other. It pulls every gram of both, all the time (Entry 19).
• 2Flight is not anti-gravity. It is lift beating weight. A butterfly’s wings push air down, and the air pushes the butterfly up. When that upward push beats its tiny weight, it climbs, not because gravity vanished, but because a stronger local force won. The instant it stops, gravity brings it straight back down.
• 3A 400-tonne airliner does the same thing. Nobody says “gravity is too weak to hold a jumbo jet” because it flies. A loaded airliner, some 400 tonnes and far heavier than any butterfly, climbs to 11 km on lift and thrust, and drops out of the sky the moment its engines quit. Flight at any size is force beating weight, not gravity switching off.
• 4The butterfly never leaves Earth. It flits a few meters up, lands, and stays fully bound to the planet. It cannot drift off into space. “Holding down” was never the same as “forbidding all motion.” You jump, a bird soars, dust blows, all rise briefly and all come back. Gravity holds them in just the sense it holds the oceans.
• 5The oceans have nothing to fly with. Water makes no lift and burns no fuel. With nothing to push it up, it just runs to the lowest point and pools there. That is not gravity being “extra strong” on water. It is water having no way to resist the same ordinary pull a butterfly beats with muscle.
• 6This is gravity being the weakest force, not the strongest. Gravity is by far the feeblest of the four fundamental forces, roughly 10³⁶ times weaker than electromagnetism (Entry 20). That weakness is why a butterfly’s muscles, a fridge magnet, or your own legs can briefly beat the whole Earth’s gravity on a small mass, while that same gentle pull, added up over every atom of the planet, inescapably governs the oceans, the air, and you.

Sources: Lift, thrust, drag and weight are four distinct forces, and an aircraft climbs when lift exceeds weight [378]; the oceans (hydrosphere) total about 1.4×10²¹ kg [379]; gravity is the weakest fundamental force, ~10³⁶× weaker than electromagnetism (Entry 20). Connects to scalars, vectors and forces (Entry 19).

• 1Weight follows mass and radius, not density. Surface gravity is g = G·M÷R²: add mass and g rises, spread that mass over a larger radius and g falls. Density never enters the formula, which is why the gravity and density columns above tell completely different stories.
• 2Saturn would float, yet it pulls like Earth. Saturn’s mean density is just 0.69 g/cm³, less than water, so a lump of it would float in a big enough ocean. Yet its surface gravity is 0.92× Earth’s. A “denser means heavier” rule would leave you nearly weightless there. Gravity (vast mass, large radius) says otherwise, and that is just what spacecraft measure.
• 3The Moon is denser than Saturn but pulls far less. The Moon (3.34 g/cm³) is nearly five times denser than Saturn (0.69), yet you weigh 0.17× on the Moon versus 0.92× on Saturn. Density predicts the order backwards. Mass and radius get it right.
• 4Mercury and Mars: same gravity, different worlds. Mercury (5.43 g/cm³) and Mars (3.93) differ in density and size, yet both pull at 3.7 m/s², 0.38× Earth, because their different masses and radii give the same G·M÷R². Buoyancy cannot produce that coincidence. The gravity formula gives it on the first try.
• 5Airless worlds still have weight. The Moon, Mercury and Ceres have essentially no atmosphere, so there is no fluid for anything to be buoyant in. Yet a dropped hammer still falls, at that body’s own g. The Apollo astronauts’ filmed one-sixth-gravity bounding matches g_Moon = 1.62 m/s² to the decimal (Entry 18).
• 6Big-G stays put, little-g varies. The gravitational constant G (6.674×10⁻¹¹) is the same everywhere in the universe, which is what makes the formula universal. What changes from world to world is little-g, the surface gravity, because M and R change. Calling surface gravity “the gravitational constant” mixes up the one number that never varies with the one that always does.
• 7From a pebble to a neutron star. One law spans the whole table: Ceres pulls at 0.28 m/s², Earth at 9.81, the Sun at 274 (28× Earth), and a neutron star at ~2×10¹² m/s², over a hundred billion times Earth’s, where a sugar-cube of its matter would outweigh a mountain. Eleven-plus orders of magnitude, one formula, no buoyancy required.

Sources: Surface gravity, weight-ratio and density for the planets, Moon and Pluto are NASA NSSDCA fact-sheet values [380]; the neutron-star figures (~2×10¹² m/s², ~10¹¹× Earth; ~10¹⁴ g/cm³) [381]. Connects to how gravity works (Entry 18), mass and weight as scalar versus vector (Entry 19), and why density governs floating rather than weight (Entry 26).

• 1Mass attracts mass, measured in a room. Two 0.73-kg lead balls sat on a rod hung from a thin wire. Two 158-kg lead spheres brought near drew them across, twisting the wire by a readable angle. The pull between a small and large ball is about a ten-millionth of a newton, tiny but clear, and purely gravitational.
• 2It has nothing to do with buoyancy. The apparatus sat in a sealed, draft-proof case, and the small balls moved sideways toward the big masses, not up or down. Buoyancy (Entry 26) only sets how objects of different density arrange themselves in a fluid. It cannot produce a horizontal attraction between two pieces of lead.
• 3It “weighed the Earth.” Comparing the big sphere’s pull on a small ball with the whole Earth’s pull on that same ball gave Earth’s mass and density without even needing a value for the gravitational constant. The result: a mean density of about 5.45–5.5 times water (modern figure 5.514).
• 4That density reveals a dense interior. Surface rock is only about 2.7 times denser than water, yet the whole Earth averages about 5.5. So the inside must be far denser, pointing to a heavy iron core. That matches what earthquake waves show on their own (Entry 15).
• 5The constant came later, the force did not. Cavendish himself never reported “big G.” That universal constant (6.674×10⁻¹¹) was pulled from his data decades afterward. What he showed in 1798 is what “density-only” claims deny: a real attractive force between masses.
• 6Anyone can repeat it. The Cavendish experiment is a standard undergraduate lab worldwide, and modern versions with a laser pointer and mirror reproduce the constant to a few percent in an afternoon. It is among the most repeated measurements in physics, not a one-off claim.
• 7Michell devised it; Cavendish credited him. The experiment and the torsion balance were the work of Reverend John Michell, who died in 1793 before he could run it. The rig passed to Cavendish, who rebuilt it, kept close to Michell’s plan, and named him in the 1798 paper. That is why it is sometimes called the Michell–Cavendish experiment. Inheriting a friend’s unfinished apparatus and crediting him is the opposite of stealing, and it changes nothing about a result that anyone can now repeat. [500]

Cavendish torsion-balance experiment & Earth’s density [353] · apparatus details, the 158-kg spheres [354]. Connects to gravity vs density (Entry 26), the mechanism of gravity (Entry 18) and Earth’s dense core (Entry 15).

• 1The mountain deflected the vertical. On each side of Schiehallion the plumb line leaned toward the bulk of the mountain, so the local “straight down” differed from the star-defined vertical. Combining both sides gave a total deflection of about 11.6 arcseconds (about 0.003°), small but firmly measured across 337 nightly star observations in 1774.
• 2It proves attraction toward mass. Only a real gravitational pull toward the mountain’s matter can swing a plumb bob sideways. This is the open-air partner to Cavendish’s lab result (Entry 23): mass attracts mass, whether it is a lead sphere or a Highland peak.
• 3Newton thought it was too small to see. Newton himself had guessed a mountain’s deflection would be too small to measure. Maskelyne’s careful astronomy proved him too pessimistic and turned a thought experiment into a hard number.
• 4It weighed the Earth. Knowing the mountain’s size and rock density from a painstaking survey, Charles Hutton worked back from the deflection to Earth’s mean density, roughly 4.5–5 times water in the 18th-century analysis, refined by later work toward the modern 5.5. The Earth came out far denser than its surface rock, ruling out a hollow interior.
• 5It invented contour lines. To compute the irregular peak’s pull, Hutton joined points of equal height on his survey. That was the first use of contour lines, now standard on every topographic map. Schiehallion is where geophysics began.
• 6“Down” is local, not absolute. Because the plumb line leans toward nearby mass, vertical is set by where the matter is, not by a fixed cosmic direction. That is why “down” points to Earth’s center everywhere on a globe, and why a mountain can nudge it aside.

Schiehallion plumb-line deflection & Earth’s density [355] · Maskelyne/Hutton, contour lines, hollow Earth ruled out [356]. Pairs with the Cavendish lab experiment (Entry 23) and Earth’s dense interior (Entry 15).

• 1Warped time: you read it off a clock. A clock deeper in gravity ticks slower. Pound and Rebka showed it up a 22.5 m tower in 1959. Hafele and Keating flew atomic clocks around the world in 1971 and saw the offset. And your phone depends on it: GPS satellite clocks run about 38 microseconds per day fast, and without that correction positions would drift roughly 10 km per day (119). Optical clocks now detect the difference across a 33 cm change in height. Curved time is built into a device in your pocket.
• 2Bent light, and the factor of two that proves space curves. The Sun deflects grazing starlight by 1.75″, confirmed at the 1919 eclipse and by modern radio interferometry to a few parts in ten thousand. That is twice the 0.87″ you would get if light simply had weight in flat space. The extra half comes from the curvature of space itself. Gravitational lensing now bends the light of distant galaxies into arcs and full Einstein rings, a routine tool for mapping mass.
• 3Ripples in spacetime, caught in the act. Curvature isn’t only static. It travels. The Hulse-Taylor binary pulsar spirals together at the exact rate it should if it radiates gravitational waves (Nobel, 1993). In 2015 LIGO directly caught two merging black holes stretch its 4 km arms by about 10⁻²¹ of their length, less than a proton’s width, matching GR’s predicted waveform. About a hundred such events have followed (Entry 41). In 2017 a neutron-star merger seen in waves and light pinned the speed of gravity to the speed of light.
• 4It has to be geometry, not a force. Every object falls the same way no matter its mass or material, tested to about one part in 10¹⁵ (the MICROSCOPE satellite). A “force” pulling on a property objects have in different amounts can’t do that. A curved stage they all move through can. And GR’s geometry predicts all of it: Mercury’s 43″-per-century drift, the Shapiro delay, Earth dragging spacetime (Gravity Probe B), black-hole shadows (the Event Horizon Telescope), with no free parameters. The one open frontier is uniting it with quantum mechanics (Entry 18), a deeper layer, not a crack.

Sources: gravitational redshift / Pound-Rebka [167] · light deflection 1.75″ vs 0.87″, 1919 & modern VLBI (very long baseline interferometry) [168] · Cassini Shapiro delay [169] · Gravity Probe B frame-dragging [170] · LIGO GW150914 [171] · GW170817, gravity at light speed [172] · Event Horizon Telescope black-hole shadows [173] · S2 orbit precession [174] · GPS time dilation (119), gravitational waves (Entry 41), the quantum-gravity frontier (Entry 18). → relativity data rows

• 1The buoyancy formula contains gravity. In its modern form, the upward force = (fluid density) × (volume displaced) × g. That g is the strength of gravity. Density and buoyancy don't compete with gravity. They are a result of it.
• 2What Archimedes really said, and when. Archimedes set out the buoyancy rule in On Floating Bodies, around 250 BC. Newton published gravity in the Principia in 1687. The gap is about nineteen centuries, and it is sometimes offered as proof that floating needs no gravity. Read the rule again. The buoyant force equals the weight of the fluid pushed aside. Weight is what gravity does to mass. Archimedes measured the effect correctly and had no account of its cause, which is the normal order of things: Kepler described the orbits in 1609 and Newton explained them in 1687. A description arriving first is not evidence the explanation was invented. And the witness is a poor one for the claim. In the same treatise Archimedes proves that water settles into a spherical surface around a center toward which everything falls, which is a round Earth with gravity, written down two thousand years early. [602]
• 3Why dense sinks. Gravity pulls all the fluid down, building a pressure that is higher at the bottom than the top. That pressure difference pushes lighter things up and lets heavier things settle. No gravity means no pressure gradient, which means no sorting by density.
• 4The clean test: free fall. In orbit or a drop tower, everything is weightless and buoyancy vanishes. A helium balloon doesn't rise, oil and water don't separate, a bubble just sits there. If "density" were a force of its own, it would still work in free fall. It doesn't, because it was gravity all along.
• 5It also can't explain the rest. Density says nothing about why g changes with latitude, why tides follow the Moon, why Cavendish's lab masses attract (Entry 17), or why GPS needs a relativistic clock fix. Gravity explains all of it. "Density" explains only the order things settle in, once gravity is already doing the pulling.
• 6Katabatic winds: gravity made visible on a continental scale. Antarctica shows it at full size. Bitterly cold air over the high ice plateau is denser than the air around it, so gravity drags it down the slope toward the coast as a ‘katabatic’ wind. The word means ‘going downhill.’ At Cape Denison in Commonwealth Bay the ice funnels this drainage into some of the strongest sustained surface winds on Earth, with gusts past 270 km/h, the ‘Home of the Blizzard’ Douglas Mawson met on his 1911–14 expedition. The cold air pours downhill the same way water would, because the same thing pulls on both: gravity.
• 7‘Density and buoyancy’ is the gravity here. The flat-Earth reply is that this is ‘just’ cold dense air sinking, density, no gravity required. But the force that pushes that dense air down the slope is the buoyancy force, and buoyancy is proportional to g: it is the weight of the displaced fluid (keypoint 1). Set gravity to zero and that force is zero. The cold air would hang in place and drain nowhere, the same reason nothing rises or sinks in orbit. Katabatic winds aren’t an escape from gravity. They need it, and they are one of its largest visible signatures.
• 8A density gradient is gravity’s signature, not its replacement. “The air is layered by density, not gravity” has it backwards. With no gravity there is no “down” for the dense to settle toward, and gases mix into a uniform blend, with no layering at all. Stratification, buoyancy, and the smooth fall of pressure with height all require a gravitational field. The barometric formula for the atmosphere’s exponential thinning (scale height H = kT/mg) has g sitting in the denominator. Density explains which way things separate. Gravity is why they separate at all.
• 9A gravity assist is gravity doing what density never could. In May 2026 NASA’s Psyche spacecraft swung past Mars and stole speed from it. The planet’s gravity bent the probe’s path and added about 1,000 miles per hour, with no fuel burned at all. Density and buoyancy cannot do that. Buoyancy needs a fluid to push against, and there is none in the vacuum between planets. Nothing about how dense the probe is reaches across 4,600 kilometers of empty space to change its speed. Only gravity, a real force that acts at a distance, does. And it was no lucky coincidence. The navigators aimed the craft years ahead to arrive where Mars would be, and used the planet’s motion around the Sun to gain the boost. That math only works if the planets orbit the Sun on the paths a Sun-centered model lays out. The boost was not asserted, it was measured: the Deep Space Network read the speed change straight off the probe’s radio signal, from its Doppler shift, in real time. You cannot slingshot a spacecraft with density. This is gravity, working as Newton wrote it, all the way out at Mars. [678] 122

Sources: Archimedes' principle [56] · gravity experiments [16],[18] · the equivalence principle (MICROSCOPE) [401] · katabatic winds in Antarctica [200],[201]. → Gravity data rows

• 1Buhler’s standing is genuine and worth stating in full. He holds a doctorate in condensed matter physics from Florida State, spent more than two decades at NASA’s Electrostatics and Surface Physics Laboratory at the Kennedy Space Center, co-founded that laboratory and now leads it, worked on electrostatic discharge safety for the Space Shuttle, the International Space Station and Hubble, and is incoming president of the Electrostatic Society of America. [704] Nothing here rests on him being a crank, because he is not one.
• 2The work is not NASA’s. Exodus Propulsion Technologies is a private company and the drive is not a NASA program or a NASA finding. [704] Headlines reading “NASA scientist” describe where he works on other things. The distinction matters, because the claim borrows an agency’s authority for a result the agency has not endorsed.
• 3He rules out the easy explanation himself, so the usual answer does not apply. Ion wind is the standard way an electric field appears to push something: charge the air, throw it backward, move forward. Buhler wants none of it, and the tests run in hard vacuum with polystyrene at 30,000 to 40,000 volts specifically to stop the gas breaking down. [705] Anyone answering this by saying “it is only ion wind” has not read the method.
• 4What is honestly unsettled. The team reports thousands of vacuum tests run between 2016 and 2023, and a theory built on third-order quantum electrodynamic perturbation with momentum drawn from the quantum vacuum. [705] The results have been presented at an alternative-propulsion conference rather than published in a reviewed journal, and the promised paper has not appeared. That is a gap in the record, not proof of error.
• 5The reported numbers do not agree with each other. Some accounts describe thrust exceeding one g; others describe forces in the millinewton range. [704] [706] Those can both be true, because one g is an acceleration and a millinewton is a force, and a light enough object accelerates at one g under a tiny push. Reporting that does not separate the two is not evidence of anything except loose reporting.
• 6Buhler names the test that would settle it, and it has not happened. He has said the decisive experiment is flying a self-contained unit in orbit, where the Earth-bound artifacts that plague this kind of measurement fall away. [705] That is the right instinct and he deserves credit for holding it. Until the flight happens, the correct description is promising and unproven.
• 7Now grant him everything, because the argument fails anyway. Suppose the effect is real, works in vacuum, and lifts more than its own weight. What has been demonstrated is a way to generate force without throwing mass out the back. That is a propulsion result. It is not a statement about why unsupported objects fall, why the oceans stay put, or what shape the planet is.
• 8Every counter-force presupposes the force it counters. A helicopter holds station against gravity. So does a hot-air balloon, a hovering drone, a table under a book and your own legs. None of them are evidence that gravity is absent; all of them are evidence that it is present and has to be worked against. Quoting a drive’s performance as a multiple of g concedes the whole point, because g is the number being fought.
• 9What would overturn gravity is a different experiment entirely. Not a device that pushes, but two ordinary uncharged masses that fail to attract each other. Cavendish measured that attraction in 1798 with lead spheres and a torsion balance Entry 23, the measurement has been repeated for two centuries with steadily better precision, and it does not care whether anyone has built a good thruster.
• 10Gravity acts on mass, electric fields act on charge, and that is a testable difference. Discharge an object, ground it, put it in a Faraday cage Entry 191, and its weight does not change by a milligram. An electric drive stops working when you take the charge away. Whatever the Exodus Effect turns out to be, it is a device that needs 40,000 volts to do anything, and gravity needs no volts at all.

• 1What was flown, with the numbers. A fixed-wing aircraft of 5 m span and 2.45 kg, carrying a battery stack and a high-voltage converter, flown ten times over a course of about 60 m indoors at MIT, at roughly 11 mph. [707] The result was published in Nature in November 2018 under the title of a flight with solid-state propulsion. The claim on this page is not that any of this is exaggerated.
• 2The wings do the lifting, and they always did. This is an aeroplane, not a hovering craft. Air flowing over a wing at speed produces lift, which is what holds the weight up. The electroaerodynamic drive supplies forward thrust in place of a propeller. Take the wings off and no amount of voltage keeps it in the air.
• 3It pushes on air, which is Newton’s third law with an unusual fan. A high voltage strips electrons from air molecules, the resulting ions are pulled toward the opposite electrode, and on the way they collide with millions of neutral molecules and drag them along. [707] Air goes backward, the craft goes forward. That is a propeller that uses an electric field instead of blades, and in vacuum it does nothing at all.
• 4The effect is about a century old. Ionic wind was described in the 1920s and hobbyists have been lifting foil triangles with it on benches for decades. What MIT achieved was the engineering, getting the thrust-to-weight high enough to fly a real airframe. That is a hard problem and they solved it. It is not new physics.
• 5It obeys gravity in the most visible way possible. A craft that had cancelled its weight would not need a 60 m run, would not fly at 11 mph, and would not need a gymnasium wall to stop it. The first flight ended in a crash. Every second it was airborne, its wings were doing work against a downward pull, and the amount of work required was set by the 2.45 kg it never stopped weighing.
• 6Charge and mass are not the same handle. The drive works because it puts charge onto air molecules and then pulls on that charge. Discharge the air and the thrust stops. Gravity has no such switch: it acts on the airframe, the battery and the pilot’s coffee regardless of whether any of them carries a net charge. That is a difference you can test on a bench Entry 191, and it is the same distinction that separates a working electric drive from a claim about gravity Entry 27.
• 7Silence is a property of the engine, not of the physics. The craft is quiet because nothing is spinning and no blade is slapping the air. A glider is quiet too. Quietness is a fact about noise, and it carries no information about what shape the planet is or whether it falls toward its own center.

• 1Gravity points to the center, everywhere. Every part of Earth’s mass pulls you toward the planet’s center of mass, so “down” is radial. It points inward from every spot on the surface (Entry 18). Drop a ball in Sydney or in London and it falls toward the same center, just from opposite sides.
• 2“Up” and “down” are local, not cosmic. Space has no absolute top or bottom, and on a sphere no point is the “real” top. We only put the Northern Hemisphere “up” on maps by habit. So calling Australia “upside down” assumes a universal up that does not exist. It is just a Northern-Hemisphere habit (62).
• 3Nobody feels inverted. A person at the North Pole and one at the South Pole are physically head-to-head, each upside down relative to the other. Yet both stand comfortably upright, because each one’s “down” is their own local vertical. “Upside down” only means something when you compare two observers. On your own patch of ground you are always right-side up.
• 4The plane’s floor turns with it. An aircraft flying “around the bottom” follows the curved surface, and gravity keeps pulling toward the center the whole way, so the cabin floor stays underfoot the entire time. The passengers never pass through an “inverted” moment, because down turns with them by the same amount the surface curves (Entry 46).
• 5A flat Earth would not even fix it. If Earth were a disc with enough mass to hold an atmosphere, its gravity would still pull toward the disc’s own center. So people away from the middle would feel pulled sideways, and those at the rim would feel they were on a steep slope or sliding toward the center. As one physicist put it, “true down is always toward the center of Earth.” The “falling off” worry is worse on a flat plane, not better.
• 6The same idea explains the flipped sky. Because your local “up” points a different way in space at every latitude, observers in different places see the Moon and stars rotated relative to one another (76). It is one consistent picture: gravity sets “down” toward the center, and orientation, the sky, and why no one falls off all follow from that (Entry 24).
• 7It is perspective, and the map is the trick. Standing on Earth you can never feel upside down. Gravity fixes your “down” toward the center, so your feet and your inner ear always agree you are upright, anywhere on the globe. From space there is no up or down at all, and an astronaut can turn the Earth any way they like. The southern hemisphere only looks like it is “hanging underneath” because of the human habit of drawing North at the top of a map. Flip the map over and Europe is on the bottom. Space has no top, so neither person is the one truly inverted. Each is right-side up in their own frame, and “upside down” just means borrowing someone else’s.
• 8“Down” is not a redefinition. It is a prediction, and it has been measured. A flat Earth needs one universal down: every plumb line on the planet hanging parallel to every other one. A globe needs the opposite. Plumb lines must converge, leaning toward each other by the same angle their two locations subtend at the center. Over 1,000 kilometers of ground that angle works out to about 9°. Two bubble levels that far apart are pointing a tenth of a right angle away from each other, and there is nothing subtle about that. Surveyors have been living with it for two centuries. Geodesy even has a name for the leftover: the deflection of the vertical, the small mismatch between where a plumb bob truly hangs and where the reference ellipsoid says down ought to be. It runs to a few arcseconds in flat country and up to about a minute of arc in mountains. Arcseconds is what remains after the globe has been subtracted out. On a flat Earth the leftover would not be arcseconds. It would be the entire difference in latitude, in degrees. [617]

Gravity points to Earth’s center, “down” is local, no absolute up/down [369] · even a flat disc’s gravity pulls to its own center (physicist) [370]. Connects to how gravity works (Entry 18), “down” proven local at Schiehallion (Entry 24) and the upside-down Moon (76).

• 1A coat of varnish. Mean ocean depth is 3,682 m, and Earth’s radius is 6,371 km. The water layer is about 0.058% of the radius. On a 30-cm classroom globe it would be a film under 0.1 mm thick, thinner than the paint.
• 2Mass, not coverage. Surface area misleads. Total ocean mass is about 1.35×10²¹ kg, roughly 1/4400 of Earth’s 5.97×10²⁴ kg, about 0.02%. It is a thin skin on a rock ball, just what gravity holds to the geoid (Entry 12).
• 3Gravity beats spin by about 290×. At the equator, surface gravity is about 9.78 m/s², while the outward (centrifugal) effect of the daily spin is only about 0.0339 m/s². Water is not flung off because gravity wins by a huge margin, the same reason you are not (Entry 45).
• 4“Seeks its level,” and finds the geoid. Water settles onto a surface of constant gravitational potential, which is curved (Entry 12). “Level” was never “planar.” A thin curved ocean film is just what a self-gravitating, spinning body produces.
• 5The bulge is real, and the sea sits on it. A spinning ball doesn’t just hold its water. It flings the equator outward, and the oceans follow. Sea level at the equator sits about 21 km farther from Earth’s center than at the poles (6,378 vs 6,357 km), the rotating oblate equipotential, made of water, mapped today to the centimeter by satellite altimetry. The single point of land farthest from the center isn’t Everest but Chimborazo in Ecuador, about 1° off the equator and about 2.1 km beyond Everest’s summit despite sitting far lower above the sea. The water never flew off. It bulged out right where rotation says it should.

Sources: ocean mass, volume & mean depth [141] · Earth mass & radius [1] · spin vs gravity [4] · the equatorial bulge & the farthest summit from Earth’s center [210]. → Water & oceans data rows

• 1What holds the ball up is a pump. The granite sphere is denser than water and never floats. It sits in a socket cut to match its curve, and a pump forces water into the gap, a film about a third of a millimeter thick, thinner than a credit card. That pressurized film is a hydrostatic bearing. Cut the pump and it drains in seconds and the ball thuds into its cup. [691]
• 2The water is pumped, contained, and pouring off the whole time. It is lifted from below under pressure, held in the paper-thin gap by the fitted socket, and spilling over the rim to be recirculated. So it clings for three reasons: a pump, pressure, and a socket. Not one of them is gravity, and a planet has none of them.
• 3A planet’s own gravity could not run a Kugel. The 29-ton granite Earth in Richmond has a surface gravity millions of times weaker than the real Earth’s, far too faint to hold a teaspoon of water. The effect is engineering, which is why a hairline crack that spoiled the sphere’s roundness once left all 29 tons sitting dead in its cradle until it was replaced. [692]
• 4The analogy runs backwards, and the honest version is stronger. Post that the Kugel shows water sticking to a spinning ball like the oceans, and the correct reply writes itself: that water is pumped and mechanically contained, and the oceans are not. Correcting a weak argument on your own side is what earns a reader’s trust.
• 5The real reason the oceans stay: gravity beats the spin about 290 to 1. Gravity pulls every water molecule toward the center at about 9.8 m/s². The outward tug of the spin at the equator, where it is strongest, is only about 0.034 m/s². To throw the oceans off you would have to spin the Earth about 17 times faster, a day near 84 minutes (Entry 45, Entry 30).
• 6What the Kugel does show honestly. Two real things. A massive sphere on a near-frictionless bearing keeps turning with almost no effort, which makes “a giant ball needs no motor to keep spinning” something you feel with your hand, and space is emptier than any water film. And the best Kugels are Earth globes you can walk up to, the Richmond sphere carved with the continents and paired with a scale Moon 250 feet away. [692]

Sources: the granite sphere fountain physics [691] · the Science Museum of Virginia Earth Kugel [692] · spin vs gravity [4].

• 1Round when gravity beats material strength. A solid body holds its shape by the strength of its rock or ice, while gravity pulls every part toward the center. Once a body is massive enough that the inward pull beats what the material can support, it deforms (over geological time) into the lowest-energy shape: a ball, or an oblate spheroid if it spins (Entry 18).
• 2The “potato radius.” Astronomers nickname the crossover the potato radius, around 200 km in radius. Below it bodies stay potato-shaped; above it they round off. It is roughly 400 km in diameter for weak icy bodies and ~600 km for stronger rock, because rock resists gravity more stubbornly than ice.
• 3The smallest sphere and the largest potato. Saturn’s icy moon Mimas (396 km across) is the smallest body known to be rounded by its own gravity; Neptune’s moon Proteus (420 km) is the largest known body that is still irregular: bigger than Mimas, yet lumpy, because the threshold depends on composition, temperature and history.
• 4Rocky holdouts, and the dwarf-planet line. The asteroids Vesta and Pallas (~520 km) are large but still potato-shaped, because rock’s greater strength keeps them from relaxing. The dwarf planet Ceres (945 km) is the smallest body confirmed round by gravity. That is why the IAU’s definition of a planet (and dwarf planet) requires “hydrostatic equilibrium,” the rounded shape self-gravity produces.
• 5Earth is far, far over the line. At ~12,742 km across, Earth sits some thirty times past the roundness threshold, where gravity overwhelms rock by a wide margin. The tallest mountain, Everest, rises ~8.8 km, under 0.14% of Earth’s radius. Scaled to a 30-cm desk globe that is a bump about a fifth of a millimeter high. A flat slab this size is not a stable shape. Gravity would pull it into a ball.
• 6The whole sky confirms the rule. Every object above the threshold we can resolve is round: the Sun, the eight planets, the large moons, Ceres, Pluto. Everything below it is an irregular lump: asteroids, comets, small moons. A flat Earth would be a lone, unexplained exception to a pattern that holds across the entire Solar System (Entry 30).

Sources: The roundness threshold and examples (Mimas, Proteus, Vesta, Ceres) [375]; the “potato radius” derived from gravity versus material strength [376]; ~200 km (icy) to ~400 km (rocky) radius for self-gravity to win [377]. Connects to how gravity works (Entry 18) and why fluids form balls in space (Entry 30).

• 1Spin makes the bulge, and it was predicted first. In the Principia (1687) Newton argued that a rotating, self-gravitating Earth must bulge at the equator and flatten at the poles, estimating the flattening near 1/230. Huygens reached a similar figure in 1690. The shape was worked out from physics before it was ever surveyed.
• 2Two expeditions settled it. To test Newton’s oblate Earth against a rival “prolate” (pole-stretched) model, France sent the Geodesic Missions of the 1730s–40s. The mathematician Maupertuis went to Lapland and the explorer La Condamine to the equator in Ecuador, each measuring the length of one degree of latitude. A degree runs longer near the poles, so Earth is flattened there, just as Newton said.
• 3The numbers are small but real. Equatorial radius 6,378.1 km, polar 6,356.8 km, a difference of ~21.4 km and a flattening of about 1/298. On a desk globe that is well under a millimeter. “Oblate spheroid” and “sphere” describe the same ball to better than half a percent. Refining a sphere to an ellipsoid is not changing the story. It is measuring it.
• 4Chimborazo, not Everest, reaches farthest into space. Because the equator bulges, the summit of Ecuador’s Chimborazo (just 1.5° south) sits about 6,384 km from Earth’s center. That is roughly 2 km farther out than Everest, even though it stands ~2.5 km lower above sea level. The bulge reorders which peak is “highest” measured from the center.
• 5The bulge shows up in your weight and in satellites. You weigh about 0.5% less at the equator, partly from the extra distance to the center and partly from the spin (the Eötvös/latitude effect, 54). Satellites feel it too: the bulge slowly drags their orbital planes around, something engineers must model. A spinning rock-and-fluid ball predicts every bit of this. A flat disc predicts none of it.
• 6“Sphere vs spheroid vs geoid vs pear” is a ~0.01% argument, and every version is round. The geoid (the mean-sea-level equipotential) departs from the ellipsoid by only about +85 m to −106 m (the Indian Ocean geoid low), a spread under 200 m on a 6,371 km body. The famous “pear shape” is smaller still: O’Keefe’s third-degree term from Vanguard 1 (1958) amounts to tens of meters between the poles, ~1,000× less than the flattening. The refinements quoted as “disagreement” are the gap between a ball and a more carefully measured ball, never between round and flat.

Sources: the equatorial bulge, the numbers & Chimborazo [399]; Newton’s prediction & the French Geodesic Missions [400] · geoid undulations & the “pear” term [446]. Builds on the “potato radius” of Entry 32 and the Eötvös effect of 54.

• 1The caption is the refutation, and it needs no outside authority. NASA’s Scientific Visualization Studio released this on 15 July 2026. The description reads that the geoid is an equipotential surface, that the height ranges from +85 m at Iceland to −106 m over southern India, and that in this visualization the geoid height is greatly exaggerated, by a factor of 10,000. [721] Anyone circulating the picture can read that by clicking the page it came from.
• 2What the numbers come to. A range of 191 meters on a radius of 6,371,000 meters is 3 parts in 100,000. The deepest point of the whole surface, the low over southern India, is 1 part in 60,000. If the planet were drawn at true scale on a screen 1,000 pixels across, the entire geoid variation would be one thirty-third of a pixel.
• 3Scaled to something you can hold. Shrink the Earth to a regulation billiard ball, 57.15 mm across. The full geoid range becomes 0.86 micrometers. The manufacturing tolerance for that ball is 0.127 mm. The bumps in the picture are around 148 times smaller than the tolerance the ball is already allowed, which is to say a real billiard ball is a far worse sphere than the Earth is.
• 4The lumps are not hills. They are gravity. The geoid is the shape an ocean at rest would take under gravity alone, not the shape of the ground. Its highs and lows come from denser and lighter rock in the mantle and crust pulling harder or less hard, so the surface is a map of mass distribution. [721] The image shows nothing about topography. Everest and the Mariana Trench, which are real relief and far larger than any geoid undulation, do not appear on it at all.
• 5Vertical exaggeration is ordinary practice and it is always declared. Sea-floor maps, cross sections of mountain ranges and weather profiles all stretch the vertical scale, because at true scale the feature of interest is a line one pixel high. The convention is to state the factor in the caption, which is what happened here. A picture with a declared exaggeration factor is not a picture that has been hidden.
• 6Where the data came from, and what that costs the claim. The model is GOCO06s, a satellite-only global gravity field built from over a billion observations gathered over 15 years by 19 satellites, including NASA’s GRACE and the European Space Agency’s GOCE. [721] The underlying model is a peer-reviewed paper in Earth System Science Data. [722] Nineteen satellites in sustained orbit, tracking each other to measure a gravity field, is not a measurement available above a plane.
• 7The same image has been recycled before. The GFZ research institute in Potsdam released a version in 2011 that has circulated ever since as the Potsdam Gravity Potato. Same physical quantity, different model, same exaggeration technique, same declared factor. A picture that resurfaces every few years as a revelation is a picture nobody was concealing.
• 8Three different surfaces, routinely blended together. The ellipsoid is the smooth mathematical figure used as a reference, flattened by 21 km between pole and equator by rotation Entry 33. The geoid is the gravity surface, departing from that ellipsoid by less than 110 m anywhere. The topographic surface is the actual ground, running from −11 km to +8.8 km. The claim takes a picture of the second, exaggerated ten thousandfold, and presents it as the shape of the third.

• 1No air, no buoyancy, and things still fall. In NASA’s Space Power Facility, the world’s largest vacuum chamber (about 30 × 37 m), a bowling ball and a feather released together drop in perfect step and land at the same moment once the ~30 tonnes of air are pumped out. Commander David Scott did the same with a hammer and a feather on the Moon in 1971. Density differences cannot explain that, because there is no fluid to be denser than.
• 2Buoyancy is downstream of gravity. A cork rises and a stone sinks in water because gravity pulls the whole fluid down, creating a pressure gradient that squeezes the least-dense things upward (Entry 26). Remove gravity and that pressure gradient vanishes, so nothing floats and nothing sinks. Buoyancy needs gravity first; it cannot stand in for it (Entry 17).
• 3Everything falls at the same rate. In a vacuum every object accelerates downward at ~9.81 m/s², no matter its mass or composition. This is the universality of free fall, tested to about one part in 10¹⁵. A “density” theory predicts no such universal acceleration. A real gravitational field does, and is measured to do so everywhere.
• 4A hammer and a feather, dropped on the Moon. At the end of the last Apollo 15 moonwalk on 2 August 1971, David Scott held out a 1.32 kg geological hammer and a 0.03 kg falcon feather and released them together, live on television. In the Moon’s vacuum there was no air to buoy the feather, and the two struck the dust at the same instant, ‘which proves that Mr. Galileo was correct.’ No air means no buoyancy, yet both still fell, pulled by gravity alone at the Moon’s ~1.62 m/s². Density decides nothing here; gravity does everything.
• 5Verified to one part in a quadrillion. The universality of free fall is not just a vacuum-chamber stunt. It is one of the most carefully tested facts in physics. The MICROSCOPE satellite (2016–2018) flew two test masses of different metals, platinum and titanium alloys, in the same orbit and measured the tiny force needed to keep them together. Its 2022 final result: they fall identically to about one part in 10¹⁵, still matched to within a tenth of a millimeter after ‘falling’ some 73 billion meters around the Earth. If density rather than gravity decided how things fall, a platinum and a titanium mass would have drifted apart long before that.

Sources: the NASA vacuum-chamber drop [116]; the Apollo 15 hammer–feather drop [117]; the equivalence principle to 1 part in 10¹⁵ [235]. Builds on the gravity of Entry 17 and the buoyancy of Entry 26. → Gravity data rows.

• 1Vapour vs droplets. Most atmospheric water is invisible vapour, individual gas molecules fully mixed and buoyant in air. A visible cloud is condensed water: countless separate droplets ~10–20 µm across, not a connected sheet.
• 2Tiny means slow. A droplet’s fall speed (Stokes’ law) scales with the square of its radius: ~0.3 cm/s at 10 µm, ~1 cm/s at 20 µm. Gravity pulls fully, but air drag, huge compared with the droplet’s mass, slows the fall to a crawl, its terminal velocity.
• 3Air does the rest. Updrafts of even 0.1–1 m/s dwarf those fall speeds and carry droplets upward, and clouds form in rising, cooling, buoyant air. The “millions of pounds” is about 0.5 g per cubic meter, ~0.04% of the air’s own mass, spread through ~10⁹ m³.
• 4Moist air is lighter. Surprisingly, humid air is less dense than dry air: a water molecule (18 g/mol) is lighter than the average air molecule (~29 g/mol), and by Avogadro’s law each water-vapour molecule that enters a parcel displaces a heavier nitrogen or oxygen one. So warm, moist air is buoyant and rises, which is right where clouds form. “Wet equals heavy” is backwards for vapour.
• 5When it does fall. Let droplets collide and grow toward 1–5 mm and the same physics flips: fall speed climbs to 4–9 m/s and the water leaves the sky as rain. Nothing was ever suspended in defiance of gravity. It was a balance that tipped.
• 6The “million tonnes” is spread very thin, and when it stops being thin, it rains. A fair-weather cumulus carries only about half a gram of liquid water per cubic meter, so its density sits a whisker above the clear air around it and its famous “weight” is smeared across a cubic kilometer. Let droplets collide and coalesce past ~0.1 mm and they fall out as rain. Precipitation is the cloud’s water losing the updraft battle and obeying gravity, right on cue.

Same lesson as Entry 17: "it floats" is buoyancy and drag operating within gravity, not evidence against it, just like a balloon, a boat, or dust in a sunbeam.

Sources: cloud microphysics & Stokes’ law [54] · the buoyancy of moist air [345] · collision–coalescence & rain formation [402]. → Water data rows.

• 1“Gas fills a vacuum” assumes no gravity. That rule is for a sealed box where gravity does not matter over the size of the box. Over a whole planet, gravity pulls every layer of air downward; pressure is just the weight of the air above you. There is nothing to “rush into” because gravity keeps pulling it back.
• 2Pressure is just weight, and it is read all the way up. At sea level the air column presses at about 101 kPa: roughly 10 tonnes bearing on every square meter, the weight of the air overhead. Barometers on balloons, airliners and mountaineers chart the smooth fall-off directly. Atop Everest the pressure is only about a third of sea level, with no jump and no wall anywhere.
• 3No edge, just an exponential fade. Air pressure halves about every 5.5 km of altitude (scale height ~8.5 km). It never hits a wall; it gets thinner and thinner until it blends into space. The “edge of space” (Kármán line, 100 km) is a chosen convention, not a surface.
• 4It is not perfectly sealed. The lightest gases, hydrogen and helium, slowly escape to space (Jeans escape). A true closed dome could not leak; a gravity-held atmosphere must, and ours does. That is the opposite of a sealed system.
• 5The ISS skims the top of it. Even at ~400 km the air is not quite nothing: the ISS feels enough drag to lose altitude steadily and must fire engines to reboost. There is no shelf where atmosphere stops and space begins. It keeps thinning, which is why low satellites slowly decay and re-enter.
• 6Every world does it. Mars, Titan, Venus, even the Sun’s million-degree corona all hold gas against the surrounding vacuum by gravity alone, thick or thin, with no container anywhere. Gravity bordering vacuum is the normal state of things, not a paradox needing a dome.
• 7The air climbs through named layers, then fades to nothing. From the ground up the air passes through the troposphere (weather), the stratosphere (the ozone layer), the mesosphere (where meteors burn), the thermosphere (aurora and the ISS), and the exosphere, where atoms grow so sparse the fastest ones escape to space. Beyond it the hydrogen thins into the geocorona, which SOHO measured reaching about 630,000 km, past the Moon. Each layer is thinner than the last, with no wall and no edge, only air fading into vacuum. [508]

Sources: atmospheric structure, scale height & escape [58] · pressure as the weight of air, and the ISS in the thin upper atmosphere [346]. → Earth/atmosphere data rows.

• 1Air is matter, and it has weight. A cubic meter of sea-level air masses about 1.2 kg. Stack kilometers of it and its own weight presses down: ~101 kPa at sea level, roughly 10 tonnes bearing on every square meter, balanced from all sides so you don't feel it. The ideal gas law, PV = nRT, ties that pressure to volume, temperature and amount of gas, and it's the workhorse behind weather, scuba tables and engines alike.
• 2The pressure gradient is gravity made visible. Pressure falls off exponentially with altitude. That is the barometric formula, with a scale height of ~8.5 km (Entry 37). Your ears pop in a lift or a climbing plane because the outside pressure drops as you rise; water boils cooler on a mountain; a sealed bag of chips puffs up at altitude. None of that happens unless gravity is pulling the gas downward into a deep, dense layer at the bottom.
• 3Heavy gases really do pool, sometimes lethally. Carbon dioxide, denser than air, collects in mine shafts, cellars and volcanic hollows; radon sinks into basements. In 1986 Lake Nyos belched a vast CO₂ cloud that flowed downhill like an invisible flood and suffocated about 1,700 people in the valleys below. Gravity sorts gases by density whenever mixing is slow. The open atmosphere stays blended only because thermal motion constantly re-stirs it.
• 4Gravity inside your chest. Stand up and gravity pulls blood and tugs lung tissue downward, so the base of each lung receives both more blood (perfusion) and more air (ventilation) than the apex. These are the classic "West zones" of respiratory physiology. It's why oxygenation shifts with posture and why the lower (dependent) lung is the better-perfused one. Every breath you take is a small, repeatable gravity experiment.
• 5Carry a barometer uphill and watch. In 1648 Blaise Pascal had his brother-in-law Florin Périer carry a mercury barometer up the Puy de Dôme. At the foot the column stood at 711 mm; about 1,000 m higher it had dropped to ~627 mm, a fall of ~85 mm (~12%), while an identical barometer left at the base never budged. Less air overhead means less weight bearing down. The test weighed the atmosphere directly, three centuries before anyone thought to dispute that air is held by gravity.
• 6Spin a gas and it sorts itself by weight. A gas centrifuge whirls uranium hexafluoride (UF6) at 50,000–70,000 rpm, producing an effective field of order 10⁵–10⁶ times gravity at the rim. The heavier U-238 molecules are flung toward the wall while the lighter U-235 collect near the axis, so the gas separates by molecular mass. It is the industrial backbone of uranium enrichment, and it works only because a gas in a gravitational field (real or spun-up) stratifies by weight, just as the still atmosphere does.
• 7The gas laws describe gases; they do not repeal gravity. A common claim holds that because Boyle’s, Charles’s and Avogadro’s laws, and the ideal gas law that combines them, contain no gravity term, gases must be unaffected by gravity, so the air needs a container. Those are equations of state. They relate a gas’s pressure, volume, temperature and amount to one another; they do not say where the pressure comes from. In the atmosphere the pressure at any height is the weight of the air stacked above it, and that weight is gravity pulling on the air’s mass. Put the ideal gas law together with that weight and you get the barometric formula, whose scale height equals kT divided by mg. The g in that denominator is gravity, and it is why pressure falls off smoothly with altitude and why denser gases settle lower. A gas with no weight would show a flat pressure profile, yet every barometer and every set of ears popping on a climb records the drop. 37

Sources: ideal gas law & kinetic theory [64] · barometric pressure & scale height [58] · pulmonary ventilation/perfusion gradient, West zones [65] · Pascal’s 1648 Puy de Dôme barometer test [155] · gas-centrifuge isotope separation [156]. See also the atmosphere/dome entry (Entry 37) and buoyancy (Entry 26). → gas & pressure data rows

• 1Gravity is the container. There is no sharp top to the air. Pressure falls off exponentially with a scale height of ~8.5 km; the atmosphere fades into near-vacuum. No surface, and no dome, is required to hold a gas in a gravity well.
• 2Light gases do leave, by design. Escape needs a molecule moving near 11.2 km/s. At the ~1000 K exobase, hydrogen averages ~5 km/s and its fastest tail escapes (Jeans escape); helium too. Earth sheds ~3 kg/s of hydrogen and ~50 g/s of helium, about 90 t/day in total.
• 3The heavy gases stay. Escape only matters when escape velocity is under about six times a gas’s typical speed, which on Earth is true only for H and He. Nitrogen and oxygen move ~1 km/s and are bound for all practical purposes forever. That is why we keep an N₂/O₂ atmosphere while H and He slowly drain away.
• 4Replenished and ancient. Helium is resupplied by radioactive decay in the crust (and pools in natural gas); oxygen is maintained by photosynthesis; volcanic outgassing adds gas. At ~93,000 t of hydrogen lost per year against 5×10¹⁸ kg of air, the atmosphere is effectively permanent on human and geological timescales.
• 5Which worlds keep an atmosphere, and which don’t. A body holds a gas only when its escape speed clears the gas’s thermal speed by a wide margin (the Jeans criterion). The Moon (escape velocity 2.38 km/s) and Mercury (4.25 km/s) are almost airless; Earth (11.2 km/s) keeps nitrogen and oxygen but slowly loses hydrogen and helium; frigid Titan, just 2.64 km/s yet only ~94 K, holds a 1.5-bar nitrogen sky because the cold keeps its molecules slow; and Jupiter (59.5 km/s) retains even hydrogen, a whole world of gas. Atmospheres sit wherever gravity is strong enough for the local temperature, never behind a wall.
• 6The Earth is not a sealed terrarium; a terrarium is a small model of the Earth. A popular claim holds that a closed glass terrarium, which runs its own water cycle under a lid, shows that the Earth is a sealed world under a dome. It shows the reverse. A terrarium is a closed system, and its water condenses on the cold glass, which is what lets the cycle close. The Earth is an open system. It takes in sunlight and radiates the same amount of heat back out to space as infrared, a balance satellites measure directly, and the second law of thermodynamics requires it. A true sealed dome would trap the Sun’s energy with nowhere for the heat to go, and the surface would cook. So the terrarium is a scale model of the Earth’s water cycle, not proof of a dome. The role the cold glass plays in the jar is the role the cold of space plays for the planet, and the planet needs no glass. 40

Sources: atmospheric escape & Jeans escape [142] · measured leak ~90 t/day, ESA Cluster [143] · atmosphere mass & scale height [1] · planetary escape velocities & Titan’s cold, thick atmosphere [157]. → Earth & retention data rows

• 1Isolated, closed, open: the words matter. The second law says the entropy of an isolated system never falls. An isolated system trades neither energy nor matter with its surroundings; a closed system trades energy but not matter; an open system trades both. Earth trades energy every second, sunlight in and infrared out, so it is not isolated, and the ‘closed-system’ premise is false. The same point settles the atmosphere in Entry 39.
• 2We measure the energy crossing the boundary. Satellites weigh Earth’s energy budget directly. About 340 W/m² of sunlight reaches the top of the atmosphere; ~30% is reflected and ~240 W/m² is absorbed; Earth then radiates ~240 W/m² of infrared back to space. The two nearly balance, with a tiny imbalance of roughly 1 W/m², the fingerprint of present-day warming. A truly sealed box could not show that throughput at all.
• 3Local order is paid for by exporting entropy. This is how a fridge, a snowflake or a living cell can grow more ordered without breaking any law. Earth absorbs a few high-energy photons from the ~5,800 K Sun and re-emits many low-energy photons at its ~255 K radiating temperature, very roughly twenty photons out for every one in. That swap is a large increase in entropy, dumped into the cold sky. Life, weather and ocean currents all ride that one-way flow of energy.
• 4Heat does flow hot-to-cold, straight through vacuum. Thermal radiation is electromagnetic, so unlike conduction or convection it needs no medium: that is why a fire warms your face across a room and the Sun warms a spacewalker in vacuum. At every step the heat runs downhill in temperature, from the ~5,800 K Sun to the ~288 K Earth to the ~2.7 K deep sky. Nothing flows from cold to hot, and no dome is needed to make the books balance.
• 5A real dome would break the second law, not save it. Seal the Earth inside an actual firmament and the thermodynamics turns against the flat model: with sunlight pouring in and no cold sky to radiate to, the trapped system would heat without limit toward equilibrium, washing out the day–night swing and any stable climate. The open Earth, steadily losing heat to a 2.7 K sky, is what keeps temperatures bounded and the second law satisfied.

Sources: Earth’s energy budget [240] · isolated/closed/open systems & the second law [241]. Extends the open-system argument of Entry 39; the Sun’s own power source is 101. → energy-budget data rows

• 1It is a Michelson interferometer, grown enormous. LIGO splits a laser down two 4 km arms set at right angles, bounces it between mirrors (Fabry–Perot cavities) to fold the path out to ~1,120 km, and recombines the beams. A passing gravitational wave stretches one arm and squeezes the other by about one part in 10²¹ of their length; the recombined light shifts from dark to bright. Same principle as Entry 48, only sensitive enough to feel spacetime itself flex.
• 2GW150914, 14 September 2015. Two black holes of ~36 and ~29 solar masses spiralled together ~1.3 billion light-years away, merging into a ~62-solar-mass hole and radiating about three Suns’ worth of mass-energy as gravitational waves in a fifth of a second. The chirp recorded at Livingston, Louisiana, and Hanford, Washington, matched each other and the relativistic prediction. The discovery won the 2017 Nobel Prize; hundreds of further mergers have been logged since.
• 3The 7-millisecond clue. The wave reached Livingston 7 ms before Hanford, the time light needs to cross the 3,002 km between them. Two widely separated machines agreeing, with a delay set by the speed of light across a curved Earth, is what a real, sky-sourced signal looks like and what local noise or wishful thinking cannot fake.
• 4Then light and waves arrived together: GW170817. On 17 August 2017 LIGO and Virgo caught two neutron stars spiralling together about 130 million light-years away; 1.7 seconds later NASA’s Fermi telescope saw a gamma-ray burst from the same patch of sky, and within hours observatories worldwide found the glowing “kilonova” in galaxy NGC 4993, later tracked in X-rays and radio. Gravitational waves and light from one event, arriving all but together after a 130-million-year journey, is proof the waves are real astrophysical signals moving at light speed, not instrument noise.
• 5A growing global network. Virgo (Italy) and KAGRA (Japan) now observe alongside the two LIGO sites; three or more detectors triangulate a source’s position from the millisecond differences in arrival time. The catalogue has grown from that first 2015 chirp to hundreds of mergers. Cross-checked across continents, this is routine astronomy now.

Built straight through the curve. The arms are so long that Earth’s curvature was a construction problem. Over each 4 km arm a straight line in vacuum departs from the ground by about 1.25 m, so the beam tubes were aligned to the light’s straight path with GPS and the WGS-84 ellipsoid, not to local level. Without that, the laser leaving the corner station would strike about a meter above the far mirror instead of hitting it. A flat Earth needs no such correction. LIGO measured the curve by having to engineer around it [476].

Sources: LIGO & the first gravitational-wave detection [66] · the GW170817 multi-messenger event [343] · the LIGO–Virgo–KAGRA network & event catalogue [344] · the 4 km arms engineered around Earth’s curvature [476]. Built on the interferometer of Entry 48; confirms the gravity of Entry 17. → LIGO data rows.

• 1Tides come from a difference in pull. Gravity weakens with distance, so the ocean facing the Moon feels more pull than the solid Earth's center, and the far ocean feels less. That difference stretches the water into two bulges. The tidal (stretching) force falls off as 1/distance³, which is why the closer, smaller Moon out-tides the far larger Sun roughly 2:1. You need a real distance and a real diameter for this math to work.
• 2The clock matches the sky. As Earth rotates under those two bulges, most places pass through both each day. You get two highs and two lows, arriving ~50 minutes later daily, tracking the Moon's own ~50-minute daily slippage. When Sun and Moon line up (new/full Moon) the bulges add for big "spring" tides; at right angles they partly cancel for "neap" tides. The pattern is the Moon's signature, written in water.
• 3Same force, bigger stage. The same tidal mechanism, differential gravity across a body, is what whips material into accretion disks, tears comets apart, and powers the volcanoes of Jupiter's moon Io. It's gravity (Entry 17) seen through a magnifying glass, and it has no working substitute in any flat-Earth account.
• 4Spring and neap tides show the Sun joining in. If something other than gravity drove the tides, they wouldn’t keep time with the Sun and Moon together, but they do. The Sun raises its own tide, about 46% as strong as the Moon’s (tidal pull falls off as 1/r³, so the Moon’s nearness beats the Sun’s mass). When Sun, Moon and Earth line up at new and full moon the two tides add and the range is largest. These are spring tides. When they sit at right angles at the quarter moons the Sun partly cancels the Moon and the range shrinks. These are neap tides. The cycle repeats every ~2 weeks and is printed in tide tables years ahead. Two pullers, one schedule, just as differential gravity predicts.
• 5The tide clock runs on Moon-time. Tides keep a 24-hour-50-minute beat, the lunar day, not the 24-hour solar day. High water arrives about 50 minutes later each day, the same 50 minutes by which the Moon itself rises later daily (Earth has to spin a little extra to ‘catch up’ to the orbiting Moon). Two highs and two lows recur every 24 h 50 m, about 12 h 25 m apart. If pressure, the Sun, or a rocking disc set the tides they would lock to the solar day. Instead they track the Moon’s transit just as differential lunar gravity demands. The Sun’s tidal pull is only ~0.46 times the Moon’s, with every planet’s contribution negligible.
• 6The textbook bulge is only the first approximation. The simple two-bulge picture, called equilibrium theory, predicts two equal high tides a day at the same clock time everywhere. Real oceans don’t oblige. The equilibrium bulge is tiny (well under a meter), and continents block it from sweeping freely around a planet that is spinning out from under it. What really happens is a forced, resonant sloshing. The tide rotates around fixed nodes called amphidromic points, where the range falls to almost zero, counter-clockwise in the Northern Hemisphere, with the Coriolis effect steering the flow. Local range is set by basin shape, depth and resonance, which is why it runs from about 10 cm in the Mediterranean to nearly 17 m in the Bay of Fundy, whose ~13-hour natural period almost matches the 12.4-hour lunar tide. Honest tide tables fold in roughly 400 harmonic terms, and every one still presupposes a real Moon orbiting a spinning, round Earth. No flat model produces an amphidromic system.
• 7The lag has a price, and it has been paid in Apollo mirrors. Earth spins faster than the Moon orbits, so friction drags the tidal bulge slightly ahead of the Moon. The Moon’s pull on that offset bulge brakes Earth’s spin and hands the lost angular momentum to the Moon’s orbit. Both effects are measured, not assumed. The day is lengthening by roughly 2 milliseconds per century, and the Moon is receding at about 3.8 cm a year (3.83 ± 0.01 cm/yr), clocked to millimeter precision by bouncing lasers off the retroreflectors Apollo 11, 14 and 15 left on the surface (121). The rocks agree. Tidal rhythmites and fossil-coral growth bands record about 400 days in a year and a ~21.9-hour day around 620 million years ago. A faster past spin, a closer Moon, a steadily lengthening day. That is what tides on a spinning round Earth must produce, and nothing a stationary disc can mimic.
• 8Even the Great Lakes have a tide. The Moon’s pull reaches every body of water, not only the oceans. A true tide rises and falls twice a day on the Great Lakes, though the largest of them is less than five centimeters, small enough that wind and weather usually hide it. A pull that reaches even a closed inland lake is a pull that falls off with distance, which a single uniform upward push cannot produce. [549]

Sources: tides & the tidal (differential gravity) force [73] · spring & neap tides and the Sun’s contribution [212]; the lunar (tidal) day [236]; dynamic theory, amphidromic systems & tidal range [386]; lunar recession by laser ranging [387]; the lengthening day in rock & coral [388]. Built on the gravity of Entry 17. → tide & gravity data rows

• 1The locks are not the argument, and saying they are loses in one move. The canal lifts ships 85 feet up to Gatun Lake and lowers them 85 feet down the far side. That is a water elevator over a range of hills, and the trade is even. Nothing about it demonstrates curvature, and anyone claiming the locks lift ships “over the curve” has handed away an argument that did not need them.
• 2The Moon cannot tell the two ends apart. The lunar distance is about 239,000 miles. The canal is 42. The difference in the Moon’s pull between one end and the other is far too small to produce any difference you could measure with a ruler, let alone a factor of fifteen. Whatever is driving the difference is on this end of the problem.
• 3The numbers, and where they come from. The Caribbean end at Cristobal runs a mean range under a foot; sources of the period put it near 0.6 ft. [723] The Pacific end at Balboa runs about 12.5 ft on an average day. [724] On many days the lowest water on the Caribbean side is still higher than the highest water on the Pacific side, and high tide reaches the Caribbean end roughly three hours before it reaches the Pacific end.
• 4What actually sets a tide is the shape of the basin. Rock a shallow baking pan and a deep bathtub with exactly the same motion at exactly the same moment and the water behaves completely differently in each, because each container has its own natural sloshing period. The Caribbean is a small, shallow basin closed off by the Antillean island chain, with an oscillation period near 24 hours, so the daily tide dominates there. The Pacific end sits at the far end of a much larger twice-daily oscillating system. [725]
• 5Which corrects the picture most people carry. A tide is not a bulge of water following the Moon around the planet like a dog on a lead. It is water sloshing inside ocean basins, driven by the Moon and the Sun but shaped by the size, depth and coastline of the basin it is trapped in. Two basins that share a Moon do not share a tide.
• 6The tide gauges at both ends have been running for over 110 years. Cristobal and Balboa are two of the longest continuous sea level records anywhere. A 2024 analysis of both found that the observed nodal modulations of the major tidal constituents are generally consistent with equilibrium tidal theory. [726] That is a prediction from lunar orbital geometry, made in advance, turning up in a century of shipping-canal paperwork.
• 7Where the 18.6 years comes from. The plane of the Moon’s orbit is tilted relative to the Earth’s orbit, and the line where the two planes cross slowly rotates all the way around, taking 18.61 years per lap. That changes how far north and south the Moon swings each month. For that wobble to move water in Panama at all, it has to matter how far north or south the Moon is relative to where the gauge is standing, which is to say it has to matter what latitude the gauge is at.
• 8The historical detail, stated with its expiry date. Across most of the record the Caribbean end shifted style, sometimes one high tide a day and sometimes two, tracking that 18.61-year cycle. The same 2024 analysis reports that this regime shift has not recurred since 1997, because of a long-term decline in the twice-daily M2 component. [726] Anyone checking a current tide table will not find it. It is a fingerprint in a century-long dataset, not a present-day prediction, and it should never be offered as one.
• 9The record shows other secular change too. The M4 and MS4 amplitudes at Cristobal have halved over the past century, tidal asymmetries there are significantly weakened, and Balboa’s tides are now almost symmetric. [726] These are measurements of a system that changes, published by people with no interest in this argument, which is what a real dataset looks like.

• 1Start with the part that is true. Pour water into anything and it settles. It runs downhill until there is no downhill left. That is the whole mechanism and it is not in question here.
• 2The argument rests on one word, and the word does not mean what it is being read to mean. “Level” is heard as “flat.” What it actually means is that no direction is downhill any more. Those are different statements, and only the first one implies a plane.
• 3So which way is downhill? Whatever direction things fall when you let go of them. There is no need to agree on what causes it to agree on where it points, and it can be checked with a plumb line in any back garden.
• 4On flat ground every “down” arrow is parallel. Water with no downhill left then settles into a flat sheet, because a flat sheet is the only surface square to every one of those parallel arrows.
• 5On a ball, “down” points toward the center. Down in Panama and down in Ohio are then at slightly different angles, and the arrows tilt with respect to each other by about one degree every 111 km, which is the same span that defines one degree of latitude. Water settles until no patch of surface has any downhill left, and a surface everywhere square to arrows converging on a center is a curved shell. “Water finds its level” is completely true and predicts a curve.
• 6Nobody notices in a glass because of the scale. The water in a drinking glass is curved too. Across three inches the deviation from flat is far below anything the eye or a straight edge could resolve, so the surface looks flat and the word “flat” gets attached to it.
• 7Digging a canal forces the question, because you need a definition of zero. Every depth, every lock sill, every dredging contract is measured from some agreed height. If sea level were one flat plane, one datum would serve the whole project.
• 8They could not use one. The canal’s published documents list a separate reference level for the Atlantic end and for the Pacific end, because the water does not match. Mean sea level on the Pacific side sits about eight inches, roughly 20 cm, above the Caribbean side. [723] [727]
• 9The Canal Zone is the only place on Earth this could be checked so directly. A 1926 survey paper notes that the Canal Zone is the only place in the world where bench marks on the mean sea level datums of different oceans can be interconnected by a short line of precise levels, and that the general impression that the surface of all oceans is one uniform level is, of course, erroneous. [723] That was written in 1926, by surveyors with no view on this argument at all.
• 10The eight inches has ordinary causes. Pacific water at that longitude is warmer and slightly less salty, so it is less dense and the same weight of it stands a little taller. Prevailing winds and currents pile water against coastlines. [727] None of that is mysterious, and none of it is needed for the point: two connected bodies of the same ocean, thirty-eight nautical miles apart, sit at different heights and both are called sea level.
• 11What this shows, and what it does not. It disproves that sea level is a single flat plane. It does not by itself prove the Earth is round, and it should never be presented as though it does. Curvature is measured elsewhere Entry 12. This is the surveyor’s receipt on the meaning of one word.

• 1Steady motion is invisible to the senses. Your inner ear (the vestibular system) senses acceleration, not velocity. The only acceleration the spin adds is a steady centripetal ~0.034 m/s² at the equator, about 0.0035 g. The otolith organs do not reliably register linear acceleration until roughly 0.06–0.1 m/s², so the spin is below threshold. Being constant, it offers nothing to detect anyway.
• 2But we DO feel acceleration when it is real, like an earthquake. The body is not numb to acceleration. A felt quake delivers abrupt, oscillating ground accelerations from a few percent of g up to ~0.25 g at Mercalli VIII (and more in severe shaking), instantly noticeable. Turbulence, hard braking and a dropping elevator are felt for the same reason. The spin goes unfelt because it is smooth and sub-threshold, not because we cannot feel motion.
• 3The oceans stay put because gravity wins ~289:1. That outward 0.034 m/s² is about 1/289 of gravity’s 9.8 m/s². Its only real effect is that you weigh ~0.3% less at the equator than at the poles, which is measurable, and the opposite of flying off. To balance gravity at the equator the Earth would have to spin about 17× faster, a day of roughly 84 minutes.
• 4The spin did leave a mark, the bulge. A spinning, self-gravitating planet settles into an oblate spheroid, and Earth is one. Its equatorial radius (6,378 km) exceeds its polar radius (6,357 km) by about 21 km, so the planet is roughly 43 km wider across the equator than pole-to-pole, about 0.3% flattening, mapped by satellite geodesy. Faster spinners bulge far more (Saturn by nearly a tenth). The rotation shows up as the planet’s shape, not as a wind.
• 5“A 1,000 mph wind” mistakes what wind is. Wind is air moving relative to the ground; the bulk atmosphere co-rotates with the surface, so rotation produces no headwind, just as you feel no 900 km/h gale walking the aisle of a cruising jet. The speed is real, but you, the air and everything around you all share it.
• 6How the inner ear works, and why the spin is beneath it. The vestibular system has two kinds of sensor, both accelerometers. Three semicircular canals in each ear register angular acceleration (rotation), and the otolith organs, the utricle and saccule, register linear acceleration and the pull of gravity. Neither responds to steady velocity. The measured thresholds are blunt. Humans can just detect a rotation of about 0.7°/s and a linear acceleration of a few hundredths of a meter per second squared (roughly 0.02–0.3 m/s²). Now hold the spin against them. Earth turns at a constant 0.0042°/s (15°/hour) with zero angular acceleration, so the canals receive nothing, and even that bare rate is more than 150× below the rotation threshold. Its only linear signature is the steady 0.034 m/s² equatorial centripetal term, about 0.3% of gravity, constant, and quietly folded into the body’s sense of ‘down’ (it just trims your weight a touch, Entry 55). The claim that we ought to “feel” the spin assumes an organ that reads velocity; biology built no such organ. [430][431]

Sources: vestibular acceleration thresholds [97]; earthquake ground acceleration [98]; Earth’s equatorial bulge [342]. The spin itself is measured directly in Entry 47 and Entry 52. → Rotation data rows.

• 1The hover trick fails on inertia. A helicopter lifting off keeps the eastward velocity it shared with the ground, so while hovering it continues east at the same rate as the land below and nothing slides past. There is no force that strips away that ~1,000 mph; stopping relative to the spinning Earth would take enormous, continuous thrust, not just rising.
• 2The “1,000 mph runway” is a non-problem. The landing aircraft is also moving east at ~1,000 mph; relative to the runway only its airspeed (a couple of hundred mph) counts. Catching a tossed peanut on a moving train works for the same reason. Everything in the carriage shares the train’s velocity. Steady motion is undetectable from inside (Entry 45).
• 3The air comes along. Gravity holds the atmosphere down and friction drags it into rotation with the surface, so it co-rotates (the small leftover differences are ordinary winds). We do detect the spin, not by being flung off but through subtle steady effects: deflected winds and shells (Coriolis, Entry 52), a turning pendulum (Foucault, Entry 47) and ring-laser gyros that sense it directly (Entry 49).
• 4Falling bodies really do shift, by the predicted amount. A weight dropped down a deep shaft lands a fraction east of straight down. The top of the shaft, farther from Earth’s axis, moves east a touch faster than the bottom. Reich measured about 8.5 mm of eastward drift in an 1833 mineshaft, against the 8.8 mm Gauss and Laplace had calculated; from a 100-m tower at the equator it is about 3.3 cm. The spin does not fling you sideways. It leaves this tiny, calculable fingerprint.
• 5The only thing that “hovers” over one spot is a satellite, and it must match the spin. A geostationary satellite stays fixed above one point only by orbiting eastward at Earth’s rotation rate, 35,786 km up (127). It cannot just stop and let the ground slide beneath it; nothing can. That a satellite has to keep pace to appear stationary is the very rule the helicopter argument forgets.
• 6Take the claim seriously and run its arithmetic. It falsifies itself. Suppose an aircraft really did leave the spinning frame the moment its wheels lifted. New York to London is 5,554 km, and the North Atlantic tracks sit near 50°N, where the ground runs east at 1,670 × cos 50° = about 1,073 km/h. An airliner does about 900 km/h. So the claim predicts, in its own numbers: flying east, the ground outruns you by 173 km/h and London recedes forever. Flying west, London rushes at you at 1,973 km/h and you land in 2.8 hours. Those are not our numbers. They are the claim’s numbers, followed to the end.
• 7Now look at the timetable. It is backwards, and it is small. Eastbound New York to London averages about 6 h 13 m. Westbound London to New York takes about 8 hours. So the eastbound leg, the one the claim says is impossible, is the faster of the two. And the gap is roughly an hour, not an eternity. Solve the two ground speeds for an airspeed and a wind and you get an aircraft doing about 794 km/h through air that is moving about 100 km/h eastward. That is a jet stream. It is not a planet. The claim is not off by a detail. It has the sign backwards and the magnitude wrong by a factor of ten. [24]
• 8Two tests the claim has to pass, and fails. Earth’s surface speed goes as cos(latitude): 1,670 km/h at the equator, 835 at 60°, near zero at the poles. So if the spin caused the east-west flight-time gap, the gap would be biggest at the equator and vanish toward the poles. It does the opposite. The gap is largest across the mid-latitudes, 30° to 60°, and is nearly absent on flights between tropical cities. That is where the jet streams live, and it is the reverse of what the claim requires. Second test: the Earth spins at the same rate in January and in July. The jet stream does not, because it is driven by the temperature difference between pole and tropics, which is sharpest in winter. So the claim predicts a fixed gap all year, and the timetables show a gap that swells in winter and shrinks in summer. Airlines publish different schedules for the two seasons. The effect tracks the weather, not the rotation.
• 9The number that settles it: Mach 0.86. On 8 February 2020, British Airways flight 112 flew New York to London, eastbound, in 4 hours 56 minutes. It is the fastest subsonic transatlantic crossing on record, and it went in the direction the claim calls impossible. Two numbers came off that aircraft, and together they close the case. Its ground speed reached 825 mph. Its Mach number was 0.86. Now do the arithmetic yourself. At 35,000 feet the air is about −54 °C, where the speed of sound is 663 mph. Mach 0.86 therefore means the aircraft was moving through the air around it at 0.86 × 663 = 570 mph, which is an utterly ordinary cruising speed. It covered ground at 825. The difference, 255 mph, is the air itself, moving. The airplane never went fast. The wind did. And that is the whole answer to the claim in one line: an aircraft does not race the ground and lose. It swims in a fluid, and the fluid came along with the planet. [626]

Sources: inertial frames & Galilean relativity [112] · the eastward deflection of falling bodies [341] · the jet stream and its effect on east-west flight times [24] · the BA112 record crossing, ground speed and Mach number [626]. See also flight times and great-circle routes (58). Builds on why a steady spin is unfelt (Entry 45) and how it is detected (Entry 47–Entry 49); a satellite that “hovers” must match the spin (127). → Rotation data rows.

• 1Foucault's pendulum (1851). A long pendulum's swing plane slowly rotates as the Earth turns beneath it, first shown publicly by Léon Foucault under the Panthéon dome. No horizontal force turns the swing; the floor (Earth) is rotating.
• 2The latitude law. Precession = 15.04°/hr × sin(latitude): a full turn per sidereal day at the poles, none at the equator, in-between elsewhere (Paris ≈31.8 hr). That sin(latitude) signature is spherical geometry. A flat spinning disc wouldn't produce it. Try the model below.
• 3Coriolis. In a rotating frame, moving objects deflect, right in the Northern Hemisphere, left in the Southern (f = 2Ω sin φ). Hence cyclones spin counter-clockwise north of the equator and clockwise south of it; trade winds, ocean gyres, and long-range artillery all show it. The hemispheric reversal is a rotating-globe fingerprint.
• 4Not your sink. Coriolis is far too weak to decide which way a basin drains (that's set by basin shape and residual motion); it governs large, long-lived systems. The genuine, measured signatures are weather systems, the Eötvös effect (you weigh slightly less moving east), and ring-laser gyros sensing 15°/hr (56).
• 5It isn’t drafts, the push, or the building. The precession can’t be blamed on air currents, the launch, or vibration, because its rate obeys a fixed law. The swing plane turns 360° × sin(latitude) per sidereal day, about 11.25° per hour at Paris, a full circle in ~32 hours, zero at the equator, once a day at the poles. No draft or building sway produces a precession tuned to your latitude and locked to the star day. Foucault released the bob by burning a thread so it got no sideways shove, and the rate is independent of the bob’s mass and amplitude. There is one real pitfall. A sloppy elliptical launch or an asymmetric suspension adds a spurious drift (Airy precession), which is why careful installations control for it, after which the result matches the latitude law that ring-laser gyroscopes confirm independently (Entry 49).

Interactive — precession by latitude

Drag the latitude. The swing plane rotates at 15.04°/hr × sin(latitude), clockwise in the north, counter-clockwise in the south, frozen on the equator. (Animation sped up to be visible.)

Sources: Foucault pendulum [45] · Coriolis effect [46] · rotation rate [6]. · precession ∝ sin(latitude), tied to the sidereal day; elliptical/Airy bias controlled for [165]. → Rotation data rows

• 11887: the experiment, in detail. In the basement of Western Reserve’s Adelbert dormitory in Cleveland, Michelson and Edward Morley floated a heavy sandstone slab on a pool of mercury so it could turn without vibration, and folded a sodium-light beam (λ 589 nanometers (nm)) back and forth to an ~11 m path in each arm. Earth’s ~30 km/s orbital motion through a fixed luminiferous aether should have shifted the fringes by about 0.4; the observed shift was under 0.02, from an apparatus sensitive to about one part in ten billion. The most famous null result in physics (his 1881 Potsdam version was already null).
• 2What the null does and does not mean. It killed the aether and any “aether wind”; it did not show a stationary Earth. Einstein’s 1905 relativity explains it. No experiment can detect uniform absolute motion, because there is no preferred frame (Lorentz invariance). “No aether wind” therefore means “absolute velocity is undetectable,” fully consistent with Earth orbiting at 30 km/s. The flat/geocentric reading stops at the null; the physics did not. The same tactic, quoting a result without the context that resolves it, runs through the documents catalogued in 146.
• 3The aether-drag escape hatch fails. One could try to save a motionless Earth by saying it drags the aether along, so there is no wind. But stellar aberration, the astronomer James Bradley’s 1729 ~20.5″ annual tilt of every star, requires the telescope to move through the light, which a fully dragged aether forbids; and the 1925 rotation result below is impossible under complete drag, as Michelson himself noted in 1904. The loophole is shut from two sides (55).
• 41925: Michelson–Gale–Pearson, light measures the spin. On a field at Clearing, Illinois, the team laid a 2010 × 1113 ft (612 × 339 m) rectangle of 12-inch evacuated pipe and sent carbon-arc light both ways around it. Rotation, unlike uniform motion, is absolute, and it produced a Sagnac shift Δ = 4AΩ sin φ ÷ λc. At latitude 41.8°N they predicted 0.236 ± 0.002 and observed 0.230 ± 0.005. That is Earth’s 15°/hour spin, read off in light. A North-Pole-centered flat-Earth map predicts ≈0.266, well outside that range.
• 5The Sagnac effect and its sin(latitude) fingerprint. Georges Sagnac (1913) showed a rotating ring interferometer shifts its fringes in proportion to enclosed area × rotation rate. The Earth-rotation signal scales as sin(latitude): zero at the equator, greatest at the poles. That latitude dependence is sphere-specific. Measure it at several latitudes and only a globe fits; a flat disc would give one wrong number everywhere (Entry 49).
• 6Ring-laser gyroscopes, the spin continuously. The same Sagnac principle, now with lasers. The “G” ring-laser gyroscope at Wettzell, Germany, a 4 m square cavity on a Zerodur block, buried for stability, tracks Earth’s rotation around the clock, pinning the length of the day to under a millisecond and resolving solid-Earth tides, polar motion, even the axis’s precession and nutation. Small ring-laser gyros in every airliner’s inertial navigation sense the 15°/hour rotation to find true north (Entry 50). Filmed for a 2018 documentary, a flat-Earth team bought a ~$20,000 ring-laser gyro, measured a 15°/hour drift, just Earth’s rotation, then tried to shield it away; the signal stayed.
• 6a1958: the man who invented the maser pointed two of them at each other and looked again. Charles Townes had just built the first maser. One of the earliest things he did with it was rerun Michelson and Morley. He and Cedarholm, Bland and Havens mounted two ammonia-beam masers with their molecular beams pointing opposite ways on a rack that turned about a vertical axis, and read the frequency difference as the whole assembly was rotated through 180 degrees, over and over, through 1958 and 1959. [714] Null again. He went on to repeat it with helium-neon lasers once lasers existed, which they did not in 1958: he and Schawlow only published the proposal for one that same year.
• 6bThat is the shape of the whole record, and it is the answer to “Michelson proved the Earth is still”. This test has been rebuilt every time somebody invented a better instrument, usually by the person who invented it. Michelson’s interferometer in 1887. Townes’s maser in 1958. His laser in the 1960s. Cryogenic optical resonators in 2003. A turntable-mounted cavity in 2009. Cryogenic sapphire in 2015. Each generation went in with a sharper tool and a Nobel waiting for whoever broke it, and Townes already had his. A result defended by its friends looks like agreement. A result attacked for 140 years by people with everything to gain from breaking it looks like this.
• 7The modern Michelson–Morley, still null to one part in 10⁻¹⁸. Since Brillet and Hall (1979), laser and optical-cavity versions compare the speed of light in perpendicular directions as the apparatus and the Earth turn. The best of them (rotating cryogenic sapphire oscillators, 2015) bound any direction-dependence of c to about 9 × 10⁻¹⁹. Light travels at the same speed in every direction to staggering precision, confirming Lorentz invariance and no preferred frame, the opposite of a special, stationary Earth.
• 7°What one of these resonators is, since the word gets used without ever being explained. A cavity resonator is a rod of glass with a mirror at each end, and a laser beam bouncing between them. The beam only builds up if a whole number of wavelengths fits exactly between the two mirrors, the way a guitar string only rings at lengths that fit a whole number of half-waves. That makes the cavity a ruler whose tick marks are wavelengths of light, about 1,064 nanometers apart, and the frequency it settles on is a direct readout of its own length. Change that length by a fraction of a wavelength and the frequency shifts by an amount you can hear as a beat note. Hold two of them at right angles and any difference between them is a difference between two directions in space.
• 7°°And that is why they are cooled. A ruler made of glass is only as good as its own stability, and glass expands and contracts with temperature by far more than the effect being looked for. Cooling the cavity to a few degrees above absolute zero all but stops that drift, which is what buys the long runs: the 2003 experiment compared its two resonators for about a year. [709] The cryogenics are not exotic physics, they are how you keep the ruler from changing length while you are reading it.
• 7aThe paper people hold up is real, and it says something narrower than they think. The one usually shown on screen is Müller, Herrmann, Braxmaier, Schiller and Peters, 2003, in Physical Review Letters. [709] It is a genuine paper by real physicists and the null result is real. What it tested was whether the speed of light depends on direction. That is a question about light, not about whether the ground is moving, and answering the first one does not answer the second.
• 7bThe same abstract that is waved as proof of a motionless Earth says the resonators were subject to the Earth’s rotation. [709] That is not a throwaway phrase. The Earth turning is what carried the apparatus through every orientation over the year the data ran, and without it there is no modulation to look for and no experiment. A paper offered as evidence that the planet does not turn is built on the planet turning.
• 7cAnd the follow-up exists because leaning on the Earth’s rotation was not good enough. Using the planet as the turntable leaves one direction, the one along the spin axis, unconstrained. So the next generation cut two cavities into a single block of fused silica and turned it on an air-bearing turntable about every 45 seconds, reaching the 10⁻¹⁷ level. [710] A field that finds a blind spot in its own method and engineers a rig to cover it is not a field hiding a result.
• 8Speed of light, where the curved Earth entered. From the 1879 Annapolis rotating-mirror work to the 1926 Mount Wilson ↔ Lookout Mountain runs (~35 km, c ≈ 299,796 km/s) and the 1930–35 mile-long evacuated-tube measurement (c ≈ 299,774 km/s), Michelson timed light over long baselines. To fix the 1926 distance, the U.S. Coast & Geodetic Survey laid the “Pasadena Base” (34.6 km) to ~1 part in 11 million and triangulated the path, a geodetic survey that reduces to the curved Earth. Curvature wasn’t a nuisance; it was built into getting the answer.
• 9Dayton Miller, the one result that seemed to disagree. Concede the strongest version first. Dayton Miller was no crank. He was a respected American physicist and an officer of national scientific societies. His ether-drift work ran for more than thirty years and millions of readings, and even critics called his data carefully kept. In 1933 he reported a steady signal near 10 kilometers per second and called it “the absolute motion of the earth.” Einstein treated it as serious, and said a confirmed Miller result would put relativity in real trouble. So the claim has a pedigree. Now weigh it. A true ether wind from the Earth’s orbit, near 30 kilometers per second, should have been three times larger and easy to see. Miller’s signal was about a third of that, and no one with a more sensitive apparatus reproduced it. In 1955 Shankland and three colleagues reanalyzed his original data sheets. They traced the pattern to two dull causes: random scatter in a hard reading, and thermal gradients in the Mount Wilson hut, worse there than anywhere else. In 2006 Thomas Roberts modeled the same data and showed that Miller’s own averaging method manufactured a false signal with the shape he was hoping for. Roberts called it “every experimenter’s nightmare,” and bounded any real motion below 6 kilometers per second, consistent with none. One careful man chased an artifact for a career. The modern null below closes the door he believed he had opened. [663] [664] [665]
• 10The Sagnac effect does not bring the aether back. This is the aetherist’s best card, so play it fair. Georges Sagnac built his rotating interferometer in 1913 for one purpose: to prove the aether and refute Einstein. He saw a real fringe shift, and he believed he had done it. The effect is real. It is also, every day, the reason a fiber-optic gyroscope in an airliner knows it is turning. But it does not need an aether, and it does not break relativity. Max von Laue had shown the effect fit relativity in 1911, before Sagnac ran it, and Paul Langevin gave the full account in 1921. The reason is plain once stated. Sagnac measures rotation, not motion through a medium. A rotating ring is not an inertial frame, and relativity never promised that rotating frames behave like resting ones. In any resting frame the two beams both travel at the speed of light. But the detector moves while they are in flight, so one beam covers a longer path and arrives late. No wind is needed, only a spinning ring. And here is the turn: this is the very effect that clocks the Earth’s own spin, in Michelson–Gale in 1925, in every ring-laser gyroscope since, and in the Sagnac correction that keeps GPS honest. The instrument built to prove a motionless aether is the one that catches the Earth turning. To offer it as proof of a still Earth points the wrong way. [666] 49 119

Sources: Michelson–Morley 1887 [48] · Michelson–Gale–Pearson 1925 [49] · the Sagnac effect [314] · ring-laser gyroscopes & the Wettzell G ring [312] · modern isotropy-of-c tests to 10⁻¹⁸ [313] · stellar aberration [77] · Michelson’s speed of light & the Pasadena Base [50]. → Rotation & optics data rows

• 1Compton's water ring (1913). As a Wooster undergraduate, Arthur Compton filled a ring tube with water and oil droplets, let it settle, then flipped it 180°. Coriolis from Earth's spin sets the water drifting by a tiny, microscope-measurable amount, giving the rotation rate, the latitude, and the direction of true north, on a tabletop. (The third classical method, after the pendulum and the gyroscope.)
• 2The gyrocompass. Since ~1908 (Anschütz, Sperry), ships find true north not magnetically but by sensing the rotation axis itself; modern ring-laser and fiber-optic gyrocompasses do the same. Navies standardized on them because they point to the geographic pole.
• 3Ring laser gyroscopes. The Wettzell "G" ring (Bavaria, 2002–) is a 4 m square laser Sagnac interferometer on a buried Zerodur block; it tracks Earth's rotation finely enough to see length-of-day changes below a millisecond, plus tides and polar motion (Nature Photonics, 2023). Every aircraft inertial system runs a smaller version (56).
• 4Entangled photons (2024). Philip Walther's group in Vienna ran a 2 km optical-fiber Sagnac interferometer with maximally entangled photon pairs and measured Earth's rotation on a two-photon quantum state, about 1000× better than prior quantum sensors and, in their words, "a century after the first observation of Earth's rotation with light" (Michelson-Gale, 1925). Classical light then, quantum light now, the same spinning planet.
• 5A ring laser can’t be “calibrated to lie.” A ring-laser gyroscope measures rotation absolutely through the Sagnac effect. Counter-propagating beams beat at a frequency set by how fast the frame turns, needing no external reference. The Wettzell “G” ring reads Earth’s spin (~15°/h) to about one part in 10⁸, in an underground lab where temperature and pressure are controlled. The decisive point against “it’s just calibrated”: the same instrument independently reproduces the tiny wobbles, polar motion, length-of-day changes, the Chandler and annual wobble, that radio-telescope VLBI sees, and ring-laser gyros steer airliners by sensing this rotation (Entry 47).

Add Foucault's pendulum (Entry 47) and the Hafele-Keating flying-clock test (1971), and the spin has been independently confirmed by mechanics, fluids, optics, atomic clocks, and quantum entanglement, every one returning the same sidereal rate and the same sin(latitude) law.

One spin, written every way. The same rotation rate can be given as a period, an angular rate, or a surface speed. These are one fact in different units, and the instruments above all return this set, never zero. A flat, non-rotating Earth predicts zero in every row. A flat disc spun about the North Pole could fake a single angular rate, but its rim speed would climb without limit toward the edge and its Coriolis pattern would come out wrong, so no consistent set fits it.

Every row is one rotation rate, converted. Surface speed uses the equatorial radius 6,378 km and shrinks with the cosine of latitude, reaching zero at the poles. The day appears twice because Earth turns once against the stars (sidereal) a little faster than once against the Sun (solar); that ~3m 56s gap is why the stars rise about four minutes earlier each night [471].

Sources: Compton generator [51] · gyrocompass / inertial nav [21] · Wettzell G ring laser [52] · Hafele-Keating [53] · Walther/Vienna entanglement [47]. · Sagnac measures absolute rotation; the Wettzell “G” ring matches VLBI polar motion & length-of-day [166]. → Rotation data rows

• 1Parts: a rotor, gimbals, and a spin axis. A gyroscope is a heavy rotor (flywheel) spun at high speed inside a set of gimbals, pivoted rings that let the housing turn freely in any direction. The rotor’s spin axis is what matters. The gimbals isolate it from the frame, so turning the case does not turn the rotor.
• 2Rigidity in space. A spinning rotor stores angular momentum L = Iω (moment of inertia × spin rate), a vector along the spin axis. With no external torque, L is conserved, so the axis keeps pointing at the same spot in space no matter how you move the frame. The gimbals swivel around it. That rigidity in space is the heart of every gyrocompass and inertial navigator.
• 3Precession: a torque turns sideways. Push on the spin axis and it does not tip the way you pushed. It moves at right angles. A torque changes angular momentum (dL = τ dt), so the axis swings perpendicular to both spin and push, precessing at Ω_p = τ/(Iω). The faster it spins, the slower it precesses, which is why a fast top stands tall and a slow one wobbles over.
• 4The spin it can’t hide: drift = 15°×sin(latitude) per hour. Because the axis stays fixed in space while the ground turns beneath it, an Earth-bound observer sees a free gyro slowly wander, in azimuth at 15°×sinφ per hour (“drift”) and in tilt at 15°×cosφ per hour (“topple”). Zero azimuth drift at the equator, a full turn a day at the pole. That is a clean latitude signature of a rotating sphere, not random error.
• 4aSpin the rotor the other way and the answer does not change, which is the control experiment. A free gyro is not pointing at anything. It is holding a direction in space, and its axis tracks a star across the sky while the ground turns beneath it. [711] Reverse the wheel and it still holds a direction in space, so the drift you measure against the local horizon is the same 15°×sin(latitude) per hour it was before. The number belongs to the Earth, not to the machine. Anyone arguing the reading is an artifact of the instrument has to explain why running the instrument backwards produces the identical artifact.
• 4bWhere the direction does matter, conceded plainly. In a gyrocompass, which adds gravity and damping to hunt for north, rotor sense matters a great deal. The wheel has to spin anticlockwise seen from the south to give the westerly precession as the north end tilts up, and if you reverse it the control precession and the damping precession fight each other unless the pendulous weight is moved from above the gyro to below it. [711] So the sense is a real engineering constraint. It sets which way you must hang the weight. It does not set the rotation rate the instrument reads.
• 4cAnd the sense of the drift flips at the equator. North of it a free gyro’s apparent drift runs clockwise; south of it, anticlockwise, with the end pointing east always tilting up and the end pointing west always tilting down. [711] That is the same sign change the rotating optical resonator shows, reached with a spinning brass wheel instead of a laser. A flat disc turning about its center has every local vertical parallel to every other, so every gyro on it would drift the same way at the same rate, with nothing to flip and no equator to flip at.
• 5That is why a gyrocompass finds true north. A gyrocompass adds a gravity-sensed control that precesses the axis until its drift falls to zero. That happens only when the axis lies in the meridian, pointing at the geographic pole. Navies and airlines adopted it because, unlike a magnetic compass, it seeks the rotation axis itself (Entry 49). On a stationary flat Earth there would be nothing for it to find.
• 6No spinning mass needed: ring lasers read Ω directly. Modern systems often drop the rotor. A ring-laser or fiber-optic gyro sends light both ways around a loop and reads the Sagnac shift, demonstrated by the physicist Georges Sagnac in 1913, which is proportional to the rotation rate. The Wettzell ring laser tracks Earth’s spin to sub-millisecond length-of-day changes, and a 2024 entangled-photon version did the same on a quantum state (Entry 49, Entry 48). Spinning mass or light, the answer is the same 15°/hr.
• 7Transport wander: the same drift, now from crossing the globe. Keypoint 4 is the drift a parked gyro shows as the Earth turns beneath it, 15 degrees per hour times the sine of the latitude. Flight adds a second piece. When an aircraft travels east or west it changes its longitude, and a change of longitude is a rotation about the Earth’s polar axis. So the heading indicator gains an extra apparent drift equal to the change of longitude times the sine of the latitude. As a rate, that is the eastward ground speed divided by the Earth’s radius, times the tangent of the latitude. Pilots call it transport wander. Flying east adds to Earth rate, and flying west works against it. Flying due north or south along a meridian adds none, because that motion turns about a level axis, not the vertical one. Two things follow. First, this is why a directional gyro must be reset to the compass every ten to fifteen minutes. An inertial system instead computes Earth rate and transport rate together and removes them without pause. Second, read the shape of the correction. Every term carries the sine or the tangent of the latitude, and every term falls to zero at the equator. That factor is spherical geometry. A flat plane holds no latitude, so it predicts no such drift and no longitude term at all. The correction a working autopilot applies, and the fact that it puts the aircraft where the chart says, is the globe writing its signature into the instrument. 49

Sources: gyroscope rigidity & precession [297] · the gyrocompass & north-seeking [298]. See also Entry 47, Entry 49, Entry 52. → Rotation data rows.

• 1What the instrument is, and how a frequency nobody can count turns into one anybody can. Build a closed loop for light, a square of mirrors or a coil of fiber, and send a beam around it both ways at once. Hold it still and the two beams behave identically. Turn the loop, and the beam going the same way as the turn has a slightly longer trip, because its finish line is moving away from it, while the beam going against the turn has a slightly shorter one. Two trips of different length mean two slightly different frequencies. Now, light at 633 nm oscillates about 474 trillion times a second and no instrument on Earth counts that. So you do not try. You shine both beams onto one detector, and the brightness rises and falls at the difference between their two frequencies. It is the same thing you hear when two guitar strings are tuned close but not quite together: the two notes are far too high to count, yet the slow throb between them is obvious. Here the throb runs at a few hundred cycles a second, which ordinary electronics counts without complaint.
• 2Two things set the size of that throb, and both are geometry. The first is the shape of the loop: the ground it encloses, divided by the distance around it. The ring resonator relation puts numbers on it. A loop that fences off a lot of area for its length of path does well, so doubling the side of a square ring doubles the signal. That is why the serious instruments are meters across rather than centimeters, and why the best of them are anchored in bedrock rather than standing on a bench.
• 3The second thing is which way the loop faces, and that is where the whole argument lives. Picture an arrow pushed through the middle of the loop at right angles to it, like the axle of a wheel. The instrument does not respond to rotation in general. It responds only to the share of the rotation that lines up with that arrow. Point the arrow straight along the axis of the turn and you get the full reading. Hold it side-on and you get nothing. Everything below follows from that, because it means moving the instrument or tipping it over has to change the reading by an amount that can be worked out beforehand.
• 4Now stand the loop flat on the ground, and watch what latitude does to it. The arrow through a level loop points straight up, away from the center of the Earth. The spin axis, meanwhile, runs through the poles. Those two directions are the same thing at the pole, at right angles to each other at the equator, and somewhere in between everywhere else, and how far in between is what latitude means. So a level ring reads full strength at the pole, nothing at all at the equator, and the reversed sign south of it, following the latitude form. Now take a flat disc turning about its center. Every horizontal loop anywhere on that disc is parallel to every other, and all of them are perpendicular to the same axis. Every laboratory would read the identical value, with no latitude term and no sign change. The measured pattern is the globe’s, not the disc’s.
• 4aThe other version of the model, where the ground is still and the sky turns above it, fares worse still. That one is usually offered as a way to keep the night sky moving without moving the Earth. It cannot survive this instrument, because a ring resonator does not measure the sky at all. It measures whether the apparatus itself is turning, by comparing two beams inside a sealed box. Hold the box still and the two beams stay in step no matter what is happening overhead, so a stationary Earth predicts a beat note of zero everywhere, in every orientation, forever. The instruments read 7.292×10⁻⁵ rad/s instead, and they read a different share of it in every city.
• 5Turn the instrument upside down and the reading changes sign. Flip the loop and the arrow through it now points the other way, so the same rotation registers at the same size but with the opposite sign, and the throb comes back reversed. That is a bench test, done in an afternoon, with no reference to the sky and nobody else’s data involved. Something responding to a vague local disturbance has no reason to care which way up the box is sitting.
• 6Three rings at right angles recover one vector. Point three resonators along three different axes and each reads a different number. Combine them and they reconstruct a single rotation vector, of one magnitude, pointing one way. That is the design of GINGER, an array of large ring lasers at the Gran Sasso underground laboratory in Italy, whose prototype GINGERino is a 3.6 m square ring sitting 1,400 m below the mountain. [708] Three independent instruments agreeing on one axis is not an assumption anybody bolted on.
• 7The box is sealed, which is the point. No star sighting, no horizon, no Sun, no satellite, no appeal to anybody else’s data. The quantity measured is a frequency difference between two beams of light inside a closed cavity, in a room with no windows, and in GINGERino’s case under more than a kilometer of rock. [708] Whatever the reading is, it was not copied from an almanac.
• 8The steelman deserves its answer. It is true that a ring laser registers seismic rotations, and true that these instruments double as rotational seismometers. But that is the wobble on top, and the wobble is what varies. Underneath it sits a constant offset that does not go away between earthquakes, matches the predicted latitude projection, reverses on inversion, and agrees with the rate obtained by radio telescopes watching quasars. Attributing the varying part to earthquakes and the constant part to the Earth turning is not an assumption. It is what the two parts do differently.
• 9The number is small, and that is why the agreement means something. Earth turns at 7.292×10⁻⁵ rad/s, about 15 degrees an hour. The instruments resolve down to tens of picoradians per second. [708] A device that sensitive, reading a value that specific, matching a prediction made from geometry alone, in a sealed room, is a hard thing to get by accident.
• 10What it does and does not show. This measures rotation, not shape. A ring laser cannot tell you the Earth is round. What it can tell you is that the planet turns once a sidereal day about a fixed axis, and that the axis sits at a different angle over every laboratory, in the pattern a sphere requires. Shape comes from elsewhere Entry 49. This is the spin, measured in the dark.

• 1Storms pick a side. Hurricanes and low-pressure systems rotate counter-clockwise in the Northern Hemisphere and clockwise in the Southern. That is Buys-Ballot's law (1857). The deflection is zero at the equator and strongest at the poles, scaling with the sine of latitude, just as rotation predicts. A flat, non-rotating plane gives no preferred spin direction and no latitude dependence.
• 2Gunners and snipers pay for it. Long-range artillery fire-control and precision rifle solutions include an explicit Coriolis correction. Ignore it past ~1,000 m and you miss measurably, and the sign of the correction flips between hemispheres. The US Army’s own gunnery manual, TC 3-09.81, successor to FM 6-40, requires firing data to account for the rotation of the Earth. The “no rotation of the earth” phrase flat-Earthers quote from it is just one of the firing table’s standard baseline conditions, listed beside “no wind” and standard air and then corrected back to reality. Militaries don't budget for a force that isn't there. [426] The same out-of-context use of “flat, non-rotating Earth” engineering documents is the subject of 146.
• 3Even a dropped stone drifts east. Because the top of a tall drop is moving east slightly faster than the bottom, falling objects land a touch to the east, measured in deep mineshafts by Ferdinand Reich in 1833 and many times since. It's a second, independent rotation signature alongside Foucault's pendulum (Entry 47) and the direct spin measurements of Entry 49; it's also why launch sites and aircraft inertial systems (56) account for it.
• 4Drive east and you weigh a little less. Move east and your speed adds to Earth’s eastward spin, so the centrifugal effect grows and your measured weight drops slightly. Move west and it rises. That is the Eötvös effect. A team measuring gravity on ships in the early 1900s saw just this: lower readings sailing east, higher sailing west, confirmed in 1908 by two ships running opposite ways across the Black Sea. A train at 300 km/h reads about 0.1% lighter, and every marine and airborne gravity survey still applies the correction. A non-spinning Earth would show none of it.
• 5Hurricanes avoid the equator. The Coriolis force is the spin that organizes a storm, and it scales with the sine of latitude, falling to zero at the equator. So even though the equatorial ocean is plenty warm, tropical cyclones almost never form within about 5° of it (fewer than two a year on average), and none has ever been observed to cross the line. A flat disc spinning about a central pole would have no such equator-centered dead zone. A rotating sphere does, right where the math puts it.
• 6The rifle proves the spin. A 1,000-yard shot near 45° latitude drifts about 3 inches (8 cm) sideways from the Coriolis force, right in the Northern Hemisphere, left in the Southern, and ballistic computers flip that correction’s sign at the equator. The vertical Eötvös component also makes an eastward shot strike high and a westward one low. Long-range artillery firing tables have carried these terms for over a century. A flat, non-rotating plane needs none of them.
• 7The equation. The sideways Coriolis acceleration is a = 2 Ω v sin φ, with Ω = 7.29×10⁻⁵ rad/s (Earth’s spin, 2π per sidereal day), v the speed and φ the latitude. The full vector form is a = −2 Ω × v. It is zero at the equator (sin 0 = 0), largest at the poles, and deflects to the right in the north and the left in the south, the very pattern seen in storms, shells and ocean currents. A flat, non-rotating Earth has Ω = 0 and predicts none of it.
• 8It only exists in a rotating frame, which is the whole point. Coriolis is an inertial (“fictitious”) force. It appears only when you do physics in a rotating reference frame. In a non-rotating frame the shell flies straight and the Earth turns beneath it. Switch to the ground’s rotating frame and you must add a −2 Ω × v term to keep Newton’s laws. That term is the Coriolis force, and its size is fixed by Ω and latitude. Both pictures agree on where the round lands, so measuring the deflection measures the frame’s rotation. A flat, stationary Earth (Ω = 0) produces none, so every gun, gyroscope and pendulum that records it is reading Earth’s spin directly.
• 9The Navy railgun fires over the curve, not through it. A viral claim holds that the electromagnetic railgun, able to reach targets 100 nautical miles or more away, must prove a flat Earth, because a line-of-sight shot could not clear the bulge. But the railgun is not a line-of-sight weapon at that range. It is ballistic. The Navy’s own program documents describe the long-range flight as predominantly exo-atmospheric, meaning the projectile arcs high above the air and falls back down onto the target, the same indirect fire naval guns have used for over a century. The claim even concedes the target sits below the horizon, which is why the round must travel up and over rather than straight across. And like all long-range gunnery, the firing solution corrects for the Earth’s curvature and its spin. [558]

Sources: Coriolis effect, Buys-Ballot's law & deflection of falling bodies (Reich) [78] · the Eötvös effect [208]; why cyclones avoid the equator [247] · ballistic Coriolis drift [414]. Joins the rotation evidence of Entry 47 and Entry 49. → rotation data rows

• 1“Downhill” is set by gravity, not the compass. Water flows toward lower gravitational potential, lower elevation above sea level, regardless of north, south, east or west. The Nile flows north for ~6,650 km, yet it descends the entire way, from highland sources over 1,000 m above the sea down to the Mediterranean. Geographers note its direction is fixed by the slope of the land, not by rotation or magnetism. “North” is not “up.” [464]
• 2Everything co-rotates, so nothing sloshes. The water, the riverbed, the banks and the air above them all turn with the Earth at the same rate, so there is no force to drive a river first one way and then the other. It is the same reason the oceans stay put and you feel no spin (Entry 45). The river just follows its bed downhill.
• 3Rotation’s real effect on a river is a constant lean, not a reversal. The Coriolis force (Entry 52) nudges a current slightly sideways, so rivers erode the right bank a little more in the Northern Hemisphere and the left in the Southern. That is Baer’s law, which Einstein wrote a paper on in 1926. It is a permanent, one-sided bias, never a daily flip. [462]
• 4The one real reversal proves the opposite. Some rivers genuinely do run backwards, roughly twice a day, in the tidal bores of the Qiantang, the Amazon’s pororoca and the Bay of Fundy, where an incoming tide funnels up a narrowing estuary. But that is the Moon’s tide (Entry 42) on a ~12-hour lunar rhythm, confined to the lowest reach near the sea, not the planet’s 24-hour spin acting on whole rivers. It is globe-and-Moon evidence, not flat-Earth evidence. [463]
• 5The gauges settle it. Every stream-gauging station on Earth records steady, one-way discharge; a river reversing on a daily cycle far from the sea would be the single most obvious signal in all of hydrology, and not one has ever been recorded.

Sources: the Nile’s northward, downhill course set by the land’s slope [464]; Baer’s law, the Coriolis erosion bias [462]; tidal bores that reverse a river’s current [463]. Joins the rotation evidence of Entry 52, the tides of Entry 42 and why we feel no spin (Entry 45).

• 1East = lighter, west = heavier. Eötvös noticed that gravity data from ships (gathered ~1901–1905 by Oskar Hecker across three oceans) read low when the ship moved east and high when it moved west. He traced it to Earth’s rotation, the vertical kick of the Coriolis force (Entry 52).
• 2Why the spin does it. You are already circling Earth’s axis at up to ~465 m/s at the equator. Head east and your total eastward speed rises, so more of gravity is “spent” supplying centripetal force and your measured weight drops. Head west and it climbs (Entry 45). On a non-rotating Earth, direction of travel could not change your weight at all.
• 3Small, but firmly measurable. At the equator the east-versus-west difference is roughly 0.03% of gravity. For example, a 1,000-g weight on a fast aircraft reads about 991 g flying east and about 999 g flying west, an ~8-g swing from nothing but heading.
• 4Confirmed by design in 1908. Eötvös predicted the size, then had it tested with two ships crossing the Black Sea in opposite directions; the eastbound and westbound gravity readings differed just as the rotation model said. In 1913 he reproduced it in the lab on a rotating balance, an independent rotation proof alongside Foucault’s pendulum (Entry 47).
• 5Still corrected for today. Every ship- and aircraft-borne gravity survey applies the Eötvös correction (proportional to speed, heading and the cosine of latitude) before the data can be used. It is not a historical curiosity. It is routine, working geophysics that silently assumes, and depends on, a spinning Earth.
• 6It points to the spin specifically. The effect vanishes for north–south travel and is largest east–west at the equator, fading to zero at the poles, the signature of rotation about the polar axis (Entry 49). No flat, stationary model has any reason to make your weight depend on compass heading.

The Eötvös effect: east–west weight change from rotation [365] · Hecker’s ship data & the 1908 Black Sea confirmation [366]. Connects to the Coriolis effect (Entry 52), why we don’t feel the spin (Entry 45) and measuring Earth’s spin (Entry 49).

• 1The rule: you feel acceleration, not motion. In a jet at 900 km/h your coffee sits flat and you can walk the aisle; you feel only takeoff, turbulence and turns, the changes. Steady velocity is undetectable from inside a closed cabin (an inertial frame). “We’d feel 1,000 mph” is the whole error. Constant speed has no feel, on a plane or a planet.
• 2Spin: ~1,674 km/h at the equator. One turn per sidereal day (23 h 56 m). This is the only motion with a felt effect, and it is tiny. The centripetal pull at the equator is ~0.034 m/s², about 0.3% of gravity, so you weigh ~0.3% less at the equator than at the poles, just as measured. To fling anyone off you would have to spin Earth ~17× faster. The spin is detected by Foucault’s pendulum, the Coriolis and Eötvös effects and ring-laser gyroscopes (Entry 47, Entry 49, Entry 52).
• 3The Moon around us: ~3,680 km/h (~1 km/s). Strictly the Moon’s own motion rather than one Earth carries us through: it laps the planet about once a month at ~1.022 km/s [385]. Earth answers with a small monthly wobble around the shared Earth–Moon center of mass, a point ~4,670 km from Earth’s core, still inside the planet, so even our companion adds a gentle motion no one feels.
• 4Orbit around the Sun: ~107,000 km/h (29.8 km/s). One lap per year, ~150 million km out. You feel nothing because Earth and everything on it are in free-fall around the Sun. The Sun’s gravity supplies the centripetal force, and in free-fall there is no sensation, just as astronauts float. The orbit is proven by stellar aberration (~20.5″, Bradley 1727) and stellar parallax (81).
• 5The Sun around the galaxy: ~828,000 km/h (~230 km/s). One galactic year is ~225–250 million years, and Earth has lived only about twenty of them. Again it is free-fall around the galactic center, with a centripetal acceleration of order 10⁻¹⁰ g, undetectable by feel, but mapped through stellar dynamics.
• 6Through the cosmos: ~370 km/s. Add every motion and our net speed through the universe’s rest frame is ~370 km/s for the Sun (~620 km/s for the whole Local Group of galaxies). This is measured directly as the CMB dipole. One half of the sky is ~0.0034 K warmer in the direction we are heading. Steady, so unfelt. Real, so detected.
• 7The slow wobbles, layered on top. Axial precession swings the axis around a cone every 25,772 years, swapping the pole star (Polaris now, Vega in ~12,000 years). Nutation adds an 18.6-year nod (~9″) driven by the Moon’s orbit. The axial tilt drifts between 22.1° and 24.5° over ~41,000 years and the orbit’s shape cycles over ~100,000 and 413,000 years (the Milankovitch climate cycles). The Chandler wobble walks the pole ~9 m every ~433 days. And Earth swings monthly around the Earth–Moon barycenter, a point ~4,671 km off-center (still inside the planet). All real, all measured, all far too slow or small to feel.
• 8Why nothing flies off or lands wrong. Put numbers to the scare: the spin’s outward pull is 0.3% of gravity, so oceans and people stay firmly down (Entry 45). The orbital and galactic motions are felt by no one because the entire planet shares them in free-fall. A freely falling frame is weightless, not crushing. A dropped ball lands at your feet for the same reason a coin dropped in a cruising jet falls straight down: it already shares the motion. Planes take off because runway, air, aircraft and pilot all move together.
• 9Inertial vs non-inertial: the one principle behind the whole list. A frame moving at constant velocity is inertial, and Galileo’s relativity says no experiment sealed inside it can reveal that motion. That is why Earth’s orbit and galactic flight are not merely unfelt but undetectable from within. Catching them takes an outside reference (stellar aberration and parallax, 81; the CMB dipole, above). Rotation is the exception, because spinning is acceleration. Earth’s surface is a non-inertial frame, and non-inertial frames give themselves away from the inside through fictitious forces: the centrifugal ~0.3% weight loss at the equator and the Coriolis deflection of winds, pendulums and shells. That asymmetry is the real answer to “why don’t we feel it?”. Steady motion is hidden by physics itself, while the single accelerating motion, the spin, leaves fingerprints we have measured many times over (Entry 47, Entry 49, Entry 52).

Sources: axial precession [303] · nutation & the Chandler wobble [304] · the galactic year & the Sun’s galactic speed [305] · our motion through the cosmos, the CMB dipole [306] · obliquity & orbital (Milankovitch) cycles [307]. See also Entry 45, Entry 47, Entry 52. → Rotation & orbit data rows.

An aircraft holds a constant barometric altitude, which is a surface of constant geopotential. That surface curves with the Earth. Lift always acts along the local vertical (straight up from the wings when they are level), and the local vertical itself rotates as you travel, at the rate v/R. At ~900 km/h that is about 0.0022°/s (~8°/hr). That is a steady, unnoticeable nose-down that the autopilot and altimeter keep without anyone touching anything. It never builds up as felt pitch, because the reference frame turns with it. On a flat plane, no such turning would ever be needed.

Interactive model: how an attitude indicator works

The instrument is a gyro-stabilized card erected to local vertical (gravity). The fixed orange aircraft symbol reads pitch (against the blue-sky / brown-ground split) and bank (against the top scale). Drag the sliders; then press play to fly around the globe and watch "level" silently re-define.

MEMS vs physical gyros, and what they reveal

• RLG/FOGOptical gyros (ring-laser, fiber-optic) use the Sagnac effect, with no moving parts. Nav-grade bias is ~0.001–0.01°/hr, sensitive enough to measure Earth's 15.04°/hr rotation and gyrocompass to true north with no GPS. The airliner's IRS doesn't hide rotation. It depends on sensing it.
• MEMSMEMS gyros are vibrating-structure Coriolis sensors: tiny, cheap, with ~1–100°/hr bias. That is too noisy to gyrocompass off Earth's 15°/hr spin in real time, so an AHRS pins their drift using gravity (accelerometers) and the magnetic field, sometimes aided by GPS. They're the standard attitude/heading source in glass-cockpit AHRS from light aircraft to business and commercial jets (which still use optical IRS for primary navigation), plus drones and phones.
• SPINSpinning-mass gyros are the classic mechanical instrument behind legacy attitude indicators. They are erected to local vertical, drift slowly, and are corrected continuously.
• INSThe instrument is tuned to Earth’s radius. Every inertial navigator is ‘Schuler-tuned’: its platform is given a natural period of about 84.4 minutes. That is the swing period of a pendulum as long as Earth’s radius (~6,371 km). That tuning is what keeps the platform pointing ‘down’ toward Earth’s center as the aircraft rides over the curve. Max Schuler worked it out in 1923, and you cannot build a working inertial nav system without baking the globe’s radius into it.

Sources: flight dynamics & attitude reference [23] · RLG/FOG & Earth-rate sensing [21] · MEMS gyros [22] · Schuler tuning & the Earth-radius period [207]. → Navigation data rows