Cities on a Spool

The Princeton Question

Mara: In 1969, Gerard O'Neill was teaching introductory physics at Princeton, and his class had just watched Apollo on television. To an advanced section of freshmen he put a question no textbook asked: is a planetary surface the right place for an expanding technological civilization? He had them calculate habitats in space, built as rotating pressure vessels, big sealed containers spun on their axis. That's how O'Neill himself later told it. And the numbers surprised him.

Eli: And the listener's question starts right there: explain the O'Neill Cylinder, how it would work, and whether it is a real possibility—basically, a colony floating in space. Today's angle is the physics paradox of Island Three.

Mara: Underneath sat a bigger bet: that open space itself, not some planet, could become a real home for a whole civilization.

Eli: And the question landed on students disillusioned with the Vietnam War and a degraded environment.

Mara: That's the mood his answer fits. Within the limits of existing materials, a habitat could be several kilometers across. That's several hundred square kilometers of inner land, at Earth-normal gravity made by spinning. Later, in The High Frontier, he was partly answering a depressing 1972 report, Limits to Growth: overpopulation, environmental damage, depletion. O'Neill framed orbital habitats as a direct answer to planetary limits, aiming to transfer heavy industry off Earth so the home planet could be preserved.

Eli: No theoretical obstacles, he found. A finding about physics, not a finished structure.

Mara: And the version people picture today is his biggest one, Island Three.

Eli: Island Three is two cylinders, not one. By one modern tally, each runs about twenty miles long. Each is about five miles across. Spin one, and anything on the inner wall gets pressed outward against it, and that press stands in for gravity. But a spinning thing that big behaves like a gyroscope, it wants to hold its direction. So the two counter-rotate, one clockwise, one counterclockwise, as a pair. That cancels the gyroscopic effects that would otherwise make it hard to keep them aimed at the Sun.

Mara: So the pair stays sunward.

Eli: Right, and the size helps too. At that radius each turns only about twenty-eight times an hour. Spin a small habitat fast and Coriolis forces kick in, that sideways shove felt when moving inside something spinning, bringing nausea and disorientation. Big and slow keeps it mild.

Mara: Which makes for a strange home: land curving up the inside of a spinning tube, and no obvious way for a sun to rise over it.

A World Inside

Eli: The way is glass and mirrors. The inner wall alternates, a long strip of land, then a long strip of window, round the whole cylinder. Outside the windows sit movable mirrors throwing sunlight in. Change their angle and the colony sets its own clock, day length and seasons both. O'Neill described it in a 1979 interview: "The sun would seem to rise in the morning and set at night, an effect accomplished through a complex array of mirrors."

Mara: That's why the pair has to stay sunward: the mirrors need it there.

Eli: And an enclosed volume that big makes its own weather: clouds and rain, which designers could in principle tune. O'Neill, a glider pilot, even checked it could give soaring pilots lift.

Mara: Not everyone favored the cylinder. In 1975 NASA ran a ten-week summer study at its Ames center, weighing different spinning shapes. The design it picked as most feasible came from Stanford students: the Stanford Torus, a giant wheel with people living around the rim, like the inside of a bicycle tire. By one engineering comparison, for the same floor seen from above, a cylinder needs roughly four times the structural mass of a torus. And it brings stronger Coriolis trouble, the kind felt when someone turns their head off the line of the spin.

Eli: So the wheel gives the same floor for far less metal.

Mara: It has its own snags. Beth Westfall points to a possible giant mirror, and a size built for a whole colony that would need scaling down.

Eli: Wheel or cylinder, either one needs a parking spot and a mountain of building material, and that's where O'Neill went next, Mara.

The Lunar Catapult

Mara: May 1974, Princeton. O'Neill's calculations go public at a one-day conference on campus. The New York Times runs it on the front page. That September, Physics Today publishes his article, The Colonization of Space, and he writes: "I believe we have now reached the point where we can, if we so choose, build new habitats far more comfortable, productive and attractive than is most of Earth."

Eli: And soon afterward, coverage followed in many other newspapers, magazines and broadcast media, in the U.S. and abroad.

Mara: Two big pieces made the plan. First, where to put it: at L4 or L5, Lagrange points, spots where the pulls of Earth and Moon balance so a thing parked there keeps its place. Out there, O'Neill and his students reasoned, the sun shines all the time, and nobody fights gravity coming and going.

Eli: One group even took its name from that orbit: the L5 Society.

Mara: Second, where to get the stuff. Early on, O'Neill saw that launching material off Earth was prohibitively expensive. The Moon, with its weaker pull, could be the quarry. Lifting a pound of lunar ore to high Earth orbit takes less than a twentieth of the energy it would take from Earth. So between 1972 and 1974 he hunted for a cheap way off the lunar surface, and landed on an electric track that would fling Moon rock into space.

Eli: The Moon's rich in oxygen for air. Nitrogen, though, it barely has.

Mara: Neither do the nearby asteroids, so nitrogen would have to come from Earth or elsewhere, and how to get megatons of it into orbit is still unclear.

Eli: So five years on from that classroom, O'Neill had a front page, a parking spot and a quarry. What he didn't have yet was hardware showing the Moon could actually ship rock. That's what the prototypes were for. O'Neill called the launcher a mass driver: a long electromagnetic track whose magnets hurl small carrier buckets along, then circulate them back for the next load. Mass Driver I pushed its load at thirty-three times Earth's gravity. Mass Driver III hit eighteen hundred times Earth's gravity. And a 1976 paper he co-wrote said the mass drivers it considered could be built with present-day technology and commercially available materials.

Mara: Eighteen hundred times gravity. A launcher for minerals, then, not people.

Eli: And a ten-week summer study concluded space colonization appeared technically feasible. The obstacles, in its words, were "principally philosophical, political, and social rather than technological."

Mara: That sharpens the puzzle. The physics checks out and the launcher ran on a test track, so whatever has kept these cylinders unbuilt must be something else.

The Trillion-Dollar Shield

Eli: It's weight. Deep space is laced with cosmic rays, particles so energetic that ordinary metal walls don't stop them. Cosmic ray shielding means sheer bulk, thick rock or lunar soil, and for a settlement, millions of tons of it. Then the 1975 study hit a snag: that much mass can't spin with the habitat. The stresses would exceed the strength of the shield's own materials. So the shield has to sit apart, turning far slower or not at all. Skeptics like Robert Zubrin weigh that heavily.

Mara: So the miles-wide vessel O'Neill's math allowed needs a coat it can't wear.

Eli: And the vessel itself has a ceiling. Its wall carries hoop stress, the pull running around the curve that tries to split it open, like a barrel straining at its hoops. Widen the cylinder and that pull grows, so the radius is capped by how strong a metal is for its weight. Cold-drawn steel, pulled through a die to strengthen it, holds about one and a quarter billion pascals of tension. Cold-formed titanium is a bit weaker, but a good deal lighter.

Mara: O'Neill's own 1976 paper hunted shielded shapes giving the most land for least mass.

Eli: Then the bill. The 1975 plan called for lifting about a million tonnes off Earth inside a decade, and O'Neill imagined the first loads riding the space shuttle.

Mara: So everything hung on what a shuttle ride would cost.

Eli: O'Neill worked with a much lower launch estimate. By 2008, the going rate to low orbit was about twenty-five times that.

Mara: Twenty-five times the price his whole plan was built on.

Eli: At those prices, The Space Review put just getting Island One, a smaller first habitat, started at one and a half to ten trillion dollars.

Mara: Even with the Moon's mass driver around Earth's prices?

Eli: Robert Zubrin ran the numbers with it anyway. Say a lunar mass driver got ore to high orbit for a ten-thousandth of the going launch price. Even then, he says, launching the raw material for an O'Neill habitat would cost four trillion dollars.

Mara: What sank the 1970s timeline was deadweight: the rock it takes to keep cosmic rays off people, and the trillions, by these estimates, to lift it.

Eli: Which leaves the bill with whoever believes enough to pay it, Mara, and one of O'Neill's readers went and founded a rocket company.

The High Frontier Revisited

Mara: Jeff Bezos has said, "I read 'The High Frontier' in high school. I read it multiple times." He founded Blue Origin in 2000, and he talks about millions of people living and working in space, with heavy-lift rockets as the enabling step. He backs O'Neill's free-space colonies over planetary surfaces, and describes them this way: "they'll have farms and rivers and universities; they could have a million people in them. They're cities."

Eli: Robert Zubrin's reading: "It's a lot easier to settle a planet than to build one."

Mara: Al Globus, an advocate, counters that Mars needs all the same things as orbit: pressure vessels, power, life support. And life support is the open one. A closed-loop biosphere, a sealed living system recycling all its own air, water and food, is still unsolved. Earth's soil took millions of years of life to build. Even O'Neill, in 1979, granted the critics' point: "Others say — quite rightly — that we're nowhere near achieving that on Earth, so how can we hope to do it in space?"

Eli: As of 2019, Astronomy magazine still called O'Neill's concepts some of the most practical options for large-scale, long-term space habitation. The rockets have changed: low-cost, fully reusable launchers existed only on paper in 1975, and by 2026 they're a reality on the launch pad. But Space.com, in 2025, called the technology for Island Three, or even the smaller spherical Island One and Two, completely untested.

Mara: So the readings split. Advocates like Bezos and Globus see it as achievable once launch costs fall and asteroid mining matures. Skeptics like Zubrin see shielding, stress, biology and cost putting it out of reach for the foreseeable future.

Eli: So O'Neill's 1969 classroom question gets a two-part answer. The cylinder holds up as physics: a pair of sealed, counter-spinning tubes giving Earth gravity with nothing but known mechanics. What it can't be, yet, is built. That waits on mining off Earth at industrial scale, and on a living system proven to last.

Mara: Three things would change the picture. Whether multi-year sealed life-support tests can run without collapsing or building up toxic gas, with no resupply from Earth. Whether working mines and mass drivers ever operate on the Moon or near-Earth asteroids. And whether peer-reviewed medicine shows mammals can reproduce and grow up, generation after generation, in spun gravity.

Eli: Until one of those lands, Island Three stays where those freshmen first found it.

Mara: Dig into Coriolis forces.

Eli: And stay with us on technology.

Mara: That's our story for today. I'm Mara.

Eli: And I'm Eli. Thanks for listening to Angle, and until next time.