Freeman Dyson thinks big. Thirty years ago, in a series of lectures in London, he calmly proposed that humans might genetically engineer trees to grow on comets. Since there would be little gravitational grab in a comet but lots of water and, of course, sunlight, his comets would soon sprout epic quantities of foliage, providing an ecosystem for a new kind of genetically engineered human, to swing in huge cometary orbit through space, as Tarzan swung across the African jungle.
He began with RAF Bomber Command, and later worked with veterans of the Manhattan Project. He and colleagues devised a really safe nuclear reactor, in the days when people thought nuclear technology could be safe. But he kept thinking on a cosmic scale. As a hedge against the ultimate energy crisis, he proposed crunching up the solar system's planets and asteroids, to build a huge rigid shell around the sun, collecting all its energy - and not just the bit aimed at the tiny Earth. What is now known as a Dyson shell would provide humanity with one enormous solar-powered living room, and its interior surface would seem utterly flat to the exponentially growing population that colonised it and everybody would have a place in the sun.
After all that, far-out ideas like Project Orion seem almost demure. In 1957, the Russians put up Sputnik I, and a panic-stricken US struggled to launch a modest competitor into low Earth orbit, just to stay in the space race. But, at the same time, Dyson and a team of physicists and engineers at General Atomic were working in secret, on something entirely different.
With air force money, they were designing a nuclear spaceship the size of an ocean liner, blasted to a notional 10,000km a second by a machine-gun fire of exploding fission or hydrogen bombs ejected behind it. The bombs, rattling down a kind of Coca-Cola bottling plant machine system - yes, they visited Coca-Cola to look for engineering solutions - would explode at the rate of two, or maybe four, a second, each heating and accelerating a propellant that would slam against a huge, heavy pusher plate mounted on hydraulic shock absorbers behind the vessel, kicking the spacecraft to ever greater velocities. The propellant would be water or excrement, or something - it didn't matter.
In July 1958, aboard a scaled-down version - the ship would be a hemisphere 12 storeys high, with big picture windows and weighing 4,000 tons - Dyson and his colleagues proposed to fly to the moon, to Mars, even the 800m miles to Saturn, eventually to touch down on Enceladus, a moon with the density of a snowball (and an escape velocity of 400mph, just to make take off easy). Nasa is beginning to look again at what it calls "pulsed propulsion". Arthur C Clarke and Stanley Kubrick toyed with the idea of a nuclear spaceship for the film 2001: A Space Odyssey.
Dyson is now approaching 80, but still busy with science, at Einstein's old haunt, the Institute for Advanced Study at Princeton. Would he really have gone? Yes, he says: you would have needed 40 people in those days to do the things that robots and automatic sensors do now.
"We would have been scanning the sky. We certainly planned to take lots of telescopes with us, scanning the skies, scanning the asteroids, scanning the planets, doing all sorts of observing, measuring fields and particles, all the kinds of things now done with machines. It wouldn't have been just sitting and cruising and getting bored. The cruise part of the trip would have mostly been astronomy, then after you landed you would do geology and the business of exploring a planet," he says, cheerfully. The motto for the Orion team was "Saturn by 1970".
These days, Freeman Dyson thinks small, not big. Manned flight now is a disappointment. The space shuttle is a freighter that tries to carry coal and passengers, but does neither well. The international space station is just a kind of high- altitude bus ride. "If I were in charge I would get rid of the space station, get rid of the shuttle and start afresh with a high-performance two-seater.
"If you want the manned programme to do something, it should be directed towards people having fun. That is essentially what it should be all about. It certainly shouldn't be sold to the public as having anything to do with science," he says. "The reason it has got into such difficulties is because it was sold under false pretences. If you started afresh with a two- seater with a high performance that could go places, that could be very successful. You could imagine a couple of people digging in on the moon and finding out how to live there. It would be a great public entertainment, it could all be on television and you could see their adventures every evening."
The story of Project Orion was teased from him - and his surviving colleagues - by George Dyson, his son, who made a career building hi-tech Aleut kayaks in the Pacific north-west before writing a best-seller history of computing called Darwin Among The Machines. He wrote it because he realised that the computer pioneers were dying without telling their stories.
"As you get older, you realise that unrecorded wisdom is the most valuable thing. So the book is intended to be a kind of a monument to those people who had all that imagination, besides my father. The other thing is, I wanted to understand what had happened. I had grown up with this tremendous curiosity about the project. Imagine being that age and hearing that your father is working on a spaceship, but he can't explain it to you."
