On August 16 1960, Captain Joe Kittinger of the US Air Force stepped into the open gondola of a helium balloon in New Mexico and went up to a height of just over 31 kilometres - three times as high as a long-haul passenger aircraft, with 80 per cent of the Earth's air beneath him. Then, in the cause of science, he jumped out.
Wearing the prototype of the pressure suits later used in the US space programme, Joe Kittinger proceeded to free-fall down to about 6,000 metres, at which point he opened his parachute. In the upper part of his fall, the air was so thin that he built up a falling speed of well over 1,000kmh, claiming a record for the fastest human without an aircraft.
Joe Kittinger is one of many courageous explorers adding to our understanding of the upper atmosphere. How far up does our atmosphere go? Could you rise somehow to the surface, hold your breath and pop your head out into empty space for a moment, like a whale emerging from the sea?
Well, it's not really like that. As you'd expect, the atmosphere gradually peters out, until it's a mere wisp of scattered molecules, merging with the gas and dust of space. Defining the final boundary, therefore, depends on what exactly you're talking about. The figure of 560km is sometimes used, but at that height, the distinction between air and space is difficult to perceive. Even at the height of Kittinger's flight, air pressure is nearly 1,000 times less than it is at sea level.
Take a famous example: Everest (or any other high mountain). You don't have to climb far before you start feeling a little short of breath. At 3,000m your brain is getting 10 per cent less oxygen than at sea level. At 6,000m you can get "mountain sickness" - extreme shortness of breath, headaches, difficulty sleeping, sometimes unconsciousness: time to strap on the oxygen cylinders. Everest, 8,850m (this new official height was set in 1999, with new technology putting it two metres higher than the previous official figure), is usually climbed with the aid of extra oxygen.
Air "gets thinner" the higher you go. But what does that mean? The answer is partly to do with density and partly with pressure. At the surface of the Earth, the air, compressed by the weight of the atmosphere above, is more dense, which means more oxygen molecules in each breath. Go higher and the density decreases. At 5,500m, it is half of what it is at sea level.
Low down, the air is not only more dense but also under greater pressure.
Go to 12,000m and above and low pressure becomes an issue, as well as lack of oxygen. Lungs start to lose their function, and gas bubbles begin to appear in the blood. You need to be either in a pressurised cabin or wearing a pressure suit.
If you try a slower ascent - actually living quite high up for a time - your body can adjust up to about 7,000m (the highest human settlement on Earth is Tada Village in Nepal, 5,769m above sea level). On a holiday flight, you're likely to be cruising at 10,000m. The air inside the cabin is pressurised to simulate a lower height (usually about 2,500m). It lulls you into thinking that the sunny environment outside might feel much the same. If you did step out, however, it would be a toss-up between freezing to death (at - 40oC) and dying from lack of oxygen. And yet, at that height there's demonstrably enough air to support the plane and feed the oxygen-hungry jet engines.
Add another 3,000m, though, and conventional planes have more difficulty finding lift in the thinning air. At 13,000m and above, a passenger jet's cruising speed gets disconcertingly close to its stalling speed - the minimum speed at which it will stay in the air. Very high altitude manoeuvres in fighter aircraft can result in uncontrolled tumbling if the wings lose their grip on what has become an insubstantial medium. At that height, you are entering the stratosphere.
For most of history, air was considered to be a God-given homogeneous substance. But experimentation and observation showed that air behaves differently in different circumstances. By the 17th century, it was known that burning something in a closed vessel used up some air, leaving behind a "different kind of air" that wouldn't support life or flame. This led to the conclusion that air may be a mixture of gases. In the 1770s, three people isolated oxygen: the Swedish pharmacist Carl Wilhelm von Scheele, Joseph Priestley in England and the French chemist Antoine Lavoisier. Who first realised that oxygen is a component of air and who stole the credit from whom are still disputed.
The belief that air is essentially a mixture of two gases - mostly nitrogen and about 20 per cent oxygen (plus a very small amount of carbon dioxide, discovered by Joseph Black in the 1750s) - held sway until the 1880s, when British physicists Wlliam Ramsay and Lord Rayleigh discovered a new element - argon - which makes up 1 per cent of air by volume. Through the 1890s, Ramsay went on to find tiny proportions in air of more elements - Jneon, krypton and xenon. He also found helium, which had been discovered for the first time on the Sun 30 years earlier, by spectroscopic analysis of the Sun's light.
Nitrogen makes up about 78 per cent of dry air by volume. Oxygen is next at 21 per cent. Argon has a little less than 1 per cent, carbon dioxide 0.03 per cent and four trace elements - helium, argon, xenon and krypton - make up the rest. There are minor variations - there is usually water vapour for example, perhaps 4 per cent, depending on temperature and location.
Other scientists soon realised that atmosphere has mass - that it presses down on the surface ofthe Earth under its own weight. The key to this discovery lay in the behaviour of the "suction" pumps commonly used to lift water from wells or mines. These work by drawing water up a pipe (just as you can draw water from a bucket with a bicycle pump) but what puzzled everyone who made one was that you could pump indefinitely and the water would not rise more than 10 metres or so above its original level. Galileo wondered why this is so and whether a heavier liquid would be even more reluctant to climb the tube.
One of his students, Evangelista Torricelli, carried out the classic experiment that gave the answer. In 1644, he filled a tube, closed at one end, with mercury - 13 times heavier than water - and put the open end into a bath of mercury. The mercury began to run out into the bath but stopped when there was still 75 centimetres of mercury left in the tube. Torricelli correctly deduced that it was the weight of the air above the bath that was holding the column of mercury up inside the tube, and that the air weighed just enough to hold the column of mercury at the level which he observed.
Later experimenters demonstrated that pressure reduces with altitude.
Crucially, this led to the conclusion that the atmosphere must be finite in extent and doesn't just go on never-endingly into space.
But to find out what the upper atmosphere is like, you have to go there - or at least send up some instruments. The story of the exploration of the upper atmosphere starts with some heroic balloon flights. By the end of the 19th century, several explorers had been above 9,000m, and a few of them had died from lack of oxygen. Experimental high-altitude balloon flights started in earnest in the 1930s, in pressurised capsules.
US and Soviet experiments took explorers higher and higher, culminating in the flight of the American "Explorer II" up to more than 22,000 metres in 1935. Experiments recommenced in the 1950s, with a series of American balloon ascents leading up to "Stratolab V", which went up to 34,668 metres in 1961, by which time orbital space flights were starting and the era of balloon exploration came to an end.
Now, our knowledge of the upper atmosphere comes from high altitude samples gathered by rockets and high flying aircraft, and by observations and analyses from orbiting satellites.
Extremely thin layer, almost a vacuum. Fades off gradually into space at about 560km. 99 per cent of the mass of the atmosphere is contained below the exosphere.
Extends to 400km. Even here, there's enough air to catch meteorites and turn them into burning shooting stars.
Up to 100km. Coldest layer. Most meteors burn up in this layer.
Most long-distance airline routes operate in the lower stratosphere, where the air is generally calm. Extends to about 50km, where the air pressure is one-thousandth of that at sea level. In the upper part are powerful winds called "jetstreams" that can reach 400kmh or more, always from west to east. The stratosphere has within it the "ozone layer", between about 15km and 40km above the Earth (see panel, right).
Rests on the Earth's surface. Extends roughly up to 16km at the equator and 9km at the poles. Most weather occurs within this turbulent layer and most of the air is contained in it - 75 per cent of the air is below 11,000m.
Friends of the Earth - information on pollution:www.foe.co.uk
NASA-basic facts and figures about the Earth's atmosphere: http:liftoff.msfc.nasa.govacademyspaceatmosphere.html
Weather information for the UK www.bbc.co.ukweather
Official site of the UK meteorological office: www.met-office.gov.uk
Official site of the World Meterological organisation: www.wmo.chindex-en.html