For example, Galileo is said to have dropped spheres of different weights from the top of the leaning tower of Pisa to demonstrate to sceptics that the spheres would strike the ground together. Instead of presenting these instances as experiments, he and his supporters claimed that it was common knowledge that the spheres fell at the same rate.

The book examines how the scientific revolution introduced experimentation into science. It is extensively researched with detailed footnotes. But it is difficult to read. Complex ideas are expressed in inaccessible language and illustrated by examples given in the lingua franca of the time, Latin. Dear expects his reader to be familiar with the works to which he refers – even though he cites more than 500.

Today, physicists believe that all matter is composed of 12 particles – six leptons, including electrons – and six quarks, which together form larger particles such as protons and neutrons. This year’s Nobel laureates discovered two of the leptons.

Perl heard about the award at 4.30 am, when a journalist phoned him at home. “I didn’t believe it,” he says, “so I asked her what her source was and she said ‘Stockholm’.” Reines, who is ill and has been in hospital for some time, has not commented publicly on his award.

One of the cornerstones of physics is the law which says that energy can neither be created nor destroyed. But in the late 1920s, physicists found that energy did not seem to be conserved during neutron decay. To resolve this problem, another Nobel prizewinning physicist, Wolfgang Pauli, proposed that the missing energy had been carried away by an invisible particle – a neutrino.

The hypothetical particle would be extremely difficult to detect because it had no electric charge, very little mass – or possibly none at all – and hardly interacted with matter. In fact it took 25 years and the invention of the nuclear reactor before Reines and his colleague, the late Clyde Cowan, could develop a technique to show whether or not neutrinos existed.

“It’s like looking for an invisible person,” says Christine Sutton of the University of Oxford. “You know that they are walking down the street, so you look for things that they bump into and knock over.” Reines’s detector was a tank of cadmium chloride solution. The scientists calculated that a neutrino colliding with a proton in the liquid would produce a distinctive series of flashes. They found just such a signature.

But if Reines’s discovery was long hoped for, Perl’s came as a shock. In 1974, physicists thought they had completed their model of the fundamental constituents of matter. Then, in 1975, “Martin Perl spoiled it all”, says Burt Richter of Stanford.

Physicists had sorted the four known leptons and four quarks into two families (see Table). But Perl, who was working at the Stanford Linear Accelerator Center, found a heavy lepton. “That had to signal that there was a still another family,” says Richter. Theorists have now added the third family of particles, including the top quark.

No fourth family has yet been discovered, and if such a family does exist it will be very different from the first three. “I think there’s lots more out there,” says Perl “but not of the same kind. In some ways I’m as puzzled as ever.”

Ultrahigh-energy cosmic rays have a million times more energy than the most energetic particles created in accelerators on Earth: above 10¹⁸, electron volts. In 1966 three physicists proposed that above this energy level (the Greisen-Zatsepin-Kuzmin cutoff) cosmic rays would be energetic enough to interact with photons from the microwave background radiation – remnants of the big bang. In doing so, they would be slowed down to lower energies. So there would be 100 times fewer cosmic rays with ultrahigh energy from extragalactic sources than otherwise predicted.

But over the past two years, scientists have found more cosmic rays with energies significantly above the cutoff point than expected. These are among the highest energy cosmic rays that detectors can pick up.

Two complementary techniques have been used to find them. One looks for the tiny flashes of light that are created when cosmic rays hit the particles of the upper atmosphere. More than 100 spherical mirrors, each 1.5 metres across, capture images of the night sky and direct this light to detectors. The second approach detects the showers of charged particles that are created along with the light. These also create flashes of light that can be detected in an array of lightproof chambers.

If these cosmic rays were lower-energy particles that had been accelerated by exploding stars, then their numbers would tail off at very high energies. But Sigl and his colleagues say that the numbers actually increase at the highest energies. Alternatively, these cosmic rays may not have had the opportunity to interact with photons, which is possible if they came from a source less than 10 million light years away. “Unfortunately, there are no clearly identifiable nearby candidate sources,” says Sigl.

Sigl’s latest study suggests another possibility. “These ultrahigh-energy events might be produced by decay from some higher energy scale rather than by acceleration,” he says (Fermilab preprint, Pub 95/148A). This could create many more particles at high energies, which would explain the recovery in numbers.

One particle that could decay to produce ultrahigh-energy cosmic rays is the hypothetical particle “X” which may have existed shortly after the big bang. Theorists suspect that at that time there was just one force holding matter together. In the lower energies of today’s Universe, this manifests itself as the four fundamental forces of gravity, electromagnetism, and the strong and weak nuclear forces.

To test this theory, physicists try to reunite the forces at high energies in particle accelerators. They have already identified the force-carrying particle that unifies the electromagnetic force with the weak nuclear force. Particle “X” could do the same for the electroweak force and the strong nuclear force. Sigl and his colleagues say folds in the Universe – similar to black holes – could harbour “X” particles left over from the big bang. These may then escape from the fold and decay to produce ultrahigh-energy cosmic rays.

Researchers led by David Sugden of the University of Edinburgh have discovered ice in the Transantarctic Mountains that they believe is at least 8.1 million years old. They say that the southern ice cap did not collapse in the Pliocene – when world temperatures were similar to those we can expect by the middle of the next century, if today’s warming trend continues.

Earlier studies in the same area found the remains of beech trees, pollen and marine diatoms – algae found in plankton – apparently dating from three million years ago. Some researchers say that the ice sheet must have melted during the Pliocene, allowing trees to cover the mountains and diatoms to thrive in the seas. According to Sugden, however, the remnants of trees, pollen and diatoms can all be explained even if the ice remained frozen. “The trees could date from much earlier – perhaps as far back as 13 million years ago. And the pollen and diatoms could have been carried there by winds,” he says.

The glacier ice found by the team, which came from a layer that began just 50 centimetres below the surface, was dated by analysing the relative abundances of isotopes of argon in a thin layer of overlying volcanic ash. Radioactive isotopes decay with time, whereas non-radioactive isotopes do not. From the current proportions of the two, the researchers dated the volcanic ash at up to 8.1 million years old. As the ice lay below the volcanic ash, Sugden’s team concluded that it must be at least as ancient (Nature, vol 376, p 212).

If Sugden is correct, the Antarctic ice sheet is less vulnerable to warming than some scientists have supposed. This is reassuring, because if the ice cap did melt completely in the near future, it would raise global sea levels by 60 metres. Nevertheless, some scientists claim that ratios of oxygen isotopes in marine fossils from the east coast of the US indicate that the Antarctic ice sheet melted at least partially during the Pliocene.

Another possibility, says Dick van der Wateren of the Free University, Amsterdam, is that the layer of volcanic ash dated by Sugden’s team slid over younger ice. “There is still room for doubt,” he cautions.

Whatever the truth, Sugden and his colleagues have already put their discovery to good use. “We used a little bit of it in a gin and tonic,” he confides.

Stars emit light over a range of wavelengths. These spectra, however, contain powerful spectral lines – intense emissions at certain wavelengths caused when the electrons of a star’s atoms drop to lower energy levels and release this energy as light. The relative intensity of these lines – compared with the rest of a star’s spectrum – depends on the star’s mass and temperature.

Before reaching astronomers’ telescopes, a star’s light also has to travel through its corona. Previously, astronomers thought that this wispy atmosphere let light through unimpeded. But Karel Schrijver and his colleagues at the Lockheed Martin Palo Alto Research Laboratories in California, working with astronomers at Utrecht University and the Dutch Space Research Organisation, now say that coronas scatter light.

Schrijver and his colleagues made their discovery using NASA’s Extreme Ultraviolet Explorer satellite. They looked at eight stars and compared the amount of light emitted in the continuous part of their spectra with that in the spectral lines. For each star, there was less light in the spectral lines than predicted.

Coronal scattering will to a small extent affect all of the light emitted by a star. But the effect will be most marked at spectral line wavelengths, because the amount of scattering by the corona depends on the intensity of the light. “It is like a street light in a fog,” says Schrijver.

The good news, however, is that the scattering of coronal light may provide a new tool for looking at stellar winds – the flow of ionised gas from stars like our Sun. Large amounts of coronal material are needed to explain the effects observed by Schrijver and his colleagues.

Under the influence of a star’s gravity alone, this material would become so dense that it would be clearly visible. “The only way to avoid that is to spread the material out over a larger height,” says Schrijver. A stellar wind would do this job and the researchers have already used data on coronal light scattering to confirm the presence of a strong stellar wind above the star Procyon.

Schrijver and his colleagues will describe their work in a future issue of Astronomy and Astrophysics.

Catalytic converters break down 90 per cent of cars’ dangerous exhaust gases – carbon monoxide, nitrogen oxides and hydrocarbons. Hazardous in themselves, these gases can also react to create ozone, the main component of photochemical smog. Engelhard’s catalytic coating is designed primarily to treat the ozone resulting from the gases that pass through conventional converters unscathed.

The coating contains platinum, which breaks down ozone into oxygen, as well as promoting the oxidation of carbon monoxide and hydrocarbons. The paint may also contain palladium and rhodium, but Engelhard’s formula is secret for now.

The coating will be painted onto car radiators and air-conditioning compressors. These warm surfaces come into contact with a huge quantity of air when the vehicle is moving. According to calculations based on figures from the California Air Resources Board, if Los Angeles’ nine million vehicles were painted, they would treat a volume of air equivalent to that covering the area up to a height of about 5 metres.

Low level ozone is created in hot, sunny conditions, so levels peak just after midday, hampering the new approach. “Most of the vehicle miles per day in Los Angeles are driven in the morning and evening rush hours,” says Geoff Dollard, manager of atmospheric measurements and processes at the National Environmental Technology Centre in Culham, Oxfordshire. “These are times when the ozone concentration will not be at its maximum, so the overall efficiency will be reduced.”

Engelhard hopes to tackle this problem by equipping cars with an extra battery or solar cells. The fan that cools the engine could then carry on working when cars are parked and the catalyst would treat the air sucked through the engine. But Dollard is sceptical. “For the vehicle to effectively remove ozone while parked, the fan will need to be very efficient,” he says.

Engelhard has contracted two companies to test the system. One is conducting road tests, while the other is simulating the effect of the paint using environmental modelling. Early results suggest peak ozone levels could be cut by between 2 and 3 per cent and carbon monoxide concentrations reduced by over 12 per cent. The paint is expected to bump up the cost of a car by $500 and $1000. But Dollard believes that mopping up measures will never be as successful as those that reduce emissions at their source.