At first, Shri Kulkarni didn't think it was a big deal. It was the middle of the night in September 1982, and he was at Arecibo Observatory in Puerto Rico, using the enormous radio dish to hunt for pulsars: the ultra-dense, rapidly spinning corpses of massive stars.

He had just detected his first pulsar, and it was rotating really fast - once every 1.5 milliseconds - which was more than 20 times faster than any known at the time.

For Kulkarni, who was still a graduate student then, the rapid rotation didn't mean much. It was just a fast pulsar, he thought. He called his project advisor, the late Don Backer, an astronomer at the University of California, Berkeley, US and delivered the news.

"There was a long silence," Kulkarni recalls. Probably because Backer knew this was big.

He reminded Kulkarni what such a fast pulsar meant: this was an object spinning 641 times per second. "Many people thought that pulsars going at that speed would break apart," says Kulkarni, now an astronomer at the California Institute of Technology in the US. Pulsars are as big as a city - about 20 km in diameter - and the assumption was that if it were rotating that fast, the centrifugal force would rip it to smithereens.

But now Kulkarni's discovery upended that assumption. It changed not only his burgeoning career, but also an entire field. The pulsar, known as PSR B1937+21, became the first of a new class of remarkable objects called millisecond pulsars.

Not only are they fast, but they also spin with such amazing regularity that they're among the most accurate clocks in the universe. Using these celestial timekeepers, astronomers are answering questions about stars, matter - and even space and time itself - that would otherwise be impossible.

Extreme objects

Even ordinary pulsars are extraordinary. They're some of the universe's most extreme objects, the remains of stars between about eight and 20 times as massive as the sun. When such a star burns up its fuel and dies, it explodes in a supernova, blowing off its outer layers of gas.

What's left is a core so dense that its electrons have fused with protons, forming a solid sphere of mostly neutrons. It's become a neutron star. These objects squeeze between about 1.2 and 2 suns worth of mass into a ball no more than 20 km in diameter. Just a teaspoonful weighs a trillion kilograms - comparable to the mass of every person on Earth.

Such density means the gravity on a neutron star's surface is extremely strong - 100 billion times greater than Earth's. If you tried standing on a neutron star (ignoring the million-degree temperatures, of course), you'd be squished, your atoms smeared across the surface. In fact, this overwhelming gravitational pull prevents the formation of any bumps greater than a few centimetres high, giving neutron stars some of the smoothest surfaces in space.

And then there are the magnetic fields, the most powerful in the universe. Even the weakest is a hundred million times stronger than Earth's - strong enough to warp the structure of an atom. At the poles, a neutron star's magnetic field accelerates charged particles - positrons and electrons stripped off the surface by powerful electric fields - and blasts them into space in the form of jets. Those particles produce beams of radiation at radio frequencies, which eventually reach radio telescopes on Earth.

It's these beams that give pulsars their namesake. When a neutron star rotates rapidly, it swings these beams around like a lighthouse. From Earth, it appears as a steady, pulsating signal, sometimes as slow as once every 10 seconds.

But they start out faster. They were cranking up the speed before they were pulsars, when they were stellar cores. As a star runs out of nuclear fuel, it can't maintain the pressure needed to hold itself up, and the core contracts due to its own gravity.

Like the way ice skaters spin faster when tucking their arms in, the core of a dying star rotates faster as it collapses. By the time the star dies and you're left with a neutron star, it can be spinning as fast as 100 times a second. Over time, its rotating magnetic field loses energy, which slows the pulsar down.

Which is why Kulkarni's discovery of a pulsar going so much faster was so astounding. To whip it up to such speeds, astronomers realised, a pulsar must receive help from a companion star in orbit. As the companion exhausts its fuel, it swells - as all stars do eventually - and its outer layers start to spill onto the pulsar, forming a disk of hot gas spiraling inward like water circling a drain. The swirling disk spins up the pulsar.

The discovery of millisecond pulsars revitalized a moribund field, which started in 1967 when Jocelyn Bell discovered the first pulsar. The field's landmark discovery came in 1974, when Russell Hulse and Joseph Taylor found two pulsars spiraling in toward each other. For that to happen, the energy of the pulsar's orbits must be dissipating in the form of gravitational waves, ripples in the fabric of space-time.

Their measurements were the clearest evidence yet that these waves exist, confirming a prediction of Einstein's theory of general relativity; they would later win the Nobel Prize in 1993. "That was the one highlight of the field," Kulkarni says. It seemed all that was left to do was find more pulsars. "By 1982," he says, "there was a sense that everything about pulsars had been discovered."

Cosmic laboratories

That changed when Kulkarni found the first millisecond pulsar. Since then, astronomers have identified about 300 more. They estimate that the Milky Way Galaxy is home to 20,000 millisecond pulsars, and about an equal number of regular pulsars - a meagre number compared to the galaxy's hundreds of billions of stars. PSR B1937+21 held the speed record until 2006, when Jason Hessels - who, like Kulkarni, was a graduate student at the time - discovered Terzan 5ad, a faint pulsar that spins 716 times per second.

With such high speeds and masses - lots of angular momentum, in physics-speak -millisecond pulsars are hard to slow down. That makes them incredibly consistent over a long period of time. When millisecond pulsars were first discovered, they rivaled the stability of atomic clocks. Today, atomic clocks have surpassed pulsars in accuracy. But if you were to compare them over a longer period of time - say, decades - pulsars can be just as good, says Hessels, who's now at the University of Amsterdam in the Netherlands. Even after billions of years, a millisecond pulsar may slow down by only a few milliseconds. But because astronomers can precisely pin down its rate of deceleration, they can compensate and still use them as clocks.

Millisecond pulsars are so stable that astronomers have measured their spin periods to an accuracy of one part in a million trillion (that's 18 decimal places). They know when a pulse arrives on Earth to a precision of 100 nanoseconds. Because the pulses are so reliable, the tiniest deviations can reveal with great detail what's going on in and around the pulsar - and in the space between the stars.

In this space is dust and gas, called the interstellar medium, which obstructs and scatters a pulsar's signals. By measuring the pulses' delay, their intensity, and how sharp they are, astronomers can probe the properties of the interstellar medium, which plays a key role in how stars and galaxies form and evolve.

Around the pulsar is the companion star that helped speed it up. The size of the star and how it evolves over time - for example, how changing magnetic activity can alter its shape - influences its orbit. Delays, modulations, or other variations in the pulses reveal what the companion star is like and how it interacts with the pulsar.

Thanks to the precision of these pulses, astronomers can detect even the most subtle gravitational tugs. In 1992, astronomers discovered a planetary system orbiting a millisecond pulsar - the first planets found outside the solar system. The gravity of the planets were causing the pulsar to wobble ever so slightly, changing the arrival times of the pulses. In the case of Kulkarni's pulsar, PSR B1937+21, these kinds of timing variations have recently suggested the presence of objects as small as asteroids.  

Detecting those pulses of radio waves - and, in some cases, X rays and gamma rays - is crucial because it's often the only way for astronomers to observe and study these exotic pulsar systems. It's also one of the only ways to study the weird structure and composition of the pulsar itself.

Pulsars are essentially giant atomic nuclei. They can have a thin atmosphere not much more than 10 cm thick made of helium, hydrogen, and carbon, and an outer crust that's mostly iron. As you go deeper, the matter becomes denser, full of neutrons (and some protons and electrons) in increasingly exotic forms, merging together to form strands and even sheets. But no one really knows what it's like inside.

Millisecond pulsars offer clues. The pulses allow scientists to precisely determine the pulsars' orbits and thus their masses - crucial data that theorists need to constrain and devise new hypotheses. Nowhere in the universe can you find matter at such high densities and pressure. For physicists, pulsars are like laboratories for exploring such extremes - and maybe discovering entirely new types of matter.

"It's almost miraculous that there's this type of star that's so useful for testing areas of physics that would otherwise be inaccessible," Hessels says.

Testing Einstein

Those areas include gravity itself. Einstein's theory of general relativity describes gravity as bends and curves in the fabric of space-time, and so far, its predictions have been proven true again and again. But the theory may work differently in the enormous densities and strong gravity of pulsars—as strong as you can get without becoming a black hole. To find out whether that's the case, researchers can look for discrepancies in the pulses.

Recently, Hessels was part of a team that discovered a millisecond pulsar in a triple system with two white dwarfs—the remnants of stars not massive enough to form neutron stars. This rare configuration gives scientists a way to test one of the hallmarks of relativity: the equivalence principle.

The principle says that gravity is the same for everyone and everything. Perhaps the most dramatic example is when astronaut Dave Scott dropped a hammer and a feather on the moon in 1971. Both hit the lunar surface at the same time, showing that the moon's gravity pulled on both equally. Likewise, researchers want to see if the gravity of one of the white dwarfs pulls on the pulsar in the same way as the other white dwarf. They haven't done the experiment yet, but the researchers say it could be the most accurate test ever of the equivalence principle.

Of course, no one has found Einstein to be wrong just yet. One of the most successful confirmations of relativity was the Hulse-Taylor binary pulsar system, the big pre-millisecond-pulsar discovery that proved gravitational waves were real. Still, the evidence was indirect, based on measurements of orbits that allowed Hulse and Taylor to infer the existence of gravitational waves. To this day, a direct detection remains elusive.

That's despite the efforts of ground-based experiments such as LIGO, the Laser Interferometer Gravitational-Wave Observatory, which is designed to detect gravitational waves from colliding neutron stars or black holes. Its first observing run between 2002 and 2010 turned up nothing. After significant upgrades, it's set to start up again in the fall of 2015.

Meanwhile, an international effort has been racing to beat LIGO using - you guessed it - millisecond pulsars. "The idea is to use them as a galactic GPS," says Hessels, who is part of the European contingent. When gravitational waves pass through Earth, the planet bobs like a buoy on the water. Those tiny motions alter the arrival times of the pulses.

Over the last few years, astronomers have continued to refine their techniques, meticulously timing a few dozen of the best cosmic clocks known. And they hope to see something soon. "There's a reasonable prospect of detecting gravitational waves in this way in the next five years or so," says Ingrid Stairs, an astronomer at the University of British Columbia in Canada and member of the North American team.

Still, Stairs thinks LIGO probably will beat them to it. But while LIGO is designed to detect waves from merging neutron stars and black holes several times as massive as the sun, the pulsar method is sensitive to collisions between supermassive black holes, which are millions to billions of times heftier than the sun. "It's looking at a totally different source of gravitational waves," she says. "Even if we're later than LIGO, it doesn't mean they've totally scooped us."

Regardless of who wins the race, the millisecond pulsar has been vital for understanding a range of cosmic phenomena. "It's nature's gift to us," Kulkarni says. "It's a precise, physical laboratory - but in the heavens." It was a gift received more than three decades ago, and if it didn't seem like a big deal then, it certainly does now.