UK Contact: Claire Bowles

US Contact: New Scientist Washington office

AT THE heart of every galaxy lies a dark despot. Unseen by its minions, it holds sway over the billions of solar systems that stretch for thousands of light years beyond. It was there before any of them were born, and it is already shaping their future.

These tyrants are giant black holes, so heavy that they are called “supermassive” by astronomers. Ever since the existence of black holes was predicted early this century, they have seemed like outlandish things, absurd objects that offend all our ideas about how light and matter ought to behave. And even when evidence for real black holes in the Universe began to trickle in, they still seemed like curiosities, freaks. Intriguing and exciting, yes. Important? Not really.

But now astronomers are beginning to suspect that black holes have stamped their authority on the Universe. Earlier this year, researchers suggested that giant black holes were the seeds from which every galaxy in the Universe grew. Last week, more evidence emerged to support their claim. And over the past two years it has become increasingly clear that black holes shape galaxies, and might even create them. Without these strange by-products of Einstein’s imagination, we might none of us be here.

The first clue that supermassive black holes exist was the discovery several decades ago of quasars-extremely bright objects in the centres of distant galaxies. Quasars can be hundreds of times as bright as their surrounding galaxy, yet they are smaller than our Solar System. What could blast out so much light and radiation from such a small space? Black holes, for one. Though they are better known for swallowing light, they can also be a brilliant light sources. Material being sucked in by a black hole forms a spiralling disc around it, and as the disc churns the friction heats the gas until it is white hot. Astronomers believe that this is what makes quasars shine.

Quasars are rare, however. Even in their heyday, billions of years ago, there were hundreds of ordinary, unexciting galaxies for every one that harboured a quasar. So it seemed that giant black holes were freaks of nature, of no importance to common, everyday galaxies. “People thought they were something gone wrong in the centre of a galaxy, like an illness,” says Doug Richstone of the University of Michigan.

So when black holes began to show up in the centres of more ordinary galaxies, it was natural to assume that they must be burnt-out quasars. Astronomers first caught a glimpse of such a dark mass in 1978 in a galaxy called M87.

Then, in 1988, Richstone and his colleague Alan Dressler (now at the Observatories of the Carnegie Institution of Washington, based in California) looked at the spiral galaxy Andromeda and the small elliptical galaxy M32, both neighbours of our own Milky Way. There was no particular reason to think that these unremarkable galaxies harboured dead quasars, but because they are only a few million light years away, it was possible to look deep into their cores. The astronomers looked at the speeds of millions of stars in the centre of each galaxy, and found they were moving remarkably quickly-so fast that they ought to fly out of the centres of their galaxies. Unless, that is, there was a giant black hole in each galaxy holding the stars in its grip.

Could finding black holes in both these neighbouring galaxies just be good luck? Probably not. “When you trip over something on your own doorstep,” says Richstone, “it’s not reasonable to think that it should be special.” And sure enough, in the past few years the Hubble Space Telescope has found evidence for giant black holes in more than 20 nearby galaxies. One of them is our own: at the centre of the Milky Way is a hole with a mass 3 million times that of the Sun.

Indeed, it is beginning to look as though all galaxies have a hole at their core. Rather than a rare and unfortunate illness, supermassive black holes could be an integral part of every galaxy. If so, what part do they play in galactic life? In January, at a meeting of the American Astronomical Society, Richstone put forward a provocative idea: giant black holes might create galaxies in the first place.

Richstone came to this conclusion after comparing the history of quasars and of star formation in the Universe. He noticed that most quasars appeared perhaps a billion years before most stars. This might mean that black holes control the formation of galaxies, an idea that turns conventional astrophysics on its head.

Most astronomers assume that each galaxy creates its own black hole. Here’s how it might work: a very massive star ends its life in a supernova explosion, leaving behind a middling-sized black hole that weighs just a few times as much as our Sun. It feeds on a rich supply of gas near the galactic centre, and grows like a pearl in an oyster. It might also swallow stars and merge with small holes from other supernovae. But for giant black holes to form this way takes a long time. If they existed before most galaxies, this can’t be right.

Smooth start

Instead, according to Gordon Garmire of Pennsylvania State University, giant black holes might have formed at the dawn of time, and been the seeds around which galaxies grew. But why would galaxies need such seeds? The reason is that the early Universe was very smooth. The leftover glow from the big bang-the so-called cosmic microwave background-shows that the density of the early Universe didn’t vary from place to place by more than about 1 part in 100 000.

To produce our Universe of galaxies and voids, these small density fluctuations must have been amplified enormously. And the amplification must have happened very quickly, as the first galaxies existed a mere billion years after the big bang. “That’s not much time for the Universe to go from smooth to rough,” says Garmire. So he suggests that giant black holes might have acted as gravitational seeds, pulling a mantle of matter around themselves that went on to form stars. In other words, a galaxy.

In that case the supermassive holes must have formed very early indeed, not long after the big bang. It’s an astonishing idea. Stephen Hawking has suggested that tiny black holes weighing only as much as a mountain might have formed in the first instant of the Universe (see p 24). But Garmire is talking about holes a trillion trillion times bigger.

And yet black holes of about the right size could fall out of a cosmological theory called supernatural inflation, developed a few years ago by Alan Guth of the Massachusetts Institute of Technology and his colleagues Lisa Randall and Marin Soljacic. According to the theory, quantum fluctuations in the first fraction of a second after the big bang naturally produce the gentle bumps and dips we see in the microwave background. But they would also have made some small regions of space especially dense. Within a few years, these regions would have been compressed by gravity and collapsed to form black holes. The masses of these black holes aren’t tightly constrained by the theory, but Guth says they would be some fraction of the mass of a galaxy, “perhaps a tenth, perhaps a thousandth”. If they are towards the bottom of this range, and if enough are produced, these could be the black holes that power quasars. And they could be Garmire’s gravitational seeds. “Our theory is on backburner,” says Guth, “but we may want to get back onto it.”

However, not everyone is convinced by these arguments. Despite Richstone’s graph, many astronomers don’t even believe that quasars came before galaxies. “Uncertainties in these histories are very great because of the obscuring effects of dust,” says Martin Rees of Cambridge University. The star-formation history used by Richstone is based on optical observations, and if early galaxies were much dustier than their modern counterparts, as many theorists believe, then their optical emissions could all be absorbed by dust. An instrument called SCUBA, on Britain’s James Clark Maxwell Telescope in Hawaii, appears to confirm this possibility. It looks at infrared light emitted by dust, which reveals signs of many more early galaxies than the optical observations do.

But although Richstone admits that there is room for doubt, he is sticking to his guns. He thinks that dustiness would affect both quasars and stars in a similar way, so that comparing their histories would still give the same answer. “If you ask me, do quasars really come before galaxies-well, I’d bet a hundred dollars on it at three to one. I just wouldn’t bet anything I couldn’t afford to lose.” And he has his supporters. “Richstone is very probably right,” says Virginia Trimble of the University of California at Irvine.

Bright sparks

NASA’s new Chandra X-ray Observatory has lent him some support, too. This week in Nature (vol 404 p 459), a group of astronomers published observations by Chandra that finally explain the cosmic X-ray background discovered in 1962. This high-energy glow turns out to be made up of little sparks from about a hundred million X-ray sources. Among them are ordinary quasars, as well as many unfamiliar objects that could be quasars obscured by dust. There’s also a third type of source that the Chandra team call “optically faint objects”.

The optically faint objects might be very distant quasars whose ordinary light has been absorbed by intergalactic gas so that only their X-rays get through, suggests Richard Mushotzky of NASA’s Goddard Space Flight Center in Maryland, the head of the group. These would be quasars in their glorious youth, at a time before most galaxies formed.

But even if black holes don’t turn out to be so venerable, they could still have had profound effects on galaxies. About two years ago, Richstone and Dressler teamed up with John Magorrian of the University of Toronto, John Kormendy of the University of Texas and eight other astronomers to pull together all the known information about black holes in nearby galaxies.

They compared the mass of each hole with the mass of the blobby part of each galaxy. Galaxies come in two broad types, spirals and ellipticals. A spiral, such as the Milky Way, looks like a fried egg, with a flat white disk full of gas and new stars and a yellowish yolk of old stars in the middle. The yolk is called a bulge, or spheroid. An elliptical galaxy is just a big blob of old stars-a separated yolk.

When Magorrian and co. plotted the mass of each black hole against the mass of its bulge, they found that the points fell on a straight line. From little spiral bulges to giant elliptical galaxies, the central black hole always has a mass about 1 per cent that of its surrounding bulge. “There seems to be an intimate link between the two masses,” says Richstone.

But why should the sizes of the bulge and the central black hole be so closely connected? In 1998, Rees and Joe Silk, then both at Cambridge University, suggested how this might happen. The radiation from a young quasar should drive a wind of charged particles out into the surrounding galaxy. As the black hole swallows more and more material and grows steadily larger, the quasar should get brighter and the wind stronger. Eventually, the wind will be so strong that it overcomes the gravity of the galaxy and blows all the gas away. With its gas supply cut off, the hole will stop growing, and so will the galaxy. Silk and Rees worked out that the hole would have to grow to around the size given by Magorrian’s mass relation before it calls a halt to growth.

Before this happens, the young quasar could have other influences. “In the first 100 million years of its life,” says Richstone, “the quasar can dominate the energy output of its galaxy.” And, by a rather complicated route, all that radiation could help trigger star formation.

Clouds of gas have to cool down before they can collapse to make stars. But the gas in early galaxies would be mostly hydrogen, and hydrogen atoms tend to hang on to their thermal energy. On the other hand, hydrogen molecules, made of two hydrogen atoms, are good emitters of heat. A blast of radiation from a quasar could turn some hydrogen atoms into ions, which would then react to form molecules and radiate their energy, speeding cloud collapse and star formation.

Quasars might stir up their galaxies, too. They squirt out powerful jets of high-speed matter, which could sweep around the galaxy, generating shockwaves that compress the surrounding gas. That compression could also encourage stars to form.

Finally, a big black hole could change its galaxy’s shape. James Binney of Oxford University worked out back in the 1970s that most elliptical galaxies ought to be an odd shape, with one long axis, one short axis, and one of intermediate length. They would look a bit like a watermelon seed, or a squashed rugby ball. But later observations showed that most of them are more symmetrical-flattened spheres, like a chocolate M&M.

In 1996, David Merritt and Tema Fridman at Rutgers University in New Jersey came up with an explanation. In a seed-shaped galaxy, stars travel on elongated orbits that plunge very close to the centre, like comets in our Solar System. Merritt and Fridman calculated that a big central black hole in the centre disturbs these orbits, making them unstable. So such a galaxy would soon settle into a more stable flattened sphere. “It’s incredible that this tiny pip in the middle forces the whole enormous object to change shape,” says Richstone.

Indeed, it’s incredible that a black hole could have any of these powers-that something smaller than our Solar System could control a vast island universe of billions of stars. But then, think of Alexander, Attila, Napoleon… Human dictators have a famous tendency to be diminutive. Perhaps it is a universal law.


Further reading: Gravity’s Fatal Attraction: Black Holes in the Universe by Mitchell Begelman & Martin Rees (W. H. Freeman)