When John Mather entered the auditorium, he was stunned by the sight that greeted him. He had expected about 50 people to turn up for his talk. Instead, it was standing room only and more than 1,000 had packed into the lecture hall.
It was 13 January 1990, and COBE had been up in space just six weeks. The American Astronomical Society was meeting in Crystal City, Virginia, and Mather had come to present COBE’s first result – a spectrum of the microwave background based on just nine minutes of looking at the sky.
Mather was determined to remain calm. He launched into the five-minute talk he had prepared, explaining the purpose of the experiment and proceeding to describe it. Finally, he put a transparency onto an overhead projector so that its image was thrown onto a large screen.
‘Here is our spectrum,’ he said. ‘The little boxes are the points we measured and here is the black body curve going through them. As you can see, all our points lie on the curve.’
‘At first, you could hear a pin drop in that hall,’ says Bruce Partridge. ‘Then there were murmurings in the audience. Next people began to applaud. Then they got to their feet, clapping wildly, enthusiastically.’1
‘I’ve never seen anything like it at a scientific meeting,’ says Charles Bennett. ‘Not before or since.’
Up there on the screen was the most perfect black body spectrum anyone had ever seen. Not a single measured point deviated by more than 1 per cent from the mesmerising curve drawn through them.
‘It was a wonderful moment,’ says Partridge. ‘The spectrum was absolutely spectacular. There had been rumours that it was going to be impressive, but the COBE team had been very good at keeping it a secret.’
Mather’s immediate reaction to the audience’s applause was not pleasure but embarrassment. ‘I was afraid they were clapping for me,’ he says. ‘I wanted to tell them I wasn’t the one that did this thing. COBE was a team effort. I played a part but thousands of other people worked on it day and night. They left behind their families just to do it.’
But Mather need not have worried. The people were not cheering for him alone. They were applauding a wonderful experiment. They were cheering because no one in that lecture hall had ever seen such perfection emerge from an experiment. Nature was simply not like that. It was messy.
COBE had seen to the very heart of things. It had stripped away all the bewildering complexity of the Universe. And there at the beginning of time was breathtaking simplicity – more beautiful than anyone had dared imagine.
‘A lot of that cheering was relief,’ says Mather. ‘The scientists were relieved that the Universe was the way everyone had hoped.’
There was no sign in the spectrum of the bump found by the Berkeley–Nagoya team. ‘It seemed that every issue of the Astrophysical Journal had three papers speculating on what caused it,’ says Bennett. ‘But none of that complicated stuff happened.’
There could have been no large release of light radiation into the Universe from the decay of microscopic particles or the explosion of an early generation of stars. Almost all of the cosmic background radiation had come straight from the Big Bang.2
The early Universe could have been complicated. Its temperature and other properties might have varied wildly from place to place. But they didn’t. The early Universe was unbelievably simple. All you needed to know was one number – its temperature – and you knew everything there was to know about it.
Not everyone who was anyone in background work was at the meeting at Crystal City. Dave Wilkinson, for instance, was back at Princeton giving a simultaneous talk on the COBE spectrum. Another notable absentee was Robert Wilson.
The irony was that the co-discoverer of the Big Bang radiation had attended the Crystal City meeting but had decided to go home a day early. And nobody on the COBE team had thought to tip him off. When Wilson finally saw the spectrum, he was bowled over by it like everyone else. ‘It was just marvellous,’ he says. ‘I never believed I’d see a spectrum that good. To my mind, it puts an end to the argument about whether this is really from the early Universe or not.’
COBE had actually beamed down the spectrum in early December, shortly after the satellite was launched. But the COBE team had kept it a secret. ‘The pressure on these guys was tremendous,’ says Partridge. ‘Everyone knew that if everything worked, once the probe was up and the cover was off the instruments they’d know within ten minutes what the spectrum was like.’
The reason the COBE team kept the spectrum under wraps was that its members had an agreement: no one was to talk about any result until everyone was good and ready. This would enable the team to check and recheck a result to make absolutely sure there was no mistake. There would also be time to prepare a rigorous scientific paper before any announcement.
Dave Wilkinson remembers his first sight of the spectrum. It was on a computer screen at Princeton. Ed Cheng, the team member who had generated the spectrum from the raw data, had sent it to him by electronic mail.
‘Seeing that spectrum after 25 years of knocking off one point at a time was just thrilling,’ says Wilkinson. ‘Each of those points had taken a graduate student’s thesis.
‘Everything on the satellite worked perfectly. After all our bitter experience with balloons, it was just amazing. That complicated thing actually worked!’
Wilkinson’s office at Princeton is next door to those of Peebles and Dicke but, because of the team’s publication policy, he was unable to show either of them the amazing spectrum. For nearly six weeks he drank coffee with Peebles and Dicke without ever spilling the beans.
‘I finally showed it to Jim a few days before the official announcement,’ says Wilkinson. Peebles was not totally surprised at the spectrum Wilkinson showed him. ‘Dave was walking round with an “Oops, I swallowed a canary” grin, so I could tell that it looked awfully good.’
But though he expected to see something good, Peebles was still not prepared for the sight of such a perfect spectrum. ‘Dave had been carrying the spectrum about in his pocket for some time,’ he says. ‘When he finally got it out and showed it to me, it had all the drama of “Take a look at this!”
It was one of those stunning moments in your life you remember for ever.
‘The COBE team kept it under cover until they were absolutely sure of the result. That shows a degree of care that you don’t normally see with scientists. Usually, they are in a hurry to get into print.’
Peebles admits he never expected to see such a perfect spectrum. ‘In the real world, when you measure any quantity in nature, there are always errors – the measurements “scatter” about the real value,’ he says. ‘The stunning thing with the spectrum was that the scatter was so small.’
Wilkinson had also been surprised when he first saw the spectrum that it was so perfect. ‘A lot of us were expecting to see the Berkeley–Nagoya distortion,’ he says. He had grilled Andrew Lange, who had worked with Paul Richards at Berkeley, but had been unable to pinpoint anything the team had done wrong.
‘They were very careful,’ says Partridge, ‘but they were simply trapped by nature.’
‘It was very hard to create that distortion,’ says Wilkinson. ‘We knew that if it was right, we’d need to invent some new physics or put a fairly dramatic chapter into the story of the Universe.’
Peebles never believed in the Berkeley–Nagoya distortion, something of which he is proud. At a meeting on the microwave background just before COBE, he remembers that people discussed the distortion at length. ‘But none of the theorists at the meeting had a convincing explanation for that effect,’ he says. ‘This makes me feel good, because the effect wasn’t there after all.’
Few thought that all this hard theorising was wasted, though. ‘Their result generated a lot of thinking about what could cause the distortions,’ says Wilkinson. ‘It was a very useful exercise.’
Partridge agrees. ‘It played the same sort of role as the steady-state theory,’ he says.
Mather had more faith in the experiment – and in nature. ‘The spectrum was pretty much what I thought it would be,’ he says. ‘The cosmic background radiation really dominated in the early Universe. For every particle of matter there were 10 billion particles of light. If you’re going to make them not be perfect, every particle of matter has got to do multiple duty. It’s hard to imagine how that would work.’
The COBE spectrum was widely referred to as the best black body ever seen in nature. But the COBE team itself was not prepared to go that far. ‘All COBE did was compare the sky with the best black body we could make,’ says Wilkinson. ‘All we proved was that the Universe is the same as our black body.’
The fact that it was the same is why the COBE team is so confident in the result. ‘If there was anything wrong with the experiment, we wouldn’t expect the sky to be like the cold load,’ he says. ‘It would be an incredible coincidence if the cold load mimicked the sky and neither were black bodies!’
If COBE had detected some kind of distortion, it would have been another story completely. ‘It would have been much longer before we told people,’ says Bennett. ‘We’d be wondering, “Is that really in the sky or is there something wrong with the cold load?” It’s because it is unlikely the sky and the cold load would match by accident that we have a great deal of confidence in the spectrum.’
The COBE team continued to measure the spectrum of the Big Bang radiation more and more precisely. ‘So far we’ve found no deviation greater than a thirtieth of a per cent,’ says Wilkinson. The cosmic background is a true black body with a temperature of 2.725 degrees above absolute zero, with no deviations greater than a thirtieth of a per cent of the peak.
Although the COBE team was confident that its spectrum was right, what was needed was confirmation by another experiment. As it happened, the confirmation would come sooner than anyone expected, from Herb Gush, Mark Halpern and Ed Wishnow at the University of British Columbia.
‘Gush is the unsung hero of cosmic background work,’ says Bruce Partridge. ‘For years he’s worked on a shoestring budget with just a handful of people.’
As mentioned earlier, during the 1970s Gush developed the technique of launching cosmic background experiments on sounding rockets. Sounding rockets basically roar up a few hundred kilometres, then plummet back down as soon as their fuel is used up. For a few minutes the instruments on board get to take a peek at the Universe from the very edge of space. In principle, when they are above the Earth’s atmosphere they can do better than instruments on a balloon that drifts on the winds for ten hours.
Gush pioneered measurements of the spectrum from rockets. His first flights were in the early 1970s, but they were plagued by problems, the most serious of which were caused by the rocket exhaust.
Rockets are very messy beasts, and all sorts of complicated molecules spew from their exhausts. ‘Unless you’re very careful, you end up looking at the Universe through a thick cloud of smoke,’ says Wilkinson.
Gush thought he had the exhaust problem solved. Along with his instrument package he included a sort of ‘ejector seat’. It was supposed to blow the experiment clear of the rocket when the right altitude was reached. But things did not go as he had hoped.
On Gush’s fourth rocket flight in 1978, the ejection mechanism proved to be too feeble. ‘The payload blasted free of the rocket all right,’ says Gush. ‘But, as it sailed on, the rocket overtook it, still burning the last of its fuel.’ During the seven minutes the experiment was above the atmosphere, it observed the background radiation through a veil of shimmering exhaust fumes.
A spectrum was radioed down to the ground from a height of 300 kilometres. It was like a black body for the most part. But at millimetre wavelengths there was a large bump. Was the bump really in the background radiation or did it come from the glowing rocket exhaust? It was impossible to tell.
In 1980, when Gush started designing his fifth rocket experiment, more bad luck and frustration were just around the corner. Until now, he had been firing his rockets from a launch pad at Churchill, Manitoba. The Canadian government ran the facility jointly with the Americans, but in the early 1980s decided to pull its money out. It would be nearly a decade before Gush would fly a rocket again. When he did, it would not be from Canada at all but from the desert of northern New Mexico.
In September 1989, two months before COBE was due for launch, Gush, Halpern and Wishnow were almost ready to go. First, though, they needed to be certain their instrument package would survive the violent vibration of a rocket launch. They took it to Bristol Aerospace in Winnipeg for a ‘shake test’. It failed.
‘Only later did we find out that the engineers at Bristol Aerospace had shaken the instrument package too vigorously!’ says Gush. Some things had broken loose. There was nothing to do but go back to Vancouver and start repairing the damage. ‘The extra work cost us five months,’ says Gush. While the team worked in its lab, COBE was launched and began observing the microwave background.
Finally, in January 1990, Gush was ready to launch. He took the experiment down to White Sands missile range in New Mexico. It was a facility run jointly by the US Navy and Army.
On the launch pad, the two-stage rocket stood more than 40 feet high, glistening in the morning light. Gush’s instruments were crammed into the nose cone, a cylindrical space just three feet high and 17 inches in diameter.
As the countdown began, Gush sat in an underground bunker close to the launch tower. It was designed to provide protection if the rocket exploded and burning metal and fuel rained down from the sky.
The countdown reached zero and the rocket whooshed into the blue New Mexico sky on a column of flame. Minutes later, it reached an altitude of 300 kilometres and the instruments were ejected successfully. Sensitive detectors, cooled by liquid helium to just a third of a degree above absolute zero, came alive as the radiation from the Big Bang poured in.
Everything worked perfectly. After 20 years of failed experiments, Gush had finally done it.
On his way back from the rocket site, Mark Halpern stopped off in Aspen, Colorado, where a meeting on the microwave background was in progress. It was just a couple of weeks after Mather had received his standing ovation in Crystal City.
‘Halpern brought with him a beautiful black body spectrum,’ says Wilkinson. ‘It was stunning.’
Gush’s team had achieved what hundreds of other experiments had tried and failed to do ever since Penzias and Wilson discovered the background radiation in 1965. And he had done it only weeks after COBE had cleaned up the field. ‘If it hadn’t have been for COBE, it would have been that spectrum that got a standing ovation,’ says Wilkinson.
‘They tried and tried again, and finally they got it right,’ says Mather.
‘My heart goes out to Herb Gush,’ says Peebles.
‘I suppose they knew they had only one more flight, so they were really careful,’ says Wilkinson.
‘These two experiments were running for more than a decade each, and yet by coincidence they came to fruition at almost exactly the same time,’ says Peebles. ‘It would have been a dramatic triumph for Herb if he had got the spectrum first. But then one measurement had to be made before the other. And one had to be the confirmation.’
The parallels with Roll and Wilkinson were hard to avoid. In 1965, they, too, had succeeded in making an epoch-making measurement of the cosmic background – but only after being scooped.
But Gush had done the community proud. ‘I think the important thing is that it was an almost instantaneous confirmation of the COBE spectrum,’ says Peebles. Now, nobody could really doubt that the radiation from the beginning of time was a perfect black body.
If there was a Nobel Prize for persistence, Herb Gush would have won it.
1. Dan Quayle, the then US vice president, gave a major address on NASA space policy at the Crystal City meeting. But his audience and the applause he received were nowhere near as impressive.
2. In fact, COBE would eventually find that the background spectrum differed by less than 0.03 per cent from a perfect black body at a temperature of 2.726 degrees above absolute zero. This implies that 99.7 per cent of the cosmic background energy was released within one year of the Big Bang.