In the spring of 1967, Dave Wilkinson and Bruce Partridge turned their rent-a-truck through the gates of a US army base in Yuma, Arizona. Back in Princeton, their experiment to measure the smoothness of the Big Bang radiation had been hampered by clouds of water vapour high in the air above New Jersey. After redesigning it, they had headed south-west to the place they had determined was the sunniest spot in the United States.
‘The US army loaned us an area of desert surrounded by a high fence,’ says Partridge. ‘It even gave us active ranks so we could use the officer’s club during our stay. Dave was a captain and I was a lieutenant.’
The site was perfect for the Princeton team’s purpose. It was sunny and dry, and the high fence stopped any animals that might be around from trampling over the equipment. The site had only one drawback: ‘It was where the army put their nerve-gas shells to see if they would leak in desert conditions,’ says Partridge.
Under a china-blue sky, surrounded by racks of nerve-gas shells gently roasting in the Sun, Wilkinson and Partridge went to work setting up their equipment. Neither of them gave too much thought to the danger. ‘If certain symptoms appeared, we were told to slap on our masks and get out of the area as fast as possible,’ says Partridge.
The electronics for the experiment went in a garden hut they had bought from Sears-Roebuck. The microwave horn was stuck out in the desert, pointing down at the ground rather than up at the sky. ‘If we’d left the horn pointing up, it would have soon filled up with dead bugs and water from condensation,’ says Partridge. A metal mirror placed underneath the horn ensured that radio waves from the sky were reflected into the horn’s flared opening.
For several weeks, Partridge and Wilkinson worked among the nerve-gas shells, riding about the desert on a moped they had bought with some of their research money. ‘The moped was a lot cheaper than a rent-a-car,’ says Partridge. ‘It saved the taxpayers a lot of money.’ The experiment was designed to be automated, so when they had finished setting things up they left it to chug away on its own in the desert.
Things did not go right. ‘The Yuma experiment ran for a year,’ says Partridge, ‘but it was a complete failure.’
It turned out that when the mirror switched the horn from looking at one part of the sky to looking at another, the horn saw different temperatures all right. But the temperature difference was not in the cosmic background radiation. It was more local than that. Different parts of the sky contained different amounts of water vapour, and the effect of this unevenness completely overwhelmed any variation there might have been in the radiation from the beginning of time.
‘When Dave and I had sat down with weather records back in Princeton, we’d figured out that Yuma, Arizona, was the sunniest place in America,’ says Partridge. ‘We were absolutely right – it was sunny. But there was still plenty of water vapour hanging in the air. It just didn’t show up as visible clouds.’
The failure of the Yuma experiment highlighted just how difficult it was to do cosmic background experiments beneath the thick blanket of the atmosphere. ‘Our experiments to probe the Big Bang would continually be defeated by the cussedness of nature,’ says Partridge.
It would become a depressing pattern over the next two decades.
Moisture hanging invisibly in the air proved to be the bane of cosmic background experiments. It was particularly troublesome when observations were made at short wavelengths of a few millimetres or less. Here, water vapour glowed so brilliantly that it overwhelmed the precious cosmic background radiation. Unfortunately, it was precisely at these short wavelengths that astronomers most wanted to observe the afterglow of creation.
The reason was that a black body at a temperature of about three degrees above absolute zero had a peak in its spectrum at a wavelength of around a millimetre. To prove once and for all that the cosmic background radiation really had come straight from the Big Bang astronomers would have to prove that its spectrum was black-body-shaped. In practice this meant looking for the peak and showing that beyond the peak the spectrum fell away sharply.
At a wavelength of about a millimetre, water vapour and other molecules in the atmosphere glowed fiercely. Beyond the peak, at ‘sub-millimetre’ wavelengths, the situation was even worse. Not only would water vapour be shining brightly, but the cosmic background radiation itself would be getting rapidly fainter. The tiny signal from the background would be completely swamped.
Observing the cosmic background at the peak of the spectrum and beyond was a formidable problem. The only solution was to take an instrument to high altitude and get above as much of the obscuring atmosphere as possible. As anyone who climbs up a mountain knows, it always gets colder as you get higher. If you go high enough, it gets so cold that the water vapour in the air turns to ice and simply drops out as snow.1
In their quest to steal a march on the atmosphere, researchers would go from deserts to mountaintops to high-flying balloons, spy planes and rockets. And, finally, one day they would even go into space.
After the Yuma debacle, Wilkinson and Partridge turned their attention back to measuring the spectrum of the cosmic background radiation and to confirming that it was indeed a black body. They had learnt a painful lesson in the desert of south-west Arizona. They would not make the same mistake again.
For the site of their new experiment, they selected the summit of White Mountain in northern California. At 12,500 feet high, it was the driest spot in the United States.
Wilkinson and Partridge were not the only ones to notice this. ‘When we arrived at White Mountain and drove onto the site, we discovered a suspicious-looking device with a microwave horn,’ says Partridge. ‘Bernie Burke and some people from MIT were on the summit doing precisely the same thing as us!’ It went to show what a boom industry measuring the cosmic background was in the early days – before it began to get hard.
Wilkinson and Partridge were now working with a colleague, Bob Stokes. They had brought three microwave horns with them, each operating at a different wavelength. The plan was to kill three cosmological birds with one stone and pin down a trio of points on the spectrum of the cosmic background.
In the wake of Penzias and Wilson’s discovery, every astronomer in the world who had access to a suitable radio telescope attempted to measure the cosmic background. By the middle of 1966, the temperature had been shown to be close to three degrees at wavelengths all the way from 21 centimetres to 2.6 millimetres, a span of almost a hundred times in wavelength.
But all these measurements had been made on only one side of the humped spectrum – the side at relatively long radio wavelengths. With their three microwave horns, the Princeton team intended to probe the spectrum at the peak and at the shorter wavelengths beyond. They worked like beavers on their experiment for a month and a half. And this time they were rewarded with success. ‘We found the first tentative evidence of a turn-down after the peak,’ says Partridge.
But it was the end of the road for conventional microwave technology. It was impossible to build microwave receivers at wavelengths as short as a millimetre. Water vapour was a known problem at short wavelengths. But the main reason radio astronomers had first filled in the long-wavelength side of the humped spectrum was because they were able to make use of tried-and-tested microwave receivers. The long-wavelength side of the spectrum was the easy side.
‘Now the easy cream had been skimmed,’ says Partridge. ‘What was left was hard.’
After the early frenzy of activity, there would be a fallow period. ‘At Princeton, at least, people went off to do other things,’ says Partridge. He himself turned to more conventional radio and optical astronomy.
In the early 1970s, very little was being done. To go any further would need entirely new technology: the technology to detect radiation at millimetre and sub-millimetre wavelengths.
One person who soldiered on in the field even during the hard times was Dave Wilkinson. That day in 1965, when Bob Dicke had burst into his lab and announced that the Universe might be filled with the afterglow of creation, had been a fateful day. Wilkinson had been bitten by the background bug, and he was infected for life.
He was not alone. The people who do cosmic background experiments are a dedicated band. They tend to stay in the field for the rest of their careers. Even Partridge, who went off to do other things, would come back to the cosmic background again and again.
Partridge knows exactly why he is so fascinated by the cosmic background radiation. ‘For me the answer is quite clear,’ he says. ‘It’s simplicity. The experiments to measure the radiation are simple to understand and simple to describe. The radiation itself is simple – it’s a black body and it has the same temperature in all directions in the sky. Once you’ve said those two things about it, you’ve said everything there is to say.
‘The simplicity of the cosmic background radiation is telling us something marvellous – that the early Universe was a remarkably uncomplicated place.’
‘It’s the only way to look back to the beginning of the Universe,’ says Bob Dicke.
Wilkinson agrees. But the cosmology is only part of the reason he is fascinated. The main reason is that he loves the challenge of designing and building experiments to outwit nature. ‘The experiments are the sort I like,’ he says. ‘They’re tough but important. You have to think hard about the unwanted effects, and there’s a novelty and cleverness in the experiments.’
It goes back to his childhood. ‘I’m a tinkerer,’ says Wilkinson. ‘I got it from my dad. He only graduated from high school but got interested in electronics. He had a workshop in our basement. When I was a kid, I was always tinkering with cars and electronics.’
But the appeal for Wilkinson does not end here. There is another thing that has always appealed to him. ‘You can carry out an important experiment with just you and a graduate student,’ he says. ‘You have complete control over an experiment. It’s small-scale, manageable science.’
Nowadays, when so much science is big science and is carried out by international teams of hundreds of scientists, it is easy to see why background work appeals so much.
During the early 1970s, Wilkinson took advantage of a new technology that was being developed: that of high-altitude balloons. ‘It was the failure of the Yuma experiment that led Dave to first think about using balloons to get above most of the water vapour in the atmosphere,’ says Partridge.
Balloons could take an instrument package to an enormous height – perhaps three or four times the height of Everest. The air at such a height would be so thin that the instruments would practically be in space. For ten hours or so, before the strong winds at such high altitudes blew the balloon over the sea or simply out of range, the instruments could get an almost unobstructed peek at the Universe.
Hoisting experiments aloft on balloons meant building cosmic background experiments that were a lot more complicated than before. Everything had to be done remotely. Even the simple operation of sticking a cold load in front of an antenna – so easy on the ground – was difficult and plagued by problems when it had to be automated.
‘Balloon experiments are not like table-top experiments,’ says Wilkinson. ‘If you find an error in one of those, you can modify the experiment and do it again. You can’t do that with one balloon flight a year.’
Unexpected things were likely to happen in the extreme environment 30 or 40 kilometres up in the air. For a start, it would be dreadfully cold. Ice could freeze up the equipment. All sorts of things could happen which would be dead easy to put right on the ground but which could wreck an experiment flying on its own beneath a balloon on the edge of space.
On balloon campaigns, Wilkinson had a secret weapon – his dad. Because his dad was retired and lived in Texas, where the balloons were launched, he often turned up to lend a hand. Wilkinson, usually working with just one graduate student, was always grateful. It took lots of work to ready a payload for launch.
Balloons were only one of the ways people found to get above the atmosphere. Some used high-flying aeroplanes. Others used sounding rockets. These pencil-thin launchers were used by meteorologists for studying the upper atmosphere. They went straight up to a height of a few hundred kilometres, then came right back down again when their fuel ran out. But in the few minutes they were in space the instruments they carried got a clear view of the Universe. The drawback was that everything had to work perfectly during those few minutes, or else years of hard work went down the drain.
What kept the field alive during the 1970s were the efforts of researchers bringing new technology to bear on the problem of the microwave background. At the beginning of the decade, there was a major breakthrough in detector technology. New detectors came in which were called ‘bolometers’. These responded to the warmth of incoming radiation. Essentially, a very small amount of heat changed their resistance to an electric current by a large amount, and this was something experimenters could easily measure.
Bolometers were much better than radio receivers at detecting faint radiation at short wavelengths of a few millimetres. And they worked at even shorter wavelengths, for which it was impossible to build radio receivers. To achieve the best results, though, bolometers had to be cooled to within a whisker of absolute zero.
Into the cosmic background field came groups of scientists specialising in using the new detectors. Ray Weiss and his colleagues at Boston’s MIT were among the first to use bolometers in the early 1970s. In Britain, a group led by John Beckman at Queen Mary College in London made a foray into the field. Another group got started at the University of California at Berkeley. It included Paul Richards, John Mather and George Smoot. Both the American groups were destined to have an important impact on the field.
But the first observations of the cosmic background radiation made with bolometers did not show the drop-off with wavelength expected for a black body. ‘The experiments were extremely difficult to do,’ says Wilkinson.
Rather than responding to one wavelength like a radio receiver, bolometers responded to all wavelengths at once. This meant that to make a measurement at any one wavelength, it was necessary to put a ‘filter’ in front of a detector. The filter was transparent to the wavelength of interest but absorbed all other wavelengths.
A familiar example of a filter is a sheet of red cellophane. This allows red light through while stopping, or ‘filtering out’, all other colours. Similarly, a blue filter is transparent only to blue light.
For observing the cosmic background radiation, scientists like Ray Weiss used a bolometer in conjunction with several filters. But though it was now possible to make measurements that were quite impossible with microwave receivers, instruments using bolometers were not without their problems.
For a start, filters absorbed radiation, and anything that absorbs must also emit (otherwise it would simply get hotter and hotter until it was white-hot). So filters were yet another source of unwanted radiation which astronomers had to learn to contend with.
Another problem with filters was that they threw away a lot of the precious cosmic background radiation, allowing through only light at one particular wavelength. This was very wasteful. But, in the late 1970s, there was another major development in the field. The second generation of experiments at short wavelengths used an instrument known as the Michelson interferometer, which allowed all wavelengths to be detected at once.
The Michelson interferometer had been invented back in the 1880s by the American physicist Albert Michelson.2 Essentially, all such an instrument does is split the light into two equal parts and then recombine it. This may seem a pretty pointless thing to do but, after the light is split and before it is recombined, the two halves are made to travel different distances – usually by bouncing them off two separate mirrors. This path difference can be changed continuously by gradually moving one of the mirrors.
Since the light entering a Michelson interferometer consists of waves of a multitude of different wavelengths mixed in together, when the light is recombined, each different wavelength combines with its other half. For a particular wavelength, if the peaks of the two halves still coincide despite having travelled different distances, then the waves reinforce each other when they recombine. But if the peaks of one half coincide with the troughs of the other, then the waves cancel each other out.
The wavelengths which reinforce and cancel will change as the path difference is changed. So when the recombined light falls on a bolometer detector, the brightness the detector registers will vary. The way in which the brightness varies with path difference is known as an ‘interferogram’. In theory, the interferogram contains all the information needed to determine the brightness of the light at a large number of wavelengths simultaneously – in other words, to determine the spectrum of the light. In practice it takes a bit of mathematical manipulation to extract the spectrum.
The details of how you do this are not important to this discussion. The key thing is that a Michelson interferometer can measure all wavelengths at once, and so waste none. And it has another feature which makes it particularly suited to measuring the spectrum of the cosmic background radiation: it can compare the sky and the cold load all the time. There is no need to keep switching back and forth between the sky and the cold load and wasting half the precious light from the sky.
By the late 1970s, Michelson interferometers equipped with bolometers and cold loads, both cooled by liquid helium, represented the state of the art in measuring the spectrum of the cosmic background radiation. But, in the 1980s, experimenters took a final step. Instead of using a small amount of liquid helium, they used a very large amount. In fact, they immersed their entire instruments, including their antennas, in large vacuum flasks full of the liquid. This cooled them to within a few degrees of absolute zero, dramatically reducing the unwanted radiation from the instruments themselves.
Despite the advances, measuring the cosmic background radiation remained tough. It was, after all, the lowest temperature in the Universe. Everything else was hotter – the ground, the sky, even the instrument making the measurement. So, whenever people measured the background, all those other things would be in there too. It was like trying to observe a faint star while standing on a searchlight.
Basically, cosmic background experiments came down to good accounting. Measure a temperature. Then think of all the possible confusing effects and estimate how big they are. Better still, go out and measure how big they are. The cosmic background is the residue, what is left over after everything else has been accounted for.
In theory it was straightforward. That was why Partridge said the experiments were simple to understand and describe. But in practice it was a lot harder. ‘The problem is, have you thought of everything?’ says Wilkinson. ‘The nature of the field is funny. You see someone’s result and you have to question everything.’
It was almost impossible to think of every spurious source of radiation. Something was always overlooked. Some astronomers accidentally measured the temperature of the plastic window through which their instruments looked at the sky. Others measured the temperature of the balloon that was hoisting their experimental package aloft.
Partridge remembers an experiment done with Norwegian colleagues at a site in northern Norway. It was 400 kilometres above the Arctic Circle. ‘We reasoned that if you went high in the Arctic, it was just like going into space,’ he says. ‘It’s cold, dark and, when there’s no Sun, the temperature is stable.’
They reasoned wrong. ‘It was a total failure,’ remembers Partridge. ‘We were killed off by the atmosphere.’ They had not reckoned on a waterfall of cold air draining off the Soviet plateau and down over the Norwegian coastal plateau. ‘We ended up observing under all this turbulence. It turns out the South Pole, not the North, is the place to go. It’s high and dry, and the air behaves because the atmosphere there is very stable.’
On another occasion, Partridge and his colleagues were flying a balloon experiment to measure the smoothness of the background radiation and ended up measuring the temperature of some cables attached to the balloon. The experiment had a horn spinning beneath the balloon. Unknown to Partridge’s team, the launch cables dropped down – the umbilicus of the balloon – so six rubber-coated wires dangled in front of the horn.
‘We’d worried ourselves sick about the four metal cables that suspended the instrument package from the balloon,’ says Partridge (in the end, they had carefully shielded them). ‘But nobody told us this was going to happen. Every time the horn swung round, it got a huge jolt of radiation emitted by the cables. On our chart recorder, we got a gigantic trace followed by a long, slow decline. The instrument had just about recovered when the horn swung round and saw the cable again!’
But, despite all the problems, the experimenters were imperturbable. The cosmic background radiation, after all, was our only window on the beginning of the Universe. By the late 1970s, all the hard work was beginning to pay off …