John and I sat on the floor of the not-yet-really-working accelerator, drinking beer and discussing the world. "What really happens to the pions that come flying off the target?" he asked suddenly. I had learned to be cautious. John was a gambler in physics as well as in horses. "Well, if the target is inside the machine [and it had to be, we didn't know how to get the accelerated protons out of the cyclotron], the powerful magnet will spray them in all directions," I answered cautiously.
JOHN: Some will come out of the machine and hit the shielding wall?
ME : Sure, but all over the place.
JOHN: Why don't we find out?
ME: How?
JOHN: We do magnetic tracing.
ME: That's work. [It was 8 P.M. on a Friday.]
JOHN: Do we have the table of measured magnetic fields?
ME: I'm supposed to go home.
JOHN: We'll use those huge rolls of brown wrapping paper and draw the paths of the pions on a scale of one to one ...
ME: Monday?
JOHN: You do the slide-rule work [this was 1950] and I'll draw the paths.
Well, by 4 A.M. Saturday we had made a fundamental discovery that would change the way cyclotrons were used. We had traced eighty or so fictional particles emerging from a target in the accelerator with plausible directions and energies—we used 40, 60, 80, and 100 MeV. To our amazement, the particles didn't "just go everywhere." Instead, because of the properties of the magnetic field near and beyond the rim of the cyclotron magnet, they curved around the machine in a tight beam. We had discovered what became known as "fringe field focusing." By rotating the large sheets of paper—that is, by picking a specific target location—we could get the beam of pions in a generous energy band around 60 MeV to head right for my brand-new cloud chamber. The only catch was the wall of concrete between the machine and the experimental area where my princess chamber sat.
No one had anticipated our discovery. Monday morning we were camped outside the director's office waiting to pounce on him with it. We had three simple requests: (1) a new target location in the machine; (2) a much thinner window between the beam vacuum chamber of the cyclotron and the outside world to minimize the influence of a one-inch-thick stainless steel plate on the emerging pions; and (3) a new hole about four inches high by ten inches wide, we guessed, through the ten-foot-thick concrete wall. All this from a lowly graduate student and a postdoc!
Our director, Professor Eugene Booth, was a Georgia gentleman and a Rhodes scholar who rarely said "gosh darn." He made an exception for us. We argued, we explained, cajoled. We painted visions of glory. He would be famous! Imagine an external pion beam, the first ever!
Booth threw us out, but after lunch he called us in again. (We had been weighing the advantages of strychnine versus arsenic.) Bernardini had dropped in, and Booth had tried out our idea on this eminent visiting professor. My guess is that the details, expressed in Booth's Georgian lilt, were too much for Gilberto, who once confided in me, "Booos, Boosth, who can pronounce dese American names?" However, Bernardini supported us with typical Latin exaggeration, and we were in.
A month later it all worked—just like the wrapping paper sketches. In a few days my cloud chamber had registered more pions than all the other labs in the world put together. Each photograph (we took one each minute) had six to ten beautiful tracks of pions. Every three or four photographs would show a kink in a pion track as it disintegrated into a muon and "something else." I used the pion decays as my thesis. Within six months we had constructed four beams, and Nevis was in full production as a data factory for the properties of pions. At the earliest opportunity, John and I went to the racetrack in Saratoga where, continuing his roll, he hit a 28-to-l shot in the eighth race, on which he had wagered our dinner and return-home gas money. I really loved that guy.
John Tinlot must have had extraordinary insight to suspect the fringe field focusing that everyone else in the cyclotron business had missed. He went on to a distinguished career as a professor at the University of Rochester but died of cancer at the age of forty-three.
A SOCIAL SCIENCE DIVERSION: THE ORIGIN OF BIG SCIENCE
World War II marked a crucial watershed between pre-WWII and post-WWII scientific research. (How's that for a controversial statement?) But it also marked a new phase in the search for the a-tom. Let's count some of the ways. The war generated a leap forward in technology, much of this centered in the United States, which was unencumbered by the loud noises of nearby explosions that Europe was experiencing. The wartime development of radar, electronics, the nuclear bomb (to use its proper name) all provided examples of what a collaboration between science and engineering could do—as long as it was unconstrained by budget considerations.
Vannevar Bush, the scientist who led U.S. science policy during the war, spelled out a new relationship between science and government in an eloquent report to President Franklin D. Roosevelt. From that time on, the U.S. government was committed to supporting basic research in the sciences. Support for research, basic and applied, climbed so rapidly that we can laugh at the $1,000 grant E. O. Lawrence worked so hard to get in the early 1930s. Even adjusting for inflation, that amount pales in contrast to federal support of basic research in 1990—some $12 billion! World War II also saw a flood of scientist refugees from Europe become a crucial part of the research boom in the United States.
In the early 1950s some twenty universities had accelerators capable of carrying out research in nuclear physics at the cutting edge. As we came to understand the nucleus better, the frontier shifted to the subnuclear domain, where larger—more expensive—machines were required. The era became one of consolidation—scientific mergers and acquisitions. Nine universities banded together to build and manage the accelerator laboratory at Brookhaven, Long Island. They commissioned a 3 GeV machine in 1952 and a 30 GeV machine in 1960. Princeton University and the University of Pennsylvania banded together to build a proton machine near Princeton. MIT and Harvard built the Cambridge Electron Accelerator, a 6 GeV electron machine.
Over the years, as the consortia grew in size, the number of front-line machines diminished. We needed ever higher energy to address the question of "What's inside?" and to search for the true a-toms—or the zero and one of our library metaphor. As new machines were proposed, older ones were phased out to free up funds, and Big Science (a term often used as an expletive by ignorant commentators) grew bigger. In the 1950s, one could do maybe two or three experiments a year with groups of two to four scientists. In the following decades, the collaborations got larger and larger, and the experiments took longer and longer, driven in part by the necessity to build ever more complex detectors. By the 1990s the Collider Detector Facility alone at Fermilab comprised 360 scientists and students from twelve universities, two national labs, and institutions in Japan and Italy. Scheduled runs stretched to a continuous year or more of data taking—with time off for Christmas, the Fourth of July, or whenever something broke down.
Supervising the evolution from a tabletop science to one based on accelerators measured in miles around was the U.S. government. The World War II bomb program gave rise to the Atomic Energy Commission (AEC), a civilian agency that oversaw nuclear weapons research, production, and stockpiling. It was also given the mission, as a national trust, of funding and overseeing basic research in nuclear and what was later to become particle physics.
The case for Democritus's a-tom even reached the halls of Congress, which created the Joint (House and Senate) Committee on Atomic Energy to provide oversight. The committee's hearings, published in dense, government-green booklets, are a Fort Knox of information for science historians. Here one reads the testimony of E. O. Lawrence, Robert Wilson, I. I. Rabi, J. Robert Oppenheimer, Hans Bethe, Enrico Fermi, Murray Gell-Mann, and many others patiently responding to questions about how the search for the ultimate particle was going—and why did it require yet another machine? The interchange at the beginning of this chapter between Fermilab's flamboyant founding
director, Robert Wilson, and Senator John Pastore was taken from one of those green books.
To complete the alphabet soup, the AEC dissolved into the ERDA (Energy Research and Development Agency), which soon gave way to the DOE (U.S. Department of Energy), which at this writing oversees the national laboratories where atom smashers operate. Presently there are five such high-energy labs in the U.S.: SLAC, Brookhaven, Cornell, Fermilab, and the Superconducting Super Collider lab, now under construction.
Accelerator labs are generally owned by the government, but operated by a contractor; which can be a university, such as Stanford in SLAC's case, or a consortium of universities and institutions, as is the case with Fermilab. The contractors appoint a director, and then they pray. The director runs the lab, makes all the important decisions, and often stays on the job too long. As Fermilab director from 1979 to 1989, my major task was to implement Robert R. Wilson's vision: the construction of the Tevatron, the first superconducting accelerator. We also had to create a proton-antiproton collider and humongous detectors that would observe head-on collisions near 2 TeV.
I worried a lot about the process of research when I was director of Fermilab. How could students and young postdocs experience the joy, the learning, the exercise of creativity experienced by Rutherford's students, by the founders of quantum theory, by my own small group of colleagues as we sweated out the problems on the floor of the Nevis cyclotron? But the more I looked into what was happening at the lab, the better I felt. The nights I visited the CDF (when old Democritus wasn't there), I found students enormously excited as they ran their experiments. On a giant screen events were flashing, reconstructed by the computer to make sense to the dozen or so physicists on shift. Occasionally, an event would be so suggestive of "new physics" that an audible gasp would be heard.
Each large research collaboration consists of many groups of five or ten people: a professor or two, several postdocs, and several graduate students. The professor looks after his brood, making sure they are not lost in the crowd. Early on they are wrapped up in the design, building, and testing of equipment. Later on comes the data analysis. There is so much data in one of these collider experiments that much of it must wait for some group to complete one analysis before getting around to tackling the next problem. The individual young scientist, perhaps advised by her professor, selects a specific problem that receives the consensual agreement of the council of group leaders. And problems abound. For example, when W+ and W− particles are produced in proton-antiproton collisions, what is the precise form of the process? How much energy do the Ws take away? At what angles are they emitted? And so on. This could be an interesting detail, or it could be a clue to a crucial mechanism in the strong and weak forces. The most exciting task for the 1990s is to find the top quark and measure its properties. Up to mid-1992 this search was carried out by four subgroups of the CDF collaboration at Fermilab doing four independent analyses.
Here the young physicists are on their own, fighting complex computer programs and the inevitable distortions introduced by an imperfect apparatus. Their problem is to extract a valid conclusion about how nature works, to establish one more piece of the jigsaw puzzle of the microworld. They have the benefit of a huge support group: experts in software, in theoretical analysis, in the art of seeking confirming evidence for tentative conclusions. If there is an interesting glitch in the way W's are thrown out of collisions, is it an artifact of the apparatus (metaphorically, a small crack in the microscope lens)? Is it a bug in the software? Or is it real? And if it is real, wouldn't colleague Harry see a similar effect in his analysis of Z particles—or perhaps Marjorie in her analysis of recoil jets?
Big Science is not the sole province of particle physicists. Astronomers share giant telescopes, pooling their observations in order to draw valid conclusions about the cosmos. Oceanographers share research ships elaborately equipped with sonar, diving vessels, and special cameras. Genome research is the microbiologists' Big Science program. Even chemists require mass spectrometers, expensive dye lasers, and huge computers. Inevitably, in one discipline after another, scientists are sharing the expensive facilities that are necessary to make progress.
Having said all this, I must emphasize that it is also extremely important for young scientists to be able to work in more traditional modes, clustered around a tabletop experiment with their peers and a professor. There they have the splendid option of pulling a switch, turning out the lights, and going home to think, perchance to sleep. "Small science" has also been a source of discovery, variety, and innovation, which contribute enormously to the advancement of knowledge. We must strike the proper balance in our science policy and be prayerfully grateful that both options exist. As for high-energy practitioners, one can tsk, tsk, and wish for the good old days when the lonely scientist sat in his folksy laboratory, mixing colorful elixirs. It's a charming vision, but it will never get us to the God Particle.
BACK TO THE MACHINES: THREE TECHNICAL BREAKTHROUGHS
Of the many technical breakthroughs that permitted acceleration to essentially unlimited energy (unlimited, that is, except by budgets) we'll look at three up close.
The first was the concept of phase stability, discovered by V. I. Veksler, a Soviet genius, and independently and simultaneously by Edwin McMillan, a Berkeley physicist. Our ubiquitous Norwegian engineer Rolf Wideröe, independently patented the idea. Phase stability is important enough to call in a metaphor. Think of two identical hemispherical bowls with very small flat bottoms. Turn one bowl upside down, and place a ball on the small flat bottom, which is now the top. Place a second ball at the bottom of the noninverted bowl. Both balls are at rest. Are both stable? No. The test is to give each ball a nudge. Ball No. 1 rolls down the outside of the bowl, changing its condition radically. That's unstable. Ball No. 2 rolls up the side a bit, returns to the bottom, overshoots, and oscillates around its equilibrium position. That's stable.
The mathematics of particles in accelerators has much in common with the two conditions. If a small disturbance—for example, a particle's gentle collision with a residual gas atom or with a fellow accelerated particle—results in large changes in motion, there is no basic stability, and sooner or later the particle will be lost. On the other hand, if these perturbations result in small oscillatory excursions around the ideal orbit, we have stability.
Progress in the design of accelerators was an exquisite mixture of analytic (now highly computerized) study and the invention of ingenious devices, many of them building on the radar technology developed during World War II. The concept of phase stability was implemented in a variety of machines by applying radio frequency (rf) electrical forces. Phase stability in an accelerator happens when we organize the accelerating radio frequency so that a particle arrives at a gap at slightly the wrong time, resulting in a slight change in the particle's trajectory; the next time the particle hits the gap, the error is corrected. An example was given earlier with the synchrotron. What actually happens is that the error is overcorrected, and the particle's phase, relative to the radio frequency, oscillates around an ideal phase in which good acceleration is achieved, like a ball at the bottom of the bowl.
The second breakthrough occurred in 1952, when Brookhaven Laboratory was completing its 3 GeV Cosmotron accelerator. The accelerator group was expecting a visit from colleagues at the CERN lab in Geneva, where a 10 GeV machine was being designed. Three physicists preparing for the meeting made an important discovery. Stanley Livingston (a student of Lawrence's), Ernest Courant, and Hartland Snyder were a new breed of cat: accelerator theorists. They hit on a principle known as strong focusing. Before I describe this second breakthrough, I should make the point that particle accelerators had become a sophisticated and scholarly discipline. It pays to review the key ideas. We have a gap, or radio-frequency cavity, which is what gives the particle its increase in energy at each crossing. To use it over and over we guide the particles into an approximate circle, using magnets. The maximum en
ergy of particles that can be achieved in an accelerator is determined by two factors: (1) the largest radius that the magnet can provide and (2) the strongest magnetic field possible at that radius. We can build higher-energy machines by making the radius bigger, by making the maximum magnetic field stronger, or by doing both.
Once these parameters are set, giving the particles too much energy would drive them outside of the magnet. Cyclotrons in 1952 could accelerate particles to no more than 1,000 MeV. Synchrotrons provided magnetic fields to guide the particles at a fixed radius. Recall that the synchrotron magnet strength starts out very low (to match the low energy of the injected particles) at the beginning of the acceleration cycle and increases gradually to its maximum value. The machine is doughnut-shaped, and the radius of the doughnut in the various machines constructed during this era varied from 10 to 50 feet. The energies achieved were up to 10 GeV.
The problem that occupied the clever theorists at Brookhaven was how to keep the particles tightly bunched and stable relative to an idealized particle moving without disturbances in magnetic fields of mathematical perfection. Since the transits are so long, extremely small disturbances and magnetic imperfections can drive the particle away from the ideal orbit. Soon we have no beam. So we must provide conditions for stable acceleration. The mathematics was complicated enough, one wag said, "to curl a rabbi's eyebrows."
The God Particle: If the Universe Is the Answer, What Is the Question? Page 29
