By that standard subatomic physics was a flourishing enterprise indeed. From the cathode-ray tube with which, in 1897, the English physicist J. J. Thomson discovered the electron, to a palm-size cyclotron the American physicist Ernest O. Lawrence built in 1931, the energy increased about 200 times-from hundreds of electron volts to almost 100,000. In 1952 the Brookhaven National Laboratory on Long Island, New York, broke the billion mark with its three-billion-electron-volt (3 GeV) Cosmotron. Two decades later, the four-mile-long proton-antiproton collider at Fermilab, in Batavia, Illinois, reached 400 GeV; since then, superconducting magnets have raised its energy to 1.8 trillion electron volts (TeV), the highest of any particle accelerator in the world.

The big numbers are not empty status symbols, like the heights of skyscrapers. Rather, they measure the capabilities of research tools built for a well-defined purpose: to create a beam of charged particles (electrons, protons, helium nuclei or something more exotic), shoot it at a target and record the results. Because any target, whether another beam or a solid object, is mostly empty space, almost all the particles in the beam hurtle through the target unobstructed. But a few strike other particles, with results that can be much more interesting than a mere collision: the matter and energy of the particles in the smashup can recombine, constrained by appropriate conservation laws, to create unheard-of or even outlandish assemblages of matter.

Those assemblages — new elementary particles — are what physicists sought in their accelerators. At first they were almost embarrassingly successful: new particles seemed to turn up virtually every week. Some of them were fleeting, lasting barely long enough to move away from the collision site before they disintegrated into showers of decay products. Theorists such as Murray Gell-Mann, then at the California Institute of Technology and Yuval Ne'eman, then at Imperial College, London, discovered order in the growing menagerie of particles. In 1964 they and George Zweig, also of Caltech, independently proposed what became known as the theory of quarks — a system that tied the known particles together in an elegant mathematical web and predicted additional particles that should exist. Experimenters, eager to fill the empty cages, made even deeper forays into the high-energy wilderness and brought back ever more exotic beasts.

Today the particle zoo numbers in the hundreds. The roster of quarks has grown from a trio to a sextet: up, down, strange, charm, bottom (or beauty) and — after long searching and several false starts — top (or truth) [see Quanta: "Now We Are Six," by Ted Anton, July/August l994]. The push toward higher energies continues, despite the recent budgetary demise of the most ambitious particle accelerator of all, the Superconducting Super Collider. At Brookhaven workers are building the Relativistic Heavy-Ion Collider, a machine that will push gold nuclei to energies of twenty trillion electron volts each. At CERN, the European Laboratory for Particle Physics outside Geneva, Switzerland, the Large Hadron Collider, a fourteen-TeV proton-proton collider, is scheduled to go on-line by 2005.

Yet in the stampede toward higher energies, the up and down quarks that make up the protons and neutrons of ordinary atoms came to seem humdrum. Next to their four glamorous cousins, the up and down quarks looked somehow mundane, common-even boring.

That is a shame, for ordinary matter has a lot to recommend it as an object of study. First, it is durable; that is why it is ordinary matter. Exotic particles may flash into existence for a few quadrillionths of a second or less; the proton, depending on whose theory you accept, lasts either forever or effectively forever — one experimentally determined lower limit on its lifetime is 10³² years. Second — not a trivial point — ordinary matter makes up almost all of the observable universe. And third, it gives rise to a tremendous variety of permutations. Chemists have not even come close to synthesizing all the compounds possible with the naturally occurring elements in the periodic table.

But the fascination goes much deeper. From the point of view of atomic theory, chemistry is just a complicated dance of clouds of electrons, wrapping, spreading, morphing and merging in accordance with physical law. In the middle of the electron swarm, occupying only one-quadrillionth the volume of a typical atom, is the nucleus. That nuclear fly-speck is made of protons and neutrons, particles collectively known as nucleons. The nucleons, in turn, enclose another world: a world almost cut off from the atomic world, ruled by its own forces, subject to its own laws and harboring its own chemistry: the chemistry of quarks.

Thirty-five years ago the richness of the nucleonic world was entirely unsuspected. Since then theorists have sketched a good general outline of what ought to be its general features. The details of its topography however, are only beginning to be glimpsed. For mapping them, the violent techniques of high-energy physics are all but useless. Instead of blasting their way in, physicists must take a subtler approach: tickling a proton or a neutron with a beam finely tuned to nudge its quarks into new configurations.

The Jefferson Lab is designed to investigate the structure of the nucleus and its nucleons at a scale at which quarks become important. It is made up mainly of an electron accelerator shaped like a racetrack seven-eighths of a mile around, buried in a tunnel twenty-five feet underground. The electrons are accelerated by a rapid series of electromagnetic pulses as they make as many as five passes around the racetrack. The process imparts as much as four billion electron volts of energy to the electrons — only one-half of one percent the energy of the nucleon-vaporizing beams generated by the big accelerators at Fermilab or CERN, but ideal for the job the Jefferson Lab was designed to do.

Electrons, according to quantum mechanics, act sometimes as particles and sometimes as waves. A billion electron volts (1 GeV) of energy give rise to electrons whose quantum-mechanical wavelengths are a little less than 10^-15 meter. Such waves in effect become the illuminating "light" of a giant microscope with just enough resolving power to probe nucleons and the quarks that make them up. One feature that makes the Jefferson Lab unique is the high intensity and quality (that is, the precisely controlled energy) of the electron beam.

In a typical experiment, physicists aim the beam at a fixed target, often a thin film of carbon or iron, or a tube of hydrogen or helium the size of a fat crayon. When the electrons strike a nucleon or a constituent quark of a nucleon, the collision can give rise to both short-lived and long-lived particles, moving in all directions much like the shards of glass from a light bulb struck by a bullet. By tracing the trajectories of the particles backward, often through several stages of decay, one can then draw conclusions about the original interaction at the collision site, as well as the cascade of subsequent reactions.

Halls A and C contain a total of four spectrometers, each the size of a house. The two machines in Hall C can measure the energy of a particle to within one part in a thousand; their counterparts in Hall A are ten times as precise. The resolution of those machines comes at a price: they are positioned to detect particles only in the directions around the collision site in which physicists most expect particles of interest to emerge. Thus there is a fifth spectrometer, in Hall B, which has a resolution of only one percent but can catch particles emerging in almost any direction. That capability makes it ideal for monitoring "High coincidence reactions" — collisions that can send particles spraying out in many directions.

The success of Bohr's idea — one of the earliest manifestations of quantum mechanics — was its immediate application to the emission lines in the electromagnetic spectrum of an atom: pure, bright colors given off when an element is heated. The discrete colors of the spectrum were explained as the discrete quantity of energy released by the atoms of the substance as the electrons orbiting the atoms dropped from a relatively high energy orbit to a relatively low energy one. The shorter the wavelength of the spectral line, or in other words the more its color tended toward the blue end of the spectrum, the longer the drop of the electrons and the greater the energy released by the atoms; that effect, among others, is what makes a blue flame much hotter than a yellow one. The color of the emitted spectral line is a precise measure of the difference in energy between the higher-energy and the lower-energy electronic orbits.

Later, Bohr's idea of discrete electronic orbits gave way to more abstract, mathematical descriptions of the energy states of atoms. Those states are harder to imagine than Bohr's simple picture (sometimes the states are described as the shape of a "probability cloud" of electrons that surround the atomic nucleus), but the central idea that spectral lines are signals of transitions among the various energy states has remained unchanged. When you couple that idea with the principle that physical systems always tend to assume their lowest available energy state, most of chemistry is laid out before you.

Each of those features has a counterpart in the world inside a nucleon. Electromagnetism holds negligible sway here; although the quarks inside the nucleon do carry electric charge, the electromagnetic force between them is overwhelmed by the much more powerful color force, also known as the strong nuclear force. (The electromagnetic force still dominates outside the atomic nucleus, because the color force acts only over an extremely short range.) The particles corresponding to the electron and the proton are the quarks. As I noted earlier, the quarks of primary interest here are the ones having what physicists call the up and down flavors. The particles corresponding to the photon are gluons, which carry the color force.

Corresponding to the color force is the so-called color charge, which is the analogue of electric charge. But unlike electric charge — which comes in two varieties, positive or negative — color charge comes in three fundamental forms, generally called red, blue and green. Quarks also have antiparticle partners — antiquarks — with negative color charge: antired, antiblue and antigreen. A final complication is that there are eight kinds of gluon, which convey the color force from quark to quark. Most gluons are themselves color charged. That too is in contrast to the single kind of photon, which is electrically neutral. One should not be misled into thinking that the names of the color charges are anything but whimsical labels; quarks and gluons have no colors, nor do they give rise to spectrums of visible colors in any ordinary sense.

The colors interact according to rules about as complicated as the instructions for a moderately challenging game of solitaire: Like colors repel. Opposite colors (such as red and antired) attract. And other pairs of colors (such as red and blue) also attract, but more weakly. Perhaps the most important rule is that the quarks must combine to make particles that are colorless or "white." For example, a color and its anticolor can combine to make a two-quark assembly

That fact raises curious questions about the nature of quarks themselves: Since they carry a net color charge, how can they themselves be observed? And if they cannot be singled out, what status do they have as real particles? The questions highlight the changing nature of scientific evidence. Throughout the history of the study of matter we physicists have been able to confirm our hypotheses about the basic constituents of matter by distilling and isolating those constituent particles. In the eighteenth and nineteenth centuries chemists were busy isolating the elements before the English chemist John Dalton could articulate his atomic theory or the Russian chemist Dmitry I. Mendeleyev could organize the elements into his periodic table. To confirm our ideas about the structure of the atom, J. J. Thomson observed the electron, and the New Zealand-born English physicist Ernest Rutherford observed the nucleus. Likewise, when the nucleus was first pulled apart, Rutherford was able to observe the proton and the English physicist James Chadwick was able to observe the neutron.

But no one has ever seen a free quark. In colliders, physicists have created fireballs with a thousand times more energy than is necessary to create or destroy a proton, yet a free quark, whose mass is less than a third that of the proton, has never been created. There are many good reasons to think one never will be.

One way of understanding such "quark confinement" is to picture the gluons as springs that always pull the quarks back together, no matter how far apart they stretch. The springs — technically flux tubes — never tangle and never break. But what would happen if two quarks were somehow pulled, say, a centimeter apart? Could they not then be observed as discrete particles, the color-charge analogues of an ionized atom or a free electron? According to the quark model, the energy needed to pull quarks a centimeter apart is so large that new quark matter would materialize out of that energy long before the separation reached a centimeter. In accordance with Einstein's famous formula that equates mass with energy the new quark matter would combine with the original quarks, and one would observe new, colorless particles flying off in various directions.

Thus, in searching for evidence of the existence of quarks, and for the ways they interact, one must be content with indirect, albeit persuasive, evidence. At the Jefferson Lab our strategy will be to build models of the proton and the neutron out of quarks and then design experiments that should give rise to the phenomena the models predict. But more than just seeing whether the models work to describe the proton and neutron, it will also be important to examine the other ways quarks can combine, to see to what extent our models really can explain those systems.

For example, there is a quark analogue, inside the nucleon, to the orbital energy levels of the electrons in an atom. Think of the three quarks in a nucleon as whirling

Such an excitation may account for a particle that has already been detected, the Roper. Named after the physicist L. David Roper, who discovered it in 1964 at the Lawrence Berkeley Laboratory in Berkeley, California, the Roper [also called the P₁₁(1440)] is a heavy particle whose mass is more than one and a half times that of the proton. Technically, it is an excited state of the proton. But something even odder may be going on. Some physicists think that what has been excited is not the quarks but the gluons. Recall that in the model the gluons act as springs connecting the quarks. But springs store energy in more than one way: they can Hex, but they can also vibrate. Gluons may play a similar role here. If so, the Roper could be made up of three quarks still in their low-energy configuration, but with the springs between them "plucked" and excited.

Whatever accounts for the Roper, at least twenty distinct energy excitations for both the proton and the neutron have been resolved, all probably analogous to the electronic energy levels of the atom. Many others may be observable in principle, though they may be too short-lived to be seen in practice. In one respect, though, the spectrum of excited states of a nucleon is quite different from the corresponding spectrum of the atom. As with the Roper particle, the excited nucleonic states are much more energetic than the ground state-the ordinary neutron and proton. Indeed, the energy differences are so great that the masses of excited states are substantially larger than the mass of the ground state. That is decidedly not the case for the electromagnetic analogue. The primary reason is that the color force is so much stronger than the electromagnetic force.

To a first approximation, the spin of a particle is like the English on a billiard ball. And like a rotating ball, in collisions a particle with spin rebounds differently from a particle without spin (that is how the spin of a particle is usually detected). A quantum-mechanical rule known as the Pauli exclusion principle dictates that no two quarks or electrons with identical spins can occupy the same space — such as an atomic orbit or energy level. (Strange as it may sound, particles with spins of zero or one — photons or gluons, for instance — can pile up without limit.) I once saw a bumper sticker that said "Time is what keeps everything from happening all at once." In a sense, spin is what keeps everything from happening in the same place.

Now think about the quark structure of a neutron, namely, two down quarks and one up quark. At first glance, it would seem that the gluonic springs all would pull with the same force. After all, the strong force between two down quarks is the same as that between an up quark and a down quark. In that case, on average, the quarks would form an equilateral triangle. As the triangle pivoted and the springs wiggled, the negative electric charges of the two down quarks (each negative 1/3) would exactly cancel the positive 2/3 electric charge of the up quark. Every spot inside the neutron would be left with an average charge of zero.

But the world is not quite so simple. It turns out that quarks with parallel spins repel each other faintly; slightly weakening the strong force between them. The spins of the two down quarks in the neutron are parallel. The spin of the up quark, however, flip-flops at random, spending only half its time parallel to each of the down quarks. Thus you can think of the springs that bind it to them as being relatively taut, compared with the spring that connects the two down quarks. The three quarks therefore form a distorted isosceles triangle, in which the up quark is pulled toward the midpoint of the spring connecting the two down quarks. The result is an asymmetric charge distribution: positive at the center of the neutron, negative farther out. just how powerfully the spin affects the strong force, and thus how much the charge is distorted, is one of the big questions the Jefferson Lab hopes to settle.

Probing that asymmetry is an ideal job for the Jefferson Lab's electron beam. Energetic electrons penetrate deep into the neutron, and so their deflection is largely a matter of scattering off the positively charged up quark. Less energetic electrons bounce off the negatively charged down quarks at the surface of the neutron. By plotting energies against scattering angles, physicists can get crucial information about how the forces inside the neutron interact.

Inside the nucleon, the effects of such spin flips are much more dramatic. The three quarks inside a nucleon each can have a component of spin parallel or antiparallel to some arbitrary direction. Thus there are two possibilities: either the quarks all are parallel, or one quark is different. The latter case holds for the ground state of the proton or the neutron: not all the quarks have parallel spin. Suppose the antiparallel quark inside a neutron flips its spin to match those of its neighbors. The neutron metamorphoses into a new particle — a delta particle — with significantly more energy than a proton or a neutron has. The hyperfine transition between the delta and the neutron releases energy equivalent to roughly 30 percent of the mass of the neutron in the ground state.

Can other, even more exotic assemblages of quarks be created? For example, can one create the analogues of various neutral atoms with many electrons? No one has ever detected particles with four, five, six or more quarks, but nothing in the quark model forbids them. At present, colleagues of mine at the University of New Hampshire are working with Russian physicists at the synchrophasotron accelerator at the Joint Institute for Nuclear Research in Dubna, north of Moscow, in a search for such unusual multiple-quark particles. If they succeed, the particles will be analogous to the synthetic chemical elements nuclear physicists create in reactors: heavy, short-lived and unknown in nature.

What about quark molecules? The most straightforward analogue is the atomic nucleus, in which the nucleons are regarded as the "atoms," and the permitted nuclei are regarded as the "molecules." That, of course, is the phenomenon that has been studied in atomic physics for the past fifty years. But neutrons and protons might, if pushed together closely enough, also form a new kind of matter, a quark-gluon plasma. How would such a plasma evolve? What new particles might condense out of it? All those ideas, and many more, will serve to guide us as we prod the quarks and gluons in nucleons and try to coax from them the secrets of their interactions.