Physicists meeting in Chicago yesterday outlined plans for a bold and risky step: the building of a 100-mile-circumference accelerator intended to allow Americans to leapfrog scientists in the rest of the world in basic atomic physics research.

The proposed machine would be the largest ever built, and would use the newly developed technology called "superconductivity," which holds the promise for industry of simultaneously increasing the power of electrical machines and cutting the power it takes to run them.

Plans for the $2 billion machine were described during the 12th International Conference on High-Energy Accelerators at the Fermi National Accelerator Laboratory near Chicago.

The proposal comes only a month after the Fermi lab began a two-part project that may bring the United States back into competition with the Europeans in high-energy physics after this country fell behind recently.

Texas already is bidding to be the site of the giant new machine, which the universities of Texas, Houston, Texas A&M and Rice have named the "Texatron." As an incentive to locate in Texas, the universities hope to pay for the land on which the machine would be built, as well as the cost of constructing the huge tunnel in which the machine will operate.

That could amount to 20 percent of the cost of the machine, according to Robert Tribble, chairman of the physics department at Texas A&M. The key technology that would allow the building of such a massive machine has been developed by the Department of Energy at Fermilab. It is the art of making what are called "superconducting" magnets.

Magnets and other electrical devices use a great deal of energy to overcome the resistance electrical energy encounters as it passes through wires. But if the wires are frozen to about 450 degrees below zero Fahrenheit--within a few degress of absolute zero--all electrical resistance disappears.

At Fermilab it was proved for the first time that large, extremely precise machines can be operated at super cold temperatures, thus saving energy and increasing the power of the machine at the same time. Superconductivity can save about half to two-thirds of the energy of running a large machine, if the Fermi accelerator is a guide.

Now physicists want to push the technology ahead, and virtually the entire field of high-energy physics is unanimous in approving the bold plan, according to several leading physicists interviewed over the weekend. A Department of Energy panel recommended that Secretary Donald P. Hodel put research money into the budget immediately.

At a news conference yesterday James Liess, Energy's associate director for high energy and nuclear physics, said that while the government has not committed itself to the project, "we are seriously considering an intensive research and development effort," the first step toward building the machine. He said the administration considers the proposal a "very exciting opportunity."

The proposed machine would take about 10 years to construct and cost between $1.5 billion and $2 billion dollars. It would be able to operate with the power of 40 trillion electron volts, about 100 times greater than any current machine.

One electron volt is the energy gained by an electron jolted with one volt of power. The energy of each particle in a fire, for example, is a few electron volts, while the energy of particles inside the sun is a few million electron volts.

Leon Lederman, director of the Fermi accelerator, which is America's most powerful machine, said the new machine has physicists excited for a number of reasons.

Many point to the power of new discoveries to revitalize industry, he said. The new machine would help push along the technology of superconductivity, which is expected to be useful to industry, as well as to provide insights into forces of nature and how they may be manipulated.

The machine is also exciting because, "We have reached a crisis in physics . . . where we are drowning in theoretical possibilities not based on a single solitary fact." He said that experimental physics has reached the point where physicists are beginning to understand a new level of matter, the level at which photons and quarks exist.

But that new understanding leaves only larger questions about what matter is, how it formed and why it behaves as it does.

The fundamental particles, the quark and the photon and their variants, turn out to come in very large families. There are 36 types of quark, and more than 100 fundamental particles. Physicists do not understand why there are so many, or why they vary so greatly. Some have virutally zero mass while others are millions of times more massive.

Physicists hope with a larger machine to peer, as if through a larger microscope, deeper into matter for answers. In the proposed machine, atomic particles will be drawn by magnets through the 100-mile circular tube, speeding them up with each circuit around, until they reach the speed of light.

Then two particles collide head-on; they annihilate one another and the resulting smear of pure energy instantly congeals into a new array of particles, giving clues to what the original particles were made of. In Lederman's metaphor it is like smashing together two speeding garbage cans. The idea is to find out what they contain. After the cans collide, physicists try to reconstruct the contents and their arrangments within the cans. Lederman also said that the simple, practical value of such a project is only part of the reason for doing it. "We are exploring outer space, and with a machine like this we can explore inner space . . . . There is a certain pride in doing these great projects . . . . If we abandon trying such things we do it at a great peril."

There are dozens of particle-accelerating machines in operation around the world, looking at different levels of matter. The new "super-collider" would be the most advanced machine to explore the unknown territory of matter. It would be like building a telescope immensely more powerful than any now in existence, to search for new objects at the edge of the universe.

Other machines, such as the accelerator planned for Newport News, Va., will instead explore closer, more familiar targets.