ITHACA, N.Y., June 22 - Off campus, energy shortages may be creating talk of a nuclear power renaissance. But on campuses around the country, the technology's infrastructure is dying.

Here at Cornell, the trustees voted unanimously last month to close the university's research reactor, the only one in New York State and the Ivy League's last. There was a petition drive, a demonstration, even offers by the nuclear staff to have other departments use the reactor, all to no avail.

The University of Michigan and the Massachusetts Institute of Technology are considering doing the same.

Where 40 campuses had reactors in 1988, there are 28 today, and only about half of those operate more than a few hours a year. Nuclear departments and programs are disappearing or being merged into electrical or mechanical engineering departments, where they fare worse in the perennial university battles for faculty slots and other resources.

The decline in nuclear engineering programs and campus reactors reflects a decline in student interest that has paralleled the industry's decline.

"It's a fact of life that kids are pretty practical these days," said Marvin M. Mendonca, who oversees licensing for nuclear reactors other than power reactors for the Nuclear Regulatory Commission. What they choose to major in, Mr. Mendonca said, depends on "what they think they can get for a job."

Some question what this will mean for the future of the industry. Nuclear departments like Cornell's have supplied the senior engineering staff and executives at the nation's 103 commercial power reactors, as well as engineers for the regulatory commission.

"If we do build new nuclear plants," said William D. Magwood IV, director of the Office of Nuclear Energy, Science and Technology at the United States Department of Energy, "we're going to need people who understand the technology and can operate the plants safely, and the places that train those people are beginning to disappear. It's very depressing."

Jeffrey Merrifield, one of the five members of the regulatory commission, said that among the agency's problems was its aging staff, with five times more staff members over age 60 than below age 30.

The Energy Department counted just 570 students nationwide majoring in nuclear engineering in 1997, down from 1,500 five years earlier. The drop in student interest is somewhat paradoxical, industry experts say, since there is substantial demand for nuclear engineers.

Even without new nuclear plants envisioned by the Bush-Cheney energy plan, the applications by dozens of reactors to extend their licenses to 60 years from 40 mean a longer future for the industry.

M.I.T. and the University of Michigan face another problem, a need for modernizing their aging reactors, requiring an investment that has led administrators to consider shutting down. The Energy Department has promised those two universities and Cornell $250,000 each, and Senator Pete V. Domenici, a New Mexico Republican who is on the Energy Committee, has proposed bigger increases.

Some hope a new generation of students might show more interest.

"We're coming out of the `nuclear is bad' phase," said Danielle K. Hauck, who worked on a petition to save the Cornell reactor and just completed her junior year here.

Ms. Hauck, who was born the year of the Three Mile Island accident and was in elementary school at the time of the Chernobyl one, says she likes the idea of new reactors.

But Cornell administrators say they have better uses for their resources and even for the real estate the reactor occupies. A billboard near the reactor building here advertises a nanotechnology center that will open in the spring of 2004.

The aging of those with nuclear expertise is a cause of decline as well as a symptom. Robert C. Richardson, vice provost for research at Cornell, wrote in a letter in April to Cornell's president that the reactor should be closed, in part because "very few if any young faculty are enthusiastic about the science, about devoting their own careers to building or improving the facility, or about utilizing the reactor heavily."

This is in part the university's policy; in the mid-1990's it ended its program in nuclear engineering and reassigned the faculty to other departments, virtually assuring that no new faculty members in the field would be hired.

The nuclear staff at Cornell fought hard for survival. They volunteered the reactor as a tool for archaeology, geology and even art history, and students collected more than 1,900 petition signatures from their classmates to keep the place open. Twenty-five people demonstrated outside a faculty senate meeting, surely one of the few pro-nuclear demonstrations on campus in history. But none of this produced sufficient allies.

The reactor director, Kenan Unlu, is working on a method to analyze silicon for dopants, which are trace contaminants the electronics industry seeks to insert in the raw material of computer chips.

A professor of dendrochronology has students slicing up wood samples from trees and from the beams of medieval buildings from the eastern Mediterranean. The wood rings are then put in test tubes and irradiated in order to date the wood.

Other experiments measure trace elements in volcanic rock. The veterinary school would like to try treating cancers in animals' brains with neutrons from the reactor.

But none of this seems likely to sustain the reactor in the absence of a nuclear engineering program.

William B. Streett, who was dean of engineering from 1984 to 1993 and is now retired, pondered the problem of allocating resources based on what the students were signing up for.

"Doesn't an institution like Cornell have a responsibility to society to support areas not currently popular or fashionable?" Dr. Streett asked. "I think it does."

At some time in the future, he said, the country is very likely to return to nuclear power. "But whether nuclear engineering and science qualifies is strictly a matter of judgment, and for better or worse, the trustees have not supported it here."