Using a novel observational technique that one scientist likened to alchemy, an 85-year-old Nobel laureate has mixed lasers with starlight and, with a sprinkling of luck, discovered that some old stars are larger than expected.

If it holds up, the finding could force astronomers to rethink theories about how stars evolve and behave.

Charles Townes, a professor emeritus at the University of California at Berkeley, led a team of researchers that peered through the deceptive gas veils of three well-known aged stars and found each of them to be larger than previous measurements.

Betelgeuse (often pronounced "beetle juice"), Antares and Mira are red giants, a class of ancient objects that are running out of fuel and preparing to kiss their stellar lives goodbye. Our own Sun will one day become a red giant, likely swallowing Earth as it swells and contracts over thousands and millions of years.

Some red giants, like Mira, also oscillate in brightness on short time scales of months or years for reasons scientists don't fully understand. Mira (officially called Omicron Ceti) is in fact the namesake for a special class of red giants whose brightness oscillates on relatively short cycles.

Mira's intensity waxes and wanes every 330 days. At its dimmest, it is invisible to the naked eye and just 1 percent as bright in the sky as when it reaches peak brightness.

"Mira has an awesome amount of material around it," Townes explained in a telephone interview, "and when it oscillates in intensity it blows off material, every year, as it gets brighter and weaker, brighter and weaker."

He says the star's diameter ought to change in step with this oscillation.

"Theory says it's probably at its biggest at about the time when it's at its brightest, but we don't know that for sure," he said, adding that research will continue later this year in an attempt to see how the size and shape of the star change over time.

Meanwhile, the leading theory that explains this short-term oscillation requires that Mira be no larger than previous estimates.

"If the [Townes] results hold up, and if current theory is at a loss, then a real problem for astrophysics faces us," Dupree told SPACE.com.

If theorists have to rethink how red giants work, they can be excused for not having had much to go on. Red giants are surrounded by so much gas and dust that scientists have difficulty observing their true size, shape and behavior.

So knowledge of these stars, which are often more than 10 billion years old, has relied in part on estimates of the type and amount of debris surrounding them, along with assumptions based on what we know about our own Sun, which is roughly 4.6 billion years old and not yet anything like a red giant.

"But what if other stars are not like the Sun?" asks Peter Tuthill, an astronomer at the University of Sydney. "The only way to really find out is of course to go out and make measurements, and the stars studied by Townes are very unlike our Sun, so the orthodox theory is on very shaky ground indeed."

With the new laser technique, the Townes team set out a decade ago to map the quantity and distribution of gas and dust around the three stars. They observed the stars in infrared light using Berkeley's Infrared Spatial Interferometer (ISI). The interferometer amounts to a pair of telescopes mounted on commercial tractor-trailers and hooked together to increase their resolving power by gathering starlight in tandem from two locations.

But as the researchers began to study the data and the method, they made the fortuitous realization that the important aspect to it all was the fact that they could peer right through the cloudy cocoon to measure the stars themselves.

"These new mid-infrared measurements are important, as this wavelength avoids some of the problems and complexities which make it difficult to interpret earlier data," Tuthill says. "Hopefully, with the new ISI measurements, a more complete model of the atmospheres of red giants might finally be constructed."

The results, if confirmed, could apply to many or most red giants, Townes said. But to a lesser extent the results might have implications for other stars.

Very young stars are surrounded by natal debris clouds and have also evaded measurement; they might be studied with the new technique. Even many "normal" or middle-aged stars might now be found to be slightly larger than thought, Townes said, adding that the difference would be minimal among these stars, however.