It makes wine smell like a barn, wet leather, horse sweat, or burned beans. It is called "brett," and it produces an often-pungent aroma in wine.

Scientists at Cornell University are starting to unravel the chemical mysteries that produce the curious aroma found in fermented beverages like wine and beer. "We're trying to understand the aroma's chemical profile, essentially the aroma chemistry of the wine and the microbiology that created it," says Jonathan Licker, a Cornell graduate student in food science.

Brett is the nickname for a wild yeast, Brettanomyces. Winemakers often find this yeast where they least want it: in barrels, production lines, bottled wine and even on the grapes themselves. Licker explains that without proper sanitation, particularly after the harvesting of grapes, Brettanomyces can thrive in a winery. "It can become the unwelcome guest that takes over your home," he says.

Licker, Terry Acree, Cornell professor of biochemistry, and Thomas Henick-Kling, Cornell associate professor of enology and microbiology, are using a procedure known as gas chromatography-olfactometry to further understand the odor-active compounds found in wine. Christoph Egli, a Cornell postdoctoral associate of molecular biology in Henick-Kling's laboratory, is attempting to genetically differentiate and identify, using DNA fingerprinting, the Brettanomyces strains. The researchers all work at Cornell's New York State Agricultural Experiment Station in Geneva.

"The chemistry of wine is very complex," says Licker, who has chemically analyzed samples of domestic cabernet sauvignon from three different years, using the odor-analyzing methods developed at the Agricultural Experiment Station. He initially found striking differences in chemical composition in the samples provided by a winery from 1989, 1992 and 1994 batches. The professional noses at the winery believed the 1989 sample was "bretty," the 1992 sample was not too bretty and the 1994 sample was fruity and did not have a bretty flavor.

The trained sensory snouts at the experiment station confirmed the results with a technical descriptive analysis of each wine. "TheBrettanomyces yeasts isolated from the bottled wine were the same yeast strain within a vintage, suggesting dominance of a particular yeast in a batch of wine," Egli and Craig Mitrakul, who received an M.S. from Cornell in 1997, wrote in a recent research paper. The same yeast strain completely dominated wines from the vintages of 1992 and 1994, suggesting that the yeasts took hold in the cellar rather than in the vineyard, they wrote.

In the past, some brewers sought a bretty aroma for their beer, and even winemakers savored it during the early part of this century, but today the yeast is regarded as overwhelming in some wineries and the wine industry is concerned that this wild yeast might be out of control.

"Our ultimate goal will be to provide winemakers with more tools to help them control and optimize brett aroma in their wine production," says Licker. "You can still find brett in some of the best French Bordeaux and domestic cabernet sauvignons. Many people -- winemakers and consumers alike -- think a little brett is a great thing."

Licker and Egli will present "Brettanomyces: Microbiology and Aromas," at the New York Wine Industry Workshop, during the wine sensory workshop, "The Good, the Bad, and the Ugly -- a Review of Wine Aromas," Saturday, April 4, at the Geneva Lakefront Ramada Inn, Geneva.