Suspicions raised by biology postdoc Margot O'Toole about a few lines of data in a 1986 paper in the journal Cell by former MIT immunologist Thereza Imanishi-Kari, Nobel laureate David Baltimore and several co-workers snowballed into the late 20th century's most public dispute over fraud and error in science. In 1991 The New York Times called it "A Scientific Watergate," but a far more apt comparison is available today in the Whitewater investigation. Both the Baltimore saga and the Whitewater inquiry were pushed forward by obsessive, heavy-handed investigators with political agendas; both cases feature prominent targets who acted by turns indignant and contrite; and both matters dragged on interminably, with investigators chasing after details decreasingly related to the original alleged offenses. And just as the Whitewater inquiry long ago grew too convoluted for most citizens to grasp, the Baltimore case hinged on immunological experiments so arcane-and produced so many conflicting interpretations of the allegedly fabricated experimental records-that impartial observers rarely knew what to believe.

Daniel Kevles knows. An eminent historian of science at Caltech (where Baltimore was named president in 1997), Kevles not only examined Imanishi-Kari's lab notebooks, which document the disputed experiments on the effect of gene transfers on antibody production in mice, but also waded through thousands of pages of testimony and interviewed more than 70 scientists and others caught up in the case, including all the principals. He concludes-as a Public Health Service appeals board did in 1996-that Imanishi-Kari is innocent of fakery. He also argues that Baltimore's famously combative defense of Imanishi-Kari before muckraking Rep. John Dingell in 1989 congressional hearings was justified, if brash.