A scientist can go bad in any number of ways. Some of them, such as trimming facts to fit theories, are lamentable but almost understandable. Others, such as making up facts altogether, are unforgivable.

One way of going bad, however, is harder to judge. A reputable, even eminent scientist discovers something unexpected and nearly undetectable. The scientist is intrigued, then enthralled, then obdurately convinced. A few fellow scientists concur, but others, unable to repeat the discovery, attack. War breaks out. The defenders claim greater perceptual acuity and explain away all findings to the contrary. The attackers finally gather enough counterevidence, and the original finding is dismissed. Gratzer, a British biophysicist and frequent contributor to Nature, calls this insistent embrace of an untenable hypothesis "communal derangement"; physicist Irving Langmuir called it "pathological science."

Around 1900, for example, the distinguished French physicist René Blondlot announced the discovery of "N-rays": nearly imperceptible electromagnetic radiation that passed through quartz but not through water. Scientists all over Europe repeated his experiments. Some saw the radiation and made further claims—one announced that N-rays heightened the sensitivity of the human retina—but others couldn’t detect it. N-ray defenders derided the critics as insufficiently perceptive. "If N-rays can only be observed by rare privileged individuals," responded one critic, "then they no longer belong to the domain of experiment." Finally, Blondlot claimed to see N-rays even after a colleague had removed an essential part of the experiment. N-rays disappeared from physics.

The Undergrowth of Science assembles case studies in pathological science: Groups of growing cells supposedly emit radiation. Changes in an animal’s body are inherited by the animal’s offspring. Implanted monkey prostate glands rejuvenate aging men. Disagreeable inherited traits, from imbecility to alcoholism to criminality, are abolished by sterilizing the people who inherited them. Radiation given off by menstruating women kills microorganisms. Fusion, the energy source of the Sun, is reproduced in a jar.

None of these case studies rose to outright fraud. Instead, they resulted from a very human combination of ambition, overcommitment to a dubious investment, hero worship, mass hysteria, and an aversion to being wrong, especially in public. Scientists, Gratzer observes, "are as much a prey to human frailty as anyone else, and their capacity for unbending objectivity is circumscribed."

Pathological science remains with us— fusion-in-a-jar dates from the late 1980s— but it can be difficult for nonscientists to recognize. Gratzer’s cases seem like the usual science news that first sounds unreasonable and then turns out to be right or wrong, either one. Throughout history, scientists have successfully defended marginal data, and theories that sounded silly have proved revolutionary. And, though Gratzer explains the experiments thoroughly and clearly, the general reader doesn’t know the principles that make, say, radiation from growing cells just plain impossible. Perhaps such principles are uncodified and unspoken. If so, readers have to take a lot on faith.

Still, they’re going to like this book. The writing is elegant and unusually intelligent. Science and politics are credibly interwoven. And the hapless scientists, clinging to their theories as the counterevidence mounts, come across as at once terribly weird and terribly normal.

—Ann Finkbeiner