One of those things was a reference that Deryagin made to sodium contamination in his original paper. “Deryagin buried that, and in a journal that, at the time, was only published in Russian,” Rousseau said, “so people outside the Soviet Union didn’t know much about it.” When Rousseau’s team performed an analysis for sodium, they found it in any polywater they were able to produce, along with calcium, potassium, and chlorine.
How did these contaminants make it into the supposedly pristine samples? Rousseau had an idea. “I used to play handball regularly, so one time, I squeezed out a sample of sweat from my T-shirt,” he said. Tests showed that the infrared spectrum absorbed by his sweat was virtually identical to the pattern produced by polywater.
He had discovered the disappointing truth: Polywater was just sweaty water. Sodium lactate, a salt derived from human sweat, was entirely responsible for polywater’s otherwise inexplicable qualities. Tiny amounts of perspiration must have dripped or seeped into the chambers then were picked up and carried into the capillaries by the water vapor as it condensed. Deryagin, Lippincott, and my great-uncle hadn’t found a strange, revolutionary form of water; they’d been fooled by the sweat that dripped from their own brows.
Rousseau’s findings, published in the January 1971 issue of Science, were so devastating that polywater’s proponents barely put up a fight. “They said that their material was clean, and ours was dirty, but that got resolved pretty quickly,” Rousseau says. Within months, Lippincott and Stromberg (my great-uncle) published their own analysis, which confirmed that contaminants were to blame; soon afterward, Deryagin’s team followed suit. Just a few years after it was born, polywater was killed. There would be no more research grants and certainly no Nobel Prizes. Polywater had been an artifact of flaws in these scientists’ experiments—and a figment of their imaginations.
* * *
You could charitably argue that the polywater saga wasn’t really a scientific failure, but a success. Scientists continually come up with new theories and disprove them—that’s the scientific method, the way we improve our understanding of the world around us. Most of this messy work stays behind the curtain of unpronounceable names and esoteric formulas. But polywater was uncommonly interesting and got picked up by the press, so its collapse occurred in the glare of the public eye.
But on the other hand, if you look closely enough at polywater, you’ll notice a disconcerting thread that connects it to a handful of other high-profile scientific failures and runs through the present day. In 1989, for instance, two teams of electrochemists independently announced that they’d achieved the holy grail of nuclear energy, cold fusion—that is, controlled nuclear reactions at room temperature. But other scientists were never able to reproduce their results. The amount of nuclear reactions they supposedly detected, it turned out, was within their instruments’ margin of error. Similarly, in 2011, a group of European researchers claimed that they’d observed subatomic particles called neutrinos moving faster than the speed of light, violating our current model of physics. Subsequent analysis showed that their experimental setup included a fiber-optic cable that was incorrectly installed and a clock that ticked at the wrong speed, causing them to time the speed of their neutrinos just a hair too fast.
Rousseau thinks of these episodes as “pathological science”: cases where tiny sample sizes (or effects that are otherwise difficult to measure) and a potentially revolutionary (and career-making) finding lead scientists astray. These aren’t instances of outright fraud, but of unconscious bias. A scientist misinterprets a small amount of data as a paradigm-shifting discovery, and once in that mindset, he or she sees all subsequent information through the same lens.
It’s hard to imagine a better example of this than polywater. “The data was right, but our interpretation was wrong,” my great-uncle told me. Technically, he’s correct: His team produced an accurate infrared spectrum reading of the tiny amounts of the sample they tested, and it didn’t match any other substance in their database. But if they had maintained a more skeptical approach, they might have considered other possibilities—such as sweat—before drawing up a novel chemical structure and giving their material a catchy new name. “The critical thing, whenever you find something that’s really strange, is to try to disprove it, rather than trying to prove it’s there,” Rousseau reminded me. In other words, in their excitement, my great-uncle’s team forgot the maxim popularized by Carl Sagan: Extraordinary claims require extraordinary evidence.
To his credit, Uncle Bob isn’t defensive or embarrassed about this mistake—he’s happy to reflect on it all these years later and readily admits that he wishes it had turned out differently. “If it really were true, this would have been Nobel Prize stuff,” he said. At the same time, he harbors no illusions about polywater: “Most mistaken hypotheses in science aren’t entirely wrong, they just have to be modified a bit,” he told me toward the end of our conversation. “I think this was unusual in that we really were 100 percent wrong.”
He’s right: His team’s error was large indeed. But like cold fusion and faster-than-light particles, it wasn’t fraudulent or ill-intentioned—it was simply a reminder that humans guide scientific research, not robots. And humans are subject to desires, which make objectivity really hard to achieve. Ultimately, the discovery of polywater didn't tell us anything accurate about the chemical nature of water, but it did reveal something fundamentally true about the character of the human mind. If you discovered polywater, you'd want it to be real, too.