She said the first thing that came into her head: “You were poisoned.”

“Yes,” he said, rising from the chair and crossing the room till he was standing right there before her, peering into her eyes as if looking for something he’d misplaced. “But how? And what does it mean?”

She couldn’t think. He was too close to her, so close she could smell the lozenge on his breath. “I don’t know,” she said. “That you’re lucky?”

He laughed aloud. “Lucky, that’s exactly it. We’ve got something here, I feel it, I really do.”

“No,” she said, taking a step back, all the precautions, all the rules, everything she’d learned in her apprenticeship and her time here as a full-fledged employee, all the horror stories about inadvertent poisonings, splashes and caustic burns flapping through her brain like flocks of black-winged birds—Never pour water into acid, All volatile materials to be handled in a hood with the exhaust fan on, Always cover up and always wear gloves. “What I mean is you’re lucky it wasn’t worse. Lucky”—she paused and felt something float up in her, some amalgam of fear and loss and love—“lucky you’re alive.”

The chemical was one of the ergot compounds Herr H. first synthesized in 1938, when she was just sixteen and working as an au pair in Neuchâtel and he an ambitious young chemist hoping to synthesize an analog to Coramine, a cardiovascular stimulant produced by Ciba, one of the company’s biggest rivals. The structure of Coramine—nicotinic acid diethylamide—was strikingly similar to that of lysergic acid, the basic constituent of the ergot alkaloids his mentor, Arthur Stoll, had isolated eighteen years earlier, and so Herr Hofmann reasoned that it would have similar properties and uses. He spent three years on his investigations, which yielded one useful product—ergobasine, marketed by the company for obstetrical use, as it promoted contraction of the uterus and stanched bleeding after birth—and a series of lysergic acid derivatives that unfortunately didn’t show much promise, including the twenty-fifth iteration, lysergic acid diethylamide. The pharmacological unit found it 30 percent less effective than ergobasine, though it did seem to have a mild stimulatory effect in animal trials, producing a degree of restlessness in rats, rabbits and dogs. But Sandoz was not in the business of marketing stimulants for the lower animals, and the compound was shelved, along with its twenty-four predecessors.