TUXEDO PARK: A Wall Street Tycoon and the Secret Palace of Science That Changed the Course of World War II.

By Jennet Conant. Simon & Schuster. 330 pp. $26

They don’t make rich nerds like they used to. Look at Bill Gates, frittering his life away in trench warfare with the Justice Department instead of using his gazillions to, say, colonize and air-condition Mars. Compare him to Alfred Loomis (1887–1975). Having made a fortune of Gatesian proportions in the electric utilities boom of the 1920s, Loomis got out just before the stock-market crash. He retreated to a castle in the cloistered New York village of Tuxedo Park to pursue his youthful passion for physics. Patron to the finest scientific minds of his generation, he assembled teams of researchers who would help win World War II by developing first radar and then the atomic bomb. And he managed to do it all without attracting the notice of journalists or historians.

Until now. Conant, a former Newsweek reporter whose grandfather and great-uncle were Loomis cronies, weaves a skillful account drawn from family correspondence and interviews with the aging remnants of the tycoon’s networks. She pierces the protective curtain the publicity-shy Loomis hung about himself, and in the process manages to make him a sympathetic character. Not easy, considering that she is writing about an investment-banker-turned-physicist, two species popularly supposed to rank with reptiles on the warmth-and-kindness scale.

Certainly Loomis’s wife found him chilly. Horrified that he refused to intervene when their three teenage sons announced plans to, variously, cross the Atlantic in a 35-foot boat and scale remote peaks in India, Ellen Loomis stormed at her husband: "Will you still believe in your theories about children if all three of them get killed this summer?" Replied Loomis: "Three is not a sufficient number to prove any scientific theory."

If Loomis’s paternal skills were uncertain, there could be no doubting his passion for physics. He spent countless millions of dollars following his whims, often with spectacular results. Fascinated by reports of a French submarine-detection device that killed any fish that swam across its beam, Loomis built a 50,000-watt oscillator and fathered the science of ultrasonics. Puttering around the lab, he designed the nation’s first working electroencephalograph, to measure and record a brain’s electrical activity.

Nothing, however, surpassed his work with the relatively new fields of microwave technology and small-particle physics. An army weapons researcher during World War I, Loomis understood the military implications well ahead of most scientists (or military men, for that matter). At a time when little government money went into scientific research, Loomis poured his own cash into the work and marshaled additional support from universities and philanthropists. Many of the key men in the development of the atomic bomb—Niels Bohr, Ernest O. Lawrence, Enrico Fermi, Arthur Compton, Vannevar Bush—were members of the Tuxedo Park team of the 1930s. And his research on microwaves was so advanced that in 1940, with London tormented by Nazi bombers and the Roosevelt administration finally awakening to the danger, Loomis was put in charge of the new government radar lab at the Massachusetts Institute of Technology.

For all the significance of Loomis’s later work, Conant especially enjoys herself when describing the early years of his Tuxedo Park lab. Like benign Dr. Frankensteins, the eminent scientists in Loomis’s castle would stay up all night boiling frogs with high-frequency beams, transplanting beating turtle hearts into petri dishes, poisoning themselves with experimental bathtub gin, and furgling one another’s wives. And, like Frankenstein, they occasionally ran disastrously amok: The bastards invented the first radar gun. Some things really are better not known to man.