On October 17, 1946, sixteen months and one day after his first wife Arline died of tuberculosis at a sanatorium in Albuquerque, Richard Feynman sat down at his desk in Ithaca, New York, and wrote her a letter.
He was 28 years old. He had already, in the previous three years, helped design the atomic bomb at Los Alamos, calculated neutron equations for nuclear reactors, watched the Trinity test through the windshield of a parked truck, and become one of the most respected young theoretical physicists in the United States. He had moved to Cornell University to teach. He was, by every external measure, a man getting on with his life.
The letter to Arline is two pages long. It addresses her as “D’Arline” — a private nickname. It tells her about his work, about the people he is meeting, about the small ordinary contents of a life she was no longer in. It says, near the end, that he loves her now more than two years after her death and that he knows she would tell him not to be silly.
And then it ends with the postscript that has, in the eight decades since, become one of the most quoted single sentences in all of American correspondence:
“PS Please excuse my not mailing this — but I don’t know your new address.”
He sealed it. He kept it for the rest of his life. It was found, still sealed, in his papers after his own death in 1988.
This is one of the things you have to understand about Richard Feynman. The man who taught the world to question everything — who picked the locks on America’s atomic bomb secrets to embarrass the people who had hired him, who exposed the cause of the Challenger disaster on national television with a glass of ice water — kept a sealed letter in a drawer for forty-three years that was addressed to a woman he could not stop loving and could not, by the operating rules of physics, deliver it to.
Both of those things are him. Neither cancels the other.
Richard Phillips Feynman was born in Queens, New York, on May 11, 1918, the son of secular Jewish parents Lucille and Melville. His father was a uniform salesman who taught him, from earliest memory, that the name of a thing is not the same as understanding the thing. He earned his bachelor’s degree at MIT, his Ph.D. at Princeton in 1942, and was recruited at age 24 to Robert Oppenheimer’s Manhattan Project. He arrived at Los Alamos in March 1943.
At Los Alamos, he became the project’s most reliable, most productive, and most flagrantly unmanageable young calculator. Hans Bethe, the head of the Theoretical Division, made him a group leader within weeks. The two of them developed what is still known as the Bethe-Feynman formula for calculating the explosive yield of a fission bomb.
And in his spare time, he picked locks.
Feynman discovered that the filing cabinets used to secure America’s atomic-bomb research could be opened with a screwdriver and a length of wire. He discovered that, of the cabinets that had supposedly been upgraded to combination locks, roughly one in five had been left set to the factory default. He discovered that the rest had been set to dates and addresses and other guessable numbers by physicists who weren’t paying attention.
And then he discovered the combination he made the most famous. He worked out, over the course of an afternoon, that the cabinet of his colleague Frederic de Hoffmann — which contained a substantial part of the project’s classified research — would be set to a combination a physicist would find easy to remember. He tried 27-18-28, the first six digits of e, the base of the natural logarithm: 2.71828.
The cabinet opened. So did the next two cabinets, which had the same combination. Inside were de Hoffmann’s notes on the design of the bomb.
Feynman left a note in the cabinet. “Guess who?” He left several more. He told everyone what he was doing. He was trying to make a single point: the security at Los Alamos was theater. The locks looked like they were doing something. They weren’t. If a casual prankster could get inside America’s nuclear secrets with a paperclip, an enemy with actual training and resources could empty the project in a weekend.
Some of the senior officers found this hilarious. Many did not. Klaus Fuchs, a physicist who often loaned Feynman his car for the weekend trips to Albuquerque, was later revealed to have been passing real bomb designs to the Soviet Union the entire time.
Feynman had been right.
He had been making those Albuquerque trips, weekend after weekend, in Fuchs’s borrowed car, to sit with his wife.
Arline Greenbaum had been Feynman’s high-school sweetheart in Far Rockaway. She was funny and irreverent and fearless. She was diagnosed with tuberculosis before they were married, and the doctors gave her a few years at most. Feynman married her anyway, in 1942, in a civil ceremony on Staten Island, against the wishes of his parents. He carried her up the steps of the city hall building because she could no longer walk.
When the Manhattan Project moved to New Mexico in 1943, Oppenheimer arranged for Arline to be admitted to the Presbyterian Sanatorium in Albuquerque, two hours from Los Alamos. Feynman drove down on weekends and spent every Saturday with her. She wrote him letters in code, knowing he loved a puzzle. She had stationery printed that read RICHARD DARLING, I LOVE YOU! POPPA across every sheet, and used it to write to him about ordinary domestic things, knowing it would make him laugh in the middle of a war.
On June 16, 1945, the call came that she was failing. Feynman drove to Albuquerque in Fuchs’s car. He sat with her for hours. She died that evening. He recorded the time in his notebook with a single word.
Death.
Then he drove back to Los Alamos and went back to work.
When colleagues asked him about it, his answer was the famous Feynman line: she was dead. How was the program going. He did not break down for weeks. He broke down, finally, in a department store in Oak Ridge, when he saw a dress in a window that he thought Arline would have liked.
Exactly one month and one day after Arline died, on July 16, 1945, the United States detonated the first atomic bomb at the Trinity site in southern New Mexico. Feynman watched the test through the windshield of a parked truck — he had reasoned, correctly, that the windshield would block ultraviolet radiation. He was the only observer who saw the explosion without protective eyewear.
He came back to civilian life, took a teaching job at Cornell, and discovered, slowly, that he could think again. He went on to win the 1965 Nobel Prize in Physics, jointly with Julian Schwinger and Sin-Itiro Tomonaga, for the formulation of quantum electrodynamics — the theoretical framework, illustrated in what became known as Feynman diagrams, that describes how light and matter interact at the subatomic scale. He spent the rest of his career at Caltech. He played the bongo drums. He learned to draw in his forties. He gave the most famous undergraduate physics lectures of the 20th century. He wrote two best-selling books of stories about his life — Surely You’re Joking, Mr. Feynman and What Do You Care What Other People Think? — that contain, alongside the lock-picking and the Trinity test and the love story, his own honest accounts of behavior toward women that has not aged well, and that should not be airbrushed when his life is described accurately.
His final public act was the moment most Americans of a certain age remember him for.
On January 28, 1986, the Space Shuttle Challenger broke apart 73 seconds after launch and fell into the Atlantic Ocean. Seven crew members died, including Christa McAuliffe, the schoolteacher who had been the public face of NASA’s Teacher in Space program. Feynman, by then dying of two rare cancers, agreed reluctantly to serve on the Rogers Commission investigating the disaster.
He was 67 years old. He was running out of time. He decided, in the quiet way of someone who knew he was running out of time, that the investigation was being managed.
On February 11, 1986, during a televised hearing of the Rogers Commission, Feynman asked the chairman for a glass of ice water and a sample of the rubber O-ring material that sealed the joints of the solid rocket boosters. He used a small C-clamp to compress a piece of the O-ring. He dropped it in the ice water. He held it down. He waited a few minutes.
Then he released it.
The rubber did not spring back. It stayed compressed. It had lost its elasticity in the cold.
Feynman looked up from the table and said, calmly, that the launch temperature on the morning of January 28 had been 36 degrees Fahrenheit.
The room went silent. The rest of the investigation was, in many ways, a formality. The O-rings had failed because they had become brittle in the cold. NASA had launched the shuttle anyway. Seven people had died because the agency had trusted its own paperwork instead of its own engineers.
Feynman wrote a personal appendix to the Commission’s final report. He had to threaten to remove his name from the main report to get the appendix included. It contained the sentence that has been carved into more engineering school walls than any other sentence Feynman ever wrote:
“For a successful technology, reality must take precedence over public relations, for nature cannot be fooled.”
He died two years later, on February 15, 1988, in Los Angeles. He was 69 years old. His sealed letter to Arline was still in his papers. It would not be opened by anyone outside his family for another seventeen years.
When it was finally published, in a 2005 collection of his letters edited by his daughter Michelle, the contents were exactly what you would expect from a man who had spent forty-three years in an unsent conversation with a woman who had been dead since the war. He told her about his life. He told her, with the slightly embarrassed honesty he reserved for the people he had loved most, that he loved her still and that he was sure she would tell him to stop being sentimental about it.
Then he ended the letter the way he ended it. Because he could not, by the operating rules of physics, do otherwise.
Please excuse my not mailing this. But I don’t know your new address.
The man who had spent his life refusing to accept the answers other people had told him to accept — who had picked locks to prove security was a fiction, who had dunked rubber in ice water to prove engineers had been ignored, who had spent forty-three years waiting for an address that was not going to arrive — kept the letter sealed.
Because some things, in the end, you do not test.
