The Highest Authority in Science

What happens to the right answer when it comes from the wrong man.

“As Lord Kelvin is the highest authority in science now living, I think we must yield to him and accept his view.”

Mark Twain, “Was the World Made for Man?”, 1903

In 1869 the most careful man in England was losing sleep over a number he could not argue with.

Charles Darwin’s theory needed time. Oceans of it, deep enough that selection could work its slow arithmetic on every living thing, and Darwin had spent his career assuming the Earth would supply it. Now a physicist in Glasgow had run the numbers and delivered a verdict: the planet was young. Far too young. Darwin called the man’s work one of his “sorest troubles.” To another correspondent he confessed he was “greatly troubled at the short duration of the world,” because his theory required an enormous run of time before the first complex animals (the Cambrian). Under the pressure, he did something scientists almost never do in public: he retreated. He had once estimated three hundred million years just for the slow carving of a single English valley, and in later editions of the Origin he cut the number out.

He didn’t doubt what he had seen in the field. He doubted whether he could survive the arithmetic of a man who was never wrong.

Lowercase k

You already know this man’s name, even if you have never once thought about him as a person. It’s in every chemistry classroom on Earth: the temperature scale that starts at absolute zero and counts upward, the one scientists use when Fahrenheit and Celsius aren’t serious enough. The kelvin. Lowercase k, because units of measure named for people lose the capital.

In 1869 he was William Thomson, not yet Lord Kelvin, and he was the closest thing Victorian science had to an infallible man. He had helped write the laws of thermodynamics, the rules that govern every engine, every star, every cup of coffee going cold on every counter. When the first transatlantic telegraph cable died in the water, it was Thomson’s mathematics and Thomson’s instruments that made the second one work, and a grateful Empire knighted him for it. He would end his life as a baron, with a seat in the House of Lords and a grave in Westminster Abbey a few steps from Isaac Newton’s. His colleagues did not treat him as a very good physicist. They treated him as the standard the others were checked against. If Thomson’s arithmetic said a thing, the conversation was over.

And in the 1860s, Thomson turned the full weight of that authority on one of the oldest questions there is: how old is the world?

The Cooling Clock

The Earth, everyone agreed, had started hot. A ball of molten rock, spun off from the young sun. And a hot thing left in cold space does one thing: it cools. Thomson realized this made the planet a clock. Miners had known for centuries that the deeper you dig, the warmer the rock gets; that temperature gradient is the planet’s leftover heat, still leaking out through the crust. Measure how fast the heat is escaping, assume you know how hot things started, and you can run the film backward to the beginning.

It’s the move every detective in every crime drama makes when they kneel next to the body: she’s been dead about six hours, based on how far she’s cooled. Thomson did it to the entire planet. He took the best measurements of his day, applied the physics he had helped invent, and read the Earth’s time of death in reverse.

The numbers came out small. In 1862 he published a range of twenty to four hundred million years. As the data improved he tightened it, and the more he worked the problem, the narrower and more confident the answer became. By the late 1860s he had settled near one hundred million years. By 1897, in his last major statement on the subject, he had squeezed it down to “more than 20 and less than 40 million years, and probably much nearer 20 than 40.” Forty years of refinement, and the margin only ever got tighter.

Twenty million years was a death sentence for Darwin’s theory. The geologists grumbled; their strata and their erosion rates whispered of something far older. Thomson had heard that objection straight from a geologist’s mouth and waved it off. He once asked the geologist Andrew Ramsay how much time Earth’s history required, and Ramsay said he could set no limit to it. “You don’t suppose geological history has run through 1 Billion years?” Thomson pressed. “Certainly I do,” Ramsay said. “10 Billion years?” “Yes.” To Thomson that was not an estimate but a refusal to make one, and putting a hard number on the Earth was exactly his project. But the geologists were readers of rocks, and this was arithmetic, delivered by the man who owned the subject. The calculation treated the Earth the way the physics of the day treated it: a rigid ball, solid all the way down, motionless, its heat crawling out by conduction, the way heat crawls through the handle of a pan.

Cobbler’s Wax

John Perry knew Thomson’s calculation the way you know a house you grew up in. He had been Thomson’s assistant in the Glasgow laboratory, one of the bright young men the great man trained, and he had gone on to a solid, unglamorous career: engineering professor, teacher of teachers, co-author of practical papers, a few years spent teaching engineering in Japan. He revered his old chief. That’s not a figure of speech; it’s in his own published words.

And in 1894, Perry noticed what the calculation was standing on.

Let the inside of the Earth move, he realized, even a little, even absurdly slowly, and the whole calculation comes apart. A stirring interior doesn’t just sit there letting heat crawl out. It ferries heat upward, hauling warmth from the deep interior toward the surface, keeping the outer rock warm for eons longer than conduction ever could. That deeper-is-warmer reading the miners measured would then say nothing about the planet’s age. The clock face was real, but it was connected to the wrong machinery.

The objection writes itself, and Perry knew it: the Earth’s interior is rock. Solid rock. Solid things don’t flow. Except they do, on the right timescale, and you have probably held the proof in your hand. Silly Putty bounces off the floor like a rubber ball if you throw it, hard and definite as any solid. Leave the same putty on the table overnight and by morning it has slumped into a puddle. Solid to a smack, liquid to a century. Rock is the same, with the timescales stretched a millionfold. Perry put it more carefully, in the pages of Nature, addressing the teacher he could not bring himself to attack: “I know that solid rock is not like cobbler’s wax, but 10⁹ years is a long time, and the forces are great.”

Run the numbers with a slowly creeping interior under a thin solid crust and the Earth’s possible age explodes: two billion years, room enough for every stratum the geologists had ever read and all of Darwin’s slow arithmetic besides. Perry had not found an error in Kelvin’s mathematics. There wasn’t one. He had found the assumption underneath the mathematics, and shown that everything rested on it.

The man who defined how the world measures heat had been beaten on a question of heat, by his own former assistant, over the specific issue of how heat moves. Kelvin was not wrong about temperature. He was wrong about transport.

There was a fair objection to Perry, and the careful men of the day raised it. Perry had shown that a moving interior could break the calculation. He had not shown that it did. He could not fully compute a churning Earth, so he approximated it, standing in for the real motion with a high “quasi-conductivity” and asking the reader to grant that solid rock, given a billion years, flows. Nobody could yet demonstrate that it does. The idea that the Earth’s solid interior could stir like something half-liquid would not be proven for another fifty years. A cautious man in 1895 could reasonably call it a strong possibility and want the proof. He would have had a point.

That same cautious man never turned the question around. Kelvin’s calculation rested on its own unproven assumption, that the interior sits rigid and still, and nobody made it clear that bar. The rigid Earth got to be the default, innocent until proven guilty. The moving Earth had to prove a mechanism it could not yet reach, guilty until proven innocent. Two assumptions sat the same distance from proof, and only one was ever asked to walk it. The difference between them was not evidence. It was the name on top.

Perry did not publish. He went to Kelvin privately, in person and by letter, and tried to change the old man’s mind quietly, the way you’d approach a father. He was brushed off. Only then, reluctantly, did he take it to the letters pages of Nature, and even there he opened with an apology: he wrote that he had usually told friends “it is hopeless to expect that Lord Kelvin should have made an error in calculation.” Kelvin published a reply that conceded nothing that mattered. And that was, essentially, that. The community watched the exchange the way you watch two specialists argue in a language you don’t speak, and then everyone went back to work. Perry was a knowledgeable engineer disagreeing with the Empire’s greatest physicist, and the Empire’s greatest physicist had already answered. The question was closed.

Perry let it drop. He spent the next twenty-five years teaching engineers and writing textbooks and, so far as the record shows, never made a public campaign of being right. The answer that would have restored deep time sat in the letters pages of Nature, volume 51, in plain sight, for anyone who cared to look.

Nobody looked for ten years.

Radium!

On the evening of May 20, 1904, a young physicist from New Zealand named Ernest Rutherford stood up to lecture at the Royal Institution in London. He was thirty-two, colonial, confident, and about to explain in public that the Earth was far older than twenty million years. In the audience, white-bearded and formidable, nearly eighty, sat Lord Kelvin.

The story of what happened next is Rutherford’s own, told and retold by him for years and written down by his biographer decades later, so take it as a great scientist’s favorite anecdote rather than a court transcript. As Rutherford told it: “I came into the room, which was half dark, and presently spotted Lord Kelvin in the audience and realized that I was in for trouble... To my relief he fell fast asleep but as I came to the important point, I saw the old bird sit up, open an eye and cock a baleful glance at me! Then sudden inspiration came, and I said Lord Kelvin had limited the age of the Earth, provided no new source of heat was discovered. That prophetic utterance refers to what we are now considering tonight, radium!”

The old man, so the story goes, beamed.

A scene that perfect has usually been improved somewhere along the years of retelling. The physics underneath it was real. Radioactivity had just been discovered, and radioactive atoms turn out to be tiny heaters. The Curies had measured it: a lump of radium salt keeps itself warm, indefinitely, from the inside. The Earth’s rock is full of them. The planet is not a hot brick cooling on a counter. It’s more like an electric blanket left switched on, warm not because it started hot but because it keeps making heat. Kelvin’s clock assumed no new heat would ever be added. Nature had been adding heat the entire time. Rutherford’s flattery gave the old man an exit: nobody had erred, a new source of heat had simply been discovered, exactly as his lordship had allowed for. Kelvin took no exit. He never accepted the new age of the Earth, and he died three years later, still sure, and was buried in the Abbey a few steps from Newton.

That lecture-hall story hardened into the official history, the one you may have absorbed without knowing where you got it: Kelvin was wrong because he didn’t know about radioactivity. It’s in the textbooks and the documentaries, a tidy parable about pride and dumb luck.

It’s also wrong. A century later, three geophysicists went back and did what nobody had ever bothered to do: they put radioactivity into Kelvin’s actual calculation to see what changes. The answer: almost nothing. Sprinkle the crust with every radioactive atom we now know is there, keep Kelvin’s one quiet assumption of a rigid, motionless interior, and his math still spits out a young Earth. The famous fix does not fix it. The thing that actually breaks Kelvin’s calculation is the thing that had been sitting in the letters pages of Nature since 1895: an interior that moves. Perry’s creeping rock, hauling heat upward. Convection.

What finally read the true number was neither flattery nor authority. It was a better instrument. In February 1907 a chemist named Bertram Boltwood measured the lead piling up in a scrap of mineral as its uranium slowly decayed, and read an age of two billion years. He was low by more than half, but the answer had finally climbed into a whole new range, tens of millions to billions. That was 1907, the year Kelvin died. The old authority went quiet and the new clock read billions in the same twelve months.

Radioactive atoms don’t just make heat; they keep time. Each unstable atom decays into a stable one at a fixed pace nothing can speed up, and in a rock, how many have decayed versus how many haven’t is a record of how long the process has run. You have stood in a kitchen and heard this exact mathematics: a bag of microwave popcorn, the pops building to a frenzy and then slowing to a straggle, and from the ratio of popped to unpopped you could tell me, without looking at a clock, roughly how long the bag has been in. Read that ratio in uranium and lead, in rock after rock, meteorite after meteorite, and the answers all point to the same number: the Earth is about four and a half billion years old. Kelvin’s final confident figure was off by more than a hundredfold. Perry’s number, published against his own teacher and then let drop, was within reach of the truth.

Bayswater, 1920

John Perry outlived Kelvin by thirteen years. He lived long enough to see radioactivity celebrated as the answer, long enough to watch the wrong correction become the official story while his right one stayed forgotten. The record shows no letter to the editor, no lecture, no late paper reminding the world of Nature, volume 51. His obituaries, when he died in London in August 1920, are warm about the man: an impulsive, enthusiastic, warm-hearted Irishman, “one of the most delightful of companions.” They praise his teaching. They barely mention the Earth.

He was right, he was ignored, and he never once said I told you so. There is no dramatic deathbed scene to give you, because none exists, and this story has had enough improvements added by people who wanted it tidier than it was.

What remains is the shape of it. A field held a wrong answer for half a century, not because the evidence was hidden, the rocks had said old the whole time, and not because the correction was never made. It was made, in the right journal, by a qualified man, with the arithmetic attached, and then a decade later a wrong correction scooped it, on nothing more than a more famous hand.

None of that makes it a story about experts being frauds. Perry was an expert; so was Boltwood. What broke the wrong answer was more science, not less: a sharper argument in 1895, a sharper instrument in 1907. The field never trusted physics too much. It mistook the prestige of a physicist for the strength of his physics.

The next time a number arrives wearing a great name, you don’t have to argue with the name. Ask instead what the number assumes is sitting still.

Paying my debts

The rehabilitation of John Perry is the work of Philip England, Peter Molnar and Frank Richter, whose 2007 papers in GSA Today and American Scientist did the arithmetic this essay leans on and named the missed opportunity. Brian Shipley’s study of the Perry–Kelvin exchange supplied the texture of Perry’s reluctance. Darwin’s dread is preserved by the Darwin Correspondence Project; Rutherford’s anecdote by A. S. Eve; Perry’s own voice survives in three letters to Nature in 1895, which anyone can read. The Silly Putty and the popcorn belong to generations of science teachers. The argument about authority, and any errors in it, are mine.