Yet as some geologists pointed out, Kelvin’s assumptions were seriously flawed. The resulting dispute got bitter.

“The fascinating impressiveness of rigorous mathematical analysis, with its atmosphere of precision and elegance, should not blind us to the defects of the premises that condition the whole process,” deadpanned University of Chicago geologist T. C. Chamberlin before proceeding to demolish Kelvin’s arguments in an 1899 paper. Chamberlin suggested Kelvin’s assumption that there is no source of new heat in the solar system was especially problematic. Though Chamberlin did not mention it, the assumption had effectively been overturned already by Henri Becquerel’s discovery of radioactivity in 1896.

Chemistry has typically been considered the study of the composition and transformation of substances. Though the triumph of atomism in the 19th century incorporated the study of bonding between atoms into these traditional concerns, the nucleus remained the province of physics rather than chemistry. Becquerel’s discovery of radioactivity upset such disciplinary distinctions, however, because the decay of one element into another changes the composition of substances, the bread-and-butter of chemists.

The branch of study that emerged, nuclear chemistry, also eroded geology’s disciplinary independence from physics and chemistry, though in a way that proved more acceptable to geologists than Kelvin’s theory of Earth’s age. For the discovery of radioactivity not only undercut the famous physicist’s theory—by identifying a source of new heat in the energy released by decay—it also provided the key to quantifying the geologic timescale.

Between 1902 and 1903 physicists Ernest Rutherford and Frederick Soddy together proposed that each type of radioactive element decays at a unique and unvarying rate. This is the concept of the radioactive half-life. If their theory were correct, it made every such element a potential clock for measuring geologic time: the conversion of one element into another by decay would act like sand steadily falling between the chambers of an hourglass.

A particularly important kind of radioactive decay to early-20th-century physicists was the breakdown of uranium isotopes U-235 and U-238 into lead isotopes Pb-207 and Pb-206, respectively. With half-lives of about 710 million and 4.5 billion years, the rate of uranium decay was sufficiently long that it could be used to measure timescales on the order of billions of years.

With this discovery, Kelvin’s theory was finished for good, and scientists seemed to have an accurate clock by which to measure Earth’s age. Now they needed to learn how to read it properly.

A crucial step in using radioactivity to date the planet required the separation and accurate measurement of the isotopes found inside minerals. Enter a young Harvard post-doc named Alfred O. C. Nier and his mass spectrometer.

Mass spectrometers separate isotopes according to mass using a combination of electric and magnetic fields. In 1936 Nier married a two-ton electromagnet to a 5-kilowatt electric generator to create a mass spectrometer powerful enough to cleanly separate isotopes of heavy elements, including lead and uranium, the grains of sand marking off Earth’s age. The instrument caught the attention of retired geologist Alfred Lane, the chairman of the National Research Council’s Committee for the Measurement of Geologic Time and the man who steered Nier into geochronology. (Nier found the chairman “peculiar,” but a grant from the NRC likely helped overcome any doubts he had about the committee’s project.)

Meanwhile, Nier was establishing rapport with Harvard chemist Gregory Baxter, a student of Nobel laureate Theodore Richards. Baxter, like Richards before him, was a world leader in measuring atomic weights by chemical methods. For someone looking to measure geologic time through the half-life of uranium, Baxter was a good connection to make. Over the course of their careers, Richards and Baxter had amassed lead samples of high purity, painstakingly refined from ores. Nier found his instrument could rapidly analyze isotope abundances in these samples and boasted that he could do in an hour what it took chemists weeks to do.

By the 1930s, dating methods based on Nier’s technology had won over all but the most hardened geologists. In 1939 he designed a streamlined spectrometer with lower power requirements; its ease of use and indisputable accuracy led to the increasing application of nuclear chemistry to geological problems. (Nier’s innovative mass spectrometers would also prove essential for building the first atomic bomb during the Manhattan Project.) 

More important changes were in the offing. A year earlier Nier had discovered the lead in Richard and Baxter’s samples wasn’t as uniform as it might appear. These ores consisted of a mixture of radiogenic lead—that is, lead generated over the course of Earth’s history from the decay of uranium—and primeval lead, which had been present at the time of Earth’s formation.

Whereas primeval lead had fixed ratios of lead isotopes, the ratios in radiogenic lead varied as a result of uranium decay. Nier discovered these variations in the Harvard samples, and hypothesized that they were caused by the addition of radiogenic lead to primeval lead.

Arthur Holmes, the ringleader of geochronology by this time, used Nier’s data to develop a geologic timescale, attaching ages to the periods and epochs geologists had ordered previously by comparing strata and fossils. In the years that followed, Holmes and others threw all delicacy to the wind, using Nier’s discovery of primeval lead to attempt calculations of Earth’s total age.

By determining how much radiogenic lead had been added to the primeval lead through uranium decay since Earth’s crust had formed, Holmes found he could, in principle, infer the corresponding time elapsed.

After one such calculation in 1946, Holmes wrote excitedly to Nier that he had calculated the planet’s age to be about “3 thousand million years.” The average of his calculations was in fact 3,015 million, but, as he jokingly noted, “we can . . . afford to neglect the odd 15.! [sic].” (Time is cheap in geology!) He considered this calculation to be the “first really reliable estimate of the age of the earth” and congratulated Nier on having made the feat possible.

Despite Holmes’s optimism, however, the debate over Earth’s age wouldn’t be settled for another decade. The problem was that Holmes was working with lead samples that were stand-ins for primeval lead. While geologically very old, these samples originated sometime after Earth’s formation and hence had been subject to uranium decay. That meant these samples did not necessarily reflect Earth’s initial isotopic composition and, as a result, led to an underestimation of the time elapsed.

Where could scientists find truly primeval lead? Why not outer space, where remnants of the original bits of matter from which Earth had formed still drifted about? More specifically, a meteorite whose uranium content, relative to lead, was so small that no significant decay could have occurred since it was formed.

It so happened that 50,000 years ago an errant asteroid tore through the desert sky and blasted out three-quarters of a mile of scrubland near Flagstaff, Arizona, scattering iron-rich debris that is now collectively known as the Canyon Diablo meteorite. Nier’s fellow Manhattan Project alum, University of Chicago geochemistry professor Harrison Brown and his post-doc, Clair Patterson, reasoned that both meteorites and Earth must have formed at the same time and from the same precursor matter. Therefore, such a meteorite could serve as representative of the primeval lead incorporated into Earth at the time of its formation. Find primeval lead in the Canyon Diablo debris, and scientists might finally nail down Earth’s age for good.

Patterson was just the man for the job. For his graduate work, also with Brown, he had been tasked with determining the lead isotope composition of zircons. Zircons are an extremely stable crystal species whose lead content is almost entirely the product of radioactive decay. (Other kinds of rocks were more vulnerable to contamination by non-radiogenic lead).

Unfortunately, zircon’s virtue was also a vice for Patterson—the amount of lead present is almost impossibly meager—1,000 times less than in any mineral seen before. The work infuriated Patterson; his exacting measurements didn’t make any sense when calibrated against rocks of known ages. Eventually he realized trace lead was drifting into his lab like a ghost, haunting his carefully prepared samples. To counter this contamination, he developed a rigorously decontaminated “clean lab,” which was a novelty at the time.

Patterson put this know-how to use in measuring the lead in the Canyon Diablo remnants. Using a modified, Nier-type spectrometer, he and Brown confirmed that it was as free of radiogenic isotopes as they expected. They further demonstrated that certain samples of deep-sea lead, which they argued were representative of modern Earth lead, were as old as Canyon Diablo and a few other meteorites as well. Mathematical analysis revealed this age to be 4.6 billion years.

Patterson recalled the experience in 1995:

The Brown–Patterson age is still accepted today. Decades after Kelvin’s contested amputation of geologic time, physics and geology were reconciled: the physicists got their mathematical rigor, and the geologists got their long timespans. More importantly, the discovery put to bed two older and grander problems.

First, geology got the number it needed to calibrate the geological timescale. Geologic time could finally be securely quantified. Second, the Brown–Patterson age cemented the secularization of Earth’s history and providing a final nail in the coffin for millennia-old religious concepts of the planet’s origins and the solar system in which it resides.

As usually happens in science, however, no sooner were these problems solved, then further ones reared their heads. In this case, an obvious question raised by the meteorite connection was, where do meteorites come from in the first place?

The Case of the Martian Meteorites

Patterson and Brown had bet on a common origin of meteorites and Earth in determining the latter’s age. But the study of meteorites would reveal an even more direct connection between Earth and the solar system, showing the planet was engaged in an eons-long conversation with its neighbors.

Once again, Al Nier’s wizardry with mass spectrometry played a key role, allowing scientists to listen in on the interplanetary exchange.

After an exceptionally successful post-doc at Harvard, in 1938 Nier took up an assistant professorship at the University of Minnesota, where he obtained tenure almost immediately. About the time he became chair of the physics department in 1953, he developed an interest in meteorites, looking at how cosmic rays in space alter their isotopic composition.

With American interest in rocketry taking off after the Sputnik launch in 1957, Nier began to obsess about how to apply mass spectrometry to space exploration. He focused on miniaturizing the mass spectrometer for space missions to other planets, where the instrument could be used to analyze the compositions of their atmospheres.

In his efforts to convince NASA senior managers of the feasibility of miniaturization, Nier built an instrument concealed in a briefcase and took it to NASA headquarters to get the support of friendly “underlings” there. By chance Nier ran into NASA’s associate administrator for research coming out of an elevator. Nier plopped open the case and gave him a demonstration on the spot. “A crowd gathered around to see this little instrument in an attaché case,” Nier later remembered, and the “little sales job” eventually paid off when he was tapped for the 1976 Viking mission to Mars.

Nier was responsible for planning and executing the elemental and isotopic measurements of the Martian atmosphere during the descent of the spacecraft. Some of the most important discoveries of the Mars missions would come from this work, including the discovery of isotopic signatures, characteristic of the Martian atmosphere, that would reveal a stunning connection between Earth and the Red Planet—some meteorites found on Earth were Martian.

This time, the detective who cracked the case was not Nier, but Donald Bogard of NASA’s Johnson Space Center.

In the late 1970s Bogard began studying an unusual category of meteorite, called shergottite, that appeared to be very young in geological terms—only about 200 million years old. Bogard came to this remarkably young age by measuring the argon produced by decay of potassium-40, a technique that had been pioneered by Nier in the 1940s to study terrestrial minerals. The potassium-argon technique, as it was known, was extremely useful for measuring the ages of young rocks, and so provided an essential complement to the uranium-lead method, which was better suited to older rocks. Bogard concluded that the young ages reflected recent crystallization formed in the intense heat of two asteroids colliding.

However, one such meteorite, named Elephant Moraine 79001 after its discovery site in Antarctica, featured unusually large amounts of melt glass, presumably from a particularly hot impact. With his colleague Pratt Johnson, Bogard measured the argon content of the glass, which yielded an apparent age of six billion years old. How could the melted parts be so much older than the rest of the meteorite? Beginning to suspect an unusual origin, Bogard and Johnson compared the argon and other gas contents of the melt to the Martian atmosphere data from Viking. Lo and behold, they matched!

Bob Pepin, a colleague of Nier’s at Minnesota, and coworkers would soon show that the ¹⁵N/¹⁴N ratio of the melt also matched Nier’s data. These results could be explained by a violent impact on the Martian surface that had melted the rock and thereby trapped atmospheric gases in the process of ejecting the rock from the surface. The first interplanetary meteorites had been discovered.

Ironically, Bogard had narrowly missed discovering the Martian origins of these rocks as a graduate student at the University of Arkansas in the 1960s. At the time, it was assumed that meteorites came exclusively from asteroids. Indeed, theoretical studies indicated that a force sufficient to eject a rock from Mars would vaporize the rock in the process.

“It never crossed our mind that these were from a totally different parent body!” he later recalled. “Sometimes you miss serendipity in science.”

Delayed though it may have been, the discovery expanded the geological horizons initiated by isotope geochronology even further. Brown and Patterson had shown that Earth and meteorites had a common origin. Elephant Moraine 79001 had now revealed that the planets, once formed, were not like the detached gods of Epicurus’s imagining, but bodies in interaction with each other through the exchange of matter.

This discovery was not entirely revolutionary, for it, in a most unexpected fashion, comforted Hutton’s uniformitarianism: Earth was shaped not (just) by sporadic dramatic events, but by a steady stream of interactions with the rest of the solar system. More importantly, it raised further questions, as do all important discoveries in science. If rocks could voyage from planet to planet, might those vessels carry more than interesting minerals and isotopic ratios?

Scientists began to wonder whether life, previously considered an actor in a purely terrestrial drama—chemical or divine—had sprouted instead from astral seeds, like the humans in the movie Prometheus. The discovery of meteorites from Mars fueled studies in the 1980s and 1990s that would detect possible traces of life on such meteorites, suggesting that perhaps life could move between planets on meteorites and challenging prevailing theories of exobiology. Unwittingly, the scientific effort to ask Mother Earth her age had morphed into a discipline-exploding expansion of geology to cosmological proportions.

What Killed Off the Dinosaurs?

The 20th century’s most publicly compelling geochemical investigation of meteorites emerged from economic anxiety.

As declining support for nuclear weapons threatened their funding in the 1970s, the U.S. national labs cast about for other research projects. At the Lawrence Berkeley Lab (LBL) at the University of California, Berkeley, the division of nuclear chemistry developed analytical techniques for answering archeological and historical questions. One of the key instrumentalists there was a scientist named Helen Michel.

Michel was first enthralled by chemistry after watching an experiment blow up on her sixth-grade teacher. After graduating high school, she set out for a career in commercial chemistry, but was stymied by the sexism of prospective employers. Instead she enrolled at her hometown university and joined what was then the Radiation Laboratory as a student assistant. After a brief and discouraging stint in graduate school, she returned to the Rad Lab, where she proved herself an expert instrumental chemist, especially in a cutting-edge technique—neutron activation analysis (NAA).

In NAA, neutrons bombard a sample, transforming some of the stable isotopes into radioactive ones. As the radioactive isotopes decay, they emit telltale gamma rays with energy levels unique to each isotope. The method is particularly useful for detecting and quantifying trace elements.

In the late 1970s Michel set this technology to a range of mysteries, including the debunking of a long-treasured hoax in the university’s collection, a brass plate purportedly cast by Francis Drake on his arrival to California in 1579. Michel’s analysis revealed the plate couldn’t be more than a century old. But her use of NAA to measure iridium in a thin layer of rock would prove far more consequential.

The question of what drove the dinosaurs into extinction was then a lively scientific debate. Some paleontologists thought the question was unsolvable. More than a few 1970s children’s books concluded with a line like, “What killed the dinosaurs? We may never know.” Those entranced by this mystery split into several camps. One leading theory suggested dinosaurs had died off gradually, possibly the result of long-term environmental changes. Another suggested emissions from massive volcanic eruptions had quickly soured the global climate, with the theory’s proponents pointing to the Deccan Traps, an enormous deposit of volcanic rocks in India created roughly 65 million years ago.

An alternative theory was advanced by Berkeley geologist Walter Alvarez and his father, Nobel laureate physicist, LBL professor emeritus, and Manhattan Project alum Luis Alvarez. The Alvarezes believed a killer asteroid had smashed into the planet and brought a sudden and catastrophic end to 75% of the world’s species.

Their theory was grounded in a puzzling half-inch-thick layer of clay in a gorge near Gubbio, Italy. On either side of the clay were thick limestone deposits studded with iron minerals and the fossilized shells of tiny aquatic animals called foraminifera. The Italian geologist Isabella Premoli Silva had tracked the evolution of these “forams” during the late Cretaceous period in the limestone below the clay. The abundance of tiny shells testified to the thriving conditions for life. Above that was a layer of clay where no fossilized shells were preserved. Next above that layer in geological time, the fossil record began again, showing just a single species of microscopic foram initially, slowly proliferating into many new species as geological time progressed.

The younger Alvarez had encountered the clay layer while attempting to use the forams, radiometric dating, and the iron deposits—a sort of fossil compass—to reconstruct a 100-million-year record of changes in Earth’s magnetic field. He took a sample of this clay boundary home and showed it to his dad. The elder Alvarez was entranced by the puzzle: why had the limestone-making mechanism in this area apparently turned off and then turned back on?

The first step to answering that question was to know how long it might have taken for this deposit to have been laid down. An unconventional way to answer that question again came from the sky.

Cosmic dust is extremely fine material left behind from the formation of the solar system that continues to rain down on the earth imperceptibly. Its higher concentration of iridium and other platinum group metals makes it measurable—with the right kind of tools. One of those tools was close at hand in the LBL’s NAA team.

The Alvarezes asked LBL colleagues Michel and Frank Asaro to measure the iridium levels in clay samples from Gubbio and a similar sample from New Zealand. Michel’s analysis revealed comparatively high concentrations of iridium in rocks from the geological strata that marked the geological boundary between the Cretaceous and the Paleogene eras, deposits now known as the K–Pg boundary. The samples didn’t reveal a steady accumulation of iridium that could function as a clock—they showed a sudden influx of an element that is usually very rare on the Earth but known to be in much higher concentrations in some meteorites.

For the Alvarezes, this was decisive evidence of a killer asteroid, and the popular press glommed onto their findings. “Comet Fire: Did it doom the dinosaurs?” Time magazine asked. The idea of annihilation raining down from the heavens was a provocation to a nation gripped by the nuclear brinkmanship of the Reagan administration. In Parade magazine Carl Sagan warned Americans of a cataclysmic nuclear winter, an idea fashioned from the Alvarezes’ theory that Cretaceous life had been snuffed out by a planet-wide cloud of debris.

But many geologists and paleontologists found the theory less compelling. Evidence from two locations was hardly enough to support a global claim. Perhaps volcanoes or some other geological process could have concentrated iridium. More importantly, where was the hole? Surely a planet-transforming meteor must have left a massive impact crater?

Nuclear chemistry stepped into the raging debate. Further sampling revealed elevated iridium levels in the K–Pg boundary at more than 100 sites around the globe. Scientists investigated other chemical signatures to test the Alvarez theory. An analysis by J.M. Luck and K.K. Turekian published in 1983 found that samples from the K–Pg boundary had nearly 1:1 ratios of osmium-187 and osmium-186, characteristic of meteorites, rather than the ratios of around 10:1 characteristic of rocks in Earth’s crust.

As compelling as this evidence was for the asteroid hypothesis, it could not rule out volcanoes. Scientists were still split.

The decisive discovery of a suitably massive crater was announced in a paper published in 1991. This paper brought together multiple strands of evidence. Two of the paper’s co-authors were geophysicists Glen Penfield and Antonio Camargo, who had worked for Mexico’s national oil company. In prospecting for oil, they had gathered magnetic anomaly data and gravity maps that suggested a 110-mile-wide ring below the surface of the Caribbean Sea at the edge of the Yucatán Peninsula, near the Mexican town of Chicxulub. Rock samples from oil wells drilled inside the crater revealed shocked rock and other debris characteristic of a meteor impact.

Two other co-authors, Adam Hildebrand and William Boynton, had studied the chemical composition of tektites embedded in the K–Pg boundary near the Brazos River basin in Texas, 1,000 miles away. Analysis showed these tektites had the same chemical and isotopic composition as shocked andesite samples from the oil wells, suggesting this rock was the origin of the ejected material.

Finally, the crater was located in a thick section of limestone, suggesting that the impact may have produced a huge pulse of carbon dioxide that could have caused a period of global warming as well. The combination of geological wreckage and chemical remnants provided compelling evidence in support of the impact theory.

In the years since, isotopic analysis has also narrowed down the time of the extinction, even to the season—spring—and revealed the probable role of acid rain in contributing to global extinction. Most recently, researchers have used isotopic analysis to find in K–Pg samples the distinctive geochemical signature produced when sulfur gases interact with ultraviolet light. This suggests the asteroid’s impact did indeed shoot a massive cloud of aerosols and dust into the stratosphere, helping set off a deep freeze that led to the mass extinction event like the nuclear winter Sagan imagined.

In Search of Lost Time

In 2007 philosopher Derek Turner illustrated the limits of scientific knowledge of the past by means of an analogy:

Three years later Turner was proven dramatically wrong: paleontologists had discovered the color—chestnut—of the small theropod dinosaur Sinosauropteryx, from the early Cretaceous period (120–131 million years ago). Other species followed. It turned out that we could know the colors of at least some animals who lived hundreds of millions of years in the past but not the contents of a lunchbox just a few decades ago. How was this possible?

Advances in nuclear chemistry and instrumentation have enabled scientists to reach into the deep past. Techniques such as scanning electron microscopy and X-ray spectroscopy now allow paleontologists to study microscopic fossil remnants, including cell organelles and trace metals. In combination with what we know about the color of current animals, those traces are sufficient to draw conclusions about what dinosaurs and prehistoric birds looked like, as well as further conclusions based on them, such as the nature of their habitats. Unfortunately, no comparable traces remain of the contents of Turner Sr.’s lunchbox.

Turner characterized his failed prediction as an “epistemic bet against science.” One moral that may be drawn from the story we’ve told here is that scientific discoveries do not just tell us how the world is, but often reveal new, unexpected opportunities for further discovery. This makes betting against science a risky proposition. The spectacular success of nuclear chemistry is a case in point: Neither Becquerel nor Rutherford and Soddy could have predicted that the discovery of isotopes would ultimately reveal the cause of the dinosaurs’ demise or the Martian origin of certain meteorites. Nor would Hutton have dreamed that meteorites, of all things, would indeed provide a “vestige of a beginning” for Earth. On the other hand, scientists are, to some extent, at the mercy of what they discover, and so some problems might never be solved. We may indeed never be able to know what was in that lunchbox.

Advances in nuclear chemistry revealed new knowledge pathways to unlock the secrets held in rocks. But to exploit them, scientists like Al Nier, Clair Patterson, and Helen Michel had to be creative in a way that illustrates the painstaking and far-flung nature of modern science: they had to not just solve theoretical puzzles and make careful observations, but to obsess about arcane details of instrument design and look far and wide for the right rocks to probe.

Correction: This article has been updated to correct Alfred Nier’s relationship to Gregory Baxter and Theodore Richards and to clarify how isotopic ratios of primordial and radiogenic lead were used as a dating method.