4. Dinosaurs

In 1976, seven years had passed since I had received my doctorate. I had stayed at Berkeley, and had followed Luie's model in pursuing several projects of my own. Two of these had become particularly successful: a measurement of cosmic microwave radiation, and the invention of a new method for radioisotope dating. I would soon be rewarded with a faculty position in the physics department at Berkeley and two prestigious national awards. A friend advised me that it was easier to get a good reputation than to keep it. I was looking for something new to do.

      I received a telephone call from Walter Alvarez, Luie's son, who was then working at Columbia University's Lamont-Doherty Geological Observatory. I had never met Walter, and this was only the second time we had spoken on the telephone. He had heard a theory that the dinosaurs had been killed by a supernova, and he thought that he might be able to prove or disprove the theory by making some measurements of a thin clay layer he had found in Italy He said he needed my help, since he had learned from his father that I had invented a sensitive radioisotope-dating technique. His ideas were flowing so quickly over the phone that I couldn't keep up. I finally slowed his pace by saying, "I thought that the dinosaurs had been killed off by smarter mammals." As soon as I'd made this comment, he realized the depth of my ignorance, and he transformed himself from an excited scientist into a patient professor. He said, "Let me tell you a little bit about the dinosaurs." I was grateful for both the change of pace and the absence of condescension in his tone.

      Most of what I knew about dinosaurs I had learned as a child, from the one book the local public library had on the subject. (In the olden days, unlike today, we were not inundated with dinosaur books, dinosaur models, and stuffed, furry, cuddly dinosaur dolls.) I knew the dinosaurs had mysteriously disappeared at the end of the Cretaceous period, 65 million years ago. Most people believed that their bodies had simply grown too big for their brains, and they had become too dumb to compete with the mammals; the word dinosaur has even become a derogatory metaphor for anything grown too big for its own good. This is what I thought I knew about dinosaurs. But as Josh Billings, the nineteenth- century humorist, once said, "The problem with most people isn't so much their ignorance; it's what they know that ain't so."

      Walter quickly corrected my misconceptions. He explained that 65 million years ago, the brightest of the mammals were no smarter than the most intelligent of the dinosaurs. The mammals had been no recent invention of Nature; they had competed unsuccessfully with the dinosaurs for 100 million years. Moreover, most of the mammals that lived back then became extinct at the same time. The popular myth that the dino saurs had been killed by clever little mammals with a taste for dinosaur eggs was no more worthy of respect than the facetious claim that the dinosaurs had died because they were too big to fit on Noah's Ark.

      Whatever killed the large lizards did not particularly specialize in dinosaurs. Over two thirds of all species living at the time disappeared along with them, including not only mammals, but also fishes, corals, shellfish, and even the microscopic single-celled animals known as forams. So much of the biomass was destroyed that paleontologists called the catastrophe a mass extinction.

      At that time the most reputable theory for the extinctions attributed them to climate change. This in turn may have been brought about by continental drift, also known as plate tectonics because it is not just the continents that move, but larger pieces of the Earth's crust, known as plates. North America and Europe are presently moving apart at a speed of about 1 millimeter per month as new seafloor is created at the mid- Atlantic ridge. "That's about the same rate that your fingernails grow," Walter said. A millimeter per month adds up to about a centimeter per year, a meter every 100 years, 10 kilometers every million years, and 10.000 kilometers every billion years, comparable to the size of the Earth. As these plates move, many other things happen. The sea level can rise or fall, the configuration of the oceans can be altered, and the climate, worldwide, can change. About 65 million years ago such motion caused the great inland seas of North America to dry up, a change sufficient by itself to trigger an ecological catastrophe. But Walter hadn't found the reputable theories of mass extinction convincing. He was particularly bothered by the abrupt disappearance of those microscopic animals known as forams, short for foraminfera. What could simul- taneously kill both the giant dinosaurs and the microscopic forams, and do it all over the globe?

      Malvin Ruderman, a professor of physics at Columbia University, had proposed that the creatures were killed by a nearby exploding star, a supernova. In such an explosion the outer layers of a star blow off, and for a few seconds the power poured into space is greater than the combined power of all the other stars in the Universe. For a few weeks the expanding plasma glows with the brightness of 10 billion stars, greater than that of all the other stars in the galaxy put together. Ruderman speculated that intense x-rays from such an explosion destroyed the ozone shield of the upper atmosphere and allowed life-killing ultraviolet radiation from the sun to reach the surface of the Earth. According to this supernova theory, the disaster lasted only a decade, the time it took the atmosphere to restore the shield. Walter noted that this period was much shorter than the millions of years required by other theories, and so it was the duration of the catastrophe that he was most interested in measuring.

       Sedimentary rock forms on the seabed as coccoliths, the remains of shells and skeletons of small sea animals, drift to the bottom and are gradually compressed by the weight of further material falling from above. Walter had been working for several years at various sites in Italy, where he had painstakingly identified all the geologic stages in the rock, and he knew exactly where to find the rock formed at the time the dinosaurs were killed. Just above the rock formed in the Cretaceous period, the last age of the dinosaurs, the continuous limestone was interrupted by a layer of clay less than a half-inch thick. Below the clay layer were abundant forams, but above it they were virtually absent. Walter speculated that the clay was laid down during the great catastro phe, when there were few sea creatures to form limestone, but clay continued to wash down from continental erosion. If we could tell how long it took for the clay to be deposited, we would know how long the catastrophe had lasted.

      Walter did not have to twist my arm to get me interested in helping. The measurement he wanted looked tough, but possible. I agreed to visit him in New York to work out some plans in detail. While there I would also give a colloquium about the cyclotron method and its applications to geology There were many experts at Lamont who could criticize our plans, and thus help us anticipate problems.

      The death of the dinosaurs had been the first unsolved mystery of science I had learned of as a child. After a while I stopped thinking about the difficulties ahead and, instead, indulged myself in the fantasy of solving the problem. It doesn't hurt to fantasize a little, and I had better do it while I could. Luie had taught me that most clever ideas are proven to be wrong within a few days. If you manage to have one good idea every week, then you can expect to find one every few months that is really worth pursuing. Every few years you might have an idea that could lead to a discovery. The important thing is to keep working, thinking, trying out crazy ideas. If you can keep up the pace, then maybe once in your lifetime you might stumble on something really important, something that could change the course of science.

      As I drove up to the Lamont-Doherty Geological Observatory, I pon dered the odd title given to this laboratory. To me, observatory meant telescopes, chilly nights spent on mountains trying to work in virtual darkness. Why had the geologists borrowed this term from the astrono mers? True, neither astronomers nor geologists can do experiments, at least not in the sense that physicists do. The stars and the rocks just sit out there, giving at most a snapshot of what the Universe is like right now. You can't change them in any significant way; you can only make detailed, careful observations of their present configuration, and try to deduce what they must have been like millions or billions of years ago. Astronomers and geologists are observers, not experimenters, observers of the experiments done by Nature. So perhaps "geological observatory" was just right.

      When I had talked to Walter on the phone I had imagined he was in some dingy office that perhaps looked out on other buildings, but in fact his office was in an elegant old mansion that had once served as the manor house on an estate. From the site high over the river, one could see the high cliffs of the Palisades and the rolling hills of the Hudson Valley. Geologists certainly know how to live.

      For some reason it had never occurred to me that Walter would look like his father. But he did. There he was, tall and lean, with the charac teristic smile and blond hair of my former thesis adviser. I have had similar experiences many times, yet it always surprises me to see the clear effects of heredity in people. Perhaps I subconsciously like to forget that I am a biological as well as a cultural creature. Walter looked like a man who had spent a good deal of his life outdoors, and he had a strong handshake. But his voice was somewhat gentler than that of his father. It had a slight twang to it that remninded me of Henry Fonda.

      My colloquium at Lamont was scheduled early in the day About a dozen geologists came to hear me talk about "accelerator mass spec trometry" and how it might be used for dating sedimentary rock. The method used a particle accelerator such as a cyclotron, commonly known as an atom smasher, to accelerate a few of the atoms in the rock to a speed near that of light. When the atoms move with this velocity, special detectors can be used to identify and count individual atoms. One par ticular atom that looked useful for geology was Be- 10, an isotope of beryllium with a total of 10 protons and neutrons. It is a radioactive atom, created when the cosmic radiation from space hits the atoms of oxygen and nitrogen in the atmosphere, breaking them into smaller pieces. Most of this Be-lO gradually falls to the surface of the Earth or into the oceans. As it settles to the seabed it gets trapped in whatever rock happens to be forming. We believe that this rain of Be-lO on the Earth is constant, since we think that the influx of cosmic radiation is constant. So if the rate of rock formation is changing, the density of entrapped Be-lO will change along with it. If a lot of rock forms in any particular year, then the density of Beryllium-lO will below. If the rock forms slowly, the density of Be-lO will be high. By measuring the density of Be-lO atoms entrapped in the sample, we could answer Walter's question about the rate of formation of the mysterious clay layer.

      Be-lO is radioactive, and the data books gave its half-life as only 2.5 million years. That means that in each 2.5-million-year interval after it is laid down half of the remaining Be-lO in the layer will disappear. In 65 million years there are 26 such intervals, so the fraction of Be-lO left after that time can be calculated as (1/2) x (1/2) x (1/2) x (1/2) x . . . with the term "(1/2)" appearing 26 times. A few quick button pushes on a pocket calculator had given Walter and me the answer during our phone conver sation: the original Be-lO had been reduced by a factor of 67 million. Virtually all of it was gone. Nevertheless, the new cyclotron method was so sensitive that it might be able to detect even this tiny remnant. The measurement looked difficult, but still possible.

      The geologists in the audience asked pertinent questions, and raised many objections. I was able to handle the physics, but I had to defer to Walter for virtually everything relating to geology. There were clearly many subtleties in this business that I wasn't even aware of. By the end of my presentation I had the strong impression that Walter knew my subject much better than I knew his.

      Afterward Walter took me to see the enormous "library" of seafloor cores that were stored at Lamont. These were samples of rocks taken by the famous oceanographic ship Glomar Challenger, which had spent years (and many millions of dollars) drilling sample holes in the bottom of the oceans all over the world. This ship was similar to, but should not be confused with, the Glomar Explorer, which achieved notoriety by being secretly run by the CIA, and which reputedly recovered parts of a Soviet submarine off the seafloor. It is ironic that the Glomar Challenger was really an explorer, and the Glomar Explorer was used, in effect, to challenge the Soviets.

      The huge core library was located in the basement of one of the larger buildings. Neatly laid out in rows over a hundred feet long were thousands of metal trays containing cylinders of rock, each one carefully marked with a notation similar to that used in the Dewey decimal system. The age of most of this rock had never been determined, but here it was, a sample of the world, sitting at Lamont waiting to be measured. It was invaluable and irreplaceable (except at great cost), full of many secrets, waiting for clever scientists to decode it. The history of the world was there, if only we could decipher it. Much progress had already been made by using fossils and remnant magnetism to identify the layers. Now Be- 10 and the accelerators might help.

      Walter didn't like to stay indoors, and we talked while walking around the beautiful grounds. He delighted in pointing out where the servants' quarters had once been, where the horses had been kept, which buildings had been used by the master of the estate for entertainment. It was characteristic of him, as I more fully appreciated when I visited him in Italy many years later, that he loved to learn the history of any place where he worked. We talked about geology and physics, and I continued to be impressed with his quick grasp of physics ideas. He was better at this than most physicists I knew, and yet he was very modest and aware of the limitations of his knowledge.

      When I told him about the work I had done with his father, he became particularly attentive. He wanted to know everything I could tell him about his father. I told him how I had apprenticed myself to Luie when I was a graduate student, and how I had once pledged myself to learn everything he was willing to teach me. The apprenticeship was still continuing; after a decade I still had a lot to learn.

      But I became self-conscious when I realized that I seemed to know Luis Alvarez far better than Walter did. His parents' marriage had broken up just as he was finishing high school, but if he had ever harbored resent ment over their divorce, he had clearly left it far behind now. He had childhood memories of his father, but knew almost nothing about his father's professional life. And in the interval he had become a scientist himself. Thus he was consumed with curiosity about his father's work. He wanted to know every detail of his father's methods, how he worked, what he was like to work with, how he thought. I surprised myself by knowing the answers to all his questions. In some sense, because I felt that I had become a surrogate son to Luie, I found it awkward talking to Walter, the real son. But my embarrassment faded as I realized that Walter harbored no jealousy, only great curiosity. I soon found myself talking to him freely. I loved telling him about the excitement of being exposed to Luie's constant barrage of new ideas.

      When I flew back to Berkeley, I thought of Walter as a new friend, potentially as a long-lost brother, and I hoped that our collaboration would grow. But the measurement he wanted wouldn't be easy. The new method, if everything worked perfectly, would barely reach back 65 million years. Perhaps years of further development would be necessary to give the required sensitivity.

      Before visiting Walter, I had given a talk at the Research Progress Meeting (or RPM) at Berkeley about the cyclotron method. My work in this field had begun when Luie had had the ingenious idea of converting a large cyclotron into a mass spectrometer, and then using it to search for quarks. With the help of William Holley and Edward Stephenson, we had succeeded in making the search, but after two years of effort we had found none. It had turned into a "null" experiment, an experiment that states with great precision that something is not there. Null experiments are the least exciting kind done by scientists. Luis' project to x-ray the pyramids had likewise proven to be a null experiment.

      At the Research Progress Meeting I had reported on our null quark search, and on my idea of applying the method to the detection of isotopes such as C-14 (radiocarbon) and Be-lO. Grant Raisbeck, a young physicist with a background in nuclear science, had been one of those in the audience. As he told me later, he spent most of the time while I was speaking "kicking himself' for not having invented the cyclotron method himself. He had been interested in Be-lO for many years, he said, and had realized its potential for geology, but he knew that the standard methods for its detection were very difficult and insensitive. He decided that I had found the key to unlock the problem. The accelerator method would open up a new field of research. Shortly after my talk, Raisbeck devoted himself fully to the exploitation of the new method. In a few years he became the world's expert in the use of the technique for Be- 10, and made hundreds of measurements, using several accelerators in France, where he continued his research .

      In July 1976, after I returned from my visit with Walter Alvarez, I found Raisbeck waiting in my office. I was pleased to see him again, particularly because I had the most exciting Be-lO experiment in the world to tell him about. Walter and I were going to try to solve the mystery of the dinosaurs' extinction. Raisbeck found this application interesting, but complained that I was using the wrong value for the half- life of Be-10. I showed him the number in the table of the isotopes: 2.5 million years.

      "I'm sorry," he said. "That number is wrong. Don't you remember? I told you after you gave the RPM. I measured the lifetime myself. It is 1.5 million years, not 2.5."

      I vaguely remembered Grant coming up to me after my RPM, but at that time I hadn't been very interested in Be-10. I felt that he couldn't be right about the half-life. The value of 2.5 had been obtained by Nobel laureate Edwin McMillan, recently retired as director of the Lawrence Radiation Laboratory. I was certain that McMillan was too good a scien tist to have published such an incorrect number. But Raisbeck was equally certain that McMillan was wrong.

      If Raisbeck was right and McMillan wrong, it would end the proposed dinosaur study with Walter. Sixty-five million years is only 26 half-lives if the value is 2.5 million years. But if the value is 1.5 million years, it would be over 43 half-lives. The additional 17 half-lives would reduce our signal by a factor of (1/2) multiplied by itself 17 times, that is, 1/131,000. The Be- 10 signal would be 131,000 times weaker than we had planned. A difficult experiment would turn into an impossible one.

      I wasn't sure what to do, and I told Luie about the discrepancy. He immediately called McMillan, who didn't take long to discover what had happened. He located his original notebooks and compared his results with those in his published paper. He found that the value in his notes was closer to Raisbeck 's "new" value of 1.5 million years than to his own published value. He traced the discrepancy to a sentence in his paper in which he calculated the mean life, a number 1.4 times larger than the half-life. To his amazement, he had left out the factor of 1.4(1 divided by the natural logarithm of 2) and accidentally published the mean life instead of the half-life. It was only a transcription error!

      McMillan chided me for having used the value given in the table of the isotopes without ever looking back at his original paper. "Sloppy physics," he said. "Physicists today are lazy. Nobody reads the literature any more. If you had read my original paper carefully you would have caught the error."

      He was right, in part. I had done a lot of work, planned a lot of research, without checking one critical number. Had I read McMillan's article I might have noticed the discrepancy. But maybe not. It's very hard to catch mistakes like that. Nobody reads the literature any more, I nastily thought, including the author. The dinosaur project was dead, dead as a project could get. Killed by a misprint.

      I called Walter and told him the tragic news. Too bad. But that is the way projects go. Very few ideas really lead to an important discovery. You mustn't sulk every time one of your ideas fails. Just keep on having ideas. Once per week. Keep at it. Don't stop thinking.