The Elephant in the Universe, page 10
Or a matter of wavelength. For beyond the apparent edge of a galaxy lie tenuous, invisible clouds of cold hydrogen gas that can only be observed at radio wavelengths.
From Kent and Ellen Ford’s farmhouse in Millboro Springs, it’s another sixty miles farther northwest, through the Allegheny Mountains and into Pocahontas County, West Virginia, to the venerable Green Bank Observatory. Originally run by the National Radio Astronomy Observatory (NRAO), Green Bank harbors the largest fully steerable radio dish in the world. It is also a Valhalla for science historians.6
Just beyond the entrance of the observatory grounds is a full-scale replica of physicist Karl Jansky’s 30-meter-diameter “merry-go-round” antenna, which made the first detection of cosmic radio waves back in 1931. Across the entrance road is the refurbished and freshly painted 9.4-meter dish constructed six years later by radio engineer Grote Reber in his mom’s backyard—the first instrument ever to yield a crude map of the radio sky. And in the observatory’s Residence Hall Lounge, a plaque commemorates the fact that here, in November 1961, staff astronomer Frank Drake presented his famous eponymous equation. (The Drake Equation estimates the number of extraterrestrial civilizations in our Milky Way galaxy from whom we could potentially detect radio emissions.)
One slightly less conspicuous exhibit on site—in fact the one that enabled the later observations by Bosma and his radio astronomy colleagues—is the original horn antenna used by Harvard grad student Doc Ewen and his thesis advisor Edward Purcell to detect so-called line emission of neutral hydrogen—radio waves with the very specific frequency of 1420.4 megahertz, corresponding to a wavelength of 21.1 centimeters. This hallmark discovery, made on March 25, 1951 (Easter Sunday; Ewen was working 24 / 7 at the time), eventually made it possible to map the outer parts of distant galaxies. And since the Doppler effect works the same way for radio waves as it does for visible light, precise observations of the 21-cm line reveal the kinematics of the hydrogen gas, including the rotation of gas clouds way beyond the visible disk of a galaxy, where stars are all but absent.
Ewen and Purcell’s search had been motivated by a prediction from Dutch astronomer Henk van de Hulst, a student of Oort’s at Leiden. In 1944, during the Second World War, van de Hulst—the son of a famous Dutch author of children’s books—was asked by his visionary thesis advisor to check if the newly discovered cosmic radio hiss might somehow contain information on the omnipresent cool neutral hydrogen in interstellar space. After some literature study and handwritten arithmetic, the twenty-five-year-old student concluded that there should be a faint hydrogen signal at a wavelength of 21 centimeters.
Right after the war, the Leiden group turned a 7.5-meter radar antenna (left behind by the Germans) into a radio telescope and started searching for the 21-cm line. But partly due to delays resulting from a fire in their receiver, they only succeeded some seven weeks after the discovery by Ewen and Purcell, who knew about van de Hulst’s prediction. Not much later, Australian radio engineers Chris Christiansen and Jim Hindman obtained a third independent detection; all three results were published in the same September 1, 1951, issue of Nature.7
By then, Oort was busy gathering funds for what would briefly be the world’s largest radio dish: the 25-meter Dwingeloo telescope. Inaugurated in April 1956, the instrument would make history by providing the first detailed map of the spiral structure of our own galaxy. However, the first 21-cm observations at Dwingeloo were not targeting our Milky Way but rather Andromeda. Under the supervision of van de Hulst, astronomers Hugo van Woerden and Ernst Raimond obtained the first ever rotation curve of another spiral galaxy based on HI observations. (HI is neutral hydrogen; as you may recall from chapter 5, HII is ionized hydrogen.)
Those were the early days of radio astronomy. For each individual fifteen-minute measurement, the huge dish and the bulky line receiver were prepared manually. Pointing corrections had to be calculated by hand. Data were written on rolls of paper by a pen recorder. When they weren’t observing or tweaking the hardware, van Woerden and Raimond slept in the guest room of Lex Muller, the telescope’s designer and administrator, whose house was right next to the dish. Mrs. Muller provided breakfast, lunch, and dinner.
The Dwingeloo results on Andromeda were published in November 1957, in the Bulletin of the Astronomical Institutes of the Netherlands.8 In a 2020 interview shortly before his death, van Woerden recalled that, indeed, rotational velocities hardly appeared to be declining with increasing distance from the galaxy’s center, but back then, no one was too surprised by that. “It wasn’t an issue at all,” he said. That started to change only slowly, when other astronomers obtained more detailed radio observations of our nearest galactic neighbor, as well as HI rotation curves for other disk galaxies.9
One of those astronomers was Seth Shostak, who would later become senior astronomer at the SETI Institute in Mountain View, California.10 (SETI stands for the Search for ExtraTerrestrial Intelligence. No, nothing has been found yet.) In the late 1960s, as part of his PhD research, Shostak spent a lot of time at Caltech’s Owens Valley Radio Observatory near Big Pine, California, close to the Nevada border. There he studied the distribution and dynamics of neutral hydrogen in three galaxies, including NGC 2403, which is some 3.5 times farther away than Andromeda.11
By the time Shostak was doing his research, radio astronomers had constructed telescopes much larger than the 25-meter Dwingeloo dish in the Netherlands or those at Owens Valley. Jodrell Bank Observatory in northern England operated a 76-meter dish (now known as the Lovell telescope, after the observatory’s first director, Bernard Lovell); Australia had a 64-meter instrument at Parkes, New South Wales, nicknamed simply The Dish; and Green Bank held the world record with its giant 90-meter telescope. But what the smaller antennas at Owens Valley lacked in size and sensitivity, they made up for with their flexibility and larger angular resolution, a measure of the amount of detail that can be seen.
The two identical antennas Shostak used were each 27.4 meters in diameter, but they could be moved on rails, and the data they collected were precisely combined (“correlated”) into one set of observations, as if the two instruments were small parts of one huge virtual dish. This kind of system, called an interferometer, not only has a higher angular resolution than a single-dish instrument but is also much more efficient. Radio telescopes generally have a very small field of view, as if you’re watching the sky through a drinking straw, so you can create a larger “picture” only by carrying out many observations in succession, usually over a period of many days or even weeks. In contrast, an interferometer can build up a two-dimensional radio image in less than a day, through a process known as aperture synthesis.
More often than not, Shostak was the only person at the observatory, spending long nights in the control room listening to the eerie grinding sounds of the antennas outside. He couldn’t help thinking about the multitude of galaxies, stars, and planets in the wider universe and about the possibility that some of those distant worlds might be populated by alien civilizations. Wouldn’t it be great to use radio telescopes to eavesdrop on their interstellar communications? In a 1959 Nature paper, physicists Giuseppe Cocconi and Philip Morrison had suggested that very idea.12 If you know where to look, it’s easy to see when Shostak first became passionate about SETI. At the end of his 1972 dissertation, he wrote, “This thesis is dedicated to NGC 2403 and its inhabitants, to whom copies can be furnished at cost.”
Between 1971 and 1973, Shostak and his thesis advisor David Rogstad, who had moved to Groningen in the Netherlands to work with the new Westerbork telescope, published papers on a total of six galaxies, including NGC 2403, M101 (also known as the Pinwheel galaxy), and M33—the third major member of the so-called Local Group, to which our Milky Way and Andromeda belong. In each case, they found that clouds of cold hydrogen gas way beyond the optical edge of the galaxy were rotating much faster than expected, indicating the presence of “low-luminosity material in the outer regions of these galaxies,” as they wrote in a September 1972 paper in The Astrophysical Journal.13
Meanwhile, NRAO astronomer Morton Roberts was studying neutral hydrogen in the Andromeda galaxy, using the then-largest radio telescope in the world—the Green Bank Telescope, which had become operational in 1962. Improving on the pioneering Dwingeloo observations by van de Hulst, Raimond, and van Woerden, Roberts published his first results in 1966—just one year after Vera Rubin started to share an office with Kent Ford at Carnegie’s Department of Terrestrial Magnetism.14 Roberts’ paper is referenced in Rubin and Ford’s 1970 publication on Andromeda. “I knew Vera pretty well,” Roberts tells me during a Zoom interview from his home in Alexandria, Virginia. “She was a very kind person, and happy to have found a male astronomer listening to her.”15
In the early 1970s—when he started to get better and better results for Andromeda, culminating in a 1975 paper with Robert Whitehurst—Roberts gave Rubin a telephone call.16 “I have something interesting for you,” he told her. “Are you around this week?” A few days later, he drove the 120 miles from NRAO’s headquarters in Charlottesville, Virginia, to Carnegie’s DTM lab in Washington, DC, where he met with Rubin, Ford, their colleague Norbert Thonnard, and Sandra Faber, a Harvard PhD student who lived in DC and had been offered a temporary DTM desk by Rubin.
“Please get me a copy of The Hubble Atlas of Galaxies,” Roberts asked Faber. She duly went off to the library to fetch the iconic 1961 book, famous for its beautiful black-and-white photographs of dozens of galaxies. Back in the meeting room, Roberts opened the atlas to the page featuring the Andromeda spiral. Using tracing paper, he plotted his newest hydrogen velocity measurements, out to a distance of some 95,000 light-years—far beyond what was pictured in the book. Even there, the galaxy’s rotation curve remained flat. Everyone in the room fell silent. When Faber asked, “So what? What’s the significance of a flat rotation curve?” they all swiveled toward her. “Don’t you see? There’s no light there!”
From then on, plots of the rotation curve of the Andromeda galaxy usually contained Roberts’s HI velocities in the outer parts of the galaxy, although he believes the famous image with the photo of Andromeda in the background—the one I saw at Kent Ford’s house—wasn’t published until 1987.
Whether or not Rubin, Ford, and Thonnard were also aware of the work by Rogstad and Shostak remains unclear—the trio did not reference it in their 1978 and 1980 publications. But Shostak just can’t imagine they didn’t know about it. “In 1972, thanks to Mort Roberts, I had a postdoc job at NRAO,” he says. “One of the summer students there was Vera’s twenty-year-old daughter Judy, who would later become an astronomer herself. I’m sure she must have discussed our work with her mother. Vera didn’t have flat rotation curves until a couple of years later; we had done it years before.”
But radio astronomy was a novel and unfamiliar technique to most astronomers, and not too many people took the results all that seriously back then, according to Roberts. “Some were outright skeptical,” he says. “Most were at least very cautious. Certainly no one made the connection to the earlier work on galaxy clusters by Fritz Zwicky.” Dark matter still had a lot of converts to win.
As it turned out, the big game changer—at least as far as rotation curves go—was the Dutch Westerbork Synthesis Radio Telescope, an interferometer like the one in Owens Valley but with fourteen dishes instead of two, each 25 meters in diameter.17 Inaugurated on June 24, 1970, initially with twelve antennas, this new brainchild of Jan Oort attracted a lot of scientists from abroad, most of whom ended up having their office at the Kapteyn Astronomical Laboratory at the University of Groningen. Rogstad spent a couple of years there instructing students—including Albert Bosma—in data analysis and computer technology. Roberts went to Groningen to work on 21-cm observations of the spiral galaxy M81 with Arnold Rots. And Shostak arrived in 1975. He stayed for thirteen years.
Bosma once described the Groningen radio astronomers as “the young Turks of the Kapteyn lab.” At a 1973 conference in Cambridge, England, their younger colleagues shamelessly boasted that the Westerbork telescope was much better than the older Cambridge Interferometer. Eminent British radio astronomer and observatory director Martin Ryle, who more or less had invented aperture synthesis, wrote a formal letter to Oort, complaining about the group’s tacky behavior. Subsequently, the “Westerbork cowboys” were allowed to attend a conference in Onsala, Sweden, only if chaperoned by an older staff member, lest things get out of hand once more.
The Westerbork Synthesis Radio Telescope in the Netherlands, used by Albert Bosma to measure rotational velocities in the outer regions of spiral galaxies.
Then again, the Westerbork telescope was indeed very powerful, as Bosma’s PhD research showed. Bosma took advantage of a new piece of technology: inspired by Rogstad, Groningen astronomer Ron Allen had developed a novel spectrometer that enabled eighty simultaneous observations at slightly different wavelengths. Using this 80-channel filter bank receiver, Bosma studied one galaxy after another, measuring Doppler shifts both within and well beyond the optical edge, charting velocities all over the place, mapping the extended distribution of neutral hydrogen, and discovering warps in the outermost parts of the hydrogen disk.
It all seemed to confirm the earlier, less precise, and less sensitive results of Rogstad and Shostak and of Roberts and Whitehurst. Eventually, no fewer than twenty-five galaxies turned out to have flat rotation curves out to very large distances from their cores, indicating the presence of large amounts of invisible mass way beyond the optical disk. Bosma presented initial results at conferences in 1976 and 1977, but the full extent of his work only became clear with the publication of his 1978 dissertation “The Distribution and Kinematics of Neutral Hydrogen in Spiral Galaxies of Various Morphological Types.”18 Later that year, Rubin, Ford, and Thonnard described their results for just ten galaxies, based on optical observations.
So, is Albert Bosma frustrated?
Well, at least it’s remarkable to read so many different stories, he says. For instance, in her 2014 book The Cosmic Cocktail, theoretical astrophysicist Katherine Freese writes, “It was the work of Rubin and Ford that clinched the case for dark matter in galaxies. Their observations persuaded astronomers that dark matter must exist … and the two deserve a Nobel Prize for this discovery.”19 Likewise, in her Nature obituary of Rubin, Princeton astrophysicist Neta Bahcall calls her “the ‘mother’ of flat rotation curves and dark matter” and states that “her groundbreaking work confirmed the existence of dark matter and demonstrated that galaxies are embedded in dark-matter halos, which we now know contain most of the mass in the universe.”20
Yes, Bosma wrote a letter to the editor in response to Bahcall’s obituary.21 The letter explains why Bahcall’s piece “oversimplifies the dark matter problem”: you really need radio data to probe the outermost regions of galaxies. However, it’s impossible to respond to every incomplete or biased publication—there’s so much interesting radio astronomy to do. “There’s a lot of reinterpretation going on,” he says, “and much of it is outright wrong. I just keep track of what everybody’s thinking. But I’m a bit hesitant about writing that book—I just haven’t got enough time.”
Shostak has mixed feelings, too. “I have a big soft spot for Albert,” he tells me, but “I’m not terribly upset.” Still, “It’s true that Vera came late to the party. All that talk about a Nobel Prize, and now a large telescope being named after her … it makes you feel kind of strange.” Then again, he adds, she never claimed priority herself. Indeed, as I have noted before, Bosma’s thesis is referenced in the 1980 publication by Rubin, Ford, and Thonnard. And in their 1978 paper, the authors make clear that “Mort Roberts and his collaborators deserve credit for first calling attention to flat rotation curves.”
Sandra Faber, who later became a distinguished professor at the University of California, Santa Cruz, believes that—contrary to what is usually the case—Rubin’s current record in history has actually been helped by the fact that she was a woman. It is a remarkable example of reverse gender inequality. “Bosma’s thesis is brilliant. Two hundred years from now,” she muses, “people will certainly realize how important his contributions have been.”
Hopefully, it won’t take that long.
PART II
Tusk
9
Into the Cold
By now you may be wondering when this book will leave the past for what it is and turn to the present. After all, we’re at one-third of the story, and we still seem to be stuck in the 1970s. But don’t worry—we’ll get there. If you want to understand the quest for solutions to the dark matter riddle, you first need to know how the problem arose in the first place. And although the mystery is almost a century old, as we saw in chapter 3, it just so happens that almost all of the most important developments took place in the roaring seventies.
As a quick recap, we’ve learned that galaxies can’t be stable, unless they’re embedded in giant, massive halos. Moreover, galaxies are much more massive than you would guess on the basis of their visible content. Rotational velocities do not decrease with increasing distance from the galaxy’s center but remain more or less constant—a sign that there is more matter in galaxies than is apparent through telescopes. The relative smoothness of the cosmic microwave background suggests that, in the moments after the big bang, weird particles must have already started to form a dark, massive scaffolding that would only later pull in the familiar baryonic matter. Finally, the big bang cannot have produced enough baryonic matter to explain the dynamical observations and the growth of cosmic structure, indicating that most of the gravitating mass in the universe must be in some unfamiliar, nonbaryonic form.
