Strange New Worlds: The Search for Alien Planets and Life Beyond Our Solar System - Ray Jayawardhana (2011)

Chapter 3. A Wobbly Start

False Starts and Death Star Planets

It is one thing to infer that planetary systems must be ubiquitous in the Galaxy from the presence of dusty disks around most newborn stars, but quite another to fnd the planets themselves. Just because the raw material for planet making is common does not have to mean the fnal products are as well. In fact, the hunt for extra-solar planets started long before astronomers had direct evidence of protoplanetary disks. Naturally the targets of these searches were the Sun’s nearest neighbors. The preferred method: looking for periodic wobbles as these stars travel across the sky, by recording their positions carefully over many years. The wobbly behavior would indicate the gravitational tug of an unseen companion, because the stars travel in straight lines otherwise. If the wobble were small enough, it could be due to a planet rather than a dim stellar partner.

On a number of occasions, starting as early as the mid-nineteenth century, astronomers thought they had hit the jackpot. Their announcements often resulted in newspaper headlines, and fed into speculations of alien life, but did not survive closer scrutiny. In every case, the culprit turned out to be nothing more than observational error. The researchers had underestimated the challenge and overrated the precision of their measurements. Meanwhile, starting in the late 1970s, astronomers developed ways to measure precisely the wobble of a star toward and away from us—that is, along our line of sight rather than across the sky—with a spectrograph. But several early surveys came up empty. The refuted claims and failed searches fueled a skeptical, if not hostile, attitude toward planet hunting in the scientifc community. Success came in 1991 from a rather surprising milieu: small, rocky worlds orbiting a fast-spinning stellar cinder called a pulsar, made almost entirely of neutrons and emitting beams of radiation like a cosmic lighthouse. But that did not satisfy those who yearned for counterparts to our own solar system, circling normal stars.

Career-ending Non-Planet

The frst known claim of an extrasolar planet detection involves the star 70 Ophiuchi. In fact, for a prosaic star in a nondescript constellation, it has drawn a lot of attention over the years. Orange in color and not particularly bright, in no way does it stand out among the thousands of stars visible to the naked eye. Viewed with a small telescope though, it turns out to be a double star, as frst reported in 1779 by William Herschel, the discoverer of planet Uranus. The two stars circle each other every eighty-eight years. Burnham’s Celestial Handbook, a perennial favorite among amateur sky watchers, describes 70 Ophiuchi as “probably among the most thoroughly studied dozen binaries in the heavens.” During the course of two centuries, many observers have recorded the relative motions of the pair. Starting with Herschel himself, a number of them suspected the presence of an unseen third body whose gravity tugs on the pair.

Undoubtedly the most controversial, and probably the most reviled, of the astronomers associated with 70 Ophiuchi is Thomas Jefferson Jackson See. Born as a farmer’s son in 1866 in post-Civil War Missouri, he excelled in science at university, graduating as the class valedictorian. It was while using a modest campus telescope to scan the heavens that he developed a lifelong interest in double stars. See continued his study of astronomy abroad, at the University of Berlin, where he learned to calculate the orbits of double stars and wrote a dissertation on their origins. He returned to the United States as an instructor at the University of Chicago astronomy department, headed by George Ellery Hale, the legendary astrophysicist who went on to build several of the world’s largest telescopes at the time. Within a few years, See was submitting papers on observations of binary stars and calculations of their orbits almost monthly to scientifc journals while also writing articles for magazines such as Popular Astronomy and Atlantic Monthly. As his reputation grew, both as a scientist and a popularizer, so apparently did his arrogance. He left Chicago in a huff in 1896 when the university failed to promote him to associate professor, the same rank held by the more prominent Hale. He moved to the Lowell Observatory in Flagstaff, Arizona, at the invitation of its founder, and later to the U.S. Naval Observatory in Washington, DC.

The brash thirty-three-year-old with a promising future committed a blunder of career-destroying proportions in 1899. It had to do with a controversy surrounding the 70 Ophiuchi binary. Back in 1855, based on visual observations of the two stars’ motions, Captain W. S. Jacob of the East India Company’s Madras Observatory had written: “There is, then, some positive evidence in favor of the existence of a planetary body in connexion with this system, enough for us to pronounce it highly probable.” Jacob’s declaration is likely the frst serious claim in a scientifc journal of detecting a planet beyond the solar system. Forty years later, See also invoked a dark companion to account for apparent anomalies in the binary orbit. “I have succeeded in showing conclusively that the system is perturbed by an unseen body,” he wrote. He believed that such a body is unlikely to be shining by its own light, but stopped short of declaring it to be “the frst case of planets . . . noticed among the fxed stars.” In 1899, Forest Ray Moulton, one of See’s former graduate students at Chicago, published a paper in the Astronomical Journal showing that the proposed triple system would be unstable also pointed out that a new orbit for 70 Ophiuchi calculated by Eric Doolittle, another former student, removed the need for a third body. See took the refutation personally. He fred off a vitriolic letter to the Astronomical Journal, which published a sanitized version. The accompanying editor’s note explained, “the remainder of Dr. See’s communication is omitted, partly because it has no bearing on Mr. Moulton’s paper.” What’s more, the editor effectively banned him from publishing in the Journal in the future—a severe rebuke, especially considering See’s prolifc record. Three years later, See suffered a breakdown, and moved to the naval station at Mare Island, California. He later switched to theoretical work and continued to publish in a German journal as well as in popular magazines, gaining fame even as he feuded with other scientists and attacked Einstein’s theory of relativity.

The claims of a planet in the 70 Ophiuchi system reappeared in the 1940s. Dirk Reuyl and Erik Holmberg of the University of Virginia caused a sensation when they inferred a 10-Jupiter-mass companion in its midst. By now, astronomers were looking for periodic wobbles in other stars, both single and binary, as unseen planets’ gravity tugs on them. The technique, known as astrom-etry, had proven successful in the study of binary stars, revealing, for example, a faint stellar cinder orbiting Sirius, the brightest star in the night sky and one of the Sun’s closest neighbors at a mere 8.6 light-years. Back in 1844, the German astronomer Friedrich Wilhelm Bessel frst noted subtle departures in the path of Sirius through the sky. He proposed that a faint companion is responsible. Bessel died two years later, but in 1862, the American telescope maker Alvan Graham Clark was able to see Sirius’s partner with a new 18-inch refracting telescope he had built. Sirius B, as it is dubbed, is about a thousand times fainter than Sirius A. It was the frst white dwarf, the collapsed core of a dead low-mass star, to be identifed and still the nearest one known. Clark’s spectacular success not only helped boost his family’s telescope business, founded by his father, but also vindicated the promise of astrometry to pick out hitherto unseen companions of stars.

But the application of astrometry to planet hunting is a lot more challenging, because planets have much lower masses than stars and thus induce much smaller wobbles. As seen from ten light-years away, Jupiter causes the Sun to wobble by a mere 1.6 milliarcseconds—some two-millionths of a degree—over its twelve-year orbit. That is an angle comparable to the thickness of a human hair seen from 3 kilometers! The other planets affect the Sun’s motion even less, because they are less massive. The size of the wobble depends not only on the mass of the planet but also on the mass of the star: the less massive the star, the more it would feel the tug of a given planet. So a red dwarf with a Jupiter would display a much bigger wobble than the Sun. The other factor that affects the wobble’s size is the distance between the star and the planet. Planets in wider orbits cause larger astrometric wobbles, because the center of gravity between the star and the planet is shifted farther out. For that reason, Neptune induces a larger astrometric wobble on the Sun than Uranus, even though both planets have similar masses. Therefore, the easiest planets to fnd with astrometry are massive ones far out. The downside is that it takes longer to confrm such planets because their periods are longer. It also helps to focus on the nearest stars, because their wobbles appear larger on the sky.

Kaj Strand of Swarthmore College near Philadelphia also announced a planet discovery at about the same time as Reuyl and Holmberg. Using photographic plates taken with the 61-centimeter Sproul telescope on campus, he reported an 8-Jupiter-mass planet in another binary star system, dubbed 61 Cygni. In fact, Swarth-more researchers remained at the forefront of planet hunting for the next few decades. The campus observatory’s director, Peter van de Kamp, a Dutch-born expert on double stars with a talent for music and a fondness for Charlie Chaplin movies, led the effort. In 1951, he and his student Sarah Lippincott announced a planet around the nearby red dwarf star Lalande 21185. The most infamous case, though, came a decade later when van de Kamp announced planets orbiting the so-called Barnard’s star.


Figure 3.1. Wobble pattern of the Sun as seen from afar, for the most part in response to the gravitational tugs of the two largest planets, Jupiter and Saturn. Credit: NASA

A Fifty-Year Saga

A red dwarf too faint to see with the naked eye, Barnard’s star also happens to be located in the constellation Ophiuchus. It had come to the attention of Edward Emerson Barnard, a pioneer in the use of photography in astronomy, back in 1916 for its rapid apparent motion across the sky. Barnard correctly surmised that it had to be nearby: at six light-years, it is one of the Sun’s closest neighbors. That fact and its low mass—one-sixth of the Sun’s—made Barnard’s star an excellent target for astrometric planet searches.

Van de Kamp began observing it in 1938 with the Sproul telescope on campus, soon after he arrived at Swarthmore. By the early 1960s, he had collected over two thousand photographic plates, and reported wiggles in the star’s path through space. The culprit, he inferred, must be a 1.6-Jupiter-mass planet in a highly elongated orbit. The periodic wobbles were about ffteen times bigger than the Sun’s due to Jupiter, and easier to detect, because Barnard’s star is so low in mass. By 1969, with more observations in hand, he found evidence of not one but two giant planets, roughly akin to Jupiter and Saturn, in twelve-and twenty-six-year orbits. Later he revised the outer planet’s period to twenty years.

Through the 1960s, his fndings were generally accepted, even appearing in a popular astronomy textbook. But Nicholas Wagman of the Allegheny Observatory at the University of Pittsburgh had his doubts. His own photographic plates, albeit covering a much shorter timeline, did not show wobbles in the star’s motion. He wondered whether Van de Kamp’s instrument was fawed. When George Gatewood arrived in Pittsburgh as a graduate student, Allegheny astronomers encouraged him to follow up on Barnard’s star. Working with Heinrich Eichhorn and using plates from Allegheny Observatory as well as Van Vleck Observatory in Connecticut, he found no wiggles in the star’s motion. In a 1973 paper, the two researchers wrote: “Thus we conclude, with disappointment, that our observations fail to con-frm the existence of a planetary companion to Barnard’s star.” They speculated whether “spurious effects” in the optical system of the Sproul telescope were responsible for Van de Kamp’s erroneous measurements, and cautioned that “There are perhaps similar instances in the past, when astrometric investigations have suggested the reality of actually unreal things.”

Van de Kamp was not ready to give in. He continued to believe that his planets were real. But even his own handpicked successor as director of the Sproul Observatory, Wulff Heintz, a quiet German with a skeptical eye, was unable to duplicate his fndings. Heintz found small variations in the photographic plates taken over the years that could well confound the subtle measurements. In 1976, he published the frst of a series of papers refuting the planet claims, declaring that “No evidence for a real, periodic motion of [Barnard’s star] is found.” Van de Kamp felt betrayed and never forgave his colleague.

Meanwhile, Barnard’s star had become a favorite locale in science fction. The English writer Michael Moorcock depicted it as a destination for people feeing social breakdown on Earth in his 1969 novel Black Corridor. In Douglas Adams’s Hitchhiker’s Guide to the Galaxy, published a decade later, it is a way station for interstellar travelers. It featured in a miniseries on Norwegian television as well as several novels of the American aerospace engineer and science fction writer Robert L. Forward.

Researchers continued to argue about the reality of orbital anomalies until about the mid-1980s. By then, most of them, with the notable exception of Van de Kamp, who had gone back to the Netherlands, had given up on the planets’ existence. Gatewood had also ruled out the companion proposed by Van de Kamp and Lippincott around Lalande 21185. Heintz went on to analyze the astrometric observations of 61 Cygni and 70 Ophiuchi, and concluded that their apparent orbital deviations are also entirely due to observational errors rather than unseen companions. Gatewood and Heintz hold the unglamorous, albeit important, distinction of striking down several of the best-known “extrasolar planets” of the twentieth century.

Cautious Pioneers

The failures of the astrometric planet searches soured the mood in the scientifc community. “It is quite hard nowadays to realize the atmosphere of skepticism and indifference in the 1980s to proposed searches for extrasolar planets. Some people felt that such an undertaking was not even a legitimate part of astronomy,” Gordon Walker, now retired from the University of British Columbia in Canada, wrote in a recent article. “One distinguished astronomer strode out of the room when I got up to talk about searching for planets in the late 80’s—seems hard to believe now,” he added in an e-mail. Still, with digital cameras and spectrographs coming into use, Walker thought that a different approach could bear fruit. Instead of measuring tiny changes in a star’s position on the sky, he intended to

1 In 1996, the same Gatewood announced, at a meeting of the American Astronomical Society in Madison, evidence for two giant planets around Lalande 21185. The claim received widespread media attention—“Data Seem to Show a Solar System Nearly in the Neighborhood” reported the New York Times—but remains unsupported. look for periodic shifts in its line-of-sight velocity induced by an unseen planet. As a star sways back and forth due to a companion’s tug, the lines imprinted on its spectrum would shift toward the red and the blue, due to the so-called Doppler effect, refecting the star’s motion perpendicular to the plane of the sky.


Figure 3.2. Doppler (or radial velocity) shifts of a star, in response to the gravitational tugs of an unseen companion.

Both astrometry and the Doppler technique favor massive planets, because they perturb the stars more. However, while the astrometric wobble—the apparent displacement of the star on the sky—is larger for planets farther out, the Doppler velocity shift is bigger for fast-moving close-in planets. Unlike astrometry, the Doppler technique does not need a star to be nearby to work: the measurement of the radial velocities does not depend on the star’s distance from the Sun. Doppler has a big disadvantage, however. It can only determine the minimum mass of a companion, not its exact mass, without knowing the inclination of its orbit relative to the plane of the sky. If the companion’s orbit is aligned exactly edge-on from our vantage point, we will see the star move toward and away from us with the highest possible velocity. But if the system is oriented precisely pole-on, we won’t see the star move at all along the line of sight. Thus, without knowing the tilt of a given system, we have to live with this ambiguity in the companion’s mass.

Stellar “radial velocities” could not be determined to better than about 1 kilometer per second from spectra registered on photographic plates. That was a far cry from what was needed to detect planets: the Sun’s velocity shift due to Jupiter is only 12 meters per second. However, digital detectors, with their much greater sensitivity, made it possible to measure minuscule shifts of stellar lines against the lines stamped in the same spectrum by gases in the Earth’s own atmosphere. The new precision, Walker realized, may be suffcient to look for extrasolar planets.

Bruce Campbell, who joined Walker as a postdoctoral fellow in 1976, improved on the idea. He proposed passing the starlight through a captive gas before it entered the spectrograph. The lines imprinted by the vapor would act as sort of a precise ruler for measuring subtle shifts of the stellar lines. At the suggestion of two colleagues, Campbell and Walker chose hydrogen fuoride (HF), with well-spaced lines similar in width to stellar lines. On December 27, 1978, the researchers took their frst observations—of the Sun—with a gas cell in front of the spectrograph on the 1.2-meter telescope at the Dominion Astrophysical Observatory in nearby Victoria. “Frankly, it was quite unsafe. HF is highly corrosive and toxic,” Walker wrote. “The cell had to be heated to 100 Celsius to prevent the HF from polymerizing and the cell windows being plexiglass warped with the heat. Nonetheless, we took a series of exposures of the Sun with the telescope mirror covers closed—enough light got through the gaps between the covers to give us a good signal,” he added. The setup worked.

The following year, Campbell joined the staff of the Canada-France-Hawaii Telescope in Hawaii. While there, he built a safer version of the HF cell, using sapphire windows rather than Plexiglas, for use on the 3.6-meter CFHT For the next twelve years, Campbell, Walker, and Stephenson Yang used the CFHT instrument to monitor the velocities of some two dozen bright stars. Since astronomers expected other planetary systems to resemble our own, with giant planets in decade-long orbits, their “early search strategies concentrated on long-term monitoring with observations spaced out over months and years,” Walker explained. They applied for telescope time twice a year and were usually given four pairs of nights a year, though the annual allocation was eventually reduced to three pairs of nights. “It really was tedious because there could be no obvious results from any one observing run and, perhaps more seriously, no publications to nourish research funds,” he added.

By 1987, the team thought it had some interesting fnds. Several stars showed long-term trends perhaps indicative of Jovian planets, while one binary star system in particular, gamma Cephei, called out for special attention. In addition to the large velocity changes resulting from the two stars’ motions, it exhibited signs of a third body, with a period of 2.7 years and a minimum mass of only 1.7 times that of Jupiter. Campbell announced their preliminary results at the American Astronomical Society meeting in Vancouver that June. The Associated Press report published in the New York Times carried the headline “Planets Outside Solar System Hinted,” refecting the excitement of the press conference. Other astronomers were much more skeptical. When Campbell, Walker, and Yang published their results in the Astrophysical Journal the following year, the tone was more muted: the gamma Cephei planet was not even mentioned in the abstract. Sometime later, unable to secure a permanent position, Campbell left astronomy to become a personal tax consultant. Walker persisted but pretty much relented on the planet claim in a follow-up paper four years later, concluding that the star’s own variability rather than a planet was likely responsible for the small velocity shifts. His caution stemmed in part from a colleague’s misclassifcation of gamma Cephei as a yellow giant, a bloated star well past its prime whose own palpitations might mimic a planet-induced wobble.

Meanwhile, in 1989, David Latham of the Harvard-Smithsonian Center for Astrophysics reported a probable brown dwarf companion to the solar-type star HD 114762. Given its hefty minimum mass of eleven times Jupiter’s, many hesitated to call it a planet. Other astronomers, notably Geoffrey Marcy and Paul Butler at San Francisco State University, were also in the game. Taking a cue from the Canadian team, they used a gas cell for their observations with the 3-meter telescope at the Lick Observatory near San Jose, California. The Canadians “invented the technique that we stole,” Marcy told the Globe & Mail recently. “They were measuring the velocities of stars, for the frst time in history, to plus or minus 10 meters a second.” However, Butler, coming from a chemistry background, opted for safe-to-handle iodine instead of toxic hydrogen fuoride. The spectrum of iodine was in some ways less suited for making precise velocity measurements of stars, but the California researchers developed sophisticated software to get around its shortcomings.

Starting in the late 1980s, Marcy and Butler targeted one hundred nearby Sun-like stars. Just like the Canadians, they had to persevere in an atmosphere of widespread skepticism. “There was literally a gravesite with lots of tombstones of planets that had come to life erroneously and then laid to rest,” Marcy told a journalist recently. Suffcient access to telescopes and grant funds were a problem too. By the early 1990s, review panels were losing patience with null results after years of investments. Marcy once showed me the evaluation report of a NASA grant selection committee from 1994. “The prior scientifc achievements of the investigators in the areas of stellar spectroscopy are world-class and they are working very hard on this project,” the panel commended. But the reviewers expressed disappointment that “present precision is no better than a decade ago.” They were also “concerned about the absence of publications” and “unconvinced that another factor of 2 or 3 in precision could be obtained by further refn-ing [the analysis software].” The grant application was denied.

Death Star Planets

Meanwhile, news of extrasolar planets had come from a totally unexpected corner. In the summer of 1991, three astronomers using the venerable 76-meter radio dish at Jodrell Bank near Manchester, England, reported a planet orbiting a pulsar dubbed PSR B1829-10.

Pulsars are fast-spinning remnants of stars that long ago exploded as supernovae. These compact stellar cinders, made almost entirely of neutrons, emit beams of radiation, like celestial lighthouses. As the pulsar rotates and the beams sweep through space, it appears to wink on and off or “pulse” as seen from the Earth. The American astronomers Walter Baade and Fritz Zwicky had predicted the existence of neutron stars back in 1934, soon after the discovery of neutrons themselves. They proposed that “a supernova represents the transition of an ordinary star into a neutron star, consisting mainly of neutrons. Such a star may possess a very small radius and an extremely high density.” When Jocelyn Bell Burnell, then a graduate student working with Anthony Hewish at Cambridge University, found the frst pulsar in 1967, it baffed scientists for a while. Its beeping every 1.3 seconds seemed so artifcial that they jokingly referred to it as LGM-1—for “Little Green Men.” It was Thomas Gold of Cornell University who suggested pulsars are fast-spinning neutron stars, and proposed a model for their radio emission. By now, astronomers have identi-fed over a thousand pulsars in the Galaxy.


Figure 3.3. A pulsar is like a cosmic lighthouse. As it spins, the beams of radiation sweep through space, so we see it pulsing on and off with clockwork precision.

Normally the regularity of their pulses rivals the best atomic clocks. But that was not the case with PSR B1829-10: it seemed to alternately speed up and slow down, as if something was nudging the pulsar back and forth, thus very slightly changing the time its radiation takes to reach us. It is from subtle periodic shifts in the time interval between these radio pulses that Andrew Lyne, Matthew Bailes, and Setnam Shemar inferred a 10-Earth-mass planet with a six-month period—sort of a Neptune in Venus’ orbit. The journal Nature declared, “First Planet Outside Our Solar System,” and scientists scratched their heads. A pulsar’s midst was a rather odd—not to mention hazardous—location for a planet. If the planet had formed before the star went supernova, the powerful explosion would surely have disrupted its orbit. Perhaps it was more likely that the planet coalesced out of the debris of a destroyed companion star. Some scientists were skeptical: they wondered if the six-month period could be an artifact caused by the Earth’s motion around the Sun.

That same summer, Alexander Wolszczan, a Polish-born astronomer then at Cornell University, was puzzling over cyclical changes in the period of another pulsar. He had been observing PSR B1257+12, located 1,300 light-years away toward the Virgo constellation, with the 300-meter Arecibo radio telescope in Puerto Rico for about a year. (You may have seen the famed radio dish, built into a natural sinkhole on the ground, in the Bond movie GoldenEye,the science fction flm Contact, or the X-Files episode “Little Green Men.”) He wondered if the cause could be planets. The idea was exotic enough but not totally new: in two earlier cases, researchers had considered pulsar planets before the errant signals turned out to be what astronomers call timing noise. “I was excited but skeptical enough to sit on my data,” Wolszczan, who is now at Pennsylvania State University, told me. “I needed more data before publishing something so out of the ordinary.”

Wolszczan received news of the Lyne group’s announcement that summer with mixed feelings. He was disappointed that someone else had gotten there frst but was encouraged that his own result may be real. Meanwhile, he had asked Dale Frail of the National Radio Astronomy Observatory in New Mexico to measure the precise position of the pulsar using the Very Large Array. The VLA consists of twenty-seven individual antennas laid out in a “Y” pattern and linked together to act as one giant radio telescope. Once the pulsar’s exact position was in hand, the orbital solution became clear. The presence of two separate cycles, at sixty-six days and ninety-eight days, implied two planets in the resonant orbits (with a period ratio of 2:3), each about three times the Earth’s mass. There were also hints of a third cycle, thus a third, even lower-mass, planet. With a good solution in hand by September, Wolszczan started giving talks about the pulsar planets at scientifc meetings. The two researchers submitted a paper to Nature in late November. The news spread quickly, and New Scientist magazine reported it in mid-December, even though their paper was not scheduled to appear until a few weeks later.

Both Lyne and Wolszczan were invited to give talks at the American Astronomical Society (AAS) meeting in Atlanta in January 1992. With less than a week to go, Lyne noticed a small difference in the assumed position of PSR B1829-10 between two observing epochs. He redid the calculations with the corrected position. “Five minutes later I froze in horror. I saw the planet evaporate,” he told Discover magazine. Once the correction was applied, the signal of a planetary companion disappeared. It was a simple, yet embarrassing, error. In his AAS talk on January 15, Lyne stood up before a packed audience, explained the mistake, and retracted the erroneous planet claim. “People were moved by Andrew’s honest admission. He confronted the evidence and explained their mistake. It was a moment of human drama,” according to Wolszczan. The sympathetic audience gave Lyne a long standing ovation. The next day, Nature was to publish a retraction from Lyne and Bailes, along with an editorial commending the directness of their admission.

Wolszczan was to speak right after Lyne. He wondered how the audience would react to his discovery now that the other pulsar planet system had turned out to be spurious. Fortunately, those in the room were able to distinguish between the two claims. For one, Wolszczan had obtained careful readings of his pulsar’s position on multiple occasions. For another, PSR B1257+12 was a different sort of beast, known as a millisecond pulsar. With a period of just six milliseconds, it is believed to have spun up by cannibalizing a companion star. As material spirals in from a normal star toward the pulsar, it adds spin energy (or angular momentum) to it, speeding up its rotation. The presence of raw material also makes it easier to imagine planets forming around a millisecond pulsar than a regular pulsar like PSR B1829-10. At the end of the AAS session, “everybody in the audience was still quite shocked. There was one dramatic retraction and then another case that seemed quite convincing . . . . People needed some time to think about things, to process what had happened,” Wolszczan recalled.

Donald Backer of the University of California at Berkeley presented independent confrmation of the PSR B1257+12 planets three months later, at a workshop in Pasadena. By late 1993, Wolszczan was able to detect evidence of the two planets interacting with each other gravitationally, laying any remaining doubts to rest. He also confrmed the presence of a third, innermost planet, weighing about as much as the Earth’s Moon. Intriguingly, if the orbital distances of the three pulsar planets are doubled, they would line up roughly with the positions of Mercury, Venus, and Earth in the solar system. In both cases, the innermost planet is the least massive while the other two are comparable in mass.

Table 3.1 Planets around the Pulsar PSR B1257+12


What the pulsar planets are made of is anybody’s guess. Most researchers expect them to be rocky and barren, baked in high-energy radiation. Their origin also continues to baffe us. One possibility is that the pulsar’s beam ripped apart a companion star, whose material ended up in a disk out of which the planets coalesced. Astronomers have since found a few such “black widow” pulsars, which have cannibalized their mates. Another theory is that two white dwarfs spiraled together and merged to form a neutron star, while leftover material went into making planets. Both scenarios have diffcul-ties, however. In the frst instance, a disk may not form at all. In the second case, the merger may lead to an explosion, which leaves nothing behind to form planets out of. Nearly two decades later, PSR B1257+12 is still the only isolated pulsar with defnitive evidence of planets orbiting it. So it may just be a special case.

Even after the discovery of pulsar planets, most scientists, and the public, continued to focus on planetary systems orbiting normal stars like our Sun. As a stringer for Science magazine, I interviewed members of several planet search teams about the status of their programs at the triennial general assembly of the International Astronomical Union in The Hague in August 1994. In a brief news item, published two weeks later under the title “No Alien Jupiters,” I wrote: “Recently astronomers have found planets where they least expected them: around pulsars, those fast-spinning remnants of stars that long ago exploded as supernovae. It would be far more intriguing, however, to fnd counterparts to our own solar system—planets circling nearby stars that resemble our Sun. But that search keeps coming up empty. It’s not for want of looking, though, as several groups reported at the IAU meeting.” One possibility, raised by Geoff Marcy was that most planetary systems might not include planets massive enough to be detectable. “The critical question [is] whether Jupiter itself is more massive than commonly occurs elsewhere,” he told me. We didn’t have to wait much longer to fnd out.

2 In 2003, astronomers reported a massive planet orbiting a pulsar-white dwarf binary system in the core of the old star cluster M4. This planet is believed to have formed around a normal star and survived the capture of that star into an orbit around the pulsar, the swelling of the star’s outer layers, and the contraction of its core into a white dwarf. See newscenter/archive/releases/2003/19/.