In April 1963, at a meeting of the American Astronomical Society in Tucson, Ariz., Peter van de Kamp made what should have been a landmark announcement. By tracking the motion of a dim, nearby star across the night sky, he had uncovered an unseen object tugging ever so slightly on the star and perturbing its motion—an exoplanet, well before that became a household word. The gravitational perturbation was so subtle that van de Kamp, a Dutch-born astronomer at Swarthmore College, had relied on almost 50 years of telescope observations to build his case for the planet orbiting what is known as Barnard’s Star.

The next day, the news appeared in the New York Times under the headline “Another Solar System Is Found 36 Trillion Miles from the Sun.” The article noted that Barnard’s Star was not the first place that astronomers had found unseen celestial objects in orbit. In fact, in the two decades prior, van de Kamp and his Swarthmore colleagues had presented similar evidence for companions orbiting two other nearby stars, known as 61 Cygni and Lalande 21185. (Nearby stars move appreciably across the night sky on timescales of decades, permitting astronomers to track their paths for any deviations from linearity [see photos above].)

But the supposed companions to 61 Cygni and Lalande 21185 were so massive that they strained the definition of a planet. The object orbiting 61 Cygni was initially estimated to be 16 times the mass of Jupiter, the largest planet in our solar system; the companion to Lalande 21185 was thought to have more than 30 times Jupiter’s mass, although both figures were later revised downward. “The probability is high that, in its internal composition, [the object orbiting 61 Cygni] resembles a star, and differs radically from that which is now attributed, on good evidence, to the major planets,” famed astronomer Henry Norris Russell wrote in 1943. Russell nonetheless concluded that “planet” was an acceptable designation for the giant body, given the casual way astronomers had deployed such terminology throughout history; today we would more likely call those objects “brown dwarfs.”

Barnard’s Star, in contrast, left little room for semantic debate. The object van de Kamp detected there had only 1.6 times the mass of Jupiter. It was clearly a planet. And its discoverer was a respected professional. By 1963 Van de Kamp had already served as the National Science Foundation’s program director for astronomy and as a vice president of the American Association for the Advancement of Science. At Swarthmore, where he had taught for a quarter century and where the observatory now bears his name, van de Kamp was known to provide piano accompaniment during campus screenings of selections from his Charlie Chaplin film collection.

The New York Times article cautioned that the world circling Barnard’s Star was probably frigid and uninhabitable but noted that the discovery “adds support to the conviction of astronomers that a great many solar systems exist, some of them possibly supporting life.”

Fifty years on, the search for a habitable, truly Earth-like exoplanet is in full swing, and the conviction that planetary systems are common has proved correct. Astronomers have now discovered nearly 900 exoplanets orbiting some 700 different stars, according to the latest figures from the online Extrasolar Planets Encyclopedia. But in that vast catalogue of known planets, you will not find a record of the worlds around 61 Cygni, Lalande 21185 or Barnard’s Star. That is because, as it now appears, they do not exist. In time, all of those claims have faded under scrutiny by other astronomers, consigning the supposed planets to the celestial scrapheap.

Unseeing the Unseen

The early planetary claims stood for some time, perhaps because of the reputation of their discoverer, and perhaps because no competing astronomer could hope to compile a data set comparable to van de Kamp’s own, which comprised thousands of photographic plates spanning several decades of telescopic observations. But by the 1970s new evidence began to cast serious doubt on all the supposed planetary systems.

In 1974, astronomer George Gatewood of the University of Pittsburgh published a study showing that his own observations had failed to confirm the existence of a planet orbiting Lalande 21185. (In a strange twist, decades later Gatewood would announce that his data hinted at the presence of a different planetary system around Lalande 21185, but other astronomers, in turn, failed to confirm those claims.)

Around the same time, Swarthmore astronomer John L. Hershey published a study on the long-term reliability of observations from the college’s Sproul 24-inch telescope, where van de Kamp and his colleagues had gathered much of their data. Changes to the telescope’s lens over the years, Hershey concluded, had introduced “one or more sudden small changes” in the apparent position of several stars. The positional shifts that van de Kamp had attributed to planets, it seemed, might therefore have a more mundane explanation.

Indeed, in 1978, yet another Swarthmore astronomer—van de Kamp’s successor as director of Sproul Observatory—chimed in with a sobering assessment of the planetary claims. Wulff Dieter Heintz reviewed the data for Barnard’s Star and the history of telescope adjustments as documented by Hershey, concluding that “the evidence is therefore that the measured nonlinear motion [of the star] is caused by the telescope optics and not by a companion.” (Gatewood and a colleague had also failed to confirm the presence of a planet orbiting Barnard’s Star.) As for 61 Cygni, Heintz also found the evidence lackluster. The original data set, he wrote, “was quite a weak basis for concluding the existence of an orbital effect… and the much stronger present data make it more likely that this result was spurious.”

By that time a handful of other claims had emerged for possible unseen orbital companions around other stars; Heintz dismissed them all. “Other reported cases of low-mass companions do not look more auspicious,” he wrote. “There is now no observational evidence for any companion less massive than the lowest-mass visible stars known.”

The Real Thing

After many false starts, it would take until 1992—nearly three decades after van de Kamp’s announcement in Tucson—for the exoplanet era to begin in earnest. That is when Alex Wolszczan, now at Penn State University, and Dale Frail of the National Radio Astronomy Observatory published a study about an exotic planetary system they had found around a pulsar (a dense, rapidly spinning remnant of a star) called PSR 1257+12. Three years later Michel Mayor and Didier Queloz of the Geneva Observatory found 51 Pegasi b, the first exoplanet known to orbit an ordinary, sunlike star. Van de Kamp lived to see the first of those achievements, but not the second. He died in 1995, six months before Mayor and Queloz’s historic discovery went public, at the age of 93.

Even though van de Kamp’s Tucson announcement ultimately proved not to be a landmark, Barnard’s Star—and the claim for a planetary system there—still captivates researchers five decades later.

In January of this year, a team that included some of the most accomplished exoplanet researchers in the field published the results of their own attempt to confirm or disprove the existence of van de Kamp’s planets. (Seemingly undeterred by his colleagues’ critiques, as late as 1982 van de Kamp published a new analysis of the motion of Barnard’s Star, this time arguing for the presence of not one but two planets, each less massive than Jupiter.) “Although significant doubt has been cast on these planets, including some by van de Kamp himself, no study has definitively ruled them out,” the researchers wrote in the Astrophysical Journal.

In the study, Jieun Choi, now a graduate student at the University of California, Santa Cruz, and her colleagues analyzed decades of data from the Keck telescopes in Hawaii and Lick Observatory in California. The researchers scoured the data for evidence of a stellar wobble imparted by the gravitational pull of orbiting planets. But whereas van de Kamp had relied on astrometry, or the precise measurement of the position of a star on the sky, Choi and her colleagues used radial velocity measurements, which gauge a star’s motion toward or away from Earth. Astronomers can now detect extremely subtle changes in a star’s radial velocity—comparable to the walking speed of a human—and have managed to attribute those changes to unseen exoplanets in hundreds of cases, including 51 Pegasi b [see figure above].

The planets van de Kamp had described would exert a considerable tug on Barnard’s Star, generating a wavelike oscillation in the radial velocity data. But Choi and her colleagues found no such signal. Still, they were careful to praise the man whose work their study had refuted. “Peter van de Kamp remains one of the most respected astrometrists of all time for his observational care, persistence, and ingenuity,” the researchers wrote. “But there can be little doubt now that van de Kamp's two putative planets do not exist.”