Van de Kamp's Planets

Summary

In 1963 Peter van de Kamp reported a planet around Barnard's Star, seen as a wobble in decades of telescope plates. The wobble was in the telescope.

He later described two planets and was still publishing on them in 1982. Plates from other observatories showed no wobble, and the Sproul telescope's own plates contained small jumps that matched the history of its lens. In 2024 and 2025 planets were found at Barnard's Star after all: four of them, each with a minimum mass well below the Earth's, circling the star in less than a week. They are not van de Kamp's planets, which have been ruled out.

What is documented

The star. In 1916 E. E. Barnard, at Yerkes Observatory, reported a faint star crossing the sky by 10.30 seconds of arc a year: "At present this is the largest known proper-motion of a star." Barnard's Star is about six light-years away, the nearest single star to the Sun.

The plates. Swarthmore College's Sproul Observatory photographed the star with its 24-inch (61-centimetre) refractor: 24 plates from 1916 to 1919, and from 1938 a large number every year. Van de Kamp, director of the observatory from 1937, used 2,413 plates carrying 8,260 exposures, taken on 609 nights by fifty observers between 1916 and 1962.

The 1963 claim. In a paper received by The Astronomical Journal on 21 June 1963 and published that September, van de Kamp reported that twenty-five consecutive years of plates showed deviations from the star's expected motion. He fitted an orbit with a period of 24 years, an eccentricity of 0.6, periastron (closest approach) in 1950, and a semi-axis of 0.0245 seconds of arc. At the plate scale he gave, 18.87 seconds of arc per millimetre, that is about 1.3 thousandths of a millimetre on the plate. Assuming a mass for the star of 0.150 solar masses, he derived a companion of 1.6 times the mass of Jupiter, "which therefore appears to be a planet".

Two planets. In 1969 he revised the orbit, and then reanalysed the motion as two planets on circular orbits. In 1975 he discussed two "systematic instrumental equations" in the plates, at 1941.82 and 1949.21, and tied the second to "the replacement of the old aluminum cell by a new cast-iron cell, accompanied by a change in emulsion". He set aside the plates from 1938 to 1949 "for the moment". From the plates of 1950 to 1974 he still reported two companions: periods of 11.5 and 22 years, and masses of 0.00094 and 0.00035 solar masses, about 1.0 and 0.4 times the mass of Jupiter. His series of papers on the planets ran to 1982.

Hershey, 1973. John L. Hershey, also at Sproul Observatory, analysed 12 stars on 423 Sproul plates taken from 1937 to 1969, in a different part of the sky. He found "small changes in the astrometric field which are related to the history of the objective lens". There was a jump of about 3.5 thousandths of a millimetre in 1949, coinciding with the new lens cell and the change of emulsion, and a weaker break coinciding with adjustments of the lens between late 1957 and early 1958. He found no effect in 1967, when the lens was removed from its cell for renovation of the telescope, and wrote that it was "difficult to ascribe any simple cause". His paper does not discuss Barnard's Star.

Gatewood and Eichhorn, 1973. George Gatewood of the Allegheny Observatory and Heinrich Eichhorn of the University of South Florida measured 241 plates of Barnard's Star from the Allegheny and Van Vleck observatories. Fitting the orbit van de Kamp had published in 1969 to their own plates, they found a semi-axis of 0.004 ± 0.003 seconds of arc, against his 0.028 ± 0.003, and a correlation of 0.06 between the motion they measured and the motion he predicted. They concluded, "with disappointment", that their observations "fail to confirm" a planetary companion. On 1 December 1973 Science News ran the headline "No planet for Barnard's star?".

Later searches. A 1999 study used the Hubble Space Telescope's Fine Guidance Sensor, and radial-velocity surveys followed. In 2013 Jieun Choi and colleagues published 248 precise Doppler measurements taken at Lick and Keck observatories over the 25 years from 1987 to 2012. They found no periodic signal above about 2 metres per second and concluded that "for the masses and inclinations found by van de Kamp, the claimed planets are clearly ruled out".

A false start. In November 2018 a team led by Ignasi Ribas reported a candidate planet of at least 3.2 Earth masses on a 233-day orbit, from 771 measurements spanning 20 years. In 2021 Jack Lubin and colleagues showed with new observations that the signal was a false positive, arising from the star's activity and its roughly 145-day rotation.

The planets that are there. On 1 October 2024 a team led by Jonay González Hernández, using the ESPRESSO spectrograph on ESO's Very Large Telescope, announced Barnard b: a planet of at least 0.37 Earth masses on a 3.15-day orbit, with three further candidates. On 11 March 2025 Ritvik Basant and colleagues, combining the MAROON-X spectrograph on the Gemini North telescope with ESPRESSO, confirmed all four. Their minimum masses are 0.19 to 0.34 Earth masses and their orbits last from 2.3 to 6.7 days. The NASA Exoplanet Archive lists the four as confirmed.

Leading explanations

Systematic error in the Sproul plates. Established. The wobble came from the Sproul plates, and plates from two other observatories showed none. The Sproul plates contain jumps tied to changes of lens cell and emulsion and to adjustments of the lens, and van de Kamp himself identified two instrumental breaks. Hershey found no single simple cause for the Sproul effects. His paper lists possible contributors, among them the figure of the lens, the alignment of its components, flexure of lens and tube, and seasonal effects.

Van de Kamp's planets. Rejected. Independent astrometry did not confirm them in 1973. A quarter of a century of radial velocities rules out planets with the masses and inclinations he gave. Van de Kamp published in their support until 1982.

The 2024 and 2025 planets as confirmation of the old claim. Rejected. The planets now known are a few tenths of an Earth mass and orbit in days. Van de Kamp's companions were 0.4 to 1.6 times the mass of Jupiter, on orbits of 11.5 to 24 years. The new planets were found by a different method, the measurement of the star's velocity along the line of sight, from signals of less than 50 centimetres per second.

Current status

Debunked. The planets van de Kamp described do not exist. The wobble he measured came from his instrument, and later searches have excluded planets of the masses and orbits he gave.

Barnard's Star does have planets, and the registry records that plainly: four, confirmed in 2024 and 2025, each with a minimum mass below the Earth's. The search for more continues. New discoveries around the star would not change this record's finding, which concerns the planets claimed from 1963 to 1982.

Last reviewed: September 2026. Records are provisional. Where the evidence changes, the entry changes. Found an error? Tell us.