how we found the first planet around another sun

The Mechanism

*Michel Gustave Édouard Mayor* (born Lausanne, Switzerland, *12 January 1942*) and *Didier Patrick Queloz* (born Geneva, *23 February 1966*), professor and PhD student respectively in the *Department of Astronomy of the University of Geneva* at the *Geneva Observatory* in Sauverny, Switzerland, had been using since *April 1994* a new high-resolution spectrograph called *ELODIE*, mounted on the *193-centimetre reflecting telescope* of the *Observatoire de Haute-Provence (OHP)* in *Saint-Michel-l'Observatoire*, in the *Alpes-de-Haute-Provence* department of southern France. ELODIE had been built between 1991 and 1993 by André Baranne, Mayor, and Queloz, and was an *échelle spectrograph* of resolution *R ≈ 42,000* with a fibre-fed input and a vacuum-stabilised optical bench that allowed Mayor and Queloz to measure the *radial velocity* — the line-of-sight velocity along the line of sight from the telescope to a target star — of any selected nearby Sun-like star to a precision of approximately *13 metres per second per measurement*, a factor of approximately 10 better than the previous generation of CORAVEL-based instruments. The *radial-velocity method* of planet detection had been proposed in the 1950s by *Otto Struve* (*The Observatory* 72: 199-200, 1952) and refined into a concrete prediction by *Stephen Dole* in 1964: a planet orbiting a star pulls the star around a common centre of mass; the star's resulting motion, projected onto the observer's line of sight, modulates the star's radial velocity periodically; the period of the modulation equals the planet's orbital period and the amplitude of the modulation depends on the planet's mass and its orbital distance. Through the 1980s several groups — *Bruce Campbell* and *Gordon Walker* at the University of British Columbia, *Geoffrey Marcy* and *Paul Butler* at San Francisco State University, *Michel Mayor* at Geneva — had been searching for radial-velocity modulations of Sun-like stars. Marcy and Butler's *iodine-cell* technique, which used the absorption lines of a temperature-stabilised iodine vapour cell as a reference grid against the star's spectrum, had reached *~3 m/s* radial-velocity precision by 1994 but had been used primarily to put *upper limits* on Jupiter-mass planets on Jupiter-period orbits — a class of object that, if present, would produce a *~12 m/s* sinusoidal modulation over 12 years and which by 1994 had been ruled out for the nearest several dozen Sun-like stars. The Geneva ELODIE survey, beginning in *April 1994*, targeted *142 nearby Sun-like stars* in the magnitude range 5-9, drawn from the *Gliese-Jahreiss catalogue* of nearby stars within roughly 50 parsecs of the Sun. Each star was scheduled for *one observation per observing run*, with observing runs of approximately one week, every two months, in the expectation that *Jupiter-mass companions on Jupiter-period (~12 year) orbits* would manifest as slow, drifting radial-velocity modulations detectable after several years of monitoring. The Geneva survey was *not* designed to find anything else. On the *6th of September 1995*, after a 17-month monitoring baseline, Queloz, then 29 and a doctoral candidate, was reducing the cumulative ELODIE radial-velocity data on *51 Pegasi* — a *G2 IV-V* solar-type star at *15.36 parsecs* from the Sun (about 50 light-years), apparent magnitude *5.49* in the constellation Pegasus. The star had been scheduled for routine, slow monitoring. Queloz's reduction software produced an unexpected result: the radial velocity of 51 Pegasi was not slowly drifting in the way a Jupiter-period orbit would predict; it was *oscillating with a clearly resolved sinusoidal modulation of period 4.231 days and semi-amplitude approximately 56 metres per second*. Mayor and Queloz checked the calibration repeatedly, suspecting instrument error or a rotational-modulation artefact of star-spot activity. Over the next two weeks they obtained additional ELODIE observations at high cadence (multiple observations per night, then every night for a week); the 4.23-day modulation persisted with strict phase coherence. They also checked, by examining the line bisector — the shape of the absorption lines at different intensities — that the modulation was not produced by stellar pulsation or star-spot modulation, both of which would produce velocity changes correlated with line-shape changes. The lines stayed sharp; the velocity modulation persisted. The interpretation, which Mayor and Queloz published in *Nature* later that year, was that a *companion of minimum mass 0.47 ± 0.02 Jupiter masses* (the *minimum* mass because the inclination of the orbit was unknown — the radial-velocity method gives only the line-of-sight component) was orbiting 51 Pegasi at a semi-major axis of *0.052 astronomical units* — *one-eighth* of Mercury's orbital distance from the Sun, with a period of *4.231 days*. This was a gas giant. On an orbit closer to its star than Mercury is to ours. With an equilibrium surface temperature of approximately *1,300 Kelvin*. The class of object had no precedent in any model of planet formation: *Saunders, Cameron 1969*, *Lin, Bodenheimer, Richardson 1996* showed that Jupiter-mass gas giants form *beyond a frost line* at a radial distance of several AU from their star, where ices can condense and provide enough solid material to gravitationally collect a hydrogen-rich envelope; nothing in the standard formation theory accounted for a Jupiter-mass planet at 0.05 AU. Mayor and Queloz announced the result at the *9th Cambridge Workshop on Cool Stars, Stellar Systems and the Sun*, held in *Florence, Italy*, on *6 October 1995* — a venue, with no formal embargo, where Mayor presented the result in a 20-minute oral talk and was met with mixed reaction (a fraction of the audience considered the result implausible). The *Nature* paper was titled *"A Jupiter-mass companion to a solar-type star"* and was received on *29 August 1995* (revised October), accepted *6 October*, and published in *Nature* vol. 378, pp. 355-359, on *23 November 1995*. The first independent confirmation came twelve days after the Florence announcement, on *18 October 1995*, when *Geoffrey Marcy* and *Paul Butler* — the rival California team — interrupted their own iodine-cell observing programme at *Lick Observatory* in California, pointed the *3-metre Shane Telescope* at 51 Pegasi for four nights of high-cadence observation, and reproduced the 4.23-day, 56 m/s modulation. Marcy and Butler's confirmation paper appeared in the *Astrophysical Journal Letters* in early 1996. By the end of 1996, Marcy and Butler had found the second hot-Jupiter (*70 Virginis b*), the third (*47 Ursae Majoris b*), and the fourth (*Tau Boötis b*). By the end of 2000, more than 50 hot Jupiters had been catalogued — none of them detectable with anything but the new high-precision radial-velocity instruments. The *Kepler Space Telescope* (launched 2009) used the *transit photometry* method — a planet crossing in front of its star dims the starlight by a fraction of a percent — to confirm by 2020 that approximately *50% of Sun-like stars have at least one planet*, and that planet-mass-and-period combinations like 51 Pegasi b's are *not anomalies of formation but a major outcome of it*. The current best understanding of how a 51 Pegasi b ends up at 0.052 AU is *orbital migration*: gas giants form at several AU from their star, then migrate inward through dynamical friction with the protoplanetary disk in the first few million years of the system's life, settling at whatever inner radius the surviving disk material happens to support. The 2019 *Nobel Prize in Physics* was awarded jointly to *James Peebles* (for theoretical cosmology — half) and to *Mayor and Queloz* (for the discovery of an exoplanet orbiting a Sun-like star — the other half). At the time of the prize, *4,099 confirmed exoplanets* had been catalogued. As of 2026, the figure exceeds *5,800*. 51 Pegasi b itself was given the proper name *"Dimidium"* in 2015 by the International Astronomical Union, in a public-naming contest; its star, 51 Pegasi, was named *"Helvetios"* (Latin for "Switzerland") to recognise the discovery's Geneva origin. The ELODIE spectrograph itself was decommissioned in *August 2006* after twelve years of service and is now on display at the Geneva Observatory; the 1.93-metre OHP telescope on which it was mounted continues to operate as one of the principal observational instruments of French astronomy.